• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1; RUN: llc -O3 -disable-peephole -mcpu=corei7-avx -mattr=+avx < %s | FileCheck %s
2
3target datalayout = "e-i64:64-f80:128-n8:16:32:64-S128"
4target triple = "x86_64-apple-macosx10.9.0"
5
6; Function Attrs: nounwind ssp uwtable
7define void @test1(float* %A, float* %C) #0 {
8  %tmp1 = bitcast float* %A to <8 x float>*
9  %tmp2 = load <8 x float>, <8 x float>* %tmp1, align 32
10  %tmp3 = bitcast <8 x float> %tmp2 to <8 x i32>
11  %tmp4 = and <8 x i32> %tmp3, <i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647>
12  %tmp5 = bitcast <8 x i32> %tmp4 to <8 x float>
13  %tmp6 = extractelement <8 x float> %tmp5, i32 0
14  store float %tmp6, float* %C
15  ret void
16
17  ; CHECK: vandps LCPI0_0(%rip), %ymm0, %ymm0
18}
19
20; Function Attrs: nounwind ssp uwtable
21define void @test2(float* %A, float* %C) #0 {
22  %tmp1 = bitcast float* %A to <8 x float>*
23  %tmp2 = load <8 x float>, <8 x float>* %tmp1, align 32
24  %tmp3 = bitcast <8 x float> %tmp2 to <8 x i32>
25  %tmp4 = or <8 x i32> %tmp3, <i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647>
26  %tmp5 = bitcast <8 x i32> %tmp4 to <8 x float>
27  %tmp6 = extractelement <8 x float> %tmp5, i32 0
28  store float %tmp6, float* %C
29  ret void
30
31  ; CHECK: vorps LCPI1_0(%rip), %ymm0, %ymm0
32}
33
34; Function Attrs: nounwind ssp uwtable
35define void @test3(float* %A, float* %C) #0 {
36  %tmp1 = bitcast float* %A to <8 x float>*
37  %tmp2 = load <8 x float>, <8 x float>* %tmp1, align 32
38  %tmp3 = bitcast <8 x float> %tmp2 to <8 x i32>
39  %tmp4 = xor <8 x i32> %tmp3, <i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647>
40  %tmp5 = bitcast <8 x i32> %tmp4 to <8 x float>
41  %tmp6 = extractelement <8 x float> %tmp5, i32 0
42  store float %tmp6, float* %C
43  ret void
44
45  ; CHECK: vxorps LCPI2_0(%rip), %ymm0, %ymm0
46}
47
48define void @test4(float* %A, float* %C) #0 {
49  %tmp1 = bitcast float* %A to <8 x float>*
50  %tmp2 = load <8 x float>, <8 x float>* %tmp1, align 32
51  %tmp3 = bitcast <8 x float> %tmp2 to <8 x i32>
52  %tmp4 = xor <8 x i32> %tmp3, <i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1, i32 -1>
53  %tmp5 = and <8 x i32> %tmp4, <i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647>
54  %tmp6 = bitcast <8 x i32> %tmp5 to <8 x float>
55  %tmp7 = extractelement <8 x float> %tmp6, i32 0
56  store float %tmp7, float * %C
57  ret void
58
59  ;CHECK: vandnps LCPI3_0(%rip), %ymm0, %ymm0
60}
61