1 ; RUN: llc < %s -mtriple=x86_64-linux | FileCheck %s
2 ; RUN: llc < %s -mtriple=x86_64-win32 | FileCheck %s
3 ; RUN: llc -O0 < %s -march=x86 -mcpu=core2 | FileCheck %s --check-prefix=CHECK_O0
5 define <4 x i32> @t00(<4 x i32>* %a0) nounwind ssp {
7 ; CHECK: movaps ({{%rdi|%rcx}}), %xmm0
8 ; CHECK: movaps %xmm0, %xmm1
9 ; CHECK-NEXT: movss %xmm2, %xmm1
10 ; CHECK-NEXT: shufps $36, %xmm1, %xmm0
11 %0 = load <4 x i32>* undef, align 16
12 %1 = load <4 x i32>* %a0, align 16
13 %2 = shufflevector <4 x i32> %1, <4 x i32> %0, <4 x i32> <i32 0, i32 1, i32 2, i32 4>
17 define void @t01(double* %a0) nounwind ssp {
19 ; CHECK_O0: movsd (%eax), %xmm0
20 ; CHECK_O0: unpcklpd %xmm0, %xmm0
21 %tmp93 = load double* %a0, align 8
22 %vecinit94 = insertelement <2 x double> undef, double %tmp93, i32 1
23 store <2 x double> %vecinit94, <2 x double>* undef
27 define void @t02(<8 x i32>* %source, <2 x i32>* %dest) nounwind noinline {
29 ; CHECK: movl 36({{%rdi|%rcx}})
30 ; CHECK-NEXT: movl 48({{%rdi|%rcx}})
32 ; CHECK: movq %xmm0, ({{%rsi|%rdx}})
33 %0 = bitcast <8 x i32>* %source to <4 x i32>*
34 %arrayidx = getelementptr inbounds <4 x i32>* %0, i64 3
35 %tmp2 = load <4 x i32>* %arrayidx, align 16
36 %tmp3 = extractelement <4 x i32> %tmp2, i32 0
37 %tmp5 = insertelement <2 x i32> <i32 undef, i32 0>, i32 %tmp3, i32 0
38 %arrayidx7 = getelementptr inbounds <8 x i32>* %source, i64 1
39 %1 = bitcast <8 x i32>* %arrayidx7 to <4 x i32>*
40 %tmp8 = load <4 x i32>* %1, align 16
41 %tmp9 = extractelement <4 x i32> %tmp8, i32 1
42 %tmp11 = insertelement <2 x i32> %tmp5, i32 %tmp9, i32 1
43 store <2 x i32> %tmp11, <2 x i32>* %dest, align 8