Skip to content

Sync with upstream (2025-10-08): bb22f7d4d63446c9095db32ca013a9b2182df7d9 #4200

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom
Open
wants to merge 946 commits into
base: master
Choose a base branch
from
Open

Sync with upstream (2025-10-08): bb22f7d4d63446c9095db32ca013a9b2182df7d9 #4200

wants to merge 946 commits into from

Conversation

dkm
Copy link
Member

@dkm dkm commented Oct 8, 2025 •
edited
Loading

This is a sync with upstream GCC:

-- gerris 🦀

@dkm dkm force-pushed the gerris/rebase/2025-10-08 branch 2 times, most recently from d65e0fc to 6b23cb7 Compare October 8, 2025 08:37
@dkm
Copy link
Member Author

dkm commented Oct 8, 2025

FTR, I've rebased/squash all my CI stuff for sending mails + added a commit to bump the macos dep.

@dkm
Copy link
Member Author

dkm commented Oct 8, 2025

@powerboat9 doesn't seem to work for macos... Do we really want to have a WIP branch in our CI?

@dkm dkm self-assigned this Oct 8, 2025
@dkm
Copy link
Member Author

dkm commented Oct 8, 2025

And FTR, before pushing this, I'll fix the mail workflow to really skip too-big PR and check that the label checking really works.

@dkm dkm force-pushed the gerris/rebase/2025-10-08 branch from 6b23cb7 to 0554e02 Compare October 8, 2025 18:43
@thesamesam
Copy link
Contributor

IMO having macOS in there is too brittle for now because we're combining 3 trees in cases like this (gccrs, gcc upstream, iain's WIP fork).

@dkm
Copy link
Member Author

dkm commented Oct 9, 2025

I'll give it one more chance. iains is going to rebase his branch and fix the conflicts. I'll then rebase my branch over a more recent revision and everything should "just work"®

philberty and others added 21 commits October 9, 2025 20:42
@philberty @dkm
Remove clang-format from bors
a1797cc
@mvvsmk @dkm
Added Remark.yml and .remarkrc
3aaf499 
Addresses #487
Remark.yml contains the workflow and .remarkrc is the linter's config file.
Signed-off-by: M V V S Manoj Kumar <[email protected]>
@tschwinge @dkm
Running clang-format locally [#242]
b8355fd 
Apparently everyone but me knows how to do that; I'm now documenting it for my
own future use.
@philberty @dkm
Add script to generate gource
2f3eb04 
This script tries to isolate our changes to GCC minus the changes to
gcc/config so that we can visualize our development history.
@philberty @dkm
add script to graph the changes over time
c711d60
@CohenArthur @dkm
dockerfile: Install cargo-gccrs alongside gccrs
1c17fb0
@CohenArthur @dkm
ci: Do not allow unresolved test cases
652ee0d
@CastilloDel @dkm
Add a check for new warnings to the CI
cfc2ca9 
This should prevent new warnings from appearing silently

Signed-off-by: Daniel del Castillo [email protected]
@tschwinge @dkm
Force 'LC_ALL=C' for all steps of '.github/workflows/ccpp.yml:jobs.bu…
27cf52d 
…ild-and-check'
@tschwinge @dkm
Force 'LC_ALL=C.UTF-8' for all steps of '.github/workflows/ccpp.yml:j…
d0bc784 
…obs.build-and-check'

That's what I use in my local development enviroment, and I suppose I'll
be the main one touching this file semi-regularly in context of
<#247> "Rebasing against GCC".
@tschwinge @dkm
Merge 'Build logs' into 'Build' in '.github/workflows/ccpp.yml:jobs.b…
7343d94 
…uild-and-check'

This avoids the supposed issue that in case that 'make' fails, the whole
'jobs.build-and-check' stops, and 'Build logs' isn't executed, and thus there's
no indication in the GitHub UI why 'make' failed.

Using a shell pipeline is OK; the exit code of 'make' isn't lost, as per
<https://docs.github.com/en/actions/using-workflows/workflow-syntax-for-github-actions#jobsjob_idstepsshell>,
'bash' is being run with '-o pipefail'.
@tschwinge @dkm
Tune '.github/workflows/ccpp.yml:jobs.build-and-check', 'Check for ne…
79279a9 
…w warnings' step

Run it in scratch directory, too, to not pollute the pristine sources
directory.  Point to <#1026> in case of
failure.
@CohenArthur @dkm
ci: Add job for building gccrs with older gcc
983e6aa 
As Thomas Schwinge pointed out, GCC 4.8 is the minimum version to be
used for building current GCC, meaning that we should make an effort to
support it before we consider upstreaming or backporting. The main
differences are probably a less powerful standard template library or
small compilation differences.
@CohenArthur @dkm
ci: Run tests with gccrs compiled under gcc-4.8
92e3199
@simonpcook @dkm
rust: Clang/macOS Testing
bb89d13 
This adds a version of the build-and-check job that runs with clang on
macOS.
@simonpcook @dkm
rust: Use -Otarget when building and logging warnings
c7ac8d5 
This will provide some synchronization for output lines, and so will
make comparisons with known warnings as part of CI more reliable.
@liushuyu @dkm
CI: do not push Docker image ...
68b0cff 
... when the workflow is ran in a forked repository

Signed-off-by: Zixing Liu <[email protected]>
@liushuyu @dkm
CI: catch malformed test cases
5546247 
Signed-off-by: Zixing Liu <[email protected]>
@liushuyu @dkm
dockerfile: change the base image ...
2e53979 
... to match Rust container image's base image

Signed-off-by: Zixing Liu <[email protected]>
@CohenArthur @dkm
docker: Add commit information to Docker image
0af2ecb
@CohenArthur @dkm
docker: Fix GCCRS_BUILD info
e53f7ce 
Sometimes, the GCCRS_BUILD file would end up empty. I believe this is
due to how Dockerfiles work in that there was a "race condition" between
the two `RUN` commands. Hopefully this should fix it.
philberty and others added 25 commits October 9, 2025 20:42
@philberty @dkm
gccrs: Remove old hack from resolve operator overload
fcdbcf4 
This was an old chunk of code to try and sort out the Self type on resolving
methods for impl blocks. But this is not required when our unify code works
properly.

gcc/rust/ChangeLog:

	* typecheck/rust-hir-type-check-expr.cc: remove old hack

Signed-off-by: Philip Herron <[email protected]>
@philberty @dkm
gccrs: Fix ICE on copied array expressions
b3be7a0 
We need to check for errors on the number of copies expression before
trying to const fold it otherwise it will just fail in the const evaluator.

Fixes #4165

gcc/rust/ChangeLog:

	* typecheck/rust-hir-type-check-expr.cc (TypeCheckExpr::visit): check for error

gcc/testsuite/ChangeLog:

	* rust/compile/issue-4165.rs: New test.

Signed-off-by: Philip Herron <[email protected]>
@philberty @dkm
gccrs: Add error diag for self params on plain functions
0236d14 
self params dont have a type unless used within impl blocks. Rustc as far as I
can tell in this senario generics a synthetic param of type Self in this senario
so that it keeps consistent error diagnostic for number of parameters but the
logic for what the parameter typpe should be seems unclear.

Fixes #3592

gcc/rust/ChangeLog:

	* hir/rust-ast-lower-item.cc (ASTLoweringItem::visit): add error diagnostic

gcc/testsuite/ChangeLog:

	* rust/compile/issue-3592.rs: New test.

Signed-off-by: Philip Herron <[email protected]>
@philberty @dkm
gccrs: Fix ICE where we expect a num enum variant
2c9be33 
This changes the assertion into a valid error diagnostic.

Fixes #3538

gcc/rust/ChangeLog:

	* backend/rust-compile-resolve-path.cc: add error diag

gcc/testsuite/ChangeLog:

	* rust/compile/issue-3538.rs: New test.

Signed-off-by: Philip Herron <[email protected]>
@dkm
ci: new workflow for sending emails after merging a PR
641b3a4 
This new workflow is used to send emails to the gcc-patches@ mailing list
after every merged PR.

For each merged PR:
- remove commits touching files not to-be-upstreamed (e.g. github
workflow, README.md, ...)
- send a mail series

Each mail contains explanation + links to commits on the github project.

Authors of commits will be put in Cc: of emails.

Configure it by setting the following secrets in github:
- PATCH_TO: the "To:" field for the emails
- PATCH_CC: optional "Cc:"
- SMTP_FROM, SMTP_PASSWORD, SMTP_PORT, SMTP_SERVER, STMP_USERNAME: self explanatory

ChangeLog:

	* .github/workflows/send-emails.yml: New file.

Signed-off-by: Marc Poulhiès <[email protected]>
@P-E-P @dkm
Prevent segfault on builtin metavars
f9fa30f 
Builtin metavars are not in the fragment, the compiler should not emit
an error message.

gcc/rust/ChangeLog:

	* expand/rust-macro-substitute-ctx.cc (is_builtin_metavariable): Add
	function to check builtin metavars knowledge.
	(SubstituteCtx::check_repetition_amount): Do not process missing
	fragment.

Signed-off-by: Pierre-Emmanuel Patry <[email protected]>
@P-E-P @dkm
Emit errors from the transcriber when they occur
d1cb2b8 
Emitting the errors later means some error could be emitted multiple
times.

gcc/rust/ChangeLog:

	* expand/rust-macro-expand.cc (transcribe_expression): Emit error
	early.

Signed-off-by: Pierre-Emmanuel Patry <[email protected]>
@P-E-P @dkm
Add origin location to expanded tokens
2ddc05c 
gcc/rust/ChangeLog:

	* expand/rust-macro-expand.cc: Forward invocation tree locus to
	substitution context.
	* expand/rust-macro-substitute-ctx.cc: Use origin location for expanded
	tokens.
	* expand/rust-macro-substitute-ctx.h (class SubstituteCtx): Save
	invocation location.

gcc/testsuite/ChangeLog:

	* rust/compile/macros/mbe/macro58.rs: New test.

Signed-off-by: Pierre-Emmanuel Patry <[email protected]>
@CohenArthur @dkm
expand: Inherit from PointerVisitor
cb91362 
gcc/rust/ChangeLog:

	* expand/rust-expand-visitor.cc: Remove old visitors.
	* expand/rust-expand-visitor.h: Likewise, plus inherit from PointerVisitor.
@powerboat9 @dkm
Handle link_name attribute
35e2a7d 
gcc/rust/ChangeLog:

	* backend/rust-compile-extern.h: Add includes.
	(CompileExternItem::visit): Use get_link_name.
	(CompileExternItem::get_link_name): New static member function.
	* util/rust-attribute-values.h (Attributes::LINK_NAME): New
	static constexpr member variable.
	* util/rust-attributes.cc (__definitions): New entry for
	LINK_NAME.
	* util/rust-ggc.cc: Include "rust-ast.h".
	(Ident::Ident): Add overload for Rust::Identifier.
	* util/rust-ggc.h (class Identifier): Forward declare.
	(Ident::Ident): Add overload for Rust::Identifier.

gcc/testsuite/ChangeLog:

	* rust/execute/torture/link-name.rs: New test.

Signed-off-by: Owen Avery <[email protected]>
@Polygonalr @dkm
Add HIR lowering support for rest pattern in struct patterns' AST
14d6664 
gcc/rust/ChangeLog:

	* ast/rust-pattern.h (StructPatternElements): Rename has_struct_pattern_etc
	to has_rest_pattern, and has_etc to has_rest to signify presense of rest
	patterns more clearly.
	* ast/rust-pattern.cc (StructPatternElements::as_string): Rename variables
	accordingly.
	* ast/rust-ast-collector.cc: Rename variables accordingly.
	* expand/rust-cfg-strip.cc: Rename variables accordingly.
	* parse/rust-parse-impl.h: Rename variable accordingly.
	* hir/tree/rust-hir-pattern.h (StructPatternElements): Add a boolean to track
	presense of rest pattern.
	* hir/rust-ast-lower-pattern.cc (visit(StructPattern)): Add support for
	lowering rest pattern to HIR.
	* typecheck/rust-hir-type-check-pattern.cc (visit(StructPattern)): Remove
	size check when rest pattern is present.

Signed-off-by: Yap Zhi Heng <[email protected]>
@Polygonalr @dkm
Fix StructPattern type checking rejecting tuple struct scrutinee
931c867 
Previously, type checking of StructPattern will throw an error if it is used
to match against a tuple struct, even though it is possible to do so in rustc.

gcc/rust/ChangeLog:

	* typecheck/rust-hir-type-check-pattern.cc (visit(StructPattern)): Add type check
	support for StructPatterns matching against a TUPLE type ADT.
	* backend/rust-compile-pattern.cc(CompilePatternBindings::visit(StructPattern)):
	Update assert to allow TUPLE type ADTs.
	* hir/tree/rust-hir.cc (StructPatternField::as_string()): Improve info dumped.

Signed-off-by: Yap Zhi Heng <[email protected]>
@powerboat9 @dkm
Test libcore
74c2043 
This adds compilation of libcore up to the AST validation phase to the
testsuite.

gcc/testsuite/ChangeLog:

	* rust/core/core.exp: New test.

Signed-off-by: Owen Avery <[email protected]>
@0xllx0 @dkm
tests: added extra parens compile test
841905d 
Adds a test to verify extra parentheses are discarded, and the source is
compiled as expected.

Resolves: #2886

gcc/testsuite/ChangeLog:

	* rust/compile/issue-4148.rs: New test.

Signed-off-by: Elle Rhumsaa <[email protected]>
@Polygonalr @dkm
Implement lowering of AST::TupleStructItemsHasRest
3b90363 
gcc/rust/ChangeLog:

	* hir/rust-ast-lower-pattern.cc (ASTLoweringPattern::visit(TupleStructPattern)):
	Implement lowering of AST::TupleStructItemsHasRest to HIR.
	* typecheck/rust-hir-type-check-pattern.cc (TypeCheckPattern::visit(TupleStructPattern)):
	Add the respective type checking for AST::TupleStructItemsHasRest
	* checks/errors/rust-hir-pattern-analysis.cc (lower_tuple_pattern):
	Add respective pattern for HAS_REST case.

Signed-off-by: Yap Zhi Heng <[email protected]>
@Polygonalr @dkm
Implement compilation support for HIR::TupleStructItemsHasRest
f1250ef 
gcc/rust/ChangeLog:

	* backend/rust-compile-pattern.cc: Add compilation support for TupleStructItemsHasRest
	in CompilePatternCheckExpr(TupleStructPattern) & CompilePatternBindings(TupleStructPattern)

Signed-off-by: Yap Zhi Heng <[email protected]>
@Polygonalr @dkm
Use rich_location for TupleStructPattern type check num field error
b03f541 
gcc/rust/ChangeLog:

	* typecheck/rust-hir-type-check-pattern.cc (visit(TupleStructPattern)):
	Update error for mismatched number of fields to use rich_location.

Signed-off-by: Yap Zhi Heng <[email protected]>
@P-E-P @dkm
Emit an error on malformed path
3f425b5 
Path must be made of a single literal item, otherwise an error should be
thrown.

gcc/rust/ChangeLog:

	* util/rust-attributes.cc (AttributeChecker::check_attribute): Recurse
	within attr input for additional attribute checking.
	(AttributeChecker::visit): Remove empty definition in favor of default
	ast visitor definition.
	* util/rust-attributes.h: Remove now unused prototypes.

Signed-off-by: Pierre-Emmanuel Patry <[email protected]>
@P-E-P @dkm
Add override modifier
7299ccf 
gcc/rust/ChangeLog:

	* ast/rust-ast.h: Add missing override modifier.
	* ast/rust-path.h: Likewise.

Signed-off-by: Pierre-Emmanuel Patry <[email protected]>
@P-E-P @dkm
Force crash when retrieving meta item location
3446dfa 
We still don't know which location should be preferred over the other,
this means that nobody should rely on this function's return value.

gcc/rust/ChangeLog:

	* ast/rust-expr.h: Force crash when retrieving locus.

Signed-off-by: Pierre-Emmanuel Patry <[email protected]>
@P-E-P @dkm
Avoid malformed attribute conversion
0da5e69 
Add location for malformed attribute errors and avoid processing of
malformed attributes.

gcc/rust/ChangeLog:

	* ast/rust-ast.cc (AttrInputMetaItemContainer::separate_cfg_attrs):
	Avoid malformed attributes.
	* util/rust-attributes.cc (AttributeChecker::visit): Change location.

gcc/testsuite/ChangeLog:

	* rust/compile/attr_malformed_path.rs: New test.

Signed-off-by: Pierre-Emmanuel Patry <[email protected]>
@P-E-P @dkm
Add doc attribute checking on traits
216f07f 
Doc attribute checking was not performed on traits.

gcc/rust/ChangeLog:

	* util/rust-attributes.cc (check_doc_attribute): Change message.
	(AttributeChecker::visit): Add doc attribute checking on traits.

gcc/testsuite/ChangeLog:

	* rust/compile/attr_malformed_doc.rs: New test.

Signed-off-by: Pierre-Emmanuel Patry <[email protected]>
@0xllx0 @dkm
fix: add early return for empty module file
1a7ddf1 
Converts an assert into an early return during AST parsing.

Resolves: #4145

gcc/rust/ChangeLog:

	* ast/rust-ast.cc (Module::process_file_path): empty module early return

Signed-off-by: Elle Rhumsaa <[email protected]>
@0xllx0 @dkm
test: add a regression test for issue #4145
43a684a 
gcc/testsuite/ChangeLog:

	* rust/compile/issue-4145.rs: New test.

Signed-off-by: Elle Rhumsaa <[email protected]>
@philberty @dkm
gccrs: Refactor TyTy::ConstType into separate types
4f5d3fe 
This patch refactors the const generic type system to follow the same pattern
as regular type parameters. The monolithic ConstType is split into four
distinct types:

  ConstParamType (generic parameter placeholder)
  ConstValueType (resolved constant value)
  ConstInferType (inference variable)
  ConstErrorType (error sentinel)

gcc/rust/ChangeLog:

	* backend/rust-compile-expr.cc (CompileExpr::array_copied_expr): refactor to new classes
	* backend/rust-compile-pattern.cc (CompilePatternCheckExpr::visit): likewise
	(CompilePatternBindings::visit): likewise
	* backend/rust-compile-type.cc (TyTyResolveCompile::visit): likewise
	* backend/rust-compile-type.h: likewise
	* typecheck/rust-hir-type-check-base.cc (TypeCheckBase::resolve_literal): likewise
	* typecheck/rust-hir-type-check-expr.cc (TypeCheckExpr::visit): likewise
	* typecheck/rust-hir-type-check-pattern.cc (TypeCheckPattern::visit): likewise
	* typecheck/rust-hir-type-check-type.cc (TypeCheckType::visit): likewise
	* typecheck/rust-substitution-mapper.cc (SubstMapperInternal::visit): likewise
	* typecheck/rust-substitution-mapper.h: likewise
	* typecheck/rust-type-util.cc (unify_site_and): remove bad delete (alpine32)
	* typecheck/rust-tyty-call.h: refactor to new classes
	* typecheck/rust-tyty-subst.cc (SubstitutionParamMapping::clone): likewise
	(SubstitutionRef::infer_substitions): likewise
	* typecheck/rust-tyty-util.cc (TyVar::get_implicit_const_infer_var): likewise
	* typecheck/rust-tyty-util.h: likewise
	* typecheck/rust-tyty-variance-analysis-private.h: likewise
	* typecheck/rust-tyty-visitor.h: likewise
	* typecheck/rust-tyty.cc (BaseType::destructure): likewise
	(BaseType::monomorphized_clone): likewise
	(BaseType::is_concrete): likewise
	(VariantDef::clone): likewise
	(VariantDef::monomorphized_clone): likewise
	(ArrayType::as_string): likewise
	(ArrayType::get_capacity): likewise
	(ArrayType::handle_substitions): likewise
	(generate_tree_str): likewise
	(ConstType::ConstType): likewise
	(ConstParamType::ConstParamType): likewise
	(ConstType::accept_vis): likewise
	(ConstParamType::const_kind): likewise
	(ConstParamType::get_symbol): likewise
	(ConstParamType::can_resolve): likewise
	(ConstParamType::resolve): likewise
	(ConstParamType::accept_vis): likewise
	(ConstType::set_value): likewise
	(ConstType::as_string): likewise
	(ConstParamType::as_string): likewise
	(ConstType::clone): likewise
	(ConstParamType::clone): likewise
	(ConstType::get_symbol): likewise
	(ConstParamType::get_name): likewise
	(ConstType::can_resolve): likewise
	(ConstParamType::is_equal): likewise
	(ConstType::resolve): likewise
	(ConstValueType::ConstValueType): likewise
	(ConstValueType::const_kind): likewise
	(ConstValueType::accept_vis): likewise
	(ConstValueType::as_string): likewise
	(ConstValueType::clone): likewise
	(ConstValueType::get_name): likewise
	(ConstValueType::is_equal): likewise
	(ConstValueType::get_value): likewise
	(ConstInferType::ConstInferType): likewise
	(ConstInferType::const_kind): likewise
	(ConstInferType::accept_vis): likewise
	(ConstType::get_name): likewise
	(ConstInferType::as_string): likewise
	(ConstInferType::clone): likewise
	(ConstInferType::get_name): likewise
	(ConstType::is_equal): likewise
	(ConstInferType::is_equal): likewise
	(ConstErrorType::ConstErrorType): likewise
	(ConstErrorType::const_kind): likewise
	(ConstType::handle_substitions): likewise
	(ConstErrorType::accept_vis): likewise
	(ConstErrorType::as_string): likewise
	(ConstErrorType::clone): likewise
	(ConstErrorType::get_name): likewise
	(ConstErrorType::is_equal): likewise
	* typecheck/rust-tyty.h (class BaseConstType): likewise
	(class ConstType): likewise
	(class ConstParamType): likewise
	(class ConstValueType): likewise
	(class ConstInferType): likewise
	(class ConstErrorType): likewise
	* typecheck/rust-unify.cc (UnifyRules::commit): likewise
	(UnifyRules::go): likewise
	(UnifyRules::expect_array): likewise
	(UnifyRules::expect_const): likewise
	* typecheck/rust-unify.h: likewise

Signed-off-by: Philip Herron <[email protected]>
@dkm dkm force-pushed the gerris/rebase/2025-10-08 branch from 0554e02 to 196a6a8 Compare October 9, 2025 18:43
dkm added 3 commits October 9, 2025 20:44
@dkm
ci: bump Apple branch
fd690d7 
Signed-off-by: Marc Poulhiès <[email protected]>
@dkm
ci: refactor conditional parts
4d4aca1 
Cap the number of commits to 30. Anything above that is suspicious and we should not send 30 mails at once.
Note that github REST API has a limit of 250 elements in arrays, so "huge" PR is hard to handle anyway.

Also change the conditional to fix them (!!) and make it clearer.

Signed-off-by: Marc Poulhiès <[email protected]>
@dkm
Merge remote-tracking branch 'upstream-gccrs/master' into gerris/reba…
b31f7a5 
…se/2025-10-08

This branch has a no-op merge as the last commit:
 - one arm is the "current" development branch from github
 - the other arm is a rebased version of the "current" master branch onto a recent GCC's master

The merge is obtained with "git merge --strategy=ours" to only keep the changes from second arm.
@dkm dkm force-pushed the gerris/rebase/2025-10-08 branch from 196a6a8 to b31f7a5 Compare October 9, 2025 18:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

@dkm dkm

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.