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
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
|
; RUN: llvm-as < %s | llc -march=arm -mattr=+neon > %t
; RUN: grep {vrshl\\.s8} %t | count 2
; RUN: grep {vrshl\\.s16} %t | count 2
; RUN: grep {vrshl\\.s32} %t | count 2
; RUN: grep {vrshl\\.s64} %t | count 2
; RUN: grep {vrshl\\.u8} %t | count 2
; RUN: grep {vrshl\\.u16} %t | count 2
; RUN: grep {vrshl\\.u32} %t | count 2
; RUN: grep {vrshl\\.u64} %t | count 2
; RUN: grep {vrshr\\.s8} %t | count 2
; RUN: grep {vrshr\\.s16} %t | count 2
; RUN: grep {vrshr\\.s32} %t | count 2
; RUN: grep {vrshr\\.s64} %t | count 2
; RUN: grep {vrshr\\.u8} %t | count 2
; RUN: grep {vrshr\\.u16} %t | count 2
; RUN: grep {vrshr\\.u32} %t | count 2
; RUN: grep {vrshr\\.u64} %t | count 2
define <8 x i8> @vrshls8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
%tmp1 = load <8 x i8>* %A
%tmp2 = load <8 x i8>* %B
%tmp3 = call <8 x i8> @llvm.arm.neon.vrshifts.v8i8(<8 x i8> %tmp1, <8 x i8> %tmp2)
ret <8 x i8> %tmp3
}
define <4 x i16> @vrshls16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
%tmp1 = load <4 x i16>* %A
%tmp2 = load <4 x i16>* %B
%tmp3 = call <4 x i16> @llvm.arm.neon.vrshifts.v4i16(<4 x i16> %tmp1, <4 x i16> %tmp2)
ret <4 x i16> %tmp3
}
define <2 x i32> @vrshls32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
%tmp1 = load <2 x i32>* %A
%tmp2 = load <2 x i32>* %B
%tmp3 = call <2 x i32> @llvm.arm.neon.vrshifts.v2i32(<2 x i32> %tmp1, <2 x i32> %tmp2)
ret <2 x i32> %tmp3
}
define <1 x i64> @vrshls64(<1 x i64>* %A, <1 x i64>* %B) nounwind {
%tmp1 = load <1 x i64>* %A
%tmp2 = load <1 x i64>* %B
%tmp3 = call <1 x i64> @llvm.arm.neon.vrshifts.v1i64(<1 x i64> %tmp1, <1 x i64> %tmp2)
ret <1 x i64> %tmp3
}
define <8 x i8> @vrshlu8(<8 x i8>* %A, <8 x i8>* %B) nounwind {
%tmp1 = load <8 x i8>* %A
%tmp2 = load <8 x i8>* %B
%tmp3 = call <8 x i8> @llvm.arm.neon.vrshiftu.v8i8(<8 x i8> %tmp1, <8 x i8> %tmp2)
ret <8 x i8> %tmp3
}
define <4 x i16> @vrshlu16(<4 x i16>* %A, <4 x i16>* %B) nounwind {
%tmp1 = load <4 x i16>* %A
%tmp2 = load <4 x i16>* %B
%tmp3 = call <4 x i16> @llvm.arm.neon.vrshiftu.v4i16(<4 x i16> %tmp1, <4 x i16> %tmp2)
ret <4 x i16> %tmp3
}
define <2 x i32> @vrshlu32(<2 x i32>* %A, <2 x i32>* %B) nounwind {
%tmp1 = load <2 x i32>* %A
%tmp2 = load <2 x i32>* %B
%tmp3 = call <2 x i32> @llvm.arm.neon.vrshiftu.v2i32(<2 x i32> %tmp1, <2 x i32> %tmp2)
ret <2 x i32> %tmp3
}
define <1 x i64> @vrshlu64(<1 x i64>* %A, <1 x i64>* %B) nounwind {
%tmp1 = load <1 x i64>* %A
%tmp2 = load <1 x i64>* %B
%tmp3 = call <1 x i64> @llvm.arm.neon.vrshiftu.v1i64(<1 x i64> %tmp1, <1 x i64> %tmp2)
ret <1 x i64> %tmp3
}
define <16 x i8> @vrshlQs8(<16 x i8>* %A, <16 x i8>* %B) nounwind {
%tmp1 = load <16 x i8>* %A
%tmp2 = load <16 x i8>* %B
%tmp3 = call <16 x i8> @llvm.arm.neon.vrshifts.v16i8(<16 x i8> %tmp1, <16 x i8> %tmp2)
ret <16 x i8> %tmp3
}
define <8 x i16> @vrshlQs16(<8 x i16>* %A, <8 x i16>* %B) nounwind {
%tmp1 = load <8 x i16>* %A
%tmp2 = load <8 x i16>* %B
%tmp3 = call <8 x i16> @llvm.arm.neon.vrshifts.v8i16(<8 x i16> %tmp1, <8 x i16> %tmp2)
ret <8 x i16> %tmp3
}
define <4 x i32> @vrshlQs32(<4 x i32>* %A, <4 x i32>* %B) nounwind {
%tmp1 = load <4 x i32>* %A
%tmp2 = load <4 x i32>* %B
%tmp3 = call <4 x i32> @llvm.arm.neon.vrshifts.v4i32(<4 x i32> %tmp1, <4 x i32> %tmp2)
ret <4 x i32> %tmp3
}
define <2 x i64> @vrshlQs64(<2 x i64>* %A, <2 x i64>* %B) nounwind {
%tmp1 = load <2 x i64>* %A
%tmp2 = load <2 x i64>* %B
%tmp3 = call <2 x i64> @llvm.arm.neon.vrshifts.v2i64(<2 x i64> %tmp1, <2 x i64> %tmp2)
ret <2 x i64> %tmp3
}
define <16 x i8> @vrshlQu8(<16 x i8>* %A, <16 x i8>* %B) nounwind {
%tmp1 = load <16 x i8>* %A
%tmp2 = load <16 x i8>* %B
%tmp3 = call <16 x i8> @llvm.arm.neon.vrshiftu.v16i8(<16 x i8> %tmp1, <16 x i8> %tmp2)
ret <16 x i8> %tmp3
}
define <8 x i16> @vrshlQu16(<8 x i16>* %A, <8 x i16>* %B) nounwind {
%tmp1 = load <8 x i16>* %A
%tmp2 = load <8 x i16>* %B
%tmp3 = call <8 x i16> @llvm.arm.neon.vrshiftu.v8i16(<8 x i16> %tmp1, <8 x i16> %tmp2)
ret <8 x i16> %tmp3
}
define <4 x i32> @vrshlQu32(<4 x i32>* %A, <4 x i32>* %B) nounwind {
%tmp1 = load <4 x i32>* %A
%tmp2 = load <4 x i32>* %B
%tmp3 = call <4 x i32> @llvm.arm.neon.vrshiftu.v4i32(<4 x i32> %tmp1, <4 x i32> %tmp2)
ret <4 x i32> %tmp3
}
define <2 x i64> @vrshlQu64(<2 x i64>* %A, <2 x i64>* %B) nounwind {
%tmp1 = load <2 x i64>* %A
%tmp2 = load <2 x i64>* %B
%tmp3 = call <2 x i64> @llvm.arm.neon.vrshiftu.v2i64(<2 x i64> %tmp1, <2 x i64> %tmp2)
ret <2 x i64> %tmp3
}
define <8 x i8> @vrshrs8(<8 x i8>* %A) nounwind {
%tmp1 = load <8 x i8>* %A
%tmp2 = call <8 x i8> @llvm.arm.neon.vrshifts.v8i8(<8 x i8> %tmp1, <8 x i8> < i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8 >)
ret <8 x i8> %tmp2
}
define <4 x i16> @vrshrs16(<4 x i16>* %A) nounwind {
%tmp1 = load <4 x i16>* %A
%tmp2 = call <4 x i16> @llvm.arm.neon.vrshifts.v4i16(<4 x i16> %tmp1, <4 x i16> < i16 -16, i16 -16, i16 -16, i16 -16 >)
ret <4 x i16> %tmp2
}
define <2 x i32> @vrshrs32(<2 x i32>* %A) nounwind {
%tmp1 = load <2 x i32>* %A
%tmp2 = call <2 x i32> @llvm.arm.neon.vrshifts.v2i32(<2 x i32> %tmp1, <2 x i32> < i32 -32, i32 -32 >)
ret <2 x i32> %tmp2
}
define <1 x i64> @vrshrs64(<1 x i64>* %A) nounwind {
%tmp1 = load <1 x i64>* %A
%tmp2 = call <1 x i64> @llvm.arm.neon.vrshifts.v1i64(<1 x i64> %tmp1, <1 x i64> < i64 -64 >)
ret <1 x i64> %tmp2
}
define <8 x i8> @vrshru8(<8 x i8>* %A) nounwind {
%tmp1 = load <8 x i8>* %A
%tmp2 = call <8 x i8> @llvm.arm.neon.vrshiftu.v8i8(<8 x i8> %tmp1, <8 x i8> < i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8 >)
ret <8 x i8> %tmp2
}
define <4 x i16> @vrshru16(<4 x i16>* %A) nounwind {
%tmp1 = load <4 x i16>* %A
%tmp2 = call <4 x i16> @llvm.arm.neon.vrshiftu.v4i16(<4 x i16> %tmp1, <4 x i16> < i16 -16, i16 -16, i16 -16, i16 -16 >)
ret <4 x i16> %tmp2
}
define <2 x i32> @vrshru32(<2 x i32>* %A) nounwind {
%tmp1 = load <2 x i32>* %A
%tmp2 = call <2 x i32> @llvm.arm.neon.vrshiftu.v2i32(<2 x i32> %tmp1, <2 x i32> < i32 -32, i32 -32 >)
ret <2 x i32> %tmp2
}
define <1 x i64> @vrshru64(<1 x i64>* %A) nounwind {
%tmp1 = load <1 x i64>* %A
%tmp2 = call <1 x i64> @llvm.arm.neon.vrshiftu.v1i64(<1 x i64> %tmp1, <1 x i64> < i64 -64 >)
ret <1 x i64> %tmp2
}
define <16 x i8> @vrshrQs8(<16 x i8>* %A) nounwind {
%tmp1 = load <16 x i8>* %A
%tmp2 = call <16 x i8> @llvm.arm.neon.vrshifts.v16i8(<16 x i8> %tmp1, <16 x i8> < i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8 >)
ret <16 x i8> %tmp2
}
define <8 x i16> @vrshrQs16(<8 x i16>* %A) nounwind {
%tmp1 = load <8 x i16>* %A
%tmp2 = call <8 x i16> @llvm.arm.neon.vrshifts.v8i16(<8 x i16> %tmp1, <8 x i16> < i16 -16, i16 -16, i16 -16, i16 -16, i16 -16, i16 -16, i16 -16, i16 -16 >)
ret <8 x i16> %tmp2
}
define <4 x i32> @vrshrQs32(<4 x i32>* %A) nounwind {
%tmp1 = load <4 x i32>* %A
%tmp2 = call <4 x i32> @llvm.arm.neon.vrshifts.v4i32(<4 x i32> %tmp1, <4 x i32> < i32 -32, i32 -32, i32 -32, i32 -32 >)
ret <4 x i32> %tmp2
}
define <2 x i64> @vrshrQs64(<2 x i64>* %A) nounwind {
%tmp1 = load <2 x i64>* %A
%tmp2 = call <2 x i64> @llvm.arm.neon.vrshifts.v2i64(<2 x i64> %tmp1, <2 x i64> < i64 -64, i64 -64 >)
ret <2 x i64> %tmp2
}
define <16 x i8> @vrshrQu8(<16 x i8>* %A) nounwind {
%tmp1 = load <16 x i8>* %A
%tmp2 = call <16 x i8> @llvm.arm.neon.vrshiftu.v16i8(<16 x i8> %tmp1, <16 x i8> < i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8, i8 -8 >)
ret <16 x i8> %tmp2
}
define <8 x i16> @vrshrQu16(<8 x i16>* %A) nounwind {
%tmp1 = load <8 x i16>* %A
%tmp2 = call <8 x i16> @llvm.arm.neon.vrshiftu.v8i16(<8 x i16> %tmp1, <8 x i16> < i16 -16, i16 -16, i16 -16, i16 -16, i16 -16, i16 -16, i16 -16, i16 -16 >)
ret <8 x i16> %tmp2
}
define <4 x i32> @vrshrQu32(<4 x i32>* %A) nounwind {
%tmp1 = load <4 x i32>* %A
%tmp2 = call <4 x i32> @llvm.arm.neon.vrshiftu.v4i32(<4 x i32> %tmp1, <4 x i32> < i32 -32, i32 -32, i32 -32, i32 -32 >)
ret <4 x i32> %tmp2
}
define <2 x i64> @vrshrQu64(<2 x i64>* %A) nounwind {
%tmp1 = load <2 x i64>* %A
%tmp2 = call <2 x i64> @llvm.arm.neon.vrshiftu.v2i64(<2 x i64> %tmp1, <2 x i64> < i64 -64, i64 -64 >)
ret <2 x i64> %tmp2
}
declare <8 x i8> @llvm.arm.neon.vrshifts.v8i8(<8 x i8>, <8 x i8>) nounwind readnone
declare <4 x i16> @llvm.arm.neon.vrshifts.v4i16(<4 x i16>, <4 x i16>) nounwind readnone
declare <2 x i32> @llvm.arm.neon.vrshifts.v2i32(<2 x i32>, <2 x i32>) nounwind readnone
declare <1 x i64> @llvm.arm.neon.vrshifts.v1i64(<1 x i64>, <1 x i64>) nounwind readnone
declare <8 x i8> @llvm.arm.neon.vrshiftu.v8i8(<8 x i8>, <8 x i8>) nounwind readnone
declare <4 x i16> @llvm.arm.neon.vrshiftu.v4i16(<4 x i16>, <4 x i16>) nounwind readnone
declare <2 x i32> @llvm.arm.neon.vrshiftu.v2i32(<2 x i32>, <2 x i32>) nounwind readnone
declare <1 x i64> @llvm.arm.neon.vrshiftu.v1i64(<1 x i64>, <1 x i64>) nounwind readnone
declare <16 x i8> @llvm.arm.neon.vrshifts.v16i8(<16 x i8>, <16 x i8>) nounwind readnone
declare <8 x i16> @llvm.arm.neon.vrshifts.v8i16(<8 x i16>, <8 x i16>) nounwind readnone
declare <4 x i32> @llvm.arm.neon.vrshifts.v4i32(<4 x i32>, <4 x i32>) nounwind readnone
declare <2 x i64> @llvm.arm.neon.vrshifts.v2i64(<2 x i64>, <2 x i64>) nounwind readnone
declare <16 x i8> @llvm.arm.neon.vrshiftu.v16i8(<16 x i8>, <16 x i8>) nounwind readnone
declare <8 x i16> @llvm.arm.neon.vrshiftu.v8i16(<8 x i16>, <8 x i16>) nounwind readnone
declare <4 x i32> @llvm.arm.neon.vrshiftu.v4i32(<4 x i32>, <4 x i32>) nounwind readnone
declare <2 x i64> @llvm.arm.neon.vrshiftu.v2i64(<2 x i64>, <2 x i64>) nounwind readnone
|
