1 ; RUN: llc < %s -mattr=+sse2 -mtriple=i686-apple-darwin -mcpu=core2 | FileCheck %s -check-prefix=SSE2
2 ; RUN: llc < %s -mattr=+sse2 -mtriple=i686-pc-mingw32 -mcpu=core2 | FileCheck %s -check-prefix=SSE2
3 ; RUN: llc < %s -mattr=+sse,-sse2 -mtriple=i686-apple-darwin -mcpu=core2 | FileCheck %s -check-prefix=SSE1
4 ; RUN: llc < %s -mattr=-sse -mtriple=i686-apple-darwin -mcpu=core2 | FileCheck %s -check-prefix=NOSSE
5 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=core2 | FileCheck %s -check-prefix=X86-64
7 @.str = internal constant [25 x i8] c"image\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00\00"
8 @.str2 = internal constant [30 x i8] c"xxxxxxxxxxxxxxxxxxxxxxxxxxxxx\00", align 4
10 define void @t1(i32 %argc, i8** %argv) nounwind {
13 ; SSE2: movsd _.str+16, %xmm0
14 ; SSE2: movsd %xmm0, 16(%esp)
15 ; SSE2: movaps _.str, %xmm0
17 ; SSE2: movb $0, 24(%esp)
20 ; SSE1: movaps _.str, %xmm0
22 ; SSE1: movb $0, 24(%esp)
23 ; SSE1: movl $0, 20(%esp)
24 ; SSE1: movl $0, 16(%esp)
33 ; NOSSE: movl $1734438249
36 ; X86-64: movaps _.str(%rip), %xmm0
37 ; X86-64: movaps %xmm0
40 %tmp1 = alloca [25 x i8]
41 %tmp2 = bitcast [25 x i8]* %tmp1 to i8*
42 call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* getelementptr inbounds ([25 x i8]* @.str, i32 0, i32 0), i32 25, i32 1, i1 false)
47 %struct.s0 = type { [2 x double] }
49 define void @t2(%struct.s0* nocapture %a, %struct.s0* nocapture %b) nounwind ssp {
52 ; SSE2: movaps (%eax), %xmm0
53 ; SSE2: movaps %xmm0, (%eax)
56 ; SSE1: movaps (%eax), %xmm0
57 ; SSE1: movaps %xmm0, (%eax)
72 ; X86-64: movaps (%rsi), %xmm0
73 ; X86-64: movaps %xmm0, (%rdi)
74 %tmp2 = bitcast %struct.s0* %a to i8* ; <i8*> [#uses=1]
75 %tmp3 = bitcast %struct.s0* %b to i8* ; <i8*> [#uses=1]
76 tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* %tmp3, i32 16, i32 16, i1 false)
80 define void @t3(%struct.s0* nocapture %a, %struct.s0* nocapture %b) nounwind ssp {
83 ; SSE2: movsd (%eax), %xmm0
84 ; SSE2: movsd 8(%eax), %xmm1
85 ; SSE2: movsd %xmm1, 8(%eax)
86 ; SSE2: movsd %xmm0, (%eax)
113 ; X86-64: movq (%rsi), %rax
114 ; X86-64: movq 8(%rsi), %rcx
115 ; X86-64: movq %rcx, 8(%rdi)
116 ; X86-64: movq %rax, (%rdi)
117 %tmp2 = bitcast %struct.s0* %a to i8* ; <i8*> [#uses=1]
118 %tmp3 = bitcast %struct.s0* %b to i8* ; <i8*> [#uses=1]
119 tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* %tmp3, i32 16, i32 8, i1 false)
123 define void @t4() nounwind {
127 ; SSE2: movl $2021161080
128 ; SSE2: movl $2021161080
129 ; SSE2: movl $2021161080
130 ; SSE2: movl $2021161080
131 ; SSE2: movl $2021161080
132 ; SSE2: movl $2021161080
133 ; SSE2: movl $2021161080
137 ; SSE1: movl $2021161080
138 ; SSE1: movl $2021161080
139 ; SSE1: movl $2021161080
140 ; SSE1: movl $2021161080
141 ; SSE1: movl $2021161080
142 ; SSE1: movl $2021161080
143 ; SSE1: movl $2021161080
147 ; NOSSE: movl $2021161080
148 ; NOSSE: movl $2021161080
149 ; NOSSE: movl $2021161080
150 ; NOSSE: movl $2021161080
151 ; NOSSE: movl $2021161080
152 ; NOSSE: movl $2021161080
153 ; NOSSE: movl $2021161080
156 ; X86-64: movabsq $8680820740569200760, %rax
161 ; X86-64: movl $2021161080
162 %tmp1 = alloca [30 x i8]
163 %tmp2 = bitcast [30 x i8]* %tmp1 to i8*
164 call void @llvm.memcpy.p0i8.p0i8.i32(i8* %tmp2, i8* getelementptr inbounds ([30 x i8]* @.str2, i32 0, i32 0), i32 30, i32 1, i1 false)
168 declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind