llvm.org GIT mirror llvm / release_37 test / CodeGen / X86 / vec_fp_to_int.ll
release_37

Tree @release_37 (Download .tar.gz)

vec_fp_to_int.ll @release_37raw · 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
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+sse2 | FileCheck %s --check-prefix=ALL --check-prefix=SSE --check-prefix=SSE2
; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx | FileCheck %s --check-prefix=ALL --check-prefix=AVX --check-prefix=AVX1
;
; 32-bit tests to make sure we're not doing anything stupid.
; RUN: llc < %s -mtriple=i686-unknown-unknown
; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+sse
; RUN: llc < %s -mtriple=i686-unknown-unknown -mattr=+sse2

;
; Double to Signed Integer
;

define <2 x i64> @fptosi_2vf64(<2 x double> %a) {
; SSE2-LABEL: fptosi_2vf64:
; SSE2:       # BB#0:
; SSE2-NEXT:    cvttsd2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm1
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    cvttsd2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE2-NEXT:    movdqa %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_2vf64:
; AVX:       # BB#0:
; AVX-NEXT:    vcvttsd2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm1
; AVX-NEXT:    vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX-NEXT:    vcvttsd2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    retq
  %cvt = fptosi <2 x double> %a to <2 x i64>
  ret <2 x i64> %cvt
}

define <4 x i32> @fptosi_2vf64_i32(<2 x double> %a) {
; SSE2-LABEL: fptosi_2vf64_i32:
; SSE2:       # BB#0:
; SSE2-NEXT:    cvttsd2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm1
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    cvttsd2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm1[0,2,2,3]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_2vf64_i32:
; AVX:       # BB#0:
; AVX-NEXT:    vcvttsd2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm1
; AVX-NEXT:    vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX-NEXT:    vcvttsd2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
; AVX-NEXT:    retq
  %cvt = fptosi <2 x double> %a to <2 x i32>
  %ext = shufflevector <2 x i32> %cvt, <2 x i32> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  ret <4 x i32> %ext
}

define <4 x i64> @fptosi_4vf64(<4 x double> %a) {
; SSE2-LABEL: fptosi_4vf64:
; SSE2:       # BB#0:
; SSE2-NEXT:    cvttsd2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm2
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    cvttsd2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE2-NEXT:    cvttsd2si %xmm1, %rax
; SSE2-NEXT:    movd %rax, %xmm3
; SSE2-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[1,0]
; SSE2-NEXT:    cvttsd2si %xmm1, %rax
; SSE2-NEXT:    movd %rax, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm3 = xmm3[0],xmm0[0]
; SSE2-NEXT:    movdqa %xmm2, %xmm0
; SSE2-NEXT:    movdqa %xmm3, %xmm1
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_4vf64:
; AVX:       # BB#0:
; AVX-NEXT:    vextractf128 $1, %ymm0, %xmm1
; AVX-NEXT:    vcvttsd2si %xmm1, %rax
; AVX-NEXT:    vmovq %rax, %xmm2
; AVX-NEXT:    vpermilpd {{.*#+}} xmm1 = xmm1[1,0]
; AVX-NEXT:    vcvttsd2si %xmm1, %rax
; AVX-NEXT:    vmovq %rax, %xmm1
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm2[0],xmm1[0]
; AVX-NEXT:    vcvttsd2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm2
; AVX-NEXT:    vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX-NEXT:    vcvttsd2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm2[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-NEXT:    retq
  %cvt = fptosi <4 x double> %a to <4 x i64>
  ret <4 x i64> %cvt
}

define <4 x i32> @fptosi_4vf64_i32(<4 x double> %a) {
; SSE2-LABEL: fptosi_4vf64_i32:
; SSE2:       # BB#0:
; SSE2-NEXT:    cvttsd2si %xmm1, %rax
; SSE2-NEXT:    movd %rax, %xmm2
; SSE2-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[1,0]
; SSE2-NEXT:    cvttsd2si %xmm1, %rax
; SSE2-NEXT:    movd %rax, %xmm1
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm2[0,2,2,3]
; SSE2-NEXT:    cvttsd2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm2
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    cvttsd2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm2[0,2,2,3]
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_4vf64_i32:
; AVX:       # BB#0:
; AVX-NEXT:    vcvttpd2dqy %ymm0, %xmm0
; AVX-NEXT:    vzeroupper
; AVX-NEXT:    retq
  %cvt = fptosi <4 x double> %a to <4 x i32>
  ret <4 x i32> %cvt
}

;
; Double to Unsigned Integer
;

define <2 x i64> @fptoui_2vf64(<2 x double> %a) {
; SSE2-LABEL: fptoui_2vf64:
; SSE2:       # BB#0:
; SSE2-NEXT:    movsd {{.*#+}} xmm2 = mem[0],zero
; SSE2-NEXT:    movapd %xmm0, %xmm1
; SSE2-NEXT:    subsd %xmm2, %xmm1
; SSE2-NEXT:    cvttsd2si %xmm1, %rax
; SSE2-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; SSE2-NEXT:    xorq %rcx, %rax
; SSE2-NEXT:    cvttsd2si %xmm0, %rdx
; SSE2-NEXT:    ucomisd %xmm2, %xmm0
; SSE2-NEXT:    cmovaeq %rax, %rdx
; SSE2-NEXT:    movd %rdx, %xmm1
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    movapd %xmm0, %xmm3
; SSE2-NEXT:    subsd %xmm2, %xmm3
; SSE2-NEXT:    cvttsd2si %xmm3, %rax
; SSE2-NEXT:    xorq %rcx, %rax
; SSE2-NEXT:    cvttsd2si %xmm0, %rcx
; SSE2-NEXT:    ucomisd %xmm2, %xmm0
; SSE2-NEXT:    cmovaeq %rax, %rcx
; SSE2-NEXT:    movd %rcx, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE2-NEXT:    movdqa %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_2vf64:
; AVX:       # BB#0:
; AVX-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
; AVX-NEXT:    vsubsd %xmm1, %xmm0, %xmm2
; AVX-NEXT:    vcvttsd2si %xmm2, %rax
; AVX-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttsd2si %xmm0, %rdx
; AVX-NEXT:    vucomisd %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rdx
; AVX-NEXT:    vmovq %rdx, %xmm2
; AVX-NEXT:    vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX-NEXT:    vsubsd %xmm1, %xmm0, %xmm3
; AVX-NEXT:    vcvttsd2si %xmm3, %rax
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttsd2si %xmm0, %rcx
; AVX-NEXT:    vucomisd %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rcx
; AVX-NEXT:    vmovq %rcx, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm2[0],xmm0[0]
; AVX-NEXT:    retq
  %cvt = fptoui <2 x double> %a to <2 x i64>
  ret <2 x i64> %cvt
}

define <4 x i32> @fptoui_2vf64_i32(<2 x double> %a) {
; SSE2-LABEL: fptoui_2vf64_i32:
; SSE2:       # BB#0:
; SSE2-NEXT:    movsd {{.*#+}} xmm1 = mem[0],zero
; SSE2-NEXT:    movapd %xmm0, %xmm2
; SSE2-NEXT:    subsd %xmm1, %xmm2
; SSE2-NEXT:    cvttsd2si %xmm2, %rax
; SSE2-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; SSE2-NEXT:    xorq %rcx, %rax
; SSE2-NEXT:    cvttsd2si %xmm0, %rdx
; SSE2-NEXT:    ucomisd %xmm1, %xmm0
; SSE2-NEXT:    cmovaeq %rax, %rdx
; SSE2-NEXT:    movd %rdx, %xmm2
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    movapd %xmm0, %xmm3
; SSE2-NEXT:    subsd %xmm1, %xmm3
; SSE2-NEXT:    cvttsd2si %xmm3, %rax
; SSE2-NEXT:    xorq %rcx, %rax
; SSE2-NEXT:    cvttsd2si %xmm0, %rcx
; SSE2-NEXT:    ucomisd %xmm1, %xmm0
; SSE2-NEXT:    cmovaeq %rax, %rcx
; SSE2-NEXT:    movd %rcx, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm0[0]
; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm2[0,2,2,3]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_2vf64_i32:
; AVX:       # BB#0:
; AVX-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
; AVX-NEXT:    vsubsd %xmm1, %xmm0, %xmm2
; AVX-NEXT:    vcvttsd2si %xmm2, %rax
; AVX-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttsd2si %xmm0, %rdx
; AVX-NEXT:    vucomisd %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rdx
; AVX-NEXT:    vmovq %rdx, %xmm2
; AVX-NEXT:    vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX-NEXT:    vsubsd %xmm1, %xmm0, %xmm3
; AVX-NEXT:    vcvttsd2si %xmm3, %rax
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttsd2si %xmm0, %rcx
; AVX-NEXT:    vucomisd %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rcx
; AVX-NEXT:    vmovq %rcx, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm2[0],xmm0[0]
; AVX-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,2,2,3]
; AVX-NEXT:    retq
  %cvt = fptoui <2 x double> %a to <2 x i32>
  %ext = shufflevector <2 x i32> %cvt, <2 x i32> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  ret <4 x i32> %ext
}

define <4 x i64> @fptoui_4vf64(<4 x double> %a) {
; SSE2-LABEL: fptoui_4vf64:
; SSE2:       # BB#0:
; SSE2-NEXT:    movapd %xmm0, %xmm2
; SSE2-NEXT:    movsd {{.*#+}} xmm3 = mem[0],zero
; SSE2-NEXT:    subsd %xmm3, %xmm0
; SSE2-NEXT:    cvttsd2si %xmm0, %rcx
; SSE2-NEXT:    movabsq $-9223372036854775808, %rax # imm = 0x8000000000000000
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttsd2si %xmm2, %rdx
; SSE2-NEXT:    ucomisd %xmm3, %xmm2
; SSE2-NEXT:    cmovaeq %rcx, %rdx
; SSE2-NEXT:    movd %rdx, %xmm0
; SSE2-NEXT:    shufpd {{.*#+}} xmm2 = xmm2[1,0]
; SSE2-NEXT:    movapd %xmm2, %xmm4
; SSE2-NEXT:    subsd %xmm3, %xmm4
; SSE2-NEXT:    cvttsd2si %xmm4, %rcx
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttsd2si %xmm2, %rdx
; SSE2-NEXT:    ucomisd %xmm3, %xmm2
; SSE2-NEXT:    cmovaeq %rcx, %rdx
; SSE2-NEXT:    movd %rdx, %xmm2
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm2[0]
; SSE2-NEXT:    movapd %xmm1, %xmm2
; SSE2-NEXT:    subsd %xmm3, %xmm2
; SSE2-NEXT:    cvttsd2si %xmm2, %rcx
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttsd2si %xmm1, %rdx
; SSE2-NEXT:    ucomisd %xmm3, %xmm1
; SSE2-NEXT:    cmovaeq %rcx, %rdx
; SSE2-NEXT:    movd %rdx, %xmm2
; SSE2-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[1,0]
; SSE2-NEXT:    movapd %xmm1, %xmm4
; SSE2-NEXT:    subsd %xmm3, %xmm4
; SSE2-NEXT:    cvttsd2si %xmm4, %rcx
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttsd2si %xmm1, %rax
; SSE2-NEXT:    ucomisd %xmm3, %xmm1
; SSE2-NEXT:    cmovaeq %rcx, %rax
; SSE2-NEXT:    movd %rax, %xmm1
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
; SSE2-NEXT:    movdqa %xmm2, %xmm1
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_4vf64:
; AVX:       # BB#0:
; AVX-NEXT:    vextractf128 $1, %ymm0, %xmm2
; AVX-NEXT:    vmovsd {{.*#+}} xmm1 = mem[0],zero
; AVX-NEXT:    vsubsd %xmm1, %xmm2, %xmm3
; AVX-NEXT:    vcvttsd2si %xmm3, %rax
; AVX-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttsd2si %xmm2, %rdx
; AVX-NEXT:    vucomisd %xmm1, %xmm2
; AVX-NEXT:    cmovaeq %rax, %rdx
; AVX-NEXT:    vmovq %rdx, %xmm3
; AVX-NEXT:    vpermilpd {{.*#+}} xmm2 = xmm2[1,0]
; AVX-NEXT:    vsubsd %xmm1, %xmm2, %xmm4
; AVX-NEXT:    vcvttsd2si %xmm4, %rax
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttsd2si %xmm2, %rdx
; AVX-NEXT:    vucomisd %xmm1, %xmm2
; AVX-NEXT:    cmovaeq %rax, %rdx
; AVX-NEXT:    vmovq %rdx, %xmm2
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
; AVX-NEXT:    vsubsd %xmm1, %xmm0, %xmm3
; AVX-NEXT:    vcvttsd2si %xmm3, %rax
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttsd2si %xmm0, %rdx
; AVX-NEXT:    vucomisd %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rdx
; AVX-NEXT:    vmovq %rdx, %xmm3
; AVX-NEXT:    vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX-NEXT:    vsubsd %xmm1, %xmm0, %xmm4
; AVX-NEXT:    vcvttsd2si %xmm4, %rax
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttsd2si %xmm0, %rcx
; AVX-NEXT:    vucomisd %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rcx
; AVX-NEXT:    vmovq %rcx, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm3[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    retq
  %cvt = fptoui <4 x double> %a to <4 x i64>
  ret <4 x i64> %cvt
}

define <4 x i32> @fptoui_4vf64_i32(<4 x double> %a) {
; SSE2-LABEL: fptoui_4vf64_i32:
; SSE2:       # BB#0:
; SSE2-NEXT:    movsd {{.*#+}} xmm2 = mem[0],zero
; SSE2-NEXT:    movapd %xmm1, %xmm3
; SSE2-NEXT:    subsd %xmm2, %xmm3
; SSE2-NEXT:    cvttsd2si %xmm3, %rcx
; SSE2-NEXT:    movabsq $-9223372036854775808, %rax # imm = 0x8000000000000000
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttsd2si %xmm1, %rdx
; SSE2-NEXT:    ucomisd %xmm2, %xmm1
; SSE2-NEXT:    cmovaeq %rcx, %rdx
; SSE2-NEXT:    movd %rdx, %xmm3
; SSE2-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[1,0]
; SSE2-NEXT:    movapd %xmm1, %xmm4
; SSE2-NEXT:    subsd %xmm2, %xmm4
; SSE2-NEXT:    cvttsd2si %xmm4, %rcx
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttsd2si %xmm1, %rdx
; SSE2-NEXT:    ucomisd %xmm2, %xmm1
; SSE2-NEXT:    cmovaeq %rcx, %rdx
; SSE2-NEXT:    movd %rdx, %xmm1
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm3 = xmm3[0],xmm1[0]
; SSE2-NEXT:    pshufd {{.*#+}} xmm1 = xmm3[0,2,2,3]
; SSE2-NEXT:    movapd %xmm0, %xmm3
; SSE2-NEXT:    subsd %xmm2, %xmm3
; SSE2-NEXT:    cvttsd2si %xmm3, %rcx
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttsd2si %xmm0, %rdx
; SSE2-NEXT:    ucomisd %xmm2, %xmm0
; SSE2-NEXT:    cmovaeq %rcx, %rdx
; SSE2-NEXT:    movd %rdx, %xmm3
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    movapd %xmm0, %xmm4
; SSE2-NEXT:    subsd %xmm2, %xmm4
; SSE2-NEXT:    cvttsd2si %xmm4, %rcx
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttsd2si %xmm0, %rax
; SSE2-NEXT:    ucomisd %xmm2, %xmm0
; SSE2-NEXT:    cmovaeq %rcx, %rax
; SSE2-NEXT:    movd %rax, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm3 = xmm3[0],xmm0[0]
; SSE2-NEXT:    pshufd {{.*#+}} xmm0 = xmm3[0,2,2,3]
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm0 = xmm0[0],xmm1[0]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_4vf64_i32:
; AVX:       # BB#0:
; AVX-NEXT:    vpermilpd {{.*#+}} xmm1 = xmm0[1,0]
; AVX-NEXT:    vcvttsd2si %xmm1, %rax
; AVX-NEXT:    vcvttsd2si %xmm0, %rcx
; AVX-NEXT:    vmovd %ecx, %xmm1
; AVX-NEXT:    vpinsrd $1, %eax, %xmm1, %xmm1
; AVX-NEXT:    vextractf128 $1, %ymm0, %xmm0
; AVX-NEXT:    vcvttsd2si %xmm0, %rax
; AVX-NEXT:    vpinsrd $2, %eax, %xmm1, %xmm1
; AVX-NEXT:    vpermilpd {{.*#+}} xmm0 = xmm0[1,0]
; AVX-NEXT:    vcvttsd2si %xmm0, %rax
; AVX-NEXT:    vpinsrd $3, %eax, %xmm1, %xmm0
; AVX-NEXT:    vzeroupper
; AVX-NEXT:    retq
  %cvt = fptoui <4 x double> %a to <4 x i32>
  ret <4 x i32> %cvt
}

;
; Float to Signed Integer
;

define <4 x i32> @fptosi_4vf32(<4 x float> %a) {
; SSE2-LABEL: fptosi_4vf32:
; SSE2:       # BB#0:
; SSE2-NEXT:    cvttps2dq %xmm0, %xmm0
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_4vf32:
; AVX:       # BB#0:
; AVX-NEXT:    vcvttps2dq %xmm0, %xmm0
; AVX-NEXT:    retq
  %cvt = fptosi <4 x float> %a to <4 x i32>
  ret <4 x i32> %cvt
}

define <2 x i64> @fptosi_4vf32_i64(<4 x float> %a) {
; SSE2-LABEL: fptosi_4vf32_i64:
; SSE2:       # BB#0:
; SSE2-NEXT:    cvttss2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,2,3]
; SSE2-NEXT:    cvttss2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE2-NEXT:    movdqa %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_4vf32_i64:
; AVX:       # BB#0:
; AVX-NEXT:    vcvttss2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm1
; AVX-NEXT:    vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX-NEXT:    vcvttss2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm1[0],xmm0[0]
; AVX-NEXT:    retq
  %shuf = shufflevector <4 x float> %a, <4 x float> %a, <2 x i32> <i32 0, i32 1>
  %cvt = fptosi <2 x float> %shuf to <2 x i64>
  ret <2 x i64> %cvt
}

define <8 x i32> @fptosi_8vf32(<8 x float> %a) {
; SSE2-LABEL: fptosi_8vf32:
; SSE2:       # BB#0:
; SSE2-NEXT:    cvttps2dq %xmm0, %xmm0
; SSE2-NEXT:    cvttps2dq %xmm1, %xmm1
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_8vf32:
; AVX:       # BB#0:
; AVX-NEXT:    vcvttps2dq %ymm0, %ymm0
; AVX-NEXT:    retq
  %cvt = fptosi <8 x float> %a to <8 x i32>
  ret <8 x i32> %cvt
}

define <4 x i64> @fptosi_8vf32_i64(<8 x float> %a) {
; SSE2-LABEL: fptosi_8vf32_i64:
; SSE2:       # BB#0:
; SSE2-NEXT:    cvttss2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm2
; SSE2-NEXT:    movaps %xmm0, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[1,1,2,3]
; SSE2-NEXT:    cvttss2si %xmm1, %rax
; SSE2-NEXT:    movd %rax, %xmm1
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm1[0]
; SSE2-NEXT:    movaps %xmm0, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,1,2,3]
; SSE2-NEXT:    cvttss2si %xmm1, %rax
; SSE2-NEXT:    movd %rax, %xmm3
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    cvttss2si %xmm0, %rax
; SSE2-NEXT:    movd %rax, %xmm1
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
; SSE2-NEXT:    movdqa %xmm2, %xmm0
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_8vf32_i64:
; AVX:       # BB#0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm1 = xmm0[3,1,2,3]
; AVX-NEXT:    vcvttss2si %xmm1, %rax
; AVX-NEXT:    vmovq %rax, %xmm1
; AVX-NEXT:    vpermilpd {{.*#+}} xmm2 = xmm0[1,0]
; AVX-NEXT:    vcvttss2si %xmm2, %rax
; AVX-NEXT:    vmovq %rax, %xmm2
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm1 = xmm2[0],xmm1[0]
; AVX-NEXT:    vcvttss2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm2
; AVX-NEXT:    vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX-NEXT:    vcvttss2si %xmm0, %rax
; AVX-NEXT:    vmovq %rax, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm2[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-NEXT:    retq
  %shuf = shufflevector <8 x float> %a, <8 x float> %a, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %cvt = fptosi <4 x float> %shuf to <4 x i64>
  ret <4 x i64> %cvt
}

;
; Float to Unsigned Integer
;

define <4 x i32> @fptoui_4vf32(<4 x float> %a) {
; SSE2-LABEL: fptoui_4vf32:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps %xmm0, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm1 = xmm1[3,1,2,3]
; SSE2-NEXT:    cvttss2si %xmm1, %rax
; SSE2-NEXT:    movd %eax, %xmm1
; SSE2-NEXT:    movaps %xmm0, %xmm2
; SSE2-NEXT:    shufps {{.*#+}} xmm2 = xmm2[1,1,2,3]
; SSE2-NEXT:    cvttss2si %xmm2, %rax
; SSE2-NEXT:    movd %eax, %xmm2
; SSE2-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
; SSE2-NEXT:    cvttss2si %xmm0, %rax
; SSE2-NEXT:    movd %eax, %xmm1
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    cvttss2si %xmm0, %rax
; SSE2-NEXT:    movd %eax, %xmm0
; SSE2-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],xmm0[0],xmm1[1],xmm0[1]
; SSE2-NEXT:    punpckldq {{.*#+}} xmm1 = xmm1[0],xmm2[0],xmm1[1],xmm2[1]
; SSE2-NEXT:    movdqa %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_4vf32:
; AVX:       # BB#0:
; AVX-NEXT:    vmovshdup {{.*#+}} xmm1 = xmm0[1,1,3,3]
; AVX-NEXT:    vcvttss2si %xmm1, %rax
; AVX-NEXT:    vcvttss2si %xmm0, %rcx
; AVX-NEXT:    vmovd %ecx, %xmm1
; AVX-NEXT:    vpinsrd $1, %eax, %xmm1, %xmm1
; AVX-NEXT:    vpermilpd {{.*#+}} xmm2 = xmm0[1,0]
; AVX-NEXT:    vcvttss2si %xmm2, %rax
; AVX-NEXT:    vpinsrd $2, %eax, %xmm1, %xmm1
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX-NEXT:    vcvttss2si %xmm0, %rax
; AVX-NEXT:    vpinsrd $3, %eax, %xmm1, %xmm0
; AVX-NEXT:    retq
  %cvt = fptoui <4 x float> %a to <4 x i32>
  ret <4 x i32> %cvt
}

define <2 x i64> @fptoui_4vf32_i64(<4 x float> %a) {
; SSE2-LABEL: fptoui_4vf32_i64:
; SSE2:       # BB#0:
; SSE2-NEXT:    movss {{.*#+}} xmm2 = mem[0],zero,zero,zero
; SSE2-NEXT:    movaps %xmm0, %xmm1
; SSE2-NEXT:    subss %xmm2, %xmm1
; SSE2-NEXT:    cvttss2si %xmm1, %rax
; SSE2-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; SSE2-NEXT:    xorq %rcx, %rax
; SSE2-NEXT:    cvttss2si %xmm0, %rdx
; SSE2-NEXT:    ucomiss %xmm2, %xmm0
; SSE2-NEXT:    cmovaeq %rax, %rdx
; SSE2-NEXT:    movd %rdx, %xmm1
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[1,1,2,3]
; SSE2-NEXT:    movaps %xmm0, %xmm3
; SSE2-NEXT:    subss %xmm2, %xmm3
; SSE2-NEXT:    cvttss2si %xmm3, %rax
; SSE2-NEXT:    xorq %rcx, %rax
; SSE2-NEXT:    cvttss2si %xmm0, %rcx
; SSE2-NEXT:    ucomiss %xmm2, %xmm0
; SSE2-NEXT:    cmovaeq %rax, %rcx
; SSE2-NEXT:    movd %rcx, %xmm0
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm0[0]
; SSE2-NEXT:    movdqa %xmm1, %xmm0
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_4vf32_i64:
; AVX:       # BB#0:
; AVX-NEXT:    vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero
; AVX-NEXT:    vsubss %xmm1, %xmm0, %xmm2
; AVX-NEXT:    vcvttss2si %xmm2, %rax
; AVX-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttss2si %xmm0, %rdx
; AVX-NEXT:    vucomiss %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rdx
; AVX-NEXT:    vmovq %rdx, %xmm2
; AVX-NEXT:    vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX-NEXT:    vsubss %xmm1, %xmm0, %xmm3
; AVX-NEXT:    vcvttss2si %xmm3, %rax
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttss2si %xmm0, %rcx
; AVX-NEXT:    vucomiss %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rcx
; AVX-NEXT:    vmovq %rcx, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm2[0],xmm0[0]
; AVX-NEXT:    retq
  %shuf = shufflevector <4 x float> %a, <4 x float> %a, <2 x i32> <i32 0, i32 1>
  %cvt = fptoui <2 x float> %shuf to <2 x i64>
  ret <2 x i64> %cvt
}

define <8 x i32> @fptoui_8vf32(<8 x float> %a) {
; SSE2-LABEL: fptoui_8vf32:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps %xmm0, %xmm2
; SSE2-NEXT:    shufps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; SSE2-NEXT:    cvttss2si %xmm0, %rax
; SSE2-NEXT:    movd %eax, %xmm0
; SSE2-NEXT:    movaps %xmm2, %xmm3
; SSE2-NEXT:    shufps {{.*#+}} xmm3 = xmm3[1,1,2,3]
; SSE2-NEXT:    cvttss2si %xmm3, %rax
; SSE2-NEXT:    movd %eax, %xmm3
; SSE2-NEXT:    punpckldq {{.*#+}} xmm3 = xmm3[0],xmm0[0],xmm3[1],xmm0[1]
; SSE2-NEXT:    cvttss2si %xmm2, %rax
; SSE2-NEXT:    movd %eax, %xmm0
; SSE2-NEXT:    shufpd {{.*#+}} xmm2 = xmm2[1,0]
; SSE2-NEXT:    cvttss2si %xmm2, %rax
; SSE2-NEXT:    movd %eax, %xmm2
; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm2[0],xmm0[1],xmm2[1]
; SSE2-NEXT:    punpckldq {{.*#+}} xmm0 = xmm0[0],xmm3[0],xmm0[1],xmm3[1]
; SSE2-NEXT:    movaps %xmm1, %xmm2
; SSE2-NEXT:    shufps {{.*#+}} xmm2 = xmm2[3,1,2,3]
; SSE2-NEXT:    cvttss2si %xmm2, %rax
; SSE2-NEXT:    movd %eax, %xmm2
; SSE2-NEXT:    movaps %xmm1, %xmm3
; SSE2-NEXT:    shufps {{.*#+}} xmm3 = xmm3[1,1,2,3]
; SSE2-NEXT:    cvttss2si %xmm3, %rax
; SSE2-NEXT:    movd %eax, %xmm3
; SSE2-NEXT:    punpckldq {{.*#+}} xmm3 = xmm3[0],xmm2[0],xmm3[1],xmm2[1]
; SSE2-NEXT:    cvttss2si %xmm1, %rax
; SSE2-NEXT:    movd %eax, %xmm2
; SSE2-NEXT:    shufpd {{.*#+}} xmm1 = xmm1[1,0]
; SSE2-NEXT:    cvttss2si %xmm1, %rax
; SSE2-NEXT:    movd %eax, %xmm1
; SSE2-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm1[0],xmm2[1],xmm1[1]
; SSE2-NEXT:    punpckldq {{.*#+}} xmm2 = xmm2[0],xmm3[0],xmm2[1],xmm3[1]
; SSE2-NEXT:    movdqa %xmm2, %xmm1
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_8vf32:
; AVX:       # BB#0:
; AVX-NEXT:    vextractf128 $1, %ymm0, %xmm1
; AVX-NEXT:    vmovshdup {{.*#+}} xmm2 = xmm1[1,1,3,3]
; AVX-NEXT:    vcvttss2si %xmm2, %rax
; AVX-NEXT:    vcvttss2si %xmm1, %rcx
; AVX-NEXT:    vmovd %ecx, %xmm2
; AVX-NEXT:    vpinsrd $1, %eax, %xmm2, %xmm2
; AVX-NEXT:    vpermilpd {{.*#+}} xmm3 = xmm1[1,0]
; AVX-NEXT:    vcvttss2si %xmm3, %rax
; AVX-NEXT:    vpinsrd $2, %eax, %xmm2, %xmm2
; AVX-NEXT:    vpermilps {{.*#+}} xmm1 = xmm1[3,1,2,3]
; AVX-NEXT:    vcvttss2si %xmm1, %rax
; AVX-NEXT:    vpinsrd $3, %eax, %xmm2, %xmm1
; AVX-NEXT:    vmovshdup {{.*#+}} xmm2 = xmm0[1,1,3,3]
; AVX-NEXT:    vcvttss2si %xmm2, %rax
; AVX-NEXT:    vcvttss2si %xmm0, %rcx
; AVX-NEXT:    vmovd %ecx, %xmm2
; AVX-NEXT:    vpinsrd $1, %eax, %xmm2, %xmm2
; AVX-NEXT:    vpermilpd {{.*#+}} xmm3 = xmm0[1,0]
; AVX-NEXT:    vcvttss2si %xmm3, %rax
; AVX-NEXT:    vpinsrd $2, %eax, %xmm2, %xmm2
; AVX-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[3,1,2,3]
; AVX-NEXT:    vcvttss2si %xmm0, %rax
; AVX-NEXT:    vpinsrd $3, %eax, %xmm2, %xmm0
; AVX-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
; AVX-NEXT:    retq
  %cvt = fptoui <8 x float> %a to <8 x i32>
  ret <8 x i32> %cvt
}

define <4 x i64> @fptoui_8vf32_i64(<8 x float> %a) {
; SSE2-LABEL: fptoui_8vf32_i64:
; SSE2:       # BB#0:
; SSE2-NEXT:    movss {{.*#+}} xmm1 = mem[0],zero,zero,zero
; SSE2-NEXT:    movaps %xmm0, %xmm2
; SSE2-NEXT:    subss %xmm1, %xmm2
; SSE2-NEXT:    cvttss2si %xmm2, %rcx
; SSE2-NEXT:    movabsq $-9223372036854775808, %rax # imm = 0x8000000000000000
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttss2si %xmm0, %rdx
; SSE2-NEXT:    ucomiss %xmm1, %xmm0
; SSE2-NEXT:    cmovaeq %rcx, %rdx
; SSE2-NEXT:    movd %rdx, %xmm2
; SSE2-NEXT:    movaps %xmm0, %xmm3
; SSE2-NEXT:    shufps {{.*#+}} xmm3 = xmm3[1,1,2,3]
; SSE2-NEXT:    movaps %xmm3, %xmm4
; SSE2-NEXT:    subss %xmm1, %xmm4
; SSE2-NEXT:    cvttss2si %xmm4, %rcx
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttss2si %xmm3, %rdx
; SSE2-NEXT:    ucomiss %xmm1, %xmm3
; SSE2-NEXT:    cmovaeq %rcx, %rdx
; SSE2-NEXT:    movd %rdx, %xmm3
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm2 = xmm2[0],xmm3[0]
; SSE2-NEXT:    movaps %xmm0, %xmm3
; SSE2-NEXT:    shufps {{.*#+}} xmm3 = xmm3[3,1,2,3]
; SSE2-NEXT:    movaps %xmm3, %xmm4
; SSE2-NEXT:    subss %xmm1, %xmm4
; SSE2-NEXT:    cvttss2si %xmm4, %rcx
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttss2si %xmm3, %rdx
; SSE2-NEXT:    ucomiss %xmm1, %xmm3
; SSE2-NEXT:    cmovaeq %rcx, %rdx
; SSE2-NEXT:    movd %rdx, %xmm3
; SSE2-NEXT:    shufpd {{.*#+}} xmm0 = xmm0[1,0]
; SSE2-NEXT:    movapd %xmm0, %xmm4
; SSE2-NEXT:    subss %xmm1, %xmm4
; SSE2-NEXT:    cvttss2si %xmm4, %rcx
; SSE2-NEXT:    xorq %rax, %rcx
; SSE2-NEXT:    cvttss2si %xmm0, %rax
; SSE2-NEXT:    ucomiss %xmm1, %xmm0
; SSE2-NEXT:    cmovaeq %rcx, %rax
; SSE2-NEXT:    movd %rax, %xmm1
; SSE2-NEXT:    punpcklqdq {{.*#+}} xmm1 = xmm1[0],xmm3[0]
; SSE2-NEXT:    movdqa %xmm2, %xmm0
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_8vf32_i64:
; AVX:       # BB#0:
; AVX-NEXT:    vpermilps {{.*#+}} xmm2 = xmm0[3,1,2,3]
; AVX-NEXT:    vmovss {{.*#+}} xmm1 = mem[0],zero,zero,zero
; AVX-NEXT:    vsubss %xmm1, %xmm2, %xmm3
; AVX-NEXT:    vcvttss2si %xmm3, %rax
; AVX-NEXT:    movabsq $-9223372036854775808, %rcx # imm = 0x8000000000000000
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttss2si %xmm2, %rdx
; AVX-NEXT:    vucomiss %xmm1, %xmm2
; AVX-NEXT:    cmovaeq %rax, %rdx
; AVX-NEXT:    vmovq %rdx, %xmm2
; AVX-NEXT:    vpermilpd {{.*#+}} xmm3 = xmm0[1,0]
; AVX-NEXT:    vsubss %xmm1, %xmm3, %xmm4
; AVX-NEXT:    vcvttss2si %xmm4, %rax
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttss2si %xmm3, %rdx
; AVX-NEXT:    vucomiss %xmm1, %xmm3
; AVX-NEXT:    cmovaeq %rax, %rdx
; AVX-NEXT:    vmovq %rdx, %xmm3
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm2 = xmm3[0],xmm2[0]
; AVX-NEXT:    vsubss %xmm1, %xmm0, %xmm3
; AVX-NEXT:    vcvttss2si %xmm3, %rax
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttss2si %xmm0, %rdx
; AVX-NEXT:    vucomiss %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rdx
; AVX-NEXT:    vmovq %rdx, %xmm3
; AVX-NEXT:    vmovshdup {{.*#+}} xmm0 = xmm0[1,1,3,3]
; AVX-NEXT:    vsubss %xmm1, %xmm0, %xmm4
; AVX-NEXT:    vcvttss2si %xmm4, %rax
; AVX-NEXT:    xorq %rcx, %rax
; AVX-NEXT:    vcvttss2si %xmm0, %rcx
; AVX-NEXT:    vucomiss %xmm1, %xmm0
; AVX-NEXT:    cmovaeq %rax, %rcx
; AVX-NEXT:    vmovq %rcx, %xmm0
; AVX-NEXT:    vpunpcklqdq {{.*#+}} xmm0 = xmm3[0],xmm0[0]
; AVX-NEXT:    vinsertf128 $1, %xmm2, %ymm0, %ymm0
; AVX-NEXT:    retq
  %shuf = shufflevector <8 x float> %a, <8 x float> %a, <4 x i32> <i32 0, i32 1, i32 2, i32 3>
  %cvt = fptoui <4 x float> %shuf to <4 x i64>
  ret <4 x i64> %cvt
}

;
; Constant Folding
;

define <2 x i64> @fptosi_2vf64c() {
; SSE2-LABEL: fptosi_2vf64c:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [1,18446744073709551615]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_2vf64c:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} xmm0 = [1,18446744073709551615]
; AVX-NEXT:    retq
  %cvt = fptosi <2 x double> <double 1.0, double -1.0> to <2 x i64>
  ret <2 x i64> %cvt
}

define <4 x i32> @fptosi_2vf64c_i32() {
; SSE2-LABEL: fptosi_2vf64c_i32:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = <4294967295,1,u,u>
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_2vf64c_i32:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} xmm0 = <4294967295,1,u,u>
; AVX-NEXT:    retq
  %cvt = fptosi <2 x double> <double -1.0, double 1.0> to <2 x i32>
  %ext = shufflevector <2 x i32> %cvt, <2 x i32> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  ret <4 x i32> %ext
}

define <4 x i64> @fptosi_4vf64c() {
; SSE2-LABEL: fptosi_4vf64c:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [1,18446744073709551615]
; SSE2-NEXT:    movaps {{.*#+}} xmm1 = [2,18446744073709551613]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_4vf64c:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [1,18446744073709551615,2,18446744073709551613]
; AVX-NEXT:    retq
  %cvt = fptosi <4 x double> <double 1.0, double -1.0, double 2.0, double -3.0> to <4 x i64>
  ret <4 x i64> %cvt
}

define <4 x i32> @fptosi_4vf64c_i32() {
; SSE2-LABEL: fptosi_4vf64c_i32:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [4294967295,1,4294967294,3]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_4vf64c_i32:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} xmm0 = [4294967295,1,4294967294,3]
; AVX-NEXT:    retq
  %cvt = fptosi <4 x double> <double -1.0, double 1.0, double -2.0, double 3.0> to <4 x i32>
  ret <4 x i32> %cvt
}

define <2 x i64> @fptoui_2vf64c() {
; SSE2-LABEL: fptoui_2vf64c:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [2,4]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_2vf64c:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} xmm0 = [2,4]
; AVX-NEXT:    retq
  %cvt = fptoui <2 x double> <double 2.0, double 4.0> to <2 x i64>
  ret <2 x i64> %cvt
}

define <4 x i32> @fptoui_2vf64c_i32(<2 x double> %a) {
; SSE2-LABEL: fptoui_2vf64c_i32:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = <2,4,u,u>
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_2vf64c_i32:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} xmm0 = <2,4,u,u>
; AVX-NEXT:    retq
  %cvt = fptoui <2 x double> <double 2.0, double 4.0> to <2 x i32>
  %ext = shufflevector <2 x i32> %cvt, <2 x i32> undef, <4 x i32> <i32 0, i32 1, i32 undef, i32 undef>
  ret <4 x i32> %ext
}

define <4 x i64> @fptoui_4vf64c(<4 x double> %a) {
; SSE2-LABEL: fptoui_4vf64c:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [2,4]
; SSE2-NEXT:    movaps {{.*#+}} xmm1 = [6,8]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_4vf64c:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [2,4,6,8]
; AVX-NEXT:    retq
  %cvt = fptoui <4 x double> <double 2.0, double 4.0, double 6.0, double 8.0> to <4 x i64>
  ret <4 x i64> %cvt
}

define <4 x i32> @fptoui_4vf64c_i32(<4 x double> %a) {
; SSE2-LABEL: fptoui_4vf64c_i32:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [2,4,6,8]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_4vf64c_i32:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} xmm0 = [2,4,6,8]
; AVX-NEXT:    retq
  %cvt = fptoui <4 x double> <double 2.0, double 4.0, double 6.0, double 8.0> to <4 x i32>
  ret <4 x i32> %cvt
}

define <4 x i32> @fptosi_4vf32c() {
; SSE2-LABEL: fptosi_4vf32c:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [1,4294967295,2,3]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_4vf32c:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} xmm0 = [1,4294967295,2,3]
; AVX-NEXT:    retq
  %cvt = fptosi <4 x float> <float 1.0, float -1.0, float 2.0, float 3.0> to <4 x i32>
  ret <4 x i32> %cvt
}

define <4 x i64> @fptosi_4vf32c_i64() {
; SSE2-LABEL: fptosi_4vf32c_i64:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [1,18446744073709551615]
; SSE2-NEXT:    movaps {{.*#+}} xmm1 = [2,3]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_4vf32c_i64:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [1,18446744073709551615,2,3]
; AVX-NEXT:    retq
  %cvt = fptosi <4 x float> <float 1.0, float -1.0, float 2.0, float 3.0> to <4 x i64>
  ret <4 x i64> %cvt
}

define <8 x i32> @fptosi_8vf32c(<8 x float> %a) {
; SSE2-LABEL: fptosi_8vf32c:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [1,4294967295,2,3]
; SSE2-NEXT:    movaps {{.*#+}} xmm1 = [6,4294967288,2,4294967295]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptosi_8vf32c:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [1,4294967295,2,3,6,4294967288,2,4294967295]
; AVX-NEXT:    retq
  %cvt = fptosi <8 x float> <float 1.0, float -1.0, float 2.0, float 3.0, float 6.0, float -8.0, float 2.0, float -1.0> to <8 x i32>
  ret <8 x i32> %cvt
}

define <4 x i32> @fptoui_4vf32c(<4 x float> %a) {
; SSE2-LABEL: fptoui_4vf32c:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [1,2,4,6]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_4vf32c:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} xmm0 = [1,2,4,6]
; AVX-NEXT:    retq
  %cvt = fptoui <4 x float> <float 1.0, float 2.0, float 4.0, float 6.0> to <4 x i32>
  ret <4 x i32> %cvt
}

define <4 x i64> @fptoui_4vf32c_i64() {
; SSE2-LABEL: fptoui_4vf32c_i64:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [1,2]
; SSE2-NEXT:    movaps {{.*#+}} xmm1 = [4,8]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_4vf32c_i64:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [1,2,4,8]
; AVX-NEXT:    retq
  %cvt = fptoui <4 x float> <float 1.0, float 2.0, float 4.0, float 8.0> to <4 x i64>
  ret <4 x i64> %cvt
}

define <8 x i32> @fptoui_8vf32c(<8 x float> %a) {
; SSE2-LABEL: fptoui_8vf32c:
; SSE2:       # BB#0:
; SSE2-NEXT:    movaps {{.*#+}} xmm0 = [1,2,4,6]
; SSE2-NEXT:    movaps {{.*#+}} xmm1 = [8,6,4,1]
; SSE2-NEXT:    retq
;
; AVX-LABEL: fptoui_8vf32c:
; AVX:       # BB#0:
; AVX-NEXT:    vmovaps {{.*#+}} ymm0 = [1,2,4,6,8,6,4,1]
; AVX-NEXT:    retq
  %cvt = fptoui <8 x float> <float 1.0, float 2.0, float 4.0, float 6.0, float 8.0, float 6.0, float 4.0, float 1.0> to <8 x i32>
  ret <8 x i32> %cvt
}