1 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
2 ; RUN: llc < %s -mtriple=x86_64-apple-darwin -mcpu=core-avx2 | FileCheck %s
4 ; rdar://11314175: SD Scheduler, BuildSchedUnits assert:
5 ; N->getNodeId() == -1 && "Node already inserted!
7 define void @func(<4 x float> %a, <16 x i8> %b, <16 x i8> %c, <8 x float> %d, <8 x float> %e, ptr %f) nounwind ssp {
10 ; CHECK-NEXT: vmovdqu 0, %xmm0
11 ; CHECK-NEXT: vpalignr {{.*#+}} xmm1 = xmm0[4,5,6,7,8,9,10,11,12,13,14,15],xmm1[0,1,2,3]
12 ; CHECK-NEXT: vmulps %xmm1, %xmm1, %xmm1
13 ; CHECK-NEXT: vmulps %xmm0, %xmm0, %xmm0
14 ; CHECK-NEXT: vaddps %xmm1, %xmm0, %xmm0
15 ; CHECK-NEXT: vaddps %xmm0, %xmm0, %xmm0
16 ; CHECK-NEXT: vmulps %xmm0, %xmm0, %xmm0
17 ; CHECK-NEXT: vxorps %xmm1, %xmm1, %xmm1
18 ; CHECK-NEXT: vinsertf128 $1, %xmm0, %ymm1, %ymm0
19 ; CHECK-NEXT: vaddps %ymm0, %ymm0, %ymm0
20 ; CHECK-NEXT: vhaddps %ymm4, %ymm0, %ymm0
21 ; CHECK-NEXT: vsubps %ymm0, %ymm0, %ymm0
22 ; CHECK-NEXT: vhaddps %ymm0, %ymm1, %ymm0
23 ; CHECK-NEXT: vmovaps %ymm0, (%rdi)
24 ; CHECK-NEXT: vzeroupper
26 %tmp = load <4 x float>, ptr null, align 1
27 %tmp14 = getelementptr <4 x float>, ptr null, i32 2
28 %tmp15 = load <4 x float>, ptr %tmp14, align 1
29 %tmp16 = shufflevector <4 x float> %tmp, <4 x float> <float 0.000000e+00, float undef, float undef, float undef>, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 4, i32 4, i32 4>
30 %tmp17 = call <8 x float> @llvm.x86.avx.vinsertf128.ps.256(<8 x float> %tmp16, <4 x float> %a, i8 1)
31 %tmp18 = bitcast <4 x float> %tmp to <16 x i8>
32 %tmp19 = shufflevector <16 x i8> %tmp18, <16 x i8> %b, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 16, i32 17, i32 18, i32 19>
33 %tmp20 = bitcast <16 x i8> %tmp19 to <4 x float>
34 %tmp21 = bitcast <4 x float> %tmp15 to <16 x i8>
35 %tmp22 = shufflevector <16 x i8> %c, <16 x i8> %tmp21, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 16, i32 17, i32 18, i32 19>
36 %tmp23 = bitcast <16 x i8> %tmp22 to <4 x float>
37 %tmp24 = shufflevector <4 x float> %tmp20, <4 x float> <float 0.000000e+00, float undef, float undef, float undef>, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 4, i32 4, i32 4>
38 %tmp25 = call <8 x float> @llvm.x86.avx.vinsertf128.ps.256(<8 x float> %tmp24, <4 x float> %tmp23, i8 1)
39 %tmp26 = fmul <8 x float> %tmp17, %tmp17
40 %tmp27 = fmul <8 x float> %tmp25, %tmp25
41 %tmp28 = fadd <8 x float> %tmp26, %tmp27
42 %tmp29 = fadd <8 x float> %tmp28, %tmp28
43 %tmp30 = shufflevector <8 x float> %tmp29, <8 x float> %d, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
44 %tmp31 = fmul <4 x float> %tmp30, %tmp30
45 %tmp32 = call <8 x float> @llvm.x86.avx.vinsertf128.ps.256(<8 x float> zeroinitializer, <4 x float> %tmp31, i8 1)
46 %tmp33 = fadd <8 x float> %tmp32, %tmp32
47 %tmp34 = call <8 x float> @llvm.x86.avx.hadd.ps.256(<8 x float> %tmp33, <8 x float> %e) nounwind
48 %tmp35 = fsub <8 x float> %tmp34, %tmp34
49 %tmp36 = call <8 x float> @llvm.x86.avx.hadd.ps.256(<8 x float> zeroinitializer, <8 x float> %tmp35) nounwind
50 store <8 x float> %tmp36, ptr %f, align 32
54 declare <8 x float> @llvm.x86.avx.vinsertf128.ps.256(<8 x float>, <4 x float>, i8) nounwind readnone
56 declare <8 x float> @llvm.x86.avx.hadd.ps.256(<8 x float>, <8 x float>) nounwind readnone