1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt < %s -mtriple=x86_64-unknown-linux -mcpu=skylake-avx512 -passes=slp-vectorizer -S | FileCheck %s
; RUN: opt < %s -mtriple=x86_64-unknown-linux -mcpu=skylake-avx512 -passes=inject-tli-mappings,slp-vectorizer -vector-library=SVML -S | FileCheck %s --check-prefix=VECLIB
; RUN: opt < %s -mtriple=x86_64-unknown-linux -mcpu=skylake-avx512 -passes=inject-tli-mappings,slp-vectorizer -vector-library=AMDLIBM -S | FileCheck %s --check-prefix=AMDLIBM
@src = common global [8 x double] zeroinitializer, align 64
@dst = common global [8 x double] zeroinitializer, align 64
declare double @llvm.sqrt.f64(double)
declare double @llvm.sin.f64(double)
define void @test() {
; CHECK-LABEL: @test(
; CHECK-NEXT: [[A0:%.*]] = load double, ptr @src, align 8
; CHECK-NEXT: [[A1:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 1), align 8
; CHECK-NEXT: [[A2:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 2), align 8
; CHECK-NEXT: [[A3:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 3), align 8
; CHECK-NEXT: [[A4:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 4), align 8
; CHECK-NEXT: [[A5:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 5), align 8
; CHECK-NEXT: [[A6:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 6), align 8
; CHECK-NEXT: [[A7:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 7), align 8
; CHECK-NEXT: [[TMP1:%.*]] = insertelement <2 x double> poison, double [[A2]], i32 0
; CHECK-NEXT: [[TMP2:%.*]] = insertelement <2 x double> [[TMP1]], double [[A6]], i32 1
; CHECK-NEXT: [[TMP3:%.*]] = call fast <2 x double> @llvm.sin.v2f64(<2 x double> [[TMP2]])
; CHECK-NEXT: [[TMP4:%.*]] = insertelement <2 x double> poison, double [[A3]], i32 0
; CHECK-NEXT: [[TMP5:%.*]] = insertelement <2 x double> [[TMP4]], double [[A7]], i32 1
; CHECK-NEXT: [[TMP6:%.*]] = call fast <2 x double> @llvm.sin.v2f64(<2 x double> [[TMP5]])
; CHECK-NEXT: [[TMP7:%.*]] = insertelement <2 x double> poison, double [[A0]], i32 0
; CHECK-NEXT: [[TMP8:%.*]] = insertelement <2 x double> [[TMP7]], double [[A4]], i32 1
; CHECK-NEXT: [[TMP9:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP8]])
; CHECK-NEXT: [[TMP10:%.*]] = insertelement <2 x double> poison, double [[A1]], i32 0
; CHECK-NEXT: [[TMP11:%.*]] = insertelement <2 x double> [[TMP10]], double [[A5]], i32 1
; CHECK-NEXT: [[TMP12:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP11]])
; CHECK-NEXT: [[TMP13:%.*]] = fadd fast <2 x double> [[TMP9]], [[TMP6]]
; CHECK-NEXT: [[TMP14:%.*]] = fadd fast <2 x double> [[TMP3]], [[TMP12]]
; CHECK-NEXT: [[TMP15:%.*]] = fadd fast <2 x double> [[TMP13]], [[TMP14]]
; CHECK-NEXT: store <2 x double> [[TMP15]], ptr @dst, align 8
; CHECK-NEXT: ret void
;
; VECLIB-LABEL: @test(
; VECLIB-NEXT: [[A0:%.*]] = load double, ptr @src, align 8
; VECLIB-NEXT: [[A1:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 1), align 8
; VECLIB-NEXT: [[A2:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 2), align 8
; VECLIB-NEXT: [[A3:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 3), align 8
; VECLIB-NEXT: [[A4:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 4), align 8
; VECLIB-NEXT: [[A5:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 5), align 8
; VECLIB-NEXT: [[A6:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 6), align 8
; VECLIB-NEXT: [[A7:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 7), align 8
; VECLIB-NEXT: [[TMP1:%.*]] = insertelement <2 x double> poison, double [[A2]], i32 0
; VECLIB-NEXT: [[TMP2:%.*]] = insertelement <2 x double> [[TMP1]], double [[A6]], i32 1
; VECLIB-NEXT: [[TMP3:%.*]] = call fast <2 x double> @__svml_sin2(<2 x double> [[TMP2]])
; VECLIB-NEXT: [[TMP4:%.*]] = insertelement <2 x double> poison, double [[A3]], i32 0
; VECLIB-NEXT: [[TMP5:%.*]] = insertelement <2 x double> [[TMP4]], double [[A7]], i32 1
; VECLIB-NEXT: [[TMP6:%.*]] = call fast <2 x double> @__svml_sin2(<2 x double> [[TMP5]])
; VECLIB-NEXT: [[TMP7:%.*]] = insertelement <2 x double> poison, double [[A0]], i32 0
; VECLIB-NEXT: [[TMP8:%.*]] = insertelement <2 x double> [[TMP7]], double [[A4]], i32 1
; VECLIB-NEXT: [[TMP9:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP8]])
; VECLIB-NEXT: [[TMP10:%.*]] = insertelement <2 x double> poison, double [[A1]], i32 0
; VECLIB-NEXT: [[TMP11:%.*]] = insertelement <2 x double> [[TMP10]], double [[A5]], i32 1
; VECLIB-NEXT: [[TMP12:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP11]])
; VECLIB-NEXT: [[TMP13:%.*]] = fadd fast <2 x double> [[TMP9]], [[TMP6]]
; VECLIB-NEXT: [[TMP14:%.*]] = fadd fast <2 x double> [[TMP3]], [[TMP12]]
; VECLIB-NEXT: [[TMP15:%.*]] = fadd fast <2 x double> [[TMP13]], [[TMP14]]
; VECLIB-NEXT: store <2 x double> [[TMP15]], ptr @dst, align 8
; VECLIB-NEXT: ret void
;
; AMDLIBM-LABEL: @test(
; AMDLIBM-NEXT: [[A0:%.*]] = load double, ptr @src, align 8
; AMDLIBM-NEXT: [[A1:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 1), align 8
; AMDLIBM-NEXT: [[A2:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 2), align 8
; AMDLIBM-NEXT: [[A3:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 3), align 8
; AMDLIBM-NEXT: [[A4:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 4), align 8
; AMDLIBM-NEXT: [[A5:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 5), align 8
; AMDLIBM-NEXT: [[A6:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 6), align 8
; AMDLIBM-NEXT: [[A7:%.*]] = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 7), align 8
; AMDLIBM-NEXT: [[TMP1:%.*]] = insertelement <2 x double> poison, double [[A2]], i32 0
; AMDLIBM-NEXT: [[TMP2:%.*]] = insertelement <2 x double> [[TMP1]], double [[A6]], i32 1
; AMDLIBM-NEXT: [[TMP3:%.*]] = call fast <2 x double> @amd_vrd2_sin(<2 x double> [[TMP2]])
; AMDLIBM-NEXT: [[TMP4:%.*]] = insertelement <2 x double> poison, double [[A3]], i32 0
; AMDLIBM-NEXT: [[TMP5:%.*]] = insertelement <2 x double> [[TMP4]], double [[A7]], i32 1
; AMDLIBM-NEXT: [[TMP6:%.*]] = call fast <2 x double> @amd_vrd2_sin(<2 x double> [[TMP5]])
; AMDLIBM-NEXT: [[TMP7:%.*]] = insertelement <2 x double> poison, double [[A0]], i32 0
; AMDLIBM-NEXT: [[TMP8:%.*]] = insertelement <2 x double> [[TMP7]], double [[A4]], i32 1
; AMDLIBM-NEXT: [[TMP9:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP8]])
; AMDLIBM-NEXT: [[TMP10:%.*]] = insertelement <2 x double> poison, double [[A1]], i32 0
; AMDLIBM-NEXT: [[TMP11:%.*]] = insertelement <2 x double> [[TMP10]], double [[A5]], i32 1
; AMDLIBM-NEXT: [[TMP12:%.*]] = call fast <2 x double> @llvm.sqrt.v2f64(<2 x double> [[TMP11]])
; AMDLIBM-NEXT: [[TMP13:%.*]] = fadd fast <2 x double> [[TMP9]], [[TMP6]]
; AMDLIBM-NEXT: [[TMP14:%.*]] = fadd fast <2 x double> [[TMP3]], [[TMP12]]
; AMDLIBM-NEXT: [[TMP15:%.*]] = fadd fast <2 x double> [[TMP13]], [[TMP14]]
; AMDLIBM-NEXT: store <2 x double> [[TMP15]], ptr @dst, align 8
; AMDLIBM-NEXT: ret void
;
%a0 = load double, ptr @src, align 8
%a1 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 1), align 8
%a2 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 2), align 8
%a3 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 3), align 8
%a4 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 4), align 8
%a5 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 5), align 8
%a6 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 6), align 8
%a7 = load double, ptr getelementptr inbounds ([8 x double], ptr @src, i32 0, i64 7), align 8
%sin0 = call fast double @llvm.sin.f64(double %a2)
%sin1 = call fast double @llvm.sin.f64(double %a3)
%sqrt0 = call fast double @llvm.sqrt.f64(double %a0)
%sqrt1 = call fast double @llvm.sqrt.f64(double %a1)
%sin2 = call fast double @llvm.sin.f64(double %a6)
%sin3 = call fast double @llvm.sin.f64(double %a7)
%sqrt2 = call fast double @llvm.sqrt.f64(double %a4)
%sqrt3 = call fast double @llvm.sqrt.f64(double %a5)
%res1 = fadd fast double %sqrt0, %sin1
%res2 = fadd fast double %sin0, %sqrt1
%res00 = fadd fast double %res1, %res2
%res3 = fadd fast double %sqrt2, %sin3
%res4 = fadd fast double %sin2, %sqrt3
%res01 = fadd fast double %res3, %res4
store double %res00, ptr @dst, align 8
store double %res01, ptr getelementptr inbounds ([8 x double], ptr @dst, i32 0, i64 1), align 8
ret void
}
|
