diff options
| -rw-r--r-- | llvm/test/CodeGen/X86/widen_mul.ll | 265 |
1 files changed, 0 insertions, 265 deletions
diff --git a/llvm/test/CodeGen/X86/widen_mul.ll b/llvm/test/CodeGen/X86/widen_mul.ll deleted file mode 100644 index 62d1b33e1f4..00000000000 --- a/llvm/test/CodeGen/X86/widen_mul.ll +++ /dev/null @@ -1,265 +0,0 @@ -; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py -; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown | FileCheck %s --check-prefix=SSE --check-prefix=SSE2 -; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+sse4.1 | FileCheck %s --check-prefix=SSE --check-prefix=SSE41 -; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx2 | FileCheck %s --check-prefix=AVX --check-prefix=AVX2 -; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512f | FileCheck %s --check-prefix=AVX --check-prefix=AVX512 --check-prefix=AVX512F -; RUN: llc < %s -x86-experimental-vector-widening-legalization -mtriple=x86_64-unknown-unknown -mattr=+avx512bw | FileCheck %s --check-prefix=AVX --check-prefix=AVX512 --check-prefix=AVX512BW - -; Test multiplies of various narrow types. - -define <2 x i8> @mul_v2i8(<2 x i8> %x, <2 x i8> %y) { -; SSE2-LABEL: mul_v2i8: -; SSE2: # %bb.0: -; SSE2-NEXT: movdqa %xmm1, %xmm2 -; SSE2-NEXT: punpckhbw {{.*#+}} xmm2 = xmm2[8],xmm0[8],xmm2[9],xmm0[9],xmm2[10],xmm0[10],xmm2[11],xmm0[11],xmm2[12],xmm0[12],xmm2[13],xmm0[13],xmm2[14],xmm0[14],xmm2[15],xmm0[15] -; SSE2-NEXT: movdqa %xmm0, %xmm3 -; SSE2-NEXT: punpckhbw {{.*#+}} xmm3 = xmm3[8],xmm0[8],xmm3[9],xmm0[9],xmm3[10],xmm0[10],xmm3[11],xmm0[11],xmm3[12],xmm0[12],xmm3[13],xmm0[13],xmm3[14],xmm0[14],xmm3[15],xmm0[15] -; SSE2-NEXT: pmullw %xmm2, %xmm3 -; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [255,255,255,255,255,255,255,255] -; SSE2-NEXT: pand %xmm2, %xmm3 -; SSE2-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7] -; SSE2-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] -; SSE2-NEXT: pmullw %xmm1, %xmm0 -; SSE2-NEXT: pand %xmm2, %xmm0 -; SSE2-NEXT: packuswb %xmm3, %xmm0 -; SSE2-NEXT: retq -; -; SSE41-LABEL: mul_v2i8: -; SSE41: # %bb.0: -; SSE41-NEXT: pmovzxbw {{.*#+}} xmm3 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero -; SSE41-NEXT: punpckhbw {{.*#+}} xmm1 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; SSE41-NEXT: pmovzxbw {{.*#+}} xmm2 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero -; SSE41-NEXT: punpckhbw {{.*#+}} xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; SSE41-NEXT: pmullw %xmm1, %xmm0 -; SSE41-NEXT: movdqa {{.*#+}} xmm1 = [255,255,255,255,255,255,255,255] -; SSE41-NEXT: pand %xmm1, %xmm0 -; SSE41-NEXT: pmullw %xmm3, %xmm2 -; SSE41-NEXT: pand %xmm1, %xmm2 -; SSE41-NEXT: packuswb %xmm0, %xmm2 -; SSE41-NEXT: movdqa %xmm2, %xmm0 -; SSE41-NEXT: retq -; -; AVX2-LABEL: mul_v2i8: -; AVX2: # %bb.0: -; AVX2-NEXT: vpmovsxbw %xmm1, %ymm1 -; AVX2-NEXT: vpmovsxbw %xmm0, %ymm0 -; AVX2-NEXT: vpmullw %ymm1, %ymm0, %ymm0 -; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1 -; AVX2-NEXT: vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u> -; AVX2-NEXT: vpshufb %xmm2, %xmm1, %xmm1 -; AVX2-NEXT: vpshufb %xmm2, %xmm0, %xmm0 -; AVX2-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0] -; AVX2-NEXT: vzeroupper -; AVX2-NEXT: retq -; -; AVX512F-LABEL: mul_v2i8: -; AVX512F: # %bb.0: -; AVX512F-NEXT: vpmovsxbw %xmm1, %ymm1 -; AVX512F-NEXT: vpmovsxbw %xmm0, %ymm0 -; AVX512F-NEXT: vpmullw %ymm1, %ymm0, %ymm0 -; AVX512F-NEXT: vpmovzxwd {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero -; AVX512F-NEXT: vpmovdb %zmm0, %xmm0 -; AVX512F-NEXT: vzeroupper -; AVX512F-NEXT: retq -; -; AVX512BW-LABEL: mul_v2i8: -; AVX512BW: # %bb.0: -; AVX512BW-NEXT: vpmovsxbw %xmm1, %ymm1 -; AVX512BW-NEXT: vpmovsxbw %xmm0, %ymm0 -; AVX512BW-NEXT: vpmullw %ymm1, %ymm0, %ymm0 -; AVX512BW-NEXT: vpmovwb %zmm0, %ymm0 -; AVX512BW-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0 -; AVX512BW-NEXT: vzeroupper -; AVX512BW-NEXT: retq - %res = mul <2 x i8> %x, %y - ret <2 x i8> %res -} - -define <4 x i8> @mul_v4i8(<4 x i8> %x, <4 x i8> %y) { -; SSE2-LABEL: mul_v4i8: -; SSE2: # %bb.0: -; SSE2-NEXT: movdqa %xmm1, %xmm2 -; SSE2-NEXT: punpckhbw {{.*#+}} xmm2 = xmm2[8],xmm0[8],xmm2[9],xmm0[9],xmm2[10],xmm0[10],xmm2[11],xmm0[11],xmm2[12],xmm0[12],xmm2[13],xmm0[13],xmm2[14],xmm0[14],xmm2[15],xmm0[15] -; SSE2-NEXT: movdqa %xmm0, %xmm3 -; SSE2-NEXT: punpckhbw {{.*#+}} xmm3 = xmm3[8],xmm0[8],xmm3[9],xmm0[9],xmm3[10],xmm0[10],xmm3[11],xmm0[11],xmm3[12],xmm0[12],xmm3[13],xmm0[13],xmm3[14],xmm0[14],xmm3[15],xmm0[15] -; SSE2-NEXT: pmullw %xmm2, %xmm3 -; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [255,255,255,255,255,255,255,255] -; SSE2-NEXT: pand %xmm2, %xmm3 -; SSE2-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7] -; SSE2-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] -; SSE2-NEXT: pmullw %xmm1, %xmm0 -; SSE2-NEXT: pand %xmm2, %xmm0 -; SSE2-NEXT: packuswb %xmm3, %xmm0 -; SSE2-NEXT: retq -; -; SSE41-LABEL: mul_v4i8: -; SSE41: # %bb.0: -; SSE41-NEXT: pmovzxbw {{.*#+}} xmm3 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero -; SSE41-NEXT: punpckhbw {{.*#+}} xmm1 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; SSE41-NEXT: pmovzxbw {{.*#+}} xmm2 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero -; SSE41-NEXT: punpckhbw {{.*#+}} xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; SSE41-NEXT: pmullw %xmm1, %xmm0 -; SSE41-NEXT: movdqa {{.*#+}} xmm1 = [255,255,255,255,255,255,255,255] -; SSE41-NEXT: pand %xmm1, %xmm0 -; SSE41-NEXT: pmullw %xmm3, %xmm2 -; SSE41-NEXT: pand %xmm1, %xmm2 -; SSE41-NEXT: packuswb %xmm0, %xmm2 -; SSE41-NEXT: movdqa %xmm2, %xmm0 -; SSE41-NEXT: retq -; -; AVX2-LABEL: mul_v4i8: -; AVX2: # %bb.0: -; AVX2-NEXT: vpmovsxbw %xmm1, %ymm1 -; AVX2-NEXT: vpmovsxbw %xmm0, %ymm0 -; AVX2-NEXT: vpmullw %ymm1, %ymm0, %ymm0 -; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1 -; AVX2-NEXT: vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u> -; AVX2-NEXT: vpshufb %xmm2, %xmm1, %xmm1 -; AVX2-NEXT: vpshufb %xmm2, %xmm0, %xmm0 -; AVX2-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0] -; AVX2-NEXT: vzeroupper -; AVX2-NEXT: retq -; -; AVX512F-LABEL: mul_v4i8: -; AVX512F: # %bb.0: -; AVX512F-NEXT: vpmovsxbw %xmm1, %ymm1 -; AVX512F-NEXT: vpmovsxbw %xmm0, %ymm0 -; AVX512F-NEXT: vpmullw %ymm1, %ymm0, %ymm0 -; AVX512F-NEXT: vpmovzxwd {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero -; AVX512F-NEXT: vpmovdb %zmm0, %xmm0 -; AVX512F-NEXT: vzeroupper -; AVX512F-NEXT: retq -; -; AVX512BW-LABEL: mul_v4i8: -; AVX512BW: # %bb.0: -; AVX512BW-NEXT: vpmovsxbw %xmm1, %ymm1 -; AVX512BW-NEXT: vpmovsxbw %xmm0, %ymm0 -; AVX512BW-NEXT: vpmullw %ymm1, %ymm0, %ymm0 -; AVX512BW-NEXT: vpmovwb %zmm0, %ymm0 -; AVX512BW-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0 -; AVX512BW-NEXT: vzeroupper -; AVX512BW-NEXT: retq - %res = mul <4 x i8> %x, %y - ret <4 x i8> %res -} - -define <8 x i8> @mul_v8i8(<8 x i8> %x, <8 x i8> %y) { -; SSE2-LABEL: mul_v8i8: -; SSE2: # %bb.0: -; SSE2-NEXT: movdqa %xmm1, %xmm2 -; SSE2-NEXT: punpckhbw {{.*#+}} xmm2 = xmm2[8],xmm0[8],xmm2[9],xmm0[9],xmm2[10],xmm0[10],xmm2[11],xmm0[11],xmm2[12],xmm0[12],xmm2[13],xmm0[13],xmm2[14],xmm0[14],xmm2[15],xmm0[15] -; SSE2-NEXT: movdqa %xmm0, %xmm3 -; SSE2-NEXT: punpckhbw {{.*#+}} xmm3 = xmm3[8],xmm0[8],xmm3[9],xmm0[9],xmm3[10],xmm0[10],xmm3[11],xmm0[11],xmm3[12],xmm0[12],xmm3[13],xmm0[13],xmm3[14],xmm0[14],xmm3[15],xmm0[15] -; SSE2-NEXT: pmullw %xmm2, %xmm3 -; SSE2-NEXT: movdqa {{.*#+}} xmm2 = [255,255,255,255,255,255,255,255] -; SSE2-NEXT: pand %xmm2, %xmm3 -; SSE2-NEXT: punpcklbw {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1],xmm1[2],xmm0[2],xmm1[3],xmm0[3],xmm1[4],xmm0[4],xmm1[5],xmm0[5],xmm1[6],xmm0[6],xmm1[7],xmm0[7] -; SSE2-NEXT: punpcklbw {{.*#+}} xmm0 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] -; SSE2-NEXT: pmullw %xmm1, %xmm0 -; SSE2-NEXT: pand %xmm2, %xmm0 -; SSE2-NEXT: packuswb %xmm3, %xmm0 -; SSE2-NEXT: retq -; -; SSE41-LABEL: mul_v8i8: -; SSE41: # %bb.0: -; SSE41-NEXT: pmovzxbw {{.*#+}} xmm3 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero -; SSE41-NEXT: punpckhbw {{.*#+}} xmm1 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; SSE41-NEXT: pmovzxbw {{.*#+}} xmm2 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero -; SSE41-NEXT: punpckhbw {{.*#+}} xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; SSE41-NEXT: pmullw %xmm1, %xmm0 -; SSE41-NEXT: movdqa {{.*#+}} xmm1 = [255,255,255,255,255,255,255,255] -; SSE41-NEXT: pand %xmm1, %xmm0 -; SSE41-NEXT: pmullw %xmm3, %xmm2 -; SSE41-NEXT: pand %xmm1, %xmm2 -; SSE41-NEXT: packuswb %xmm0, %xmm2 -; SSE41-NEXT: movdqa %xmm2, %xmm0 -; SSE41-NEXT: retq -; -; AVX2-LABEL: mul_v8i8: -; AVX2: # %bb.0: -; AVX2-NEXT: vpmovsxbw %xmm1, %ymm1 -; AVX2-NEXT: vpmovsxbw %xmm0, %ymm0 -; AVX2-NEXT: vpmullw %ymm1, %ymm0, %ymm0 -; AVX2-NEXT: vextracti128 $1, %ymm0, %xmm1 -; AVX2-NEXT: vmovdqa {{.*#+}} xmm2 = <0,2,4,6,8,10,12,14,u,u,u,u,u,u,u,u> -; AVX2-NEXT: vpshufb %xmm2, %xmm1, %xmm1 -; AVX2-NEXT: vpshufb %xmm2, %xmm0, %xmm0 -; AVX2-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0] -; AVX2-NEXT: vzeroupper -; AVX2-NEXT: retq -; -; AVX512F-LABEL: mul_v8i8: -; AVX512F: # %bb.0: -; AVX512F-NEXT: vpmovsxbw %xmm1, %ymm1 -; AVX512F-NEXT: vpmovsxbw %xmm0, %ymm0 -; AVX512F-NEXT: vpmullw %ymm1, %ymm0, %ymm0 -; AVX512F-NEXT: vpmovzxwd {{.*#+}} zmm0 = ymm0[0],zero,ymm0[1],zero,ymm0[2],zero,ymm0[3],zero,ymm0[4],zero,ymm0[5],zero,ymm0[6],zero,ymm0[7],zero,ymm0[8],zero,ymm0[9],zero,ymm0[10],zero,ymm0[11],zero,ymm0[12],zero,ymm0[13],zero,ymm0[14],zero,ymm0[15],zero -; AVX512F-NEXT: vpmovdb %zmm0, %xmm0 -; AVX512F-NEXT: vzeroupper -; AVX512F-NEXT: retq -; -; AVX512BW-LABEL: mul_v8i8: -; AVX512BW: # %bb.0: -; AVX512BW-NEXT: vpmovsxbw %xmm1, %ymm1 -; AVX512BW-NEXT: vpmovsxbw %xmm0, %ymm0 -; AVX512BW-NEXT: vpmullw %ymm1, %ymm0, %ymm0 -; AVX512BW-NEXT: vpmovwb %zmm0, %ymm0 -; AVX512BW-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0 -; AVX512BW-NEXT: vzeroupper -; AVX512BW-NEXT: retq - %res = mul <8 x i8> %x, %y - ret <8 x i8> %res -} - -define <2 x i16> @mul_v2i16(<2 x i16> %x, <2 x i16> %y) { -; SSE-LABEL: mul_v2i16: -; SSE: # %bb.0: -; SSE-NEXT: pmullw %xmm1, %xmm0 -; SSE-NEXT: retq -; -; AVX-LABEL: mul_v2i16: -; AVX: # %bb.0: -; AVX-NEXT: vpmullw %xmm1, %xmm0, %xmm0 -; AVX-NEXT: retq - %res = mul <2 x i16> %x, %y - ret <2 x i16> %res -} - -define <4 x i16> @mul_v4i16(<4 x i16> %x, <4 x i16> %y) { -; SSE-LABEL: mul_v4i16: -; SSE: # %bb.0: -; SSE-NEXT: pmullw %xmm1, %xmm0 -; SSE-NEXT: retq -; -; AVX-LABEL: mul_v4i16: -; AVX: # %bb.0: -; AVX-NEXT: vpmullw %xmm1, %xmm0, %xmm0 -; AVX-NEXT: retq - %res = mul <4 x i16> %x, %y - ret <4 x i16> %res -} - -define <2 x i32> @mul_v2i32(<2 x i32> %x, <2 x i32> %y) { -; SSE2-LABEL: mul_v2i32: -; SSE2: # %bb.0: -; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm0[1,1,3,3] -; SSE2-NEXT: pmuludq %xmm1, %xmm0 -; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3] -; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[1,1,3,3] -; SSE2-NEXT: pmuludq %xmm2, %xmm1 -; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3] -; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] -; SSE2-NEXT: retq -; -; SSE41-LABEL: mul_v2i32: -; SSE41: # %bb.0: -; SSE41-NEXT: pmulld %xmm1, %xmm0 -; SSE41-NEXT: retq -; -; AVX-LABEL: mul_v2i32: -; AVX: # %bb.0: -; AVX-NEXT: vpmulld %xmm1, %xmm0, %xmm0 -; AVX-NEXT: retq - %res = mul <2 x i32> %x, %y - ret <2 x i32> %res -} |
