diff options
Diffstat (limited to 'test/CodeGen/X86/sandybridge-loads.ll')
-rw-r--r-- | test/CodeGen/X86/sandybridge-loads.ll | 42 |
1 files changed, 24 insertions, 18 deletions
diff --git a/test/CodeGen/X86/sandybridge-loads.ll b/test/CodeGen/X86/sandybridge-loads.ll index 2e31154068fc..8570fe7fe7ba 100644 --- a/test/CodeGen/X86/sandybridge-loads.ll +++ b/test/CodeGen/X86/sandybridge-loads.ll @@ -1,13 +1,20 @@ -; RUN: llc -march=x86-64 -mtriple=x86_64-apple-darwin -mcpu=corei7-avx -o - < %s | FileCheck %s - -;CHECK-LABEL: wideloads: -;CHECK: vmovaps -;CHECK: vinsertf128 -;CHECK: vmovaps -;CHECK-NOT: vinsertf128 -;CHECK: ret +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py +; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mcpu=corei7-avx | FileCheck %s define void @wideloads(<8 x float>* %a, <8 x float>* %b, <8 x float>* %c) nounwind uwtable noinline ssp { +; CHECK-LABEL: wideloads: +; CHECK: # BB#0: +; CHECK-NEXT: vmovaps (%rdi), %xmm0 +; CHECK-NEXT: vinsertf128 $1, 16(%rdi), %ymm0, %ymm0 +; CHECK-NEXT: vmovaps (%rsi), %ymm1 +; CHECK-NEXT: vcmpltps %ymm0, %ymm1, %ymm1 +; CHECK-NEXT: vmovaps (%rdx), %ymm2 +; CHECK-NEXT: vcmpltps %ymm0, %ymm2, %ymm0 +; CHECK-NEXT: vandps %ymm1, %ymm0, %ymm0 +; CHECK-NEXT: vandps {{.*}}(%rip), %ymm0, %ymm0 +; CHECK-NEXT: vmovaps %ymm0, (%rax) +; CHECK-NEXT: vzeroupper +; CHECK-NEXT: retq %v0 = load <8 x float>, <8 x float>* %a, align 16 ; <---- unaligned! %v1 = load <8 x float>, <8 x float>* %b, align 32 ; <---- aligned! %m0 = fcmp olt <8 x float> %v1, %v0 @@ -19,17 +26,16 @@ define void @wideloads(<8 x float>* %a, <8 x float>* %b, <8 x float>* %c) nounwi ret void } -; CHECK: widestores -; loads: -; CHECK: vmovaps -; CHECK: vmovaps -; stores: -; CHECK: vmovaps -; CHECK: vextractf128 -; CHECK: vmovaps -;CHECK: ret - define void @widestores(<8 x float>* %a, <8 x float>* %b, <8 x float>* %c) nounwind uwtable noinline ssp { +; CHECK-LABEL: widestores: +; CHECK: # BB#0: +; CHECK-NEXT: vmovaps (%rdi), %ymm0 +; CHECK-NEXT: vmovaps (%rsi), %ymm1 +; CHECK-NEXT: vmovaps %ymm0, (%rsi) +; CHECK-NEXT: vextractf128 $1, %ymm1, 16(%rdi) +; CHECK-NEXT: vmovaps %xmm1, (%rdi) +; CHECK-NEXT: vzeroupper +; CHECK-NEXT: retq %v0 = load <8 x float>, <8 x float>* %a, align 32 %v1 = load <8 x float>, <8 x float>* %b, align 32 store <8 x float> %v0, <8 x float>* %b, align 32 ; <--- aligned |