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
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
|
; NOTE: Assertions have been autogenerated by utils/update_analyze_test_checks.py
; RUN: opt < %s -passes="print<cost-model>" -mtriple=x86_64-apple-darwin 2>&1 -disable-output -mattr=+sse2 | FileCheck %s --check-prefixes=SSE
; RUN: opt < %s -passes="print<cost-model>" -mtriple=x86_64-apple-darwin 2>&1 -disable-output -mattr=+ssse3 | FileCheck %s --check-prefixes=SSE
; RUN: opt < %s -passes="print<cost-model>" -mtriple=x86_64-apple-darwin 2>&1 -disable-output -mattr=+sse4.1 | FileCheck %s --check-prefixes=SSE
; RUN: opt < %s -passes="print<cost-model>" -mtriple=x86_64-apple-darwin 2>&1 -disable-output -mattr=+sse4.2 | FileCheck %s --check-prefixes=SSE
; RUN: opt < %s -passes="print<cost-model>" -mtriple=x86_64-apple-darwin 2>&1 -disable-output -mattr=+avx | FileCheck %s --check-prefixes=AVX,AVX1
; RUN: opt < %s -passes="print<cost-model>" -mtriple=x86_64-apple-darwin 2>&1 -disable-output -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX2
; RUN: opt < %s -passes="print<cost-model>" -mtriple=x86_64-apple-darwin 2>&1 -disable-output -mattr=+avx512f | FileCheck %s --check-prefixes=AVX512
; RUN: opt < %s -passes="print<cost-model>" -mtriple=x86_64-apple-darwin 2>&1 -disable-output -mattr=+avx512f,+avx512bw | FileCheck %s --check-prefixes=AVX512
; RUN: opt < %s -passes="print<cost-model>" -mtriple=x86_64-apple-darwin 2>&1 -disable-output -mattr=+avx512f,+avx512dq | FileCheck %s --check-prefixes=AVX512
define i32 @reduce_f64(i32 %arg) {
; SSE-LABEL: 'reduce_f64'
; SSE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call double @llvm.vector.reduce.fmax.v1f64(<1 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V2 = call double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V4 = call double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %V8 = call double @llvm.vector.reduce.fmax.v8f64(<8 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 33 for instruction: %V16 = call double @llvm.vector.reduce.fmax.v16f64(<16 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_f64'
; AVX1-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call double @llvm.vector.reduce.fmax.v1f64(<1 x double> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V2 = call double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4 = call double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 13 for instruction: %V8 = call double @llvm.vector.reduce.fmax.v8f64(<8 x double> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 23 for instruction: %V16 = call double @llvm.vector.reduce.fmax.v16f64(<16 x double> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX2-LABEL: 'reduce_f64'
; AVX2-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call double @llvm.vector.reduce.fmax.v1f64(<1 x double> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4 = call double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V8 = call double @llvm.vector.reduce.fmax.v8f64(<8 x double> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 15 for instruction: %V16 = call double @llvm.vector.reduce.fmax.v16f64(<16 x double> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX512-LABEL: 'reduce_f64'
; AVX512-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call double @llvm.vector.reduce.fmax.v1f64(<1 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4 = call double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %V8 = call double @llvm.vector.reduce.fmax.v8f64(<8 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %V16 = call double @llvm.vector.reduce.fmax.v16f64(<16 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
%V1 = call double @llvm.vector.reduce.fmax.v1f64(<1 x double> undef)
%V2 = call double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
%V4 = call double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
%V8 = call double @llvm.vector.reduce.fmax.v8f64(<8 x double> undef)
%V16 = call double @llvm.vector.reduce.fmax.v16f64(<16 x double> undef)
ret i32 undef
}
define i32 @reduce_f32(i32 %arg) {
; SSE-LABEL: 'reduce_f32'
; SSE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.vector.reduce.fmax.v1f32(<1 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V2 = call float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 10 for instruction: %V4 = call float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 14 for instruction: %V8 = call float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 22 for instruction: %V16 = call float @llvm.vector.reduce.fmax.v16f32(<16 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 38 for instruction: %V32 = call float @llvm.vector.reduce.fmax.v32f32(<32 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX1-LABEL: 'reduce_f32'
; AVX1-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.vector.reduce.fmax.v1f32(<1 x float> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V2 = call float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V4 = call float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V8 = call float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 17 for instruction: %V16 = call float @llvm.vector.reduce.fmax.v16f32(<16 x float> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 27 for instruction: %V32 = call float @llvm.vector.reduce.fmax.v32f32(<32 x float> undef)
; AVX1-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX2-LABEL: 'reduce_f32'
; AVX2-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.vector.reduce.fmax.v1f32(<1 x float> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V2 = call float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V4 = call float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V8 = call float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 12 for instruction: %V16 = call float @llvm.vector.reduce.fmax.v16f32(<16 x float> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 18 for instruction: %V32 = call float @llvm.vector.reduce.fmax.v32f32(<32 x float> undef)
; AVX2-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX512-LABEL: 'reduce_f32'
; AVX512-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call float @llvm.vector.reduce.fmax.v1f32(<1 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4 = call float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8 = call float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V16 = call float @llvm.vector.reduce.fmax.v16f32(<16 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 13 for instruction: %V32 = call float @llvm.vector.reduce.fmax.v32f32(<32 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
%V1 = call float @llvm.vector.reduce.fmax.v1f32(<1 x float> undef)
%V2 = call float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
%V4 = call float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
%V8 = call float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
%V16 = call float @llvm.vector.reduce.fmax.v16f32(<16 x float> undef)
%V32 = call float @llvm.vector.reduce.fmax.v32f32(<32 x float> undef)
ret i32 undef
}
; Fast Reductions
define i32 @reduce_f64_fast(i32 %arg) {
; SSE-LABEL: 'reduce_f64_fast'
; SSE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call fast double @llvm.vector.reduce.fmax.v1f64(<1 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call fast double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 3 for instruction: %V4 = call fast double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V8 = call fast double @llvm.vector.reduce.fmax.v8f64(<8 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V16 = call fast double @llvm.vector.reduce.fmax.v16f64(<16 x double> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX-LABEL: 'reduce_f64_fast'
; AVX-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call fast double @llvm.vector.reduce.fmax.v1f64(<1 x double> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call fast double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4 = call fast double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V8 = call fast double @llvm.vector.reduce.fmax.v8f64(<8 x double> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %V16 = call fast double @llvm.vector.reduce.fmax.v16f64(<16 x double> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX512-LABEL: 'reduce_f64_fast'
; AVX512-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call fast double @llvm.vector.reduce.fmax.v1f64(<1 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call fast double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4 = call fast double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8 = call fast double @llvm.vector.reduce.fmax.v8f64(<8 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %V16 = call fast double @llvm.vector.reduce.fmax.v16f64(<16 x double> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
%V1 = call fast double @llvm.vector.reduce.fmax.v1f64(<1 x double> undef)
%V2 = call fast double @llvm.vector.reduce.fmax.v2f64(<2 x double> undef)
%V4 = call fast double @llvm.vector.reduce.fmax.v4f64(<4 x double> undef)
%V8 = call fast double @llvm.vector.reduce.fmax.v8f64(<8 x double> undef)
%V16 = call fast double @llvm.vector.reduce.fmax.v16f64(<16 x double> undef)
ret i32 undef
}
define i32 @reduce_f32_fast(i32 %arg) {
; SSE-LABEL: 'reduce_f32_fast'
; SSE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call fast float @llvm.vector.reduce.fmax.v1f32(<1 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call fast float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4 = call fast float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 5 for instruction: %V8 = call fast float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %V16 = call fast float @llvm.vector.reduce.fmax.v16f32(<16 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 11 for instruction: %V32 = call fast float @llvm.vector.reduce.fmax.v32f32(<32 x float> undef)
; SSE-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX-LABEL: 'reduce_f32_fast'
; AVX-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call fast float @llvm.vector.reduce.fmax.v1f32(<1 x float> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call fast float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4 = call fast float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8 = call fast float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 7 for instruction: %V16 = call fast float @llvm.vector.reduce.fmax.v16f32(<16 x float> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V32 = call fast float @llvm.vector.reduce.fmax.v32f32(<32 x float> undef)
; AVX-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
; AVX512-LABEL: 'reduce_f32_fast'
; AVX512-NEXT: Cost Model: Found an estimated cost of 0 for instruction: %V1 = call fast float @llvm.vector.reduce.fmax.v1f32(<1 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 2 for instruction: %V2 = call fast float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 4 for instruction: %V4 = call fast float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 6 for instruction: %V8 = call fast float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 8 for instruction: %V16 = call fast float @llvm.vector.reduce.fmax.v16f32(<16 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 9 for instruction: %V32 = call fast float @llvm.vector.reduce.fmax.v32f32(<32 x float> undef)
; AVX512-NEXT: Cost Model: Found an estimated cost of 0 for instruction: ret i32 undef
;
%V1 = call fast float @llvm.vector.reduce.fmax.v1f32(<1 x float> undef)
%V2 = call fast float @llvm.vector.reduce.fmax.v2f32(<2 x float> undef)
%V4 = call fast float @llvm.vector.reduce.fmax.v4f32(<4 x float> undef)
%V8 = call fast float @llvm.vector.reduce.fmax.v8f32(<8 x float> undef)
%V16 = call fast float @llvm.vector.reduce.fmax.v16f32(<16 x float> undef)
%V32 = call fast float @llvm.vector.reduce.fmax.v32f32(<32 x float> undef)
ret i32 undef
}
declare double @llvm.vector.reduce.fmax.v1f64(<1 x double>)
declare double @llvm.vector.reduce.fmax.v2f64(<2 x double>)
declare double @llvm.vector.reduce.fmax.v4f64(<4 x double>)
declare double @llvm.vector.reduce.fmax.v8f64(<8 x double>)
declare double @llvm.vector.reduce.fmax.v16f64(<16 x double>)
declare float @llvm.vector.reduce.fmax.v1f32(<1 x float>)
declare float @llvm.vector.reduce.fmax.v2f32(<2 x float>)
declare float @llvm.vector.reduce.fmax.v4f32(<4 x float>)
declare float @llvm.vector.reduce.fmax.v8f32(<8 x float>)
declare float @llvm.vector.reduce.fmax.v16f32(<16 x float>)
declare float @llvm.vector.reduce.fmax.v32f32(<32 x float>)
|
