1 ; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
2 ; RUN: opt -passes=slp-vectorizer -mattr=+avx512f -mtriple=x86_64 -S < %s | FileCheck %s
4 define void @test(ptr noalias %0, ptr %p) {
6 ; CHECK-NEXT: [[TMP2:%.*]] = insertelement <8 x ptr> poison, ptr [[P:%.*]], i32 0
7 ; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <8 x ptr> [[TMP2]], <8 x ptr> poison, <8 x i32> zeroinitializer
8 ; CHECK-NEXT: [[TMP4:%.*]] = getelementptr float, <8 x ptr> [[TMP3]], <8 x i64> <i64 15, i64 4, i64 5, i64 0, i64 2, i64 6, i64 7, i64 8>
9 ; CHECK-NEXT: [[TMP5:%.*]] = getelementptr inbounds float, ptr [[TMP0:%.*]], i64 2
10 ; CHECK-NEXT: [[TMP6:%.*]] = call <8 x float> @llvm.masked.gather.v8f32.v8p0(<8 x ptr> [[TMP4]], i32 4, <8 x i1> <i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true, i1 true>, <8 x float> poison)
11 ; CHECK-NEXT: [[TMP7:%.*]] = shufflevector <8 x float> [[TMP6]], <8 x float> poison, <16 x i32> <i32 4, i32 3, i32 0, i32 1, i32 2, i32 0, i32 1, i32 2, i32 0, i32 2, i32 5, i32 6, i32 7, i32 5, i32 6, i32 7>
12 ; CHECK-NEXT: [[TMP8:%.*]] = shufflevector <8 x float> [[TMP6]], <8 x float> poison, <16 x i32> <i32 0, i32 1, i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
13 ; CHECK-NEXT: [[TMP9:%.*]] = shufflevector <16 x float> [[TMP8]], <16 x float> <float poison, float poison, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00, float 0.000000e+00>, <16 x i32> <i32 0, i32 1, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31>
14 ; CHECK-NEXT: [[TMP10:%.*]] = fadd reassoc nsz arcp contract afn <16 x float> [[TMP7]], [[TMP9]]
15 ; CHECK-NEXT: [[TMP11:%.*]] = shufflevector <16 x float> [[TMP10]], <16 x float> poison, <16 x i32> <i32 2, i32 3, i32 4, i32 5, i32 6, i32 7, i32 8, i32 1, i32 9, i32 0, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15>
16 ; CHECK-NEXT: store <16 x float> [[TMP11]], ptr [[TMP5]], align 4
17 ; CHECK-NEXT: ret void
19 %2 = getelementptr inbounds float, ptr %p, i64 2
20 %3 = getelementptr inbounds float, ptr %p, i64 4
21 %4 = load float, ptr %3, align 4
22 %5 = getelementptr inbounds float, ptr %p, i64 5
23 %6 = load float, ptr %5, align 16
24 %7 = getelementptr inbounds float, ptr %p, i64 15
25 %8 = load float, ptr %7, align 4
26 %9 = fadd reassoc nsz arcp contract afn float %8, 0.000000e+00
27 %10 = getelementptr inbounds float, ptr %0, i64 2
28 store float %9, ptr %10, align 4
29 %11 = fadd reassoc nsz arcp contract afn float %4, 0.000000e+00
30 %12 = getelementptr inbounds float, ptr %0, i64 3
31 store float %11, ptr %12, align 4
32 %13 = fadd reassoc nsz arcp contract afn float %6, 0.000000e+00
33 %14 = getelementptr inbounds float, ptr %0, i64 4
34 store float %13, ptr %14, align 4
35 %15 = fadd reassoc nsz arcp contract afn float %8, 0.000000e+00
36 %16 = getelementptr inbounds float, ptr %0, i64 5
37 store float %15, ptr %16, align 4
38 %17 = fadd reassoc nsz arcp contract afn float %4, 0.000000e+00
39 %18 = load float, ptr %p, align 16
40 %19 = getelementptr inbounds float, ptr %0, i64 6
41 store float %17, ptr %19, align 4
42 %20 = fadd reassoc nsz arcp contract afn float %6, 0.000000e+00
43 %21 = getelementptr inbounds float, ptr %0, i64 7
44 store float %20, ptr %21, align 4
45 %22 = fadd reassoc nsz arcp contract afn float %8, 0.000000e+00
46 %23 = load float, ptr %2, align 8
47 %24 = getelementptr inbounds float, ptr %0, i64 8
48 store float %22, ptr %24, align 4
49 %25 = fadd reassoc nsz arcp contract afn float %4, %18
50 %26 = getelementptr inbounds float, ptr %0, i64 9
51 store float %25, ptr %26, align 4
52 %27 = fadd reassoc nsz arcp contract afn float %6, 0.000000e+00
53 %28 = getelementptr inbounds float, ptr %0, i64 10
54 store float %27, ptr %28, align 4
55 %29 = fadd reassoc nsz arcp contract afn float %8, %23
56 %30 = getelementptr inbounds float, ptr %0, i64 11
57 store float %29, ptr %30, align 4
58 %31 = getelementptr inbounds float, ptr %p, i64 6
59 %32 = load float, ptr %31, align 4
60 %33 = fadd reassoc nsz arcp contract afn float %32, 0.000000e+00
61 %34 = getelementptr inbounds float, ptr %0, i64 12
62 store float %33, ptr %34, align 4
63 %35 = getelementptr inbounds float, ptr %p, i64 7
64 %36 = load float, ptr %35, align 8
65 %37 = fadd reassoc nsz arcp contract afn float %36, 0.000000e+00
66 %38 = getelementptr inbounds float, ptr %0, i64 13
67 store float %37, ptr %38, align 4
68 %39 = getelementptr inbounds float, ptr %p, i64 8
69 %40 = load float, ptr %39, align 4
70 %41 = fadd reassoc nsz arcp contract afn float %40, 0.000000e+00
71 %42 = getelementptr inbounds float, ptr %0, i64 14
72 store float %41, ptr %42, align 4
73 %43 = fadd reassoc nsz arcp contract afn float %32, 0.000000e+00
74 %44 = getelementptr inbounds float, ptr %0, i64 15
75 store float %43, ptr %44, align 4
76 %45 = fadd reassoc nsz arcp contract afn float %36, 0.000000e+00
77 %46 = getelementptr inbounds float, ptr %0, i64 16
78 store float %45, ptr %46, align 4
79 %47 = fadd reassoc nsz arcp contract afn float %40, 0.000000e+00
80 %48 = getelementptr inbounds float, ptr %0, i64 17
81 store float %47, ptr %48, align 4