llvm.org GIT mirror llvm / stable test / CodeGen / AMDGPU / max.i16.ll
stable

Tree @stable (Download .tar.gz)

max.i16.ll @stableraw · history · blame

  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
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -mtriple=amdgcn-- -mcpu=fiji -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck %s -check-prefixes=GCN,VIPLUS,VI
; RUN: llc -mtriple=amdgcn-- -mcpu=gfx900 -mattr=-flat-for-global -verify-machineinstrs < %s | FileCheck %s -check-prefixes=GCN,VIPLUS,GFX9

; FIXME: Need to handle non-uniform case for function below (load without gep).
define amdgpu_kernel void @v_test_imax_sge_i16(i16 addrspace(1)* %out, i16 addrspace(1)* %aptr, i16 addrspace(1)* %bptr) nounwind {
; VI-LABEL: v_test_imax_sge_i16:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; VI-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; VI-NEXT:    v_lshlrev_b32_e32 v4, 1, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s7
; VI-NEXT:    v_add_u32_e32 v0, vcc, s6, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_add_u32_e32 v2, vcc, s0, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_load_ushort v2, v[2:3]
; VI-NEXT:    flat_load_ushort v3, v[0:1]
; VI-NEXT:    v_mov_b32_e32 v1, s5
; VI-NEXT:    v_add_u32_e32 v0, vcc, s4, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    v_max_i16_e32 v2, v3, v2
; VI-NEXT:    flat_store_short v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: v_test_imax_sge_i16:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; GFX9-NEXT:    v_lshlrev_b32_e32 v4, 1, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s7
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s6, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_add_co_u32_e32 v2, vcc, s0, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v3, vcc, 0, v3, vcc
; GFX9-NEXT:    global_load_ushort v2, v[2:3], off
; GFX9-NEXT:    global_load_ushort v3, v[0:1], off
; GFX9-NEXT:    v_mov_b32_e32 v1, s5
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s4, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    v_max_i16_e32 v2, v3, v2
; GFX9-NEXT:    global_store_short v[0:1], v2, off
; GFX9-NEXT:    s_endpgm
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr i16, i16 addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr i16, i16 addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr i16, i16 addrspace(1)* %out, i32 %tid
  %a = load i16, i16 addrspace(1)* %gep0, align 4
  %b = load i16, i16 addrspace(1)* %gep1, align 4
  %cmp = icmp sge i16 %a, %b
  %val = select i1 %cmp, i16 %a, i16 %b
  store i16 %val, i16 addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
define amdgpu_kernel void @v_test_imax_sge_v2i16(<2 x i16> addrspace(1)* %out, <2 x i16> addrspace(1)* %aptr, <2 x i16> addrspace(1)* %bptr) nounwind {
; VI-LABEL: v_test_imax_sge_v2i16:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; VI-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; VI-NEXT:    v_lshlrev_b32_e32 v4, 2, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s7
; VI-NEXT:    v_add_u32_e32 v0, vcc, s6, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_add_u32_e32 v2, vcc, s0, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_load_dword v2, v[2:3]
; VI-NEXT:    flat_load_dword v3, v[0:1]
; VI-NEXT:    v_add_u32_e32 v0, vcc, s4, v4
; VI-NEXT:    v_mov_b32_e32 v1, s5
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    v_max_i16_e32 v4, v3, v2
; VI-NEXT:    v_max_i16_sdwa v2, v3, v2 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
; VI-NEXT:    v_or_b32_e32 v2, v4, v2
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: v_test_imax_sge_v2i16:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; GFX9-NEXT:    v_lshlrev_b32_e32 v4, 2, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s7
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s6, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_add_co_u32_e32 v2, vcc, s0, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v3, vcc, 0, v3, vcc
; GFX9-NEXT:    global_load_dword v2, v[2:3], off
; GFX9-NEXT:    global_load_dword v3, v[0:1], off
; GFX9-NEXT:    v_mov_b32_e32 v1, s5
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s4, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    v_pk_max_i16 v2, v3, v2
; GFX9-NEXT:    global_store_dword v[0:1], v2, off
; GFX9-NEXT:    s_endpgm
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr <2 x i16>, <2 x i16> addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr <2 x i16>, <2 x i16> addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr <2 x i16>, <2 x i16> addrspace(1)* %out, i32 %tid
  %a = load <2 x i16>, <2 x i16> addrspace(1)* %gep0, align 4
  %b = load <2 x i16>, <2 x i16> addrspace(1)* %gep1, align 4
  %cmp = icmp sge <2 x i16> %a, %b
  %val = select <2 x i1> %cmp, <2 x i16> %a, <2 x i16> %b
  store <2 x i16> %val, <2 x i16> addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
define amdgpu_kernel void @v_test_imax_sge_v3i16(<3 x i16> addrspace(1)* %out, <3 x i16> addrspace(1)* %aptr, <3 x i16> addrspace(1)* %bptr) nounwind {
; VI-LABEL: v_test_imax_sge_v3i16:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; VI-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; VI-NEXT:    v_lshlrev_b32_e32 v8, 3, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v3, s7
; VI-NEXT:    v_mov_b32_e32 v1, s1
; VI-NEXT:    v_add_u32_e32 v0, vcc, s0, v8
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_add_u32_e32 v2, vcc, s6, v8
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    v_add_u32_e32 v4, vcc, 4, v0
; VI-NEXT:    v_addc_u32_e32 v5, vcc, 0, v1, vcc
; VI-NEXT:    v_add_u32_e32 v6, vcc, 4, v2
; VI-NEXT:    flat_load_dword v9, v[0:1]
; VI-NEXT:    v_addc_u32_e32 v7, vcc, 0, v3, vcc
; VI-NEXT:    flat_load_ushort v4, v[4:5]
; VI-NEXT:    flat_load_dword v5, v[2:3]
; VI-NEXT:    flat_load_ushort v6, v[6:7]
; VI-NEXT:    v_mov_b32_e32 v1, s5
; VI-NEXT:    v_add_u32_e32 v0, vcc, s4, v8
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_add_u32_e32 v2, vcc, 4, v0
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v1, vcc
; VI-NEXT:    s_waitcnt vmcnt(1) lgkmcnt(1)
; VI-NEXT:    v_max_i16_e32 v7, v5, v9
; VI-NEXT:    v_max_i16_sdwa v5, v5, v9 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    v_max_i16_e32 v4, v6, v4
; VI-NEXT:    v_or_b32_e32 v5, v7, v5
; VI-NEXT:    flat_store_dword v[0:1], v5
; VI-NEXT:    flat_store_short v[2:3], v4
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: v_test_imax_sge_v3i16:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; GFX9-NEXT:    v_lshlrev_b32_e32 v4, 3, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s7
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s6, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_add_co_u32_e32 v2, vcc, s0, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v3, vcc, 0, v3, vcc
; GFX9-NEXT:    global_load_dword v6, v[2:3], off
; GFX9-NEXT:    global_load_dword v7, v[0:1], off
; GFX9-NEXT:    v_mov_b32_e32 v5, s5
; GFX9-NEXT:    v_add_co_u32_e32 v4, vcc, s4, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v5, vcc, 0, v5, vcc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v8, v7
; GFX9-NEXT:    v_pk_max_i16 v7, v7, v6
; GFX9-NEXT:    global_load_short_d16 v6, v[2:3], off offset:4
; GFX9-NEXT:    global_load_short_d16 v8, v[0:1], off offset:4
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    v_pk_max_i16 v0, v8, v6
; GFX9-NEXT:    global_store_dword v[4:5], v7, off
; GFX9-NEXT:    global_store_short v[4:5], v0, off offset:4
; GFX9-NEXT:    s_endpgm
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr <3 x i16>, <3 x i16> addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr <3 x i16>, <3 x i16> addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr <3 x i16>, <3 x i16> addrspace(1)* %out, i32 %tid
  %a = load <3 x i16>, <3 x i16> addrspace(1)* %gep0, align 4
  %b = load <3 x i16>, <3 x i16> addrspace(1)* %gep1, align 4
  %cmp = icmp sge <3 x i16> %a, %b
  %val = select <3 x i1> %cmp, <3 x i16> %a, <3 x i16> %b
  store <3 x i16> %val, <3 x i16> addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
define amdgpu_kernel void @v_test_imax_sge_v4i16(<4 x i16> addrspace(1)* %out, <4 x i16> addrspace(1)* %aptr, <4 x i16> addrspace(1)* %bptr) nounwind {
; VI-LABEL: v_test_imax_sge_v4i16:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; VI-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; VI-NEXT:    v_lshlrev_b32_e32 v4, 3, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s7
; VI-NEXT:    v_add_u32_e32 v0, vcc, s6, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_add_u32_e32 v2, vcc, s0, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_load_dwordx2 v[2:3], v[2:3]
; VI-NEXT:    flat_load_dwordx2 v[0:1], v[0:1]
; VI-NEXT:    v_mov_b32_e32 v5, s5
; VI-NEXT:    v_add_u32_e32 v4, vcc, s4, v4
; VI-NEXT:    v_addc_u32_e32 v5, vcc, 0, v5, vcc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    v_max_i16_e32 v6, v1, v3
; VI-NEXT:    v_max_i16_sdwa v1, v1, v3 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
; VI-NEXT:    v_max_i16_e32 v3, v0, v2
; VI-NEXT:    v_max_i16_sdwa v0, v0, v2 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
; VI-NEXT:    v_or_b32_e32 v1, v6, v1
; VI-NEXT:    v_or_b32_e32 v0, v3, v0
; VI-NEXT:    flat_store_dwordx2 v[4:5], v[0:1]
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: v_test_imax_sge_v4i16:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; GFX9-NEXT:    v_lshlrev_b32_e32 v4, 3, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s7
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s6, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_add_co_u32_e32 v2, vcc, s0, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v3, vcc, 0, v3, vcc
; GFX9-NEXT:    global_load_dwordx2 v[2:3], v[2:3], off
; GFX9-NEXT:    global_load_dwordx2 v[0:1], v[0:1], off
; GFX9-NEXT:    v_mov_b32_e32 v5, s5
; GFX9-NEXT:    v_add_co_u32_e32 v4, vcc, s4, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v5, vcc, 0, v5, vcc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    v_pk_max_i16 v1, v1, v3
; GFX9-NEXT:    v_pk_max_i16 v0, v0, v2
; GFX9-NEXT:    global_store_dwordx2 v[4:5], v[0:1], off
; GFX9-NEXT:    s_endpgm
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr <4 x i16>, <4 x i16> addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr <4 x i16>, <4 x i16> addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr <4 x i16>, <4 x i16> addrspace(1)* %out, i32 %tid
  %a = load <4 x i16>, <4 x i16> addrspace(1)* %gep0, align 4
  %b = load <4 x i16>, <4 x i16> addrspace(1)* %gep1, align 4
  %cmp = icmp sge <4 x i16> %a, %b
  %val = select <4 x i1> %cmp, <4 x i16> %a, <4 x i16> %b
  store <4 x i16> %val, <4 x i16> addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
define amdgpu_kernel void @v_test_imax_sgt_i16(i16 addrspace(1)* %out, i16 addrspace(1)* %aptr, i16 addrspace(1)* %bptr) nounwind {
; VI-LABEL: v_test_imax_sgt_i16:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; VI-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; VI-NEXT:    v_lshlrev_b32_e32 v4, 1, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s7
; VI-NEXT:    v_add_u32_e32 v0, vcc, s6, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_add_u32_e32 v2, vcc, s0, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_load_ushort v2, v[2:3]
; VI-NEXT:    flat_load_ushort v3, v[0:1]
; VI-NEXT:    v_mov_b32_e32 v1, s5
; VI-NEXT:    v_add_u32_e32 v0, vcc, s4, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    v_max_i16_e32 v2, v3, v2
; VI-NEXT:    flat_store_short v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: v_test_imax_sgt_i16:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; GFX9-NEXT:    v_lshlrev_b32_e32 v4, 1, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s7
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s6, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_add_co_u32_e32 v2, vcc, s0, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v3, vcc, 0, v3, vcc
; GFX9-NEXT:    global_load_ushort v2, v[2:3], off
; GFX9-NEXT:    global_load_ushort v3, v[0:1], off
; GFX9-NEXT:    v_mov_b32_e32 v1, s5
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s4, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    v_max_i16_e32 v2, v3, v2
; GFX9-NEXT:    global_store_short v[0:1], v2, off
; GFX9-NEXT:    s_endpgm
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr i16, i16 addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr i16, i16 addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr i16, i16 addrspace(1)* %out, i32 %tid
  %a = load i16, i16 addrspace(1)* %gep0, align 4
  %b = load i16, i16 addrspace(1)* %gep1, align 4
  %cmp = icmp sgt i16 %a, %b
  %val = select i1 %cmp, i16 %a, i16 %b
  store i16 %val, i16 addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
define amdgpu_kernel void @v_test_umax_uge_i16(i16 addrspace(1)* %out, i16 addrspace(1)* %aptr, i16 addrspace(1)* %bptr) nounwind {
; VI-LABEL: v_test_umax_uge_i16:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; VI-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; VI-NEXT:    v_lshlrev_b32_e32 v4, 1, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s7
; VI-NEXT:    v_add_u32_e32 v0, vcc, s6, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_add_u32_e32 v2, vcc, s0, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_load_ushort v2, v[2:3]
; VI-NEXT:    flat_load_ushort v3, v[0:1]
; VI-NEXT:    v_mov_b32_e32 v1, s5
; VI-NEXT:    v_add_u32_e32 v0, vcc, s4, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    v_max_u16_e32 v2, v3, v2
; VI-NEXT:    flat_store_short v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: v_test_umax_uge_i16:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; GFX9-NEXT:    v_lshlrev_b32_e32 v4, 1, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s7
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s6, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_add_co_u32_e32 v2, vcc, s0, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v3, vcc, 0, v3, vcc
; GFX9-NEXT:    global_load_ushort v2, v[2:3], off
; GFX9-NEXT:    global_load_ushort v3, v[0:1], off
; GFX9-NEXT:    v_mov_b32_e32 v1, s5
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s4, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    v_max_u16_e32 v2, v3, v2
; GFX9-NEXT:    global_store_short v[0:1], v2, off
; GFX9-NEXT:    s_endpgm
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr i16, i16 addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr i16, i16 addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr i16, i16 addrspace(1)* %out, i32 %tid
  %a = load i16, i16 addrspace(1)* %gep0, align 4
  %b = load i16, i16 addrspace(1)* %gep1, align 4
  %cmp = icmp uge i16 %a, %b
  %val = select i1 %cmp, i16 %a, i16 %b
  store i16 %val, i16 addrspace(1)* %outgep, align 4
  ret void
}

; FIXME: Need to handle non-uniform case for function below (load without gep).
define amdgpu_kernel void @v_test_umax_ugt_i16(i16 addrspace(1)* %out, i16 addrspace(1)* %aptr, i16 addrspace(1)* %bptr) nounwind {
; VI-LABEL: v_test_umax_ugt_i16:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; VI-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; VI-NEXT:    v_lshlrev_b32_e32 v4, 1, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s7
; VI-NEXT:    v_add_u32_e32 v0, vcc, s6, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_add_u32_e32 v2, vcc, s0, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_load_ushort v2, v[2:3]
; VI-NEXT:    flat_load_ushort v3, v[0:1]
; VI-NEXT:    v_mov_b32_e32 v1, s5
; VI-NEXT:    v_add_u32_e32 v0, vcc, s4, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    v_max_u16_e32 v2, v3, v2
; VI-NEXT:    flat_store_short v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: v_test_umax_ugt_i16:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; GFX9-NEXT:    v_lshlrev_b32_e32 v4, 1, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s7
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s6, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_add_co_u32_e32 v2, vcc, s0, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v3, vcc, 0, v3, vcc
; GFX9-NEXT:    global_load_ushort v2, v[2:3], off
; GFX9-NEXT:    global_load_ushort v3, v[0:1], off
; GFX9-NEXT:    v_mov_b32_e32 v1, s5
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s4, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    v_max_u16_e32 v2, v3, v2
; GFX9-NEXT:    global_store_short v[0:1], v2, off
; GFX9-NEXT:    s_endpgm
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr i16, i16 addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr i16, i16 addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr i16, i16 addrspace(1)* %out, i32 %tid
  %a = load i16, i16 addrspace(1)* %gep0, align 4
  %b = load i16, i16 addrspace(1)* %gep1, align 4
  %cmp = icmp ugt i16 %a, %b
  %val = select i1 %cmp, i16 %a, i16 %b
  store i16 %val, i16 addrspace(1)* %outgep, align 4
  ret void
}

define amdgpu_kernel void @v_test_umax_ugt_v2i16(<2 x i16> addrspace(1)* %out, <2 x i16> addrspace(1)* %aptr, <2 x i16> addrspace(1)* %bptr) nounwind {
; VI-LABEL: v_test_umax_ugt_v2i16:
; VI:       ; %bb.0:
; VI-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; VI-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; VI-NEXT:    v_lshlrev_b32_e32 v4, 2, v0
; VI-NEXT:    s_waitcnt lgkmcnt(0)
; VI-NEXT:    v_mov_b32_e32 v1, s7
; VI-NEXT:    v_add_u32_e32 v0, vcc, s6, v4
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    v_mov_b32_e32 v3, s1
; VI-NEXT:    v_add_u32_e32 v2, vcc, s0, v4
; VI-NEXT:    v_addc_u32_e32 v3, vcc, 0, v3, vcc
; VI-NEXT:    flat_load_dword v2, v[2:3]
; VI-NEXT:    flat_load_dword v3, v[0:1]
; VI-NEXT:    v_add_u32_e32 v0, vcc, s4, v4
; VI-NEXT:    v_mov_b32_e32 v1, s5
; VI-NEXT:    v_addc_u32_e32 v1, vcc, 0, v1, vcc
; VI-NEXT:    s_waitcnt vmcnt(0) lgkmcnt(0)
; VI-NEXT:    v_max_u16_e32 v4, v3, v2
; VI-NEXT:    v_max_u16_sdwa v2, v3, v2 dst_sel:WORD_1 dst_unused:UNUSED_PAD src0_sel:WORD_1 src1_sel:WORD_1
; VI-NEXT:    v_or_b32_e32 v2, v4, v2
; VI-NEXT:    flat_store_dword v[0:1], v2
; VI-NEXT:    s_endpgm
;
; GFX9-LABEL: v_test_umax_ugt_v2i16:
; GFX9:       ; %bb.0:
; GFX9-NEXT:    s_load_dwordx4 s[4:7], s[0:1], 0x24
; GFX9-NEXT:    s_load_dwordx2 s[0:1], s[0:1], 0x34
; GFX9-NEXT:    v_lshlrev_b32_e32 v4, 2, v0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_mov_b32_e32 v1, s7
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s6, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    v_mov_b32_e32 v3, s1
; GFX9-NEXT:    v_add_co_u32_e32 v2, vcc, s0, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v3, vcc, 0, v3, vcc
; GFX9-NEXT:    global_load_dword v2, v[2:3], off
; GFX9-NEXT:    global_load_dword v3, v[0:1], off
; GFX9-NEXT:    v_mov_b32_e32 v1, s5
; GFX9-NEXT:    v_add_co_u32_e32 v0, vcc, s4, v4
; GFX9-NEXT:    v_addc_co_u32_e32 v1, vcc, 0, v1, vcc
; GFX9-NEXT:    s_waitcnt vmcnt(0)
; GFX9-NEXT:    v_pk_max_u16 v2, v3, v2
; GFX9-NEXT:    global_store_dword v[0:1], v2, off
; GFX9-NEXT:    s_endpgm
  %tid = call i32 @llvm.amdgcn.workitem.id.x() nounwind readnone
  %gep0 = getelementptr <2 x i16>, <2 x i16> addrspace(1)* %aptr, i32 %tid
  %gep1 = getelementptr <2 x i16>, <2 x i16> addrspace(1)* %bptr, i32 %tid
  %outgep = getelementptr <2 x i16>, <2 x i16> addrspace(1)* %out, i32 %tid
  %a = load <2 x i16>, <2 x i16> addrspace(1)* %gep0, align 4
  %b = load <2 x i16>, <2 x i16> addrspace(1)* %gep1, align 4
  %cmp = icmp ugt <2 x i16> %a, %b
  %val = select <2 x i1> %cmp, <2 x i16> %a, <2 x i16> %b
  store <2 x i16> %val, <2 x i16> addrspace(1)* %outgep, align 4
  ret void
}

declare i32 @llvm.amdgcn.workitem.id.x() nounwind readnone