X-Git-Url: http://plrg.eecs.uci.edu/git/?a=blobdiff_plain;f=test%2FCodeGen%2FX86%2Fvec_extract.ll;h=3b478880590d21db913da58b8f1eaf5d9e50e100;hb=43421abda8b7134f409ccbdc4be130a5ba9ccfab;hp=88f5a585b9fd83926960a9adabcf0ae2743edbb3;hpb=4aa8bdaa46b455d446c10325533dcbdc65eeecdd;p=oota-llvm.git diff --git a/test/CodeGen/X86/vec_extract.ll b/test/CodeGen/X86/vec_extract.ll index 88f5a585b9f..3b478880590 100644 --- a/test/CodeGen/X86/vec_extract.ll +++ b/test/CodeGen/X86/vec_extract.ll @@ -1,11 +1,18 @@ -; RUN: llc < %s -mcpu=corei7 -march=x86 -mattr=+sse2,-sse4.1 -o %t -; RUN: grep movss %t | count 4 -; RUN: grep movhlps %t | count 1 -; RUN: not grep pshufd %t -; RUN: grep unpckhpd %t | count 1 +; RUN: llc < %s -mcpu=corei7 -march=x86 -mattr=+sse2,-sse4.1 | FileCheck %s + +target triple = "x86_64-unknown-linux-gnu" define void @test1(<4 x float>* %F, float* %f) nounwind { - %tmp = load <4 x float>* %F ; <<4 x float>> [#uses=2] +; CHECK-LABEL: test1: +; CHECK: # BB#0: # %entry +; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax +; CHECK-NEXT: movl {{[0-9]+}}(%esp), %ecx +; CHECK-NEXT: movaps (%ecx), %xmm0 +; CHECK-NEXT: addps %xmm0, %xmm0 +; CHECK-NEXT: movss %xmm0, (%eax) +; CHECK-NEXT: retl +entry: + %tmp = load <4 x float>, <4 x float>* %F ; <<4 x float>> [#uses=2] %tmp7 = fadd <4 x float> %tmp, %tmp ; <<4 x float>> [#uses=1] %tmp2 = extractelement <4 x float> %tmp7, i32 0 ; [#uses=1] store float %tmp2, float* %f @@ -13,20 +20,51 @@ define void @test1(<4 x float>* %F, float* %f) nounwind { } define float @test2(<4 x float>* %F, float* %f) nounwind { - %tmp = load <4 x float>* %F ; <<4 x float>> [#uses=2] +; CHECK-LABEL: test2: +; CHECK: # BB#0: # %entry +; CHECK-NEXT: pushl %eax +; CHECK-NEXT: movl {{[0-9]+}}(%esp), %eax +; CHECK-NEXT: movaps (%eax), %xmm0 +; CHECK-NEXT: addps %xmm0, %xmm0 +; CHECK-NEXT: shufpd {{.*#+}} xmm0 = xmm0[1,0] +; CHECK-NEXT: movss %xmm0, (%esp) +; CHECK-NEXT: flds (%esp) +; CHECK-NEXT: popl %eax +; CHECK-NEXT: retl +entry: + %tmp = load <4 x float>, <4 x float>* %F ; <<4 x float>> [#uses=2] %tmp7 = fadd <4 x float> %tmp, %tmp ; <<4 x float>> [#uses=1] %tmp2 = extractelement <4 x float> %tmp7, i32 2 ; [#uses=1] ret float %tmp2 } define void @test3(float* %R, <4 x float>* %P1) nounwind { - %X = load <4 x float>* %P1 ; <<4 x float>> [#uses=1] +; CHECK-LABEL: test3: +; CHECK: # BB#0: # %entry +; 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 +entry: + %X = load <4 x float>, <4 x float>* %P1 ; <<4 x float>> [#uses=1] %tmp = extractelement <4 x float> %X, i32 3 ; [#uses=1] store float %tmp, float* %R ret void } define double @test4(double %A) nounwind { +; CHECK-LABEL: test4: +; CHECK: # BB#0: # %entry +; CHECK-NEXT: subl $12, %esp +; CHECK-NEXT: calll foo +; CHECK-NEXT: shufpd {{.*#+}} xmm0 = xmm0[1,0] +; CHECK-NEXT: addsd {{[0-9]+}}(%esp), %xmm0 +; CHECK-NEXT: movsd %xmm0, (%esp) +; CHECK-NEXT: fldl (%esp) +; CHECK-NEXT: addl $12, %esp +; CHECK-NEXT: retl +entry: %tmp1 = call <2 x double> @foo( ) ; <<2 x double>> [#uses=1] %tmp2 = extractelement <2 x double> %tmp1, i32 1 ; [#uses=1] %tmp3 = fadd double %tmp2, %A ; [#uses=1]