diff options
| -rw-r--r-- | llvm/lib/Target/X86/X86ISelLowering.cpp | 3 | ||||
| -rw-r--r-- | llvm/test/CodeGen/X86/avx512-shuffle.ll | 18 |
2 files changed, 20 insertions, 1 deletions
diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp index 96f01c8b245..98eedb8ec6e 100644 --- a/llvm/lib/Target/X86/X86ISelLowering.cpp +++ b/llvm/lib/Target/X86/X86ISelLowering.cpp @@ -3872,7 +3872,8 @@ static bool isAlignrMask(ArrayRef<int> Mask, MVT VT, bool InterLane) { static bool isPALIGNRMask(ArrayRef<int> Mask, MVT VT, const X86Subtarget *Subtarget) { if ((VT.is128BitVector() && !Subtarget->hasSSSE3()) || - (VT.is256BitVector() && !Subtarget->hasInt256())) + (VT.is256BitVector() && !Subtarget->hasInt256()) || + VT.is512BitVector()) // FIXME: Add AVX512BW. return false; diff --git a/llvm/test/CodeGen/X86/avx512-shuffle.ll b/llvm/test/CodeGen/X86/avx512-shuffle.ll index bae99d97b21..964971e6f62 100644 --- a/llvm/test/CodeGen/X86/avx512-shuffle.ll +++ b/llvm/test/CodeGen/X86/avx512-shuffle.ll @@ -199,6 +199,24 @@ define <16 x i32> @test15(<16 x i32> %a) { %b = shufflevector <16 x i32> %a, <16 x i32> undef, <16 x i32><i32 1, i32 0, i32 3, i32 2, i32 5, i32 4, i32 7, i32 6, i32 9, i32 8, i32 11, i32 10, i32 13, i32 12, i32 15, i32 14> ret <16 x i32> %b } + +; CHECK-LABEL: valign_test_v16f32 +; CHECK: valignd $2, %zmm0, %zmm0 +; CHECK: ret +define <16 x float> @valign_test_v16f32(<16 x float> %a, <16 x float> %b) nounwind { + %c = shufflevector <16 x float> %a, <16 x float> %b, <16 x i32><i32 2, i32 3, i32 undef, i32 undef, i32 6, i32 7, i32 undef, i32 undef, i32 10, i32 11, i32 undef, i32 undef, i32 14, i32 15, i32 undef, i32 undef> + ret <16 x float> %c +} + +; CHECK-LABEL: valign_test_v16i32 +; CHECK: valignd $2, %zmm0, %zmm0 +; CHECK: ret +define <16 x i32> @valign_test_v16i32(<16 x i32> %a, <16 x i32> %b) nounwind { + %c = shufflevector <16 x i32> %a, <16 x i32> %b, <16 x i32><i32 2, i32 3, i32 undef, i32 undef, i32 6, i32 7, i32 undef, i32 undef, i32 10, i32 11, i32 undef, i32 undef, i32 14, i32 15, i32 undef, i32 undef> + ret <16 x i32> %c +} + + ; CHECK-LABEL: test16 ; CHECK: valignq $2, %zmm0, %zmm1 ; CHECK: ret |
