diff options
| author | Simon Pilgrim <llvm-dev@redking.me.uk> | 2016-05-10 16:55:20 +0000 |
|---|---|---|
| committer | Simon Pilgrim <llvm-dev@redking.me.uk> | 2016-05-10 16:55:20 +0000 |
| commit | d86b2d6b1e4efd858fc04bae7cf3884ef4f5bb03 (patch) | |
| tree | c5d233223b62d93aa29c18f7f149c70be1635b44 | |
| parent | a356bb7fa46dc3f62c1136a186dd690f0f0cff6c (diff) | |
| download | llvm-d86b2d6b1e4efd858fc04bae7cf3884ef4f5bb03.zip llvm-d86b2d6b1e4efd858fc04bae7cf3884ef4f5bb03.tar.gz llvm-d86b2d6b1e4efd858fc04bae7cf3884ef4f5bb03.tar.bz2 | |
[X86][AVX512] Added another masked shuffle combine from load test
llvm-svn: 269077
| -rw-r--r-- | llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll | 13 |
1 files changed, 13 insertions, 0 deletions
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll b/llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll index 23e1896..4070c00 100644 --- a/llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll +++ b/llvm/test/CodeGen/X86/vector-shuffle-combining-avx512bw.ll @@ -139,6 +139,19 @@ define <16 x float> @combine_vpermt2var_16f32_vmovddup_mask(<16 x float> %x0, <1 %res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 %m) ret <16 x float> %res0 } +define <16 x float> @combine_vpermt2var_16f32_vmovddup_mask_load(<16 x float> *%p0, <16 x float> %x1, i16 %m) { +; CHECK-LABEL: combine_vpermt2var_16f32_vmovddup_mask_load: +; CHECK: # BB#0: +; CHECK-NEXT: kmovw %esi, %k1 +; CHECK-NEXT: vmovaps (%rdi), %zmm1 +; CHECK-NEXT: vmovdqa32 {{.*#+}} zmm2 = [0,1,0,1,4,5,4,5,8,9,8,9,12,13,12,13] +; CHECK-NEXT: vpermt2ps %zmm0, %zmm2, %zmm1 {%k1} {z} +; CHECK-NEXT: vmovaps %zmm1, %zmm0 +; CHECK-NEXT: retq + %x0 = load <16 x float>, <16 x float> *%p0 + %res0 = call <16 x float> @llvm.x86.avx512.maskz.vpermt2var.ps.512(<16 x i32> <i32 0, i32 1, i32 0, i32 1, i32 4, i32 5, i32 4, i32 5, i32 8, i32 9, i32 8, i32 9, i32 12, i32 13, i32 12, i32 13>, <16 x float> %x0, <16 x float> %x1, i16 %m) + ret <16 x float> %res0 +} define <16 x float> @combine_vpermt2var_16f32_vmovshdup(<16 x float> %x0, <16 x float> %x1) { ; CHECK-LABEL: combine_vpermt2var_16f32_vmovshdup: |