[Git][ghc/ghc][wip/issue-23821] 18 commits: Remove ScopedTypeVariables => TypeAbstractions

Gergő Érdi (@cactus) gitlab at gitlab.haskell.org
Thu Aug 31 06:44:01 UTC 2023



Gergő Érdi pushed to branch wip/issue-23821 at Glasgow Haskell Compiler / GHC


Commits:
9eecdf33 by sheaf at 2023-08-28T18:54:06+00:00
Remove ScopedTypeVariables => TypeAbstractions

This commit implements [amendment 604](https://github.com/ghc-proposals/ghc-proposals/pull/604/)
to [GHC proposal 448](https://github.com/ghc-proposals/ghc-proposals/pull/448)
by removing the implication of language extensions

  ScopedTypeVariables => TypeAbstractions

To limit breakage, we now allow type arguments in constructor patterns
when both ScopedTypeVariables and TypeApplications are enabled, but
we emit a warning notifying the user that this is deprecated behaviour
that will go away starting in GHC 9.12.

Fixes #23776

- - - - -
fadd5b4d by sheaf at 2023-08-28T18:54:06+00:00
.stderr: ScopedTypeVariables =/> TypeAbstractions

This commit accepts testsuite changes for the changes in the previous
commit, which mean that TypeAbstractions is no longer implied by
ScopedTypeVariables.

- - - - -
4f5fb500 by Greg Steuck at 2023-08-29T07:55:13-04:00
Repair `codes` test on OpenBSD by explicitly requesting extended RE

- - - - -
6bbde581 by Vasily Sterekhov at 2023-08-29T12:06:58-04:00
Add test for #23540

`T23540.hs` makes use of `explainEv` from `HieQueries.hs`, so
`explainEv` has been moved to `TestUtils.hs`.

- - - - -
257bb3bd by Vasily Sterekhov at 2023-08-29T12:06:58-04:00
Add test for #23120

- - - - -
4f192947 by Vasily Sterekhov at 2023-08-29T12:06:58-04:00
Make some evidence uses reachable by toHie

Resolves #23540, #23120

This adds spans to certain expressions in the typechecker and renamer,
and lets 'toHie' make use of those spans. Therefore the relevant
evidence uses for the following syntax will now show up under the
expected nodes in 'HieAst's:

- Overloaded literals ('IsString', 'Num', 'Fractional')

- Natural patterns and N+k patterns ('Eq', 'Ord', and instances from the
  overloaded literals being matched on)

- Arithmetic sequences ('Enum')

- Monadic bind statements ('Monad')

- Monadic body statements ('Monad', 'Alternative')

- ApplicativeDo ('Applicative', 'Functor')

- Overloaded lists ('IsList')

Also see Note [Source locations for implicit function calls]

In the process of handling overloaded lists I added an extra 'SrcSpan'
field to 'VAExpansion' - this allows us to more accurately reconstruct
the locations from the renamer in 'rebuildHsApps'. This also happens to
fix #23120.

See the additions to Note [Looking through HsExpanded]

- - - - -
fe9fcf9d by Sylvain Henry at 2023-08-29T12:07:50-04:00
ghc-heap: rename C file (fix #23898)

- - - - -
b60d6576 by Krzysztof Gogolewski at 2023-08-29T12:08:29-04:00
Misc cleanup

- Builtin.PrimOps: ReturnsAlg was used only for unboxed tuples.
  Rename to ReturnsTuple.
- Builtin.Utils: use SDoc for a panic message.
  The comment about <<details unavailable>> was obsoleted by e8d356773b56.
- TagCheck: fix wrong logic. It was zipping a list 'args' with its
  version 'args_cmm' after filtering.
- Core.Type: remove an outdated 1999 comment about unlifted polymorphic types
- hadrian: remove leftover debugging print

- - - - -
3054fd6d by Krzysztof Gogolewski at 2023-08-29T12:09:08-04:00
Add a regression test for #23903

The bug has been fixed by commit bad2f8b8aa8424.

- - - - -
21584b12 by Ben Gamari at 2023-08-29T19:52:02-04:00
README: Refer to ghc-hq repository for contributor and governance information

- - - - -
e542d590 by sheaf at 2023-08-29T19:52:40-04:00
Export setInertSet from GHC.Tc.Solver.Monad

We used to export getTcSInerts and setTcSInerts from GHC.Tc.Solver.Monad.
These got renamed to getInertSet/setInertSet in e1590ddc. That commit also removed
the export of setInertSet, but that function is useful for the GHC API.

- - - - -
694ec5b1 by sheaf at 2023-08-30T10:18:32-04:00
Don't bundle children for non-parent Avails

We used to bundle all children of the parent Avail with things that
aren't the parent, e.g. with

  class C a where
    type T a
    meth :: ..

we would bundle the whole Avail (C, T, meth) with all of C, T and meth,
instead of only with C.

Avoiding this fixes #23570

- - - - -
d926380d by Krzysztof Gogolewski at 2023-08-30T10:19:08-04:00
Fix typos

- - - - -
d07080d2 by Josh Meredith at 2023-08-30T19:42:32-04:00
JS: Implement missing C functions `rename`, `realpath`, and `getcwd` (#23806)

- - - - -
e2940272 by David Binder at 2023-08-30T19:43:08-04:00
Bump submodules of hpc and hpc-bin to version 0.7.0.0

hpc 0.7.0.0 dropped SafeHaskell safety guarantees in order to simplify
compatibility with newer versions of the directory package which
dropped all SafeHaskell guarantees.

- - - - -
5d56d05c by David Binder at 2023-08-30T19:43:08-04:00
Bump hpc bound in ghc.cabal.in

- - - - -
44d967f2 by Gergő Érdi at 2023-08-31T04:44:01+01:00
If we have multiple defaulting plugins, then we should zonk in between them

after any defaulting has taken place, to avoid a defaulting plugin seeing
a metavariable that has already been filled.

Fixes #23821.

- - - - -
b26097b3 by Gergő Érdi at 2023-08-31T07:43:50+01:00
Improvements to the documentation of defaulting plugins

Based on @simonpj's draft and comments in !11117

- - - - -


30 changed files:

- .gitlab/ci.sh
- .gitlab/rel_eng/upload.sh
- README.md
- compiler/GHC/Builtin/PrimOps.hs
- compiler/GHC/Builtin/Types.hs
- compiler/GHC/Builtin/Types/Prim.hs
- compiler/GHC/Builtin/Utils.hs
- compiler/GHC/CmmToAsm.hs
- compiler/GHC/CmmToAsm/Reg/Graph/TrivColorable.hs
- compiler/GHC/CmmToAsm/X86/Instr.hs
- compiler/GHC/CmmToAsm/X86/Regs.hs
- compiler/GHC/CmmToLlvm/Base.hs
- compiler/GHC/Core/Coercion/Opt.hs
- compiler/GHC/Core/Make.hs
- compiler/GHC/Core/Opt/CSE.hs
- compiler/GHC/Core/Opt/ConstantFold.hs
- compiler/GHC/Core/Opt/DmdAnal.hs
- compiler/GHC/Core/Opt/SetLevels.hs
- compiler/GHC/Core/Opt/Specialise.hs
- compiler/GHC/Core/Stats.hs
- compiler/GHC/Core/TyCo/Rep.hs
- compiler/GHC/Core/TyCon.hs
- compiler/GHC/Core/Type.hs
- compiler/GHC/CoreToStg.hs
- compiler/GHC/Driver/Config/HsToCore.hs
- compiler/GHC/Driver/DynFlags.hs
- compiler/GHC/Driver/Make.hs
- compiler/GHC/Driver/Session.hs
- compiler/GHC/Hs/Utils.hs
- compiler/GHC/HsToCore/Usage.hs


The diff was not included because it is too large.


View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/79789b1ddd27b66184110b4eb5eaee67f32ab1cd...b26097b3d3889c809d56c8d41c8e35b0231a13d2

-- 
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/compare/79789b1ddd27b66184110b4eb5eaee67f32ab1cd...b26097b3d3889c809d56c8d41c8e35b0231a13d2
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/20230831/2c425d21/attachment-0001.html>


More information about the ghc-commits mailing list