Coverage Report

Created: 2026-06-30 06:42

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
/src/gmp/mpn/gcdext_lehmer.c
Line
Count
Source
1
/* mpn_gcdext -- Extended Greatest Common Divisor.
2
3
Copyright 1996, 1998, 2000-2005, 2008, 2009, 2012 Free Software Foundation,
4
Inc.
5
6
This file is part of the GNU MP Library.
7
8
The GNU MP Library is free software; you can redistribute it and/or modify
9
it under the terms of either:
10
11
  * the GNU Lesser General Public License as published by the Free
12
    Software Foundation; either version 3 of the License, or (at your
13
    option) any later version.
14
15
or
16
17
  * the GNU General Public License as published by the Free Software
18
    Foundation; either version 2 of the License, or (at your option) any
19
    later version.
20
21
or both in parallel, as here.
22
23
The GNU MP Library is distributed in the hope that it will be useful, but
24
WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
25
or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License
26
for more details.
27
28
You should have received copies of the GNU General Public License and the
29
GNU Lesser General Public License along with the GNU MP Library.  If not,
30
see https://www.gnu.org/licenses/.  */
31
32
#include "gmp-impl.h"
33
#include "longlong.h"
34
35
/* Here, d is the index of the cofactor to update. FIXME: Could use qn
36
   = 0 for the common case q = 1. */
37
void
38
mpn_gcdext_hook (void *p, mp_srcptr gp, mp_size_t gn,
39
     mp_srcptr qp, mp_size_t qn, int d)
40
18.1k
{
41
18.1k
  struct gcdext_ctx *ctx = (struct gcdext_ctx *) p;
42
18.1k
  mp_size_t un = ctx->un;
43
44
18.1k
  if (gp)
45
411
    {
46
411
      mp_srcptr up;
47
48
411
      ASSERT (gn > 0);
49
411
      ASSERT (gp[gn-1] > 0);
50
51
411
      MPN_COPY (ctx->gp, gp, gn);
52
411
      ctx->gn = gn;
53
54
411
      if (d < 0)
55
215
  {
56
215
    int c;
57
58
    /* Must return the smallest cofactor, +u1 or -u0 */
59
215
    MPN_CMP (c, ctx->u0, ctx->u1, un);
60
215
    ASSERT (c != 0 || (un == 1 && ctx->u0[0] == 1 && ctx->u1[0] == 1));
61
62
215
    d = c < 0;
63
215
  }
64
65
411
      up = d ? ctx->u0 : ctx->u1;
66
67
411
      MPN_NORMALIZE (up, un);
68
411
      MPN_COPY (ctx->up, up, un);
69
70
411
      *ctx->usize = d ? -un : un;
71
411
    }
72
17.7k
  else
73
17.7k
    {
74
17.7k
      mp_limb_t cy;
75
17.7k
      mp_ptr u0 = ctx->u0;
76
17.7k
      mp_ptr u1 = ctx->u1;
77
78
17.7k
      ASSERT (d >= 0);
79
80
17.7k
      if (d)
81
8.34k
  MP_PTR_SWAP (u0, u1);
82
83
17.7k
      qn -= (qp[qn-1] == 0);
84
85
      /* Update u0 += q  * u1 */
86
17.7k
      if (qn == 1)
87
11.4k
  {
88
11.4k
    mp_limb_t q = qp[0];
89
90
11.4k
    if (q == 1)
91
      /* A common case. */
92
8.98k
      cy = mpn_add_n (u0, u0, u1, un);
93
2.48k
    else
94
2.48k
      cy = mpn_addmul_1 (u0, u1, un, q);
95
11.4k
  }
96
6.30k
      else
97
6.30k
  {
98
6.30k
    mp_size_t u1n;
99
6.30k
    mp_ptr tp;
100
101
6.30k
    u1n = un;
102
6.30k
    MPN_NORMALIZE (u1, u1n);
103
104
6.30k
    if (u1n == 0)
105
0
      return;
106
107
    /* Should always have u1n == un here, and u1 >= u0. The
108
       reason is that we alternate adding u0 to u1 and u1 to u0
109
       (corresponding to subtractions a - b and b - a), and we
110
       can get a large quotient only just after a switch, which
111
       means that we'll add (a multiple of) the larger u to the
112
       smaller. */
113
114
6.30k
    tp = ctx->tp;
115
116
6.30k
    if (qn > u1n)
117
1.77k
      mpn_mul (tp, qp, qn, u1, u1n);
118
4.53k
    else
119
4.53k
      mpn_mul (tp, u1, u1n, qp, qn);
120
121
6.30k
    u1n += qn;
122
6.30k
    u1n -= tp[u1n-1] == 0;
123
124
6.30k
    if (u1n >= un)
125
6.30k
      {
126
6.30k
        cy = mpn_add (u0, tp, u1n, u0, un);
127
6.30k
        un = u1n;
128
6.30k
      }
129
0
    else
130
      /* Note: Unlikely case, maybe never happens? */
131
0
      cy = mpn_add (u0, u0, un, tp, u1n);
132
133
6.30k
  }
134
17.7k
      u0[un] = cy;
135
17.7k
      ctx->un = un + (cy > 0);
136
17.7k
    }
137
18.1k
}
138
139
/* Temporary storage: 3*(n+1) for u. If hgcd2 succeeds, we need n for
140
   the matrix-vector multiplication adjusting a, b. If hgcd fails, we
141
   need at most n for the quotient and n+1 for the u update (reusing
142
   the extra u). In all, 4n + 3. */
143
144
mp_size_t
145
mpn_gcdext_lehmer_n (mp_ptr gp, mp_ptr up, mp_size_t *usize,
146
         mp_ptr ap, mp_ptr bp, mp_size_t n,
147
         mp_ptr tp)
148
26.6k
{
149
26.6k
  mp_size_t ualloc = n + 1;
150
151
  /* Keeps track of the second row of the reduction matrix
152
   *
153
   *   M = (v0, v1 ; u0, u1)
154
   *
155
   * which correspond to the first column of the inverse
156
   *
157
   *   M^{-1} = (u1, -v1; -u0, v0)
158
   *
159
   * This implies that
160
   *
161
   *   a =  u1 A (mod B)
162
   *   b = -u0 A (mod B)
163
   *
164
   * where A, B denotes the input values.
165
   */
166
167
26.6k
  struct gcdext_ctx ctx;
168
26.6k
  mp_size_t un;
169
26.6k
  mp_ptr u0;
170
26.6k
  mp_ptr u1;
171
26.6k
  mp_ptr u2;
172
173
26.6k
  MPN_ZERO (tp, 3*ualloc);
174
26.6k
  u0 = tp; tp += ualloc;
175
26.6k
  u1 = tp; tp += ualloc;
176
26.6k
  u2 = tp; tp += ualloc;
177
178
26.6k
  u1[0] = 1; un = 1;
179
180
26.6k
  ctx.gp = gp;
181
26.6k
  ctx.up = up;
182
26.6k
  ctx.usize = usize;
183
184
  /* FIXME: Handle n == 2 differently, after the loop? */
185
494k
  while (n >= 2)
186
467k
    {
187
467k
      struct hgcd_matrix1 M;
188
467k
      mp_limb_t ah, al, bh, bl;
189
467k
      mp_limb_t mask;
190
191
467k
      mask = ap[n-1] | bp[n-1];
192
467k
      ASSERT (mask > 0);
193
194
467k
      if (mask & GMP_NUMB_HIGHBIT)
195
18.0k
  {
196
18.0k
    ah = ap[n-1]; al = ap[n-2];
197
18.0k
    bh = bp[n-1]; bl = bp[n-2];
198
18.0k
  }
199
449k
      else if (n == 2)
200
23.6k
  {
201
    /* We use the full inputs without truncation, so we can
202
       safely shift left. */
203
23.6k
    int shift;
204
205
23.6k
    count_leading_zeros (shift, mask);
206
23.6k
    ah = MPN_EXTRACT_NUMB (shift, ap[1], ap[0]);
207
23.6k
    al = ap[0] << shift;
208
23.6k
    bh = MPN_EXTRACT_NUMB (shift, bp[1], bp[0]);
209
23.6k
    bl = bp[0] << shift;
210
23.6k
  }
211
426k
      else
212
426k
  {
213
426k
    int shift;
214
215
426k
    count_leading_zeros (shift, mask);
216
426k
    ah = MPN_EXTRACT_NUMB (shift, ap[n-1], ap[n-2]);
217
426k
    al = MPN_EXTRACT_NUMB (shift, ap[n-2], ap[n-3]);
218
426k
    bh = MPN_EXTRACT_NUMB (shift, bp[n-1], bp[n-2]);
219
426k
    bl = MPN_EXTRACT_NUMB (shift, bp[n-2], bp[n-3]);
220
426k
  }
221
222
      /* Try an mpn_nhgcd2 step */
223
467k
      if (mpn_hgcd2 (ah, al, bh, bl, &M))
224
458k
  {
225
458k
    n = mpn_matrix22_mul1_inverse_vector (&M, tp, ap, bp, n);
226
458k
    MP_PTR_SWAP (ap, tp);
227
458k
    un = mpn_hgcd_mul_matrix1_vector(&M, u2, u0, u1, un);
228
458k
    MP_PTR_SWAP (u0, u2);
229
458k
  }
230
9.20k
      else
231
9.20k
  {
232
    /* mpn_hgcd2 has failed. Then either one of a or b is very
233
       small, or the difference is very small. Perform one
234
       subtraction followed by one division. */
235
9.20k
    ctx.u0 = u0;
236
9.20k
    ctx.u1 = u1;
237
9.20k
    ctx.tp = u2;
238
9.20k
    ctx.un = un;
239
240
    /* Temporary storage n for the quotient and ualloc for the
241
       new cofactor. */
242
9.20k
    n = mpn_gcd_subdiv_step (ap, bp, n, 0, mpn_gcdext_hook, &ctx, tp);
243
9.20k
    if (n == 0)
244
411
      return ctx.gn;
245
246
8.79k
    un = ctx.un;
247
8.79k
  }
248
467k
    }
249
26.2k
  ASSERT_ALWAYS (ap[0] > 0);
250
26.2k
  ASSERT_ALWAYS (bp[0] > 0);
251
252
26.2k
  if (ap[0] == bp[0])
253
1.02k
    {
254
1.02k
      int c;
255
256
      /* Which cofactor to return now? Candidates are +u1 and -u0,
257
   depending on which of a and b was most recently reduced,
258
   which we don't keep track of. So compare and get the smallest
259
   one. */
260
261
1.02k
      gp[0] = ap[0];
262
263
1.02k
      MPN_CMP (c, u0, u1, un);
264
1.02k
      ASSERT (c != 0 || (un == 1 && u0[0] == 1 && u1[0] == 1));
265
1.02k
      if (c < 0)
266
532
  {
267
532
    MPN_NORMALIZE (u0, un);
268
532
    MPN_COPY (up, u0, un);
269
532
    *usize = -un;
270
532
  }
271
494
      else
272
494
  {
273
494
    MPN_NORMALIZE_NOT_ZERO (u1, un);
274
494
    MPN_COPY (up, u1, un);
275
494
    *usize = un;
276
494
  }
277
1.02k
      return 1;
278
1.02k
    }
279
25.2k
  else
280
25.2k
    {
281
25.2k
      mp_limb_t uh, vh;
282
25.2k
      mp_limb_signed_t u;
283
25.2k
      mp_limb_signed_t v;
284
25.2k
      int negate;
285
286
25.2k
      gp[0] = mpn_gcdext_1 (&u, &v, ap[0], bp[0]);
287
288
      /* Set up = u u1 - v u0. Keep track of size, un grows by one or
289
   two limbs. */
290
291
25.2k
      if (u == 0)
292
77
  {
293
77
    ASSERT (v == 1);
294
77
    MPN_NORMALIZE (u0, un);
295
77
    MPN_COPY (up, u0, un);
296
77
    *usize = -un;
297
77
    return 1;
298
77
  }
299
25.1k
      else if (v == 0)
300
332
  {
301
332
    ASSERT (u == 1);
302
332
    MPN_NORMALIZE (u1, un);
303
332
    MPN_COPY (up, u1, un);
304
332
    *usize = un;
305
332
    return 1;
306
332
  }
307
24.7k
      else if (u > 0)
308
18.1k
  {
309
18.1k
    negate = 0;
310
18.1k
    ASSERT (v < 0);
311
18.1k
    v = -v;
312
18.1k
  }
313
6.68k
      else
314
6.68k
  {
315
6.68k
    negate = 1;
316
6.68k
    ASSERT (v > 0);
317
6.68k
    u = -u;
318
6.68k
  }
319
320
24.7k
      uh = mpn_mul_1 (up, u1, un, u);
321
24.7k
      vh = mpn_addmul_1 (up, u0, un, v);
322
323
24.7k
      if ( (uh | vh) > 0)
324
4.12k
  {
325
4.12k
    uh += vh;
326
4.12k
    up[un++] = uh;
327
4.12k
    if (uh < vh)
328
0
      up[un++] = 1;
329
4.12k
  }
330
331
24.7k
      MPN_NORMALIZE_NOT_ZERO (up, un);
332
333
24.7k
      *usize = negate ? -un : un;
334
24.7k
      return 1;
335
25.2k
    }
336
26.2k
}