diff options
Diffstat (limited to 'llvm/test/CodeGen/X86/vector-reduce-mul.ll')
| -rw-r--r-- | llvm/test/CodeGen/X86/vector-reduce-mul.ll | 267 |
1 files changed, 132 insertions, 135 deletions
diff --git a/llvm/test/CodeGen/X86/vector-reduce-mul.ll b/llvm/test/CodeGen/X86/vector-reduce-mul.ll index 8fa6baa44df..79ac2195664 100644 --- a/llvm/test/CodeGen/X86/vector-reduce-mul.ll +++ b/llvm/test/CodeGen/X86/vector-reduce-mul.ll @@ -19,8 +19,8 @@ define i64 @test_v2i64(<2 x i64> %a0) { ; SSE-NEXT: movdqa %xmm0, %xmm2 ; SSE-NEXT: psrlq $32, %xmm2 ; SSE-NEXT: pmuludq %xmm1, %xmm2 -; SSE-NEXT: movdqa %xmm1, %xmm3 -; SSE-NEXT: psrlq $32, %xmm3 +; SSE-NEXT: movdqa %xmm0, %xmm3 +; SSE-NEXT: psrldq {{.*#+}} xmm3 = xmm3[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero ; SSE-NEXT: pmuludq %xmm0, %xmm3 ; SSE-NEXT: paddq %xmm2, %xmm3 ; SSE-NEXT: psllq $32, %xmm3 @@ -34,7 +34,7 @@ define i64 @test_v2i64(<2 x i64> %a0) { ; AVX-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] ; AVX-NEXT: vpsrlq $32, %xmm0, %xmm2 ; AVX-NEXT: vpmuludq %xmm1, %xmm2, %xmm2 -; AVX-NEXT: vpsrlq $32, %xmm1, %xmm3 +; AVX-NEXT: vpsrldq {{.*#+}} xmm3 = xmm0[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero ; AVX-NEXT: vpmuludq %xmm3, %xmm0, %xmm3 ; AVX-NEXT: vpaddq %xmm2, %xmm3, %xmm2 ; AVX-NEXT: vpsllq $32, %xmm2, %xmm2 @@ -48,7 +48,7 @@ define i64 @test_v2i64(<2 x i64> %a0) { ; AVX512BW-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] ; AVX512BW-NEXT: vpsrlq $32, %xmm0, %xmm2 ; AVX512BW-NEXT: vpmuludq %xmm1, %xmm2, %xmm2 -; AVX512BW-NEXT: vpsrlq $32, %xmm1, %xmm3 +; AVX512BW-NEXT: vpsrldq {{.*#+}} xmm3 = xmm0[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero ; AVX512BW-NEXT: vpmuludq %xmm3, %xmm0, %xmm3 ; AVX512BW-NEXT: vpaddq %xmm2, %xmm3, %xmm2 ; AVX512BW-NEXT: vpsllq $32, %xmm2, %xmm2 @@ -107,8 +107,8 @@ define i64 @test_v4i64(<4 x i64> %a0) { ; SSE-NEXT: movdqa %xmm0, %xmm2 ; SSE-NEXT: psrlq $32, %xmm2 ; SSE-NEXT: pmuludq %xmm1, %xmm2 -; SSE-NEXT: movdqa %xmm1, %xmm3 -; SSE-NEXT: psrlq $32, %xmm3 +; SSE-NEXT: movdqa %xmm0, %xmm3 +; SSE-NEXT: psrldq {{.*#+}} xmm3 = xmm3[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero ; SSE-NEXT: pmuludq %xmm0, %xmm3 ; SSE-NEXT: paddq %xmm2, %xmm3 ; SSE-NEXT: psllq $32, %xmm3 @@ -153,11 +153,11 @@ define i64 @test_v4i64(<4 x i64> %a0) { ; AVX2-NEXT: vpmuludq %ymm1, %ymm0, %ymm0 ; AVX2-NEXT: vpaddq %ymm2, %ymm0, %ymm0 ; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; AVX2-NEXT: vpsrlq $32, %ymm0, %ymm2 -; AVX2-NEXT: vpmuludq %ymm1, %ymm2, %ymm2 -; AVX2-NEXT: vpsrlq $32, %ymm1, %ymm3 -; AVX2-NEXT: vpmuludq %ymm3, %ymm0, %ymm3 -; AVX2-NEXT: vpaddq %ymm2, %ymm3, %ymm2 +; AVX2-NEXT: vpsrldq {{.*#+}} ymm2 = ymm0[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,ymm0[28,29,30,31],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero +; AVX2-NEXT: vpmuludq %ymm2, %ymm0, %ymm2 +; AVX2-NEXT: vpsrlq $32, %ymm0, %ymm3 +; AVX2-NEXT: vpmuludq %ymm1, %ymm3, %ymm3 +; AVX2-NEXT: vpaddq %ymm3, %ymm2, %ymm2 ; AVX2-NEXT: vpsllq $32, %ymm2, %ymm2 ; AVX2-NEXT: vpmuludq %ymm1, %ymm0, %ymm0 ; AVX2-NEXT: vpaddq %ymm2, %ymm0, %ymm0 @@ -177,11 +177,11 @@ define i64 @test_v4i64(<4 x i64> %a0) { ; AVX512BW-NEXT: vpmuludq %ymm1, %ymm0, %ymm0 ; AVX512BW-NEXT: vpaddq %ymm2, %ymm0, %ymm0 ; AVX512BW-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; AVX512BW-NEXT: vpsrlq $32, %ymm0, %ymm2 -; AVX512BW-NEXT: vpmuludq %ymm1, %ymm2, %ymm2 -; AVX512BW-NEXT: vpsrlq $32, %ymm1, %ymm3 -; AVX512BW-NEXT: vpmuludq %ymm3, %ymm0, %ymm3 -; AVX512BW-NEXT: vpaddq %ymm2, %ymm3, %ymm2 +; AVX512BW-NEXT: vpsrldq {{.*#+}} ymm2 = ymm0[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,ymm0[28,29,30,31],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero +; AVX512BW-NEXT: vpmuludq %ymm2, %ymm0, %ymm2 +; AVX512BW-NEXT: vpsrlq $32, %ymm0, %ymm3 +; AVX512BW-NEXT: vpmuludq %ymm1, %ymm3, %ymm3 +; AVX512BW-NEXT: vpaddq %ymm3, %ymm2, %ymm2 ; AVX512BW-NEXT: vpsllq $32, %ymm2, %ymm2 ; AVX512BW-NEXT: vpmuludq %ymm1, %ymm0, %ymm0 ; AVX512BW-NEXT: vpaddq %ymm2, %ymm0, %ymm0 @@ -274,8 +274,8 @@ define i64 @test_v8i64(<8 x i64> %a0) { ; SSE-NEXT: movdqa %xmm0, %xmm2 ; SSE-NEXT: psrlq $32, %xmm2 ; SSE-NEXT: pmuludq %xmm1, %xmm2 -; SSE-NEXT: movdqa %xmm1, %xmm3 -; SSE-NEXT: psrlq $32, %xmm3 +; SSE-NEXT: movdqa %xmm0, %xmm3 +; SSE-NEXT: psrldq {{.*#+}} xmm3 = xmm3[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero ; SSE-NEXT: pmuludq %xmm0, %xmm3 ; SSE-NEXT: paddq %xmm2, %xmm3 ; SSE-NEXT: psllq $32, %xmm3 @@ -345,11 +345,11 @@ define i64 @test_v8i64(<8 x i64> %a0) { ; AVX2-NEXT: vpmuludq %ymm1, %ymm0, %ymm0 ; AVX2-NEXT: vpaddq %ymm2, %ymm0, %ymm0 ; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; AVX2-NEXT: vpsrlq $32, %ymm0, %ymm2 -; AVX2-NEXT: vpmuludq %ymm1, %ymm2, %ymm2 -; AVX2-NEXT: vpsrlq $32, %ymm1, %ymm3 -; AVX2-NEXT: vpmuludq %ymm3, %ymm0, %ymm3 -; AVX2-NEXT: vpaddq %ymm2, %ymm3, %ymm2 +; AVX2-NEXT: vpsrldq {{.*#+}} ymm2 = ymm0[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,ymm0[28,29,30,31],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero +; AVX2-NEXT: vpmuludq %ymm2, %ymm0, %ymm2 +; AVX2-NEXT: vpsrlq $32, %ymm0, %ymm3 +; AVX2-NEXT: vpmuludq %ymm1, %ymm3, %ymm3 +; AVX2-NEXT: vpaddq %ymm3, %ymm2, %ymm2 ; AVX2-NEXT: vpsllq $32, %ymm2, %ymm2 ; AVX2-NEXT: vpmuludq %ymm1, %ymm0, %ymm0 ; AVX2-NEXT: vpaddq %ymm2, %ymm0, %ymm0 @@ -527,8 +527,8 @@ define i64 @test_v16i64(<16 x i64> %a0) { ; SSE-NEXT: movdqa %xmm0, %xmm2 ; SSE-NEXT: psrlq $32, %xmm2 ; SSE-NEXT: pmuludq %xmm1, %xmm2 -; SSE-NEXT: movdqa %xmm1, %xmm3 -; SSE-NEXT: psrlq $32, %xmm3 +; SSE-NEXT: movdqa %xmm0, %xmm3 +; SSE-NEXT: psrldq {{.*#+}} xmm3 = xmm3[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero ; SSE-NEXT: pmuludq %xmm0, %xmm3 ; SSE-NEXT: paddq %xmm2, %xmm3 ; SSE-NEXT: psllq $32, %xmm3 @@ -648,11 +648,11 @@ define i64 @test_v16i64(<16 x i64> %a0) { ; AVX2-NEXT: vpmuludq %ymm1, %ymm0, %ymm0 ; AVX2-NEXT: vpaddq %ymm2, %ymm0, %ymm0 ; AVX2-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; AVX2-NEXT: vpsrlq $32, %ymm0, %ymm2 -; AVX2-NEXT: vpmuludq %ymm1, %ymm2, %ymm2 -; AVX2-NEXT: vpsrlq $32, %ymm1, %ymm3 -; AVX2-NEXT: vpmuludq %ymm3, %ymm0, %ymm3 -; AVX2-NEXT: vpaddq %ymm2, %ymm3, %ymm2 +; AVX2-NEXT: vpsrldq {{.*#+}} ymm2 = ymm0[12,13,14,15],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,ymm0[28,29,30,31],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero +; AVX2-NEXT: vpmuludq %ymm2, %ymm0, %ymm2 +; AVX2-NEXT: vpsrlq $32, %ymm0, %ymm3 +; AVX2-NEXT: vpmuludq %ymm1, %ymm3, %ymm3 +; AVX2-NEXT: vpaddq %ymm3, %ymm2, %ymm2 ; AVX2-NEXT: vpsllq $32, %ymm2, %ymm2 ; AVX2-NEXT: vpmuludq %ymm1, %ymm0, %ymm0 ; AVX2-NEXT: vpaddq %ymm2, %ymm0, %ymm0 @@ -779,13 +779,13 @@ define i32 @test_v4i32(<4 x i32> %a0) { ; SSE2-LABEL: test_v4i32: ; SSE2: # %bb.0: ; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; 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: pshufd {{.*#+}} xmm2 = xmm0[3,3,1,1] +; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3] +; SSE2-NEXT: pmuludq %xmm2, %xmm3 +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm3[0,2,2,3] +; SSE2-NEXT: pmuludq %xmm0, %xmm1 +; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] ; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,2,3] ; SSE2-NEXT: pmuludq %xmm0, %xmm1 ; SSE2-NEXT: movd %xmm1, %eax @@ -829,13 +829,13 @@ define i32 @test_v8i32(<8 x i32> %a0) { ; 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: pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; 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: pshufd {{.*#+}} xmm2 = xmm1[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm0[2,3,0,1] +; SSE2-NEXT: pmuludq %xmm0, %xmm2 +; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm2[0,2,2,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm1[2,2,0,0] +; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,0,2,2] ; 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] @@ -901,29 +901,29 @@ define i32 @test_v16i32(<16 x i32> %a0) { ; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3] ; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3] ; SSE2-NEXT: pmuludq %xmm4, %xmm3 -; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[0,2,2,3] -; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] -; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm3[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm4[0],xmm1[1],xmm4[1] +; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm0[1,1,3,3] ; SSE2-NEXT: pmuludq %xmm2, %xmm0 ; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3] ; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3] -; SSE2-NEXT: pmuludq %xmm3, %xmm2 -; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[0,2,2,3] -; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] -; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm0[1,1,3,3] +; SSE2-NEXT: pmuludq %xmm4, %xmm2 +; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm2[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1] ; 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: pshufd {{.*#+}} xmm1 = xmm3[0,0,2,2] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[0,0,2,2] +; SSE2-NEXT: pmuludq %xmm1, %xmm2 +; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm2[0,2,2,3] ; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] ; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; 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: pmuludq %xmm0, %xmm1 +; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[0,2,2,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm2[2,2,0,0] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[0,0,2,2] +; SSE2-NEXT: pmuludq %xmm1, %xmm2 +; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm2[0,2,2,3] ; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1] ; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm0[1,1,2,3] ; SSE2-NEXT: pmuludq %xmm0, %xmm1 @@ -993,63 +993,63 @@ define i32 @test_v32i32(<32 x i32> %a0) { ; SSE2-NEXT: pshufd {{.*#+}} xmm8 = xmm2[1,1,3,3] ; SSE2-NEXT: pmuludq %xmm6, %xmm2 ; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[0,2,2,3] -; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm6[1,1,3,3] -; SSE2-NEXT: pmuludq %xmm8, %xmm6 -; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm6[0,2,2,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm9 = xmm6[1,1,3,3] +; SSE2-NEXT: pmuludq %xmm8, %xmm9 +; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm9[0,2,2,3] ; SSE2-NEXT: punpckldq {{.*#+}} xmm2 = xmm2[0],xmm6[0],xmm2[1],xmm6[1] -; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm0[1,1,3,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm8 = xmm0[1,1,3,3] ; SSE2-NEXT: pmuludq %xmm4, %xmm0 -; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3] -; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[1,1,3,3] -; SSE2-NEXT: pmuludq %xmm6, %xmm4 -; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm4[0,2,2,3] -; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm4[0],xmm0[1],xmm4[1] -; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm3[1,1,3,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm0[0,2,2,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm4[1,1,3,3] +; SSE2-NEXT: pmuludq %xmm8, %xmm0 +; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm0[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm6 = xmm6[0],xmm4[0],xmm6[1],xmm4[1] +; SSE2-NEXT: pmuludq %xmm2, %xmm6 +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm3[1,1,3,3] ; SSE2-NEXT: pmuludq %xmm7, %xmm3 ; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[0,2,2,3] -; SSE2-NEXT: pshufd {{.*#+}} xmm6 = xmm7[1,1,3,3] -; SSE2-NEXT: pmuludq %xmm4, %xmm6 -; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm6[0,2,2,3] -; SSE2-NEXT: punpckldq {{.*#+}} xmm3 = xmm3[0],xmm4[0],xmm3[1],xmm4[1] -; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm1[1,1,3,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm7[1,1,3,3] +; SSE2-NEXT: pmuludq %xmm2, %xmm4 +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm4[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm3 = xmm3[0],xmm2[0],xmm3[1],xmm2[1] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm1[1,1,3,3] ; SSE2-NEXT: pmuludq %xmm5, %xmm1 ; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3] ; SSE2-NEXT: pshufd {{.*#+}} xmm5 = xmm5[1,1,3,3] -; SSE2-NEXT: pmuludq %xmm4, %xmm5 -; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm5[0,2,2,3] -; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm4[0],xmm1[1],xmm4[1] -; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm1[1,1,3,3] +; SSE2-NEXT: pmuludq %xmm2, %xmm5 +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm5[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] ; SSE2-NEXT: pmuludq %xmm3, %xmm1 ; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm1[0,2,2,3] -; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[1,1,3,3] -; SSE2-NEXT: pmuludq %xmm4, %xmm3 -; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm3[0,2,2,3] -; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm3[0],xmm1[1],xmm3[1] -; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm0[1,1,3,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm4[0,0,2,2] +; SSE2-NEXT: pshufd {{.*#+}} xmm3 = xmm5[0,0,2,2] +; SSE2-NEXT: pmuludq %xmm2, %xmm3 +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm3[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm6[0,2,2,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm9[0,0,2,2] +; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,0,2,2] +; SSE2-NEXT: pmuludq %xmm4, %xmm0 +; SSE2-NEXT: pshufd {{.*#+}} xmm4 = xmm0[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm2 = xmm2[0],xmm4[0],xmm2[1],xmm4[1] +; SSE2-NEXT: pmuludq %xmm1, %xmm2 +; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm2[0,2,2,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm3[0,0,2,2] +; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,0,2,2] +; SSE2-NEXT: pmuludq %xmm2, %xmm0 +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm0[0,2,2,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm1[2,3,0,1] +; SSE2-NEXT: pmuludq %xmm1, %xmm2 +; SSE2-NEXT: pshufd {{.*#+}} xmm1 = xmm2[0,2,2,3] +; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm0[2,2,0,0] +; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,0,2,2] ; SSE2-NEXT: pmuludq %xmm2, %xmm0 ; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm0[0,2,2,3] -; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[1,1,3,3] -; SSE2-NEXT: pmuludq %xmm3, %xmm2 -; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm2[0,2,2,3] -; SSE2-NEXT: punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1] -; SSE2-NEXT: pshufd {{.*#+}} xmm2 = xmm0[1,1,3,3] +; SSE2-NEXT: punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1] +; SSE2-NEXT: pshufd {{.*#+}} xmm0 = xmm1[1,1,2,3] ; 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: pshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; 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: pshufd {{.*#+}} xmm1 = xmm0[1,1,2,3] -; SSE2-NEXT: pmuludq %xmm0, %xmm1 -; SSE2-NEXT: movd %xmm1, %eax +; SSE2-NEXT: movd %xmm0, %eax ; SSE2-NEXT: retq ; ; SSE41-LABEL: test_v32i32: @@ -1841,9 +1841,9 @@ define i8 @test_v32i8(<32 x i8> %a0) { ; AVX1-NEXT: vpand %xmm1, %xmm0, %xmm0 ; AVX1-NEXT: vpackuswb %xmm3, %xmm0, %xmm0 ; AVX1-NEXT: vpshufd {{.*#+}} xmm2 = xmm0[2,3,0,1] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm4 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpmullw %xmm4, %xmm3, %xmm3 +; AVX1-NEXT: vpunpcklbw {{.*#+}} xmm3 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] +; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm4 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] +; AVX1-NEXT: vpmullw %xmm3, %xmm4, %xmm3 ; AVX1-NEXT: vpand %xmm1, %xmm3, %xmm3 ; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero ; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm2 = xmm2[0],zero,xmm2[1],zero,xmm2[2],zero,xmm2[3],zero,xmm2[4],zero,xmm2[5],zero,xmm2[6],zero,xmm2[7],zero @@ -1851,15 +1851,14 @@ define i8 @test_v32i8(<32 x i8> %a0) { ; AVX1-NEXT: vpand %xmm1, %xmm0, %xmm0 ; AVX1-NEXT: vpackuswb %xmm3, %xmm0, %xmm0 ; AVX1-NEXT: vpshufd {{.*#+}} xmm2 = xmm0[1,1,2,3] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm4 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpmullw %xmm4, %xmm3, %xmm3 -; AVX1-NEXT: vpand %xmm1, %xmm3, %xmm3 -; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero +; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm3 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero ; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm2 = xmm2[0],zero,xmm2[1],zero,xmm2[2],zero,xmm2[3],zero,xmm2[4],zero,xmm2[5],zero,xmm2[6],zero,xmm2[7],zero -; AVX1-NEXT: vpmullw %xmm2, %xmm0, %xmm0 +; AVX1-NEXT: vpmullw %xmm2, %xmm3, %xmm2 +; AVX1-NEXT: vpand %xmm1, %xmm2, %xmm2 +; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] +; AVX1-NEXT: vpmullw %xmm0, %xmm0, %xmm0 ; AVX1-NEXT: vpand %xmm1, %xmm0, %xmm0 -; AVX1-NEXT: vpackuswb %xmm3, %xmm0, %xmm0 +; AVX1-NEXT: vpackuswb %xmm0, %xmm2, %xmm0 ; AVX1-NEXT: vpsrld $16, %xmm0, %xmm2 ; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] ; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm4 = xmm2[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] @@ -2229,9 +2228,9 @@ define i8 @test_v64i8(<64 x i8> %a0) { ; AVX1-NEXT: vpand %xmm2, %xmm0, %xmm0 ; AVX1-NEXT: vpackuswb %xmm1, %xmm0, %xmm0 ; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm4 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpmullw %xmm4, %xmm3, %xmm3 +; AVX1-NEXT: vpunpcklbw {{.*#+}} xmm3 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] +; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm4 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] +; AVX1-NEXT: vpmullw %xmm3, %xmm4, %xmm3 ; AVX1-NEXT: vpand %xmm2, %xmm3, %xmm3 ; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero ; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero @@ -2239,15 +2238,14 @@ define i8 @test_v64i8(<64 x i8> %a0) { ; AVX1-NEXT: vpand %xmm2, %xmm0, %xmm0 ; AVX1-NEXT: vpackuswb %xmm3, %xmm0, %xmm0 ; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm4 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpmullw %xmm4, %xmm3, %xmm3 -; AVX1-NEXT: vpand %xmm2, %xmm3, %xmm3 -; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero +; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm3 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero ; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero -; AVX1-NEXT: vpmullw %xmm1, %xmm0, %xmm0 +; AVX1-NEXT: vpmullw %xmm1, %xmm3, %xmm1 +; AVX1-NEXT: vpand %xmm2, %xmm1, %xmm1 +; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] +; AVX1-NEXT: vpmullw %xmm0, %xmm0, %xmm0 ; AVX1-NEXT: vpand %xmm2, %xmm0, %xmm0 -; AVX1-NEXT: vpackuswb %xmm3, %xmm0, %xmm0 +; AVX1-NEXT: vpackuswb %xmm0, %xmm1, %xmm0 ; AVX1-NEXT: vpsrld $16, %xmm0, %xmm1 ; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] ; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm4 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] @@ -2782,9 +2780,9 @@ define i8 @test_v128i8(<128 x i8> %a0) { ; AVX1-NEXT: vpand %xmm4, %xmm0, %xmm0 ; AVX1-NEXT: vpackuswb %xmm2, %xmm0, %xmm0 ; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[2,3,0,1] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm2 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpmullw %xmm3, %xmm2, %xmm2 +; AVX1-NEXT: vpunpcklbw {{.*#+}} xmm2 = xmm0[0,0,1,1,2,2,3,3,4,4,5,5,6,6,7,7] +; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] +; AVX1-NEXT: vpmullw %xmm2, %xmm3, %xmm2 ; AVX1-NEXT: vpand %xmm4, %xmm2, %xmm2 ; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero ; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero @@ -2792,15 +2790,14 @@ define i8 @test_v128i8(<128 x i8> %a0) { ; AVX1-NEXT: vpand %xmm4, %xmm0, %xmm0 ; AVX1-NEXT: vpackuswb %xmm2, %xmm0, %xmm0 ; AVX1-NEXT: vpshufd {{.*#+}} xmm1 = xmm0[1,1,2,3] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm2 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] -; AVX1-NEXT: vpmullw %xmm3, %xmm2, %xmm2 -; AVX1-NEXT: vpand %xmm4, %xmm2, %xmm2 -; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero,xmm0[4],zero,xmm0[5],zero,xmm0[6],zero,xmm0[7],zero +; AVX1-NEXT: vpmovzxbw {{.*#+}} 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 ; AVX1-NEXT: vpmovzxbw {{.*#+}} xmm1 = xmm1[0],zero,xmm1[1],zero,xmm1[2],zero,xmm1[3],zero,xmm1[4],zero,xmm1[5],zero,xmm1[6],zero,xmm1[7],zero -; AVX1-NEXT: vpmullw %xmm1, %xmm0, %xmm0 +; AVX1-NEXT: vpmullw %xmm1, %xmm2, %xmm1 +; AVX1-NEXT: vpand %xmm4, %xmm1, %xmm1 +; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm0 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] +; AVX1-NEXT: vpmullw %xmm0, %xmm0, %xmm0 ; AVX1-NEXT: vpand %xmm4, %xmm0, %xmm0 -; AVX1-NEXT: vpackuswb %xmm2, %xmm0, %xmm0 +; AVX1-NEXT: vpackuswb %xmm0, %xmm1, %xmm0 ; AVX1-NEXT: vpsrld $16, %xmm0, %xmm1 ; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm2 = xmm0[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] ; AVX1-NEXT: vpunpckhbw {{.*#+}} xmm3 = xmm1[8,8,9,9,10,10,11,11,12,12,13,13,14,14,15,15] |
