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
|
.file "erf.s"
// Copyright (c) 2001 - 2003, Intel Corporation
// All rights reserved.
//
// Contributed 2001 by the Intel Numerics Group, Intel Corporation
//
// Redistribution and use in source and binary forms, with or without
// modification, are permitted provided that the following conditions are
// met:
//
// * Redistributions of source code must retain the above copyright
// notice, this list of conditions and the following disclaimer.
//
// * Redistributions in binary form must reproduce the above copyright
// notice, this list of conditions and the following disclaimer in the
// documentation and/or other materials provided with the distribution.
//
// * The name of Intel Corporation may not be used to endorse or promote
// products derived from this software without specific prior written
// permission.
// THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
// "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
// LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
// A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL INTEL OR ITS
// CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
// EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
// PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
// PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
// OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY OR TORT (INCLUDING
// NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
// SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
//
// Intel Corporation is the author of this code, and requests that all
// problem reports or change requests be submitted to it directly at
// http://www.intel.com/software/products/opensource/libraries/num.htm.
//
// History
//==============================================================
// 08/15/01 Initial version
// 05/20/02 Cleaned up namespace and sf0 syntax
// 02/06/03 Reordered header: .section, .global, .proc, .align
//
// API
//==============================================================
// double erf(double)
//
// Overview of operation
//==============================================================
// Background
//
//
// There are 9 paths:
// 1. x = +/-0.0
// Return erf(x) = +/-0.0
//
// 2. 0.0 < |x| < 0.5
// Return erf(x) = x *Pol9(x^2)
//
// 3. For several subranges of 0.5 <= |x| < 5.90625
// Return erf(x) = sign(x)*Pol19(y),
// where y = (|x|-b)/a, Pol19(y) = A0 + A1*y^1 + A2*y^2 + ... + A19*y^19
//
// For each subrange there is particular set of coefficients.
// Below is the list of subranges:
// 3.1 0.5 <= |x| < 1.0 b = a = 0.5
// 3.2 1.0 <= |x| < 2.0, b = a = 1.0
// 3.3 2.0 <= |x| < 3.25 b = a = 2.0
// 3.4 4.0 <= |x| < 5.90625 b = 4.0, a = 2.0
//
// 4. 3.25 <= |x| < 4.0
// Return erf(x) = sign(x)*Pol14(|x| - 3.25)
//
// 5. 5.90625 <= |x| < +INF
// Return erf(x) = sign(x)*(1.0d - 2^(-63))
//
// 6. |x| = INF
// Return erf(x) = sign(x) * 1.0
//
// 7. x = [S,Q]NaN
// Return erf(x) = QNaN
//
// 8. x is positive denormal
// Return erf(x) = A0*x - x^2,
// where A0 = 2.0/sqrt(Pi)
//
// 9. x is negative denormal
// Return erf(x) = A0*x + x^2,
// where A0 = 2.0/sqrt(Pi)
//
// Registers used
//==============================================================
// Floating Point registers used:
// f8, input, output
// f32 -> f63
// General registers used:
// r32 -> r48, r2, r3
// Predicate registers used:
// p0, p6 -> p15
// p6 to filter out case when x = denormal
// p7 to filter out case when x = [Q,S]NaN or +/-0,
// used also to process denormals
// p8 to filter out case when 3.25 <= |x| < 4.0,
// used also to process denormals
// p9 to filter out case when |x| = inf
// p10 to filter out case when |x| < 0.5
// p11 set when |x| < 3.25 or |x| > 4.0
// p12 to filter out case when |x| >= 5.90625
// p13 set if 4.0 <=|x| < 5.90625
// p14 set to 1 for positive x
// p15 set to 1 for negative x
// Assembly macros
//==============================================================
rDataPtr = r2
rDataPtr1 = r3
rBias = r33
rCoeffAddr3 = r34
rThreeAndQ = r35
rCoeffAddr2 = r36
rMask = r37
rArg = r38
rSignBit = r39
rAbsArg = r40
rSaturation = r41
rIndex = r42
rCoeffAddr1 = r43
rCoeffAddr4 = r44
rShiftedArg = r45
rShiftedArgMasked = r46
rBiasedExpOf4 = r47
rShiftedAbsArg = r48
//==============================================================
fA0 = f32
fA1 = f33
fA2 = f34
fA3 = f35
fA4 = f36
fA5 = f37
fA6 = f38
fA7 = f39
fA8 = f40
fA9 = f41
fA10 = f42
fA11 = f43
fA12 = f44
fA13 = f45
fA14 = f46
fA15 = f47
fA16 = f48
fA17 = f49
fA18 = f50
fA19 = f51
fArgSqr = f52
fArgAbsNorm = f53
fSignumX = f54
fRes = f55
fThreeAndQ = f56
fArgAbs = f57
fTSqr = f58
fTQuadr = f59
fTDeg3 = f60
fTDeg7 = f61
fArgAbsNormSgn = f62
fTQuadrSgn = f63
// Data tables
//==============================================================
RODATA
.align 64
LOCAL_OBJECT_START(erf_data)
// Coefficients ##0..15
// Polynomial coefficients for the erf(x), 0.5 <= |x| < 1.0
data8 0xB69AC40646D1F6C1, 0x00003FD2 //A19
data8 0x90AD48C0118FA10C, 0x00003FD7 //A18
data8 0x826FBAD055EA4AB8, 0x0000BFDB //A17
data8 0x8DAB171246CC2B89, 0x00003FDC //A16
data8 0xC0B1D6662F8A7564, 0x00003FDF //A15
data8 0xA46374AC35099BAF, 0x0000BFE1 //A14
data8 0xB2F230996346EF27, 0x0000BFE4 //A13
data8 0xCDEC50950FACE04A, 0x00003FE6 //A12
data8 0x826014649396E9D2, 0x00003FE9 //A11
data8 0xCDB787DC718B13F9, 0x0000BFEB //A10
data8 0x8E0B23C24EE0C8EE, 0x0000BFED //A9
data8 0xA49EA40A4E5A3F76, 0x00003FF0 //A8
data8 0xB11E30BE912617D3, 0x00003FF0 //A7
data8 0xCCF89D9351CE26E3, 0x0000BFF4 //A6
data8 0xEFF75AD1F0F22809, 0x00003FF2 //A5
data8 0xBB793EF404C09A22, 0x00003FF8 //A4
// Polynomial coefficients for the erf(x), 1.0 <= |x| < 2.0
data8 0xBAE93FF4174EA59B, 0x00003FE6 //A19
data8 0x8A0FD46092F95D44, 0x0000BFEA //A18
data8 0xA37B3242B7809E12, 0x00003FEC //A17
data8 0xA0330A5CD2E91689, 0x0000BFED //A16
data8 0x8E34A678F3497D17, 0x0000BFEC //A15
data8 0xAC185D45A2772384, 0x00003FEF //A14
data8 0xB0C11347CE7EEDE8, 0x00003FEF //A13
data8 0xD3330DC14EA0E4EB, 0x0000BFF2 //A12
data8 0xB4A6DFDE578A428F, 0x00003FF1 //A11
data8 0xA0B4034310D2D9CB, 0x00003FF5 //A10
data8 0xF71662D3132B7759, 0x0000BFF5 //A9
data8 0x9C88BF157695E9EC, 0x0000BFF7 //A8
data8 0xF84B80EFCA43895D, 0x00003FF8 //A7
data8 0x9722D22DA628A17B, 0x00003FF7 //A6
data8 0x8DB0A586F8F3381F, 0x0000BFFB //A5
data8 0x8DB0A5879F87E5BE, 0x00003FFB //A4
// Polynomial coefficients for the erf(x), 2.0 <= |x| < 3.25
data8 0x9C4AF1F3A4B21AFC, 0x00003FF6 //A19
data8 0x8D40D5D5DB741AB8, 0x0000BFF9 //A18
data8 0xDEBE7099E0A75BA4, 0x00003FFA //A17
data8 0xB99A33294D32429D, 0x0000BFFB //A16
data8 0x8109D9C7197BC7C9, 0x00003FFB //A15
data8 0xC30DE8E2EFC2D760, 0x00003FFA //A14
data8 0x80DDA28C5B35DC73, 0x0000BFFC //A13
data8 0x9BE4DE5095BACE0D, 0x00003FF9 //A12
data8 0xDA4092509EE7D111, 0x00003FFC //A11
data8 0x89D98C561B0C9040, 0x0000BFFD //A10
data8 0xD20B26EB2F0881D4, 0x0000BFF9 //A9
data8 0xD089C56948731561, 0x00003FFD //A8
data8 0xDD704DEFFB21B7E7, 0x0000BFFD //A7
data8 0xF0C9A6BBDE469115, 0x00003FF9 //A6
data8 0xD673A02CB5766633, 0x00003FFD //A5
data8 0x8D162CBAD8A12649, 0x0000BFFE //A4
// Polynomial coefficients for the erf(x), 4.0 <= |x| < 6.0
data8 0xD4428B75C6FE8FD1, 0x0000BFFC //A19
data8 0xF76BE1935675D5C8, 0x00003FFE //A18
data8 0xFD6BB3B14AA7A8E6, 0x0000BFFF //A17
data8 0x8BE8F573D348DDA4, 0x00004000 //A16
data8 0x81E91923A1030502, 0x0000BFFF //A15
data8 0xCE7FE87B26CFD286, 0x0000BFFE //A14
data8 0x84EF6B4E17404384, 0x00004000 //A13
data8 0x91FEF33015404991, 0x0000C000 //A12
data8 0xDEDF6A9370747E56, 0x00003FFF //A11
data8 0x8397E6FF56CDFD9D, 0x0000BFFF //A10
data8 0xFAD1CE912473937B, 0x00003FFD //A9
data8 0xC48C1EA8AAA624EA, 0x0000BFFC //A8
data8 0xFECAF0097ACF981B, 0x00003FFA //A7
data8 0x8829A394065E4B95, 0x0000BFF9 //A6
data8 0xED3003E477A53EE7, 0x00003FF6 //A5
data8 0xA4C07E9BB3FCB0F3, 0x0000BFF4 //A4
//
// Coefficients ##16..19
// Polynomial coefficients for the erf(x), 0.5 <= |x| < 1.0
data8 0x95FA98C337005D13, 0x0000BFF9 //A3
data8 0xE0F7E524D2808A97, 0x0000BFFB //A2
data8 0xE0F7E524D2808A98, 0x00003FFD //A1
data8 0x853F7AE0C76E915F, 0x00003FFE //A0
// Polynomial coefficients for the erf(x), 1.0 <= |x| < 2.0
data8 0x8DB0A587A96ABCF0, 0x00003FFC //A3
data8 0xD488F84B7DE18DA8, 0x0000BFFD //A2
data8 0xD488F84B7DE12E9C, 0x00003FFD //A1
data8 0xD7BB3D3A08445636, 0x00003FFE //A0
// Polynomial coefficients for the erf(x), 2.0 <= |x| < 3.25
data8 0xC58571D23D5C4B3A, 0x00003FFD //A3
data8 0xA94DCF467CD6AFF3, 0x0000BFFC //A2
data8 0xA94DCF467CD10A16, 0x00003FFA //A1
data8 0xFECD70A13CAF1997, 0x00003FFE //A0
// Polynomial coefficients for the erf(x), 4.0 <= |x| < 6.0
data8 0xB01D2B4F0D5AB8B0, 0x00003FF1 //A3
data8 0x8858A465CE594BD1, 0x0000BFEE //A2
data8 0x8858A447456DE61D, 0x00003FEA //A1
data8 0xFFFFFFBDC88BB107, 0x00003FFE //A0
// Polynomial coefficients for the erf(x), 0.0 <= |x| < 0.5
data8 0xBE839EDBB36C7FCE //A9
data8 0x3EBB7745A18DD242 //A8
data8 0xBF4C02DB238F2AFC //A5
data8 0x3F7565BCD0A9A3EA //A4
data8 0xC093A3581BCF3333, 0x0000BFFD //A1
data8 0xBEEF4BB82AD8AE22 //A7
data8 0x3F1F9A2A57A218CD //A6
data8 0xBF9B82CE3127F4E4 //A3
data8 0x3FBCE2F21A042B25 //A2
data8 0x906EBA8214DB688D, 0x00003FFF //A0
// 1.0 - 2^(-63)
data8 0xFFFFFFFFFFFFFFFF, 0x00003FFE
// Polynomial coefficients for the erf(x), 3.25 <= |x| < 4.0
data8 0x95E91576C7A12250, 0x00003FE7 //A14
data8 0x8E5E0D0E1F5D3CB5, 0x0000BFEA //A13
data8 0xED761DAFAF814DE9, 0x00003FEB //A12
data8 0xB3A77D921D0ACFC7, 0x0000BFEC //A11
data8 0xA662D27096B08D7C, 0x0000BFEC //A10
data8 0xDA0F410AE6233EA5, 0x00003FEF //A9
data8 0xAB4A8B16B3124327, 0x0000BFF1 //A8
data8 0xB241E236A5EDCED3, 0x00003FF2 //A7
data8 0x8A2A65BA1F551F77, 0x0000BFF3 //A6
data8 0xA4852D0B1D87000A, 0x00003FF3 //A5
data8 0x963EB00039489476, 0x0000BFF3 //A4
data8 0xCD5244FF4F7313A5, 0x00003FF2 //A3
data8 0xC6F1E695363BCB26, 0x0000BFF1 //A2
data8 0xF4DAF4680DA54C02, 0x00003FEF //A1
data8 0xFFFFB7CFB3F2ABBE, 0x00003FFE //A0
// A = 2.0/sqrt(Pi)
data8 0x906EBA8214DB688D, 0x00003FFF
LOCAL_OBJECT_END(erf_data)
.section .text
GLOBAL_LIBM_ENTRY(erf)
{ .mfi
alloc r32 = ar.pfs, 0, 17, 0, 0
fmerge.se fArgAbsNorm = f1, f8 // normalized x
adds rSignBit = 0x1, r0
}
{ .mfi
addl rDataPtr = @ltoff(erf_data), gp
fma.s1 fArgSqr = f8, f8, f0 // x^2
addl rThreeAndQ = 0x400A0, r0 // shifted bits of 3.25
}
;;
{ .mfi
getf.d rArg = f8 // x in GR
fclass.m p6,p0 = f8, 0x0b // is x denormal ?
shl rThreeAndQ = rThreeAndQ, 44 // bits of 3.25
}
{ .mfi
ld8 rDataPtr = [rDataPtr]
nop.f 0
addl rBiasedExpOf4 = 0x40100, r0 // shifted bits of 4.0
}
;;
{ .mfi
addl rSaturation = 0x4017A, r0 // shifted bits of 5.90625
fclass.m p7,p0 = f8, 0xc7 // is x [S,Q]NaN or +/-0 ?
shl rSignBit = rSignBit, 63 // mask for sign bit
}
{ .mfi
addl rMask = 0x7FF00, r0 // Mask for index bits
nop.f 0
addl rBias = 0x3FE00, r0 // bias of 0.5 << 8
}
;;
{ .mfi
setf.d fThreeAndQ = rThreeAndQ // 3.25 if FP register
fclass.m p9,p0 = f8, 0x23 // is x +/- inf?
shr.u rShiftedArg = rArg, 44
}
{ .mfb
andcm rAbsArg = rArg, rSignBit // |x| in GR
nop.f 0
(p6) br.cond.spnt erf_denormal // branch out if x is denormal
}
;;
{ .mfi
and rShiftedArgMasked = rShiftedArg, rMask // bias of x << 8
fmerge.s fArgAbs = f1, f8 // |x|
shr rShiftedAbsArg = rAbsArg, 44
}
{ .mfb
cmp.lt p8, p11 = rThreeAndQ, rAbsArg // p8 = 1 if |x| >= 3.25
(p7) fma.d.s0 f8 = f8,f1,f8 // NaN or +/-0
(p7) br.ret.spnt b0 // exit for x = NaN or +/-0
}
;;
{ .mfi
sub rIndex = rShiftedArgMasked, rBias // index << 8
nop.f 0
cmp.lt p10, p0 = rShiftedArgMasked, rBias // p10 = 1 if |x| < 0.5
}
{ .mfb
// p8 = 1 if 3.25 <= |x| < 4.0
(p8) cmp.lt p8, p11 = rShiftedAbsArg, rBiasedExpOf4
fms.s1 fArgAbsNorm = fArgAbsNorm, f1, f1
(p10) br.cond.spnt erf_near_zero // branch out if |x| < 0.5
}
;;
.pred.rel "mutex", p8, p11
{ .mfi
(p8) adds rCoeffAddr1 = 1392, rDataPtr // coeff. for 3.25 <=|x|<4.0
(p9) fmerge.s f8 = f8,f1 // +/- inf
nop.i 0
}
{ .mfb
(p11) add rCoeffAddr1 = rDataPtr, rIndex// coeff. ##0,2,..14
nop.f 0
(p9) br.ret.spnt b0 // exit for x = +/- inf
}
;;
{ .mfi
adds rCoeffAddr2 = 16, rCoeffAddr1
fmerge.s fSignumX = f8, f1 // signum(x)
nop.i 0
}
{ .mfb
cmp.lt p12, p0 = rSaturation, rShiftedAbsArg // |x| > 5.90625?
nop.f 0
(p12) br.cond.spnt erf_saturation // branch out if x |x| >= 6.0
}
;;
// Here if paths #3,4
// if path #4 we'll branch out after loading of 14 necessary coefficients
{.mfi
ldfe fA19 = [rCoeffAddr1], 32
nop.f 0
nop.i 0
}
{.mfi
ldfe fA18 = [rCoeffAddr2], 32
nop.f 0
adds rCoeffAddr3 = 1024, rDataPtr
}
;;
{.mfi
ldfe fA17 = [rCoeffAddr1], 32
nop.f 0
nop.i 0
}
{.mfi
ldfe fA16 = [rCoeffAddr2], 32
nop.f 0
nop.i 0
}
;;
{.mfi
ldfe fA15 = [rCoeffAddr1], 32
fma.s1 fTSqr = fArgAbsNorm, fArgAbsNorm, f0
shr.u rIndex = rIndex, 2
}
{.mfi
ldfe fA14 = [rCoeffAddr2], 32
nop.f 0
adds rCoeffAddr4 = 16, r0
}
;;
{.mfi
ldfe fA13 = [rCoeffAddr1], 32
nop.f 0
// address of coefficients ##16..23
add rCoeffAddr3 = rCoeffAddr3, rIndex
}
{.mfi
ldfe fA12 = [rCoeffAddr2], 32
nop.f 0
cmp.lt p15, p14 = rArg, r0
}
;;
{.mfi
ldfe fA11 = [rCoeffAddr1], 32
nop.f 0
add rCoeffAddr4 = rCoeffAddr3, rCoeffAddr4
}
{.mfi
ldfe fA10 = [rCoeffAddr2], 32
nop.f 0
nop.i 0
}
;;
{.mfi
ldfe fA9 = [rCoeffAddr1], 32
nop.f 0
nop.i 0
}
{.mfi
ldfe fA8 = [rCoeffAddr2], 32
nop.f 0
nop.i 0
}
;;
{.mfi
ldfe fA7 = [rCoeffAddr1], 32
fms.s1 fArgAbs = fArgAbs, f1, fThreeAndQ
nop.i 0
}
{.mfb
ldfe fA6 = [rCoeffAddr2], 32
nop.f 0
(p8) br.cond.spnt erf_3q_4 // branch out if 3.25 < |x| < 4.0
}
;;
{.mfi
ldfe fA5 = [rCoeffAddr1], 32
fma.s1 fTDeg3 = fArgAbsNorm, fTSqr, f0
nop.i 0
}
{.mfi
ldfe fA4 = [rCoeffAddr2], 32
fma.s1 fTQuadr = fTSqr, fTSqr, f0
nop.i 0
}
;;
// Path #3 Polynomial Pol19(y) computation; y = fArgAbsNorm
{.mfi
ldfe fA3 = [rCoeffAddr3], 32
fma.s1 fArgAbsNormSgn = fArgAbsNorm, fSignumX, f0
nop.i 0
}
{.mfi
ldfe fA2 = [rCoeffAddr4], 32
nop.f 0
nop.i 0
}
;;
{.mfi
ldfe fA1 = [rCoeffAddr3], 32
fma.s1 fRes = fA19, fArgAbsNorm, fA18
nop.i 0
}
{.mfi
ldfe fA0 = [rCoeffAddr4], 32
nop.f 0
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA17 = fA17, fArgAbsNorm, fA16
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA15 = fA15, fArgAbsNorm, fA14
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fTDeg7 = fTDeg3, fTQuadr, f0
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fA13 = fA13, fArgAbsNorm, fA12
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA11 = fA11, fArgAbsNorm, fA10
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA9 = fA9, fArgAbsNorm, fA8
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fRes = fRes, fTSqr, fA17
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fA7 = fA7, fArgAbsNorm, fA6
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA5 = fA5, fArgAbsNorm, f0
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA15 = fA15, fTSqr, fA13
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fA4 = fA4, fArgAbsNorm, fA3
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA2 = fA2, fArgAbsNorm, fA1
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA11 = fA11, fTSqr, fA9
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA7 = fA7, fTSqr, fA5
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fRes = fRes, fTQuadr, fA15
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA4 = fA4, fTSqr, fA2
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fRes = fRes, fTQuadr, fA11
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA4 = fA7, fTDeg3, fA4
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fRes = fRes, fTDeg7, fA4
nop.i 0
}
;;
{ .mfi
nop.m 0
// result for negative argument
(p15) fms.d.s0 f8 = fRes, fArgAbsNormSgn, fA0
nop.i 0
}
{ .mfb
nop.m 0
// result for positive argument
(p14) fma.d.s0 f8 = fRes, fArgAbsNormSgn, fA0
br.ret.sptk b0
}
// Here if 3.25 < |x| < 4.0
.align 32
erf_3q_4:
.pred.rel "mutex", p14, p15
{ .mfi
ldfe fA5 = [rCoeffAddr1], 32
fma.s1 fTSqr = fArgAbs, fArgAbs, f0
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fRes = fA19, fArgAbs, fA18
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA17 = fA17, fArgAbs, fA16
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fA15 = fA15, fArgAbs, fA14
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA13 = fA13, fArgAbs, fA12
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fA11 = fA11, fArgAbs, fA10
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA9 = fA9, fArgAbs, fA8
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fArgAbsNormSgn = fArgAbs, fSignumX, f0
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fTQuadr = fTSqr, fTSqr, f0
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fRes = fRes, fTSqr, fA17
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA15 = fA15, fTSqr, fA13
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA11 = fA11, fTSqr, fA9
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fA7 = fA7, fArgAbs, fA6
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fTDeg7 = fTQuadr, fTSqr, f0
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fRes = fRes, fTQuadr, fA15
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA11 = fA11, fTSqr, fA7
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fRes = fRes, fTDeg7, fA11
nop.i 0
}
;;
{ .mfi
nop.m 0
// result for negative argument
(p15) fms.d.s0 f8 = fRes, fArgAbsNormSgn, fA5
nop.i 0
}
{ .mfb
nop.m 0
// result for positive argument
(p14) fma.d.s0 f8 = fRes, fArgAbsNormSgn, fA5
br.ret.sptk b0
}
;;
// Here if |x| < 0.5
.align 32
erf_near_zero:
{ .mfi
adds rCoeffAddr1 = 1280, rDataPtr // address of A9
fma.s1 fTSqr = fArgSqr, fArgSqr, f0 // x^4
nop.i 0
}
{ .mfi
adds rCoeffAddr2 = 1328, rDataPtr // address of A7
nop.f 0
nop.i 0
}
;;
{ .mfi
ldfpd fA9, fA8 = [rCoeffAddr1], 16
nop.f 0
nop.i 0
}
{ .mfi
ldfpd fA7, fA6 = [rCoeffAddr2], 16
nop.f 0
nop.i 0
}
;;
{ .mfi
ldfpd fA5, fA4 = [rCoeffAddr1], 16
nop.f 0
nop.i 0
}
{ .mfi
ldfpd fA3, fA2 = [rCoeffAddr2], 16
nop.f 0
nop.i 0
}
;;
{ .mfi
ldfe fA1 = [rCoeffAddr1]
nop.f 0
nop.i 0
}
{ .mfi
ldfe fA0 = [rCoeffAddr2]
nop.f 0
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fTQuadr = fTSqr, fTSqr, f0
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fRes = fA9, fArgSqr, fA8
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fA7 = fA7, fArgSqr, fA6
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA3 = fA3, fArgSqr, fA2
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fA5 = fA5, fArgSqr, fA4
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA1 = fA1, fArgSqr, fA0
nop.i 0
}
{ .mfi
nop.m 0
fma.s1 fTQuadrSgn = fTQuadr, f8, f0
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fRes = fRes, fTSqr, fA7
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA1 = fA3, fTSqr, fA1
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fRes = fRes, fTSqr, fA5
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA1 = fA1, f8, f0
nop.i 0
}
;;
{ .mfb
nop.m 0
fma.d.s0 f8 = fRes, fTQuadrSgn, fA1 // x*Pol9(x^2)
br.ret.sptk b0 // Exit for |x| < 0.5
};;
// Here if 5.90625 <= |x| < +inf
.align 32
erf_saturation:
{ .mfi
adds rDataPtr = 1376, rDataPtr // address of A0
nop.f 0
nop.i 0
}
;;
{ .mfi
ldfe fA0 = [rDataPtr]
nop.f 0
nop.i 0
}
;;
{ .mfb
nop.m 0
fma.d.s0 f8 = fA0, fSignumX, f0 // sign(x)*(1.0 - 2^(-63))
// Exit for 5.90625 <= |x| < +inf
br.ret.sptk b0 // Exit for 5.90625 <=|x|< +inf
}
;;
// Here if x is double precision denormal
.align 32
erf_denormal:
{ .mfi
adds rDataPtr = 1632, rDataPtr // address of A0
fclass.m p7,p8 = f8, 0x0a // is x -denormal ?
nop.i 0
}
;;
{ .mfi
ldfe fA0 = [rDataPtr] // A0
nop.f 0
nop.i 0
}
;;
{ .mfi
nop.m 0
fma.s1 fA0 = fA0,f8,f0 // A0*x
nop.i 0
}
;;
{ .mfi
nop.m 0
(p7) fma.d.s0 f8 = f8,f8,fA0 // -denormal
nop.i 0
}
{ .mfb
nop.m 0
(p8) fnma.d.s0 f8 = f8,f8,fA0 // +denormal
br.ret.sptk b0 // Exit for denormal
}
;;
GLOBAL_LIBM_END(erf)
|
