diff options
Diffstat (limited to 'llvm/test/CodeGen')
| -rw-r--r-- | llvm/test/CodeGen/X86/avx2-intrinsics-x86.ll | 10 | ||||
| -rw-r--r-- | llvm/test/CodeGen/X86/stack-folding-int-avx2.ll | 4 |
2 files changed, 10 insertions, 4 deletions
diff --git a/llvm/test/CodeGen/X86/avx2-intrinsics-x86.ll b/llvm/test/CodeGen/X86/avx2-intrinsics-x86.ll index b7e2b7412ff..7b2d3585df9 100644 --- a/llvm/test/CodeGen/X86/avx2-intrinsics-x86.ll +++ b/llvm/test/CodeGen/X86/avx2-intrinsics-x86.ll @@ -746,7 +746,10 @@ declare <4 x i64> @llvm.x86.avx2.pbroadcastq.256(<2 x i64>) nounwind readonly define <8 x i32> @test_x86_avx2_permd(<8 x i32> %a0, <8 x i32> %a1) { - ; CHECK: vpermd + ; Check that the arguments are swapped between the intrinsic definition + ; and its lowering. Indeed, the offsets are the first source in + ; the instruction. + ; CHECK: vpermd %ymm0, %ymm1, %ymm0 %res = call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a0, <8 x i32> %a1) ; <<8 x i32>> [#uses=1] ret <8 x i32> %res } @@ -754,7 +757,10 @@ declare <8 x i32> @llvm.x86.avx2.permd(<8 x i32>, <8 x i32>) nounwind readonly define <8 x float> @test_x86_avx2_permps(<8 x float> %a0, <8 x float> %a1) { - ; CHECK: vpermps + ; Check that the arguments are swapped between the intrinsic definition + ; and its lowering. Indeed, the offsets are the first source in + ; the instruction. + ; CHECK: vpermps %ymm0, %ymm1, %ymm0 %res = call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a0, <8 x float> %a1) ; <<8 x float>> [#uses=1] ret <8 x float> %res } diff --git a/llvm/test/CodeGen/X86/stack-folding-int-avx2.ll b/llvm/test/CodeGen/X86/stack-folding-int-avx2.ll index fa81bf104b4..7f384940d85 100644 --- a/llvm/test/CodeGen/X86/stack-folding-int-avx2.ll +++ b/llvm/test/CodeGen/X86/stack-folding-int-avx2.ll @@ -442,7 +442,7 @@ define <8 x i32> @stack_fold_permd(<8 x i32> %a0, <8 x i32> %a1) { ;CHECK-LABEL: stack_fold_permd ;CHECK: vpermd {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"() - %2 = call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a0, <8 x i32> %a1) + %2 = call <8 x i32> @llvm.x86.avx2.permd(<8 x i32> %a1, <8 x i32> %a0) ret <8 x i32> %2 } declare <8 x i32> @llvm.x86.avx2.permd(<8 x i32>, <8 x i32>) nounwind readonly @@ -461,7 +461,7 @@ define <8 x float> @stack_fold_permps(<8 x float> %a0, <8 x float> %a1) { ;CHECK-LABEL: stack_fold_permps ;CHECK: vpermps {{-?[0-9]*}}(%rsp), {{%ymm[0-9][0-9]*}}, {{%ymm[0-9][0-9]*}} {{.*#+}} 32-byte Folded Reload %1 = tail call <2 x i64> asm sideeffect "nop", "=x,~{xmm2},~{xmm3},~{xmm4},~{xmm5},~{xmm6},~{xmm7},~{xmm8},~{xmm9},~{xmm10},~{xmm11},~{xmm12},~{xmm13},~{xmm14},~{xmm15},~{flags}"() - %2 = call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a0, <8 x float> %a1) + %2 = call <8 x float> @llvm.x86.avx2.permps(<8 x float> %a1, <8 x float> %a0) ret <8 x float> %2 } declare <8 x float> @llvm.x86.avx2.permps(<8 x float>, <8 x float>) nounwind readonly |
