1 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=corei7-avx -mattr=+avx | FileCheck %s
2 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=corei7-avx -mattr=+avx | FileCheck -check-prefix=CHECK-SSE %s
5 ; CHECK: vinsertf128 $1
6 define <8 x float> @A(<8 x float> %a) nounwind uwtable readnone ssp {
8 %shuffle = shufflevector <8 x float> %a, <8 x float> undef, <8 x i32> <i32 8, i32 8, i32 8, i32 8, i32 0, i32 1, i32 2, i32 3>
9 ret <8 x float> %shuffle
13 ; CHECK: vinsertf128 $1
14 define <4 x double> @B(<4 x double> %a) nounwind uwtable readnone ssp {
16 %shuffle = shufflevector <4 x double> %a, <4 x double> undef, <4 x i32> <i32 4, i32 4, i32 0, i32 1>
17 ret <4 x double> %shuffle
20 declare <2 x double> @llvm.x86.sse2.min.pd(<2 x double>, <2 x double>) nounwind readnone
22 declare <2 x double> @llvm.x86.sse2.min.sd(<2 x double>, <2 x double>) nounwind readnone
24 ; Just check that no crash happens
25 ; CHECK-SSE: _insert_crash
26 define void @insert_crash() nounwind {
28 %v1.i.i451 = shufflevector <4 x double> zeroinitializer, <4 x double> undef, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
29 %ret_0a.i.i.i452 = shufflevector <4 x double> %v1.i.i451, <4 x double> undef, <2 x i32> <i32 0, i32 1>
30 %vret_0.i.i.i454 = tail call <2 x double> @llvm.x86.sse2.min.pd(<2 x double> %ret_0a.i.i.i452, <2 x double> undef) nounwind
31 %ret_val.i.i.i463 = tail call <2 x double> @llvm.x86.sse2.min.sd(<2 x double> %vret_0.i.i.i454, <2 x double> undef) nounwind
32 %ret.i1.i.i464 = extractelement <2 x double> %ret_val.i.i.i463, i32 0
33 %double2float = fptrunc double %ret.i1.i.i464 to float
34 %smearinsert50 = insertelement <4 x float> undef, float %double2float, i32 3
35 %blendAsInt.i503 = bitcast <4 x float> %smearinsert50 to <4 x i32>
36 store <4 x i32> %blendAsInt.i503, <4 x i32>* undef, align 4
40 ;; DAG Combine must remove useless vinsertf128 instructions
43 ; CHECK-NOT: vinsertf128 $1
44 define <4 x i32> @DAGCombineA(<4 x i32> %v1) nounwind readonly {
45 %1 = shufflevector <4 x i32> %v1, <4 x i32> undef, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7>
46 %2 = shufflevector <8 x i32> %1, <8 x i32> undef, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
52 ; CHECK-NOT: vinsertf128 $1
54 define <8 x i32> @DAGCombineB(<8 x i32> %v1, <8 x i32> %v2) nounwind readonly {
55 %1 = add <8 x i32> %v1, %v2
56 %2 = add <8 x i32> %1, %v1
61 define <4 x double> @insert_pd(<4 x double> %a0, <2 x double> %a1) {
63 %res = call <4 x double> @llvm.x86.avx.vinsertf128.pd.256(<4 x double> %a0, <2 x double> %a1, i8 0)
67 ; CHECK: insert_undef_pd
68 define <4 x double> @insert_undef_pd(<4 x double> %a0, <2 x double> %a1) {
69 ; CHECK: vmovaps %ymm1, %ymm0
70 %res = call <4 x double> @llvm.x86.avx.vinsertf128.pd.256(<4 x double> undef, <2 x double> %a1, i8 0)
73 declare <4 x double> @llvm.x86.avx.vinsertf128.pd.256(<4 x double>, <2 x double>, i8) nounwind readnone
77 define <8 x float> @insert_ps(<8 x float> %a0, <4 x float> %a1) {
79 %res = call <8 x float> @llvm.x86.avx.vinsertf128.ps.256(<8 x float> %a0, <4 x float> %a1, i8 0)
83 ; CHECK: insert_undef_ps
84 define <8 x float> @insert_undef_ps(<8 x float> %a0, <4 x float> %a1) {
85 ; CHECK: vmovaps %ymm1, %ymm0
86 %res = call <8 x float> @llvm.x86.avx.vinsertf128.ps.256(<8 x float> undef, <4 x float> %a1, i8 0)
89 declare <8 x float> @llvm.x86.avx.vinsertf128.ps.256(<8 x float>, <4 x float>, i8) nounwind readnone
93 define <8 x i32> @insert_si(<8 x i32> %a0, <4 x i32> %a1) {
95 %res = call <8 x i32> @llvm.x86.avx.vinsertf128.si.256(<8 x i32> %a0, <4 x i32> %a1, i8 0)
99 ; CHECK: insert_undef_si
100 define <8 x i32> @insert_undef_si(<8 x i32> %a0, <4 x i32> %a1) {
101 ; CHECK: vmovaps %ymm1, %ymm0
102 %res = call <8 x i32> @llvm.x86.avx.vinsertf128.si.256(<8 x i32> undef, <4 x i32> %a1, i8 0)
105 declare <8 x i32> @llvm.x86.avx.vinsertf128.si.256(<8 x i32>, <4 x i32>, i8) nounwind readnone
108 ; CHECK: vinsertf128_combine
109 define <8 x float> @vinsertf128_combine(float* nocapture %f) nounwind uwtable readonly ssp {
113 %add.ptr = getelementptr inbounds float* %f, i64 4
114 %0 = bitcast float* %add.ptr to <4 x float>*
115 %1 = load <4 x float>* %0, align 16
116 %2 = tail call <8 x float> @llvm.x86.avx.vinsertf128.ps.256(<8 x float> undef, <4 x float> %1, i8 1)
121 ; CHECK: vinsertf128_ucombine
122 define <8 x float> @vinsertf128_ucombine(float* nocapture %f) nounwind uwtable readonly ssp {
126 %add.ptr = getelementptr inbounds float* %f, i64 4
127 %0 = bitcast float* %add.ptr to <4 x float>*
128 %1 = load <4 x float>* %0, align 8
129 %2 = tail call <8 x float> @llvm.x86.avx.vinsertf128.ps.256(<8 x float> undef, <4 x float> %1, i8 1)