sve2-intrinsics-binary-narrowing-add-sub.ll
12.8 KB
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
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
; RUN: llc -mtriple=aarch64-linux-gnu -mattr=+sve2 < %s | FileCheck %s
; ADDHNB
define <vscale x 16 x i8> @addhnb_h(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b) {
; CHECK-LABEL: addhnb_h:
; CHECK: addhnb z0.b, z0.h, z1.h
; CHECK-NEXT: ret
%out = call <vscale x 16 x i8> @llvm.aarch64.sve.addhnb.nxv8i16(<vscale x 8 x i16> %a,
<vscale x 8 x i16> %b)
ret <vscale x 16 x i8> %out
}
define <vscale x 8 x i16> @addhnb_s(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b) {
; CHECK-LABEL: addhnb_s:
; CHECK: addhnb z0.h, z0.s, z1.s
; CHECK-NEXT: ret
%out = call <vscale x 8 x i16> @llvm.aarch64.sve.addhnb.nxv4i32(<vscale x 4 x i32> %a,
<vscale x 4 x i32> %b)
ret <vscale x 8 x i16> %out
}
define <vscale x 4 x i32> @addhnb_d(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b) {
; CHECK-LABEL: addhnb_d:
; CHECK: addhnb z0.s, z0.d, z1.d
; CHECK-NEXT: ret
%out = call <vscale x 4 x i32> @llvm.aarch64.sve.addhnb.nxv2i64(<vscale x 2 x i64> %a,
<vscale x 2 x i64> %b)
ret <vscale x 4 x i32> %out
}
; ADDHNT
define <vscale x 16 x i8> @addhnt_h(<vscale x 16 x i8> %a, <vscale x 8 x i16> %b, <vscale x 8 x i16> %c) {
; CHECK-LABEL: addhnt_h:
; CHECK: addhnt z0.b, z1.h, z2.h
; CHECK-NEXT: ret
%out = call <vscale x 16 x i8> @llvm.aarch64.sve.addhnt.nxv8i16(<vscale x 16 x i8> %a,
<vscale x 8 x i16> %b,
<vscale x 8 x i16> %c)
ret <vscale x 16 x i8> %out
}
define <vscale x 8 x i16> @addhnt_s(<vscale x 8 x i16> %a, <vscale x 4 x i32> %b, <vscale x 4 x i32> %c) {
; CHECK-LABEL: addhnt_s:
; CHECK: addhnt z0.h, z1.s, z2.s
; CHECK-NEXT: ret
%out = call <vscale x 8 x i16> @llvm.aarch64.sve.addhnt.nxv4i32(<vscale x 8 x i16> %a,
<vscale x 4 x i32> %b,
<vscale x 4 x i32> %c)
ret <vscale x 8 x i16> %out
}
define <vscale x 4 x i32> @addhnt_d(<vscale x 4 x i32> %a, <vscale x 2 x i64> %b, <vscale x 2 x i64> %c) {
; CHECK-LABEL: addhnt_d:
; CHECK: addhnt z0.s, z1.d, z2.d
; CHECK-NEXT: ret
%out = call <vscale x 4 x i32> @llvm.aarch64.sve.addhnt.nxv2i64(<vscale x 4 x i32> %a,
<vscale x 2 x i64> %b,
<vscale x 2 x i64> %c)
ret <vscale x 4 x i32> %out
}
; RADDHNB
define <vscale x 16 x i8> @raddhnb_h(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b) {
; CHECK-LABEL: raddhnb_h:
; CHECK: raddhnb z0.b, z0.h, z1.h
; CHECK-NEXT: ret
%out = call <vscale x 16 x i8> @llvm.aarch64.sve.raddhnb.nxv8i16(<vscale x 8 x i16> %a,
<vscale x 8 x i16> %b)
ret <vscale x 16 x i8> %out
}
define <vscale x 8 x i16> @raddhnb_s(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b) {
; CHECK-LABEL: raddhnb_s:
; CHECK: raddhnb z0.h, z0.s, z1.s
; CHECK-NEXT: ret
%out = call <vscale x 8 x i16> @llvm.aarch64.sve.raddhnb.nxv4i32(<vscale x 4 x i32> %a,
<vscale x 4 x i32> %b)
ret <vscale x 8 x i16> %out
}
define <vscale x 4 x i32> @raddhnb_d(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b) {
; CHECK-LABEL: raddhnb_d:
; CHECK: raddhnb z0.s, z0.d, z1.d
; CHECK-NEXT: ret
%out = call <vscale x 4 x i32> @llvm.aarch64.sve.raddhnb.nxv2i64(<vscale x 2 x i64> %a,
<vscale x 2 x i64> %b)
ret <vscale x 4 x i32> %out
}
; RADDHNT
define <vscale x 16 x i8> @raddhnt_h(<vscale x 16 x i8> %a, <vscale x 8 x i16> %b, <vscale x 8 x i16> %c) {
; CHECK-LABEL: raddhnt_h:
; CHECK: raddhnt z0.b, z1.h, z2.h
; CHECK-NEXT: ret
%out = call <vscale x 16 x i8> @llvm.aarch64.sve.raddhnt.nxv8i16(<vscale x 16 x i8> %a,
<vscale x 8 x i16> %b,
<vscale x 8 x i16> %c)
ret <vscale x 16 x i8> %out
}
define <vscale x 8 x i16> @raddhnt_s(<vscale x 8 x i16> %a, <vscale x 4 x i32> %b, <vscale x 4 x i32> %c) {
; CHECK-LABEL: raddhnt_s:
; CHECK: raddhnt z0.h, z1.s, z2.s
; CHECK-NEXT: ret
%out = call <vscale x 8 x i16> @llvm.aarch64.sve.raddhnt.nxv4i32(<vscale x 8 x i16> %a,
<vscale x 4 x i32> %b,
<vscale x 4 x i32> %c)
ret <vscale x 8 x i16> %out
}
define <vscale x 4 x i32> @raddhnt_d(<vscale x 4 x i32> %a, <vscale x 2 x i64> %b, <vscale x 2 x i64> %c) {
; CHECK-LABEL: raddhnt_d:
; CHECK: raddhnt z0.s, z1.d, z2.d
; CHECK-NEXT: ret
%out = call <vscale x 4 x i32> @llvm.aarch64.sve.raddhnt.nxv2i64(<vscale x 4 x i32> %a,
<vscale x 2 x i64> %b,
<vscale x 2 x i64> %c)
ret <vscale x 4 x i32> %out
}
; RSUBHNB
define <vscale x 16 x i8> @rsubhnb_h(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b) {
; CHECK-LABEL: rsubhnb_h:
; CHECK: rsubhnb z0.b, z0.h, z1.h
; CHECK-NEXT: ret
%out = call <vscale x 16 x i8> @llvm.aarch64.sve.rsubhnb.nxv8i16(<vscale x 8 x i16> %a,
<vscale x 8 x i16> %b)
ret <vscale x 16 x i8> %out
}
define <vscale x 8 x i16> @rsubhnb_s(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b) {
; CHECK-LABEL: rsubhnb_s:
; CHECK: rsubhnb z0.h, z0.s, z1.s
; CHECK-NEXT: ret
%out = call <vscale x 8 x i16> @llvm.aarch64.sve.rsubhnb.nxv4i32(<vscale x 4 x i32> %a,
<vscale x 4 x i32> %b)
ret <vscale x 8 x i16> %out
}
define <vscale x 4 x i32> @rsubhnb_d(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b) {
; CHECK-LABEL: rsubhnb_d:
; CHECK: rsubhnb z0.s, z0.d, z1.d
; CHECK-NEXT: ret
%out = call <vscale x 4 x i32> @llvm.aarch64.sve.rsubhnb.nxv2i64(<vscale x 2 x i64> %a,
<vscale x 2 x i64> %b)
ret <vscale x 4 x i32> %out
}
; RSUBHNT
define <vscale x 16 x i8> @rsubhnt_h(<vscale x 16 x i8> %a, <vscale x 8 x i16> %b, <vscale x 8 x i16> %c) {
; CHECK-LABEL: rsubhnt_h:
; CHECK: rsubhnt z0.b, z1.h, z2.h
; CHECK-NEXT: ret
%out = call <vscale x 16 x i8> @llvm.aarch64.sve.rsubhnt.nxv8i16(<vscale x 16 x i8> %a,
<vscale x 8 x i16> %b,
<vscale x 8 x i16> %c)
ret <vscale x 16 x i8> %out
}
define <vscale x 8 x i16> @rsubhnt_s(<vscale x 8 x i16> %a, <vscale x 4 x i32> %b, <vscale x 4 x i32> %c) {
; CHECK-LABEL: rsubhnt_s:
; CHECK: rsubhnt z0.h, z1.s, z2.s
; CHECK-NEXT: ret
%out = call <vscale x 8 x i16> @llvm.aarch64.sve.rsubhnt.nxv4i32(<vscale x 8 x i16> %a,
<vscale x 4 x i32> %b,
<vscale x 4 x i32> %c)
ret <vscale x 8 x i16> %out
}
define <vscale x 4 x i32> @rsubhnt_d(<vscale x 4 x i32> %a, <vscale x 2 x i64> %b, <vscale x 2 x i64> %c) {
; CHECK-LABEL: rsubhnt_d:
; CHECK: rsubhnt z0.s, z1.d, z2.d
; CHECK-NEXT: ret
%out = call <vscale x 4 x i32> @llvm.aarch64.sve.rsubhnt.nxv2i64(<vscale x 4 x i32> %a,
<vscale x 2 x i64> %b,
<vscale x 2 x i64> %c)
ret <vscale x 4 x i32> %out
}
; SUBHNB
define <vscale x 16 x i8> @subhnb_h(<vscale x 8 x i16> %a, <vscale x 8 x i16> %b) {
; CHECK-LABEL: subhnb_h:
; CHECK: subhnb z0.b, z0.h, z1.h
; CHECK-NEXT: ret
%out = call <vscale x 16 x i8> @llvm.aarch64.sve.subhnb.nxv8i16(<vscale x 8 x i16> %a,
<vscale x 8 x i16> %b)
ret <vscale x 16 x i8> %out
}
define <vscale x 8 x i16> @subhnb_s(<vscale x 4 x i32> %a, <vscale x 4 x i32> %b) {
; CHECK-LABEL: subhnb_s:
; CHECK: subhnb z0.h, z0.s, z1.s
; CHECK-NEXT: ret
%out = call <vscale x 8 x i16> @llvm.aarch64.sve.subhnb.nxv4i32(<vscale x 4 x i32> %a,
<vscale x 4 x i32> %b)
ret <vscale x 8 x i16> %out
}
define <vscale x 4 x i32> @subhnb_d(<vscale x 2 x i64> %a, <vscale x 2 x i64> %b) {
; CHECK-LABEL: subhnb_d:
; CHECK: subhnb z0.s, z0.d, z1.d
; CHECK-NEXT: ret
%out = call <vscale x 4 x i32> @llvm.aarch64.sve.subhnb.nxv2i64(<vscale x 2 x i64> %a,
<vscale x 2 x i64> %b)
ret <vscale x 4 x i32> %out
}
; SUBHNT
define <vscale x 16 x i8> @subhnt_h(<vscale x 16 x i8> %a, <vscale x 8 x i16> %b, <vscale x 8 x i16> %c) {
; CHECK-LABEL: subhnt_h:
; CHECK: subhnt z0.b, z1.h, z2.h
; CHECK-NEXT: ret
%out = call <vscale x 16 x i8> @llvm.aarch64.sve.subhnt.nxv8i16(<vscale x 16 x i8> %a,
<vscale x 8 x i16> %b,
<vscale x 8 x i16> %c)
ret <vscale x 16 x i8> %out
}
define <vscale x 8 x i16> @subhnt_s(<vscale x 8 x i16> %a, <vscale x 4 x i32> %b, <vscale x 4 x i32> %c) {
; CHECK-LABEL: subhnt_s:
; CHECK: subhnt z0.h, z1.s, z2.s
; CHECK-NEXT: ret
%out = call <vscale x 8 x i16> @llvm.aarch64.sve.subhnt.nxv4i32(<vscale x 8 x i16> %a,
<vscale x 4 x i32> %b,
<vscale x 4 x i32> %c)
ret <vscale x 8 x i16> %out
}
define <vscale x 4 x i32> @subhnt_d(<vscale x 4 x i32> %a, <vscale x 2 x i64> %b, <vscale x 2 x i64> %c) {
; CHECK-LABEL: subhnt_d:
; CHECK: subhnt z0.s, z1.d, z2.d
; CHECK-NEXT: ret
%out = call <vscale x 4 x i32> @llvm.aarch64.sve.subhnt.nxv2i64(<vscale x 4 x i32> %a,
<vscale x 2 x i64> %b,
<vscale x 2 x i64> %c)
ret <vscale x 4 x i32> %out
}
declare <vscale x 16 x i8> @llvm.aarch64.sve.addhnb.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>)
declare <vscale x 8 x i16> @llvm.aarch64.sve.addhnb.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.addhnb.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>)
declare <vscale x 16 x i8> @llvm.aarch64.sve.addhnt.nxv8i16(<vscale x 16 x i8>, <vscale x 8 x i16>, <vscale x 8 x i16>)
declare <vscale x 8 x i16> @llvm.aarch64.sve.addhnt.nxv4i32(<vscale x 8 x i16>, <vscale x 4 x i32>, <vscale x 4 x i32>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.addhnt.nxv2i64(<vscale x 4 x i32>, <vscale x 2 x i64>, <vscale x 2 x i64>)
declare <vscale x 16 x i8> @llvm.aarch64.sve.raddhnb.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>)
declare <vscale x 8 x i16> @llvm.aarch64.sve.raddhnb.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.raddhnb.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>)
declare <vscale x 16 x i8> @llvm.aarch64.sve.raddhnt.nxv8i16(<vscale x 16 x i8>, <vscale x 8 x i16>, <vscale x 8 x i16>)
declare <vscale x 8 x i16> @llvm.aarch64.sve.raddhnt.nxv4i32(<vscale x 8 x i16>, <vscale x 4 x i32>, <vscale x 4 x i32>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.raddhnt.nxv2i64(<vscale x 4 x i32>, <vscale x 2 x i64>, <vscale x 2 x i64>)
declare <vscale x 16 x i8> @llvm.aarch64.sve.subhnb.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>)
declare <vscale x 8 x i16> @llvm.aarch64.sve.subhnb.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.subhnb.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>)
declare <vscale x 16 x i8> @llvm.aarch64.sve.subhnt.nxv8i16(<vscale x 16 x i8>, <vscale x 8 x i16>, <vscale x 8 x i16>)
declare <vscale x 8 x i16> @llvm.aarch64.sve.subhnt.nxv4i32(<vscale x 8 x i16>, <vscale x 4 x i32>, <vscale x 4 x i32>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.subhnt.nxv2i64(<vscale x 4 x i32>, <vscale x 2 x i64>, <vscale x 2 x i64>)
declare <vscale x 16 x i8> @llvm.aarch64.sve.rsubhnb.nxv8i16(<vscale x 8 x i16>, <vscale x 8 x i16>)
declare <vscale x 8 x i16> @llvm.aarch64.sve.rsubhnb.nxv4i32(<vscale x 4 x i32>, <vscale x 4 x i32>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.rsubhnb.nxv2i64(<vscale x 2 x i64>, <vscale x 2 x i64>)
declare <vscale x 16 x i8> @llvm.aarch64.sve.rsubhnt.nxv8i16(<vscale x 16 x i8>, <vscale x 8 x i16>, <vscale x 8 x i16>)
declare <vscale x 8 x i16> @llvm.aarch64.sve.rsubhnt.nxv4i32(<vscale x 8 x i16>, <vscale x 4 x i32>, <vscale x 4 x i32>)
declare <vscale x 4 x i32> @llvm.aarch64.sve.rsubhnt.nxv2i64(<vscale x 4 x i32>, <vscale x 2 x i64>, <vscale x 2 x i64>)