summaryrefslogtreecommitdiffstats
path: root/src/compiler/nir/nir_opt_algebraic.py
blob: bdf70787937c19cb75cc90ed3e737e5ea0e9af87 (plain)
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
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
#
# Copyright (C) 2014 Intel Corporation
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the "Software"),
# to deal in the Software without restriction, including without limitation
# the rights to use, copy, modify, merge, publish, distribute, sublicense,
# and/or sell copies of the Software, and to permit persons to whom the
# Software is furnished to do so, subject to the following conditions:
#
# The above copyright notice and this permission notice (including the next
# paragraph) shall be included in all copies or substantial portions of the
# Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
# THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
# IN THE SOFTWARE.
#
# Authors:
#    Jason Ekstrand (jason@jlekstrand.net)

from __future__ import print_function

from collections import OrderedDict
import nir_algebraic
from nir_opcodes import type_sizes
import itertools

# Convenience variables
a = 'a'
b = 'b'
c = 'c'
d = 'd'

# Written in the form (<search>, <replace>) where <search> is an expression
# and <replace> is either an expression or a value.  An expression is
# defined as a tuple of the form ([~]<op>, <src0>, <src1>, <src2>, <src3>)
# where each source is either an expression or a value.  A value can be
# either a numeric constant or a string representing a variable name.
#
# If the opcode in a search expression is prefixed by a '~' character, this
# indicates that the operation is inexact.  Such operations will only get
# applied to SSA values that do not have the exact bit set.  This should be
# used by by any optimizations that are not bit-for-bit exact.  It should not,
# however, be used for backend-requested lowering operations as those need to
# happen regardless of precision.
#
# Variable names are specified as "[#]name[@type][(cond)]" where "#" inicates
# that the given variable will only match constants and the type indicates that
# the given variable will only match values from ALU instructions with the
# given output type, and (cond) specifies an additional condition function
# (see nir_search_helpers.h).
#
# For constants, you have to be careful to make sure that it is the right
# type because python is unaware of the source and destination types of the
# opcodes.
#
# All expression types can have a bit-size specified.  For opcodes, this
# looks like "op@32", for variables it is "a@32" or "a@uint32" to specify a
# type and size.  In the search half of the expression this indicates that it
# should only match that particular bit-size.  In the replace half of the
# expression this indicates that the constructed value should have that
# bit-size.

optimizations = [

   (('imul', a, '#b@32(is_pos_power_of_two)'), ('ishl', a, ('find_lsb', b))),
   (('imul', a, '#b@32(is_neg_power_of_two)'), ('ineg', ('ishl', a, ('find_lsb', ('iabs', b))))),
   (('unpack_64_2x32_split_x', ('imul_2x32_64(is_used_once)', a, b)), ('imul', a, b)),
   (('unpack_64_2x32_split_x', ('umul_2x32_64(is_used_once)', a, b)), ('imul', a, b)),
   (('imul_2x32_64', a, b), ('pack_64_2x32_split', ('imul', a, b), ('imul_high', a, b)), 'options->lower_mul_2x32_64'),
   (('umul_2x32_64', a, b), ('pack_64_2x32_split', ('imul', a, b), ('umul_high', a, b)), 'options->lower_mul_2x32_64'),
   (('udiv', a, 1), a),
   (('idiv', a, 1), a),
   (('umod', a, 1), 0),
   (('imod', a, 1), 0),
   (('udiv', a, '#b@32(is_pos_power_of_two)'), ('ushr', a, ('find_lsb', b))),
   (('idiv', a, '#b@32(is_pos_power_of_two)'), ('imul', ('isign', a), ('ushr', ('iabs', a), ('find_lsb', b))), 'options->lower_idiv'),
   (('idiv', a, '#b@32(is_neg_power_of_two)'), ('ineg', ('imul', ('isign', a), ('ushr', ('iabs', a), ('find_lsb', ('iabs', b))))), 'options->lower_idiv'),
   (('umod', a, '#b(is_pos_power_of_two)'),    ('iand', a, ('isub', b, 1))),

   (('fneg', ('fneg', a)), a),
   (('ineg', ('ineg', a)), a),
   (('fabs', ('fabs', a)), ('fabs', a)),
   (('fabs', ('fneg', a)), ('fabs', a)),
   (('fabs', ('u2f', a)), ('u2f', a)),
   (('iabs', ('iabs', a)), ('iabs', a)),
   (('iabs', ('ineg', a)), ('iabs', a)),
   (('f2b', ('fneg', a)), ('f2b', a)),
   (('i2b', ('ineg', a)), ('i2b', a)),
   (('~fadd', a, 0.0), a),
   (('iadd', a, 0), a),
   (('usadd_4x8', a, 0), a),
   (('usadd_4x8', a, ~0), ~0),
   (('~fadd', ('fmul', a, b), ('fmul', a, c)), ('fmul', a, ('fadd', b, c))),
   (('iadd', ('imul', a, b), ('imul', a, c)), ('imul', a, ('iadd', b, c))),
   (('~fadd', ('fneg', a), a), 0.0),
   (('iadd', ('ineg', a), a), 0),
   (('iadd', ('ineg', a), ('iadd', a, b)), b),
   (('iadd', a, ('iadd', ('ineg', a), b)), b),
   (('~fadd', ('fneg', a), ('fadd', a, b)), b),
   (('~fadd', a, ('fadd', ('fneg', a), b)), b),
   (('~fmul', a, 0.0), 0.0),
   (('imul', a, 0), 0),
   (('umul_unorm_4x8', a, 0), 0),
   (('umul_unorm_4x8', a, ~0), a),
   (('fmul', a, 1.0), a),
   (('imul', a, 1), a),
   (('fmul', a, -1.0), ('fneg', a)),
   (('imul', a, -1), ('ineg', a)),
   # If a < 0: fsign(a)*a*a => -1*a*a => -a*a => abs(a)*a
   # If a > 0: fsign(a)*a*a => 1*a*a => a*a => abs(a)*a
   # If a == 0: fsign(a)*a*a => 0*0*0 => abs(0)*0
   (('fmul', ('fsign', a), ('fmul', a, a)), ('fmul', ('fabs', a), a)),
   (('fmul', ('fmul', ('fsign', a), a), a), ('fmul', ('fabs', a), a)),
   (('~ffma', 0.0, a, b), b),
   (('~ffma', a, 0.0, b), b),
   (('~ffma', a, b, 0.0), ('fmul', a, b)),
   (('ffma', a, 1.0, b), ('fadd', a, b)),
   (('ffma', 1.0, a, b), ('fadd', a, b)),
   (('~flrp', a, b, 0.0), a),
   (('~flrp', a, b, 1.0), b),
   (('~flrp', a, a, b), a),
   (('~flrp', 0.0, a, b), ('fmul', a, b)),
   (('~flrp', a, b, ('b2f', 'c@1')), ('bcsel', c, b, a), 'options->lower_flrp32'),
   (('~flrp', a, 0.0, c), ('fadd', ('fmul', ('fneg', a), c), a)),
   (('flrp@16', a, b, c), ('fadd', ('fmul', c, ('fsub', b, a)), a), 'options->lower_flrp16'),
   (('flrp@32', a, b, c), ('fadd', ('fmul', c, ('fsub', b, a)), a), 'options->lower_flrp32'),
   (('flrp@64', a, b, c), ('fadd', ('fmul', c, ('fsub', b, a)), a), 'options->lower_flrp64'),
   (('ffloor', a), ('fsub', a, ('ffract', a)), 'options->lower_ffloor'),
   (('fadd', a, ('fneg', ('ffract', a))), ('ffloor', a), '!options->lower_ffloor'),
   (('ffract', a), ('fsub', a, ('ffloor', a)), 'options->lower_ffract'),
   (('fceil', a), ('fneg', ('ffloor', ('fneg', a))), 'options->lower_fceil'),
   (('~fadd', ('fmul', a, ('fadd', 1.0, ('fneg', ('b2f', 'c@1')))), ('fmul', b, ('b2f', c))), ('bcsel', c, b, a), 'options->lower_flrp32'),
   (('~fadd@32', ('fmul', a, ('fadd', 1.0, ('fneg',         c ))), ('fmul', b,         c )), ('flrp', a, b, c), '!options->lower_flrp32'),
   (('~fadd@64', ('fmul', a, ('fadd', 1.0, ('fneg',         c ))), ('fmul', b,         c )), ('flrp', a, b, c), '!options->lower_flrp64'),
   (('~fadd', a, ('fmul', ('b2f', 'c@1'), ('fadd', b, ('fneg', a)))), ('bcsel', c, b, a), 'options->lower_flrp32'),
   (('~fadd@32', a, ('fmul',         c , ('fadd', b, ('fneg', a)))), ('flrp', a, b, c), '!options->lower_flrp32'),
   (('~fadd@64', a, ('fmul',         c , ('fadd', b, ('fneg', a)))), ('flrp', a, b, c), '!options->lower_flrp64'),
   (('ffma', a, b, c), ('fadd', ('fmul', a, b), c), 'options->lower_ffma'),
   (('~fadd', ('fmul', a, b), c), ('ffma', a, b, c), 'options->fuse_ffma'),

   (('fdot4', ('vec4', a, b,   c,   1.0), d), ('fdph',  ('vec3', a, b, c), d)),
   (('fdot4', ('vec4', a, 0.0, 0.0, 0.0), b), ('fmul', a, b)),
   (('fdot4', ('vec4', a, b,   0.0, 0.0), c), ('fdot2', ('vec2', a, b), c)),
   (('fdot4', ('vec4', a, b,   c,   0.0), d), ('fdot3', ('vec3', a, b, c), d)),

   (('fdot3', ('vec3', a, 0.0, 0.0), b), ('fmul', a, b)),
   (('fdot3', ('vec3', a, b,   0.0), c), ('fdot2', ('vec2', a, b), c)),

   # (a * #b + #c) << #d
   # ((a * #b) << #d) + (#c << #d)
   # (a * (#b << #d)) + (#c << #d)
   (('ishl', ('iadd', ('imul', a, '#b'), '#c'), '#d'),
    ('iadd', ('imul', a, ('ishl', b, d)), ('ishl', c, d))),

   # (a * #b) << #c
   # a * (#b << #c)
   (('ishl', ('imul', a, '#b'), '#c'), ('imul', a, ('ishl', b, c))),

   # Comparison simplifications
   (('~inot', ('flt', a, b)), ('fge', a, b)),
   (('~inot', ('fge', a, b)), ('flt', a, b)),
   (('~inot', ('feq', a, b)), ('fne', a, b)),
   (('~inot', ('fne', a, b)), ('feq', a, b)),
   (('inot', ('ilt', a, b)), ('ige', a, b)),
   (('inot', ('ult', a, b)), ('uge', a, b)),
   (('inot', ('ige', a, b)), ('ilt', a, b)),
   (('inot', ('uge', a, b)), ('ult', a, b)),
   (('inot', ('ieq', a, b)), ('ine', a, b)),
   (('inot', ('ine', a, b)), ('ieq', a, b)),

   # 0.0 >= b2f(a)
   # b2f(a) <= 0.0
   # b2f(a) == 0.0 because b2f(a) can only be 0 or 1
   # inot(a)
   (('fge', 0.0, ('b2f', 'a@1')), ('inot', a)),

   (('fge', ('fneg', ('b2f', 'a@1')), 0.0), ('inot', a)),

   (('fne', ('fadd', ('b2f', 'a@1'), ('b2f', 'b@1')), 0.0), ('ior', a, b)),
   (('fne', ('fmax', ('b2f', 'a@1'), ('b2f', 'b@1')), 0.0), ('ior', a, b)),
   (('fne', ('bcsel', a, 1.0, ('b2f', 'b@1'))   , 0.0), ('ior', a, b)),
   (('fne', ('b2f', 'a@1'), ('fneg', ('b2f', 'b@1'))),      ('ior', a, b)),
   (('fne', ('fmul', ('b2f', 'a@1'), ('b2f', 'b@1')), 0.0), ('iand', a, b)),
   (('fne', ('fmin', ('b2f', 'a@1'), ('b2f', 'b@1')), 0.0), ('iand', a, b)),
   (('fne', ('bcsel', a, ('b2f', 'b@1'), 0.0)   , 0.0), ('iand', a, b)),
   (('fne', ('fadd', ('b2f', 'a@1'), ('fneg', ('b2f', 'b@1'))), 0.0), ('ixor', a, b)),
   (('fne',          ('b2f', 'a@1') ,          ('b2f', 'b@1') ),      ('ixor', a, b)),
   (('fne', ('fneg', ('b2f', 'a@1')), ('fneg', ('b2f', 'b@1'))),      ('ixor', a, b)),
   (('feq', ('fadd', ('b2f', 'a@1'), ('b2f', 'b@1')), 0.0), ('inot', ('ior', a, b))),
   (('feq', ('fmax', ('b2f', 'a@1'), ('b2f', 'b@1')), 0.0), ('inot', ('ior', a, b))),
   (('feq', ('bcsel', a, 1.0, ('b2f', 'b@1'))   , 0.0), ('inot', ('ior', a, b))),
   (('feq', ('b2f', 'a@1'), ('fneg', ('b2f', 'b@1'))),      ('inot', ('ior', a, b))),
   (('feq', ('fmul', ('b2f', 'a@1'), ('b2f', 'b@1')), 0.0), ('inot', ('iand', a, b))),
   (('feq', ('fmin', ('b2f', 'a@1'), ('b2f', 'b@1')), 0.0), ('inot', ('iand', a, b))),
   (('feq', ('bcsel', a, ('b2f', 'b@1'), 0.0)   , 0.0), ('inot', ('iand', a, b))),
   (('feq', ('fadd', ('b2f', 'a@1'), ('fneg', ('b2f', 'b@1'))), 0.0), ('ieq', a, b)),
   (('feq',          ('b2f', 'a@1') ,          ('b2f', 'b@1') ),      ('ieq', a, b)),
   (('feq', ('fneg', ('b2f', 'a@1')), ('fneg', ('b2f', 'b@1'))),      ('ieq', a, b)),

   # -(b2f(a) + b2f(b)) < 0
   # 0 < b2f(a) + b2f(b)
   # 0 != b2f(a) + b2f(b)       b2f must be 0 or 1, so the sum is non-negative
   # a || b
   (('flt', ('fneg', ('fadd', ('b2f', 'a@1'), ('b2f', 'b@1'))), 0.0), ('ior', a, b)),
   (('flt', 0.0, ('fadd', ('b2f', 'a@1'), ('b2f', 'b@1'))), ('ior', a, b)),

   # -(b2f(a) + b2f(b)) >= 0
   # 0 >= b2f(a) + b2f(b)
   # 0 == b2f(a) + b2f(b)       b2f must be 0 or 1, so the sum is non-negative
   # !(a || b)
   (('fge', ('fneg', ('fadd', ('b2f', 'a@1'), ('b2f', 'b@1'))), 0.0), ('inot', ('ior', a, b))),
   (('fge', 0.0, ('fadd', ('b2f', 'a@1'), ('b2f', 'b@1'))), ('inot', ('ior', a, b))),

   # Some optimizations (below) convert things like (a < b || c < b) into
   # (min(a, c) < b).  However, this interfers with the previous optimizations
   # that try to remove comparisons with negated sums of b2f.  This just
   # breaks that apart.
   (('flt', ('fmin', c, ('fneg', ('fadd', ('b2f', 'a@1'), ('b2f', 'b@1')))), 0.0),
    ('ior', ('flt', c, 0.0), ('ior', a, b))),

   (('~flt', ('fadd', a, b), a), ('flt', b, 0.0)),
   (('~fge', ('fadd', a, b), a), ('fge', b, 0.0)),
   (('~feq', ('fadd', a, b), a), ('feq', b, 0.0)),
   (('~fne', ('fadd', a, b), a), ('fne', b, 0.0)),

   # Cannot remove the addition from ilt or ige due to overflow.
   (('ieq', ('iadd', a, b), a), ('ieq', b, 0)),
   (('ine', ('iadd', a, b), a), ('ine', b, 0)),

   # fmin(-b2f(a), b) >= 0.0
   # -b2f(a) >= 0.0 && b >= 0.0
   # -b2f(a) == 0.0 && b >= 0.0    -b2f can only be 0 or -1, never >0
   # b2f(a) == 0.0 && b >= 0.0
   # a == False && b >= 0.0
   # !a && b >= 0.0
   #
   # The fge in the second replacement is not a typo.  I leave the proof that
   # "fmin(-b2f(a), b) >= 0 <=> fmin(-b2f(a), b) == 0" as an exercise for the
   # reader.
   (('fge', ('fmin', ('fneg', ('b2f', 'a@1')), 'b@1'), 0.0), ('iand', ('inot', a), ('fge', b, 0.0))),
   (('feq', ('fmin', ('fneg', ('b2f', 'a@1')), 'b@1'), 0.0), ('iand', ('inot', a), ('fge', b, 0.0))),

   (('feq', ('b2f', 'a@1'), 0.0), ('inot', a)),
   (('fne', ('b2f', 'a@1'), 0.0), a),
   (('ieq', ('b2i', 'a@1'), 0),   ('inot', a)),
   (('ine', ('b2i', 'a@1'), 0),   a),

   (('fne', ('u2f', a), 0.0), ('ine', a, 0)),
   (('feq', ('u2f', a), 0.0), ('ieq', a, 0)),
   (('fge', ('u2f', a), 0.0), True),
   (('fge', 0.0, ('u2f', a)), ('uge', 0, a)),    # ieq instead?
   (('flt', ('u2f', a), 0.0), False),
   (('flt', 0.0, ('u2f', a)), ('ult', 0, a)),    # ine instead?
   (('fne', ('i2f', a), 0.0), ('ine', a, 0)),
   (('feq', ('i2f', a), 0.0), ('ieq', a, 0)),
   (('fge', ('i2f', a), 0.0), ('ige', a, 0)),
   (('fge', 0.0, ('i2f', a)), ('ige', 0, a)),
   (('flt', ('i2f', a), 0.0), ('ilt', a, 0)),
   (('flt', 0.0, ('i2f', a)), ('ilt', 0, a)),

   # 0.0 < fabs(a)
   # fabs(a) > 0.0
   # fabs(a) != 0.0 because fabs(a) must be >= 0
   # a != 0.0
   (('~flt', 0.0, ('fabs', a)), ('fne', a, 0.0)),

   # -fabs(a) < 0.0
   # fabs(a) > 0.0
   (('~flt', ('fneg', ('fabs', a)), 0.0), ('fne', a, 0.0)),

   # 0.0 >= fabs(a)
   # 0.0 == fabs(a)   because fabs(a) must be >= 0
   # 0.0 == a
   (('fge', 0.0, ('fabs', a)), ('feq', a, 0.0)),

   # -fabs(a) >= 0.0
   # 0.0 >= fabs(a)
   (('fge', ('fneg', ('fabs', a)), 0.0), ('feq', a, 0.0)),

   (('fmax',                        ('b2f(is_used_once)', 'a@1'),           ('b2f', 'b@1')),           ('b2f', ('ior', a, b))),
   (('fmax', ('fneg(is_used_once)', ('b2f(is_used_once)', 'a@1')), ('fneg', ('b2f', 'b@1'))), ('fneg', ('b2f', ('ior', a, b)))),
   (('fmin',                        ('b2f(is_used_once)', 'a@1'),           ('b2f', 'b@1')),           ('b2f', ('iand', a, b))),
   (('fmin', ('fneg(is_used_once)', ('b2f(is_used_once)', 'a@1')), ('fneg', ('b2f', 'b@1'))), ('fneg', ('b2f', ('iand', a, b)))),

   # fmin(b2f(a), b)
   # bcsel(a, fmin(b2f(a), b), fmin(b2f(a), b))
   # bcsel(a, fmin(b2f(True), b), fmin(b2f(False), b))
   # bcsel(a, fmin(1.0, b), fmin(0.0, b))
   #
   # Since b is a constant, constant folding will eliminate the fmin and the
   # fmax.  If b is > 1.0, the bcsel will be replaced with a b2f.
   (('fmin', ('b2f', 'a@1'), '#b'), ('bcsel', a, ('fmin', b, 1.0), ('fmin', b, 0.0))),

   (('flt', ('fadd(is_used_once)', a, ('fneg', b)), 0.0), ('flt', a, b)),

   (('fge', ('fneg', ('fabs', a)), 0.0), ('feq', a, 0.0)),
   (('~bcsel', ('flt', b, a), b, a), ('fmin', a, b)),
   (('~bcsel', ('flt', a, b), b, a), ('fmax', a, b)),
   (('~bcsel', ('fge', a, b), b, a), ('fmin', a, b)),
   (('~bcsel', ('fge', b, a), b, a), ('fmax', a, b)),
   (('bcsel', ('i2b', a), b, c), ('bcsel', ('ine', a, 0), b, c)),
   (('bcsel', ('inot', a), b, c), ('bcsel', a, c, b)),
   (('bcsel', a, ('bcsel', a, b, c), d), ('bcsel', a, b, d)),
   (('bcsel', a, b, ('bcsel', a, c, d)), ('bcsel', a, b, d)),
   (('bcsel', a, ('bcsel', b, c, d), ('bcsel(is_used_once)', b, c, 'e')), ('bcsel', b, c, ('bcsel', a, d, 'e'))),
   (('bcsel', a, ('bcsel(is_used_once)', b, c, d), ('bcsel', b, c, 'e')), ('bcsel', b, c, ('bcsel', a, d, 'e'))),
   (('bcsel', a, ('bcsel', b, c, d), ('bcsel(is_used_once)', b, 'e', d)), ('bcsel', b, ('bcsel', a, c, 'e'), d)),
   (('bcsel', a, ('bcsel(is_used_once)', b, c, d), ('bcsel', b, 'e', d)), ('bcsel', b, ('bcsel', a, c, 'e'), d)),
   (('bcsel', a, True, 'b@bool'), ('ior', a, b)),
   (('fmin', a, a), a),
   (('fmax', a, a), a),
   (('imin', a, a), a),
   (('imax', a, a), a),
   (('umin', a, a), a),
   (('umax', a, a), a),
   (('fmax', ('fmax', a, b), b), ('fmax', a, b)),
   (('umax', ('umax', a, b), b), ('umax', a, b)),
   (('imax', ('imax', a, b), b), ('imax', a, b)),
   (('fmin', ('fmin', a, b), b), ('fmin', a, b)),
   (('umin', ('umin', a, b), b), ('umin', a, b)),
   (('imin', ('imin', a, b), b), ('imin', a, b)),
   (('fmax', a, ('fneg', a)), ('fabs', a)),
   (('imax', a, ('ineg', a)), ('iabs', a)),
   (('fmin', a, ('fneg', a)), ('fneg', ('fabs', a))),
   (('imin', a, ('ineg', a)), ('ineg', ('iabs', a))),
   (('fmin', a, ('fneg', ('fabs', a))), ('fneg', ('fabs', a))),
   (('imin', a, ('ineg', ('iabs', a))), ('ineg', ('iabs', a))),
   (('fmin', a, ('fabs', a)), a),
   (('imin', a, ('iabs', a)), a),
   (('fmax', a, ('fneg', ('fabs', a))), a),
   (('imax', a, ('ineg', ('iabs', a))), a),
   (('fmax', a, ('fabs', a)), ('fabs', a)),
   (('imax', a, ('iabs', a)), ('iabs', a)),
   (('fmax', a, ('fneg', a)), ('fabs', a)),
   (('imax', a, ('ineg', a)), ('iabs', a)),
   (('~fmin', ('fmax', a, 0.0), 1.0), ('fsat', a), '!options->lower_fsat'),
   (('~fmax', ('fmin', a, 1.0), 0.0), ('fsat', a), '!options->lower_fsat'),
   (('fsat', ('fsign', a)), ('b2f', ('flt', 0.0, a))),
   (('fsat', ('b2f', a)), ('b2f', a)),
   (('fsat', a), ('fmin', ('fmax', a, 0.0), 1.0), 'options->lower_fsat'),
   (('fsat', ('fsat', a)), ('fsat', a)),
   (('fmin', ('fmax', ('fmin', ('fmax', a, b), c), b), c), ('fmin', ('fmax', a, b), c)),
   (('imin', ('imax', ('imin', ('imax', a, b), c), b), c), ('imin', ('imax', a, b), c)),
   (('umin', ('umax', ('umin', ('umax', a, b), c), b), c), ('umin', ('umax', a, b), c)),
   (('fmax', ('fsat', a), '#b@32(is_zero_to_one)'), ('fsat', ('fmax', a, b))),
   (('fmin', ('fsat', a), '#b@32(is_zero_to_one)'), ('fsat', ('fmin', a, b))),
   (('extract_u8', ('imin', ('imax', a, 0), 0xff), 0), ('imin', ('imax', a, 0), 0xff)),
   (('~ior', ('flt(is_used_once)', a, b), ('flt', a, c)), ('flt', a, ('fmax', b, c))),
   (('~ior', ('flt(is_used_once)', a, c), ('flt', b, c)), ('flt', ('fmin', a, b), c)),
   (('~ior', ('fge(is_used_once)', a, b), ('fge', a, c)), ('fge', a, ('fmin', b, c))),
   (('~ior', ('fge(is_used_once)', a, c), ('fge', b, c)), ('fge', ('fmax', a, b), c)),
   (('~ior', ('flt', a, '#b'), ('flt', a, '#c')), ('flt', a, ('fmax', b, c))),
   (('~ior', ('flt', '#a', c), ('flt', '#b', c)), ('flt', ('fmin', a, b), c)),
   (('~ior', ('fge', a, '#b'), ('fge', a, '#c')), ('fge', a, ('fmin', b, c))),
   (('~ior', ('fge', '#a', c), ('fge', '#b', c)), ('fge', ('fmax', a, b), c)),
   (('~iand', ('flt(is_used_once)', a, b), ('flt', a, c)), ('flt', a, ('fmin', b, c))),
   (('~iand', ('flt(is_used_once)', a, c), ('flt', b, c)), ('flt', ('fmax', a, b), c)),
   (('~iand', ('fge(is_used_once)', a, b), ('fge', a, c)), ('fge', a, ('fmax', b, c))),
   (('~iand', ('fge(is_used_once)', a, c), ('fge', b, c)), ('fge', ('fmin', a, b), c)),
   (('~iand', ('flt', a, '#b'), ('flt', a, '#c')), ('flt', a, ('fmin', b, c))),
   (('~iand', ('flt', '#a', c), ('flt', '#b', c)), ('flt', ('fmax', a, b), c)),
   (('~iand', ('fge', a, '#b'), ('fge', a, '#c')), ('fge', a, ('fmax', b, c))),
   (('~iand', ('fge', '#a', c), ('fge', '#b', c)), ('fge', ('fmin', a, b), c)),

   (('ior', ('ilt(is_used_once)', a, b), ('ilt', a, c)), ('ilt', a, ('imax', b, c))),
   (('ior', ('ilt(is_used_once)', a, c), ('ilt', b, c)), ('ilt', ('imin', a, b), c)),
   (('ior', ('ige(is_used_once)', a, b), ('ige', a, c)), ('ige', a, ('imin', b, c))),
   (('ior', ('ige(is_used_once)', a, c), ('ige', b, c)), ('ige', ('imax', a, b), c)),
   (('ior', ('ult(is_used_once)', a, b), ('ult', a, c)), ('ult', a, ('umax', b, c))),
   (('ior', ('ult(is_used_once)', a, c), ('ult', b, c)), ('ult', ('umin', a, b), c)),
   (('ior', ('uge(is_used_once)', a, b), ('uge', a, c)), ('uge', a, ('umin', b, c))),
   (('ior', ('uge(is_used_once)', a, c), ('uge', b, c)), ('uge', ('umax', a, b), c)),
   (('iand', ('ilt(is_used_once)', a, b), ('ilt', a, c)), ('ilt', a, ('imin', b, c))),
   (('iand', ('ilt(is_used_once)', a, c), ('ilt', b, c)), ('ilt', ('imax', a, b), c)),
   (('iand', ('ige(is_used_once)', a, b), ('ige', a, c)), ('ige', a, ('imax', b, c))),
   (('iand', ('ige(is_used_once)', a, c), ('ige', b, c)), ('ige', ('imin', a, b), c)),
   (('iand', ('ult(is_used_once)', a, b), ('ult', a, c)), ('ult', a, ('umin', b, c))),
   (('iand', ('ult(is_used_once)', a, c), ('ult', b, c)), ('ult', ('umax', a, b), c)),
   (('iand', ('uge(is_used_once)', a, b), ('uge', a, c)), ('uge', a, ('umax', b, c))),
   (('iand', ('uge(is_used_once)', a, c), ('uge', b, c)), ('uge', ('umin', a, b), c)),

   # Common pattern like 'if (i == 0 || i == 1 || ...)'
   (('ior', ('ieq', a, 0), ('ieq', a, 1)), ('uge', 1, a)),
   (('ior', ('uge', 1, a), ('ieq', a, 2)), ('uge', 2, a)),
   (('ior', ('uge', 2, a), ('ieq', a, 3)), ('uge', 3, a)),

   (('ior', 'a@bool', ('ieq', a, False)), True),
   (('ior', a, ('inot', a)), -1),

   (('iand', ('ieq', 'a@32', 0), ('ieq', 'b@32', 0)), ('ieq', ('ior', 'a@32', 'b@32'), 0)),

   # These patterns can result when (a < b || a < c) => (a < min(b, c))
   # transformations occur before constant propagation and loop-unrolling.
   (('~flt', a, ('fmax', b, a)), ('flt', a, b)),
   (('~flt', ('fmin', a, b), a), ('flt', b, a)),
   (('~fge', a, ('fmin', b, a)), True),
   (('~fge', ('fmax', a, b), a), True),
   (('~flt', a, ('fmin', b, a)), False),
   (('~flt', ('fmax', a, b), a), False),
   (('~fge', a, ('fmax', b, a)), ('fge', a, b)),
   (('~fge', ('fmin', a, b), a), ('fge', b, a)),

   (('ilt', a, ('imax', b, a)), ('ilt', a, b)),
   (('ilt', ('imin', a, b), a), ('ilt', b, a)),
   (('ige', a, ('imin', b, a)), True),
   (('ige', ('imax', a, b), a), True),
   (('ult', a, ('umax', b, a)), ('ult', a, b)),
   (('ult', ('umin', a, b), a), ('ult', b, a)),
   (('uge', a, ('umin', b, a)), True),
   (('uge', ('umax', a, b), a), True),
   (('ilt', a, ('imin', b, a)), False),
   (('ilt', ('imax', a, b), a), False),
   (('ige', a, ('imax', b, a)), ('ige', a, b)),
   (('ige', ('imin', a, b), a), ('ige', b, a)),
   (('ult', a, ('umin', b, a)), False),
   (('ult', ('umax', a, b), a), False),
   (('uge', a, ('umax', b, a)), ('uge', a, b)),
   (('uge', ('umin', a, b), a), ('uge', b, a)),

   (('ilt', '#a', ('imax', '#b', c)), ('ior', ('ilt', a, b), ('ilt', a, c))),
   (('ilt', ('imin', '#a', b), '#c'), ('ior', ('ilt', a, c), ('ilt', b, c))),
   (('ige', '#a', ('imin', '#b', c)), ('ior', ('ige', a, b), ('ige', a, c))),
   (('ige', ('imax', '#a', b), '#c'), ('ior', ('ige', a, c), ('ige', b, c))),
   (('ult', '#a', ('umax', '#b', c)), ('ior', ('ult', a, b), ('ult', a, c))),
   (('ult', ('umin', '#a', b), '#c'), ('ior', ('ult', a, c), ('ult', b, c))),
   (('uge', '#a', ('umin', '#b', c)), ('ior', ('uge', a, b), ('uge', a, c))),
   (('uge', ('umax', '#a', b), '#c'), ('ior', ('uge', a, c), ('uge', b, c))),
   (('ilt', '#a', ('imin', '#b', c)), ('iand', ('ilt', a, b), ('ilt', a, c))),
   (('ilt', ('imax', '#a', b), '#c'), ('iand', ('ilt', a, c), ('ilt', b, c))),
   (('ige', '#a', ('imax', '#b', c)), ('iand', ('ige', a, b), ('ige', a, c))),
   (('ige', ('imin', '#a', b), '#c'), ('iand', ('ige', a, c), ('ige', b, c))),
   (('ult', '#a', ('umin', '#b', c)), ('iand', ('ult', a, b), ('ult', a, c))),
   (('ult', ('umax', '#a', b), '#c'), ('iand', ('ult', a, c), ('ult', b, c))),
   (('uge', '#a', ('umax', '#b', c)), ('iand', ('uge', a, b), ('uge', a, c))),
   (('uge', ('umin', '#a', b), '#c'), ('iand', ('uge', a, c), ('uge', b, c))),

   # Thanks to sign extension, the ishr(a, b) is negative if and only if a is
   # negative.
   (('bcsel', ('ilt', a, 0), ('ineg', ('ishr', a, b)), ('ishr', a, b)),
    ('iabs', ('ishr', a, b))),
   (('iabs', ('ishr', ('iabs', a), b)), ('ishr', ('iabs', a), b)),

   (('fabs', ('slt', a, b)), ('slt', a, b)),
   (('fabs', ('sge', a, b)), ('sge', a, b)),
   (('fabs', ('seq', a, b)), ('seq', a, b)),
   (('fabs', ('sne', a, b)), ('sne', a, b)),
   (('slt', a, b), ('b2f', ('flt', a, b)), 'options->lower_scmp'),
   (('sge', a, b), ('b2f', ('fge', a, b)), 'options->lower_scmp'),
   (('seq', a, b), ('b2f', ('feq', a, b)), 'options->lower_scmp'),
   (('sne', a, b), ('b2f', ('fne', a, b)), 'options->lower_scmp'),
   (('fne', ('fneg', a), a), ('fne', a, 0.0)),
   (('feq', ('fneg', a), a), ('feq', a, 0.0)),
   # Emulating booleans
   (('imul', ('b2i', 'a@1'), ('b2i', 'b@1')), ('b2i', ('iand', a, b))),
   (('fmul', ('b2f', 'a@1'), ('b2f', 'b@1')), ('b2f', ('iand', a, b))),
   (('fsat', ('fadd', ('b2f', 'a@1'), ('b2f', 'b@1'))), ('b2f', ('ior', a, b))),
   (('iand', 'a@bool32', 1.0), ('b2f', a)),
   # True/False are ~0 and 0 in NIR.  b2i of True is 1, and -1 is ~0 (True).
   (('ineg', ('b2i32', 'a@32')), a),
   (('flt', ('fneg', ('b2f', 'a@1')), 0), a), # Generated by TGSI KILL_IF.
   (('flt', ('fsub', 0.0, ('b2f', 'a@1')), 0), a), # Generated by TGSI KILL_IF.
   # Comparison with the same args.  Note that these are not done for
   # the float versions because NaN always returns false on float
   # inequalities.
   (('ilt', a, a), False),
   (('ige', a, a), True),
   (('ieq', a, a), True),
   (('ine', a, a), False),
   (('ult', a, a), False),
   (('uge', a, a), True),
   # Logical and bit operations
   (('fand', a, 0.0), 0.0),
   (('iand', a, a), a),
   (('iand', a, ~0), a),
   (('iand', a, 0), 0),
   (('ior', a, a), a),
   (('ior', a, 0), a),
   (('ior', a, True), True),
   (('fxor', a, a), 0.0),
   (('ixor', a, a), 0),
   (('ixor', a, 0), a),
   (('inot', ('inot', a)), a),
   (('ior', ('iand', a, b), b), b),
   (('ior', ('ior', a, b), b), ('ior', a, b)),
   (('iand', ('ior', a, b), b), b),
   (('iand', ('iand', a, b), b), ('iand', a, b)),
   # DeMorgan's Laws
   (('iand', ('inot', a), ('inot', b)), ('inot', ('ior',  a, b))),
   (('ior',  ('inot', a), ('inot', b)), ('inot', ('iand', a, b))),
   # Shift optimizations
   (('ishl', 0, a), 0),
   (('ishl', a, 0), a),
   (('ishr', 0, a), 0),
   (('ishr', a, 0), a),
   (('ushr', 0, a), 0),
   (('ushr', a, 0), a),
   (('iand', 0xff, ('ushr@32', a, 24)), ('ushr', a, 24)),
   (('iand', 0xffff, ('ushr@32', a, 16)), ('ushr', a, 16)),
   # Exponential/logarithmic identities
   (('~fexp2', ('flog2', a)), a), # 2^lg2(a) = a
   (('~flog2', ('fexp2', a)), a), # lg2(2^a) = a
   (('fpow', a, b), ('fexp2', ('fmul', ('flog2', a), b)), 'options->lower_fpow'), # a^b = 2^(lg2(a)*b)
   (('~fexp2', ('fmul', ('flog2', a), b)), ('fpow', a, b), '!options->lower_fpow'), # 2^(lg2(a)*b) = a^b
   (('~fexp2', ('fadd', ('fmul', ('flog2', a), b), ('fmul', ('flog2', c), d))),
    ('~fmul', ('fpow', a, b), ('fpow', c, d)), '!options->lower_fpow'), # 2^(lg2(a) * b + lg2(c) + d) = a^b * c^d
   (('~fexp2', ('fmul', ('flog2', a), 2.0)), ('fmul', a, a)),
   (('~fexp2', ('fmul', ('flog2', a), 4.0)), ('fmul', ('fmul', a, a), ('fmul', a, a))),
   (('~fpow', a, 1.0), a),
   (('~fpow', a, 2.0), ('fmul', a, a)),
   (('~fpow', a, 4.0), ('fmul', ('fmul', a, a), ('fmul', a, a))),
   (('~fpow', 2.0, a), ('fexp2', a)),
   (('~fpow', ('fpow', a, 2.2), 0.454545), a),
   (('~fpow', ('fabs', ('fpow', a, 2.2)), 0.454545), ('fabs', a)),
   (('~fsqrt', ('fexp2', a)), ('fexp2', ('fmul', 0.5, a))),
   (('~frcp', ('fexp2', a)), ('fexp2', ('fneg', a))),
   (('~frsq', ('fexp2', a)), ('fexp2', ('fmul', -0.5, a))),
   (('~flog2', ('fsqrt', a)), ('fmul', 0.5, ('flog2', a))),
   (('~flog2', ('frcp', a)), ('fneg', ('flog2', a))),
   (('~flog2', ('frsq', a)), ('fmul', -0.5, ('flog2', a))),
   (('~flog2', ('fpow', a, b)), ('fmul', b, ('flog2', a))),
   (('~fmul', ('fexp2(is_used_once)', a), ('fexp2(is_used_once)', b)), ('fexp2', ('fadd', a, b))),
   # Division and reciprocal
   (('~fdiv', 1.0, a), ('frcp', a)),
   (('fdiv', a, b), ('fmul', a, ('frcp', b)), 'options->lower_fdiv'),
   (('~frcp', ('frcp', a)), a),
   (('~frcp', ('fsqrt', a)), ('frsq', a)),
   (('fsqrt', a), ('frcp', ('frsq', a)), 'options->lower_fsqrt'),
   (('~frcp', ('frsq', a)), ('fsqrt', a), '!options->lower_fsqrt'),
   # Boolean simplifications
   (('i2b32(is_used_by_if)', a), ('ine32', a, 0)),
   (('i2b1(is_used_by_if)', a), ('ine', a, 0)),
   (('ieq', 'a@bool', True), a),
   (('ine(is_not_used_by_if)', 'a@bool', True), ('inot', a)),
   (('ine', 'a@bool', False), a),
   (('ieq(is_not_used_by_if)', 'a@bool', False), ('inot', 'a')),
   (('bcsel', a, True, False), a),
   (('bcsel', a, False, True), ('inot', a)),
   (('bcsel@32', a, 1.0, 0.0), ('b2f', a)),
   (('bcsel@32', a, 0.0, 1.0), ('b2f', ('inot', a))),
   (('bcsel@32', a, -1.0, -0.0), ('fneg', ('b2f', a))),
   (('bcsel@32', a, -0.0, -1.0), ('fneg', ('b2f', ('inot', a)))),
   (('bcsel', True, b, c), b),
   (('bcsel', False, b, c), c),
   (('bcsel', a, ('b2f(is_used_once)', 'b@32'), ('b2f', 'c@32')), ('b2f', ('bcsel', a, b, c))),
   # The result of this should be hit by constant propagation and, in the
   # next round of opt_algebraic, get picked up by one of the above two.
   (('bcsel', '#a', b, c), ('bcsel', ('ine', 'a', 0), b, c)),

   (('bcsel', a, b, b), b),
   (('fcsel', a, b, b), b),

   # D3D Boolean emulation
   (('bcsel', a, -1, 0), ('ineg', ('b2i', 'a@1'))),
   (('bcsel', a, 0, -1), ('ineg', ('b2i', ('inot', a)))),
   (('iand', ('ineg', ('b2i', 'a@1')), ('ineg', ('b2i', 'b@1'))),
    ('ineg', ('b2i', ('iand', a, b)))),
   (('ior', ('ineg', ('b2i','a@1')), ('ineg', ('b2i', 'b@1'))),
    ('ineg', ('b2i', ('ior', a, b)))),
   (('ieq', ('ineg', ('b2i', 'a@1')), 0), ('inot', a)),
   (('ieq', ('ineg', ('b2i', 'a@1')), -1), a),
   (('ine', ('ineg', ('b2i', 'a@1')), 0), a),
   (('ine', ('ineg', ('b2i', 'a@1')), -1), ('inot', a)),
   (('iand', ('ineg', ('b2i', a)), 1.0), ('b2f', a)),

   # SM5 32-bit shifts are defined to use the 5 least significant bits
   (('ishl', 'a@32', ('iand', 31, b)), ('ishl', a, b)),
   (('ishr', 'a@32', ('iand', 31, b)), ('ishr', a, b)),
   (('ushr', 'a@32', ('iand', 31, b)), ('ushr', a, b)),

   # Conversions
   (('i2b32', ('b2i', 'a@32')), a),
   (('f2i', ('ftrunc', a)), ('f2i', a)),
   (('f2u', ('ftrunc', a)), ('f2u', a)),
   (('i2b', ('ineg', a)), ('i2b', a)),
   (('i2b', ('iabs', a)), ('i2b', a)),
   (('fabs', ('b2f', a)), ('b2f', a)),
   (('iabs', ('b2i', a)), ('b2i', a)),
   (('inot', ('f2b1', a)), ('feq', a, 0.0)),

   # Ironically, mark these as imprecise because removing the conversions may
   # preserve more precision than doing the conversions (e.g.,
   # uint(float(0x81818181u)) == 0x81818200).
   (('~f2i32', ('i2f', 'a@32')), a),
   (('~f2i32', ('u2f', 'a@32')), a),
   (('~f2u32', ('i2f', 'a@32')), a),
   (('~f2u32', ('u2f', 'a@32')), a),

   # Section 5.4.1 (Conversion and Scalar Constructors) of the GLSL 4.60 spec
   # says:
   #
   #    It is undefined to convert a negative floating-point value to an
   #    uint.
   #
   # Assuming that (uint)some_float behaves like (uint)(int)some_float allows
   # some optimizations in the i965 backend to proceed.
   (('ige', ('f2u', a), b), ('ige', ('f2i', a), b)),
   (('ige', b, ('f2u', a)), ('ige', b, ('f2i', a))),
   (('ilt', ('f2u', a), b), ('ilt', ('f2i', a), b)),
   (('ilt', b, ('f2u', a)), ('ilt', b, ('f2i', a))),

   # Packing and then unpacking does nothing
   (('unpack_64_2x32_split_x', ('pack_64_2x32_split', a, b)), a),
   (('unpack_64_2x32_split_y', ('pack_64_2x32_split', a, b)), b),
   (('pack_64_2x32_split', ('unpack_64_2x32_split_x', a),
                           ('unpack_64_2x32_split_y', a)), a),

   # Byte extraction
   (('ushr', 'a@16',  8), ('extract_u8', a, 1), '!options->lower_extract_byte'),
   (('ushr', 'a@32', 24), ('extract_u8', a, 3), '!options->lower_extract_byte'),
   (('ushr', 'a@64', 56), ('extract_u8', a, 7), '!options->lower_extract_byte'),
   (('ishr', 'a@16',  8), ('extract_i8', a, 1), '!options->lower_extract_byte'),
   (('ishr', 'a@32', 24), ('extract_i8', a, 3), '!options->lower_extract_byte'),
   (('ishr', 'a@64', 56), ('extract_i8', a, 7), '!options->lower_extract_byte'),
   (('iand', 0xff, a), ('extract_u8', a, 0), '!options->lower_extract_byte')
]

# After the ('extract_u8', a, 0) pattern, above, triggers, there will be
# patterns like those below.
for op in ('ushr', 'ishr'):
   optimizations.extend([(('extract_u8', (op, 'a@16',  8),     0), ('extract_u8', a, 1))])
   optimizations.extend([(('extract_u8', (op, 'a@32',  8 * i), 0), ('extract_u8', a, i)) for i in range(1, 4)])
   optimizations.extend([(('extract_u8', (op, 'a@64',  8 * i), 0), ('extract_u8', a, i)) for i in range(1, 8)])

optimizations.extend([(('extract_u8', ('extract_u16', a, 1), 0), ('extract_u8', a, 2))])

# After the ('extract_[iu]8', a, 3) patterns, above, trigger, there will be
# patterns like those below.
for op in ('extract_u8', 'extract_i8'):
   optimizations.extend([((op, ('ishl', 'a@16',      8),     1), (op, a, 0))])
   optimizations.extend([((op, ('ishl', 'a@32', 24 - 8 * i), 3), (op, a, i)) for i in range(2, -1, -1)])
   optimizations.extend([((op, ('ishl', 'a@64', 56 - 8 * i), 7), (op, a, i)) for i in range(6, -1, -1)])

optimizations.extend([
    # Word extraction
   (('ushr', ('ishl', 'a@32', 16), 16), ('extract_u16', a, 0), '!options->lower_extract_word'),
   (('ushr', 'a@32', 16), ('extract_u16', a, 1), '!options->lower_extract_word'),
   (('ishr', ('ishl', 'a@32', 16), 16), ('extract_i16', a, 0), '!options->lower_extract_word'),
   (('ishr', 'a@32', 16), ('extract_i16', a, 1), '!options->lower_extract_word'),
   (('iand', 0xffff, a), ('extract_u16', a, 0), '!options->lower_extract_word'),

   # Subtracts
   (('~fsub', a, ('fsub', 0.0, b)), ('fadd', a, b)),
   (('isub', a, ('isub', 0, b)), ('iadd', a, b)),
   (('ussub_4x8', a, 0), a),
   (('ussub_4x8', a, ~0), 0),
   (('fsub', a, b), ('fadd', a, ('fneg', b)), 'options->lower_sub'),
   (('isub', a, b), ('iadd', a, ('ineg', b)), 'options->lower_sub'),
   (('fneg', a), ('fsub', 0.0, a), 'options->lower_negate'),
   (('ineg', a), ('isub', 0, a), 'options->lower_negate'),
   (('~fadd', a, ('fsub', 0.0, b)), ('fsub', a, b)),
   (('iadd', a, ('isub', 0, b)), ('isub', a, b)),
   (('fabs', ('fsub', 0.0, a)), ('fabs', a)),
   (('iabs', ('isub', 0, a)), ('iabs', a)),

   # Propagate negation up multiplication chains
   (('fmul', ('fneg', a), b), ('fneg', ('fmul', a, b))),
   (('imul', ('ineg', a), b), ('ineg', ('imul', a, b))),

   # Propagate constants up multiplication chains
   (('~fmul(is_used_once)', ('fmul(is_used_once)', 'a(is_not_const)', 'b(is_not_const)'), '#c'), ('fmul', ('fmul', a, c), b)),
   (('imul(is_used_once)', ('imul(is_used_once)', 'a(is_not_const)', 'b(is_not_const)'), '#c'), ('imul', ('imul', a, c), b)),
   (('~fadd(is_used_once)', ('fadd(is_used_once)', 'a(is_not_const)', 'b(is_not_const)'), '#c'), ('fadd', ('fadd', a, c), b)),
   (('iadd(is_used_once)', ('iadd(is_used_once)', 'a(is_not_const)', 'b(is_not_const)'), '#c'), ('iadd', ('iadd', a, c), b)),

   # Reassociate constants in add/mul chains so they can be folded together.
   # For now, we mostly only handle cases where the constants are separated by
   # a single non-constant.  We could do better eventually.
   (('~fmul', '#a', ('fmul', 'b(is_not_const)', '#c')), ('fmul', ('fmul', a, c), b)),
   (('imul', '#a', ('imul', 'b(is_not_const)', '#c')), ('imul', ('imul', a, c), b)),
   (('~fadd', '#a',          ('fadd', 'b(is_not_const)', '#c')),  ('fadd', ('fadd', a,          c),           b)),
   (('~fadd', '#a', ('fneg', ('fadd', 'b(is_not_const)', '#c'))), ('fadd', ('fadd', a, ('fneg', c)), ('fneg', b))),
   (('iadd', '#a', ('iadd', 'b(is_not_const)', '#c')), ('iadd', ('iadd', a, c), b)),

   # By definition...
   (('bcsel', ('ige', ('find_lsb', a), 0), ('find_lsb', a), -1), ('find_lsb', a)),
   (('bcsel', ('ige', ('ifind_msb', a), 0), ('ifind_msb', a), -1), ('ifind_msb', a)),
   (('bcsel', ('ige', ('ufind_msb', a), 0), ('ufind_msb', a), -1), ('ufind_msb', a)),

   (('bcsel', ('ine', a, 0), ('find_lsb', a), -1), ('find_lsb', a)),
   (('bcsel', ('ine', a, 0), ('ifind_msb', a), -1), ('ifind_msb', a)),
   (('bcsel', ('ine', a, 0), ('ufind_msb', a), -1), ('ufind_msb', a)),

   (('bcsel', ('ine', a, -1), ('ifind_msb', a), -1), ('ifind_msb', a)),

   # Misc. lowering
   (('fmod@16', a, b), ('fsub', a, ('fmul', b, ('ffloor', ('fdiv', a, b)))), 'options->lower_fmod16'),
   (('fmod@32', a, b), ('fsub', a, ('fmul', b, ('ffloor', ('fdiv', a, b)))), 'options->lower_fmod32'),
   (('fmod@64', a, b), ('fsub', a, ('fmul', b, ('ffloor', ('fdiv', a, b)))), 'options->lower_fmod64'),
   (('frem', a, b), ('fsub', a, ('fmul', b, ('ftrunc', ('fdiv', a, b)))), 'options->lower_fmod32'),
   (('uadd_carry@32', a, b), ('b2i', ('ult', ('iadd', a, b), a)), 'options->lower_uadd_carry'),
   (('usub_borrow@32', a, b), ('b2i', ('ult', a, b)), 'options->lower_usub_borrow'),

   (('bitfield_insert', 'base', 'insert', 'offset', 'bits'),
    ('bcsel', ('ilt', 31, 'bits'), 'insert',
              ('bfi', ('bfm', 'bits', 'offset'), 'insert', 'base')),
    'options->lower_bitfield_insert'),
   (('ihadd', a, b), ('iadd', ('iand', a, b), ('ishr', ('ixor', a, b), 1)), 'options->lower_hadd'),
   (('uhadd', a, b), ('iadd', ('iand', a, b), ('ushr', ('ixor', a, b), 1)), 'options->lower_hadd'),
   (('irhadd', a, b), ('isub', ('ior', a, b), ('ishr', ('ixor', a, b), 1)), 'options->lower_hadd'),
   (('urhadd', a, b), ('isub', ('ior', a, b), ('ushr', ('ixor', a, b), 1)), 'options->lower_hadd'),
   (('uadd_sat', a, b), ('bcsel', ('ult', ('iadd', a, b), a), -1, ('iadd', a, b)), 'options->lower_add_sat'),
   (('usub_sat', a, b), ('bcsel', ('ult', a, b), 0, ('isub', a, b)), 'options->lower_add_sat'),

   # Alternative lowering that doesn't rely on bfi.
   (('bitfield_insert', 'base', 'insert', 'offset', 'bits'),
    ('bcsel', ('ilt', 31, 'bits'),
     'insert',
     ('ior',
      ('iand', 'base', ('inot', ('bfm', 'bits', 'offset'))),
      ('iand', ('ishl', 'insert', 'offset'), ('bfm', 'bits', 'offset')))),
    'options->lower_bitfield_insert_to_shifts'),

   # bfm lowering -- note that the NIR opcode is undefined if either arg is 32.
   (('bfm', 'bits', 'offset'),
    ('ishl', ('isub', ('ishl', 1, 'bits'), 1), 'offset'),
    'options->lower_bfm'),

   (('ibitfield_extract', 'value', 'offset', 'bits'),
    ('bcsel', ('ilt', 31, 'bits'), 'value',
              ('ibfe', 'value', 'offset', 'bits')),
    'options->lower_bitfield_extract'),

   (('ubitfield_extract', 'value', 'offset', 'bits'),
    ('bcsel', ('ult', 31, 'bits'), 'value',
              ('ubfe', 'value', 'offset', 'bits')),
    'options->lower_bitfield_extract'),

   (('ibitfield_extract', 'value', 'offset', 'bits'),
    ('bcsel', ('ieq', 0, 'bits'),
     0,
     ('ishr',
       ('ishl', 'value', ('isub', ('isub', 32, 'bits'), 'offset')),
       ('isub', 32, 'bits'))),
    'options->lower_bitfield_extract_to_shifts'),

   (('ubitfield_extract', 'value', 'offset', 'bits'),
    ('iand',
     ('ushr', 'value', 'offset'),
     ('bcsel', ('ieq', 'bits', 32),
      0xffffffff,
      ('bfm', 'bits', 0))),
    'options->lower_bitfield_extract_to_shifts'),

   (('ifind_msb', 'value'),
    ('ufind_msb', ('bcsel', ('ilt', 'value', 0), ('inot', 'value'), 'value')),
    'options->lower_ifind_msb'),

   (('find_lsb', 'value'),
    ('ufind_msb', ('iand', 'value', ('ineg', 'value'))),
    'options->lower_find_lsb'),

   (('extract_i8', a, 'b@32'),
    ('ishr', ('ishl', a, ('imul', ('isub', 3, b), 8)), 24),
    'options->lower_extract_byte'),

   (('extract_u8', a, 'b@32'),
    ('iand', ('ushr', a, ('imul', b, 8)), 0xff),
    'options->lower_extract_byte'),

   (('extract_i16', a, 'b@32'),
    ('ishr', ('ishl', a, ('imul', ('isub', 1, b), 16)), 16),
    'options->lower_extract_word'),

   (('extract_u16', a, 'b@32'),
    ('iand', ('ushr', a, ('imul', b, 16)), 0xffff),
    'options->lower_extract_word'),

    (('pack_unorm_2x16', 'v'),
     ('pack_uvec2_to_uint',
        ('f2u32', ('fround_even', ('fmul', ('fsat', 'v'), 65535.0)))),
     'options->lower_pack_unorm_2x16'),

    (('pack_unorm_4x8', 'v'),
     ('pack_uvec4_to_uint',
        ('f2u32', ('fround_even', ('fmul', ('fsat', 'v'), 255.0)))),
     'options->lower_pack_unorm_4x8'),

    (('pack_snorm_2x16', 'v'),
     ('pack_uvec2_to_uint',
        ('f2i32', ('fround_even', ('fmul', ('fmin', 1.0, ('fmax', -1.0, 'v')), 32767.0)))),
     'options->lower_pack_snorm_2x16'),

    (('pack_snorm_4x8', 'v'),
     ('pack_uvec4_to_uint',
        ('f2i32', ('fround_even', ('fmul', ('fmin', 1.0, ('fmax', -1.0, 'v')), 127.0)))),
     'options->lower_pack_snorm_4x8'),

    (('unpack_unorm_2x16', 'v'),
     ('fdiv', ('u2f32', ('vec2', ('extract_u16', 'v', 0),
                                  ('extract_u16', 'v', 1))),
              65535.0),
     'options->lower_unpack_unorm_2x16'),

    (('unpack_unorm_4x8', 'v'),
     ('fdiv', ('u2f32', ('vec4', ('extract_u8', 'v', 0),
                                  ('extract_u8', 'v', 1),
                                  ('extract_u8', 'v', 2),
                                  ('extract_u8', 'v', 3))),
              255.0),
     'options->lower_unpack_unorm_4x8'),

    (('unpack_snorm_2x16', 'v'),
     ('fmin', 1.0, ('fmax', -1.0, ('fdiv', ('i2f', ('vec2', ('extract_i16', 'v', 0),
                                                            ('extract_i16', 'v', 1))),
                                           32767.0))),
     'options->lower_unpack_snorm_2x16'),

    (('unpack_snorm_4x8', 'v'),
     ('fmin', 1.0, ('fmax', -1.0, ('fdiv', ('i2f', ('vec4', ('extract_i8', 'v', 0),
                                                            ('extract_i8', 'v', 1),
                                                            ('extract_i8', 'v', 2),
                                                            ('extract_i8', 'v', 3))),
                                           127.0))),
     'options->lower_unpack_snorm_4x8'),

   (('isign', a), ('imin', ('imax', a, -1), 1), 'options->lower_isign'),
])

# bit_size dependent lowerings
for bit_size in [8, 16, 32, 64]:
   # convenience constants
   intmax = (1 << (bit_size - 1)) - 1
   intmin = 1 << (bit_size - 1)

   optimizations += [
      (('iadd_sat@' + str(bit_size), a, b),
       ('bcsel', ('ige', b, 1), ('bcsel', ('ilt', ('iadd', a, b), a), intmax, ('iadd', a, b)),
                                ('bcsel', ('ilt', a, ('iadd', a, b)), intmin, ('iadd', a, b))), 'options->lower_add_sat'),
      (('isub_sat@' + str(bit_size), a, b),
       ('bcsel', ('ilt', b, 0), ('bcsel', ('ilt', ('isub', a, b), a), intmax, ('isub', a, b)),
                                ('bcsel', ('ilt', a, ('isub', a, b)), intmin, ('isub', a, b))), 'options->lower_add_sat'),
   ]

invert = OrderedDict([('feq', 'fne'), ('fne', 'feq'), ('fge', 'flt'), ('flt', 'fge')])

for left, right in itertools.combinations_with_replacement(invert.keys(), 2):
   optimizations.append((('inot', ('ior(is_used_once)', (left, a, b), (right, c, d))),
                         ('iand', (invert[left], a, b), (invert[right], c, d))))
   optimizations.append((('inot', ('iand(is_used_once)', (left, a, b), (right, c, d))),
                         ('ior', (invert[left], a, b), (invert[right], c, d))))

# Optimize x2bN(b2x(x)) -> x
for size in type_sizes('bool'):
    aN = 'a@' + str(size)
    f2bN = 'f2b' + str(size)
    i2bN = 'i2b' + str(size)
    optimizations.append(((f2bN, ('b2f', aN)), a))
    optimizations.append(((i2bN, ('b2i', aN)), a))

# Optimize x2yN(b2x(x)) -> b2y
for x, y in itertools.product(['f', 'u', 'i'], ['f', 'u', 'i']):
   if x != 'f' and y != 'f' and x != y:
      continue

   b2x = 'b2f' if x == 'f' else 'b2i'
   b2y = 'b2f' if y == 'f' else 'b2i'
   x2yN = '{}2{}'.format(x, y)
   optimizations.append(((x2yN, (b2x, a)), (b2y, a)))

def fexp2i(exp, bits):
   # We assume that exp is already in the right range.
   if bits == 16:
      return ('i2i16', ('ishl', ('iadd', exp, 15), 10))
   elif bits == 32:
      return ('ishl', ('iadd', exp, 127), 23)
   elif bits == 64:
      return ('pack_64_2x32_split', 0, ('ishl', ('iadd', exp, 1023), 20))
   else:
      assert False

def ldexp(f, exp, bits):
   # First, we clamp exp to a reasonable range.  The maximum possible range
   # for a normal exponent is [-126, 127] and, throwing in denormals, you get
   # a maximum range of [-149, 127].  This means that we can potentially have
   # a swing of +-276.  If you start with FLT_MAX, you actually have to do
   # ldexp(FLT_MAX, -278) to get it to flush all the way to zero.  The GLSL
   # spec, on the other hand, only requires that we handle an exponent value
   # in the range [-126, 128].  This implementation is *mostly* correct; it
   # handles a range on exp of [-252, 254] which allows you to create any
   # value (including denorms if the hardware supports it) and to adjust the
   # exponent of any normal value to anything you want.
   if bits == 16:
      exp = ('imin', ('imax', exp, -28), 30)
   elif bits == 32:
      exp = ('imin', ('imax', exp, -252), 254)
   elif bits == 64:
      exp = ('imin', ('imax', exp, -2044), 2046)
   else:
      assert False

   # Now we compute two powers of 2, one for exp/2 and one for exp-exp/2.
   # (We use ishr which isn't the same for -1, but the -1 case still works
   # since we use exp-exp/2 as the second exponent.)  While the spec
   # technically defines ldexp as f * 2.0^exp, simply multiplying once doesn't
   # work with denormals and doesn't allow for the full swing in exponents
   # that you can get with normalized values.  Instead, we create two powers
   # of two and multiply by them each in turn.  That way the effective range
   # of our exponent is doubled.
   pow2_1 = fexp2i(('ishr', exp, 1), bits)
   pow2_2 = fexp2i(('isub', exp, ('ishr', exp, 1)), bits)
   return ('fmul', ('fmul', f, pow2_1), pow2_2)

optimizations += [
   (('ldexp@16', 'x', 'exp'), ldexp('x', 'exp', 16), 'options->lower_ldexp'),
   (('ldexp@32', 'x', 'exp'), ldexp('x', 'exp', 32), 'options->lower_ldexp'),
   (('ldexp@64', 'x', 'exp'), ldexp('x', 'exp', 64), 'options->lower_ldexp'),
]

# Unreal Engine 4 demo applications open-codes bitfieldReverse()
def bitfield_reverse(u):
    step1 = ('ior', ('ishl', u, 16), ('ushr', u, 16))
    step2 = ('ior', ('ishl', ('iand', step1, 0x00ff00ff), 8), ('ushr', ('iand', step1, 0xff00ff00), 8))
    step3 = ('ior', ('ishl', ('iand', step2, 0x0f0f0f0f), 4), ('ushr', ('iand', step2, 0xf0f0f0f0), 4))
    step4 = ('ior', ('ishl', ('iand', step3, 0x33333333), 2), ('ushr', ('iand', step3, 0xcccccccc), 2))
    step5 = ('ior', ('ishl', ('iand', step4, 0x55555555), 1), ('ushr', ('iand', step4, 0xaaaaaaaa), 1))

    return step5

optimizations += [(bitfield_reverse('x@32'), ('bitfield_reverse', 'x'))]

# For any float comparison operation, "cmp", if you have "a == a && a cmp b"
# then the "a == a" is redundant because it's equivalent to "a is not NaN"
# and, if a is a NaN then the second comparison will fail anyway.
for op in ['flt', 'fge', 'feq']:
   optimizations += [
      (('iand', ('feq', a, a), (op, a, b)), (op, a, b)),
      (('iand', ('feq', a, a), (op, b, a)), (op, b, a)),
   ]

# Add optimizations to handle the case where the result of a ternary is
# compared to a constant.  This way we can take things like
#
# (a ? 0 : 1) > 0
#
# and turn it into
#
# a ? (0 > 0) : (1 > 0)
#
# which constant folding will eat for lunch.  The resulting ternary will
# further get cleaned up by the boolean reductions above and we will be
# left with just the original variable "a".
for op in ['flt', 'fge', 'feq', 'fne',
           'ilt', 'ige', 'ieq', 'ine', 'ult', 'uge']:
   optimizations += [
      ((op, ('bcsel', 'a', '#b', '#c'), '#d'),
       ('bcsel', 'a', (op, 'b', 'd'), (op, 'c', 'd'))),
      ((op, '#d', ('bcsel', a, '#b', '#c')),
       ('bcsel', 'a', (op, 'd', 'b'), (op, 'd', 'c'))),
   ]


# For example, this converts things like
#
#    1 + mix(0, a - 1, condition)
#
# into
#
#    mix(1, (a-1)+1, condition)
#
# Other optimizations will rearrange the constants.
for op in ['fadd', 'fmul', 'iadd', 'imul']:
   optimizations += [
      ((op, ('bcsel(is_used_once)', a, '#b', c), '#d'), ('bcsel', a, (op, b, d), (op, c, d)))
   ]

# This section contains "late" optimizations that should be run before
# creating ffmas and calling regular optimizations for the final time.
# Optimizations should go here if they help code generation and conflict
# with the regular optimizations.
before_ffma_optimizations = [
   # Propagate constants down multiplication chains
   (('~fmul(is_used_once)', ('fmul(is_used_once)', 'a(is_not_const)', '#b'), 'c(is_not_const)'), ('fmul', ('fmul', a, c), b)),
   (('imul(is_used_once)', ('imul(is_used_once)', 'a(is_not_const)', '#b'), 'c(is_not_const)'), ('imul', ('imul', a, c), b)),
   (('~fadd(is_used_once)', ('fadd(is_used_once)', 'a(is_not_const)', '#b'), 'c(is_not_const)'), ('fadd', ('fadd', a, c), b)),
   (('iadd(is_used_once)', ('iadd(is_used_once)', 'a(is_not_const)', '#b'), 'c(is_not_const)'), ('iadd', ('iadd', a, c), b)),

   (('~fadd', ('fmul', a, b), ('fmul', a, c)), ('fmul', a, ('fadd', b, c))),
   (('iadd', ('imul', a, b), ('imul', a, c)), ('imul', a, ('iadd', b, c))),
   (('~fadd', ('fneg', a), a), 0.0),
   (('iadd', ('ineg', a), a), 0),
   (('iadd', ('ineg', a), ('iadd', a, b)), b),
   (('iadd', a, ('iadd', ('ineg', a), b)), b),
   (('~fadd', ('fneg', a), ('fadd', a, b)), b),
   (('~fadd', a, ('fadd', ('fneg', a), b)), b),
]

# This section contains "late" optimizations that should be run after the
# regular optimizations have finished.  Optimizations should go here if
# they help code generation but do not necessarily produce code that is
# more easily optimizable.
late_optimizations = [
   # Most of these optimizations aren't quite safe when you get infinity or
   # Nan involved but the first one should be fine.
   (('flt',          ('fadd', a, b),  0.0), ('flt',          a, ('fneg', b))),
   (('flt', ('fneg', ('fadd', a, b)), 0.0), ('flt', ('fneg', a),         b)),
   (('~fge',          ('fadd', a, b),  0.0), ('fge',          a, ('fneg', b))),
   (('~fge', ('fneg', ('fadd', a, b)), 0.0), ('fge', ('fneg', a),         b)),
   (('~feq', ('fadd', a, b), 0.0), ('feq', a, ('fneg', b))),
   (('~fne', ('fadd', a, b), 0.0), ('fne', a, ('fneg', b))),

   (('~fge', ('fmin(is_used_once)', ('fadd(is_used_once)', a, b), ('fadd', c, d)), 0.0), ('iand', ('fge', a, ('fneg', b)), ('fge', c, ('fneg', d)))),

   (('fdot2', a, b), ('fdot_replicated2', a, b), 'options->fdot_replicates'),
   (('fdot3', a, b), ('fdot_replicated3', a, b), 'options->fdot_replicates'),
   (('fdot4', a, b), ('fdot_replicated4', a, b), 'options->fdot_replicates'),
   (('fdph', a, b), ('fdph_replicated', a, b), 'options->fdot_replicates'),

   (('b2f(is_used_more_than_once)', ('inot', 'a@1')), ('bcsel', a, 0.0, 1.0)),
   (('fneg(is_used_more_than_once)', ('b2f', ('inot', 'a@1'))), ('bcsel', a, -0.0, -1.0)),

   # we do these late so that we don't get in the way of creating ffmas
   (('fmin', ('fadd(is_used_once)', '#c', a), ('fadd(is_used_once)', '#c', b)), ('fadd', c, ('fmin', a, b))),
   (('fmax', ('fadd(is_used_once)', '#c', a), ('fadd(is_used_once)', '#c', b)), ('fadd', c, ('fmax', a, b))),

   (('bcsel', 'a@bool', 0, ('b2f32', ('inot', 'b@bool'))), ('b2f32', ('inot', ('ior', a, b)))),
]

print(nir_algebraic.AlgebraicPass("nir_opt_algebraic", optimizations).render())
print(nir_algebraic.AlgebraicPass("nir_opt_algebraic_before_ffma",
                                  before_ffma_optimizations).render())
print(nir_algebraic.AlgebraicPass("nir_opt_algebraic_late",
                                  late_optimizations).render())