Since the mangling of the second lambda changed (previously we counted all
lambdas, now we only count lambdas with the same signature), we
generate_mangling_alias for handler<lambda2> for backward compatibility.
Since handler is COMDAT, resolve_alias puts the alias in the same comdat
group as handler itself. Then create_dispatcher_calls tries to add the
alias to the same comdat group as the dispatcher, but it's already in a
same_comdat_group, so we ICE.
It seems like we're just missing a remove_from_same_comdat_group before
add_to_same_comdat_group.
PR c++/114992
gcc/ChangeLog:
* multiple_target.cc (create_dispatcher_calls):
remove_from_same_comdat_group before add_to_same_comdat_group.
gcc/testsuite/ChangeLog:
* g++.target/i386/mangling-alias1.C: New test.
2025-03-21 Paul Thomas <pault@gcc.gnu.org>
gcc/fortran
PR fortran/85836
* check.cc (get_ul_from_cst_cl): New function used in
check_operation.
(check_operation): New function used in check_reduce and
check_co_reduce.
(gfc_check_co_reduce): Use it.
(gfc_check_reduce): New function.
(gfc_check_rename): Add prototype for intrinsic with 6 arguments.
* gfortran.h : Add isym id for reduce and prototype for f6.
* intrinsic.cc (do_check): Add another argument expression and use
it in the call to the six argument specific check.
(add_sym_6): New function.
(add_functions): Add the discription of the reduce intrinsic and
add it to the intrinsic list.
* intrinsic.h : Add prototypes for gfc_check_reduce and
gfc_resolve_reduce.
* iresolve.cc (generate_reduce_op_wrapper): Generate a wrapper
subroutine for the 'operation' function to enable the library
implementation to be type agnostic and use pointer arithmetic
throughout.
(gfc_resolve_reduce): New function.
* trans-expr.cc (gfc_conv_procedure_call): Add flag for scalar
reduce. Generate a return variable 'sr' for scalar reduce, pass its
address to the library function and return it as the scalar result.
* trans-intrinsic.cc (gfc_conv_intrinsic_function): Array valued
reduce is called in same way as reshape. Fall through for call to
the scalar version.
gcc/testsuite/
PR fortran/85836
* gfortran.dg/reduce_1.f90: New test
* gfortran.dg/reduce_2.f90: New test
libgfortran/
PR libfortran/85836
* Makefile.am : Add reduce.c
* Makefile.in : Regenerated
* gfortran.map : Add _gfortran_reduce, _gfortran_reduce_scalar,
_gfortran_reduce_c and _gfortran_reduce_scalar_c to the list.
* intrinsics/reduce.c (reduce, reduce_scalar, reduce_c,
reduce_scalar_c): New functions and prototypes
These tests have been looking for a very specific instruction sequence
which has the tendency to be fairly unstable as a result. But what is
more interesting is that the the tests must not contain instructions
that can't be used for unaligned data, and whether or not the copy is
executed correctly.
So make these tests executable and scan the assembler only to confirm
the absence of instructions that must not be used when the data is not
aligned.
These tests also used to be restricted to targets that support
unaligned accesses (because you get very different code otherwise).
But now we've made the tests executable and to check for the absence
of problem instructions, just falling back to memcpy *is* an
acceptable implementation. So remove the requirement for unaligned
accesses.
gcc/testsuite:
PR target/91614
* gcc.target/arm/unaligned-memcpy-1.c: Make the test executable.
Only scan for the absence of instructions that cannot access
misaligned data. Remove constraint of having unaligned accesses.
* gcc.target/arm/unaligned-memcpy-2.c: Likewise.
* gcc.target/arm/unaligned-memcpy-3.c: Likewise.
* gcc.target/arm/unaligned-memcpy-4.c: Likewise.
This test is designed to check that if one of the operands is
aligned (but the other isn't) we expand to a sensible sequence and
bypass most of the overhead of doing a memcpy. But on targets without
unaligned accessess, we still end up calling memcpy. It's then a
lottery as to whether the prologue and epilogue code, plus the
set-up for the memcpy itself, generate instructions that match the
scan patterns.
Since in those cases we're not actually testing what the test is looking
for anyway, just skip the test on strict-alignment targets.
gcc/testsuite:
* gcc.target/arm/memcpy-aligned-1.c: Require unaligned accesses.
We were lying about the return type, but that's not necessary; we already
need to handle a COND_EXPR where one side is void for THROW_EXPR.
This fixes an execution failure on nvptx:
error: Prototype doesn't match for '__cxa_throw_bad_array_new_length'
gcc/cp/ChangeLog:
* init.cc (throw_bad_array_new_length): Returns void.
In 2001 Subversion r40924 (Git commit 52a11cbfcf)
"IA-64 ABI Exception Handling", '__cxa_bad_cast' changed from 'void *' to
'void' return type:
--- libstdc++-v3/libsupc++/exception_support.cc
+++ /dev/null
@@ -1,388 +0,0 @@
-[...]
-// Helpers for rtti. Although these don't return, we give them return types so
-// that the type system is not broken.
-extern "C" void *
-__cxa_bad_cast ()
-{
- [...]
-}
-[...]
--- /dev/null
+++ libstdc++-v3/libsupc++/unwind-cxx.h
@@ -0,0 +1,163 @@
+[...]
+extern "C" void __cxa_bad_cast ();
+[...]
--- /dev/null
+++ libstdc++-v3/libsupc++/eh_aux_runtime.cc
@@ -0,0 +1,56 @@
+[...]
+extern "C" void
+__cxa_bad_cast ()
+{
+ [...]
+}
+[...]
The implicit prototype in the C++ front end however wasn't likewise adjusted,
and so for nvptx we generate code for 'void *' return type:
// BEGIN GLOBAL FUNCTION DECL: __cxa_bad_cast
.extern .func (.param .u64 %value_out) __cxa_bad_cast;
{
.param .u64 %value_in;
call (%value_in),__cxa_bad_cast;
trap;
// (noreturn)
exit;
// (noreturn)
ld.param.u64 %r30,[%value_in];
}
..., which is in conflict with the library code with 'void' return type:
// BEGIN GLOBAL FUNCTION DECL: __cxa_bad_cast
.visible .func __cxa_bad_cast;
// BEGIN GLOBAL FUNCTION DEF: __cxa_bad_cast
.visible .func __cxa_bad_cast
{
[...]
}
..., and we thus get execution test FAIL for 'g++.dg/rtti/dyncast2.C':
error : Prototype doesn't match for '__cxa_bad_cast' in 'input file 7 at offset 51437', first defined in 'input file 7 at offset 51437'
nvptx-run: cuLinkAddData failed: device kernel image is invalid (CUDA_ERROR_INVALID_SOURCE, 300)
With this patched, we get the expected:
// BEGIN GLOBAL FUNCTION DECL: __cxa_bad_cast
-.extern .func (.param .u64 %value_out) __cxa_bad_cast;
+.extern .func __cxa_bad_cast;
{
-.param .u64 %value_in;
-call (%value_in),__cxa_bad_cast;
+call __cxa_bad_cast;
trap;
// (noreturn)
exit;
// (noreturn)
-ld.param.u64 %r30,[%value_in];
}
..., and execution test PASS.
gcc/cp/
* rtti.cc (throw_bad_cast): Adjust implicit '__cxa_bad_cast'
prototype to reality.
... next to '-malias' variant: commit a1865fd338
"Add 'g++.target/nvptx/alias-g++.dg_init_dtor2-1.C'", to document what we're
doing to '-mno-alias'.
gcc/testsuite/
* g++.target/nvptx/alias-g++.dg_init_dtor2-2.C: New.
libgomp.exp added -fno-diagnostics-show-caret and -fdiagnostics-color=never
as 'additional_flags' for compilation. However, it turned out that this now
is insufficient as the [...] part of diagnostics have a hyperlink URL.
Solution: Use the -fdiagnostics-plain-output flag instead, added in commit
r11-2701-g129a1319c0ab73. This flag currently implies the following flags:
-fno-diagnostics-show-caret
-fno-diagnostics-show-line-numbers
-fdiagnostics-color=never
-fdiagnostics-urls=never
-fdiagnostics-path-format=separate-events
-fdiagnostics-text-art-charset=none
-fno-diagnostics-show-event-links
libgomp/ChangeLog:
* testsuite/lib/libgomp.exp (libgomp_init): Add
-fdiagnostics-plain-output to additional_flags; remove
-fno-diagnostics-show-caret and -fdiagnostics-color=never.
gcc/testsuite/ChangeLog:
* rust/compile/nr2/exclude: Add failing lang item typepaths tests.
* rust/execute/torture/derive_macro4.rs: Mark Clone as lang item.
gcc/rust/ChangeLog:
* expand/rust-derive-clone.cc (DeriveClone::clone_enum_struct): New function for deriving
enum struct variants.
(DeriveClone::visit_enum): Call into the new function.
gcc/testsuite/ChangeLog:
* rust/compile/nr2/exclude:
* rust/compile/derive_clone_enum1.rs: New test.
* rust/compile/derive_clone_enum2.rs: New test.
* rust/compile/derive_clone_enum3.rs: New test.
* rust/execute/torture/derive_clone_enum1.rs: New test.
gcc/rust/ChangeLog:
* ast/rust-ast-builder.cc: Add new methods for constructing struct exprs.
* ast/rust-ast-builder.h: Mention how to build tuple expressions.
This method errors out if the lang item has not been declared yet.
gcc/rust/ChangeLog:
* util/rust-hir-map.cc (Mappings::get_lang_item_node): New.
* util/rust-hir-map.h: New function.
Which formats a lang item as it appears in source code.
gcc/rust/ChangeLog:
* util/rust-lang-item.cc (LangItem::PrettyString): New.
* util/rust-lang-item.h: New.
Lang item typepaths were not handled properly, and required a complete overhaul.
All old classes that concerned lang item paths are now modified to use a simpler
version of `AST::LangItemPath`, which has been removed. TypePath segments can now
be lang items, as this is requied for having generic lang item paths such as
PhantomData<T>.
gcc/rust/ChangeLog:
* ast/rust-path.h: Rework how lang item paths are represented.
* ast/rust-path.cc: Likewise.
* ast/rust-item.h: Likewise.
* ast/rust-ast.cc: Likewise.
* ast/rust-ast-collector.cc: Adapt to new lang item path system.
* ast/rust-ast-collector.h: Likewise.
* ast/rust-ast-visitor.cc (DefaultASTVisitor::visit): Likewise.
* ast/rust-ast-visitor.h: Likewise.
* expand/rust-derive-copy.cc: Likewise.
* expand/rust-derive.h: Likewise.
* hir/rust-ast-lower-base.cc (ASTLoweringBase::visit): Likewise.
* hir/rust-ast-lower-base.h: Likewise.
* hir/rust-ast-lower-type.cc (ASTLowerTypePath::translate): Likewise.
(ASTLowerTypePath::visit): Likewise.
* hir/rust-ast-lower-type.h: Likewise.
* resolve/rust-ast-resolve-base.cc (ResolverBase::visit): Likewise.
* resolve/rust-ast-resolve-base.h: Likewise.
* resolve/rust-ast-resolve-item.cc (ResolveItem::visit): Likewise.
* resolve/rust-ast-resolve-type.h: Likewise.
* resolve/rust-ast-resolve-type.cc (ResolveRelativeTypePath::go): Likewise.
* resolve/rust-late-name-resolver-2.0.cc (Late::visit): Likewise.
* resolve/rust-late-name-resolver-2.0.h: Likewise.
* hir/tree/rust-hir-path.cc (TypePathSegment::TypePathSegment): Likewise.
(TypePathSegmentGeneric::TypePathSegmentGeneric): Likewise.
* hir/tree/rust-hir-path.h: Likewise.
* typecheck/rust-hir-type-check-type.cc (TypeCheckType::resolve_root_path): Likewise.
* ast/rust-ast-builder.cc: Likewise.
* ast/rust-ast-builder.h: Likewise.
We need to add name resolution and hir lowering for items as part of blocks
in order to typecheck and compile them correctly.
Fixes Rust-GCC#3350
gcc/rust/ChangeLog:
* hir/rust-ast-lower-stmt.cc (ASTLoweringStmt::visit): hir lowering
* hir/rust-ast-lower-stmt.h: likewise
* resolve/rust-ast-resolve-stmt.cc (ResolveStmt::visit): name resolution
* resolve/rust-ast-resolve-stmt.h: likewise
gcc/testsuite/ChangeLog:
* rust/compile/issue-3350.rs: New test.
Signed-off-by: Philip Herron <herron.philip@googlemail.com>
gcc/testsuite/ChangeLog:
* rust/compile/macros/builtin/eager1.rs: Switch to scan-assembler directive as the
GIMPLE dump does not contain strings on LE.
* rust/compile/macros/builtin/recurse2.rs: Likewise.
Some failures occur in the testsuite because we
did not account for the little-endian case.
gcc/testsuite/ChangeLog:
* rust/compile/issue-1446.rs: Add swap_bytes function.
* rust/compile/iterators1.rs: Remove unused {to, from}_le functions.
ForeverStackStore is meant to partially unify the internal states of
per-namespace ForeverStack instances. This commit does not contain
modifications to ForeverStack which would allow it to rely on a
ForeverStackStore to store nodes, but a future commit should address
this.
gcc/rust/ChangeLog:
* Make-lang.in: Handle rust-forever-stack.cc.
* resolve/rust-forever-stack.h
(class ForeverStackStore): Add.
* resolve/rust-forever-stack.cc: New file, based on
rust-forever-stack.hxx.
Signed-off-by: Owen Avery <powerboat9.gamer@gmail.com>
gcc/testsuite/ChangeLog:
* rust/compile/try-catch-unwind-old.rs: add a test to test the older
try intrinsics from plain old Rust to v1.78.0
* rust/compile/try-catch-unwind-new.rs: add a test to test the newer
catch_unwind instrinsics since Rust v1.78.0
