diff options
Diffstat (limited to 'test/CodeGen/X86/vec_extract-sse4.ll')
| -rw-r--r-- | test/CodeGen/X86/vec_extract-sse4.ll | 115 |
1 files changed, 67 insertions, 48 deletions
diff --git a/test/CodeGen/X86/vec_extract-sse4.ll b/test/CodeGen/X86/vec_extract-sse4.ll index 9f4210f7847ed..f073f1538d2ea 100644 --- a/test/CodeGen/X86/vec_extract-sse4.ll +++ b/test/CodeGen/X86/vec_extract-sse4.ll @@ -1,60 +1,79 @@ -; RUN: llc < %s -mcpu=corei7 -march=x86 -mattr=+sse4.1 | FileCheck %s +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py +; RUN: llc < %s -mtriple=i686-unknown -mattr=+sse4.1 | FileCheck %s --check-prefix=X32 +; RUN: llc < %s -mtriple=x86_64-unknown -mattr=+sse4.1 | FileCheck %s --check-prefix=X64 define void @t1(float* %R, <4 x float>* %P1) nounwind { -; CHECK-LABEL: t1: -; CHECK: # BB#0: -; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax -; CHECK-NEXT: movl {{[0-9]+}}(%esp), %ecx -; CHECK-NEXT: movss 12(%ecx), %xmm0 -; CHECK-NEXT: movss %xmm0, (%eax) -; CHECK-NEXT: retl - - %X = load <4 x float>, <4 x float>* %P1 - %tmp = extractelement <4 x float> %X, i32 3 - store float %tmp, float* %R - ret void +; X32-LABEL: t1: +; X32: # BB#0: +; X32-NEXT: movl {{[0-9]+}}(%esp), %eax +; X32-NEXT: movl {{[0-9]+}}(%esp), %ecx +; X32-NEXT: movss {{.*#+}} xmm0 = mem[0],zero,zero,zero +; X32-NEXT: movss %xmm0, (%eax) +; X32-NEXT: retl +; +; X64-LABEL: t1: +; X64: # BB#0: +; X64-NEXT: movss {{.*#+}} xmm0 = mem[0],zero,zero,zero +; X64-NEXT: movss %xmm0, (%rdi) +; X64-NEXT: retq + %X = load <4 x float>, <4 x float>* %P1 + %tmp = extractelement <4 x float> %X, i32 3 + store float %tmp, float* %R + ret void } define float @t2(<4 x float>* %P1) nounwind { -; CHECK-LABEL: t2: -; CHECK: # BB#0: -; CHECK-NEXT: pushl %eax -; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax -; CHECK-NEXT: movapd (%eax), %xmm0 -; CHECK-NEXT: shufpd {{.*#+}} xmm0 = xmm0[1,0] -; CHECK-NEXT: movss %xmm0, (%esp) -; CHECK-NEXT: flds (%esp) -; CHECK-NEXT: popl %eax -; CHECK-NEXT: retl - - %X = load <4 x float>, <4 x float>* %P1 - %tmp = extractelement <4 x float> %X, i32 2 - ret float %tmp +; X32-LABEL: t2: +; X32: # BB#0: +; X32-NEXT: pushl %eax +; X32-NEXT: movl {{[0-9]+}}(%esp), %eax +; X32-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] +; X32-NEXT: movss %xmm0, (%esp) +; X32-NEXT: flds (%esp) +; X32-NEXT: popl %eax +; X32-NEXT: retl +; +; X64-LABEL: t2: +; X64: # BB#0: +; X64-NEXT: movddup {{.*#+}} xmm0 = mem[0,0] +; X64-NEXT: retq + %X = load <4 x float>, <4 x float>* %P1 + %tmp = extractelement <4 x float> %X, i32 2 + ret float %tmp } define void @t3(i32* %R, <4 x i32>* %P1) nounwind { -; CHECK-LABEL: t3: -; CHECK: # BB#0: -; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax -; CHECK-NEXT: movl {{[0-9]+}}(%esp), %ecx -; CHECK-NEXT: movl 12(%ecx), %ecx -; CHECK-NEXT: movl %ecx, (%eax) -; CHECK-NEXT: retl - - %X = load <4 x i32>, <4 x i32>* %P1 - %tmp = extractelement <4 x i32> %X, i32 3 - store i32 %tmp, i32* %R - ret void +; X32-LABEL: t3: +; X32: # BB#0: +; X32-NEXT: movl {{[0-9]+}}(%esp), %eax +; X32-NEXT: movl {{[0-9]+}}(%esp), %ecx +; X32-NEXT: movl 12(%ecx), %ecx +; X32-NEXT: movl %ecx, (%eax) +; X32-NEXT: retl +; +; X64-LABEL: t3: +; X64: # BB#0: +; X64-NEXT: movl 12(%rsi), %eax +; X64-NEXT: movl %eax, (%rdi) +; X64-NEXT: retq + %X = load <4 x i32>, <4 x i32>* %P1 + %tmp = extractelement <4 x i32> %X, i32 3 + store i32 %tmp, i32* %R + ret void } define i32 @t4(<4 x i32>* %P1) nounwind { -; CHECK-LABEL: t4: -; CHECK: # BB#0: -; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax -; CHECK-NEXT: movl 12(%eax), %eax -; CHECK-NEXT: retl - - %X = load <4 x i32>, <4 x i32>* %P1 - %tmp = extractelement <4 x i32> %X, i32 3 - ret i32 %tmp +; X32-LABEL: t4: +; X32: # BB#0: +; X32-NEXT: movl {{[0-9]+}}(%esp), %eax +; X32-NEXT: movl 12(%eax), %eax +; X32-NEXT: retl +; +; X64-LABEL: t4: +; X64: # BB#0: +; X64-NEXT: movl 12(%rdi), %eax +; X64-NEXT: retq + %X = load <4 x i32>, <4 x i32>* %P1 + %tmp = extractelement <4 x i32> %X, i32 3 + ret i32 %tmp } |