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
|
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
; RUN: opt -instcombine -S -mtriple=thumbv8.1m.main-none-eabi -mattr=+mve.fp -o - %s | FileCheck %s
target datalayout = "e-m:e-p:32:32-Fi8-i64:64-v128:64:128-a:0:32-n32-S64"
; Various patterns testing v8i16
define <8 x i16> @test_shrn_v8i16_t1(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d) {
; CHECK-LABEL: @test_shrn_v8i16_t1(
; CHECK-NEXT: [[X:%.*]] = add <8 x i16> [[A:%.*]], <i16 1, i16 poison, i16 1, i16 poison, i16 1, i16 poison, i16 1, i16 poison>
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[X]], <4 x i32> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, <i16 1, i16 -1, i16 1, i16 -1, i16 1, i16 -1, i16 1, i16 -1>
%z = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %x, <4 x i32> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
ret <8 x i16> %z
}
define <8 x i16> @test_shrn_v8i16_t2(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d) {
; CHECK-LABEL: @test_shrn_v8i16_t2(
; CHECK-NEXT: [[X:%.*]] = add <8 x i16> [[A:%.*]], <i16 -1, i16 poison, i16 -1, i16 poison, i16 -1, i16 poison, i16 -1, i16 poison>
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[X]], <4 x i32> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, <i16 -1, i16 1, i16 -1, i16 1, i16 -1, i16 1, i16 -1, i16 1>
%z = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %x, <4 x i32> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
ret <8 x i16> %z
}
define <8 x i16> @test_shrn_v8i16_b1(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d) {
; CHECK-LABEL: @test_shrn_v8i16_b1(
; CHECK-NEXT: [[X:%.*]] = add <8 x i16> [[A:%.*]], <i16 poison, i16 -1, i16 poison, i16 -1, i16 poison, i16 -1, i16 poison, i16 -1>
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[X]], <4 x i32> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, <i16 1, i16 -1, i16 1, i16 -1, i16 1, i16 -1, i16 1, i16 -1>
%z = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %x, <4 x i32> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
ret <8 x i16> %z
}
define <8 x i16> @test_shrn_v8i16_b2(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d) {
; CHECK-LABEL: @test_shrn_v8i16_b2(
; CHECK-NEXT: [[X:%.*]] = add <8 x i16> [[A:%.*]], <i16 poison, i16 1, i16 poison, i16 1, i16 poison, i16 1, i16 poison, i16 1>
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[X]], <4 x i32> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, <i16 -1, i16 1, i16 -1, i16 1, i16 -1, i16 1, i16 -1, i16 1>
%z = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %x, <4 x i32> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
ret <8 x i16> %z
}
define <8 x i16> @test_shrn_v8i16_bt(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d) {
; CHECK-LABEL: @test_shrn_v8i16_bt(
; CHECK-NEXT: [[Y:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> poison, <4 x i32> [[C:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[Y]], <4 x i32> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, %b
%y = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %x, <4 x i32> %c, i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
%z = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %y, <4 x i32> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
ret <8 x i16> %z
}
define <8 x i16> @test_shrn_v8i16_tb(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d) {
; CHECK-LABEL: @test_shrn_v8i16_tb(
; CHECK-NEXT: [[Y:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> poison, <4 x i32> [[C:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[Y]], <4 x i32> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, %b
%y = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %x, <4 x i32> %c, i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
%z = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %y, <4 x i32> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
ret <8 x i16> %z
}
define <8 x i16> @test_shrn_v8i16_bb(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d) {
; CHECK-LABEL: @test_shrn_v8i16_bb(
; CHECK-NEXT: [[X:%.*]] = add <8 x i16> [[A:%.*]], [[B:%.*]]
; CHECK-NEXT: [[Y:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[X]], <4 x i32> [[C:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[Y]], <4 x i32> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, %b
%y = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %x, <4 x i32> %c, i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
%z = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %y, <4 x i32> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
ret <8 x i16> %z
}
define <8 x i16> @test_shrn_v8i16_tt(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d) {
; CHECK-LABEL: @test_shrn_v8i16_tt(
; CHECK-NEXT: [[X:%.*]] = add <8 x i16> [[A:%.*]], [[B:%.*]]
; CHECK-NEXT: [[Y:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[X]], <4 x i32> [[C:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> [[Y]], <4 x i32> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, %b
%y = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %x, <4 x i32> %c, i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
%z = call <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16> %y, <4 x i32> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
ret <8 x i16> %z
}
; Other types and sizes
define <16 x i8> @test_shrn_v16i8_bt(<16 x i8> %a, <16 x i8> %b, <8 x i16> %c, <8 x i16> %d) {
; CHECK-LABEL: @test_shrn_v16i8_bt(
; CHECK-NEXT: [[Y:%.*]] = call <16 x i8> @llvm.arm.mve.vshrn.v16i8.v8i16(<16 x i8> poison, <8 x i16> [[C:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
; CHECK-NEXT: [[Z:%.*]] = call <16 x i8> @llvm.arm.mve.vshrn.v16i8.v8i16(<16 x i8> [[Y]], <8 x i16> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
; CHECK-NEXT: ret <16 x i8> [[Z]]
;
%x = add <16 x i8> %a, %b
%y = call <16 x i8> @llvm.arm.mve.vshrn.v16i8.v8i16(<16 x i8> %x, <8 x i16> %c, i32 16, i32 0, i32 0, i32 0, i32 0, i32 0)
%z = call <16 x i8> @llvm.arm.mve.vshrn.v16i8.v8i16(<16 x i8> %y, <8 x i16> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 1)
ret <16 x i8> %z
}
define <8 x i16> @test_shrnp_v8i16_bt(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d, <4 x i1> %p) {
; CHECK-LABEL: @test_shrnp_v8i16_bt(
; CHECK-NEXT: [[X:%.*]] = add <8 x i16> [[A:%.*]], [[B:%.*]]
; CHECK-NEXT: [[Y:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.predicated.v8i16.v4i32.v4i1(<8 x i16> [[X]], <4 x i32> [[C:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 0, <4 x i1> [[P:%.*]])
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vshrn.predicated.v8i16.v4i32.v4i1(<8 x i16> [[Y]], <4 x i32> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 1, <4 x i1> [[P]])
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, %b
%y = call <8 x i16> @llvm.arm.mve.vshrn.predicated.v8i16.v4i32.v4i1(<8 x i16> %x, <4 x i32> %c, i32 16, i32 0, i32 0, i32 0, i32 0, i32 0, <4 x i1> %p)
%z = call <8 x i16> @llvm.arm.mve.vshrn.predicated.v8i16.v4i32.v4i1(<8 x i16> %y, <4 x i32> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 1, <4 x i1> %p)
ret <8 x i16> %z
}
define <16 x i8> @test_shrnp_v16i8_bt(<16 x i8> %a, <16 x i8> %b, <8 x i16> %c, <8 x i16> %d, <8 x i1> %p) {
; CHECK-LABEL: @test_shrnp_v16i8_bt(
; CHECK-NEXT: [[X:%.*]] = add <16 x i8> [[A:%.*]], [[B:%.*]]
; CHECK-NEXT: [[Y:%.*]] = call <16 x i8> @llvm.arm.mve.vshrn.predicated.v16i8.v8i16.v8i1(<16 x i8> [[X]], <8 x i16> [[C:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 0, <8 x i1> [[P:%.*]])
; CHECK-NEXT: [[Z:%.*]] = call <16 x i8> @llvm.arm.mve.vshrn.predicated.v16i8.v8i16.v8i1(<16 x i8> [[Y]], <8 x i16> [[D:%.*]], i32 16, i32 0, i32 0, i32 0, i32 0, i32 1, <8 x i1> [[P]])
; CHECK-NEXT: ret <16 x i8> [[Z]]
;
%x = add <16 x i8> %a, %b
%y = call <16 x i8> @llvm.arm.mve.vshrn.predicated.v16i8.v8i16.v8i1(<16 x i8> %x, <8 x i16> %c, i32 16, i32 0, i32 0, i32 0, i32 0, i32 0, <8 x i1> %p)
%z = call <16 x i8> @llvm.arm.mve.vshrn.predicated.v16i8.v8i16.v8i1(<16 x i8> %y, <8 x i16> %d, i32 16, i32 0, i32 0, i32 0, i32 0, i32 1, <8 x i1> %p)
ret <16 x i8> %z
}
define <8 x i16> @test_movnp_v8i16_bt(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d, <4 x i1> %p) {
; CHECK-LABEL: @test_movnp_v8i16_bt(
; CHECK-NEXT: [[X:%.*]] = add <8 x i16> [[A:%.*]], [[B:%.*]]
; CHECK-NEXT: [[Y:%.*]] = call <8 x i16> @llvm.arm.mve.vmovn.predicated.v8i16.v4i32.v4i1(<8 x i16> [[X]], <4 x i32> [[C:%.*]], i32 0, <4 x i1> [[P:%.*]])
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vmovn.predicated.v8i16.v4i32.v4i1(<8 x i16> [[Y]], <4 x i32> [[D:%.*]], i32 1, <4 x i1> [[P]])
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, %b
%y = call <8 x i16> @llvm.arm.mve.vmovn.predicated.v8i16.v4i32.v4i1(<8 x i16> %x, <4 x i32> %c, i32 0, <4 x i1> %p)
%z = call <8 x i16> @llvm.arm.mve.vmovn.predicated.v8i16.v4i32.v4i1(<8 x i16> %y, <4 x i32> %d, i32 1, <4 x i1> %p)
ret <8 x i16> %z
}
define <16 x i8> @test_movnp_v16i8_bt(<16 x i8> %a, <16 x i8> %b, <8 x i16> %c, <8 x i16> %d, <8 x i1> %p) {
; CHECK-LABEL: @test_movnp_v16i8_bt(
; CHECK-NEXT: [[X:%.*]] = add <16 x i8> [[A:%.*]], [[B:%.*]]
; CHECK-NEXT: [[Y:%.*]] = call <16 x i8> @llvm.arm.mve.vmovn.predicated.v16i8.v8i16.v8i1(<16 x i8> [[X]], <8 x i16> [[C:%.*]], i32 0, <8 x i1> [[P:%.*]])
; CHECK-NEXT: [[Z:%.*]] = call <16 x i8> @llvm.arm.mve.vmovn.predicated.v16i8.v8i16.v8i1(<16 x i8> [[Y]], <8 x i16> [[D:%.*]], i32 1, <8 x i1> [[P]])
; CHECK-NEXT: ret <16 x i8> [[Z]]
;
%x = add <16 x i8> %a, %b
%y = call <16 x i8> @llvm.arm.mve.vmovn.predicated.v16i8.v8i16.v8i1(<16 x i8> %x, <8 x i16> %c, i32 0, <8 x i1> %p)
%z = call <16 x i8> @llvm.arm.mve.vmovn.predicated.v16i8.v8i16.v8i1(<16 x i8> %y, <8 x i16> %d, i32 1, <8 x i1> %p)
ret <16 x i8> %z
}
define <8 x i16> @test_qmovn_v8i16_bt(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d) {
; CHECK-LABEL: @test_qmovn_v8i16_bt(
; CHECK-NEXT: [[Y:%.*]] = call <8 x i16> @llvm.arm.mve.vqmovn.v8i16.v4i32(<8 x i16> poison, <4 x i32> [[C:%.*]], i32 0, i32 0, i32 0)
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vqmovn.v8i16.v4i32(<8 x i16> [[Y]], <4 x i32> [[D:%.*]], i32 0, i32 0, i32 1)
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, %b
%y = call <8 x i16> @llvm.arm.mve.vqmovn.v8i16.v4i32(<8 x i16> %x, <4 x i32> %c, i32 0, i32 0, i32 0)
%z = call <8 x i16> @llvm.arm.mve.vqmovn.v8i16.v4i32(<8 x i16> %y, <4 x i32> %d, i32 0, i32 0, i32 1)
ret <8 x i16> %z
}
define <16 x i8> @test_qmovn_v16i8_bt(<16 x i8> %a, <16 x i8> %b, <8 x i16> %c, <8 x i16> %d) {
; CHECK-LABEL: @test_qmovn_v16i8_bt(
; CHECK-NEXT: [[Y:%.*]] = call <16 x i8> @llvm.arm.mve.vqmovn.v16i8.v8i16(<16 x i8> poison, <8 x i16> [[C:%.*]], i32 0, i32 0, i32 0)
; CHECK-NEXT: [[Z:%.*]] = call <16 x i8> @llvm.arm.mve.vqmovn.v16i8.v8i16(<16 x i8> [[Y]], <8 x i16> [[D:%.*]], i32 0, i32 0, i32 1)
; CHECK-NEXT: ret <16 x i8> [[Z]]
;
%x = add <16 x i8> %a, %b
%y = call <16 x i8> @llvm.arm.mve.vqmovn.v16i8.v8i16(<16 x i8> %x, <8 x i16> %c, i32 0, i32 0, i32 0)
%z = call <16 x i8> @llvm.arm.mve.vqmovn.v16i8.v8i16(<16 x i8> %y, <8 x i16> %d, i32 0, i32 0, i32 1)
ret <16 x i8> %z
}
define <8 x i16> @test_qmovnp_v8i16_bt(<8 x i16> %a, <8 x i16> %b, <4 x i32> %c, <4 x i32> %d, <4 x i1> %p) {
; CHECK-LABEL: @test_qmovnp_v8i16_bt(
; CHECK-NEXT: [[X:%.*]] = add <8 x i16> [[A:%.*]], [[B:%.*]]
; CHECK-NEXT: [[Y:%.*]] = call <8 x i16> @llvm.arm.mve.vqmovn.predicated.v8i16.v4i32.v4i1(<8 x i16> [[X]], <4 x i32> [[C:%.*]], i32 0, i32 0, i32 0, <4 x i1> [[P:%.*]])
; CHECK-NEXT: [[Z:%.*]] = call <8 x i16> @llvm.arm.mve.vqmovn.predicated.v8i16.v4i32.v4i1(<8 x i16> [[Y]], <4 x i32> [[D:%.*]], i32 0, i32 0, i32 1, <4 x i1> [[P]])
; CHECK-NEXT: ret <8 x i16> [[Z]]
;
%x = add <8 x i16> %a, %b
%y = call <8 x i16> @llvm.arm.mve.vqmovn.predicated.v8i16.v4i32.v4i1(<8 x i16> %x, <4 x i32> %c, i32 0, i32 0, i32 0, <4 x i1> %p)
%z = call <8 x i16> @llvm.arm.mve.vqmovn.predicated.v8i16.v4i32.v4i1(<8 x i16> %y, <4 x i32> %d, i32 0, i32 0, i32 1, <4 x i1> %p)
ret <8 x i16> %z
}
define <16 x i8> @test_qmovnp_v16i8_bt(<16 x i8> %a, <16 x i8> %b, <8 x i16> %c, <8 x i16> %d, <8 x i1> %p) {
; CHECK-LABEL: @test_qmovnp_v16i8_bt(
; CHECK-NEXT: [[X:%.*]] = add <16 x i8> [[A:%.*]], [[B:%.*]]
; CHECK-NEXT: [[Y:%.*]] = call <16 x i8> @llvm.arm.mve.vqmovn.predicated.v16i8.v8i16.v8i1(<16 x i8> [[X]], <8 x i16> [[C:%.*]], i32 0, i32 0, i32 0, <8 x i1> [[P:%.*]])
; CHECK-NEXT: [[Z:%.*]] = call <16 x i8> @llvm.arm.mve.vqmovn.predicated.v16i8.v8i16.v8i1(<16 x i8> [[Y]], <8 x i16> [[D:%.*]], i32 0, i32 0, i32 1, <8 x i1> [[P]])
; CHECK-NEXT: ret <16 x i8> [[Z]]
;
%x = add <16 x i8> %a, %b
%y = call <16 x i8> @llvm.arm.mve.vqmovn.predicated.v16i8.v8i16.v8i1(<16 x i8> %x, <8 x i16> %c, i32 0, i32 0, i32 0, <8 x i1> %p)
%z = call <16 x i8> @llvm.arm.mve.vqmovn.predicated.v16i8.v8i16.v8i1(<16 x i8> %y, <8 x i16> %d, i32 0, i32 0, i32 1, <8 x i1> %p)
ret <16 x i8> %z
}
define <8 x half> @test_cvtn_v8i16_bt(<8 x half> %a, <8 x half> %b, <4 x float> %c, <4 x float> %d) {
; CHECK-LABEL: @test_cvtn_v8i16_bt(
; CHECK-NEXT: [[Y:%.*]] = call <8 x half> @llvm.arm.mve.vcvt.narrow(<8 x half> poison, <4 x float> [[C:%.*]], i32 0)
; CHECK-NEXT: [[Z:%.*]] = call <8 x half> @llvm.arm.mve.vcvt.narrow(<8 x half> [[Y]], <4 x float> [[D:%.*]], i32 1)
; CHECK-NEXT: ret <8 x half> [[Z]]
;
%x = fadd <8 x half> %a, %b
%y = call <8 x half> @llvm.arm.mve.vcvt.narrow(<8 x half> %x, <4 x float> %c, i32 0)
%z = call <8 x half> @llvm.arm.mve.vcvt.narrow(<8 x half> %y, <4 x float> %d, i32 1)
ret <8 x half> %z
}
define <8 x half> @test_cvtnp_v8i16_bt(<8 x half> %a, <8 x half> %b, <4 x float> %c, <4 x float> %d, <4 x i1> %p) {
; CHECK-LABEL: @test_cvtnp_v8i16_bt(
; CHECK-NEXT: [[X:%.*]] = fadd <8 x half> [[A:%.*]], [[B:%.*]]
; CHECK-NEXT: [[Y:%.*]] = call <8 x half> @llvm.arm.mve.vcvt.narrow.predicated(<8 x half> [[X]], <4 x float> [[C:%.*]], i32 0, <4 x i1> [[P:%.*]])
; CHECK-NEXT: [[Z:%.*]] = call <8 x half> @llvm.arm.mve.vcvt.narrow.predicated(<8 x half> [[Y]], <4 x float> [[D:%.*]], i32 1, <4 x i1> [[P]])
; CHECK-NEXT: ret <8 x half> [[Z]]
;
%x = fadd <8 x half> %a, %b
%y = call <8 x half> @llvm.arm.mve.vcvt.narrow.predicated(<8 x half> %x, <4 x float> %c, i32 0, <4 x i1> %p)
%z = call <8 x half> @llvm.arm.mve.vcvt.narrow.predicated(<8 x half> %y, <4 x float> %d, i32 1, <4 x i1> %p)
ret <8 x half> %z
}
declare <8 x i16> @llvm.arm.mve.vshrn.v8i16.v4i32(<8 x i16>, <4 x i32>, i32, i32, i32, i32, i32, i32)
declare <8 x i16> @llvm.arm.mve.vshrn.predicated.v8i16.v4i32.v4i1(<8 x i16>, <4 x i32>, i32, i32, i32, i32, i32, i32, <4 x i1>)
declare <16 x i8> @llvm.arm.mve.vshrn.v16i8.v8i16(<16 x i8>, <8 x i16>, i32, i32, i32, i32, i32, i32)
declare <16 x i8> @llvm.arm.mve.vshrn.predicated.v16i8.v8i16.v8i1(<16 x i8>, <8 x i16>, i32, i32, i32, i32, i32, i32, <8 x i1>)
declare <8 x i16> @llvm.arm.mve.vmovn.predicated.v8i16.v4i32.v4i1(<8 x i16>, <4 x i32>, i32, <4 x i1>)
declare <16 x i8> @llvm.arm.mve.vmovn.predicated.v16i8.v8i16.v8i1(<16 x i8>, <8 x i16>, i32, <8 x i1>)
declare <8 x i16> @llvm.arm.mve.vqmovn.v8i16.v4i32(<8 x i16>, <4 x i32>, i32, i32, i32)
declare <8 x i16> @llvm.arm.mve.vqmovn.predicated.v8i16.v4i32.v4i1(<8 x i16>, <4 x i32>, i32, i32, i32, <4 x i1>)
declare <16 x i8> @llvm.arm.mve.vqmovn.v16i8.v8i16(<16 x i8>, <8 x i16>, i32, i32, i32)
declare <16 x i8> @llvm.arm.mve.vqmovn.predicated.v16i8.v8i16.v8i1(<16 x i8>, <8 x i16>, i32, i32, i32, <8 x i1>)
declare <8 x half> @llvm.arm.mve.vcvt.narrow(<8 x half>, <4 x float>, i32)
declare <8 x half> @llvm.arm.mve.vcvt.narrow.predicated(<8 x half>, <4 x float>, i32, <4 x i1>)
|
