forked from micropython/micropython
-
Notifications
You must be signed in to change notification settings - Fork 4
/
mpz.c
1752 lines (1503 loc) · 47 KB
/
mpz.c
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
/*
* This file is part of the MicroPython project, http://micropython.org/
*
* The MIT License (MIT)
*
* Copyright (c) 2013, 2014 Damien P. George
*
* 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 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.
*/
#include <string.h>
#include <assert.h>
#include "py/mpz.h"
#if MICROPY_LONGINT_IMPL == MICROPY_LONGINT_IMPL_MPZ
#define DIG_SIZE (MPZ_DIG_SIZE)
#define DIG_MASK ((MPZ_LONG_1 << DIG_SIZE) - 1)
#define DIG_MSB (MPZ_LONG_1 << (DIG_SIZE - 1))
#define DIG_BASE (MPZ_LONG_1 << DIG_SIZE)
/*
mpz is an arbitrary precision integer type with a public API.
mpn functions act on non-negative integers represented by an array of generalised
digits (eg a word per digit). You also need to specify separately the length of the
array. There is no public API for mpn. Rather, the functions are used by mpz to
implement its features.
Integer values are stored little endian (first digit is first in memory).
Definition of normalise: ?
*/
static size_t mpn_remove_trailing_zeros(mpz_dig_t *oidig, mpz_dig_t *idig) {
for (--idig; idig >= oidig && *idig == 0; --idig) {
}
return idig + 1 - oidig;
}
/* compares i with j
returns sign(i - j)
assumes i, j are normalised
*/
static int mpn_cmp(const mpz_dig_t *idig, size_t ilen, const mpz_dig_t *jdig, size_t jlen) {
if (ilen < jlen) {
return -1;
}
if (ilen > jlen) {
return 1;
}
for (idig += ilen, jdig += ilen; ilen > 0; --ilen) {
mpz_dbl_dig_signed_t cmp = (mpz_dbl_dig_t)*(--idig) - (mpz_dbl_dig_t)*(--jdig);
if (cmp < 0) {
return -1;
}
if (cmp > 0) {
return 1;
}
}
return 0;
}
/* computes i = j << n
returns number of digits in i
assumes enough memory in i; assumes normalised j; assumes n > 0
can have i, j pointing to same memory
*/
static size_t mpn_shl(mpz_dig_t *idig, mpz_dig_t *jdig, size_t jlen, mp_uint_t n) {
mp_uint_t n_whole = (n + DIG_SIZE - 1) / DIG_SIZE;
mp_uint_t n_part = n % DIG_SIZE;
if (n_part == 0) {
n_part = DIG_SIZE;
}
// start from the high end of the digit arrays
idig += jlen + n_whole - 1;
jdig += jlen - 1;
// shift the digits
mpz_dbl_dig_t d = 0;
for (size_t i = jlen; i > 0; i--, idig--, jdig--) {
d |= *jdig;
*idig = (d >> (DIG_SIZE - n_part)) & DIG_MASK;
d <<= DIG_SIZE;
}
// store remaining bits
*idig = (d >> (DIG_SIZE - n_part)) & DIG_MASK;
idig -= n_whole - 1;
memset(idig, 0, (n_whole - 1) * sizeof(mpz_dig_t));
// work out length of result
jlen += n_whole;
while (jlen != 0 && idig[jlen - 1] == 0) {
jlen--;
}
// return length of result
return jlen;
}
/* computes i = j >> n
returns number of digits in i
assumes enough memory in i; assumes normalised j; assumes n > 0
can have i, j pointing to same memory
*/
static size_t mpn_shr(mpz_dig_t *idig, mpz_dig_t *jdig, size_t jlen, mp_uint_t n) {
mp_uint_t n_whole = n / DIG_SIZE;
mp_uint_t n_part = n % DIG_SIZE;
if (n_whole >= jlen) {
return 0;
}
jdig += n_whole;
jlen -= n_whole;
for (size_t i = jlen; i > 0; i--, idig++, jdig++) {
mpz_dbl_dig_t d = *jdig;
if (i > 1) {
d |= (mpz_dbl_dig_t)jdig[1] << DIG_SIZE;
}
d >>= n_part;
*idig = d & DIG_MASK;
}
if (idig[-1] == 0) {
jlen--;
}
return jlen;
}
/* computes i = j + k
returns number of digits in i
assumes enough memory in i; assumes normalised j, k; assumes jlen >= klen
can have i, j, k pointing to same memory
*/
static size_t mpn_add(mpz_dig_t *idig, const mpz_dig_t *jdig, size_t jlen, const mpz_dig_t *kdig, size_t klen) {
mpz_dig_t *oidig = idig;
mpz_dbl_dig_t carry = 0;
jlen -= klen;
for (; klen > 0; --klen, ++idig, ++jdig, ++kdig) {
carry += (mpz_dbl_dig_t)*jdig + (mpz_dbl_dig_t)*kdig;
*idig = carry & DIG_MASK;
carry >>= DIG_SIZE;
}
for (; jlen > 0; --jlen, ++idig, ++jdig) {
carry += *jdig;
*idig = carry & DIG_MASK;
carry >>= DIG_SIZE;
}
if (carry != 0) {
*idig++ = carry;
}
return idig - oidig;
}
/* computes i = j - k
returns number of digits in i
assumes enough memory in i; assumes normalised j, k; assumes j >= k
can have i, j, k pointing to same memory
*/
static size_t mpn_sub(mpz_dig_t *idig, const mpz_dig_t *jdig, size_t jlen, const mpz_dig_t *kdig, size_t klen) {
mpz_dig_t *oidig = idig;
mpz_dbl_dig_signed_t borrow = 0;
jlen -= klen;
for (; klen > 0; --klen, ++idig, ++jdig, ++kdig) {
borrow += (mpz_dbl_dig_t)*jdig - (mpz_dbl_dig_t)*kdig;
*idig = borrow & DIG_MASK;
borrow >>= DIG_SIZE;
}
for (; jlen > 0; --jlen, ++idig, ++jdig) {
borrow += *jdig;
*idig = borrow & DIG_MASK;
borrow >>= DIG_SIZE;
}
return mpn_remove_trailing_zeros(oidig, idig);
}
#if MICROPY_OPT_MPZ_BITWISE
/* computes i = j & k
returns number of digits in i
assumes enough memory in i; assumes normalised j, k; assumes jlen >= klen (jlen argument not needed)
can have i, j, k pointing to same memory
*/
static size_t mpn_and(mpz_dig_t *idig, const mpz_dig_t *jdig, const mpz_dig_t *kdig, size_t klen) {
mpz_dig_t *oidig = idig;
for (; klen > 0; --klen, ++idig, ++jdig, ++kdig) {
*idig = *jdig & *kdig;
}
return mpn_remove_trailing_zeros(oidig, idig);
}
#endif
/* i = -((-j) & (-k)) = ~((~j + 1) & (~k + 1)) + 1
i = (j & (-k)) = (j & (~k + 1)) = ( j & (~k + 1))
i = ((-j) & k) = ((~j + 1) & k) = ((~j + 1) & k )
computes general form:
i = (im ^ (((j ^ jm) + jc) & ((k ^ km) + kc))) + ic where Xm = Xc == 0 ? 0 : DIG_MASK
returns number of digits in i
assumes enough memory in i; assumes normalised j, k; assumes length j >= length k
can have i, j, k pointing to same memory
*/
static size_t mpn_and_neg(mpz_dig_t *idig, const mpz_dig_t *jdig, size_t jlen, const mpz_dig_t *kdig, size_t klen,
mpz_dbl_dig_t carryi, mpz_dbl_dig_t carryj, mpz_dbl_dig_t carryk) {
mpz_dig_t *oidig = idig;
mpz_dig_t imask = (0 == carryi) ? 0 : DIG_MASK;
mpz_dig_t jmask = (0 == carryj) ? 0 : DIG_MASK;
mpz_dig_t kmask = (0 == carryk) ? 0 : DIG_MASK;
for (; jlen > 0; ++idig, ++jdig) {
carryj += *jdig ^ jmask;
carryk += (--klen <= --jlen) ? (*kdig++ ^ kmask) : kmask;
carryi += ((carryj & carryk) ^ imask) & DIG_MASK;
*idig = carryi & DIG_MASK;
carryk >>= DIG_SIZE;
carryj >>= DIG_SIZE;
carryi >>= DIG_SIZE;
}
if (0 != carryi) {
*idig++ = carryi;
}
return mpn_remove_trailing_zeros(oidig, idig);
}
#if MICROPY_OPT_MPZ_BITWISE
/* computes i = j | k
returns number of digits in i
assumes enough memory in i; assumes normalised j, k; assumes jlen >= klen
can have i, j, k pointing to same memory
*/
static size_t mpn_or(mpz_dig_t *idig, const mpz_dig_t *jdig, size_t jlen, const mpz_dig_t *kdig, size_t klen) {
mpz_dig_t *oidig = idig;
jlen -= klen;
for (; klen > 0; --klen, ++idig, ++jdig, ++kdig) {
*idig = *jdig | *kdig;
}
for (; jlen > 0; --jlen, ++idig, ++jdig) {
*idig = *jdig;
}
return idig - oidig;
}
#endif
/* i = -((-j) | (-k)) = ~((~j + 1) | (~k + 1)) + 1
i = -(j | (-k)) = -(j | (~k + 1)) = ~( j | (~k + 1)) + 1
i = -((-j) | k) = -((~j + 1) | k) = ~((~j + 1) | k ) + 1
computes general form:
i = ~(((j ^ jm) + jc) | ((k ^ km) + kc)) + 1 where Xm = Xc == 0 ? 0 : DIG_MASK
returns number of digits in i
assumes enough memory in i; assumes normalised j, k; assumes length j >= length k
can have i, j, k pointing to same memory
*/
#if MICROPY_OPT_MPZ_BITWISE
static size_t mpn_or_neg(mpz_dig_t *idig, const mpz_dig_t *jdig, size_t jlen, const mpz_dig_t *kdig, size_t klen,
mpz_dbl_dig_t carryj, mpz_dbl_dig_t carryk) {
mpz_dig_t *oidig = idig;
mpz_dbl_dig_t carryi = 1;
mpz_dig_t jmask = (0 == carryj) ? 0 : DIG_MASK;
mpz_dig_t kmask = (0 == carryk) ? 0 : DIG_MASK;
for (; jlen > 0; ++idig, ++jdig) {
carryj += *jdig ^ jmask;
carryk += (--klen <= --jlen) ? (*kdig++ ^ kmask) : kmask;
carryi += ((carryj | carryk) ^ DIG_MASK) & DIG_MASK;
*idig = carryi & DIG_MASK;
carryk >>= DIG_SIZE;
carryj >>= DIG_SIZE;
carryi >>= DIG_SIZE;
}
// At least one of j,k must be negative so the above for-loop runs at least
// once. For carryi to be non-zero here it must be equal to 1 at the end of
// each iteration of the loop. So the accumulation of carryi must overflow
// each time, ie carryi += 0xff..ff. So carryj|carryk must be 0 in the
// DIG_MASK bits on each iteration. But considering all cases of signs of
// j,k one sees that this is not possible.
assert(carryi == 0);
return mpn_remove_trailing_zeros(oidig, idig);
}
#else
static size_t mpn_or_neg(mpz_dig_t *idig, const mpz_dig_t *jdig, size_t jlen, const mpz_dig_t *kdig, size_t klen,
mpz_dbl_dig_t carryi, mpz_dbl_dig_t carryj, mpz_dbl_dig_t carryk) {
mpz_dig_t *oidig = idig;
mpz_dig_t imask = (0 == carryi) ? 0 : DIG_MASK;
mpz_dig_t jmask = (0 == carryj) ? 0 : DIG_MASK;
mpz_dig_t kmask = (0 == carryk) ? 0 : DIG_MASK;
for (; jlen > 0; ++idig, ++jdig) {
carryj += *jdig ^ jmask;
carryk += (--klen <= --jlen) ? (*kdig++ ^ kmask) : kmask;
carryi += ((carryj | carryk) ^ imask) & DIG_MASK;
*idig = carryi & DIG_MASK;
carryk >>= DIG_SIZE;
carryj >>= DIG_SIZE;
carryi >>= DIG_SIZE;
}
// See comment in above mpn_or_neg for why carryi must be 0.
assert(carryi == 0);
return mpn_remove_trailing_zeros(oidig, idig);
}
#endif
#if MICROPY_OPT_MPZ_BITWISE
/* computes i = j ^ k
returns number of digits in i
assumes enough memory in i; assumes normalised j, k; assumes jlen >= klen
can have i, j, k pointing to same memory
*/
static size_t mpn_xor(mpz_dig_t *idig, const mpz_dig_t *jdig, size_t jlen, const mpz_dig_t *kdig, size_t klen) {
mpz_dig_t *oidig = idig;
jlen -= klen;
for (; klen > 0; --klen, ++idig, ++jdig, ++kdig) {
*idig = *jdig ^ *kdig;
}
for (; jlen > 0; --jlen, ++idig, ++jdig) {
*idig = *jdig;
}
return mpn_remove_trailing_zeros(oidig, idig);
}
#endif
/* i = (-j) ^ (-k) = ~(j - 1) ^ ~(k - 1) = (j - 1) ^ (k - 1)
i = -(j ^ (-k)) = -(j ^ ~(k - 1)) = ~(j ^ ~(k - 1)) + 1 = (j ^ (k - 1)) + 1
i = -((-j) ^ k) = -(~(j - 1) ^ k) = ~(~(j - 1) ^ k) + 1 = ((j - 1) ^ k) + 1
computes general form:
i = ((j - 1 + jc) ^ (k - 1 + kc)) + ic
returns number of digits in i
assumes enough memory in i; assumes normalised j, k; assumes length j >= length k
can have i, j, k pointing to same memory
*/
static size_t mpn_xor_neg(mpz_dig_t *idig, const mpz_dig_t *jdig, size_t jlen, const mpz_dig_t *kdig, size_t klen,
mpz_dbl_dig_t carryi, mpz_dbl_dig_t carryj, mpz_dbl_dig_t carryk) {
mpz_dig_t *oidig = idig;
for (; jlen > 0; ++idig, ++jdig) {
carryj += *jdig + DIG_MASK;
carryk += (--klen <= --jlen) ? (*kdig++ + DIG_MASK) : DIG_MASK;
carryi += (carryj ^ carryk) & DIG_MASK;
*idig = carryi & DIG_MASK;
carryk >>= DIG_SIZE;
carryj >>= DIG_SIZE;
carryi >>= DIG_SIZE;
}
if (0 != carryi) {
*idig++ = carryi;
}
return mpn_remove_trailing_zeros(oidig, idig);
}
/* computes i = i * d1 + d2
returns number of digits in i
assumes enough memory in i; assumes normalised i; assumes dmul != 0
*/
static size_t mpn_mul_dig_add_dig(mpz_dig_t *idig, size_t ilen, mpz_dig_t dmul, mpz_dig_t dadd) {
mpz_dig_t *oidig = idig;
mpz_dbl_dig_t carry = dadd;
for (; ilen > 0; --ilen, ++idig) {
carry += (mpz_dbl_dig_t)*idig * (mpz_dbl_dig_t)dmul; // will never overflow so long as DIG_SIZE <= 8*sizeof(mpz_dbl_dig_t)/2
*idig = carry & DIG_MASK;
carry >>= DIG_SIZE;
}
if (carry != 0) {
*idig++ = carry;
}
return idig - oidig;
}
/* computes i = j * k
returns number of digits in i
assumes enough memory in i; assumes i is zeroed; assumes normalised j, k
can have j, k point to same memory
*/
static size_t mpn_mul(mpz_dig_t *idig, mpz_dig_t *jdig, size_t jlen, mpz_dig_t *kdig, size_t klen) {
mpz_dig_t *oidig = idig;
size_t ilen = 0;
for (; klen > 0; --klen, ++idig, ++kdig) {
mpz_dig_t *id = idig;
mpz_dbl_dig_t carry = 0;
size_t jl = jlen;
for (mpz_dig_t *jd = jdig; jl > 0; --jl, ++jd, ++id) {
carry += (mpz_dbl_dig_t)*id + (mpz_dbl_dig_t)*jd * (mpz_dbl_dig_t)*kdig; // will never overflow so long as DIG_SIZE <= 8*sizeof(mpz_dbl_dig_t)/2
*id = carry & DIG_MASK;
carry >>= DIG_SIZE;
}
if (carry != 0) {
*id++ = carry;
}
ilen = id - oidig;
}
return ilen;
}
/* natural_div - quo * den + new_num = old_num (ie num is replaced with rem)
assumes den != 0
assumes num_dig has enough memory to be extended by 1 digit
assumes quo_dig has enough memory (as many digits as num)
assumes quo_dig is filled with zeros
*/
static void mpn_div(mpz_dig_t *num_dig, size_t *num_len, const mpz_dig_t *den_dig, size_t den_len, mpz_dig_t *quo_dig, size_t *quo_len) {
mpz_dig_t *orig_num_dig = num_dig;
mpz_dig_t *orig_quo_dig = quo_dig;
mpz_dig_t norm_shift = 0;
mpz_dbl_dig_t lead_den_digit;
// handle simple cases
{
int cmp = mpn_cmp(num_dig, *num_len, den_dig, den_len);
if (cmp == 0) {
*num_len = 0;
quo_dig[0] = 1;
*quo_len = 1;
return;
} else if (cmp < 0) {
// numerator remains the same
*quo_len = 0;
return;
}
}
// We need to normalise the denominator (leading bit of leading digit is 1)
// so that the division routine works. Since the denominator memory is
// read-only we do the normalisation on the fly, each time a digit of the
// denominator is needed. We need to know is how many bits to shift by.
// count number of leading zeros in leading digit of denominator
{
mpz_dig_t d = den_dig[den_len - 1];
while ((d & DIG_MSB) == 0) {
d <<= 1;
++norm_shift;
}
}
// now need to shift numerator by same amount as denominator
// first, increase length of numerator in case we need more room to shift
num_dig[*num_len] = 0;
++(*num_len);
for (mpz_dig_t *num = num_dig, carry = 0; num < num_dig + *num_len; ++num) {
mpz_dig_t n = *num;
*num = ((n << norm_shift) | carry) & DIG_MASK;
carry = (mpz_dbl_dig_t)n >> (DIG_SIZE - norm_shift);
}
// cache the leading digit of the denominator
lead_den_digit = (mpz_dbl_dig_t)den_dig[den_len - 1] << norm_shift;
if (den_len >= 2) {
lead_den_digit |= (mpz_dbl_dig_t)den_dig[den_len - 2] >> (DIG_SIZE - norm_shift);
}
// point num_dig to last digit in numerator
num_dig += *num_len - 1;
// calculate number of digits in quotient
*quo_len = *num_len - den_len;
// point to last digit to store for quotient
quo_dig += *quo_len - 1;
// keep going while we have enough digits to divide
while (*num_len > den_len) {
mpz_dbl_dig_t quo = ((mpz_dbl_dig_t)*num_dig << DIG_SIZE) | num_dig[-1];
// get approximate quotient
quo /= lead_den_digit;
// Multiply quo by den and subtract from num to get remainder.
// Must be careful with overflow of the borrow variable. Both
// borrow and low_digs are signed values and need signed right-shift,
// but x is unsigned and may take a full-range value.
const mpz_dig_t *d = den_dig;
mpz_dbl_dig_t d_norm = 0;
mpz_dbl_dig_signed_t borrow = 0;
for (mpz_dig_t *n = num_dig - den_len; n < num_dig; ++n, ++d) {
// Get the next digit in (den).
d_norm = ((mpz_dbl_dig_t)*d << norm_shift) | (d_norm >> DIG_SIZE);
// Multiply the next digit in (quo * den).
mpz_dbl_dig_t x = (mpz_dbl_dig_t)quo * (d_norm & DIG_MASK);
// Compute the low DIG_MASK bits of the next digit in (num - quo * den)
mpz_dbl_dig_signed_t low_digs = (borrow & DIG_MASK) + *n - (x & DIG_MASK);
// Store the digit result for (num).
*n = low_digs & DIG_MASK;
// Compute the borrow, shifted right before summing to avoid overflow.
borrow = (borrow >> DIG_SIZE) - (x >> DIG_SIZE) + (low_digs >> DIG_SIZE);
}
// At this point we have either:
//
// 1. quo was the correct value and the most-sig-digit of num is exactly
// cancelled by borrow (borrow + *num_dig == 0). In this case there is
// nothing more to do.
//
// 2. quo was too large, we subtracted too many den from num, and the
// most-sig-digit of num is less than needed (borrow + *num_dig < 0).
// In this case we must reduce quo and add back den to num until the
// carry from this operation cancels out the borrow.
//
borrow += *num_dig;
for (; borrow != 0; --quo) {
d = den_dig;
d_norm = 0;
mpz_dbl_dig_t carry = 0;
for (mpz_dig_t *n = num_dig - den_len; n < num_dig; ++n, ++d) {
d_norm = ((mpz_dbl_dig_t)*d << norm_shift) | (d_norm >> DIG_SIZE);
carry += (mpz_dbl_dig_t)*n + (d_norm & DIG_MASK);
*n = carry & DIG_MASK;
carry >>= DIG_SIZE;
}
borrow += carry;
}
// store this digit of the quotient
*quo_dig = quo & DIG_MASK;
--quo_dig;
// move down to next digit of numerator
--num_dig;
--(*num_len);
}
// unnormalise numerator (remainder now)
for (mpz_dig_t *num = orig_num_dig + *num_len - 1, carry = 0; num >= orig_num_dig; --num) {
mpz_dig_t n = *num;
*num = ((n >> norm_shift) | carry) & DIG_MASK;
carry = (mpz_dbl_dig_t)n << (DIG_SIZE - norm_shift);
}
// strip trailing zeros
while (*quo_len > 0 && orig_quo_dig[*quo_len - 1] == 0) {
--(*quo_len);
}
while (*num_len > 0 && orig_num_dig[*num_len - 1] == 0) {
--(*num_len);
}
}
#define MIN_ALLOC (2)
void mpz_init_zero(mpz_t *z) {
z->neg = 0;
z->fixed_dig = 0;
z->alloc = 0;
z->len = 0;
z->dig = NULL;
}
void mpz_init_from_int(mpz_t *z, mp_int_t val) {
mpz_init_zero(z);
mpz_set_from_int(z, val);
}
void mpz_init_fixed_from_int(mpz_t *z, mpz_dig_t *dig, size_t alloc, mp_int_t val) {
z->neg = 0;
z->fixed_dig = 1;
z->alloc = alloc;
z->len = 0;
z->dig = dig;
mpz_set_from_int(z, val);
}
void mpz_deinit(mpz_t *z) {
if (z != NULL && !z->fixed_dig) {
m_del(mpz_dig_t, z->dig, z->alloc);
}
}
#if 0
these functions are unused
mpz_t *mpz_zero(void) {
mpz_t *z = m_new_obj(mpz_t);
mpz_init_zero(z);
return z;
}
mpz_t *mpz_from_int(mp_int_t val) {
mpz_t *z = mpz_zero();
mpz_set_from_int(z, val);
return z;
}
mpz_t *mpz_from_ll(long long val, bool is_signed) {
mpz_t *z = mpz_zero();
mpz_set_from_ll(z, val, is_signed);
return z;
}
#if MICROPY_PY_BUILTINS_FLOAT
mpz_t *mpz_from_float(mp_float_t val) {
mpz_t *z = mpz_zero();
mpz_set_from_float(z, val);
return z;
}
#endif
mpz_t *mpz_from_str(const char *str, size_t len, bool neg, unsigned int base) {
mpz_t *z = mpz_zero();
mpz_set_from_str(z, str, len, neg, base);
return z;
}
#endif
static void mpz_free(mpz_t *z) {
if (z != NULL) {
m_del(mpz_dig_t, z->dig, z->alloc);
m_del_obj(mpz_t, z);
}
}
static void mpz_need_dig(mpz_t *z, size_t need) {
if (need < MIN_ALLOC) {
need = MIN_ALLOC;
}
if (z->dig == NULL || z->alloc < need) {
// if z has fixed digit buffer there's not much we can do as the caller will
// be expecting a buffer with at least "need" bytes (but it shouldn't happen)
assert(!z->fixed_dig);
z->dig = m_renew(mpz_dig_t, z->dig, z->alloc, need);
z->alloc = need;
}
}
static mpz_t *mpz_clone(const mpz_t *src) {
assert(src->alloc != 0);
mpz_t *z = m_new_obj(mpz_t);
z->neg = src->neg;
z->fixed_dig = 0;
z->alloc = src->alloc;
z->len = src->len;
z->dig = m_new(mpz_dig_t, z->alloc);
memcpy(z->dig, src->dig, src->alloc * sizeof(mpz_dig_t));
return z;
}
/* sets dest = src
can have dest, src the same
*/
void mpz_set(mpz_t *dest, const mpz_t *src) {
mpz_need_dig(dest, src->len);
dest->neg = src->neg;
dest->len = src->len;
memcpy(dest->dig, src->dig, src->len * sizeof(mpz_dig_t));
}
void mpz_set_from_int(mpz_t *z, mp_int_t val) {
if (val == 0) {
z->neg = 0;
z->len = 0;
return;
}
mpz_need_dig(z, MPZ_NUM_DIG_FOR_INT);
mp_uint_t uval;
if (val < 0) {
z->neg = 1;
uval = -val;
} else {
z->neg = 0;
uval = val;
}
z->len = 0;
while (uval > 0) {
z->dig[z->len++] = uval & DIG_MASK;
uval >>= DIG_SIZE;
}
}
void mpz_set_from_ll(mpz_t *z, long long val, bool is_signed) {
mpz_need_dig(z, MPZ_NUM_DIG_FOR_LL);
unsigned long long uval;
if (is_signed && val < 0) {
z->neg = 1;
uval = -(unsigned long long)val;
} else {
z->neg = 0;
uval = val;
}
z->len = 0;
while (uval > 0) {
z->dig[z->len++] = uval & DIG_MASK;
uval >>= DIG_SIZE;
}
}
#if MICROPY_PY_BUILTINS_FLOAT
void mpz_set_from_float(mpz_t *z, mp_float_t src) {
mp_float_union_t u = {src};
z->neg = u.p.sgn;
if (u.p.exp == 0) {
// value == 0 || value < 1
mpz_set_from_int(z, 0);
} else if (u.p.exp == ((1 << MP_FLOAT_EXP_BITS) - 1)) {
// u.p.frc == 0 indicates inf, else NaN
// should be handled by caller
mpz_set_from_int(z, 0);
} else {
const int adj_exp = (int)u.p.exp - MP_FLOAT_EXP_BIAS;
if (adj_exp < 0) {
// value < 1 , truncates to 0
mpz_set_from_int(z, 0);
} else if (adj_exp == 0) {
// 1 <= value < 2 , so truncates to 1
mpz_set_from_int(z, 1);
} else {
// 2 <= value
const int dig_cnt = (adj_exp + 1 + (DIG_SIZE - 1)) / DIG_SIZE;
const unsigned int rem = adj_exp % DIG_SIZE;
int dig_ind, shft;
mp_float_uint_t frc = u.p.frc | ((mp_float_uint_t)1 << MP_FLOAT_FRAC_BITS);
if (adj_exp < MP_FLOAT_FRAC_BITS) {
shft = 0;
dig_ind = 0;
frc >>= MP_FLOAT_FRAC_BITS - adj_exp;
} else {
shft = (rem - MP_FLOAT_FRAC_BITS) % DIG_SIZE;
dig_ind = (adj_exp - MP_FLOAT_FRAC_BITS) / DIG_SIZE;
}
mpz_need_dig(z, dig_cnt);
z->len = dig_cnt;
if (dig_ind != 0) {
memset(z->dig, 0, dig_ind * sizeof(mpz_dig_t));
}
if (shft != 0) {
z->dig[dig_ind++] = (frc << shft) & DIG_MASK;
frc >>= DIG_SIZE - shft;
}
#if DIG_SIZE < (MP_FLOAT_FRAC_BITS + 1)
while (dig_ind != dig_cnt) {
z->dig[dig_ind++] = frc & DIG_MASK;
frc >>= DIG_SIZE;
}
#else
if (dig_ind != dig_cnt) {
z->dig[dig_ind] = frc;
}
#endif
}
}
}
#endif
// returns number of bytes from str that were processed
size_t mpz_set_from_str(mpz_t *z, const char *str, size_t len, bool neg, unsigned int base) {
assert(base <= 36);
const char *cur = str;
const char *top = str + len;
mpz_need_dig(z, len * 8 / DIG_SIZE + 1);
if (neg) {
z->neg = 1;
} else {
z->neg = 0;
}
z->len = 0;
for (; cur < top; ++cur) { // XXX UTF8 next char
// mp_uint_t v = char_to_numeric(cur#); // XXX UTF8 get char
mp_uint_t v = *cur;
if ('0' <= v && v <= '9') {
v -= '0';
} else if ('A' <= v && v <= 'Z') {
v -= 'A' - 10;
} else if ('a' <= v && v <= 'z') {
v -= 'a' - 10;
} else {
break;
}
if (v >= base) {
break;
}
z->len = mpn_mul_dig_add_dig(z->dig, z->len, base, v);
}
return cur - str;
}
void mpz_set_from_bytes(mpz_t *z, bool big_endian, size_t len, const byte *buf) {
int delta = 1;
if (big_endian) {
buf += len - 1;
delta = -1;
}
mpz_need_dig(z, (len * 8 + DIG_SIZE - 1) / DIG_SIZE);
mpz_dig_t d = 0;
int num_bits = 0;
z->neg = 0;
z->len = 0;
while (len) {
while (len && num_bits < DIG_SIZE) {
d |= *buf << num_bits;
num_bits += 8;
buf += delta;
len--;
}
z->dig[z->len++] = d & DIG_MASK;
// Need this #if because it's C undefined behavior to do: uint32_t >> 32
#if DIG_SIZE != 8 && DIG_SIZE != 16 && DIG_SIZE != 32
d >>= DIG_SIZE;
#else
d = 0;
#endif
num_bits -= DIG_SIZE;
}
z->len = mpn_remove_trailing_zeros(z->dig, z->dig + z->len);
}
#if 0
these functions are unused
bool mpz_is_pos(const mpz_t *z) {
return z->len > 0 && z->neg == 0;
}
bool mpz_is_odd(const mpz_t *z) {
return z->len > 0 && (z->dig[0] & 1) != 0;
}
bool mpz_is_even(const mpz_t *z) {
return z->len == 0 || (z->dig[0] & 1) == 0;
}
#endif
int mpz_cmp(const mpz_t *z1, const mpz_t *z2) {
int cmp = (int)z2->neg - (int)z1->neg;
if (cmp != 0) {
return cmp;
}
cmp = mpn_cmp(z1->dig, z1->len, z2->dig, z2->len);
if (z1->neg != 0) {
cmp = -cmp;
}
return cmp;
}
#if 0
// obsolete
// compares mpz with an integer that fits within DIG_SIZE bits
mp_int_t mpz_cmp_sml_int(const mpz_t *z, mp_int_t sml_int) {
mp_int_t cmp;
if (z->neg == 0) {
if (sml_int < 0) {
return 1;
}
if (sml_int == 0) {
if (z->len == 0) {
return 0;
}
return 1;
}
if (z->len == 0) {
return -1;
}
assert(sml_int < (1 << DIG_SIZE));
if (z->len != 1) {
return 1;
}
cmp = z->dig[0] - sml_int;
} else {
if (sml_int > 0) {
return -1;
}
if (sml_int == 0) {
if (z->len == 0) {
return 0;
}
return -1;
}
if (z->len == 0) {
return 1;
}
assert(sml_int > -(1 << DIG_SIZE));
if (z->len != 1) {
return -1;
}
cmp = -z->dig[0] - sml_int;
}
if (cmp < 0) {
return -1;
}
if (cmp > 0) {
return 1;
}
return 0;
}
#endif
#if 0
these functions are unused
/* returns abs(z)
*/
mpz_t *mpz_abs(const mpz_t *z) {
// TODO: handle case of z->alloc=0
mpz_t *z2 = mpz_clone(z);
z2->neg = 0;
return z2;
}
/* returns -z
*/
mpz_t *mpz_neg(const mpz_t *z) {
// TODO: handle case of z->alloc=0
mpz_t *z2 = mpz_clone(z);
z2->neg = 1 - z2->neg;
return z2;
}
/* returns lhs + rhs
can have lhs, rhs the same
*/
mpz_t *mpz_add(const mpz_t *lhs, const mpz_t *rhs) {
mpz_t *z = mpz_zero();
mpz_add_inpl(z, lhs, rhs);
return z;
}
/* returns lhs - rhs
can have lhs, rhs the same
*/
mpz_t *mpz_sub(const mpz_t *lhs, const mpz_t *rhs) {
mpz_t *z = mpz_zero();