[Git][ghc/ghc][wip/js-staging] StgToJS.Apply: Remove FIXMEs
doyougnu (@doyougnu)
gitlab at gitlab.haskell.org
Thu Aug 25 22:54:38 UTC 2022
doyougnu pushed to branch wip/js-staging at Glasgow Haskell Compiler / GHC
Commits:
13fe0a11 by doyougnu at 2022-08-25T18:54:13-04:00
StgToJS.Apply: Remove FIXMEs
StgToJS.FFI: remove FIXMEs
StgToJS.Expr: remove FIXMEs
StgToJS: Remove FIXMEs
StgToJS: Remove FIXMEs
- - - - -
18 changed files:
- compiler/GHC/StgToJS.hs
- compiler/GHC/StgToJS/Apply.hs
- compiler/GHC/StgToJS/Arg.hs
- compiler/GHC/StgToJS/CodeGen.hs
- compiler/GHC/StgToJS/CoreUtils.hs
- compiler/GHC/StgToJS/DataCon.hs
- compiler/GHC/StgToJS/Deps.hs
- compiler/GHC/StgToJS/Expr.hs
- compiler/GHC/StgToJS/ExprCtx.hs
- compiler/GHC/StgToJS/FFI.hs
- compiler/GHC/StgToJS/Literal.hs
- compiler/GHC/StgToJS/Object.hs
- compiler/GHC/StgToJS/Prim.hs
- compiler/GHC/StgToJS/Printer.hs
- compiler/GHC/StgToJS/Profiling.hs
- compiler/GHC/StgToJS/Regs.hs
- compiler/GHC/StgToJS/Stack.hs
- compiler/GHC/StgToJS/Utils.hs
Changes:
=====================================
compiler/GHC/StgToJS.hs
=====================================
@@ -70,13 +70,6 @@ import GHC.StgToJS.CodeGen
-- Primitives that are represented as multiple values (Int64#, Word64#, Addr#)
-- are passed to FFI functions with multiple arguments.
--
--- FIXME: specify argument order:
--- high then low (Int64#/Word64#)?
--- array then offset(Addr#)?
--- StablePtr#: do we pass the array?
--- FIXME: how do we return them from FFI? With h$retN variables as for
--- unboex tuples?
---
-- Interruptible convention: FFI imports with the "interruptible" calling
-- convention are passed an extra argument (usually named "$c") that is a
-- continuation function. The FFI function must call this function to return to
@@ -128,8 +121,8 @@ import GHC.StgToJS.CodeGen
-- , i -- (array) fields layout (empty if variable layout)
-- , n -- (string) object name for easier dubugging
-- , a -- constructor tag / fun arity
--- , r -- FIXME
--- , s -- static references? FIXME
+-- , r -- ??
+-- , s -- static references?
-- , m -- GC mark?
-- }
--
=====================================
compiler/GHC/StgToJS/Apply.hs
=====================================
@@ -107,7 +107,6 @@ genApp ctx i args
, [top] <- concatMap typex_expr (ctxTarget ctx)
, getUnique i == unpackCStringAppendIdKey
, d <- utf8DecodeByteString bs
- -- FIXME (Sylvain, 2022/02): we assume that it decodes but it may not (e.g. embedded file)
= do
prof <- csProf <$> getSettings
let profArg = if prof then [jCafCCS] else []
@@ -222,7 +221,7 @@ genApp ctx i args
| n <- length args
, n /= 0
, idFunRepArity i == n
- , not (isLocalId i) -- FIXME (Sylvain 2022-08): why are we testing this here and not in the oversaturated case below?
+ , not (isLocalId i)
, isStrictId i
= do
as' <- concatMapM genArg args
@@ -550,9 +549,6 @@ genericStackApply cfg = closure info body
, ifS (newAp .===. var "h$ap_gen")
((sp |= sp - needed_regs) <> (stack .! (sp - 1) |= newTag))
(sp |= sp - needed_regs - 1)
- -- FIXME (Sylvain 2022-08): this is fragile and probably inefficient.
- -- Instead of filling h$apply array with h$ap_gen, we should leave
- -- it with empty items and match "undefined" here.
-- Push generic application function as continuation
, stack .! sp |= newAp
@@ -760,17 +756,12 @@ stackApply s fun_name nargs nvars =
(ifS (toJExpr nargs .>. arity)
(traceRts s (toJExpr (fun_name <> ": oversat")) <> oversatCase c arity0 arity)
(traceRts s (toJExpr (fun_name <> ": undersat"))
- <> mkPap s pap r1 (toJExpr nargs) stackArgs -- FIXME do we want double pap?
+ <> mkPap s pap r1 (toJExpr nargs) stackArgs
<> (sp |= sp - toJExpr (nvars + 1))
<> (r1 |= toJExpr pap)
<> returnStack))
]
- _ -> mempty -- FIXME: Jeff (2022,03), just quieting non-exhaustive
- -- patterns. That the code wants to do this
- -- means we should be encoding that funCase is
- -- only callable on ValExpr (JVar pap)'s in
- -- the type system, perhaps with a GADT or
- -- phantom
+ _ -> mempty
funCase :: JExpr -> JStat
@@ -789,12 +780,7 @@ stackApply s fun_name nargs nvars =
<> (r1 |= toJExpr pap)
<> returnStack))
]
- _ -> mempty -- FIXME: Jeff (2022,03), just quieting non-exhaustive
- -- patterns. That the code wants to do this
- -- means we should be encoding that funCase is
- -- only callable on ValExpr (JVar pap)'s in
- -- the type system, perhaps with a GADT or
- -- phantom
+ _ -> mempty
-- oversat: call the function but keep enough on the stack for the next
@@ -843,7 +829,6 @@ fastApply s fun_name nargs nvars = func ||= body0
jVar \c farity arity ->
[ c |= closureEntry r1
, traceRts s (toJExpr (fun_name <> ": sp ") + sp)
- -- TODO: Jeff (2022,03): factor our and dry out this code
, SwitchStat (entryClosureType c)
[(toJExpr Fun, traceRts s (toJExpr (fun_name <> ": ")
+ clName c
@@ -872,12 +857,7 @@ fastApply s fun_name nargs nvars = func ||= body0
<> (r1 |= toJExpr pap)
<> returnStack))
]
- _ -> mempty -- FIXME: Jeff (2022,03), just quieting non-exhaustive
- -- patterns. That the code wants to do this
- -- means we should be encoding that funCase is
- -- only callable on ValExpr (JVar pap)'s in
- -- the type system, perhaps with a GADT or
- -- phantom
+ _ -> mempty
oversatCase :: JExpr -> JExpr -> JStat
oversatCase c arity =
@@ -1130,13 +1110,11 @@ papGen cfg =
-- general utilities
-- move the first n registers, starting at R2, m places up (do not use with negative m)
--- FIXME (Jeff, 2022/03): pick a better name, e.g., `r2moveRegs`
moveRegs2 :: JStat
moveRegs2 = TxtI "h$moveRegs2" ||= jLam moveSwitch
where
moveSwitch n m = SwitchStat ((n .<<. 8) .|. m) switchCases (defaultCase n m)
-- fast cases
- -- TODO: tune the parameteters for performance and size
switchCases = [switchCase n m | n <- [1..5], m <- [1..4]]
switchCase :: Int -> Int -> (JExpr, JStat)
switchCase n m = (toJExpr $
=====================================
compiler/GHC/StgToJS/Arg.hs
=====================================
@@ -158,7 +158,7 @@ genArg a = case a of
as <- concat <$> mapM genArg args
e <- varForDataConWorker dc
inl_alloc <- csInlineAlloc <$> getSettings
- return [allocDynamicE inl_alloc e as Nothing] -- FIXME: ccs
+ return [allocDynamicE inl_alloc e as Nothing]
x -> pprPanic "genArg: unexpected unfloated expression" (pprStgExpr panicStgPprOpts x)
genIdArg :: HasDebugCallStack => Id -> G [JExpr]
@@ -258,7 +258,6 @@ jsStaticArg = \case
StaticLitArg l -> toJExpr l
StaticObjArg t -> ValExpr (JVar (TxtI t))
StaticConArg c args ->
- -- FIXME: cost-centre stack
allocDynamicE False (ValExpr . JVar . TxtI $ c) (map jsStaticArg args) Nothing
-- | Generate JS code corresponding to a list of static args
=====================================
compiler/GHC/StgToJS/CodeGen.hs
=====================================
@@ -138,7 +138,7 @@ genUnits m ss spt_entries foreign_stubs
staticInit <-
initStaticPtrs spt_entries
(st', _, bs) <- serializeLinkableUnit m st [] [] []
- ( -- FIXME (Sylvain, 2022/02): optimizer disabled: O.optimize .
+ ( -- O.optimize .
jsSaturate (Just $ modulePrefix m 1)
$ mconcat (reverse glbl) <> staticInit) "" [] []
return ( st'
@@ -218,8 +218,7 @@ genUnits m ss spt_entries foreign_stubs
let allDeps = collectIds unf decl
topDeps = collectTopIds decl
required = hasExport decl
- stat = -- FIXME (Sylvain 2022/02): optimizer disabled:
- -- {-decl -} Opt.optimize .
+ stat = -- Opt.optimize .
jsSaturate (Just $ modulePrefix m n)
$ mconcat (reverse extraTl) <> tl
(st', _ss, bs) <- serializeLinkableUnit m st topDeps ci si stat mempty [] fRefs
=====================================
compiler/GHC/StgToJS/CoreUtils.hs
=====================================
@@ -203,9 +203,9 @@ primTypeVt t = case tyConAppTyCon_maybe (unwrapType t) of
| tc == mutVarPrimTyCon -> RtsObjV
| tc == mVarPrimTyCon -> RtsObjV
| tc == tVarPrimTyCon -> RtsObjV
- | tc == bcoPrimTyCon -> RtsObjV -- fixme what do we need here?
+ | tc == bcoPrimTyCon -> RtsObjV -- unsupported?
| tc == stackSnapshotPrimTyCon -> RtsObjV
- | tc == ioPortPrimTyCon -> RtsObjV -- FIXME: Jeff (2022, 05) IOPort, how to handle in JS?
+ | tc == ioPortPrimTyCon -> RtsObjV -- unsupported?
| tc == anyTyCon -> PtrV
| tc == compactPrimTyCon -> ObjV -- unsupported?
| tc == eqPrimTyCon -> VoidV -- coercion token?
=====================================
compiler/GHC/StgToJS/DataCon.hs
=====================================
@@ -43,9 +43,6 @@ genCon ctx con args
| [ValExpr (JVar ctxi)] <- concatMap typex_expr (ctxTarget ctx)
= allocCon ctxi con currentCCS args
- -- FIXME: (Sylvain 2022-03-11) Do we support e.g. "data T = MkT Word64"? It
- -- would return two JExprs
-
| xs <- concatMap typex_expr (ctxTarget ctx)
= pprPanic "genCon: unhandled DataCon" (ppr (con, args, xs))
=====================================
compiler/GHC/StgToJS/Deps.hs
=====================================
@@ -39,7 +39,7 @@ import Control.Monad
import Control.Monad.Trans.Class
import Control.Monad.Trans.State
-data DependencyDataCache = DDC -- FIXME Sylvain 2022-02: use UniqFM
+data DependencyDataCache = DDC
{ ddcModule :: !(IntMap Unit) -- ^ Unique Module -> Object.Package
, ddcId :: !(IntMap Object.ExportedFun) -- ^ Unique Id -> Object.ExportedFun (only to other modules)
, ddcOther :: !(Map OtherSymb Object.ExportedFun)
=====================================
compiler/GHC/StgToJS/Expr.hs
=====================================
@@ -331,18 +331,10 @@ resultSize xxs@(_:xs) t
| otherwise = [(LiftedRep, 1)] -- possibly newtype family, must be boxed
resultSize [] t
- -- FIXME: Jeff (2022,05): Is this check actually needed? If we have a runtime
- -- rep kinded type can't we just call typePrimReps to get the PrimReps and
- -- then primRep size just like in the catchall case? I don't see why this
- -- doesn't work.
| isRuntimeRepKindedTy t' = pprPanic "resultSize: Type was RuntimeRepKinded don't know the size! " (ppr t')
-
-- Note that RuntimeRep from Builtins.Types hits this case. A singleton of
-- (LiftedRep, 1) is exactly what's returned by the otherwise case for
-- RuntimeRep.
- -- FIXME: Luite (2022,07): typeLevity_maybe can panic, doesn't the next case
- -- give us the right answer?
- -- Nothing <- typeLevity_maybe t' = [(LiftedRep, 1)]
| otherwise = fmap (\p -> (p, slotCount (primRepSize p))) (typePrimReps t)
where
t' = unwrapType t
@@ -403,9 +395,6 @@ popLneFrame inEntry size ctx = do
let ctx' = ctxLneShrinkStack ctx size
let gen_id_slot (i,n) = do
- -- FIXME (Sylvain 2022-08): do we really need to generate all the Idents here
- -- to only select one? Is it because we need the side effect that consists in
- -- filling the GlobalId cache?
ids <- identsForId i
let !id_n = ids !! (n-1)
pure (id_n, SlotId i n)
@@ -698,8 +687,6 @@ genAlts ctx e at me alts = do
return (s, r)
_ -> error "genAlts: invalid branches for Bool"
- -- FIXME: add all alts
-
AlgAlt _tc -> do
ei <- varForId e
(r, brs) <- normalizeBranches ctx <$>
@@ -923,11 +910,7 @@ allocDynAll haveDecl middle cls = do
fillObjs = mconcat $ map fillObj cls
fillObj (i,_,es,_)
- | csInlineAlloc settings || length es > 24 = -- FIXME (Jeff, 2022/03): the call to length means `es`
- -- should be something other than
- -- a list. Also why is 24
- -- important? And 24 should be a
- -- constant such as `fooThreshold`
+ | csInlineAlloc settings || length es > 24 =
case es of
[] -> mempty
[ex] -> toJExpr i .^ closureField1_ |= toJExpr ex
=====================================
compiler/GHC/StgToJS/ExprCtx.hs
=====================================
@@ -127,7 +127,6 @@ ctxUpdateLneFrame new_spilled_vars new_lne_ids ctx =
{ ctxLneFrameBs = addListToUFM (ctxLneFrameBs ctx) (map (,new_frame_size) new_lne_ids)
, ctxLneFrameSize = new_frame_size
, ctxLneFrameVars = ctxLneFrameVars ctx ++ new_spilled_vars
- -- FIXME: could we use a stack? (i.e. cons new variables)
}
-- | Remove information about the current LNE frame
=====================================
compiler/GHC/StgToJS/FFI.hs
=====================================
@@ -50,7 +50,6 @@ import Control.Monad
import Control.Applicative
import qualified Text.ParserCombinators.ReadP as P
--- FIXME: what if the call returns a thunk?
genPrimCall :: ExprCtx -> PrimCall -> [StgArg] -> Type -> G (JStat, ExprResult)
genPrimCall ctx (PrimCall lbl _) args t = do
j <- parseFFIPattern False False False ("h$" ++ unpackFS lbl) t (concatMap typex_expr $ ctxTarget ctx) args
@@ -140,8 +139,7 @@ parseFFIPattern' :: Maybe JExpr -- ^ Nothing for sync, Just callback for async
parseFFIPattern' callback javascriptCc pat t ret args
| not javascriptCc = mkApply pat
| otherwise =
- if True -- FIXME (Sylvain 2022-03): we don't support parsing of JS imports.
- -- So we assume that we can directly apply to them...
+ if True
then mkApply pat
else do
u <- freshUnique
@@ -278,13 +276,9 @@ callbackPlaceholders (Just e) = [((TxtI "$c"), e)]
parseFfiJME :: String -> Int -> Either String JExpr
parseFfiJME _xs _u = Left "parseFfiJME not yet implemented"
- -- FIXME (Sylvain 2022-02): removed temporarily for the codegen merge. Need to
- -- decide which syntax we support
parseFfiJM :: String -> Int -> Either String JStat
parseFfiJM _xs _u = Left "parseFfiJM not yet implemented"
- -- FIXME (Sylvain 2022-02): removed temporarily for the codegen merge. Need to
- -- decide which syntax we support
saturateFFI :: JMacro a => Int -> a -> a
saturateFFI u = jsSaturate (Just . mkFastString $ "ghcjs_ffi_sat_" ++ show u)
=====================================
compiler/GHC/StgToJS/Literal.hs
=====================================
@@ -73,12 +73,6 @@ genStaticLit = \case
LitChar c -> return [ IntLit (fromIntegral $ ord c) ]
LitString str
| True -> return [ StringLit (mkFastStringByteString str), IntLit 0]
- -- FIXME: documentation for LitString says it's always UTF8 encoded but it's
- -- not true (e.g. for embedded files).
- -- 1) We should add a decoding function that detects errors in
- -- GHC.Utils.Encoding
- -- 2) We should perhaps add a different LitBin constructor that would
- -- benefit other backends?
-- \| invalid UTF8 -> return [ BinLit str, IntLit 0]
LitNullAddr -> return [ NullLit, IntLit 0 ]
LitNumber nt v -> case nt of
@@ -97,7 +91,6 @@ genStaticLit = \case
LitDouble r -> return [ DoubleLit . SaneDouble . r2d $ r ]
LitLabel name _size fod -> return [ LabelLit (fod == IsFunction) (mkFastString $ "h$" ++ unpackFS name)
, IntLit 0 ]
- -- FIXME: handle other LitRubbish, etc.
l -> pprPanic "genStaticLit" (ppr l)
-- make an unsigned 32 bit number from this unsigned one, lower 32 bits
=====================================
compiler/GHC/StgToJS/Object.hs
=====================================
@@ -6,7 +6,7 @@
{-# LANGUAGE ScopedTypeVariables #-}
{-# LANGUAGE TupleSections #-}
--- only for DB.Binary instances on Module see FIXME below
+-- only for DB.Binary instances on Module
{-# OPTIONS_GHC -fno-warn-orphans #-}
-----------------------------------------------------------------------------
@@ -36,16 +36,7 @@
-- - dependency info
-- - closureinfo index
-- - closureinfo data (offsets described by index)
-
--- FIXME: Jeff (2022,03): There are orphan instances for DB.Binary Module and
--- ModuleName. These are needed in StgToJS.Linker.Types for @Base@ serialization
--- in @putBase at . We end up in this situation because Base now holds a @Module@
--- type instead of GHCJS's previous @Package@ type. In addition to this GHC uses
--- GHC.Utils.Binary for binary instances rather than Data.Binary (even though
--- Data.Binary is a boot lib) so to fix the situation we must:
--- - 1. Choose to use GHC.Utils.Binary or Data.Binary
--- - 2. Remove Binary since this is redundant
--- - 3. Adapt the Linker types, like Base to the new Binary methods
+--
-----------------------------------------------------------------------------
module GHC.StgToJS.Object
@@ -174,7 +165,6 @@ trim = let f = dropWhile isSpace . reverse in f . f
isGlobalUnit :: Int -> Bool
isGlobalUnit n = n == 0
--- fixme document, exports unit is always linked
isExportsUnit :: Int -> Bool
isExportsUnit n = n == 1
@@ -492,9 +482,6 @@ putSymbolTable (SymbolTable _ hm) = st
st = DB.runPut $ do
DB.putWord32le (fromIntegral $ length xs)
mapM_ DB.put xs
- -- fixme: this is a workaround for some weird issue sometimes causing zero-length
- -- strings when using the Data.Text instance directly
- -- mapM_ (DB.put . TE.encodeUtf8) xs
xs :: [FastString]
xs = map fst . sortBy (compare `on` snd) . nonDetEltsUniqMap $ hm
-- We can use `nonDetEltsUniqMap` because the paired `Int`s introduce ordering.
=====================================
compiler/GHC/StgToJS/Prim.hs
=====================================
@@ -578,7 +578,6 @@ genPrim prof ty op = case op of
, a .! i |= new
, s |= zero_
]
- -- fixme both new?
[ s |= one_
, o |= x
]
@@ -705,7 +704,6 @@ genPrim prof ty op = case op of
PrimInline $ r |= app "h$compareByteArrays" [a1,o1,a2,o2,n]
CopyByteArrayOp -> \[] [a1,o1,a2,o2,n] ->
- -- FIXME: we can do faster by copying 32 bit ints or doubles
PrimInline $ loopBlockS (Sub n one_) (.>=. zero_) \i ->
[ u8_ a2 (Add i o2) |= u8_ a1 (Add i o1)
, postDecrS i
@@ -742,7 +740,7 @@ genPrim prof ty op = case op of
AddrSubOp -> \[i] [_a1,o1,_a2,o2] -> PrimInline $ i |= Sub o1 o2
AddrRemOp -> \[r] [_a,o,i] -> PrimInline $ r |= Mod o i
AddrToIntOp -> \[i] [_a,o] -> PrimInline $ i |= o -- only usable for comparisons within one range
- IntToAddrOp -> \[a,o] [i] -> PrimInline $ mconcat [a |= null_, o |= i] -- FIXME: unsupported
+ IntToAddrOp -> \[a,o] [i] -> PrimInline $ mconcat [a |= null_, o |= i]
AddrGtOp -> \[r] [a1,o1,a2,o2] -> PrimInline $ r |= if10 (app "h$comparePointer" [a1,o1,a2,o2] .>. zero_)
AddrGeOp -> \[r] [a1,o1,a2,o2] -> PrimInline $ r |= if10 (app "h$comparePointer" [a1,o1,a2,o2] .>=. zero_)
AddrEqOp -> \[r] [a1,o1,a2,o2] -> PrimInline $ r |= if10 (app "h$comparePointer" [a1,o1,a2,o2] .===. zero_)
@@ -920,13 +918,13 @@ genPrim prof ty op = case op of
MkWeakOp -> \[r] [o,b,c] -> PrimInline $ r |= app "h$makeWeak" [o,b,c]
MkWeakNoFinalizerOp -> \[r] [o,b] -> PrimInline $ r |= app "h$makeWeakNoFinalizer" [o,b]
- AddCFinalizerToWeakOp -> \[r] [_a1,_a1o,_a2,_a2o,_i,_a3,_a3o,_w] -> PrimInline $ r |= one_ -- fixme
+ AddCFinalizerToWeakOp -> \[r] [_a1,_a1o,_a2,_a2o,_i,_a3,_a3o,_w] -> PrimInline $ r |= one_
DeRefWeakOp -> \[f,v] [w] -> PrimInline $ mconcat
[ v |= w .^ "val"
, f |= if01 (v .===. null_)
]
FinalizeWeakOp -> \[fl,fin] [w] -> PrimInline $ appT [fin, fl] "h$finalizeWeak" [w]
- TouchOp -> \[] [_e] -> PrimInline mempty -- fixme what to do?
+ TouchOp -> \[] [_e] -> PrimInline mempty
KeepAliveOp -> \[_r] [x, f] -> PRPrimCall $ ReturnStat (app "h$keepAlive" [x, f])
=====================================
compiler/GHC/StgToJS/Printer.hs
=====================================
@@ -63,8 +63,6 @@ ghcjsRenderJsV r (JHash m)
quoteIfRequired x
| isUnquotedKey x' = text x'
| otherwise = PP.squotes (text x')
- -- FIXME: Jeff (2022,03): remove the deserialization to String. We are only
- -- converting from ShortText to String here to call @all@ and @tail at .
where x' = unpackFS x
isUnquotedKey :: String -> Bool
@@ -74,7 +72,6 @@ ghcjsRenderJsV r (JHash m)
&& all validOtherIdent (tail x)
- -- fixme, this will quote some idents that don't really need to be quoted
validFirstIdent c = c == '_' || c == '$' || isAlpha c
validOtherIdent c = isAlpha c || isDigit c
ghcjsRenderJsV r v = renderJsV defaultRenderJs r v
@@ -119,7 +116,7 @@ prettyBlock' r ( (DeclStat i)
)
| i == i' = (text "var" <+> jsToDocR r i <+> char '=' <+> jsToDocR r v) : prettyBlock' r xs
--- modify/assign operators (fixme this should be more general, but beware of side effects like PPostExpr)
+-- modify/assign operators
prettyBlock' r ( (AssignStat (ValExpr (JVar i)) (InfixExpr AddOp (ValExpr (JVar i')) (ValExpr (JInt 1))))
: xs
)
=====================================
compiler/GHC/StgToJS/Profiling.hs
=====================================
@@ -88,7 +88,6 @@ enterCostCentreThunk :: JStat
enterCostCentreThunk = ApplStat (var "h$enterThunkCCS") [r1 .^ "cc"]
setCC :: CostCentre -> Bool -> Bool -> G JStat
--- FIXME: ignoring tick flags for now
setCC cc _tick True = do
ccI@(TxtI _ccLbl) <- costCentreLbl cc
addDependency $ OtherSymb (cc_mod cc)
@@ -147,7 +146,7 @@ costCentreStackLbl' :: CostCentreStack -> G (Maybe String)
costCentreStackLbl' ccs = do
ifProfilingM f
where
- f | isCurrentCCS ccs = return $ Just "h$currentThread.ccs" -- FIXME
+ f | isCurrentCCS ccs = return $ Just "h$currentThread.ccs"
| dontCareCCS == ccs = return $ Just "h$CCS_DONT_CARE"
| otherwise =
case maybeSingletonCCS ccs of
=====================================
compiler/GHC/StgToJS/Regs.hs
=====================================
@@ -29,18 +29,6 @@ import GHC.Data.FastString
import Data.Array
import Data.Char
--- FIXME: Perf: Jeff (2022,03): as far as I can tell, we never pattern match on
--- these registers and make heavy use of the Enum, Bounded, and Ix, instances.
--- This heavily implies to me that we should be using something like: StgReg =
--- StgReg { unStgReg :: Int8# } and then store two nibbles in a single byte. Not
--- only would this be more memory efficient, but it would also allow for
--- optimizations such as pointer tagging and avoiding chasing the info table,
--- although I'm not sure if this would really benefit the backend as currently
--- written. Other than that a newtype wrapper with a custom bounded instance
--- (hand written or deriving via) would be better. In almost all functions that
--- take an StgReg we use either the Bounded or the Enum methods, thus we likely
--- don't gain anything from having these registers explicitly represented in
--- data constructors.
-- | General purpose "registers"
--
-- The JS backend arbitrarily supports 128 registers
@@ -103,8 +91,6 @@ r3 = toJExpr R3
r4 = toJExpr R4
--- FIXME: Jeff (2022,03): remove these serialization functions after adding a
--- StgReg type with a proper bounded and enum instance
jsRegToInt :: StgReg -> Int
jsRegToInt = (+1) . fromEnum
=====================================
compiler/GHC/StgToJS/Stack.hs
=====================================
@@ -182,9 +182,6 @@ pushOptimized' xs = do
pushOptimized =<< (zipWithM f xs (slots++repeat SlotUnknown))
where
f (i1,n1) xs2 = do
- -- FIXME (Sylvain 2022-08): do we really need to generate all the Idents here
- -- to only select one? Is it because we need the side effect that consists in
- -- filling the GlobalId cache?
xs <- varsForId i1
let !id_n1 = xs !! (n1-1)
=====================================
compiler/GHC/StgToJS/Utils.hs
=====================================
@@ -52,7 +52,6 @@ assignCoerce (TypedExpr AddrRep [a_val, a_off]) (TypedExpr UnliftedRep [sptr]) =
, a_off |= sptr
]
assignCoerce (TypedExpr UnliftedRep [sptr]) (TypedExpr AddrRep [_a_val, a_off]) =
- -- FIXME: (Sylvain 2022-03-11): why can we ignore a_val?
sptr |= a_off
assignCoerce p1 p2 = assignTypedExprs [p1] [p2]
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/13fe0a1148e9c9410739422b3f80cd0778b6b7ba
--
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/13fe0a1148e9c9410739422b3f80cd0778b6b7ba
You're receiving this email because of your account on gitlab.haskell.org.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-commits/attachments/20220825/b2dd4543/attachment-0001.html>
More information about the ghc-commits
mailing list