llvm-project.git/clang, branch main Unnamed repository; edit this file 'description' to name the repository. Don't mark lambda non-dependent if nested in a generic lambda. (#149121) 2025-11-23T03:11:58+00:00 Daniel M. Katz katzdm@gmail.com 2025-11-23T03:11:58+00:00 525e68e9e9a44e88eb88ef2d6f058a482972c989 Fixes #118187 Fixes #156579 An instantiated `LambdaExpr` can currently be marked as `LDK_NeverDependent` if it's nested within a generic lambda. If that `LambdaExpr` in fact depends on template parameters introduced by the enclosing generic lambda, then its dependence will be misreported as "never dependent" and spurious diagnostics can result. The fix here proposed is a bit ugly, but the condition that it's being bolted onto already seems like a bit of a hack, so this seems no worse for wear. Note that #89702 surfaced this change because it caused the inner lambda expression to (correctly) be considered in a constant-evaluated context. The affected check for whether to mark the inner lambda as `LDK_NeverDependent` therefore started to apply, whereas it didn't before. **Tested**: `check-clang` and `check-cxx`. 
Fixes #118187
Fixes #156579

An instantiated `LambdaExpr` can currently be marked as
`LDK_NeverDependent` if it's nested within a generic lambda. If that
`LambdaExpr` in fact depends on template parameters introduced by the
enclosing generic lambda, then its dependence will be misreported as
"never dependent" and spurious diagnostics can result.

The fix here proposed is a bit ugly, but the condition that it's being
bolted onto already seems like a bit of a hack, so this seems no worse
for wear.

Note that #89702 surfaced this change because it caused the inner lambda
expression to (correctly) be considered in a constant-evaluated context.
The affected check for whether to mark the inner lambda as
`LDK_NeverDependent` therefore started to apply, whereas it didn't
before.

**Tested**: `check-clang` and `check-cxx`.
 [StaticAnalyzer] Use a range-based for loop (NFC) (#169175) 2025-11-22T23:30:59+00:00 Kazu Hirata kazu@google.com 2025-11-22T23:30:59+00:00 1c05dfb0e2b14d783e2e08d5c02abdb8cf73877f Identified with modernize-loop-convert. 
Identified with modernize-loop-convert.
 [Lex] Use a range-based for loop (NFC) (#169174) 2025-11-22T23:30:51+00:00 Kazu Hirata kazu@google.com 2025-11-22T23:30:51+00:00 c5939937a4d1b7da4477762389dee0afa756bd9c Identified with modernize-loop-convert. 
Identified with modernize-loop-convert.
 [clang] Use llvm::equal (NFC) (#169172) 2025-11-22T23:30:35+00:00 Kazu Hirata kazu@google.com 2025-11-22T23:30:35+00:00 637299e7e079b7d80f508691103092753d282ffa Identified with llvm-use-ranges. 
Identified with llvm-use-ranges.
 [X86] Move _mm512_mask_blend_pd/ps tests to avx512f-builtins.c (#169170) 2025-11-22T17:26:13+00:00 Simon Pilgrim llvm-dev@redking.me.uk 2025-11-22T17:26:13+00:00 b00c620b3504565d9769a434bc7d4e97854cd788 These aren't AVX512VL tests 
These aren't AVX512VL tests
 [Clang] VectorExprEvaluator::VisitCallExpr / InterpretBuiltin - Allow AVX512 VPSHUFBITQMB intrinsics to be used in constexpr (#168100) 2025-11-22T17:19:27+00:00 NagaChaitanya Vellanki pnagato@protonmail.com 2025-11-22T17:19:27+00:00 456ca91815c3fdb60b5ca695c8bb05b75016a343 Resolves #161337 
Resolves #161337
 [clang][bytecode][NFC] Remove VariableScope::emitDestruction (#169148) 2025-11-22T15:41:00+00:00 Timm Baeder tbaeder@redhat.com 2025-11-22T15:41:00+00:00 cc4dd015ad4a1b33d43fbac00d62f6b309a96ff4 destroyLocals() does the same thing. 
destroyLocals() does the same thing.
 [Clang][X86] VectorExprEvaluator::VisitCallExpr / InterpretBuiltin - allow VPERMILPD/S variable mask intrinsics to be used in constexpr (#168861) 2025-11-22T14:37:50+00:00 stomfaig 55883018+stomfaig@users.noreply.github.com 2025-11-22T14:37:50+00:00 2e424deeb6180d112323f4df955c8034eb56780c Allowing VPERMILPD/S intrinsics to be used in constexpr Closes #167878 
Allowing VPERMILPD/S intrinsics to be used in constexpr

Closes #167878
 [Flang] Add -ffast-real-mod back for further control of MOD optimizations (#167118) 2025-11-22T11:55:32+00:00 Michael Klemm michael.klemm@amd.com 2025-11-22T11:55:32+00:00 9daf4345ec836d50740805c878bd570dd4093354 It turns out that having `-ffast-math` as the only option to control optimizations for MOD for REAL kinds (PR #160660) is too coarse-grained for some applications. Thus, this PR adds back `-ffast-real-mod` to have more control over the optimization. The `-ffast-math` flag will still enable the optimization, and `-fno-fast-real-mod` allows one to disable it. 
It turns out that having `-ffast-math` as the only option to control
optimizations for MOD for REAL kinds (PR #160660) is too coarse-grained
for some applications. Thus, this PR adds back `-ffast-real-mod` to have
more control over the optimization. The `-ffast-math` flag will still
enable the optimization, and `-fno-fast-real-mod` allows one to disable
it.
 [clang][X86] Allow VALIGND/Q element-shift intrinsics in constexpr evaluation (#168206) 2025-11-22T11:53:34+00:00 Muhammad Abdul alilo.ghazali@gmail.com 2025-11-22T11:53:34+00:00 7305b6eb5458b5cea62b3ab70da95b790cf988f3 Fixes #167681 
Fixes #167681