diff options
Diffstat (limited to 'llvm/test/CodeGen/X86/vector-shuffle-combining-ssse3.ll')
| -rw-r--r-- | llvm/test/CodeGen/X86/vector-shuffle-combining-ssse3.ll | 41 |
1 files changed, 29 insertions, 12 deletions
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-combining-ssse3.ll b/llvm/test/CodeGen/X86/vector-shuffle-combining-ssse3.ll index c286fd426b1..16b2efe88a4 100644 --- a/llvm/test/CodeGen/X86/vector-shuffle-combining-ssse3.ll +++ b/llvm/test/CodeGen/X86/vector-shuffle-combining-ssse3.ll @@ -49,16 +49,23 @@ define <2 x double> @combine_pshufb_as_movsd(<2 x double> %a0, <2 x double> %a1) ; ; SSE41-LABEL: combine_pshufb_as_movsd: ; SSE41: # BB#0: -; SSE41-NEXT: shufpd {{.*#+}} xmm1 = xmm1[1],xmm0[0] -; SSE41-NEXT: pshufb {{.*#+}} xmm1 = xmm1[8,9,10,11,12,13,14,15,0,1,2,3,4,5,6,7] -; SSE41-NEXT: movdqa %xmm1, %xmm0 +; SSE41-NEXT: blendpd {{.*#+}} xmm0 = xmm0[0],xmm1[1] ; SSE41-NEXT: retq ; -; AVX-LABEL: combine_pshufb_as_movsd: -; AVX: # BB#0: -; AVX-NEXT: vshufpd {{.*#+}} xmm0 = xmm1[1],xmm0[0] -; AVX-NEXT: vpermilpd {{.*#+}} xmm0 = xmm0[1,0] -; AVX-NEXT: retq +; AVX1-LABEL: combine_pshufb_as_movsd: +; AVX1: # BB#0: +; AVX1-NEXT: vblendpd {{.*#+}} xmm0 = xmm0[0],xmm1[1] +; AVX1-NEXT: retq +; +; AVX2-LABEL: combine_pshufb_as_movsd: +; AVX2: # BB#0: +; AVX2-NEXT: vblendpd {{.*#+}} xmm0 = xmm0[0],xmm1[1] +; AVX2-NEXT: retq +; +; AVX512F-LABEL: combine_pshufb_as_movsd: +; AVX512F: # BB#0: +; AVX512F-NEXT: vmovsd {{.*#+}} xmm0 = xmm0[0],xmm1[1] +; AVX512F-NEXT: retq %1 = shufflevector <2 x double> %a0, <2 x double> %a1, <2 x i32> <i32 3, i32 0> %2 = bitcast <2 x double> %1 to <16 x i8> %3 = tail call <16 x i8> @llvm.x86.ssse3.pshuf.b.128(<16 x i8> %2, <16 x i8> <i8 8, i8 9, i8 10, i8 11, i8 12, i8 13, i8 14, i8 15, i8 0, i8 1, i8 2, i8 3, i8 4, i8 5, i8 6, i8 7>) @@ -77,10 +84,20 @@ define <4 x float> @combine_pshufb_as_movss(<4 x float> %a0, <4 x float> %a1) { ; SSE41-NEXT: blendps {{.*#+}} xmm0 = xmm1[0],xmm0[1,2,3] ; SSE41-NEXT: retq ; -; AVX-LABEL: combine_pshufb_as_movss: -; AVX: # BB#0: -; AVX-NEXT: vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1,2,3] -; AVX-NEXT: retq +; AVX1-LABEL: combine_pshufb_as_movss: +; AVX1: # BB#0: +; AVX1-NEXT: vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1,2,3] +; AVX1-NEXT: retq +; +; AVX2-LABEL: combine_pshufb_as_movss: +; AVX2: # BB#0: +; AVX2-NEXT: vblendps {{.*#+}} xmm0 = xmm1[0],xmm0[1,2,3] +; AVX2-NEXT: retq +; +; AVX512F-LABEL: combine_pshufb_as_movss: +; AVX512F: # BB#0: +; AVX512F-NEXT: vmovss {{.*#+}} xmm0 = xmm1[0],xmm0[1,2,3] +; AVX512F-NEXT: retq %1 = shufflevector <4 x float> %a0, <4 x float> %a1, <4 x i32> <i32 4, i32 3, i32 2, i32 1> %2 = bitcast <4 x float> %1 to <16 x i8> %3 = tail call <16 x i8> @llvm.x86.ssse3.pshuf.b.128(<16 x i8> %2, <16 x i8> <i8 0, i8 1, i8 2, i8 3, i8 12, i8 13, i8 14, i8 15, i8 8, i8 9, i8 10, i8 11, i8 4, i8 5, i8 6, i8 7>) |
