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
|
; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
; RUN: llc < %s -mtriple=x86_64-- -mattr=+sse2 | FileCheck %s --check-prefixes=SSE,SSE2
; RUN: llc < %s -mtriple=x86_64-- -mattr=+sse4.1 | FileCheck %s --check-prefixes=SSE,SSE4
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx | FileCheck %s --check-prefixes=AVX,AVX1
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s --check-prefixes=AVX,AVX2
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512f | FileCheck %s --check-prefixes=AVX,AVX512F
; RUN: llc < %s -mtriple=x86_64-- -mattr=+avx512vl | FileCheck %s --check-prefixes=AVX,AVX512VL
define <4 x float> @sitofp_v4i32_v4f32(<2 x i32> %x, <2 x i32> %y) {
; SSE-LABEL: sitofp_v4i32_v4f32:
; SSE: # %bb.0:
; SSE-NEXT: movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT: cvtdq2ps %xmm0, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: sitofp_v4i32_v4f32:
; AVX: # %bb.0:
; AVX-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT: vcvtdq2ps %xmm0, %xmm0
; AVX-NEXT: retq
%s0 = sitofp <2 x i32> %x to <2 x float>
%s1 = sitofp <2 x i32> %y to <2 x float>
%r = shufflevector <2 x float> %s0, <2 x float> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x float> %r
}
define <4 x float> @uitofp_v4i32_v4f32(<2 x i32> %x, <2 x i32> %y) {
; SSE2-LABEL: uitofp_v4i32_v4f32:
; SSE2: # %bb.0:
; SSE2-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE2-NEXT: movdqa {{.*#+}} xmm1 = [65535,65535,65535,65535]
; SSE2-NEXT: pand %xmm0, %xmm1
; SSE2-NEXT: por {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm1
; SSE2-NEXT: psrld $16, %xmm0
; SSE2-NEXT: por {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE2-NEXT: subps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE2-NEXT: addps %xmm1, %xmm0
; SSE2-NEXT: retq
;
; SSE4-LABEL: uitofp_v4i32_v4f32:
; SSE4: # %bb.0:
; SSE4-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE4-NEXT: movdqa {{.*#+}} xmm1 = [1258291200,1258291200,1258291200,1258291200]
; SSE4-NEXT: pblendw {{.*#+}} xmm1 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
; SSE4-NEXT: psrld $16, %xmm0
; SSE4-NEXT: pblendw {{.*#+}} xmm0 = xmm0[0],mem[1],xmm0[2],mem[3],xmm0[4],mem[5],xmm0[6],mem[7]
; SSE4-NEXT: subps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE4-NEXT: addps %xmm1, %xmm0
; SSE4-NEXT: retq
;
; AVX1-LABEL: uitofp_v4i32_v4f32:
; AVX1: # %bb.0:
; AVX1-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX1-NEXT: vpblendw {{.*#+}} xmm1 = xmm0[0],mem[1],xmm0[2],mem[3],xmm0[4],mem[5],xmm0[6],mem[7]
; AVX1-NEXT: vpsrld $16, %xmm0, %xmm0
; AVX1-NEXT: vpblendw {{.*#+}} xmm0 = xmm0[0],mem[1],xmm0[2],mem[3],xmm0[4],mem[5],xmm0[6],mem[7]
; AVX1-NEXT: vsubps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX1-NEXT: vaddps %xmm0, %xmm1, %xmm0
; AVX1-NEXT: retq
;
; AVX2-LABEL: uitofp_v4i32_v4f32:
; AVX2: # %bb.0:
; AVX2-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX2-NEXT: vpbroadcastd {{.*#+}} xmm1 = [1258291200,1258291200,1258291200,1258291200]
; AVX2-NEXT: vpblendw {{.*#+}} xmm1 = xmm0[0],xmm1[1],xmm0[2],xmm1[3],xmm0[4],xmm1[5],xmm0[6],xmm1[7]
; AVX2-NEXT: vpsrld $16, %xmm0, %xmm0
; AVX2-NEXT: vpbroadcastd {{.*#+}} xmm2 = [1392508928,1392508928,1392508928,1392508928]
; AVX2-NEXT: vpblendw {{.*#+}} xmm0 = xmm0[0],xmm2[1],xmm0[2],xmm2[3],xmm0[4],xmm2[5],xmm0[6],xmm2[7]
; AVX2-NEXT: vbroadcastss {{.*#+}} xmm2 = [5.49764202E+11,5.49764202E+11,5.49764202E+11,5.49764202E+11]
; AVX2-NEXT: vsubps %xmm2, %xmm0, %xmm0
; AVX2-NEXT: vaddps %xmm0, %xmm1, %xmm0
; AVX2-NEXT: retq
;
; AVX512F-LABEL: uitofp_v4i32_v4f32:
; AVX512F: # %bb.0:
; AVX512F-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX512F-NEXT: vcvtudq2ps %zmm0, %zmm0
; AVX512F-NEXT: # kill: def $xmm0 killed $xmm0 killed $zmm0
; AVX512F-NEXT: vzeroupper
; AVX512F-NEXT: retq
;
; AVX512VL-LABEL: uitofp_v4i32_v4f32:
; AVX512VL: # %bb.0:
; AVX512VL-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX512VL-NEXT: vcvtudq2ps %xmm0, %xmm0
; AVX512VL-NEXT: retq
%s0 = uitofp <2 x i32> %x to <2 x float>
%s1 = uitofp <2 x i32> %y to <2 x float>
%r = shufflevector <2 x float> %s0, <2 x float> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x float> %r
}
define <4 x i32> @fptosi_v4f32_v4i32(<2 x float> %x, <2 x float> %y) {
; SSE-LABEL: fptosi_v4f32_v4i32:
; SSE: # %bb.0:
; SSE-NEXT: movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT: cvttps2dq %xmm0, %xmm0
; SSE-NEXT: retq
;
; AVX-LABEL: fptosi_v4f32_v4i32:
; AVX: # %bb.0:
; AVX-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT: vcvttps2dq %xmm0, %xmm0
; AVX-NEXT: retq
%s0 = fptosi <2 x float> %x to <2 x i32>
%s1 = fptosi <2 x float> %y to <2 x i32>
%r = shufflevector <2 x i32> %s0, <2 x i32> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x i32> %r
}
define <4 x i32> @fptoui_v4f32_v4i32(<2 x float> %x, <2 x float> %y) {
; SSE-LABEL: fptoui_v4f32_v4i32:
; SSE: # %bb.0:
; SSE-NEXT: movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT: cvttps2dq %xmm0, %xmm1
; SSE-NEXT: movdqa %xmm1, %xmm2
; SSE-NEXT: psrad $31, %xmm2
; SSE-NEXT: subps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0
; SSE-NEXT: cvttps2dq %xmm0, %xmm0
; SSE-NEXT: pand %xmm2, %xmm0
; SSE-NEXT: por %xmm1, %xmm0
; SSE-NEXT: retq
;
; AVX1-LABEL: fptoui_v4f32_v4i32:
; AVX1: # %bb.0:
; AVX1-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX1-NEXT: vcvttps2dq %xmm0, %xmm1
; AVX1-NEXT: vpsrad $31, %xmm1, %xmm2
; AVX1-NEXT: vsubps {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %xmm0, %xmm0
; AVX1-NEXT: vcvttps2dq %xmm0, %xmm0
; AVX1-NEXT: vpand %xmm2, %xmm0, %xmm0
; AVX1-NEXT: vpor %xmm0, %xmm1, %xmm0
; AVX1-NEXT: retq
;
; AVX2-LABEL: fptoui_v4f32_v4i32:
; AVX2: # %bb.0:
; AVX2-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX2-NEXT: vcvttps2dq %xmm0, %xmm1
; AVX2-NEXT: vpsrad $31, %xmm1, %xmm2
; AVX2-NEXT: vbroadcastss {{.*#+}} xmm3 = [2.14748365E+9,2.14748365E+9,2.14748365E+9,2.14748365E+9]
; AVX2-NEXT: vsubps %xmm3, %xmm0, %xmm0
; AVX2-NEXT: vcvttps2dq %xmm0, %xmm0
; AVX2-NEXT: vpand %xmm2, %xmm0, %xmm0
; AVX2-NEXT: vpor %xmm0, %xmm1, %xmm0
; AVX2-NEXT: retq
;
; AVX512F-LABEL: fptoui_v4f32_v4i32:
; AVX512F: # %bb.0:
; AVX512F-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX512F-NEXT: vcvttps2udq %zmm0, %zmm0
; AVX512F-NEXT: # kill: def $xmm0 killed $xmm0 killed $zmm0
; AVX512F-NEXT: vzeroupper
; AVX512F-NEXT: retq
;
; AVX512VL-LABEL: fptoui_v4f32_v4i32:
; AVX512VL: # %bb.0:
; AVX512VL-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX512VL-NEXT: vcvttps2udq %xmm0, %xmm0
; AVX512VL-NEXT: retq
%s0 = fptoui <2 x float> %x to <2 x i32>
%s1 = fptoui <2 x float> %y to <2 x i32>
%r = shufflevector <2 x i32> %s0, <2 x i32> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x i32> %r
}
define <4 x double> @sitofp_v4i32_v4f64(<2 x i32> %x, <2 x i32> %y) {
; SSE-LABEL: sitofp_v4i32_v4f64:
; SSE: # %bb.0:
; SSE-NEXT: cvtdq2pd %xmm0, %xmm0
; SSE-NEXT: cvtdq2pd %xmm1, %xmm1
; SSE-NEXT: retq
;
; AVX-LABEL: sitofp_v4i32_v4f64:
; AVX: # %bb.0:
; AVX-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT: vcvtdq2pd %xmm0, %ymm0
; AVX-NEXT: retq
%s0 = sitofp <2 x i32> %x to <2 x double>
%s1 = sitofp <2 x i32> %y to <2 x double>
%r = shufflevector <2 x double> %s0, <2 x double> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x double> %r
}
define <4 x double> @uitofp_v4i32_v4f64(<2 x i32> %x, <2 x i32> %y) {
; SSE2-LABEL: uitofp_v4i32_v4f64:
; SSE2: # %bb.0:
; SSE2-NEXT: xorpd %xmm2, %xmm2
; SSE2-NEXT: unpcklps {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
; SSE2-NEXT: movapd {{.*#+}} xmm3 = [4.503599627370496E+15,4.503599627370496E+15]
; SSE2-NEXT: orpd %xmm3, %xmm0
; SSE2-NEXT: subpd %xmm3, %xmm0
; SSE2-NEXT: unpcklps {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
; SSE2-NEXT: orpd %xmm3, %xmm1
; SSE2-NEXT: subpd %xmm3, %xmm1
; SSE2-NEXT: retq
;
; SSE4-LABEL: uitofp_v4i32_v4f64:
; SSE4: # %bb.0:
; SSE4-NEXT: pmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
; SSE4-NEXT: movdqa {{.*#+}} xmm2 = [4.503599627370496E+15,4.503599627370496E+15]
; SSE4-NEXT: por %xmm2, %xmm0
; SSE4-NEXT: subpd %xmm2, %xmm0
; SSE4-NEXT: pmovzxdq {{.*#+}} xmm1 = xmm1[0],zero,xmm1[1],zero
; SSE4-NEXT: por %xmm2, %xmm1
; SSE4-NEXT: subpd %xmm2, %xmm1
; SSE4-NEXT: retq
;
; AVX1-LABEL: uitofp_v4i32_v4f64:
; AVX1: # %bb.0:
; AVX1-NEXT: vpmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
; AVX1-NEXT: vpmovzxdq {{.*#+}} xmm1 = xmm1[0],zero,xmm1[1],zero
; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX1-NEXT: vbroadcastsd {{.*#+}} ymm1 = [4.503599627370496E+15,4.503599627370496E+15,4.503599627370496E+15,4.503599627370496E+15]
; AVX1-NEXT: vorpd %ymm1, %ymm0, %ymm0
; AVX1-NEXT: vsubpd %ymm1, %ymm0, %ymm0
; AVX1-NEXT: retq
;
; AVX2-LABEL: uitofp_v4i32_v4f64:
; AVX2: # %bb.0:
; AVX2-NEXT: vpunpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX2-NEXT: vpmovzxdq {{.*#+}} ymm0 = xmm0[0],zero,xmm0[1],zero,xmm0[2],zero,xmm0[3],zero
; AVX2-NEXT: vpbroadcastq {{.*#+}} ymm1 = [4.503599627370496E+15,4.503599627370496E+15,4.503599627370496E+15,4.503599627370496E+15]
; AVX2-NEXT: vpor %ymm1, %ymm0, %ymm0
; AVX2-NEXT: vsubpd %ymm1, %ymm0, %ymm0
; AVX2-NEXT: retq
;
; AVX512F-LABEL: uitofp_v4i32_v4f64:
; AVX512F: # %bb.0:
; AVX512F-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX512F-NEXT: vcvtudq2pd %ymm0, %zmm0
; AVX512F-NEXT: # kill: def $ymm0 killed $ymm0 killed $zmm0
; AVX512F-NEXT: retq
;
; AVX512VL-LABEL: uitofp_v4i32_v4f64:
; AVX512VL: # %bb.0:
; AVX512VL-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX512VL-NEXT: vcvtudq2pd %xmm0, %ymm0
; AVX512VL-NEXT: retq
%s0 = uitofp <2 x i32> %x to <2 x double>
%s1 = uitofp <2 x i32> %y to <2 x double>
%r = shufflevector <2 x double> %s0, <2 x double> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x double> %r
}
define <4 x i32> @fptosi_v4f64_v4i32(<2 x double> %x, <2 x double> %y) {
; SSE-LABEL: fptosi_v4f64_v4i32:
; SSE: # %bb.0:
; SSE-NEXT: cvttpd2dq %xmm0, %xmm0
; SSE-NEXT: cvttpd2dq %xmm1, %xmm1
; SSE-NEXT: unpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT: retq
;
; AVX-LABEL: fptosi_v4f64_v4i32:
; AVX: # %bb.0:
; AVX-NEXT: # kill: def $xmm0 killed $xmm0 def $ymm0
; AVX-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-NEXT: vcvttpd2dq %ymm0, %xmm0
; AVX-NEXT: vzeroupper
; AVX-NEXT: retq
%s0 = fptosi <2 x double> %x to <2 x i32>
%s1 = fptosi <2 x double> %y to <2 x i32>
%r = shufflevector <2 x i32> %s0, <2 x i32> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x i32> %r
}
define <4 x i32> @fptoui_v4f64_v4i32(<2 x double> %x, <2 x double> %y) {
; SSE-LABEL: fptoui_v4f64_v4i32:
; SSE: # %bb.0:
; SSE-NEXT: movapd {{.*#+}} xmm2 = [2.147483648E+9,2.147483648E+9]
; SSE-NEXT: cvttpd2dq %xmm0, %xmm3
; SSE-NEXT: subpd %xmm2, %xmm0
; SSE-NEXT: cvttpd2dq %xmm0, %xmm4
; SSE-NEXT: movapd %xmm3, %xmm0
; SSE-NEXT: psrad $31, %xmm0
; SSE-NEXT: pand %xmm4, %xmm0
; SSE-NEXT: por %xmm3, %xmm0
; SSE-NEXT: cvttpd2dq %xmm1, %xmm3
; SSE-NEXT: subpd %xmm2, %xmm1
; SSE-NEXT: cvttpd2dq %xmm1, %xmm1
; SSE-NEXT: movapd %xmm3, %xmm2
; SSE-NEXT: psrad $31, %xmm2
; SSE-NEXT: pand %xmm1, %xmm2
; SSE-NEXT: por %xmm3, %xmm2
; SSE-NEXT: punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
; SSE-NEXT: retq
;
; AVX1-LABEL: fptoui_v4f64_v4i32:
; AVX1: # %bb.0:
; AVX1-NEXT: # kill: def $xmm0 killed $xmm0 def $ymm0
; AVX1-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX1-NEXT: vcvttpd2dq %ymm0, %xmm1
; AVX1-NEXT: vpsrad $31, %xmm1, %xmm2
; AVX1-NEXT: vsubpd {{\.?LCPI[0-9]+_[0-9]+}}(%rip), %ymm0, %ymm0
; AVX1-NEXT: vcvttpd2dq %ymm0, %xmm0
; AVX1-NEXT: vandpd %xmm2, %xmm0, %xmm0
; AVX1-NEXT: vorpd %xmm0, %xmm1, %xmm0
; AVX1-NEXT: vzeroupper
; AVX1-NEXT: retq
;
; AVX2-LABEL: fptoui_v4f64_v4i32:
; AVX2: # %bb.0:
; AVX2-NEXT: # kill: def $xmm0 killed $xmm0 def $ymm0
; AVX2-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX2-NEXT: vcvttpd2dq %ymm0, %xmm1
; AVX2-NEXT: vpsrad $31, %xmm1, %xmm2
; AVX2-NEXT: vbroadcastsd {{.*#+}} ymm3 = [2.147483648E+9,2.147483648E+9,2.147483648E+9,2.147483648E+9]
; AVX2-NEXT: vsubpd %ymm3, %ymm0, %ymm0
; AVX2-NEXT: vcvttpd2dq %ymm0, %xmm0
; AVX2-NEXT: vandpd %xmm2, %xmm0, %xmm0
; AVX2-NEXT: vorpd %xmm0, %xmm1, %xmm0
; AVX2-NEXT: vzeroupper
; AVX2-NEXT: retq
;
; AVX512F-LABEL: fptoui_v4f64_v4i32:
; AVX512F: # %bb.0:
; AVX512F-NEXT: # kill: def $xmm0 killed $xmm0 def $ymm0
; AVX512F-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX512F-NEXT: vcvttpd2udq %zmm0, %ymm0
; AVX512F-NEXT: # kill: def $xmm0 killed $xmm0 killed $ymm0
; AVX512F-NEXT: vzeroupper
; AVX512F-NEXT: retq
;
; AVX512VL-LABEL: fptoui_v4f64_v4i32:
; AVX512VL: # %bb.0:
; AVX512VL-NEXT: # kill: def $xmm0 killed $xmm0 def $ymm0
; AVX512VL-NEXT: vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX512VL-NEXT: vcvttpd2udq %ymm0, %xmm0
; AVX512VL-NEXT: vzeroupper
; AVX512VL-NEXT: retq
%s0 = fptoui <2 x double> %x to <2 x i32>
%s1 = fptoui <2 x double> %y to <2 x i32>
%r = shufflevector <2 x i32> %s0, <2 x i32> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x i32> %r
}
; Negative test
define <4 x float> @mismatch_tofp_v4i32_v4f32(<2 x i32> %x, <2 x i32> %y) {
; SSE2-LABEL: mismatch_tofp_v4i32_v4f32:
; SSE2: # %bb.0:
; SSE2-NEXT: xorpd %xmm2, %xmm2
; SSE2-NEXT: unpcklps {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
; SSE2-NEXT: movapd {{.*#+}} xmm2 = [4.503599627370496E+15,4.503599627370496E+15]
; SSE2-NEXT: orpd %xmm2, %xmm0
; SSE2-NEXT: subpd %xmm2, %xmm0
; SSE2-NEXT: cvtpd2ps %xmm0, %xmm0
; SSE2-NEXT: cvtdq2ps %xmm1, %xmm1
; SSE2-NEXT: unpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE2-NEXT: retq
;
; SSE4-LABEL: mismatch_tofp_v4i32_v4f32:
; SSE4: # %bb.0:
; SSE4-NEXT: pmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
; SSE4-NEXT: movdqa {{.*#+}} xmm2 = [4.503599627370496E+15,4.503599627370496E+15]
; SSE4-NEXT: por %xmm2, %xmm0
; SSE4-NEXT: subpd %xmm2, %xmm0
; SSE4-NEXT: cvtpd2ps %xmm0, %xmm0
; SSE4-NEXT: cvtdq2ps %xmm1, %xmm1
; SSE4-NEXT: unpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE4-NEXT: retq
;
; AVX1-LABEL: mismatch_tofp_v4i32_v4f32:
; AVX1: # %bb.0:
; AVX1-NEXT: vpmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
; AVX1-NEXT: vmovddup {{.*#+}} xmm2 = [4.503599627370496E+15,4.503599627370496E+15]
; AVX1-NEXT: # xmm2 = mem[0,0]
; AVX1-NEXT: vpor %xmm2, %xmm0, %xmm0
; AVX1-NEXT: vsubpd %xmm2, %xmm0, %xmm0
; AVX1-NEXT: vcvtpd2ps %xmm0, %xmm0
; AVX1-NEXT: vcvtdq2ps %xmm1, %xmm1
; AVX1-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX1-NEXT: retq
;
; AVX2-LABEL: mismatch_tofp_v4i32_v4f32:
; AVX2: # %bb.0:
; AVX2-NEXT: vpmovzxdq {{.*#+}} xmm0 = xmm0[0],zero,xmm0[1],zero
; AVX2-NEXT: vpbroadcastq {{.*#+}} xmm2 = [4.503599627370496E+15,4.503599627370496E+15]
; AVX2-NEXT: vpor %xmm2, %xmm0, %xmm0
; AVX2-NEXT: vsubpd %xmm2, %xmm0, %xmm0
; AVX2-NEXT: vcvtpd2ps %xmm0, %xmm0
; AVX2-NEXT: vcvtdq2ps %xmm1, %xmm1
; AVX2-NEXT: vunpcklpd {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX2-NEXT: retq
;
; AVX512F-LABEL: mismatch_tofp_v4i32_v4f32:
; AVX512F: # %bb.0:
; AVX512F-NEXT: # kill: def $xmm0 killed $xmm0 def $zmm0
; AVX512F-NEXT: vcvtudq2ps %zmm0, %zmm0
; AVX512F-NEXT: vcvtdq2ps %xmm1, %xmm1
; AVX512F-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX512F-NEXT: vzeroupper
; AVX512F-NEXT: retq
;
; AVX512VL-LABEL: mismatch_tofp_v4i32_v4f32:
; AVX512VL: # %bb.0:
; AVX512VL-NEXT: vcvtudq2ps %xmm0, %xmm0
; AVX512VL-NEXT: vcvtdq2ps %xmm1, %xmm1
; AVX512VL-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX512VL-NEXT: retq
%s0 = uitofp <2 x i32> %x to <2 x float>
%s1 = sitofp <2 x i32> %y to <2 x float>
%r = shufflevector <2 x float> %s0, <2 x float> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x float> %r
}
; Negative test
define <4 x float> @sitofp_v4i32_v4f32_extra_use(<2 x i32> %x, <2 x i32> %y, ptr %p) {
; SSE-LABEL: sitofp_v4i32_v4f32_extra_use:
; SSE: # %bb.0:
; SSE-NEXT: cvtdq2ps %xmm0, %xmm0
; SSE-NEXT: cvtdq2ps %xmm1, %xmm1
; SSE-NEXT: movlps %xmm1, (%rdi)
; SSE-NEXT: movlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE-NEXT: retq
;
; AVX-LABEL: sitofp_v4i32_v4f32_extra_use:
; AVX: # %bb.0:
; AVX-NEXT: vcvtdq2ps %xmm0, %xmm0
; AVX-NEXT: vcvtdq2ps %xmm1, %xmm1
; AVX-NEXT: vmovlps %xmm1, (%rdi)
; AVX-NEXT: vmovlhps {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; AVX-NEXT: retq
%s0 = sitofp <2 x i32> %x to <2 x float>
%s1 = sitofp <2 x i32> %y to <2 x float>
store <2 x float> %s1, ptr %p
%r = shufflevector <2 x float> %s0, <2 x float> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x float> %r
}
define <4 x float> @PR45794(<2 x i64> %x, <2 x i64> %y) {
; SSE-LABEL: PR45794:
; SSE: # %bb.0:
; SSE-NEXT: psrad $16, %xmm0
; SSE-NEXT: psrad $16, %xmm1
; SSE-NEXT: shufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; SSE-NEXT: cvtdq2ps %xmm0, %xmm0
; SSE-NEXT: retq
;
; AVX1-LABEL: PR45794:
; AVX1: # %bb.0:
; AVX1-NEXT: vpsrad $16, %xmm0, %xmm0
; AVX1-NEXT: vpsrad $16, %xmm1, %xmm1
; AVX1-NEXT: vshufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; AVX1-NEXT: vcvtdq2ps %xmm0, %xmm0
; AVX1-NEXT: retq
;
; AVX2-LABEL: PR45794:
; AVX2: # %bb.0:
; AVX2-NEXT: vpsrad $16, %xmm0, %xmm0
; AVX2-NEXT: vpsrad $16, %xmm1, %xmm1
; AVX2-NEXT: vshufps {{.*#+}} xmm0 = xmm0[1,3],xmm1[1,3]
; AVX2-NEXT: vcvtdq2ps %xmm0, %xmm0
; AVX2-NEXT: retq
;
; AVX512F-LABEL: PR45794:
; AVX512F: # %bb.0:
; AVX512F-NEXT: # kill: def $xmm1 killed $xmm1 def $zmm1
; AVX512F-NEXT: # kill: def $xmm0 killed $xmm0 def $zmm0
; AVX512F-NEXT: vpsraq $48, %zmm0, %zmm0
; AVX512F-NEXT: vpsraq $48, %zmm1, %zmm1
; AVX512F-NEXT: vpackssdw %xmm1, %xmm0, %xmm0
; AVX512F-NEXT: vcvtdq2ps %xmm0, %xmm0
; AVX512F-NEXT: vzeroupper
; AVX512F-NEXT: retq
;
; AVX512VL-LABEL: PR45794:
; AVX512VL: # %bb.0:
; AVX512VL-NEXT: vpsraq $48, %xmm0, %xmm0
; AVX512VL-NEXT: vpsraq $48, %xmm1, %xmm1
; AVX512VL-NEXT: vpackssdw %xmm1, %xmm0, %xmm0
; AVX512VL-NEXT: vcvtdq2ps %xmm0, %xmm0
; AVX512VL-NEXT: retq
%a0 = ashr <2 x i64> %x, <i64 48, i64 48>
%s0 = sitofp <2 x i64> %a0 to <2 x float>
%a1 = ashr <2 x i64> %y, <i64 48, i64 48>
%s1 = sitofp <2 x i64> %a1 to <2 x float>
%r = shufflevector <2 x float> %s0, <2 x float> %s1, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
ret <4 x float> %r
}
|
