Author: ctopper Date: Sun Jun 10 10:27:05 2018 New Revision: 334366 URL: http://llvm.org/viewvc/llvm-project?rev=334366&view=rev Log: [X86] Use target independent masked expandload and compressstore intrinsics to implement expandload/compressstore builtins.
Summary: We've had these target independent intrinsics for at least a year and a half. Looks like they do exactly what we need here and the backend already supports them. Reviewers: RKSimon, delena, spatel, GBuella Reviewed By: RKSimon Subscribers: cfe-commits, llvm-commits Differential Revision: https://reviews.llvm.org/D47693 Modified: cfe/trunk/lib/CodeGen/CGBuiltin.cpp cfe/trunk/test/CodeGen/avx512f-builtins.c cfe/trunk/test/CodeGen/avx512vbmi2-builtins.c cfe/trunk/test/CodeGen/avx512vl-builtins.c cfe/trunk/test/CodeGen/avx512vlvbmi2-builtins.c Modified: cfe/trunk/lib/CodeGen/CGBuiltin.cpp URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/lib/CodeGen/CGBuiltin.cpp?rev=334366&r1=334365&r2=334366&view=diff ============================================================================== --- cfe/trunk/lib/CodeGen/CGBuiltin.cpp (original) +++ cfe/trunk/lib/CodeGen/CGBuiltin.cpp Sun Jun 10 10:27:05 2018 @@ -8496,6 +8496,40 @@ static Value *EmitX86MaskedLoad(CodeGenF return CGF.Builder.CreateMaskedLoad(Ptr, Align, MaskVec, Ops[1]); } +static Value *EmitX86ExpandLoad(CodeGenFunction &CGF, + ArrayRef<Value *> Ops) { + llvm::Type *ResultTy = Ops[1]->getType(); + llvm::Type *PtrTy = ResultTy->getVectorElementType(); + + // Cast the pointer to element type. + Value *Ptr = CGF.Builder.CreateBitCast(Ops[0], + llvm::PointerType::getUnqual(PtrTy)); + + Value *MaskVec = getMaskVecValue(CGF, Ops[2], + ResultTy->getVectorNumElements()); + + llvm::Function *F = CGF.CGM.getIntrinsic(Intrinsic::masked_expandload, + ResultTy); + return CGF.Builder.CreateCall(F, { Ptr, MaskVec, Ops[1] }); +} + +static Value *EmitX86CompressStore(CodeGenFunction &CGF, + ArrayRef<Value *> Ops) { + llvm::Type *ResultTy = Ops[1]->getType(); + llvm::Type *PtrTy = ResultTy->getVectorElementType(); + + // Cast the pointer to element type. + Value *Ptr = CGF.Builder.CreateBitCast(Ops[0], + llvm::PointerType::getUnqual(PtrTy)); + + Value *MaskVec = getMaskVecValue(CGF, Ops[2], + ResultTy->getVectorNumElements()); + + llvm::Function *F = CGF.CGM.getIntrinsic(Intrinsic::masked_compressstore, + ResultTy); + return CGF.Builder.CreateCall(F, { Ops[1], Ptr, MaskVec }); +} + static Value *EmitX86MaskLogic(CodeGenFunction &CGF, Instruction::BinaryOps Opc, unsigned NumElts, ArrayRef<Value *> Ops, bool InvertLHS = false) { @@ -9219,6 +9253,46 @@ Value *CodeGenFunction::EmitX86BuiltinEx return EmitX86MaskedLoad(*this, Ops, Align); } + case X86::BI__builtin_ia32_expandloaddf128_mask: + case X86::BI__builtin_ia32_expandloaddf256_mask: + case X86::BI__builtin_ia32_expandloaddf512_mask: + case X86::BI__builtin_ia32_expandloadsf128_mask: + case X86::BI__builtin_ia32_expandloadsf256_mask: + case X86::BI__builtin_ia32_expandloadsf512_mask: + case X86::BI__builtin_ia32_expandloaddi128_mask: + case X86::BI__builtin_ia32_expandloaddi256_mask: + case X86::BI__builtin_ia32_expandloaddi512_mask: + case X86::BI__builtin_ia32_expandloadsi128_mask: + case X86::BI__builtin_ia32_expandloadsi256_mask: + case X86::BI__builtin_ia32_expandloadsi512_mask: + case X86::BI__builtin_ia32_expandloadhi128_mask: + case X86::BI__builtin_ia32_expandloadhi256_mask: + case X86::BI__builtin_ia32_expandloadhi512_mask: + case X86::BI__builtin_ia32_expandloadqi128_mask: + case X86::BI__builtin_ia32_expandloadqi256_mask: + case X86::BI__builtin_ia32_expandloadqi512_mask: + return EmitX86ExpandLoad(*this, Ops); + + case X86::BI__builtin_ia32_compressstoredf128_mask: + case X86::BI__builtin_ia32_compressstoredf256_mask: + case X86::BI__builtin_ia32_compressstoredf512_mask: + case X86::BI__builtin_ia32_compressstoresf128_mask: + case X86::BI__builtin_ia32_compressstoresf256_mask: + case X86::BI__builtin_ia32_compressstoresf512_mask: + case X86::BI__builtin_ia32_compressstoredi128_mask: + case X86::BI__builtin_ia32_compressstoredi256_mask: + case X86::BI__builtin_ia32_compressstoredi512_mask: + case X86::BI__builtin_ia32_compressstoresi128_mask: + case X86::BI__builtin_ia32_compressstoresi256_mask: + case X86::BI__builtin_ia32_compressstoresi512_mask: + case X86::BI__builtin_ia32_compressstorehi128_mask: + case X86::BI__builtin_ia32_compressstorehi256_mask: + case X86::BI__builtin_ia32_compressstorehi512_mask: + case X86::BI__builtin_ia32_compressstoreqi128_mask: + case X86::BI__builtin_ia32_compressstoreqi256_mask: + case X86::BI__builtin_ia32_compressstoreqi512_mask: + return EmitX86CompressStore(*this, Ops); + case X86::BI__builtin_ia32_storehps: case X86::BI__builtin_ia32_storelps: { llvm::Type *PtrTy = llvm::PointerType::getUnqual(Int64Ty); Modified: cfe/trunk/test/CodeGen/avx512f-builtins.c URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGen/avx512f-builtins.c?rev=334366&r1=334365&r2=334366&view=diff ============================================================================== --- cfe/trunk/test/CodeGen/avx512f-builtins.c (original) +++ cfe/trunk/test/CodeGen/avx512f-builtins.c Sun Jun 10 10:27:05 2018 @@ -7293,40 +7293,52 @@ __m512i test_mm512_maskz_expand_epi64(__ } __m512i test_mm512_mask_expandloadu_epi64(__m512i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm512_mask_expandloadu_epi64 - // CHECK: @llvm.x86.avx512.mask.expand.load.q.512 + // CHECK: @llvm.masked.expandload.v8i64(i64* %{{.*}}, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) return _mm512_mask_expandloadu_epi64(__W, __U, __P); } __m512i test_mm512_maskz_expandloadu_epi64(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm512_maskz_expandloadu_epi64 - // CHECK: @llvm.x86.avx512.mask.expand.load.q.512 + // CHECK: @llvm.masked.expandload.v8i64(i64* %{{.*}}, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) return _mm512_maskz_expandloadu_epi64(__U, __P); } __m512d test_mm512_mask_expandloadu_pd(__m512d __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm512_mask_expandloadu_pd - // CHECK: @llvm.x86.avx512.mask.expand.load.pd.512 + // CHECK: @llvm.masked.expandload.v8f64(double* %{{.*}}, <8 x i1> %{{.*}}, <8 x double> %{{.*}}) return _mm512_mask_expandloadu_pd(__W, __U, __P); } __m512d test_mm512_maskz_expandloadu_pd(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm512_maskz_expandloadu_pd - // CHECK: @llvm.x86.avx512.mask.expand.load.pd.512 + // CHECK: @llvm.masked.expandload.v8f64(double* %{{.*}}, <8 x i1> %{{.*}}, <8 x double> %{{.*}}) return _mm512_maskz_expandloadu_pd(__U, __P); } __m512i test_mm512_mask_expandloadu_epi32(__m512i __W, __mmask16 __U, void const *__P) { // CHECK-LABEL: @test_mm512_mask_expandloadu_epi32 - // CHECK: @llvm.x86.avx512.mask.expand.load.d.512 + // CHECK: @llvm.masked.expandload.v16i32(i32* %{{.*}}, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) return _mm512_mask_expandloadu_epi32(__W, __U, __P); } __m512i test_mm512_maskz_expandloadu_epi32(__mmask16 __U, void const *__P) { // CHECK-LABEL: @test_mm512_maskz_expandloadu_epi32 - // CHECK: @llvm.x86.avx512.mask.expand.load.d.512 + // CHECK: @llvm.masked.expandload.v16i32(i32* %{{.*}}, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) return _mm512_maskz_expandloadu_epi32(__U, __P); } +__m512 test_mm512_mask_expandloadu_ps(__m512 __W, __mmask16 __U, void const *__P) { + // CHECK-LABEL: @test_mm512_mask_expandloadu_ps + // CHECK: @llvm.masked.expandload.v16f32(float* %{{.*}}, <16 x i1> %{{.*}}, <16 x float> %{{.*}}) + return _mm512_mask_expandloadu_ps(__W, __U, __P); +} + +__m512 test_mm512_maskz_expandloadu_ps(__mmask16 __U, void const *__P) { + // CHECK-LABEL: @test_mm512_maskz_expandloadu_ps + // CHECK: @llvm.masked.expandload.v16f32(float* %{{.*}}, <16 x i1> %{{.*}}, <16 x float> %{{.*}}) + return _mm512_maskz_expandloadu_ps(__U, __P); +} + __m512 test_mm512_mask_expand_ps(__m512 __W, __mmask16 __U, __m512 __A) { // CHECK-LABEL: @test_mm512_mask_expand_ps // CHECK: @llvm.x86.avx512.mask.expand.ps.512 @@ -7428,25 +7440,25 @@ __m512 test_mm512_maskz_mov_ps(__mmask16 void test_mm512_mask_compressstoreu_pd(void *__P, __mmask8 __U, __m512d __A) { // CHECK-LABEL: @test_mm512_mask_compressstoreu_pd - // CHECK: @llvm.x86.avx512.mask.compress.store.pd.512 + // CHECK: @llvm.masked.compressstore.v8f64(<8 x double> %{{.*}}, double* %{{.*}}, <8 x i1> %{{.*}}) return _mm512_mask_compressstoreu_pd(__P, __U, __A); } void test_mm512_mask_compressstoreu_epi64(void *__P, __mmask8 __U, __m512i __A) { // CHECK-LABEL: @test_mm512_mask_compressstoreu_epi64 - // CHECK: @llvm.x86.avx512.mask.compress.store.q.512 + // CHECK: @llvm.masked.compressstore.v8i64(<8 x i64> %{{.*}}, i64* %{{.*}}, <8 x i1> %{{.*}}) return _mm512_mask_compressstoreu_epi64(__P, __U, __A); } void test_mm512_mask_compressstoreu_ps(void *__P, __mmask16 __U, __m512 __A) { // CHECK-LABEL: @test_mm512_mask_compressstoreu_ps - // CHECK: @llvm.x86.avx512.mask.compress.store.ps.512 + // CHECK: @llvm.masked.compressstore.v16f32(<16 x float> %{{.*}}, float* %{{.*}}, <16 x i1> %{{.*}}) return _mm512_mask_compressstoreu_ps(__P, __U, __A); } void test_mm512_mask_compressstoreu_epi32(void *__P, __mmask16 __U, __m512i __A) { // CHECK-LABEL: @test_mm512_mask_compressstoreu_epi32 - // CHECK: @llvm.x86.avx512.mask.compress.store.d.512 + // CHECK: @llvm.masked.compressstore.v16i32(<16 x i32> %{{.*}}, i32* %{{.*}}, <16 x i1> %{{.*}}) return _mm512_mask_compressstoreu_epi32(__P, __U, __A); } Modified: cfe/trunk/test/CodeGen/avx512vbmi2-builtins.c URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGen/avx512vbmi2-builtins.c?rev=334366&r1=334365&r2=334366&view=diff ============================================================================== --- cfe/trunk/test/CodeGen/avx512vbmi2-builtins.c (original) +++ cfe/trunk/test/CodeGen/avx512vbmi2-builtins.c Sun Jun 10 10:27:05 2018 @@ -28,13 +28,13 @@ __m512i test_mm512_maskz_compress_epi8(_ void test_mm512_mask_compressstoreu_epi16(void *__P, __mmask32 __U, __m512i __D) { // CHECK-LABEL: @test_mm512_mask_compressstoreu_epi16 - // CHECK: @llvm.x86.avx512.mask.compress.store.w.512 + // CHECK: @llvm.masked.compressstore.v32i16(<32 x i16> %{{.*}}, i16* %{{.*}}, <32 x i1> %{{.*}}) _mm512_mask_compressstoreu_epi16(__P, __U, __D); } void test_mm512_mask_compressstoreu_epi8(void *__P, __mmask64 __U, __m512i __D) { // CHECK-LABEL: @test_mm512_mask_compressstoreu_epi8 - // CHECK: @llvm.x86.avx512.mask.compress.store.b.512 + // CHECK: @llvm.masked.compressstore.v64i8(<64 x i8> %{{.*}}, i8* %{{.*}}, <64 x i1> %{{.*}}) _mm512_mask_compressstoreu_epi8(__P, __U, __D); } @@ -64,25 +64,25 @@ __m512i test_mm512_maskz_expand_epi8(__m __m512i test_mm512_mask_expandloadu_epi16(__m512i __S, __mmask32 __U, void const* __P) { // CHECK-LABEL: @test_mm512_mask_expandloadu_epi16 - // CHECK: @llvm.x86.avx512.mask.expand.load.w.512 + // CHECK: @llvm.masked.expandload.v32i16(i16* %{{.*}}, <32 x i1> %{{.*}}, <32 x i16> %{{.*}}) return _mm512_mask_expandloadu_epi16(__S, __U, __P); } __m512i test_mm512_maskz_expandloadu_epi16(__mmask32 __U, void const* __P) { // CHECK-LABEL: @test_mm512_maskz_expandloadu_epi16 - // CHECK: @llvm.x86.avx512.mask.expand.load.w.512 + // CHECK: @llvm.masked.expandload.v32i16(i16* %{{.*}}, <32 x i1> %{{.*}}, <32 x i16> %{{.*}}) return _mm512_maskz_expandloadu_epi16(__U, __P); } __m512i test_mm512_mask_expandloadu_epi8(__m512i __S, __mmask64 __U, void const* __P) { // CHECK-LABEL: @test_mm512_mask_expandloadu_epi8 - // CHECK: @llvm.x86.avx512.mask.expand.load.b.512 + // CHECK: @llvm.masked.expandload.v64i8(i8* %{{.*}}, <64 x i1> %{{.*}}, <64 x i8> %{{.*}}) return _mm512_mask_expandloadu_epi8(__S, __U, __P); } __m512i test_mm512_maskz_expandloadu_epi8(__mmask64 __U, void const* __P) { // CHECK-LABEL: @test_mm512_maskz_expandloadu_epi8 - // CHECK: @llvm.x86.avx512.mask.expand.load.b.512 + // CHECK: @llvm.masked.expandload.v64i8(i8* %{{.*}}, <64 x i1> %{{.*}}, <64 x i8> %{{.*}}) return _mm512_maskz_expandloadu_epi8(__U, __P); } Modified: cfe/trunk/test/CodeGen/avx512vl-builtins.c URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGen/avx512vl-builtins.c?rev=334366&r1=334365&r2=334366&view=diff ============================================================================== --- cfe/trunk/test/CodeGen/avx512vl-builtins.c (original) +++ cfe/trunk/test/CodeGen/avx512vl-builtins.c Sun Jun 10 10:27:05 2018 @@ -2043,42 +2043,42 @@ __m256i test_mm256_maskz_compress_epi32( } void test_mm_mask_compressstoreu_pd(void *__P, __mmask8 __U, __m128d __A) { // CHECK-LABEL: @test_mm_mask_compressstoreu_pd - // CHECK: @llvm.x86.avx512.mask.compress.store.pd.128 + // CHECK: @llvm.masked.compressstore.v2f64(<2 x double> %{{.*}}, double* %{{.*}}, <2 x i1> %{{.*}}) return _mm_mask_compressstoreu_pd(__P,__U,__A); } void test_mm256_mask_compressstoreu_pd(void *__P, __mmask8 __U, __m256d __A) { // CHECK-LABEL: @test_mm256_mask_compressstoreu_pd - // CHECK: @llvm.x86.avx512.mask.compress.store.pd.256 + // CHECK: @llvm.masked.compressstore.v4f64(<4 x double> %{{.*}}, double* %{{.*}}, <4 x i1> %{{.*}}) return _mm256_mask_compressstoreu_pd(__P,__U,__A); } void test_mm_mask_compressstoreu_epi64(void *__P, __mmask8 __U, __m128i __A) { // CHECK-LABEL: @test_mm_mask_compressstoreu_epi64 - // CHECK: @llvm.x86.avx512.mask.compress.store.q.128 + // CHECK: @llvm.masked.compressstore.v2i64(<2 x i64> %{{.*}}, i64* %{{.*}}, <2 x i1> %{{.*}}) return _mm_mask_compressstoreu_epi64(__P,__U,__A); } void test_mm256_mask_compressstoreu_epi64(void *__P, __mmask8 __U, __m256i __A) { // CHECK-LABEL: @test_mm256_mask_compressstoreu_epi64 - // CHECK: @llvm.x86.avx512.mask.compress.store.q.256 + // CHECK: @llvm.masked.compressstore.v4i64(<4 x i64> %{{.*}}, i64* %{{.*}}, <4 x i1> %{{.*}}) return _mm256_mask_compressstoreu_epi64(__P,__U,__A); } void test_mm_mask_compressstoreu_ps(void *__P, __mmask8 __U, __m128 __A) { // CHECK-LABEL: @test_mm_mask_compressstoreu_ps - // CHECK: @llvm.x86.avx512.mask.compress.store.ps.128 + // CHECK: @llvm.masked.compressstore.v4f32(<4 x float> %{{.*}}, float* %{{.*}}, <4 x i1> %{{.*}}) return _mm_mask_compressstoreu_ps(__P,__U,__A); } void test_mm256_mask_compressstoreu_ps(void *__P, __mmask8 __U, __m256 __A) { // CHECK-LABEL: @test_mm256_mask_compressstoreu_ps - // CHECK: @llvm.x86.avx512.mask.compress.store.ps.256 + // CHECK: @llvm.masked.compressstore.v8f32(<8 x float> %{{.*}}, float* %{{.*}}, <8 x i1> %{{.*}}) return _mm256_mask_compressstoreu_ps(__P,__U,__A); } void test_mm_mask_compressstoreu_epi32(void *__P, __mmask8 __U, __m128i __A) { // CHECK-LABEL: @test_mm_mask_compressstoreu_epi32 - // CHECK: @llvm.x86.avx512.mask.compress.store.d.128 + // CHECK: @llvm.masked.compressstore.v4i32(<4 x i32> %{{.*}}, i32* %{{.*}}, <4 x i1> %{{.*}}) return _mm_mask_compressstoreu_epi32(__P,__U,__A); } void test_mm256_mask_compressstoreu_epi32(void *__P, __mmask8 __U, __m256i __A) { // CHECK-LABEL: @test_mm256_mask_compressstoreu_epi32 - // CHECK: @llvm.x86.avx512.mask.compress.store.d.256 + // CHECK: @llvm.masked.compressstore.v8i32(<8 x i32> %{{.*}}, i32* %{{.*}}, <8 x i1> %{{.*}}) return _mm256_mask_compressstoreu_epi32(__P,__U,__A); } __m128d test_mm_mask_cvtepi32_pd(__m128d __W, __mmask8 __U, __m128i __A) { @@ -2550,82 +2550,82 @@ __m256i test_mm256_maskz_expand_epi64(__ } __m128d test_mm_mask_expandloadu_pd(__m128d __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm_mask_expandloadu_pd - // CHECK: @llvm.x86.avx512.mask.expand.load.pd.128 + // CHECK: @llvm.masked.expandload.v2f64(double* %{{.*}}, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) return _mm_mask_expandloadu_pd(__W,__U,__P); } __m128d test_mm_maskz_expandloadu_pd(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm_maskz_expandloadu_pd - // CHECK: @llvm.x86.avx512.mask.expand.load.pd.128 + // CHECK: @llvm.masked.expandload.v2f64(double* %{{.*}}, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) return _mm_maskz_expandloadu_pd(__U,__P); } __m256d test_mm256_mask_expandloadu_pd(__m256d __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm256_mask_expandloadu_pd - // CHECK: @llvm.x86.avx512.mask.expand.load.pd.256 + // CHECK: @llvm.masked.expandload.v4f64(double* %{{.*}}, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) return _mm256_mask_expandloadu_pd(__W,__U,__P); } __m256d test_mm256_maskz_expandloadu_pd(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm256_maskz_expandloadu_pd - // CHECK: @llvm.x86.avx512.mask.expand.load.pd.256 + // CHECK: @llvm.masked.expandload.v4f64(double* %{{.*}}, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) return _mm256_maskz_expandloadu_pd(__U,__P); } __m128i test_mm_mask_expandloadu_epi64(__m128i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm_mask_expandloadu_epi64 - // CHECK: @llvm.x86.avx512.mask.expand.load.q.128 + // CHECK: @llvm.masked.expandload.v2i64(i64* %{{.*}}, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) return _mm_mask_expandloadu_epi64(__W,__U,__P); } __m128i test_mm_maskz_expandloadu_epi64(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm_maskz_expandloadu_epi64 - // CHECK: @llvm.x86.avx512.mask.expand.load.q.128 + // CHECK: @llvm.masked.expandload.v2i64(i64* %{{.*}}, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) return _mm_maskz_expandloadu_epi64(__U,__P); } __m256i test_mm256_mask_expandloadu_epi64(__m256i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm256_mask_expandloadu_epi64 - // CHECK: @llvm.x86.avx512.mask.expand.load.q.256 + // CHECK: @llvm.masked.expandload.v4i64(i64* %{{.*}}, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) return _mm256_mask_expandloadu_epi64(__W,__U,__P); } __m256i test_mm256_maskz_expandloadu_epi64(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm256_maskz_expandloadu_epi64 - // CHECK: @llvm.x86.avx512.mask.expand.load.q.256 + // CHECK: @llvm.masked.expandload.v4i64(i64* %{{.*}}, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) return _mm256_maskz_expandloadu_epi64(__U,__P); } __m128 test_mm_mask_expandloadu_ps(__m128 __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm_mask_expandloadu_ps - // CHECK: @llvm.x86.avx512.mask.expand.load.ps.128 + // CHECK: @llvm.masked.expandload.v4f32(float* %{{.*}}, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) return _mm_mask_expandloadu_ps(__W,__U,__P); } __m128 test_mm_maskz_expandloadu_ps(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm_maskz_expandloadu_ps - // CHECK: @llvm.x86.avx512.mask.expand.load.ps.128 + // CHECK: @llvm.masked.expandload.v4f32(float* %{{.*}}, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) return _mm_maskz_expandloadu_ps(__U,__P); } __m256 test_mm256_mask_expandloadu_ps(__m256 __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm256_mask_expandloadu_ps - // CHECK: @llvm.x86.avx512.mask.expand.load.ps.256 + // CHECK: @llvm.masked.expandload.v8f32(float* %{{.*}}, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) return _mm256_mask_expandloadu_ps(__W,__U,__P); } __m256 test_mm256_maskz_expandloadu_ps(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm256_maskz_expandloadu_ps - // CHECK: @llvm.x86.avx512.mask.expand.load.ps.256 + // CHECK: @llvm.masked.expandload.v8f32(float* %{{.*}}, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) return _mm256_maskz_expandloadu_ps(__U,__P); } __m128i test_mm_mask_expandloadu_epi32(__m128i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm_mask_expandloadu_epi32 - // CHECK: @llvm.x86.avx512.mask.expand.load.d.128 + // CHECK: @llvm.masked.expandload.v4i32(i32* %{{.*}}, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) return _mm_mask_expandloadu_epi32(__W,__U,__P); } __m128i test_mm_maskz_expandloadu_epi32(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm_maskz_expandloadu_epi32 - // CHECK: @llvm.x86.avx512.mask.expand.load.d.128 + // CHECK: @llvm.masked.expandload.v4i32(i32* %{{.*}}, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) return _mm_maskz_expandloadu_epi32(__U,__P); } __m256i test_mm256_mask_expandloadu_epi32(__m256i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm256_mask_expandloadu_epi32 - // CHECK: @llvm.x86.avx512.mask.expand.load.d.256 + // CHECK: @llvm.masked.expandload.v8i32(i32* %{{.*}}, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) return _mm256_mask_expandloadu_epi32(__W,__U,__P); } __m256i test_mm256_maskz_expandloadu_epi32(__mmask8 __U, void const *__P) { // CHECK-LABEL: @test_mm256_maskz_expandloadu_epi32 - // CHECK: @llvm.x86.avx512.mask.expand.load.d.256 + // CHECK: @llvm.masked.expandload.v8i32(i32* %{{.*}}, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) return _mm256_maskz_expandloadu_epi32(__U,__P); } __m128 test_mm_mask_expand_ps(__m128 __W, __mmask8 __U, __m128 __A) { Modified: cfe/trunk/test/CodeGen/avx512vlvbmi2-builtins.c URL: http://llvm.org/viewvc/llvm-project/cfe/trunk/test/CodeGen/avx512vlvbmi2-builtins.c?rev=334366&r1=334365&r2=334366&view=diff ============================================================================== --- cfe/trunk/test/CodeGen/avx512vlvbmi2-builtins.c (original) +++ cfe/trunk/test/CodeGen/avx512vlvbmi2-builtins.c Sun Jun 10 10:27:05 2018 @@ -28,13 +28,13 @@ __m128i test_mm_maskz_compress_epi8(__mm void test_mm_mask_compressstoreu_epi16(void *__P, __mmask8 __U, __m128i __D) { // CHECK-LABEL: @test_mm_mask_compressstoreu_epi16 - // CHECK: @llvm.x86.avx512.mask.compress.store.w.128 + // CHECK: @llvm.masked.compressstore.v8i16(<8 x i16> %{{.*}}, i16* %{{.*}}, <8 x i1> %{{.*}}) _mm_mask_compressstoreu_epi16(__P, __U, __D); } void test_mm_mask_compressstoreu_epi8(void *__P, __mmask16 __U, __m128i __D) { // CHECK-LABEL: @test_mm_mask_compressstoreu_epi8 - // CHECK: @llvm.x86.avx512.mask.compress.store.b.128 + // CHECK: @llvm.masked.compressstore.v16i8(<16 x i8> %{{.*}}, i8* %{{.*}}, <16 x i1> %{{.*}}) _mm_mask_compressstoreu_epi8(__P, __U, __D); } @@ -64,25 +64,25 @@ __m128i test_mm_maskz_expand_epi8(__mmas __m128i test_mm_mask_expandloadu_epi16(__m128i __S, __mmask8 __U, void const* __P) { // CHECK-LABEL: @test_mm_mask_expandloadu_epi16 - // CHECK: @llvm.x86.avx512.mask.expand.load.w.128 + // CHECK: @llvm.masked.expandload.v8i16(i16* %{{.*}}, <8 x i1> %{{.*}}, <8 x i16> %{{.*}}) return _mm_mask_expandloadu_epi16(__S, __U, __P); } __m128i test_mm_maskz_expandloadu_epi16(__mmask8 __U, void const* __P) { // CHECK-LABEL: @test_mm_maskz_expandloadu_epi16 - // CHECK: @llvm.x86.avx512.mask.expand.load.w.128 + // CHECK: @llvm.masked.expandload.v8i16(i16* %{{.*}}, <8 x i1> %{{.*}}, <8 x i16> %{{.*}}) return _mm_maskz_expandloadu_epi16(__U, __P); } __m128i test_mm_mask_expandloadu_epi8(__m128i __S, __mmask16 __U, void const* __P) { // CHECK-LABEL: @test_mm_mask_expandloadu_epi8 - // CHECK: @llvm.x86.avx512.mask.expand.load.b.128 + // CHECK: @llvm.masked.expandload.v16i8(i8* %{{.*}}, <16 x i1> %{{.*}}, <16 x i8> %{{.*}}) return _mm_mask_expandloadu_epi8(__S, __U, __P); } __m128i test_mm_maskz_expandloadu_epi8(__mmask16 __U, void const* __P) { // CHECK-LABEL: @test_mm_maskz_expandloadu_epi8 - // CHECK: @llvm.x86.avx512.mask.expand.load.b.128 + // CHECK: @llvm.masked.expandload.v16i8(i8* %{{.*}}, <16 x i1> %{{.*}}, <16 x i8> %{{.*}}) return _mm_maskz_expandloadu_epi8(__U, __P); } @@ -112,13 +112,13 @@ __m256i test_mm256_maskz_compress_epi8(_ void test_mm256_mask_compressstoreu_epi16(void *__P, __mmask16 __U, __m256i __D) { // CHECK-LABEL: @test_mm256_mask_compressstoreu_epi16 - // CHECK: @llvm.x86.avx512.mask.compress.store.w.256 + // CHECK: @llvm.masked.compressstore.v16i16(<16 x i16> %{{.*}}, i16* %{{.*}}, <16 x i1> %{{.*}}) _mm256_mask_compressstoreu_epi16(__P, __U, __D); } void test_mm256_mask_compressstoreu_epi8(void *__P, __mmask32 __U, __m256i __D) { // CHECK-LABEL: @test_mm256_mask_compressstoreu_epi8 - // CHECK: @llvm.x86.avx512.mask.compress.store.b.256 + // CHECK: @llvm.masked.compressstore.v32i8(<32 x i8> %{{.*}}, i8* %{{.*}}, <32 x i1> %{{.*}}) _mm256_mask_compressstoreu_epi8(__P, __U, __D); } @@ -148,25 +148,25 @@ __m256i test_mm256_maskz_expand_epi8(__m __m256i test_mm256_mask_expandloadu_epi16(__m256i __S, __mmask16 __U, void const* __P) { // CHECK-LABEL: @test_mm256_mask_expandloadu_epi16 - // CHECK: @llvm.x86.avx512.mask.expand.load.w.256 + // CHECK: @llvm.masked.expandload.v16i16(i16* %{{.*}}, <16 x i1> %{{.*}}, <16 x i16> %{{.*}}) return _mm256_mask_expandloadu_epi16(__S, __U, __P); } __m256i test_mm256_maskz_expandloadu_epi16(__mmask16 __U, void const* __P) { // CHECK-LABEL: @test_mm256_maskz_expandloadu_epi16 - // CHECK: @llvm.x86.avx512.mask.expand.load.w.256 + // CHECK: @llvm.masked.expandload.v16i16(i16* %{{.*}}, <16 x i1> %{{.*}}, <16 x i16> %{{.*}}) return _mm256_maskz_expandloadu_epi16(__U, __P); } __m256i test_mm256_mask_expandloadu_epi8(__m256i __S, __mmask32 __U, void const* __P) { // CHECK-LABEL: @test_mm256_mask_expandloadu_epi8 - // CHECK: @llvm.x86.avx512.mask.expand.load.b.256 + // CHECK: @llvm.masked.expandload.v32i8(i8* %{{.*}}, <32 x i1> %{{.*}}, <32 x i8> %{{.*}}) return _mm256_mask_expandloadu_epi8(__S, __U, __P); } __m256i test_mm256_maskz_expandloadu_epi8(__mmask32 __U, void const* __P) { // CHECK-LABEL: @test_mm256_maskz_expandloadu_epi8 - // CHECK: @llvm.x86.avx512.mask.expand.load.b.256 + // CHECK: @llvm.masked.expandload.v32i8(i8* %{{.*}}, <32 x i1> %{{.*}}, <32 x i8> %{{.*}}) return _mm256_maskz_expandloadu_epi8(__U, __P); } _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
