llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT--> @llvm/pr-subscribers-clang @llvm/pr-subscribers-backend-x86 Author: Simon Pilgrim (RKSimon) <details> <summary>Changes</summary> This should allow us to constexpr many avx512 predicated intrinsics where they wrap basic intrinsics that are already constexpr Fixes #<!-- -->152321 --- Patch is 45.67 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/152705.diff 14 Files Affected: - (modified) clang/include/clang/Basic/BuiltinsX86.td (+24-24) - (modified) clang/lib/AST/ExprConstant.cpp (+45) - (modified) clang/lib/Headers/avx512bitalgintrin.h (+4-4) - (modified) clang/lib/Headers/avx512fintrin.h (+4-4) - (modified) clang/lib/Headers/avx512vlbitalgintrin.h (+8-8) - (modified) clang/lib/Headers/avx512vlintrin.h (+18-9) - (modified) clang/lib/Headers/avx512vpopcntdqintrin.h (+8-8) - (modified) clang/lib/Headers/avx512vpopcntdqvlintrin.h (+15-13) - (modified) clang/test/CodeGen/X86/avx512bitalg-builtins.c (+6) - (modified) clang/test/CodeGen/X86/avx512f-builtins.c (+5) - (modified) clang/test/CodeGen/X86/avx512vl-builtins.c (+9) - (modified) clang/test/CodeGen/X86/avx512vlbitalg-builtins.c (+12) - (modified) clang/test/CodeGen/X86/avx512vpopcntdq-builtins.c (+4) - (modified) clang/test/CodeGen/X86/avx512vpopcntdqvl-builtins.c (+8) ``````````diff diff --git a/clang/include/clang/Basic/BuiltinsX86.td b/clang/include/clang/Basic/BuiltinsX86.td index 3efc0be8fa698..4f1dfddb53402 100644 --- a/clang/include/clang/Basic/BuiltinsX86.td +++ b/clang/include/clang/Basic/BuiltinsX86.td @@ -4241,99 +4241,99 @@ let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVecto def vfcmulcph512_mask : X86Builtin<"_Vector<16, float>(_Vector<16, float>, _Vector<16, float>, _Vector<16, float>, unsigned short, _Constant int)">; } -let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { +let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def selectb_128 : X86Builtin<"_Vector<16, char>(unsigned short, _Vector<16, char>, _Vector<16, char>)">; } -let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def selectb_256 : X86Builtin<"_Vector<32, char>(unsigned int, _Vector<32, char>, _Vector<32, char>)">; } -let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { +let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def selectb_512 : X86Builtin<"_Vector<64, char>(unsigned long long int, _Vector<64, char>, _Vector<64, char>)">; } -let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { +let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def selectw_128 : X86Builtin<"_Vector<8, short>(unsigned char, _Vector<8, short>, _Vector<8, short>)">; } -let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx512bw,avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def selectw_256 : X86Builtin<"_Vector<16, short>(unsigned short, _Vector<16, short>, _Vector<16, short>)">; } -let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { +let Features = "avx512bw,evex512", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def selectw_512 : X86Builtin<"_Vector<32, short>(unsigned int, _Vector<32, short>, _Vector<32, short>)">; } -let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { +let Features = "avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def selectd_128 : X86Builtin<"_Vector<4, int>(unsigned char, _Vector<4, int>, _Vector<4, int>)">; } -let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def selectd_256 : X86Builtin<"_Vector<8, int>(unsigned char, _Vector<8, int>, _Vector<8, int>)">; } -let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { +let Features = "avx512f,evex512", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def selectd_512 : X86Builtin<"_Vector<16, int>(unsigned short, _Vector<16, int>, _Vector<16, int>)">; } -let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { +let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def selectph_128 : X86Builtin<"_Vector<8, _Float16>(unsigned char, _Vector<8, _Float16>, _Vector<8, _Float16>)">; } -let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx512fp16,avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def selectph_256 : X86Builtin<"_Vector<16, _Float16>(unsigned short, _Vector<16, _Float16>, _Vector<16, _Float16>)">; } -let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { +let Features = "avx512fp16,evex512", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def selectph_512 : X86Builtin<"_Vector<32, _Float16>(unsigned int, _Vector<32, _Float16>, _Vector<32, _Float16>)">; } -let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { +let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def selectpbf_128 : X86Builtin<"_Vector<8, __bf16>(unsigned char, _Vector<8, __bf16>, _Vector<8, __bf16>)">; } -let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx512bf16,avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def selectpbf_256 : X86Builtin<"_Vector<16, __bf16>(unsigned short, _Vector<16, __bf16>, _Vector<16, __bf16>)">; } -let Features = "avx512bf16,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { +let Features = "avx512bf16,evex512", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def selectpbf_512 : X86Builtin<"_Vector<32, __bf16>(unsigned int, _Vector<32, __bf16>, _Vector<32, __bf16>)">; } -let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { +let Features = "avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def selectq_128 : X86Builtin<"_Vector<2, long long int>(unsigned char, _Vector<2, long long int>, _Vector<2, long long int>)">; } -let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def selectq_256 : X86Builtin<"_Vector<4, long long int>(unsigned char, _Vector<4, long long int>, _Vector<4, long long int>)">; } -let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { +let Features = "avx512f,evex512", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def selectq_512 : X86Builtin<"_Vector<8, long long int>(unsigned char, _Vector<8, long long int>, _Vector<8, long long int>)">; } -let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { +let Features = "avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def selectps_128 : X86Builtin<"_Vector<4, float>(unsigned char, _Vector<4, float>, _Vector<4, float>)">; } -let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def selectps_256 : X86Builtin<"_Vector<8, float>(unsigned char, _Vector<8, float>, _Vector<8, float>)">; } -let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { +let Features = "avx512f,evex512", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def selectps_512 : X86Builtin<"_Vector<16, float>(unsigned short, _Vector<16, float>, _Vector<16, float>)">; } -let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<128>] in { +let Features = "avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<128>] in { def selectpd_128 : X86Builtin<"_Vector<2, double>(unsigned char, _Vector<2, double>, _Vector<2, double>)">; } -let Features = "avx512vl", Attributes = [NoThrow, Const, RequiredVectorWidth<256>] in { +let Features = "avx512vl", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<256>] in { def selectpd_256 : X86Builtin<"_Vector<4, double>(unsigned char, _Vector<4, double>, _Vector<4, double>)">; } -let Features = "avx512f,evex512", Attributes = [NoThrow, Const, RequiredVectorWidth<512>] in { +let Features = "avx512f,evex512", Attributes = [NoThrow, Const, Constexpr, RequiredVectorWidth<512>] in { def selectpd_512 : X86Builtin<"_Vector<8, double>(unsigned char, _Vector<8, double>, _Vector<8, double>)">; } diff --git a/clang/lib/AST/ExprConstant.cpp b/clang/lib/AST/ExprConstant.cpp index 5cf2b9a0355ab..6595d8c69f0b2 100644 --- a/clang/lib/AST/ExprConstant.cpp +++ b/clang/lib/AST/ExprConstant.cpp @@ -11688,6 +11688,51 @@ bool VectorExprEvaluator::VisitCallExpr(const CallExpr *E) { return Success(APValue(ResultElements.data(), ResultElements.size()), E); } + case X86::BI__builtin_ia32_selectb_128: + case X86::BI__builtin_ia32_selectb_256: + case X86::BI__builtin_ia32_selectb_512: + case X86::BI__builtin_ia32_selectw_128: + case X86::BI__builtin_ia32_selectw_256: + case X86::BI__builtin_ia32_selectw_512: + case X86::BI__builtin_ia32_selectd_128: + case X86::BI__builtin_ia32_selectd_256: + case X86::BI__builtin_ia32_selectd_512: + case X86::BI__builtin_ia32_selectq_128: + case X86::BI__builtin_ia32_selectq_256: + case X86::BI__builtin_ia32_selectq_512: + case X86::BI__builtin_ia32_selectph_128: + case X86::BI__builtin_ia32_selectph_256: + case X86::BI__builtin_ia32_selectph_512: + case X86::BI__builtin_ia32_selectpbf_128: + case X86::BI__builtin_ia32_selectpbf_256: + case X86::BI__builtin_ia32_selectpbf_512: + case X86::BI__builtin_ia32_selectps_128: + case X86::BI__builtin_ia32_selectps_256: + case X86::BI__builtin_ia32_selectps_512: + case X86::BI__builtin_ia32_selectpd_128: + case X86::BI__builtin_ia32_selectpd_256: + case X86::BI__builtin_ia32_selectpd_512: { + // AVX512 predicated move: "Result = Mask[] ? LHS[] : RHS[]". + APValue SourceMask, SourceLHS, SourceRHS; + if (!EvaluateAsRValue(Info, E->getArg(0), SourceMask) || + !EvaluateAsRValue(Info, E->getArg(1), SourceLHS) || + !EvaluateAsRValue(Info, E->getArg(2), SourceRHS)) + return false; + + APSInt Mask = SourceMask.getInt(); + QualType DestEltTy = E->getType()->castAs<VectorType>()->getElementType(); + unsigned SourceLen = SourceLHS.getVectorLength(); + SmallVector<APValue, 4> ResultElements; + ResultElements.reserve(SourceLen); + + for (unsigned EltNum = 0; EltNum < SourceLen; ++EltNum) { + const APValue &LHS = SourceLHS.getVectorElt(EltNum); + const APValue &RHS = SourceRHS.getVectorElt(EltNum); + ResultElements.push_back(Mask[EltNum] ? LHS : RHS); + } + + return Success(APValue(ResultElements.data(), ResultElements.size()), E); + } } } diff --git a/clang/lib/Headers/avx512bitalgintrin.h b/clang/lib/Headers/avx512bitalgintrin.h index 9a1ff8f39734f..76c1a158b223f 100644 --- a/clang/lib/Headers/avx512bitalgintrin.h +++ b/clang/lib/Headers/avx512bitalgintrin.h @@ -32,7 +32,7 @@ _mm512_popcnt_epi16(__m512i __A) return (__m512i)__builtin_elementwise_popcount((__v32hu)__A); } -static __inline__ __m512i __DEFAULT_FN_ATTRS +static __inline__ __m512i __DEFAULT_FN_ATTRS_CONSTEXPR _mm512_mask_popcnt_epi16(__m512i __A, __mmask32 __U, __m512i __B) { return (__m512i) __builtin_ia32_selectw_512((__mmask32) __U, @@ -40,7 +40,7 @@ _mm512_mask_popcnt_epi16(__m512i __A, __mmask32 __U, __m512i __B) (__v32hi) __A); } -static __inline__ __m512i __DEFAULT_FN_ATTRS +static __inline__ __m512i __DEFAULT_FN_ATTRS_CONSTEXPR _mm512_maskz_popcnt_epi16(__mmask32 __U, __m512i __B) { return _mm512_mask_popcnt_epi16((__m512i) _mm512_setzero_si512(), @@ -54,7 +54,7 @@ _mm512_popcnt_epi8(__m512i __A) return (__m512i)__builtin_elementwise_popcount((__v64qu)__A); } -static __inline__ __m512i __DEFAULT_FN_ATTRS +static __inline__ __m512i __DEFAULT_FN_ATTRS_CONSTEXPR _mm512_mask_popcnt_epi8(__m512i __A, __mmask64 __U, __m512i __B) { return (__m512i) __builtin_ia32_selectb_512((__mmask64) __U, @@ -62,7 +62,7 @@ _mm512_mask_popcnt_epi8(__m512i __A, __mmask64 __U, __m512i __B) (__v64qi) __A); } -static __inline__ __m512i __DEFAULT_FN_ATTRS +static __inline__ __m512i __DEFAULT_FN_ATTRS_CONSTEXPR _mm512_maskz_popcnt_epi8(__mmask64 __U, __m512i __B) { return _mm512_mask_popcnt_epi8((__m512i) _mm512_setzero_si512(), diff --git a/clang/lib/Headers/avx512fintrin.h b/clang/lib/Headers/avx512fintrin.h index 95b80cc59bb02..00ef9bb0ace9e 100644 --- a/clang/lib/Headers/avx512fintrin.h +++ b/clang/lib/Headers/avx512fintrin.h @@ -8962,7 +8962,7 @@ _mm512_mask_cvtpslo_pd (__m512d __W, __mmask8 __U, __m512 __A) return (__m512d) _mm512_mask_cvtps_pd(__W, __U, _mm512_castps512_ps256(__A)); } -static __inline__ __m512d __DEFAULT_FN_ATTRS512 +static __inline__ __m512d __DEFAULT_FN_ATTRS512_CONSTEXPR _mm512_mask_mov_pd (__m512d __W, __mmask8 __U, __m512d __A) { return (__m512d) __builtin_ia32_selectpd_512 ((__mmask8) __U, @@ -8970,7 +8970,7 @@ _mm512_mask_mov_pd (__m512d __W, __mmask8 __U, __m512d __A) (__v8df) __W); } -static __inline__ __m512d __DEFAULT_FN_ATTRS512 +static __inline__ __m512d __DEFAULT_FN_ATTRS512_CONSTEXPR _mm512_maskz_mov_pd (__mmask8 __U, __m512d __A) { return (__m512d) __builtin_ia32_selectpd_512 ((__mmask8) __U, @@ -8978,7 +8978,7 @@ _mm512_maskz_mov_pd (__mmask8 __U, __m512d __A) (__v8df) _mm512_setzero_pd ()); } -static __inline__ __m512 __DEFAULT_FN_ATTRS512 +static __inline__ __m512 __DEFAULT_FN_ATTRS512_CONSTEXPR _mm512_mask_mov_ps (__m512 __W, __mmask16 __U, __m512 __A) { return (__m512) __builtin_ia32_selectps_512 ((__mmask16) __U, @@ -8986,7 +8986,7 @@ _mm512_mask_mov_ps (__m512 __W, __mmask16 __U, __m512 __A) (__v16sf) __W); } -static __inline__ __m512 __DEFAULT_FN_ATTRS512 +static __inline__ __m512 __DEFAULT_FN_ATTRS512_CONSTEXPR _mm512_maskz_mov_ps (__mmask16 __U, __m512 __A) { return (__m512) __builtin_ia32_selectps_512 ((__mmask16) __U, diff --git a/clang/lib/Headers/avx512vlbitalgintrin.h b/clang/lib/Headers/avx512vlbitalgintrin.h index 739e78aab753d..e29a149df5f90 100644 --- a/clang/lib/Headers/avx512vlbitalgintrin.h +++ b/clang/lib/Headers/avx512vlbitalgintrin.h @@ -38,7 +38,7 @@ _mm256_popcnt_epi16(__m256i __A) return (__m256i)__builtin_elementwise_popcount((__v16hu)__A); } -static __inline__ __m256i __DEFAULT_FN_ATTRS256 +static __inline__ __m256i __DEFAULT_FN_ATTRS256_CONSTEXPR _mm256_mask_popcnt_epi16(__m256i __A, __mmask16 __U, __m256i __B) { return (__m256i) __builtin_ia32_selectw_256((__mmask16) __U, @@ -46,7 +46,7 @@ _mm256_mask_popcnt_epi16(__m256i __A, __mmask16 __U, __m256i __B) (__v16hi) __A); } -static __inline__ __m256i __DEFAULT_FN_ATTRS256 +static __inline__ __m256i __DEFAULT_FN_ATTRS256_CONSTEXPR _mm256_maskz_popcnt_epi16(__mmask16 __U, __m256i __B) { return _mm256_mask_popcnt_epi16((__m256i) _mm256_setzero_si256(), @@ -60,7 +60,7 @@ _mm_popcnt_epi16(__m128i __A) return (__m128i)__builtin_elementwise_popcount((__v8hu)__A); } -static __inline__ __m128i __DEFAULT_FN_ATTRS128 +static __inline__ __m128i __DEFAULT_FN_ATTRS128_CONSTEXPR _mm_mask_popcnt_epi16(__m128i __A, __mmask8 __U, __m128i __B) { return (__m128i) __builtin_ia32_selectw_128((__mmask8) __U, @@ -68,7 +68,7 @@ _mm_mask_popcnt_epi16(__m128i __A, __mmask8 __U, __m128i __B) (__v8hi) __A); } -static __inline__ __m128i __DEFAULT_FN_ATTRS128 +static __inline__ __m128i __DEFAULT_FN_ATTRS128_CONSTEXPR _mm_maskz_popcnt_epi16(__mmask8 __U, __m128i __B) { return _mm_mask_popcnt_epi16((__m128i) _mm_setzero_si128(), @@ -82,7 +82,7 @@ _mm256_popcnt_epi8(__m256i __A) return (__m256i)__builtin_elementwise_popcount((__v32qu)__A); } -static __inline__ __m256i __DEFAULT_FN_ATTRS256 +static __inline__ __m256i __DEFAULT_FN_ATTRS256_CONSTEXPR _mm256_mask_popcnt_epi8(__m256i __A, __mmask32 __U, __m256i __B) { return (__m256i) __builtin_ia32_selectb_256((__mmask32) __U, @@ -90,7 +90,7 @@ _mm256_mask_popcnt_epi8(__m256i __A, __mmask32 __U, __m256i __B) (__v32qi) __A); } -static __inline__ __m256i __DEFAULT_FN_ATTRS256 +static __inline__ __m256i __DEFAULT_FN_ATTRS256_CONSTEXPR _mm256_maskz_popcnt_epi8(__mmask32 __U, __m256i __B) { return _mm256_mask_popcnt_epi8((__m256i) _mm256_setzero_si256(), @@ -104,7 +104,7 @@ _mm_popcnt_epi8(__m128i __A) return (__m128i)__builtin_elementwise_popcount((__v16qu)__A); } -static __inline__ __m128i __DEFAULT_FN_ATTRS128 +static __inline__ __m128i __DEFAULT_FN_ATTRS128_CONSTEXPR _mm_mask_popcnt_epi8(__m128i __A, __mmask16 __U, __m128i __B) { return (__m128i) __builtin_ia32_selectb_128((__mmask16) __U, @@ -112,7 +112,7 @@ _mm_mask_popcnt_epi8(__m128i __A, __mmask16 __U, __m128i __B) (__v16qi) __A); } -static __inline__ __m128i __DEFAULT_FN_ATTRS128 +static __inline__ __m128i __DEFAULT_FN_ATTRS128_CONSTEXPR _mm_maskz_popcnt_epi8(__mmask16 __U, __m128i __B) { return _mm_mask_popcnt_epi8((__m128i) _mm_setzero_si128(), diff --git a/clang/lib/Headers/avx512vlintrin.h b/clang/lib/Headers/avx512vlintrin.h index cbad39acad84f..fd1bd291ecbcb 100644 --- a/clang/lib/Headers/avx512vlintrin.h +++ b/clang/lib/Headers/avx512vlintrin.h @@ -23,6 +23,14 @@ __target__("avx512vl,no-evex512"), \ __min_vector_width__(256))) +#if defined(__cplusplus) && (__cplusplus >= 201103L) +#define __DEFAULT_FN_ATTRS128_CONSTEXPR __DEFAULT_FN_ATTRS128 constexpr +#define __DEFAULT_FN_ATTRS256_CONSTEXPR __DEFAULT_FN_ATTRS256 constexpr +#else +#define __DEFAULT_FN_ATTRS128_CONSTEXPR __DEFAULT_FN_ATTRS128 +#define __DEFAULT_FN_ATTRS256_CONSTEXPR __DEFAULT_FN_ATTRS256 +#endif + typedef short __v2hi __attribute__((__vector_size__(4))); typedef char __v4qi __attribute__((__vector_size__(4))); typedef char __v2qi __attribute__((__vector_size__(2))); @@ -8258,7 +8266,7 @@ _mm256_maskz_moveldup_ps (__mmask8 __U, __m256 __A) (__v4si)_mm_shuffle_epi32((A), (I)), \ (__v4si)_mm_setzero_si128())) -static __inline__ __m128d __DEFAULT_FN_ATTRS128 +static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR _mm_mask_mov_pd (__m128d __W, __mmask8 __U, __m128d __A) { return (__m128d) __builtin_ia32_selectpd_128 ((__mmask8) __U, @@ -8266,7 +8274,7 @@ _mm_mask_mov_pd (__m128d __W, __mmask8 __U, __m128d __A) (__v2df) __W); } -static __inline__ __m128d __DEFAULT_FN_ATTRS128 +static __inline__ __m128d __DEFAULT_FN_ATTRS128_CONSTEXPR _mm_maskz_mov_pd (__mmask8 __U, __m128d __A) { return (__m128d) __builtin_ia32_selectpd_128 ((__mmask8) __U, @@ -8274,7 +8282,7 @@ _mm_maskz_mov_pd (__mmask8 __U, __m128d __A) (__v2df) _mm_setzero_pd ()); } -static __inline__ __m256d __DEFAULT_FN_ATTRS256 +static __inline__ __m256d __DEFAULT_FN_ATTRS256_CONSTEXPR _mm256_mask_mov_pd (__m256d __W, __mmask8 __U, __m256d __A) { return (__m256d) __builtin_ia32_selectpd_256 ((__mmask8) __U, @@ -8282,7 +8290,7 @@ _mm256_mask_mov_pd (__m256d __W, __mmask8 __U, __m256d __A) (__v4df) __W); } -static __inline__ __m256d __DEFAULT_FN_ATTRS256 +static __inline__ __m256d __DEFAULT_FN_ATTRS256_CONSTEXPR _mm256_maskz_mov_pd (__mmask8 __U, __m256d __A) { return (__m256d) __builtin_ia32_selectpd_256 ((__mmask8) __U, @@ -8290,7 +8298,7 @@ _mm256_maskz_mov_pd (__mmask8 __U, __m256d __A) (__v4df) _mm256_setzero_pd ()); } -static __inline__ __m128 __DEFAULT_FN_ATTRS128 +static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR _mm_mask_mov_ps (__m128 __W, __mmask8 __U, __m128 __A) { return (__m128) __builtin_ia32_selectps_128 ((__mmask8) __U, @@ -8298,7 +8306,7 @@ _mm_mask_mov_ps (__m128 __W, __mmask8 __U, __m128 __A) (__v4sf) __W); } -static __inline__ __m128 __DEFAULT_FN_ATTRS128 +static __inline__ __m128 __DEFAULT_FN_ATTRS128_CONSTEXPR _mm_maskz_mov_ps (__mmask8 __U, __m128 __A) { return (__m128) __builtin_ia32_selectps_128 ((__mmask8) __U, @@ -8306,7 +8314,7 @@ _mm_maskz_mov_ps (__mmask8 __U, __m128 __A) (__v4sf) _mm_setzero_ps ()); } -static __inline__ __m256 __DEFAULT_FN_ATTRS256 +static __inline__ __m256 __DEFAULT_FN_ATTRS256_CONSTEXPR _mm256_mask_mov_ps (__m256 __W, __mmask8 __U, __m256 __A) { return (__m256) __builtin_ia32_selectps_256 ((__mmask8) __U, @@ -8314,7 +8322,7 @@ _mm256_mask_mov_ps (__m256 __W, __mmask8 __U, __m256 __A) (__v8sf) __W); } -static __inline__ __m256 __DEFAULT_FN_ATTRS256 +static __inline__ __m256 __DEFAULT_FN_ATTRS256_CONSTEXPR _mm256_maskz_mov_ps (__mmask8 __... [truncated] `````````` </details> https://github.com/llvm/llvm-project/pull/152705 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
