llvm.org GIT mirror llvm / 743f98a test / CodeGen / AMDGPU / idiv-licm.ll
743f98a

Tree @743f98a (Download .tar.gz)

idiv-licm.ll @743f98araw · 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
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc -march=amdgcn -mcpu=gfx900 -verify-machineinstrs < %s | FileCheck -enable-var-scope -check-prefix=GFX9 %s

define amdgpu_kernel void @udiv32_invariant_denom(i32 addrspace(1)* nocapture %arg, i32 %arg1) {
; GFX9-LABEL: udiv32_invariant_denom:
; GFX9:       ; %bb.0: ; %bb
; GFX9-NEXT:    s_load_dword s4, s[0:1], 0x2c
; GFX9-NEXT:    s_load_dwordx2 s[6:7], s[0:1], 0x24
; GFX9-NEXT:    s_mov_b32 s5, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_cvt_f32_u32_e32 v0, s4
; GFX9-NEXT:    s_sub_i32 s8, 0, s4
; GFX9-NEXT:    v_rcp_iflag_f32_e32 v0, v0
; GFX9-NEXT:    v_mul_f32_e32 v0, 0x4f800000, v0
; GFX9-NEXT:    v_cvt_u32_f32_e32 v0, v0
; GFX9-NEXT:    v_mul_lo_u32 v1, v0, s4
; GFX9-NEXT:    v_mul_hi_u32 v2, v0, s4
; GFX9-NEXT:    v_sub_u32_e32 v3, 0, v1
; GFX9-NEXT:    v_cmp_eq_u32_e32 vcc, 0, v2
; GFX9-NEXT:    v_cndmask_b32_e32 v1, v1, v3, vcc
; GFX9-NEXT:    v_mul_hi_u32 v1, v1, v0
; GFX9-NEXT:    v_add_u32_e32 v2, v0, v1
; GFX9-NEXT:    v_sub_u32_e32 v0, v0, v1
; GFX9-NEXT:    v_cndmask_b32_e32 v0, v0, v2, vcc
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:  BB0_1: ; %bb3
; GFX9-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT:    v_mul_lo_u32 v3, s8, v2
; GFX9-NEXT:    v_mul_lo_u32 v4, v2, s4
; GFX9-NEXT:    v_add_u32_e32 v6, 1, v2
; GFX9-NEXT:    v_add_u32_e32 v5, -1, v2
; GFX9-NEXT:    v_add_u32_e32 v3, s5, v3
; GFX9-NEXT:    v_cmp_ge_u32_e32 vcc, s5, v4
; GFX9-NEXT:    v_cmp_le_u32_e64 s[0:1], s4, v3
; GFX9-NEXT:    v_mov_b32_e32 v3, s6
; GFX9-NEXT:    s_and_b64 s[0:1], s[0:1], vcc
; GFX9-NEXT:    s_add_i32 s5, s5, 1
; GFX9-NEXT:    v_mov_b32_e32 v4, s7
; GFX9-NEXT:    s_add_u32 s6, s6, 4
; GFX9-NEXT:    s_addc_u32 s7, s7, 0
; GFX9-NEXT:    v_cndmask_b32_e64 v6, v2, v6, s[0:1]
; GFX9-NEXT:    v_add_co_u32_e64 v1, s[2:3], v1, v0
; GFX9-NEXT:    v_cndmask_b32_e32 v5, v5, v6, vcc
; GFX9-NEXT:    v_addc_co_u32_e64 v2, s[0:1], 0, v2, s[2:3]
; GFX9-NEXT:    s_cmpk_eq_i32 s5, 0x400
; GFX9-NEXT:    global_store_dword v[3:4], v5, off
; GFX9-NEXT:    s_cbranch_scc0 BB0_1
; GFX9-NEXT:  ; %bb.2: ; %bb2
; GFX9-NEXT:    s_endpgm
bb:
  br label %bb3

bb2:                                              ; preds = %bb3
  ret void

bb3:                                              ; preds = %bb3, %bb
  %tmp = phi i32 [ 0, %bb ], [ %tmp7, %bb3 ]
  %tmp4 = udiv i32 %tmp, %arg1
  %tmp5 = zext i32 %tmp to i64
  %tmp6 = getelementptr inbounds i32, i32 addrspace(1)* %arg, i64 %tmp5
  store i32 %tmp4, i32 addrspace(1)* %tmp6, align 4
  %tmp7 = add nuw nsw i32 %tmp, 1
  %tmp8 = icmp eq i32 %tmp7, 1024
  br i1 %tmp8, label %bb2, label %bb3
}

define amdgpu_kernel void @urem32_invariant_denom(i32 addrspace(1)* nocapture %arg, i32 %arg1) {
; GFX9-LABEL: urem32_invariant_denom:
; GFX9:       ; %bb.0: ; %bb
; GFX9-NEXT:    s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT:    s_load_dwordx2 s[4:5], s[0:1], 0x24
; GFX9-NEXT:    s_mov_b32 s3, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    v_cvt_f32_u32_e32 v0, s2
; GFX9-NEXT:    s_sub_i32 s6, 0, s2
; GFX9-NEXT:    v_rcp_iflag_f32_e32 v0, v0
; GFX9-NEXT:    v_mul_f32_e32 v0, 0x4f800000, v0
; GFX9-NEXT:    v_cvt_u32_f32_e32 v0, v0
; GFX9-NEXT:    v_mul_lo_u32 v1, v0, s2
; GFX9-NEXT:    v_mul_hi_u32 v2, v0, s2
; GFX9-NEXT:    v_sub_u32_e32 v3, 0, v1
; GFX9-NEXT:    v_cmp_eq_u32_e32 vcc, 0, v2
; GFX9-NEXT:    v_cndmask_b32_e32 v1, v1, v3, vcc
; GFX9-NEXT:    v_mul_hi_u32 v1, v1, v0
; GFX9-NEXT:    v_add_u32_e32 v2, v0, v1
; GFX9-NEXT:    v_sub_u32_e32 v0, v0, v1
; GFX9-NEXT:    v_cndmask_b32_e32 v0, v0, v2, vcc
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:  BB1_1: ; %bb3
; GFX9-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT:    v_mul_lo_u32 v5, s6, v2
; GFX9-NEXT:    v_sub_u32_e32 v6, 1, v2
; GFX9-NEXT:    v_not_b32_e32 v7, v2
; GFX9-NEXT:    v_mul_lo_u32 v8, v2, s2
; GFX9-NEXT:    v_mul_lo_u32 v6, s2, v6
; GFX9-NEXT:    v_mul_lo_u32 v7, s2, v7
; GFX9-NEXT:    v_add_co_u32_e32 v1, vcc, v1, v0
; GFX9-NEXT:    v_addc_co_u32_e32 v2, vcc, 0, v2, vcc
; GFX9-NEXT:    v_add_u32_e32 v5, s3, v5
; GFX9-NEXT:    v_cmp_le_u32_e32 vcc, s2, v5
; GFX9-NEXT:    v_cmp_ge_u32_e64 s[0:1], s3, v8
; GFX9-NEXT:    v_mov_b32_e32 v3, s4
; GFX9-NEXT:    v_add_u32_e32 v6, s3, v6
; GFX9-NEXT:    v_add_u32_e32 v7, s3, v7
; GFX9-NEXT:    s_and_b64 vcc, vcc, s[0:1]
; GFX9-NEXT:    s_add_i32 s3, s3, 1
; GFX9-NEXT:    v_mov_b32_e32 v4, s5
; GFX9-NEXT:    s_add_u32 s4, s4, 4
; GFX9-NEXT:    v_cndmask_b32_e32 v5, v5, v7, vcc
; GFX9-NEXT:    s_addc_u32 s5, s5, 0
; GFX9-NEXT:    v_cndmask_b32_e64 v5, v6, v5, s[0:1]
; GFX9-NEXT:    s_cmpk_eq_i32 s3, 0x400
; GFX9-NEXT:    global_store_dword v[3:4], v5, off
; GFX9-NEXT:    s_cbranch_scc0 BB1_1
; GFX9-NEXT:  ; %bb.2: ; %bb2
; GFX9-NEXT:    s_endpgm
bb:
  br label %bb3

bb2:                                              ; preds = %bb3
  ret void

bb3:                                              ; preds = %bb3, %bb
  %tmp = phi i32 [ 0, %bb ], [ %tmp7, %bb3 ]
  %tmp4 = urem i32 %tmp, %arg1
  %tmp5 = zext i32 %tmp to i64
  %tmp6 = getelementptr inbounds i32, i32 addrspace(1)* %arg, i64 %tmp5
  store i32 %tmp4, i32 addrspace(1)* %tmp6, align 4
  %tmp7 = add nuw nsw i32 %tmp, 1
  %tmp8 = icmp eq i32 %tmp7, 1024
  br i1 %tmp8, label %bb2, label %bb3
}

define amdgpu_kernel void @sdiv32_invariant_denom(i32 addrspace(1)* nocapture %arg, i32 %arg1) {
; GFX9-LABEL: sdiv32_invariant_denom:
; GFX9:       ; %bb.0: ; %bb
; GFX9-NEXT:    s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT:    s_load_dwordx2 s[6:7], s[0:1], 0x24
; GFX9-NEXT:    s_mov_b32 s8, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_ashr_i32 s4, s2, 31
; GFX9-NEXT:    s_add_i32 s2, s2, s4
; GFX9-NEXT:    s_xor_b32 s5, s2, s4
; GFX9-NEXT:    v_cvt_f32_u32_e32 v0, s5
; GFX9-NEXT:    v_rcp_iflag_f32_e32 v0, v0
; GFX9-NEXT:    v_mul_f32_e32 v0, 0x4f800000, v0
; GFX9-NEXT:    v_cvt_u32_f32_e32 v0, v0
; GFX9-NEXT:    v_mul_lo_u32 v1, v0, s5
; GFX9-NEXT:    v_mul_hi_u32 v2, v0, s5
; GFX9-NEXT:    v_sub_u32_e32 v3, 0, v1
; GFX9-NEXT:    v_cmp_eq_u32_e32 vcc, 0, v2
; GFX9-NEXT:    v_cndmask_b32_e32 v1, v1, v3, vcc
; GFX9-NEXT:    v_mul_hi_u32 v1, v1, v0
; GFX9-NEXT:    v_add_u32_e32 v2, v0, v1
; GFX9-NEXT:    v_sub_u32_e32 v0, v0, v1
; GFX9-NEXT:    v_cndmask_b32_e32 v0, v0, v2, vcc
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:  BB2_1: ; %bb3
; GFX9-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT:    v_mul_lo_u32 v5, s5, v2
; GFX9-NEXT:    v_add_u32_e32 v6, 1, v2
; GFX9-NEXT:    v_mov_b32_e32 v3, s6
; GFX9-NEXT:    v_add_u32_e32 v7, -1, v2
; GFX9-NEXT:    v_sub_u32_e32 v8, s8, v5
; GFX9-NEXT:    v_cmp_ge_u32_e64 s[0:1], s8, v5
; GFX9-NEXT:    v_cmp_le_u32_e64 s[2:3], s5, v8
; GFX9-NEXT:    s_and_b64 s[2:3], s[2:3], s[0:1]
; GFX9-NEXT:    v_cndmask_b32_e64 v5, v2, v6, s[2:3]
; GFX9-NEXT:    s_add_i32 s8, s8, 1
; GFX9-NEXT:    v_cndmask_b32_e64 v5, v7, v5, s[0:1]
; GFX9-NEXT:    v_mov_b32_e32 v4, s7
; GFX9-NEXT:    s_add_u32 s6, s6, 4
; GFX9-NEXT:    v_add_co_u32_e32 v1, vcc, v1, v0
; GFX9-NEXT:    v_xor_b32_e32 v5, s4, v5
; GFX9-NEXT:    s_addc_u32 s7, s7, 0
; GFX9-NEXT:    v_subrev_u32_e32 v5, s4, v5
; GFX9-NEXT:    v_addc_co_u32_e32 v2, vcc, 0, v2, vcc
; GFX9-NEXT:    s_cmpk_eq_i32 s8, 0x400
; GFX9-NEXT:    global_store_dword v[3:4], v5, off
; GFX9-NEXT:    s_cbranch_scc0 BB2_1
; GFX9-NEXT:  ; %bb.2: ; %bb2
; GFX9-NEXT:    s_endpgm
bb:
  br label %bb3

bb2:                                              ; preds = %bb3
  ret void

bb3:                                              ; preds = %bb3, %bb
  %tmp = phi i32 [ 0, %bb ], [ %tmp7, %bb3 ]
  %tmp4 = sdiv i32 %tmp, %arg1
  %tmp5 = zext i32 %tmp to i64
  %tmp6 = getelementptr inbounds i32, i32 addrspace(1)* %arg, i64 %tmp5
  store i32 %tmp4, i32 addrspace(1)* %tmp6, align 4
  %tmp7 = add nuw nsw i32 %tmp, 1
  %tmp8 = icmp eq i32 %tmp7, 1024
  br i1 %tmp8, label %bb2, label %bb3
}

define amdgpu_kernel void @srem32_invariant_denom(i32 addrspace(1)* nocapture %arg, i32 %arg1) {
; GFX9-LABEL: srem32_invariant_denom:
; GFX9:       ; %bb.0: ; %bb
; GFX9-NEXT:    s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT:    s_load_dwordx2 s[4:5], s[0:1], 0x24
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_ashr_i32 s3, s2, 31
; GFX9-NEXT:    s_add_i32 s2, s2, s3
; GFX9-NEXT:    s_xor_b32 s2, s2, s3
; GFX9-NEXT:    v_cvt_f32_u32_e32 v0, s2
; GFX9-NEXT:    s_mov_b32 s3, 0
; GFX9-NEXT:    v_rcp_iflag_f32_e32 v0, v0
; GFX9-NEXT:    v_mul_f32_e32 v0, 0x4f800000, v0
; GFX9-NEXT:    v_cvt_u32_f32_e32 v0, v0
; GFX9-NEXT:    v_mul_lo_u32 v1, v0, s2
; GFX9-NEXT:    v_mul_hi_u32 v2, v0, s2
; GFX9-NEXT:    v_sub_u32_e32 v3, 0, v1
; GFX9-NEXT:    v_cmp_eq_u32_e32 vcc, 0, v2
; GFX9-NEXT:    v_cndmask_b32_e32 v1, v1, v3, vcc
; GFX9-NEXT:    v_mul_hi_u32 v1, v1, v0
; GFX9-NEXT:    v_add_u32_e32 v2, v0, v1
; GFX9-NEXT:    v_sub_u32_e32 v0, v0, v1
; GFX9-NEXT:    v_cndmask_b32_e32 v0, v0, v2, vcc
; GFX9-NEXT:    v_mov_b32_e32 v1, 0
; GFX9-NEXT:    v_mov_b32_e32 v2, 0
; GFX9-NEXT:  BB3_1: ; %bb3
; GFX9-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT:    v_mul_lo_u32 v5, s2, v2
; GFX9-NEXT:    v_sub_u32_e32 v6, 1, v2
; GFX9-NEXT:    v_not_b32_e32 v7, v2
; GFX9-NEXT:    v_mul_lo_u32 v6, s2, v6
; GFX9-NEXT:    v_mul_lo_u32 v7, s2, v7
; GFX9-NEXT:    v_add_co_u32_e32 v1, vcc, v1, v0
; GFX9-NEXT:    v_addc_co_u32_e32 v2, vcc, 0, v2, vcc
; GFX9-NEXT:    v_sub_u32_e32 v8, s3, v5
; GFX9-NEXT:    v_cmp_le_u32_e32 vcc, s2, v8
; GFX9-NEXT:    v_cmp_ge_u32_e64 s[0:1], s3, v5
; GFX9-NEXT:    v_mov_b32_e32 v3, s4
; GFX9-NEXT:    v_add_u32_e32 v6, s3, v6
; GFX9-NEXT:    v_add_u32_e32 v7, s3, v7
; GFX9-NEXT:    s_and_b64 vcc, vcc, s[0:1]
; GFX9-NEXT:    s_add_i32 s3, s3, 1
; GFX9-NEXT:    v_mov_b32_e32 v4, s5
; GFX9-NEXT:    s_add_u32 s4, s4, 4
; GFX9-NEXT:    v_cndmask_b32_e32 v5, v8, v7, vcc
; GFX9-NEXT:    s_addc_u32 s5, s5, 0
; GFX9-NEXT:    v_cndmask_b32_e64 v5, v6, v5, s[0:1]
; GFX9-NEXT:    s_cmpk_eq_i32 s3, 0x400
; GFX9-NEXT:    global_store_dword v[3:4], v5, off
; GFX9-NEXT:    s_cbranch_scc0 BB3_1
; GFX9-NEXT:  ; %bb.2: ; %bb2
; GFX9-NEXT:    s_endpgm
bb:
  br label %bb3

bb2:                                              ; preds = %bb3
  ret void

bb3:                                              ; preds = %bb3, %bb
  %tmp = phi i32 [ 0, %bb ], [ %tmp7, %bb3 ]
  %tmp4 = srem i32 %tmp, %arg1
  %tmp5 = zext i32 %tmp to i64
  %tmp6 = getelementptr inbounds i32, i32 addrspace(1)* %arg, i64 %tmp5
  store i32 %tmp4, i32 addrspace(1)* %tmp6, align 4
  %tmp7 = add nuw nsw i32 %tmp, 1
  %tmp8 = icmp eq i32 %tmp7, 1024
  br i1 %tmp8, label %bb2, label %bb3
}

define amdgpu_kernel void @udiv16_invariant_denom(i16 addrspace(1)* nocapture %arg, i16 %arg1) {
; GFX9-LABEL: udiv16_invariant_denom:
; GFX9:       ; %bb.0: ; %bb
; GFX9-NEXT:    s_load_dword s3, s[0:1], 0x2c
; GFX9-NEXT:    s_mov_b32 s2, 0xffff
; GFX9-NEXT:    s_load_dwordx2 s[4:5], s[0:1], 0x24
; GFX9-NEXT:    v_mov_b32_e32 v3, 0
; GFX9-NEXT:    v_mov_b32_e32 v4, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_and_b32 s3, s3, s2
; GFX9-NEXT:    v_cvt_f32_u32_e32 v0, s3
; GFX9-NEXT:    s_movk_i32 s3, 0x400
; GFX9-NEXT:    v_rcp_iflag_f32_e32 v1, v0
; GFX9-NEXT:  BB4_1: ; %bb3
; GFX9-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT:    v_and_b32_e32 v2, s2, v4
; GFX9-NEXT:    v_cvt_f32_u32_e32 v7, v2
; GFX9-NEXT:    v_lshlrev_b64 v[5:6], 1, v[2:3]
; GFX9-NEXT:    v_mov_b32_e32 v2, s5
; GFX9-NEXT:    v_add_co_u32_e64 v5, s[0:1], s4, v5
; GFX9-NEXT:    v_addc_co_u32_e64 v6, s[0:1], v2, v6, s[0:1]
; GFX9-NEXT:    v_mul_f32_e32 v2, v7, v1
; GFX9-NEXT:    v_trunc_f32_e32 v2, v2
; GFX9-NEXT:    v_cvt_u32_f32_e32 v8, v2
; GFX9-NEXT:    v_add_u16_e32 v4, 1, v4
; GFX9-NEXT:    v_mad_f32 v2, -v2, v0, v7
; GFX9-NEXT:    v_cmp_ge_f32_e64 s[0:1], |v2|, v0
; GFX9-NEXT:    v_cmp_eq_u16_e32 vcc, s3, v4
; GFX9-NEXT:    v_addc_co_u32_e64 v2, s[0:1], 0, v8, s[0:1]
; GFX9-NEXT:    s_and_b64 vcc, exec, vcc
; GFX9-NEXT:    global_store_short v[5:6], v2, off
; GFX9-NEXT:    s_cbranch_vccz BB4_1
; GFX9-NEXT:  ; %bb.2: ; %bb2
; GFX9-NEXT:    s_endpgm
bb:
  br label %bb3

bb2:                                              ; preds = %bb3
  ret void

bb3:                                              ; preds = %bb3, %bb
  %tmp = phi i16 [ 0, %bb ], [ %tmp7, %bb3 ]
  %tmp4 = udiv i16 %tmp, %arg1
  %tmp5 = zext i16 %tmp to i64
  %tmp6 = getelementptr inbounds i16, i16 addrspace(1)* %arg, i64 %tmp5
  store i16 %tmp4, i16 addrspace(1)* %tmp6, align 2
  %tmp7 = add nuw nsw i16 %tmp, 1
  %tmp8 = icmp eq i16 %tmp7, 1024
  br i1 %tmp8, label %bb2, label %bb3
}

define amdgpu_kernel void @urem16_invariant_denom(i16 addrspace(1)* nocapture %arg, i16 %arg1) {
; GFX9-LABEL: urem16_invariant_denom:
; GFX9:       ; %bb.0: ; %bb
; GFX9-NEXT:    s_load_dword s3, s[0:1], 0x2c
; GFX9-NEXT:    s_mov_b32 s2, 0xffff
; GFX9-NEXT:    s_load_dwordx2 s[4:5], s[0:1], 0x24
; GFX9-NEXT:    v_mov_b32_e32 v3, 0
; GFX9-NEXT:    s_movk_i32 s6, 0x400
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_and_b32 s3, s3, s2
; GFX9-NEXT:    v_cvt_f32_u32_e32 v0, s3
; GFX9-NEXT:    v_mov_b32_e32 v4, 0
; GFX9-NEXT:    v_rcp_iflag_f32_e32 v1, v0
; GFX9-NEXT:  BB5_1: ; %bb3
; GFX9-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT:    v_and_b32_e32 v2, s2, v4
; GFX9-NEXT:    v_cvt_f32_u32_e32 v7, v2
; GFX9-NEXT:    v_lshlrev_b64 v[5:6], 1, v[2:3]
; GFX9-NEXT:    v_mov_b32_e32 v8, s5
; GFX9-NEXT:    v_add_co_u32_e64 v5, s[0:1], s4, v5
; GFX9-NEXT:    v_addc_co_u32_e64 v6, s[0:1], v8, v6, s[0:1]
; GFX9-NEXT:    v_mul_f32_e32 v8, v7, v1
; GFX9-NEXT:    v_trunc_f32_e32 v8, v8
; GFX9-NEXT:    v_cvt_u32_f32_e32 v9, v8
; GFX9-NEXT:    v_mad_f32 v7, -v8, v0, v7
; GFX9-NEXT:    v_cmp_ge_f32_e64 s[0:1], |v7|, v0
; GFX9-NEXT:    v_add_u16_e32 v4, 1, v4
; GFX9-NEXT:    v_addc_co_u32_e64 v7, s[0:1], 0, v9, s[0:1]
; GFX9-NEXT:    v_mul_lo_u32 v7, v7, s3
; GFX9-NEXT:    v_cmp_eq_u16_e32 vcc, s6, v4
; GFX9-NEXT:    s_and_b64 vcc, exec, vcc
; GFX9-NEXT:    v_sub_u32_e32 v2, v2, v7
; GFX9-NEXT:    global_store_short v[5:6], v2, off
; GFX9-NEXT:    s_cbranch_vccz BB5_1
; GFX9-NEXT:  ; %bb.2: ; %bb2
; GFX9-NEXT:    s_endpgm
bb:
  br label %bb3

bb2:                                              ; preds = %bb3
  ret void

bb3:                                              ; preds = %bb3, %bb
  %tmp = phi i16 [ 0, %bb ], [ %tmp7, %bb3 ]
  %tmp4 = urem i16 %tmp, %arg1
  %tmp5 = zext i16 %tmp to i64
  %tmp6 = getelementptr inbounds i16, i16 addrspace(1)* %arg, i64 %tmp5
  store i16 %tmp4, i16 addrspace(1)* %tmp6, align 2
  %tmp7 = add nuw nsw i16 %tmp, 1
  %tmp8 = icmp eq i16 %tmp7, 1024
  br i1 %tmp8, label %bb2, label %bb3
}

define amdgpu_kernel void @sdiv16_invariant_denom(i16 addrspace(1)* nocapture %arg, i16 %arg1) {
; GFX9-LABEL: sdiv16_invariant_denom:
; GFX9:       ; %bb.0: ; %bb
; GFX9-NEXT:    s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT:    s_load_dwordx2 s[4:5], s[0:1], 0x24
; GFX9-NEXT:    v_mov_b32_e32 v4, 0
; GFX9-NEXT:    s_movk_i32 s3, 0x400
; GFX9-NEXT:    v_mov_b32_e32 v5, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_sext_i32_i16 s2, s2
; GFX9-NEXT:    v_cvt_f32_i32_e32 v0, s2
; GFX9-NEXT:    v_rcp_iflag_f32_e32 v1, v0
; GFX9-NEXT:    v_and_b32_e32 v2, 0x7fffffff, v0
; GFX9-NEXT:  BB6_1: ; %bb3
; GFX9-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT:    v_bfe_i32 v6, v5, 0, 16
; GFX9-NEXT:    v_and_b32_e32 v3, 0xffff, v5
; GFX9-NEXT:    v_cvt_f32_i32_e32 v10, v6
; GFX9-NEXT:    v_xor_b32_e32 v9, s2, v6
; GFX9-NEXT:    v_lshlrev_b64 v[6:7], 1, v[3:4]
; GFX9-NEXT:    v_mov_b32_e32 v8, s5
; GFX9-NEXT:    v_add_co_u32_e64 v6, s[0:1], s4, v6
; GFX9-NEXT:    v_addc_co_u32_e64 v7, s[0:1], v8, v7, s[0:1]
; GFX9-NEXT:    v_mul_f32_e32 v8, v10, v1
; GFX9-NEXT:    v_trunc_f32_e32 v8, v8
; GFX9-NEXT:    v_ashrrev_i32_e32 v3, 30, v9
; GFX9-NEXT:    v_cvt_i32_f32_e32 v9, v8
; GFX9-NEXT:    v_mad_f32 v8, -v8, v0, v10
; GFX9-NEXT:    v_add_u16_e32 v5, 1, v5
; GFX9-NEXT:    v_or_b32_e32 v3, 1, v3
; GFX9-NEXT:    v_cmp_ge_f32_e64 s[0:1], |v8|, v2
; GFX9-NEXT:    v_cmp_eq_u16_e32 vcc, s3, v5
; GFX9-NEXT:    v_cndmask_b32_e64 v3, 0, v3, s[0:1]
; GFX9-NEXT:    v_add_u32_e32 v3, v9, v3
; GFX9-NEXT:    s_and_b64 vcc, exec, vcc
; GFX9-NEXT:    global_store_short v[6:7], v3, off
; GFX9-NEXT:    s_cbranch_vccz BB6_1
; GFX9-NEXT:  ; %bb.2: ; %bb2
; GFX9-NEXT:    s_endpgm
bb:
  br label %bb3

bb2:                                              ; preds = %bb3
  ret void

bb3:                                              ; preds = %bb3, %bb
  %tmp = phi i16 [ 0, %bb ], [ %tmp7, %bb3 ]
  %tmp4 = sdiv i16 %tmp, %arg1
  %tmp5 = zext i16 %tmp to i64
  %tmp6 = getelementptr inbounds i16, i16 addrspace(1)* %arg, i64 %tmp5
  store i16 %tmp4, i16 addrspace(1)* %tmp6, align 2
  %tmp7 = add nuw nsw i16 %tmp, 1
  %tmp8 = icmp eq i16 %tmp7, 1024
  br i1 %tmp8, label %bb2, label %bb3
}

define amdgpu_kernel void @srem16_invariant_denom(i16 addrspace(1)* nocapture %arg, i16 %arg1) {
; GFX9-LABEL: srem16_invariant_denom:
; GFX9:       ; %bb.0: ; %bb
; GFX9-NEXT:    s_load_dword s2, s[0:1], 0x2c
; GFX9-NEXT:    s_load_dwordx2 s[4:5], s[0:1], 0x24
; GFX9-NEXT:    v_mov_b32_e32 v4, 0
; GFX9-NEXT:    s_movk_i32 s3, 0x400
; GFX9-NEXT:    v_mov_b32_e32 v5, 0
; GFX9-NEXT:    s_waitcnt lgkmcnt(0)
; GFX9-NEXT:    s_sext_i32_i16 s2, s2
; GFX9-NEXT:    v_cvt_f32_i32_e32 v0, s2
; GFX9-NEXT:    v_rcp_iflag_f32_e32 v1, v0
; GFX9-NEXT:    v_and_b32_e32 v2, 0x7fffffff, v0
; GFX9-NEXT:  BB7_1: ; %bb3
; GFX9-NEXT:    ; =>This Inner Loop Header: Depth=1
; GFX9-NEXT:    v_bfe_i32 v8, v5, 0, 16
; GFX9-NEXT:    v_and_b32_e32 v3, 0xffff, v5
; GFX9-NEXT:    v_cvt_f32_i32_e32 v11, v8
; GFX9-NEXT:    v_lshlrev_b64 v[6:7], 1, v[3:4]
; GFX9-NEXT:    v_mov_b32_e32 v9, s5
; GFX9-NEXT:    v_add_co_u32_e64 v6, s[0:1], s4, v6
; GFX9-NEXT:    v_addc_co_u32_e64 v7, s[0:1], v9, v7, s[0:1]
; GFX9-NEXT:    v_mul_f32_e32 v9, v11, v1
; GFX9-NEXT:    v_xor_b32_e32 v10, s2, v8
; GFX9-NEXT:    v_trunc_f32_e32 v9, v9
; GFX9-NEXT:    v_ashrrev_i32_e32 v3, 30, v10
; GFX9-NEXT:    v_cvt_i32_f32_e32 v10, v9
; GFX9-NEXT:    v_mad_f32 v9, -v9, v0, v11
; GFX9-NEXT:    v_or_b32_e32 v3, 1, v3
; GFX9-NEXT:    v_cmp_ge_f32_e64 s[0:1], |v9|, v2
; GFX9-NEXT:    v_cndmask_b32_e64 v3, 0, v3, s[0:1]
; GFX9-NEXT:    v_add_u32_e32 v3, v10, v3
; GFX9-NEXT:    v_mul_lo_u32 v3, v3, s2
; GFX9-NEXT:    v_add_u16_e32 v5, 1, v5
; GFX9-NEXT:    v_cmp_eq_u16_e32 vcc, s3, v5
; GFX9-NEXT:    s_and_b64 vcc, exec, vcc
; GFX9-NEXT:    v_sub_u32_e32 v3, v8, v3
; GFX9-NEXT:    global_store_short v[6:7], v3, off
; GFX9-NEXT:    s_cbranch_vccz BB7_1
; GFX9-NEXT:  ; %bb.2: ; %bb2
; GFX9-NEXT:    s_endpgm
bb:
  br label %bb3

bb2:                                              ; preds = %bb3
  ret void

bb3:                                              ; preds = %bb3, %bb
  %tmp = phi i16 [ 0, %bb ], [ %tmp7, %bb3 ]
  %tmp4 = srem i16 %tmp, %arg1
  %tmp5 = zext i16 %tmp to i64
  %tmp6 = getelementptr inbounds i16, i16 addrspace(1)* %arg, i64 %tmp5
  store i16 %tmp4, i16 addrspace(1)* %tmp6, align 2
  %tmp7 = add nuw nsw i16 %tmp, 1
  %tmp8 = icmp eq i16 %tmp7, 1024
  br i1 %tmp8, label %bb2, label %bb3
}