File: legalize-llvm.amdgcn.image.store.2d.d16.ll

package info (click to toggle)
swiftlang 6.0.3-2
  • links: PTS, VCS
  • area: main
  • in suites: forky, sid, trixie
  • size: 2,519,992 kB
  • sloc: cpp: 9,107,863; ansic: 2,040,022; asm: 1,135,751; python: 296,500; objc: 82,456; f90: 60,502; lisp: 34,951; pascal: 19,946; sh: 18,133; perl: 7,482; ml: 4,937; javascript: 4,117; makefile: 3,840; awk: 3,535; xml: 914; fortran: 619; cs: 573; ruby: 573
file content (415 lines) | stat: -rw-r--r-- 31,432 bytes parent folder | download | duplicates (2) 
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
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
 
; NOTE: Assertions have been autogenerated by utils/update_mir_test_checks.py
; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=tonga -stop-after=legalizer -o - %s | FileCheck -check-prefix=UNPACKED %s
; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx810 -stop-after=legalizer -o - %s | FileCheck -check-prefix=GFX81 %s
; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx900 -stop-after=legalizer -o - %s | FileCheck -check-prefix=GFX9 %s
; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1010 -stop-after=legalizer -o - %s | FileCheck -check-prefix=GFX10 %s
; RUN: llc -global-isel -mtriple=amdgcn-mesa-mesa3d -mcpu=gfx1100 -stop-after=legalizer -o - %s | FileCheck -check-prefix=GFX10 %s

define amdgpu_ps void @image_store_f16(<8 x i32> inreg %rsrc, i32 %s, i32 %t, half %data) {
  ; UNPACKED-LABEL: name: image_store_f16
  ; UNPACKED: bb.1 (%ir-block.0):
  ; UNPACKED-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2
  ; UNPACKED-NEXT: {{  $}}
  ; UNPACKED-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; UNPACKED-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; UNPACKED-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; UNPACKED-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; UNPACKED-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; UNPACKED-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; UNPACKED-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; UNPACKED-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; UNPACKED-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; UNPACKED-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; UNPACKED-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; UNPACKED-NEXT:   [[COPY10:%[0-9]+]]:_(s32) = COPY $vgpr2
  ; UNPACKED-NEXT:   [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[COPY10]](s32)
  ; UNPACKED-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; UNPACKED-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[TRUNC]](s16), 1, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (s16), addrspace 8)
  ; UNPACKED-NEXT:   S_ENDPGM 0
  ; GFX81-LABEL: name: image_store_f16
  ; GFX81: bb.1 (%ir-block.0):
  ; GFX81-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2
  ; GFX81-NEXT: {{  $}}
  ; GFX81-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX81-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX81-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX81-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX81-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX81-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX81-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX81-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX81-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX81-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX81-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX81-NEXT:   [[COPY10:%[0-9]+]]:_(s32) = COPY $vgpr2
  ; GFX81-NEXT:   [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[COPY10]](s32)
  ; GFX81-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX81-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[TRUNC]](s16), 1, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (s16), addrspace 8)
  ; GFX81-NEXT:   S_ENDPGM 0
  ; GFX9-LABEL: name: image_store_f16
  ; GFX9: bb.1 (%ir-block.0):
  ; GFX9-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2
  ; GFX9-NEXT: {{  $}}
  ; GFX9-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX9-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX9-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX9-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX9-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX9-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX9-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX9-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX9-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX9-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX9-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX9-NEXT:   [[COPY10:%[0-9]+]]:_(s32) = COPY $vgpr2
  ; GFX9-NEXT:   [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[COPY10]](s32)
  ; GFX9-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX9-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[TRUNC]](s16), 1, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (s16), addrspace 8)
  ; GFX9-NEXT:   S_ENDPGM 0
  ; GFX10-LABEL: name: image_store_f16
  ; GFX10: bb.1 (%ir-block.0):
  ; GFX10-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2
  ; GFX10-NEXT: {{  $}}
  ; GFX10-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX10-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX10-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX10-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX10-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX10-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX10-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX10-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX10-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX10-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX10-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX10-NEXT:   [[COPY10:%[0-9]+]]:_(s32) = COPY $vgpr2
  ; GFX10-NEXT:   [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[COPY10]](s32)
  ; GFX10-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX10-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[TRUNC]](s16), 1, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (s16), addrspace 8)
  ; GFX10-NEXT:   S_ENDPGM 0
  call void @llvm.amdgcn.image.store.2d.f16.i32(half %data, i32 1, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
  ret void
}

define amdgpu_ps void @image_store_v2f16(<8 x i32> inreg %rsrc, i32 %s, i32 %t, <2 x half> %in) {
  ; UNPACKED-LABEL: name: image_store_v2f16
  ; UNPACKED: bb.1 (%ir-block.0):
  ; UNPACKED-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2
  ; UNPACKED-NEXT: {{  $}}
  ; UNPACKED-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; UNPACKED-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; UNPACKED-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; UNPACKED-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; UNPACKED-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; UNPACKED-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; UNPACKED-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; UNPACKED-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; UNPACKED-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; UNPACKED-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; UNPACKED-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; UNPACKED-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; UNPACKED-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; UNPACKED-NEXT:   [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[COPY10]](<2 x s16>)
  ; UNPACKED-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
  ; UNPACKED-NEXT:   [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C]](s32)
  ; UNPACKED-NEXT:   [[BUILD_VECTOR2:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[BITCAST]](s32), [[LSHR]](s32)
  ; UNPACKED-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[BUILD_VECTOR2]](<2 x s32>), 3, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<2 x s16>), addrspace 8)
  ; UNPACKED-NEXT:   S_ENDPGM 0
  ; GFX81-LABEL: name: image_store_v2f16
  ; GFX81: bb.1 (%ir-block.0):
  ; GFX81-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2
  ; GFX81-NEXT: {{  $}}
  ; GFX81-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX81-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX81-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX81-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX81-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX81-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX81-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX81-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX81-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX81-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX81-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX81-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; GFX81-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX81-NEXT:   [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[COPY10]](<2 x s16>)
  ; GFX81-NEXT:   [[DEF:%[0-9]+]]:_(s32) = G_IMPLICIT_DEF
  ; GFX81-NEXT:   [[BUILD_VECTOR2:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[BITCAST]](s32), [[DEF]](s32)
  ; GFX81-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[BUILD_VECTOR2]](<2 x s32>), 3, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<2 x s16>), addrspace 8)
  ; GFX81-NEXT:   S_ENDPGM 0
  ; GFX9-LABEL: name: image_store_v2f16
  ; GFX9: bb.1 (%ir-block.0):
  ; GFX9-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2
  ; GFX9-NEXT: {{  $}}
  ; GFX9-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX9-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX9-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX9-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX9-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX9-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX9-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX9-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX9-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX9-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX9-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX9-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; GFX9-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX9-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[COPY10]](<2 x s16>), 3, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<2 x s16>), addrspace 8)
  ; GFX9-NEXT:   S_ENDPGM 0
  ; GFX10-LABEL: name: image_store_v2f16
  ; GFX10: bb.1 (%ir-block.0):
  ; GFX10-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2
  ; GFX10-NEXT: {{  $}}
  ; GFX10-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX10-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX10-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX10-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX10-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX10-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX10-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX10-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX10-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX10-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX10-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX10-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; GFX10-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX10-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[COPY10]](<2 x s16>), 3, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<2 x s16>), addrspace 8)
  ; GFX10-NEXT:   S_ENDPGM 0
  call void @llvm.amdgcn.image.store.2d.v2f16.i32(<2 x half> %in, i32 3, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
  ret void
}

define amdgpu_ps void @image_store_v3f16(<8 x i32> inreg %rsrc, i32 %s, i32 %t, <3 x half> %in) {
  ; UNPACKED-LABEL: name: image_store_v3f16
  ; UNPACKED: bb.1 (%ir-block.0):
  ; UNPACKED-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2, $vgpr3
  ; UNPACKED-NEXT: {{  $}}
  ; UNPACKED-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; UNPACKED-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; UNPACKED-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; UNPACKED-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; UNPACKED-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; UNPACKED-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; UNPACKED-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; UNPACKED-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; UNPACKED-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; UNPACKED-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; UNPACKED-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; UNPACKED-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; UNPACKED-NEXT:   [[COPY11:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr3
  ; UNPACKED-NEXT:   [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[COPY10]](<2 x s16>)
  ; UNPACKED-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
  ; UNPACKED-NEXT:   [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C]](s32)
  ; UNPACKED-NEXT:   [[BITCAST1:%[0-9]+]]:_(s32) = G_BITCAST [[COPY11]](<2 x s16>)
  ; UNPACKED-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; UNPACKED-NEXT:   [[BUILD_VECTOR2:%[0-9]+]]:_(<3 x s32>) = G_BUILD_VECTOR [[BITCAST]](s32), [[LSHR]](s32), [[BITCAST1]](s32)
  ; UNPACKED-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[BUILD_VECTOR2]](<3 x s32>), 7, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<3 x s16>), align 8, addrspace 8)
  ; UNPACKED-NEXT:   S_ENDPGM 0
  ; GFX81-LABEL: name: image_store_v3f16
  ; GFX81: bb.1 (%ir-block.0):
  ; GFX81-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2, $vgpr3
  ; GFX81-NEXT: {{  $}}
  ; GFX81-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX81-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX81-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX81-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX81-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX81-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX81-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX81-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX81-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX81-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX81-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX81-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; GFX81-NEXT:   [[COPY11:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr3
  ; GFX81-NEXT:   [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[COPY10]](<2 x s16>)
  ; GFX81-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
  ; GFX81-NEXT:   [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C]](s32)
  ; GFX81-NEXT:   [[BITCAST1:%[0-9]+]]:_(s32) = G_BITCAST [[COPY11]](<2 x s16>)
  ; GFX81-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX81-NEXT:   [[C1:%[0-9]+]]:_(s32) = G_CONSTANT i32 65535
  ; GFX81-NEXT:   [[AND:%[0-9]+]]:_(s32) = G_AND [[BITCAST]], [[C1]]
  ; GFX81-NEXT:   [[AND1:%[0-9]+]]:_(s32) = G_AND [[LSHR]], [[C1]]
  ; GFX81-NEXT:   [[SHL:%[0-9]+]]:_(s32) = G_SHL [[AND1]], [[C]](s32)
  ; GFX81-NEXT:   [[OR:%[0-9]+]]:_(s32) = G_OR [[AND]], [[SHL]]
  ; GFX81-NEXT:   [[BITCAST2:%[0-9]+]]:_(<2 x s16>) = G_BITCAST [[OR]](s32)
  ; GFX81-NEXT:   [[AND2:%[0-9]+]]:_(s32) = G_AND [[BITCAST1]], [[C1]]
  ; GFX81-NEXT:   [[C2:%[0-9]+]]:_(s32) = G_CONSTANT i32 0
  ; GFX81-NEXT:   [[SHL1:%[0-9]+]]:_(s32) = G_SHL [[C2]], [[C]](s32)
  ; GFX81-NEXT:   [[OR1:%[0-9]+]]:_(s32) = G_OR [[AND2]], [[SHL1]]
  ; GFX81-NEXT:   [[BITCAST3:%[0-9]+]]:_(<2 x s16>) = G_BITCAST [[OR1]](s32)
  ; GFX81-NEXT:   [[OR2:%[0-9]+]]:_(s32) = G_OR [[C2]], [[SHL1]]
  ; GFX81-NEXT:   [[BITCAST4:%[0-9]+]]:_(<2 x s16>) = G_BITCAST [[OR2]](s32)
  ; GFX81-NEXT:   [[CONCAT_VECTORS:%[0-9]+]]:_(<6 x s16>) = G_CONCAT_VECTORS [[BITCAST2]](<2 x s16>), [[BITCAST3]](<2 x s16>), [[BITCAST4]](<2 x s16>)
  ; GFX81-NEXT:   [[BITCAST5:%[0-9]+]]:_(<3 x s32>) = G_BITCAST [[CONCAT_VECTORS]](<6 x s16>)
  ; GFX81-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[BITCAST5]](<3 x s32>), 7, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<3 x s16>), align 8, addrspace 8)
  ; GFX81-NEXT:   S_ENDPGM 0
  ; GFX9-LABEL: name: image_store_v3f16
  ; GFX9: bb.1 (%ir-block.0):
  ; GFX9-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2, $vgpr3
  ; GFX9-NEXT: {{  $}}
  ; GFX9-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX9-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX9-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX9-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX9-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX9-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX9-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX9-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX9-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX9-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX9-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX9-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; GFX9-NEXT:   [[COPY11:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr3
  ; GFX9-NEXT:   [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[COPY10]](<2 x s16>)
  ; GFX9-NEXT:   [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[BITCAST]](s32)
  ; GFX9-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
  ; GFX9-NEXT:   [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C]](s32)
  ; GFX9-NEXT:   [[TRUNC1:%[0-9]+]]:_(s16) = G_TRUNC [[LSHR]](s32)
  ; GFX9-NEXT:   [[BITCAST1:%[0-9]+]]:_(s32) = G_BITCAST [[COPY11]](<2 x s16>)
  ; GFX9-NEXT:   [[TRUNC2:%[0-9]+]]:_(s16) = G_TRUNC [[BITCAST1]](s32)
  ; GFX9-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX9-NEXT:   [[DEF:%[0-9]+]]:_(s16) = G_IMPLICIT_DEF
  ; GFX9-NEXT:   [[BUILD_VECTOR2:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR [[TRUNC]](s16), [[TRUNC1]](s16)
  ; GFX9-NEXT:   [[BUILD_VECTOR3:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR [[TRUNC2]](s16), [[DEF]](s16)
  ; GFX9-NEXT:   [[CONCAT_VECTORS:%[0-9]+]]:_(<4 x s16>) = G_CONCAT_VECTORS [[BUILD_VECTOR2]](<2 x s16>), [[BUILD_VECTOR3]](<2 x s16>)
  ; GFX9-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[CONCAT_VECTORS]](<4 x s16>), 7, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<3 x s16>), align 8, addrspace 8)
  ; GFX9-NEXT:   S_ENDPGM 0
  ; GFX10-LABEL: name: image_store_v3f16
  ; GFX10: bb.1 (%ir-block.0):
  ; GFX10-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2, $vgpr3
  ; GFX10-NEXT: {{  $}}
  ; GFX10-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX10-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX10-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX10-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX10-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX10-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX10-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX10-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX10-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX10-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX10-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX10-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; GFX10-NEXT:   [[COPY11:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr3
  ; GFX10-NEXT:   [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[COPY10]](<2 x s16>)
  ; GFX10-NEXT:   [[TRUNC:%[0-9]+]]:_(s16) = G_TRUNC [[BITCAST]](s32)
  ; GFX10-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
  ; GFX10-NEXT:   [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C]](s32)
  ; GFX10-NEXT:   [[TRUNC1:%[0-9]+]]:_(s16) = G_TRUNC [[LSHR]](s32)
  ; GFX10-NEXT:   [[BITCAST1:%[0-9]+]]:_(s32) = G_BITCAST [[COPY11]](<2 x s16>)
  ; GFX10-NEXT:   [[TRUNC2:%[0-9]+]]:_(s16) = G_TRUNC [[BITCAST1]](s32)
  ; GFX10-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX10-NEXT:   [[DEF:%[0-9]+]]:_(s16) = G_IMPLICIT_DEF
  ; GFX10-NEXT:   [[BUILD_VECTOR2:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR [[TRUNC]](s16), [[TRUNC1]](s16)
  ; GFX10-NEXT:   [[BUILD_VECTOR3:%[0-9]+]]:_(<2 x s16>) = G_BUILD_VECTOR [[TRUNC2]](s16), [[DEF]](s16)
  ; GFX10-NEXT:   [[CONCAT_VECTORS:%[0-9]+]]:_(<4 x s16>) = G_CONCAT_VECTORS [[BUILD_VECTOR2]](<2 x s16>), [[BUILD_VECTOR3]](<2 x s16>)
  ; GFX10-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[CONCAT_VECTORS]](<4 x s16>), 7, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<3 x s16>), align 8, addrspace 8)
  ; GFX10-NEXT:   S_ENDPGM 0
  call void @llvm.amdgcn.image.store.2d.v3f16.i32(<3 x half> %in, i32 7, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
  ret void
}

define amdgpu_ps void @image_store_v4f16(<8 x i32> inreg %rsrc, i32 %s, i32 %t, <4 x half> %in) {
  ; UNPACKED-LABEL: name: image_store_v4f16
  ; UNPACKED: bb.1 (%ir-block.0):
  ; UNPACKED-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2, $vgpr3
  ; UNPACKED-NEXT: {{  $}}
  ; UNPACKED-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; UNPACKED-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; UNPACKED-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; UNPACKED-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; UNPACKED-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; UNPACKED-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; UNPACKED-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; UNPACKED-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; UNPACKED-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; UNPACKED-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; UNPACKED-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; UNPACKED-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; UNPACKED-NEXT:   [[COPY11:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr3
  ; UNPACKED-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; UNPACKED-NEXT:   [[BITCAST:%[0-9]+]]:_(s32) = G_BITCAST [[COPY10]](<2 x s16>)
  ; UNPACKED-NEXT:   [[C:%[0-9]+]]:_(s32) = G_CONSTANT i32 16
  ; UNPACKED-NEXT:   [[LSHR:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST]], [[C]](s32)
  ; UNPACKED-NEXT:   [[BITCAST1:%[0-9]+]]:_(s32) = G_BITCAST [[COPY11]](<2 x s16>)
  ; UNPACKED-NEXT:   [[LSHR1:%[0-9]+]]:_(s32) = G_LSHR [[BITCAST1]], [[C]](s32)
  ; UNPACKED-NEXT:   [[BUILD_VECTOR2:%[0-9]+]]:_(<4 x s32>) = G_BUILD_VECTOR [[BITCAST]](s32), [[LSHR]](s32), [[BITCAST1]](s32), [[LSHR1]](s32)
  ; UNPACKED-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[BUILD_VECTOR2]](<4 x s32>), 15, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<4 x s16>), addrspace 8)
  ; UNPACKED-NEXT:   S_ENDPGM 0
  ; GFX81-LABEL: name: image_store_v4f16
  ; GFX81: bb.1 (%ir-block.0):
  ; GFX81-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2, $vgpr3
  ; GFX81-NEXT: {{  $}}
  ; GFX81-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX81-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX81-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX81-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX81-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX81-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX81-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX81-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX81-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX81-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX81-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX81-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; GFX81-NEXT:   [[COPY11:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr3
  ; GFX81-NEXT:   [[CONCAT_VECTORS:%[0-9]+]]:_(<4 x s16>) = G_CONCAT_VECTORS [[COPY10]](<2 x s16>), [[COPY11]](<2 x s16>)
  ; GFX81-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX81-NEXT:   [[BITCAST:%[0-9]+]]:_(<2 x s32>) = G_BITCAST [[CONCAT_VECTORS]](<4 x s16>)
  ; GFX81-NEXT:   [[UV:%[0-9]+]]:_(s32), [[UV1:%[0-9]+]]:_(s32) = G_UNMERGE_VALUES [[BITCAST]](<2 x s32>)
  ; GFX81-NEXT:   [[DEF:%[0-9]+]]:_(s32) = G_IMPLICIT_DEF
  ; GFX81-NEXT:   [[BUILD_VECTOR2:%[0-9]+]]:_(<4 x s32>) = G_BUILD_VECTOR [[UV]](s32), [[UV1]](s32), [[DEF]](s32), [[DEF]](s32)
  ; GFX81-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[BUILD_VECTOR2]](<4 x s32>), 15, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<4 x s16>), addrspace 8)
  ; GFX81-NEXT:   S_ENDPGM 0
  ; GFX9-LABEL: name: image_store_v4f16
  ; GFX9: bb.1 (%ir-block.0):
  ; GFX9-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2, $vgpr3
  ; GFX9-NEXT: {{  $}}
  ; GFX9-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX9-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX9-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX9-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX9-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX9-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX9-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX9-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX9-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX9-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX9-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX9-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; GFX9-NEXT:   [[COPY11:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr3
  ; GFX9-NEXT:   [[CONCAT_VECTORS:%[0-9]+]]:_(<4 x s16>) = G_CONCAT_VECTORS [[COPY10]](<2 x s16>), [[COPY11]](<2 x s16>)
  ; GFX9-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX9-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[CONCAT_VECTORS]](<4 x s16>), 15, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<4 x s16>), addrspace 8)
  ; GFX9-NEXT:   S_ENDPGM 0
  ; GFX10-LABEL: name: image_store_v4f16
  ; GFX10: bb.1 (%ir-block.0):
  ; GFX10-NEXT:   liveins: $sgpr2, $sgpr3, $sgpr4, $sgpr5, $sgpr6, $sgpr7, $sgpr8, $sgpr9, $vgpr0, $vgpr1, $vgpr2, $vgpr3
  ; GFX10-NEXT: {{  $}}
  ; GFX10-NEXT:   [[COPY:%[0-9]+]]:_(s32) = COPY $sgpr2
  ; GFX10-NEXT:   [[COPY1:%[0-9]+]]:_(s32) = COPY $sgpr3
  ; GFX10-NEXT:   [[COPY2:%[0-9]+]]:_(s32) = COPY $sgpr4
  ; GFX10-NEXT:   [[COPY3:%[0-9]+]]:_(s32) = COPY $sgpr5
  ; GFX10-NEXT:   [[COPY4:%[0-9]+]]:_(s32) = COPY $sgpr6
  ; GFX10-NEXT:   [[COPY5:%[0-9]+]]:_(s32) = COPY $sgpr7
  ; GFX10-NEXT:   [[COPY6:%[0-9]+]]:_(s32) = COPY $sgpr8
  ; GFX10-NEXT:   [[COPY7:%[0-9]+]]:_(s32) = COPY $sgpr9
  ; GFX10-NEXT:   [[BUILD_VECTOR:%[0-9]+]]:_(<8 x s32>) = G_BUILD_VECTOR [[COPY]](s32), [[COPY1]](s32), [[COPY2]](s32), [[COPY3]](s32), [[COPY4]](s32), [[COPY5]](s32), [[COPY6]](s32), [[COPY7]](s32)
  ; GFX10-NEXT:   [[COPY8:%[0-9]+]]:_(s32) = COPY $vgpr0
  ; GFX10-NEXT:   [[COPY9:%[0-9]+]]:_(s32) = COPY $vgpr1
  ; GFX10-NEXT:   [[COPY10:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr2
  ; GFX10-NEXT:   [[COPY11:%[0-9]+]]:_(<2 x s16>) = COPY $vgpr3
  ; GFX10-NEXT:   [[CONCAT_VECTORS:%[0-9]+]]:_(<4 x s16>) = G_CONCAT_VECTORS [[COPY10]](<2 x s16>), [[COPY11]](<2 x s16>)
  ; GFX10-NEXT:   [[BUILD_VECTOR1:%[0-9]+]]:_(<2 x s32>) = G_BUILD_VECTOR [[COPY8]](s32), [[COPY9]](s32)
  ; GFX10-NEXT:   G_AMDGPU_INTRIN_IMAGE_STORE_D16 intrinsic(@llvm.amdgcn.image.store.2d), [[CONCAT_VECTORS]](<4 x s16>), 15, [[BUILD_VECTOR1]](<2 x s32>), $noreg, [[BUILD_VECTOR]](<8 x s32>), 0, 0, 0 :: (dereferenceable store (<4 x s16>), addrspace 8)
  ; GFX10-NEXT:   S_ENDPGM 0
  call void @llvm.amdgcn.image.store.2d.v4f16.i32(<4 x half> %in, i32 15, i32 %s, i32 %t, <8 x i32> %rsrc, i32 0, i32 0)
  ret void
}

declare void @llvm.amdgcn.image.store.2d.f16.i32(half, i32 immarg, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0
declare void @llvm.amdgcn.image.store.2d.v2f16.i32(<2 x half>, i32 immarg, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0
declare void @llvm.amdgcn.image.store.2d.v3f16.i32(<3 x half>, i32 immarg, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0
declare void @llvm.amdgcn.image.store.2d.v4f16.i32(<4 x half>, i32 immarg, i32, i32, <8 x i32>, i32 immarg, i32 immarg) #0

attributes #0 = { nounwind writeonly }