xref: /freebsd/crypto/openssl/crypto/bn/bn_asm.c (revision b077aed33b7b6aefca7b17ddb250cf521f938613)
1 /*
2  * Copyright 1995-2023 The OpenSSL Project Authors. All Rights Reserved.
3  *
4  * Licensed under the Apache License 2.0 (the "License").  You may not use
5  * this file except in compliance with the License.  You can obtain a copy
6  * in the file LICENSE in the source distribution or at
7  * https://www.openssl.org/source/license.html
8  */
9 
10 #include <assert.h>
11 #include <openssl/crypto.h>
12 #include "internal/cryptlib.h"
13 #include "bn_local.h"
14 
15 #if defined(BN_LLONG) || defined(BN_UMULT_HIGH)
16 
bn_mul_add_words(BN_ULONG * rp,const BN_ULONG * ap,int num,BN_ULONG w)17 BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num,
18                           BN_ULONG w)
19 {
20     BN_ULONG c1 = 0;
21 
22     assert(num >= 0);
23     if (num <= 0)
24         return c1;
25 
26 # ifndef OPENSSL_SMALL_FOOTPRINT
27     while (num & ~3) {
28         mul_add(rp[0], ap[0], w, c1);
29         mul_add(rp[1], ap[1], w, c1);
30         mul_add(rp[2], ap[2], w, c1);
31         mul_add(rp[3], ap[3], w, c1);
32         ap += 4;
33         rp += 4;
34         num -= 4;
35     }
36 # endif
37     while (num) {
38         mul_add(rp[0], ap[0], w, c1);
39         ap++;
40         rp++;
41         num--;
42     }
43 
44     return c1;
45 }
46 
bn_mul_words(BN_ULONG * rp,const BN_ULONG * ap,int num,BN_ULONG w)47 BN_ULONG bn_mul_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
48 {
49     BN_ULONG c1 = 0;
50 
51     assert(num >= 0);
52     if (num <= 0)
53         return c1;
54 
55 # ifndef OPENSSL_SMALL_FOOTPRINT
56     while (num & ~3) {
57         mul(rp[0], ap[0], w, c1);
58         mul(rp[1], ap[1], w, c1);
59         mul(rp[2], ap[2], w, c1);
60         mul(rp[3], ap[3], w, c1);
61         ap += 4;
62         rp += 4;
63         num -= 4;
64     }
65 # endif
66     while (num) {
67         mul(rp[0], ap[0], w, c1);
68         ap++;
69         rp++;
70         num--;
71     }
72     return c1;
73 }
74 
bn_sqr_words(BN_ULONG * r,const BN_ULONG * a,int n)75 void bn_sqr_words(BN_ULONG *r, const BN_ULONG *a, int n)
76 {
77     assert(n >= 0);
78     if (n <= 0)
79         return;
80 
81 # ifndef OPENSSL_SMALL_FOOTPRINT
82     while (n & ~3) {
83         sqr(r[0], r[1], a[0]);
84         sqr(r[2], r[3], a[1]);
85         sqr(r[4], r[5], a[2]);
86         sqr(r[6], r[7], a[3]);
87         a += 4;
88         r += 8;
89         n -= 4;
90     }
91 # endif
92     while (n) {
93         sqr(r[0], r[1], a[0]);
94         a++;
95         r += 2;
96         n--;
97     }
98 }
99 
100 #else                           /* !(defined(BN_LLONG) ||
101                                  * defined(BN_UMULT_HIGH)) */
102 
bn_mul_add_words(BN_ULONG * rp,const BN_ULONG * ap,int num,BN_ULONG w)103 BN_ULONG bn_mul_add_words(BN_ULONG *rp, const BN_ULONG *ap, int num,
104                           BN_ULONG w)
105 {
106     BN_ULONG c = 0;
107     BN_ULONG bl, bh;
108 
109     assert(num >= 0);
110     if (num <= 0)
111         return (BN_ULONG)0;
112 
113     bl = LBITS(w);
114     bh = HBITS(w);
115 
116 # ifndef OPENSSL_SMALL_FOOTPRINT
117     while (num & ~3) {
118         mul_add(rp[0], ap[0], bl, bh, c);
119         mul_add(rp[1], ap[1], bl, bh, c);
120         mul_add(rp[2], ap[2], bl, bh, c);
121         mul_add(rp[3], ap[3], bl, bh, c);
122         ap += 4;
123         rp += 4;
124         num -= 4;
125     }
126 # endif
127     while (num) {
128         mul_add(rp[0], ap[0], bl, bh, c);
129         ap++;
130         rp++;
131         num--;
132     }
133     return c;
134 }
135 
bn_mul_words(BN_ULONG * rp,const BN_ULONG * ap,int num,BN_ULONG w)136 BN_ULONG bn_mul_words(BN_ULONG *rp, const BN_ULONG *ap, int num, BN_ULONG w)
137 {
138     BN_ULONG carry = 0;
139     BN_ULONG bl, bh;
140 
141     assert(num >= 0);
142     if (num <= 0)
143         return (BN_ULONG)0;
144 
145     bl = LBITS(w);
146     bh = HBITS(w);
147 
148 # ifndef OPENSSL_SMALL_FOOTPRINT
149     while (num & ~3) {
150         mul(rp[0], ap[0], bl, bh, carry);
151         mul(rp[1], ap[1], bl, bh, carry);
152         mul(rp[2], ap[2], bl, bh, carry);
153         mul(rp[3], ap[3], bl, bh, carry);
154         ap += 4;
155         rp += 4;
156         num -= 4;
157     }
158 # endif
159     while (num) {
160         mul(rp[0], ap[0], bl, bh, carry);
161         ap++;
162         rp++;
163         num--;
164     }
165     return carry;
166 }
167 
bn_sqr_words(BN_ULONG * r,const BN_ULONG * a,int n)168 void bn_sqr_words(BN_ULONG *r, const BN_ULONG *a, int n)
169 {
170     assert(n >= 0);
171     if (n <= 0)
172         return;
173 
174 # ifndef OPENSSL_SMALL_FOOTPRINT
175     while (n & ~3) {
176         sqr64(r[0], r[1], a[0]);
177         sqr64(r[2], r[3], a[1]);
178         sqr64(r[4], r[5], a[2]);
179         sqr64(r[6], r[7], a[3]);
180         a += 4;
181         r += 8;
182         n -= 4;
183     }
184 # endif
185     while (n) {
186         sqr64(r[0], r[1], a[0]);
187         a++;
188         r += 2;
189         n--;
190     }
191 }
192 
193 #endif                          /* !(defined(BN_LLONG) ||
194                                  * defined(BN_UMULT_HIGH)) */
195 
196 #if defined(BN_LLONG) && defined(BN_DIV2W)
197 
bn_div_words(BN_ULONG h,BN_ULONG l,BN_ULONG d)198 BN_ULONG bn_div_words(BN_ULONG h, BN_ULONG l, BN_ULONG d)
199 {
200     return ((BN_ULONG)(((((BN_ULLONG) h) << BN_BITS2) | l) / (BN_ULLONG) d));
201 }
202 
203 #else
204 
205 /* Divide h,l by d and return the result. */
206 /* I need to test this some more :-( */
bn_div_words(BN_ULONG h,BN_ULONG l,BN_ULONG d)207 BN_ULONG bn_div_words(BN_ULONG h, BN_ULONG l, BN_ULONG d)
208 {
209     BN_ULONG dh, dl, q, ret = 0, th, tl, t;
210     int i, count = 2;
211 
212     if (d == 0)
213         return BN_MASK2;
214 
215     i = BN_num_bits_word(d);
216     assert((i == BN_BITS2) || (h <= (BN_ULONG)1 << i));
217 
218     i = BN_BITS2 - i;
219     if (h >= d)
220         h -= d;
221 
222     if (i) {
223         d <<= i;
224         h = (h << i) | (l >> (BN_BITS2 - i));
225         l <<= i;
226     }
227     dh = (d & BN_MASK2h) >> BN_BITS4;
228     dl = (d & BN_MASK2l);
229     for (;;) {
230         if ((h >> BN_BITS4) == dh)
231             q = BN_MASK2l;
232         else
233             q = h / dh;
234 
235         th = q * dh;
236         tl = dl * q;
237         for (;;) {
238             t = h - th;
239             if ((t & BN_MASK2h) ||
240                 ((tl) <= ((t << BN_BITS4) | ((l & BN_MASK2h) >> BN_BITS4))))
241                 break;
242             q--;
243             th -= dh;
244             tl -= dl;
245         }
246         t = (tl >> BN_BITS4);
247         tl = (tl << BN_BITS4) & BN_MASK2h;
248         th += t;
249 
250         if (l < tl)
251             th++;
252         l -= tl;
253         if (h < th) {
254             h += d;
255             q--;
256         }
257         h -= th;
258 
259         if (--count == 0)
260             break;
261 
262         ret = q << BN_BITS4;
263         h = ((h << BN_BITS4) | (l >> BN_BITS4)) & BN_MASK2;
264         l = (l & BN_MASK2l) << BN_BITS4;
265     }
266     ret |= q;
267     return ret;
268 }
269 #endif                          /* !defined(BN_LLONG) && defined(BN_DIV2W) */
270 
271 #ifdef BN_LLONG
bn_add_words(BN_ULONG * r,const BN_ULONG * a,const BN_ULONG * b,int n)272 BN_ULONG bn_add_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b,
273                       int n)
274 {
275     BN_ULLONG ll = 0;
276 
277     assert(n >= 0);
278     if (n <= 0)
279         return (BN_ULONG)0;
280 
281 # ifndef OPENSSL_SMALL_FOOTPRINT
282     while (n & ~3) {
283         ll += (BN_ULLONG) a[0] + b[0];
284         r[0] = (BN_ULONG)ll & BN_MASK2;
285         ll >>= BN_BITS2;
286         ll += (BN_ULLONG) a[1] + b[1];
287         r[1] = (BN_ULONG)ll & BN_MASK2;
288         ll >>= BN_BITS2;
289         ll += (BN_ULLONG) a[2] + b[2];
290         r[2] = (BN_ULONG)ll & BN_MASK2;
291         ll >>= BN_BITS2;
292         ll += (BN_ULLONG) a[3] + b[3];
293         r[3] = (BN_ULONG)ll & BN_MASK2;
294         ll >>= BN_BITS2;
295         a += 4;
296         b += 4;
297         r += 4;
298         n -= 4;
299     }
300 # endif
301     while (n) {
302         ll += (BN_ULLONG) a[0] + b[0];
303         r[0] = (BN_ULONG)ll & BN_MASK2;
304         ll >>= BN_BITS2;
305         a++;
306         b++;
307         r++;
308         n--;
309     }
310     return (BN_ULONG)ll;
311 }
312 #else                           /* !BN_LLONG */
bn_add_words(BN_ULONG * r,const BN_ULONG * a,const BN_ULONG * b,int n)313 BN_ULONG bn_add_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b,
314                       int n)
315 {
316     BN_ULONG c, l, t;
317 
318     assert(n >= 0);
319     if (n <= 0)
320         return (BN_ULONG)0;
321 
322     c = 0;
323 # ifndef OPENSSL_SMALL_FOOTPRINT
324     while (n & ~3) {
325         t = a[0];
326         t = (t + c) & BN_MASK2;
327         c = (t < c);
328         l = (t + b[0]) & BN_MASK2;
329         c += (l < t);
330         r[0] = l;
331         t = a[1];
332         t = (t + c) & BN_MASK2;
333         c = (t < c);
334         l = (t + b[1]) & BN_MASK2;
335         c += (l < t);
336         r[1] = l;
337         t = a[2];
338         t = (t + c) & BN_MASK2;
339         c = (t < c);
340         l = (t + b[2]) & BN_MASK2;
341         c += (l < t);
342         r[2] = l;
343         t = a[3];
344         t = (t + c) & BN_MASK2;
345         c = (t < c);
346         l = (t + b[3]) & BN_MASK2;
347         c += (l < t);
348         r[3] = l;
349         a += 4;
350         b += 4;
351         r += 4;
352         n -= 4;
353     }
354 # endif
355     while (n) {
356         t = a[0];
357         t = (t + c) & BN_MASK2;
358         c = (t < c);
359         l = (t + b[0]) & BN_MASK2;
360         c += (l < t);
361         r[0] = l;
362         a++;
363         b++;
364         r++;
365         n--;
366     }
367     return (BN_ULONG)c;
368 }
369 #endif                          /* !BN_LLONG */
370 
bn_sub_words(BN_ULONG * r,const BN_ULONG * a,const BN_ULONG * b,int n)371 BN_ULONG bn_sub_words(BN_ULONG *r, const BN_ULONG *a, const BN_ULONG *b,
372                       int n)
373 {
374     BN_ULONG t1, t2;
375     int c = 0;
376 
377     assert(n >= 0);
378     if (n <= 0)
379         return (BN_ULONG)0;
380 
381 #ifndef OPENSSL_SMALL_FOOTPRINT
382     while (n & ~3) {
383         t1 = a[0];
384         t2 = (t1 - c) & BN_MASK2;
385         c  = (t2 > t1);
386         t1 = b[0];
387         t1 = (t2 - t1) & BN_MASK2;
388         r[0] = t1;
389         c += (t1 > t2);
390         t1 = a[1];
391         t2 = (t1 - c) & BN_MASK2;
392         c  = (t2 > t1);
393         t1 = b[1];
394         t1 = (t2 - t1) & BN_MASK2;
395         r[1] = t1;
396         c += (t1 > t2);
397         t1 = a[2];
398         t2 = (t1 - c) & BN_MASK2;
399         c  = (t2 > t1);
400         t1 = b[2];
401         t1 = (t2 - t1) & BN_MASK2;
402         r[2] = t1;
403         c += (t1 > t2);
404         t1 = a[3];
405         t2 = (t1 - c) & BN_MASK2;
406         c  = (t2 > t1);
407         t1 = b[3];
408         t1 = (t2 - t1) & BN_MASK2;
409         r[3] = t1;
410         c += (t1 > t2);
411         a += 4;
412         b += 4;
413         r += 4;
414         n -= 4;
415     }
416 #endif
417     while (n) {
418         t1 = a[0];
419         t2 = (t1 - c) & BN_MASK2;
420         c  = (t2 > t1);
421         t1 = b[0];
422         t1 = (t2 - t1) & BN_MASK2;
423         r[0] = t1;
424         c += (t1 > t2);
425         a++;
426         b++;
427         r++;
428         n--;
429     }
430     return c;
431 }
432 
433 #if defined(BN_MUL_COMBA) && !defined(OPENSSL_SMALL_FOOTPRINT)
434 
435 # undef bn_mul_comba8
436 # undef bn_mul_comba4
437 # undef bn_sqr_comba8
438 # undef bn_sqr_comba4
439 
440 /* mul_add_c(a,b,c0,c1,c2)  -- c+=a*b for three word number c=(c2,c1,c0) */
441 /* mul_add_c2(a,b,c0,c1,c2) -- c+=2*a*b for three word number c=(c2,c1,c0) */
442 /* sqr_add_c(a,i,c0,c1,c2)  -- c+=a[i]^2 for three word number c=(c2,c1,c0) */
443 /*
444  * sqr_add_c2(a,i,c0,c1,c2) -- c+=2*a[i]*a[j] for three word number
445  * c=(c2,c1,c0)
446  */
447 
448 # ifdef BN_LLONG
449 /*
450  * Keep in mind that additions to multiplication result can not
451  * overflow, because its high half cannot be all-ones.
452  */
453 #  define mul_add_c(a,b,c0,c1,c2)       do {    \
454         BN_ULONG hi;                            \
455         BN_ULLONG t = (BN_ULLONG)(a)*(b);       \
456         t += c0;                /* no carry */  \
457         c0 = (BN_ULONG)Lw(t);                   \
458         hi = (BN_ULONG)Hw(t);                   \
459         c1 = (c1+hi)&BN_MASK2; c2 += (c1<hi);   \
460         } while(0)
461 
462 #  define mul_add_c2(a,b,c0,c1,c2)      do {    \
463         BN_ULONG hi;                            \
464         BN_ULLONG t = (BN_ULLONG)(a)*(b);       \
465         BN_ULLONG tt = t+c0;    /* no carry */  \
466         c0 = (BN_ULONG)Lw(tt);                  \
467         hi = (BN_ULONG)Hw(tt);                  \
468         c1 = (c1+hi)&BN_MASK2; c2 += (c1<hi);   \
469         t += c0;                /* no carry */  \
470         c0 = (BN_ULONG)Lw(t);                   \
471         hi = (BN_ULONG)Hw(t);                   \
472         c1 = (c1+hi)&BN_MASK2; c2 += (c1<hi);   \
473         } while(0)
474 
475 #  define sqr_add_c(a,i,c0,c1,c2)       do {    \
476         BN_ULONG hi;                            \
477         BN_ULLONG t = (BN_ULLONG)a[i]*a[i];     \
478         t += c0;                /* no carry */  \
479         c0 = (BN_ULONG)Lw(t);                   \
480         hi = (BN_ULONG)Hw(t);                   \
481         c1 = (c1+hi)&BN_MASK2; c2 += (c1<hi);   \
482         } while(0)
483 
484 #  define sqr_add_c2(a,i,j,c0,c1,c2) \
485         mul_add_c2((a)[i],(a)[j],c0,c1,c2)
486 
487 # elif defined(BN_UMULT_LOHI)
488 /*
489  * Keep in mind that additions to hi can not overflow, because
490  * the high word of a multiplication result cannot be all-ones.
491  */
492 #  define mul_add_c(a,b,c0,c1,c2)       do {    \
493         BN_ULONG ta = (a), tb = (b);            \
494         BN_ULONG lo, hi;                        \
495         BN_UMULT_LOHI(lo,hi,ta,tb);             \
496         c0 += lo; hi += (c0<lo);                \
497         c1 += hi; c2 += (c1<hi);                \
498         } while(0)
499 
500 #  define mul_add_c2(a,b,c0,c1,c2)      do {    \
501         BN_ULONG ta = (a), tb = (b);            \
502         BN_ULONG lo, hi, tt;                    \
503         BN_UMULT_LOHI(lo,hi,ta,tb);             \
504         c0 += lo; tt = hi + (c0<lo);            \
505         c1 += tt; c2 += (c1<tt);                \
506         c0 += lo; hi += (c0<lo);                \
507         c1 += hi; c2 += (c1<hi);                \
508         } while(0)
509 
510 #  define sqr_add_c(a,i,c0,c1,c2)       do {    \
511         BN_ULONG ta = (a)[i];                   \
512         BN_ULONG lo, hi;                        \
513         BN_UMULT_LOHI(lo,hi,ta,ta);             \
514         c0 += lo; hi += (c0<lo);                \
515         c1 += hi; c2 += (c1<hi);                \
516         } while(0)
517 
518 #  define sqr_add_c2(a,i,j,c0,c1,c2)    \
519         mul_add_c2((a)[i],(a)[j],c0,c1,c2)
520 
521 # elif defined(BN_UMULT_HIGH)
522 /*
523  * Keep in mind that additions to hi can not overflow, because
524  * the high word of a multiplication result cannot be all-ones.
525  */
526 #  define mul_add_c(a,b,c0,c1,c2)       do {    \
527         BN_ULONG ta = (a), tb = (b);            \
528         BN_ULONG lo = ta * tb;                  \
529         BN_ULONG hi = BN_UMULT_HIGH(ta,tb);     \
530         c0 += lo; hi += (c0<lo);                \
531         c1 += hi; c2 += (c1<hi);                \
532         } while(0)
533 
534 #  define mul_add_c2(a,b,c0,c1,c2)      do {    \
535         BN_ULONG ta = (a), tb = (b), tt;        \
536         BN_ULONG lo = ta * tb;                  \
537         BN_ULONG hi = BN_UMULT_HIGH(ta,tb);     \
538         c0 += lo; tt = hi + (c0<lo);            \
539         c1 += tt; c2 += (c1<tt);                \
540         c0 += lo; hi += (c0<lo);                \
541         c1 += hi; c2 += (c1<hi);                \
542         } while(0)
543 
544 #  define sqr_add_c(a,i,c0,c1,c2)       do {    \
545         BN_ULONG ta = (a)[i];                   \
546         BN_ULONG lo = ta * ta;                  \
547         BN_ULONG hi = BN_UMULT_HIGH(ta,ta);     \
548         c0 += lo; hi += (c0<lo);                \
549         c1 += hi; c2 += (c1<hi);                \
550         } while(0)
551 
552 #  define sqr_add_c2(a,i,j,c0,c1,c2)      \
553         mul_add_c2((a)[i],(a)[j],c0,c1,c2)
554 
555 # else                          /* !BN_LLONG */
556 /*
557  * Keep in mind that additions to hi can not overflow, because
558  * the high word of a multiplication result cannot be all-ones.
559  */
560 #  define mul_add_c(a,b,c0,c1,c2)       do {    \
561         BN_ULONG lo = LBITS(a), hi = HBITS(a);  \
562         BN_ULONG bl = LBITS(b), bh = HBITS(b);  \
563         mul64(lo,hi,bl,bh);                     \
564         c0 = (c0+lo)&BN_MASK2; hi += (c0<lo);   \
565         c1 = (c1+hi)&BN_MASK2; c2 += (c1<hi);   \
566         } while(0)
567 
568 #  define mul_add_c2(a,b,c0,c1,c2)      do {    \
569         BN_ULONG tt;                            \
570         BN_ULONG lo = LBITS(a), hi = HBITS(a);  \
571         BN_ULONG bl = LBITS(b), bh = HBITS(b);  \
572         mul64(lo,hi,bl,bh);                     \
573         tt = hi;                                \
574         c0 = (c0+lo)&BN_MASK2; tt += (c0<lo);   \
575         c1 = (c1+tt)&BN_MASK2; c2 += (c1<tt);   \
576         c0 = (c0+lo)&BN_MASK2; hi += (c0<lo);   \
577         c1 = (c1+hi)&BN_MASK2; c2 += (c1<hi);   \
578         } while(0)
579 
580 #  define sqr_add_c(a,i,c0,c1,c2)       do {    \
581         BN_ULONG lo, hi;                        \
582         sqr64(lo,hi,(a)[i]);                    \
583         c0 = (c0+lo)&BN_MASK2; hi += (c0<lo);   \
584         c1 = (c1+hi)&BN_MASK2; c2 += (c1<hi);   \
585         } while(0)
586 
587 #  define sqr_add_c2(a,i,j,c0,c1,c2) \
588         mul_add_c2((a)[i],(a)[j],c0,c1,c2)
589 # endif                         /* !BN_LLONG */
590 
bn_mul_comba8(BN_ULONG * r,BN_ULONG * a,BN_ULONG * b)591 void bn_mul_comba8(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
592 {
593     BN_ULONG c1, c2, c3;
594 
595     c1 = 0;
596     c2 = 0;
597     c3 = 0;
598     mul_add_c(a[0], b[0], c1, c2, c3);
599     r[0] = c1;
600     c1 = 0;
601     mul_add_c(a[0], b[1], c2, c3, c1);
602     mul_add_c(a[1], b[0], c2, c3, c1);
603     r[1] = c2;
604     c2 = 0;
605     mul_add_c(a[2], b[0], c3, c1, c2);
606     mul_add_c(a[1], b[1], c3, c1, c2);
607     mul_add_c(a[0], b[2], c3, c1, c2);
608     r[2] = c3;
609     c3 = 0;
610     mul_add_c(a[0], b[3], c1, c2, c3);
611     mul_add_c(a[1], b[2], c1, c2, c3);
612     mul_add_c(a[2], b[1], c1, c2, c3);
613     mul_add_c(a[3], b[0], c1, c2, c3);
614     r[3] = c1;
615     c1 = 0;
616     mul_add_c(a[4], b[0], c2, c3, c1);
617     mul_add_c(a[3], b[1], c2, c3, c1);
618     mul_add_c(a[2], b[2], c2, c3, c1);
619     mul_add_c(a[1], b[3], c2, c3, c1);
620     mul_add_c(a[0], b[4], c2, c3, c1);
621     r[4] = c2;
622     c2 = 0;
623     mul_add_c(a[0], b[5], c3, c1, c2);
624     mul_add_c(a[1], b[4], c3, c1, c2);
625     mul_add_c(a[2], b[3], c3, c1, c2);
626     mul_add_c(a[3], b[2], c3, c1, c2);
627     mul_add_c(a[4], b[1], c3, c1, c2);
628     mul_add_c(a[5], b[0], c3, c1, c2);
629     r[5] = c3;
630     c3 = 0;
631     mul_add_c(a[6], b[0], c1, c2, c3);
632     mul_add_c(a[5], b[1], c1, c2, c3);
633     mul_add_c(a[4], b[2], c1, c2, c3);
634     mul_add_c(a[3], b[3], c1, c2, c3);
635     mul_add_c(a[2], b[4], c1, c2, c3);
636     mul_add_c(a[1], b[5], c1, c2, c3);
637     mul_add_c(a[0], b[6], c1, c2, c3);
638     r[6] = c1;
639     c1 = 0;
640     mul_add_c(a[0], b[7], c2, c3, c1);
641     mul_add_c(a[1], b[6], c2, c3, c1);
642     mul_add_c(a[2], b[5], c2, c3, c1);
643     mul_add_c(a[3], b[4], c2, c3, c1);
644     mul_add_c(a[4], b[3], c2, c3, c1);
645     mul_add_c(a[5], b[2], c2, c3, c1);
646     mul_add_c(a[6], b[1], c2, c3, c1);
647     mul_add_c(a[7], b[0], c2, c3, c1);
648     r[7] = c2;
649     c2 = 0;
650     mul_add_c(a[7], b[1], c3, c1, c2);
651     mul_add_c(a[6], b[2], c3, c1, c2);
652     mul_add_c(a[5], b[3], c3, c1, c2);
653     mul_add_c(a[4], b[4], c3, c1, c2);
654     mul_add_c(a[3], b[5], c3, c1, c2);
655     mul_add_c(a[2], b[6], c3, c1, c2);
656     mul_add_c(a[1], b[7], c3, c1, c2);
657     r[8] = c3;
658     c3 = 0;
659     mul_add_c(a[2], b[7], c1, c2, c3);
660     mul_add_c(a[3], b[6], c1, c2, c3);
661     mul_add_c(a[4], b[5], c1, c2, c3);
662     mul_add_c(a[5], b[4], c1, c2, c3);
663     mul_add_c(a[6], b[3], c1, c2, c3);
664     mul_add_c(a[7], b[2], c1, c2, c3);
665     r[9] = c1;
666     c1 = 0;
667     mul_add_c(a[7], b[3], c2, c3, c1);
668     mul_add_c(a[6], b[4], c2, c3, c1);
669     mul_add_c(a[5], b[5], c2, c3, c1);
670     mul_add_c(a[4], b[6], c2, c3, c1);
671     mul_add_c(a[3], b[7], c2, c3, c1);
672     r[10] = c2;
673     c2 = 0;
674     mul_add_c(a[4], b[7], c3, c1, c2);
675     mul_add_c(a[5], b[6], c3, c1, c2);
676     mul_add_c(a[6], b[5], c3, c1, c2);
677     mul_add_c(a[7], b[4], c3, c1, c2);
678     r[11] = c3;
679     c3 = 0;
680     mul_add_c(a[7], b[5], c1, c2, c3);
681     mul_add_c(a[6], b[6], c1, c2, c3);
682     mul_add_c(a[5], b[7], c1, c2, c3);
683     r[12] = c1;
684     c1 = 0;
685     mul_add_c(a[6], b[7], c2, c3, c1);
686     mul_add_c(a[7], b[6], c2, c3, c1);
687     r[13] = c2;
688     c2 = 0;
689     mul_add_c(a[7], b[7], c3, c1, c2);
690     r[14] = c3;
691     r[15] = c1;
692 }
693 
bn_mul_comba4(BN_ULONG * r,BN_ULONG * a,BN_ULONG * b)694 void bn_mul_comba4(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
695 {
696     BN_ULONG c1, c2, c3;
697 
698     c1 = 0;
699     c2 = 0;
700     c3 = 0;
701     mul_add_c(a[0], b[0], c1, c2, c3);
702     r[0] = c1;
703     c1 = 0;
704     mul_add_c(a[0], b[1], c2, c3, c1);
705     mul_add_c(a[1], b[0], c2, c3, c1);
706     r[1] = c2;
707     c2 = 0;
708     mul_add_c(a[2], b[0], c3, c1, c2);
709     mul_add_c(a[1], b[1], c3, c1, c2);
710     mul_add_c(a[0], b[2], c3, c1, c2);
711     r[2] = c3;
712     c3 = 0;
713     mul_add_c(a[0], b[3], c1, c2, c3);
714     mul_add_c(a[1], b[2], c1, c2, c3);
715     mul_add_c(a[2], b[1], c1, c2, c3);
716     mul_add_c(a[3], b[0], c1, c2, c3);
717     r[3] = c1;
718     c1 = 0;
719     mul_add_c(a[3], b[1], c2, c3, c1);
720     mul_add_c(a[2], b[2], c2, c3, c1);
721     mul_add_c(a[1], b[3], c2, c3, c1);
722     r[4] = c2;
723     c2 = 0;
724     mul_add_c(a[2], b[3], c3, c1, c2);
725     mul_add_c(a[3], b[2], c3, c1, c2);
726     r[5] = c3;
727     c3 = 0;
728     mul_add_c(a[3], b[3], c1, c2, c3);
729     r[6] = c1;
730     r[7] = c2;
731 }
732 
bn_sqr_comba8(BN_ULONG * r,const BN_ULONG * a)733 void bn_sqr_comba8(BN_ULONG *r, const BN_ULONG *a)
734 {
735     BN_ULONG c1, c2, c3;
736 
737     c1 = 0;
738     c2 = 0;
739     c3 = 0;
740     sqr_add_c(a, 0, c1, c2, c3);
741     r[0] = c1;
742     c1 = 0;
743     sqr_add_c2(a, 1, 0, c2, c3, c1);
744     r[1] = c2;
745     c2 = 0;
746     sqr_add_c(a, 1, c3, c1, c2);
747     sqr_add_c2(a, 2, 0, c3, c1, c2);
748     r[2] = c3;
749     c3 = 0;
750     sqr_add_c2(a, 3, 0, c1, c2, c3);
751     sqr_add_c2(a, 2, 1, c1, c2, c3);
752     r[3] = c1;
753     c1 = 0;
754     sqr_add_c(a, 2, c2, c3, c1);
755     sqr_add_c2(a, 3, 1, c2, c3, c1);
756     sqr_add_c2(a, 4, 0, c2, c3, c1);
757     r[4] = c2;
758     c2 = 0;
759     sqr_add_c2(a, 5, 0, c3, c1, c2);
760     sqr_add_c2(a, 4, 1, c3, c1, c2);
761     sqr_add_c2(a, 3, 2, c3, c1, c2);
762     r[5] = c3;
763     c3 = 0;
764     sqr_add_c(a, 3, c1, c2, c3);
765     sqr_add_c2(a, 4, 2, c1, c2, c3);
766     sqr_add_c2(a, 5, 1, c1, c2, c3);
767     sqr_add_c2(a, 6, 0, c1, c2, c3);
768     r[6] = c1;
769     c1 = 0;
770     sqr_add_c2(a, 7, 0, c2, c3, c1);
771     sqr_add_c2(a, 6, 1, c2, c3, c1);
772     sqr_add_c2(a, 5, 2, c2, c3, c1);
773     sqr_add_c2(a, 4, 3, c2, c3, c1);
774     r[7] = c2;
775     c2 = 0;
776     sqr_add_c(a, 4, c3, c1, c2);
777     sqr_add_c2(a, 5, 3, c3, c1, c2);
778     sqr_add_c2(a, 6, 2, c3, c1, c2);
779     sqr_add_c2(a, 7, 1, c3, c1, c2);
780     r[8] = c3;
781     c3 = 0;
782     sqr_add_c2(a, 7, 2, c1, c2, c3);
783     sqr_add_c2(a, 6, 3, c1, c2, c3);
784     sqr_add_c2(a, 5, 4, c1, c2, c3);
785     r[9] = c1;
786     c1 = 0;
787     sqr_add_c(a, 5, c2, c3, c1);
788     sqr_add_c2(a, 6, 4, c2, c3, c1);
789     sqr_add_c2(a, 7, 3, c2, c3, c1);
790     r[10] = c2;
791     c2 = 0;
792     sqr_add_c2(a, 7, 4, c3, c1, c2);
793     sqr_add_c2(a, 6, 5, c3, c1, c2);
794     r[11] = c3;
795     c3 = 0;
796     sqr_add_c(a, 6, c1, c2, c3);
797     sqr_add_c2(a, 7, 5, c1, c2, c3);
798     r[12] = c1;
799     c1 = 0;
800     sqr_add_c2(a, 7, 6, c2, c3, c1);
801     r[13] = c2;
802     c2 = 0;
803     sqr_add_c(a, 7, c3, c1, c2);
804     r[14] = c3;
805     r[15] = c1;
806 }
807 
bn_sqr_comba4(BN_ULONG * r,const BN_ULONG * a)808 void bn_sqr_comba4(BN_ULONG *r, const BN_ULONG *a)
809 {
810     BN_ULONG c1, c2, c3;
811 
812     c1 = 0;
813     c2 = 0;
814     c3 = 0;
815     sqr_add_c(a, 0, c1, c2, c3);
816     r[0] = c1;
817     c1 = 0;
818     sqr_add_c2(a, 1, 0, c2, c3, c1);
819     r[1] = c2;
820     c2 = 0;
821     sqr_add_c(a, 1, c3, c1, c2);
822     sqr_add_c2(a, 2, 0, c3, c1, c2);
823     r[2] = c3;
824     c3 = 0;
825     sqr_add_c2(a, 3, 0, c1, c2, c3);
826     sqr_add_c2(a, 2, 1, c1, c2, c3);
827     r[3] = c1;
828     c1 = 0;
829     sqr_add_c(a, 2, c2, c3, c1);
830     sqr_add_c2(a, 3, 1, c2, c3, c1);
831     r[4] = c2;
832     c2 = 0;
833     sqr_add_c2(a, 3, 2, c3, c1, c2);
834     r[5] = c3;
835     c3 = 0;
836     sqr_add_c(a, 3, c1, c2, c3);
837     r[6] = c1;
838     r[7] = c2;
839 }
840 
841 # ifdef OPENSSL_NO_ASM
842 #  ifdef OPENSSL_BN_ASM_MONT
843 #   include <alloca.h>
844 /*
845  * This is essentially reference implementation, which may or may not
846  * result in performance improvement. E.g. on IA-32 this routine was
847  * observed to give 40% faster rsa1024 private key operations and 10%
848  * faster rsa4096 ones, while on AMD64 it improves rsa1024 sign only
849  * by 10% and *worsens* rsa4096 sign by 15%. Once again, it's a
850  * reference implementation, one to be used as starting point for
851  * platform-specific assembler. Mentioned numbers apply to compiler
852  * generated code compiled with and without -DOPENSSL_BN_ASM_MONT and
853  * can vary not only from platform to platform, but even for compiler
854  * versions. Assembler vs. assembler improvement coefficients can
855  * [and are known to] differ and are to be documented elsewhere.
856  */
bn_mul_mont(BN_ULONG * rp,const BN_ULONG * ap,const BN_ULONG * bp,const BN_ULONG * np,const BN_ULONG * n0p,int num)857 int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp,
858                 const BN_ULONG *np, const BN_ULONG *n0p, int num)
859 {
860     BN_ULONG c0, c1, ml, *tp, n0;
861 #   ifdef mul64
862     BN_ULONG mh;
863 #   endif
864     volatile BN_ULONG *vp;
865     int i = 0, j;
866 
867 #   if 0                        /* template for platform-specific
868                                  * implementation */
869     if (ap == bp)
870         return bn_sqr_mont(rp, ap, np, n0p, num);
871 #   endif
872     vp = tp = alloca((num + 2) * sizeof(BN_ULONG));
873 
874     n0 = *n0p;
875 
876     c0 = 0;
877     ml = bp[0];
878 #   ifdef mul64
879     mh = HBITS(ml);
880     ml = LBITS(ml);
881     for (j = 0; j < num; ++j)
882         mul(tp[j], ap[j], ml, mh, c0);
883 #   else
884     for (j = 0; j < num; ++j)
885         mul(tp[j], ap[j], ml, c0);
886 #   endif
887 
888     tp[num] = c0;
889     tp[num + 1] = 0;
890     goto enter;
891 
892     for (i = 0; i < num; i++) {
893         c0 = 0;
894         ml = bp[i];
895 #   ifdef mul64
896         mh = HBITS(ml);
897         ml = LBITS(ml);
898         for (j = 0; j < num; ++j)
899             mul_add(tp[j], ap[j], ml, mh, c0);
900 #   else
901         for (j = 0; j < num; ++j)
902             mul_add(tp[j], ap[j], ml, c0);
903 #   endif
904         c1 = (tp[num] + c0) & BN_MASK2;
905         tp[num] = c1;
906         tp[num + 1] = (c1 < c0 ? 1 : 0);
907  enter:
908         c1 = tp[0];
909         ml = (c1 * n0) & BN_MASK2;
910         c0 = 0;
911 #   ifdef mul64
912         mh = HBITS(ml);
913         ml = LBITS(ml);
914         mul_add(c1, np[0], ml, mh, c0);
915 #   else
916         mul_add(c1, ml, np[0], c0);
917 #   endif
918         for (j = 1; j < num; j++) {
919             c1 = tp[j];
920 #   ifdef mul64
921             mul_add(c1, np[j], ml, mh, c0);
922 #   else
923             mul_add(c1, ml, np[j], c0);
924 #   endif
925             tp[j - 1] = c1 & BN_MASK2;
926         }
927         c1 = (tp[num] + c0) & BN_MASK2;
928         tp[num - 1] = c1;
929         tp[num] = tp[num + 1] + (c1 < c0 ? 1 : 0);
930     }
931 
932     if (tp[num] != 0 || tp[num - 1] >= np[num - 1]) {
933         c0 = bn_sub_words(rp, tp, np, num);
934         if (tp[num] != 0 || c0 == 0) {
935             for (i = 0; i < num + 2; i++)
936                 vp[i] = 0;
937             return 1;
938         }
939     }
940     for (i = 0; i < num; i++)
941         rp[i] = tp[i], vp[i] = 0;
942     vp[num] = 0;
943     vp[num + 1] = 0;
944     return 1;
945 }
946 #  else
947 /*
948  * Return value of 0 indicates that multiplication/convolution was not
949  * performed to signal the caller to fall down to alternative/original
950  * code-path.
951  */
bn_mul_mont(BN_ULONG * rp,const BN_ULONG * ap,const BN_ULONG * bp,const BN_ULONG * np,const BN_ULONG * n0,int num)952 int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp,
953                 const BN_ULONG *np, const BN_ULONG *n0, int num)
954 {
955     return 0;
956 }
957 #  endif                        /* OPENSSL_BN_ASM_MONT */
958 # endif
959 
960 #else                           /* !BN_MUL_COMBA */
961 
962 /* hmm... is it faster just to do a multiply? */
963 # undef bn_sqr_comba4
964 # undef bn_sqr_comba8
bn_sqr_comba4(BN_ULONG * r,const BN_ULONG * a)965 void bn_sqr_comba4(BN_ULONG *r, const BN_ULONG *a)
966 {
967     BN_ULONG t[8];
968     bn_sqr_normal(r, a, 4, t);
969 }
970 
bn_sqr_comba8(BN_ULONG * r,const BN_ULONG * a)971 void bn_sqr_comba8(BN_ULONG *r, const BN_ULONG *a)
972 {
973     BN_ULONG t[16];
974     bn_sqr_normal(r, a, 8, t);
975 }
976 
bn_mul_comba4(BN_ULONG * r,BN_ULONG * a,BN_ULONG * b)977 void bn_mul_comba4(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
978 {
979     r[4] = bn_mul_words(&(r[0]), a, 4, b[0]);
980     r[5] = bn_mul_add_words(&(r[1]), a, 4, b[1]);
981     r[6] = bn_mul_add_words(&(r[2]), a, 4, b[2]);
982     r[7] = bn_mul_add_words(&(r[3]), a, 4, b[3]);
983 }
984 
bn_mul_comba8(BN_ULONG * r,BN_ULONG * a,BN_ULONG * b)985 void bn_mul_comba8(BN_ULONG *r, BN_ULONG *a, BN_ULONG *b)
986 {
987     r[8] = bn_mul_words(&(r[0]), a, 8, b[0]);
988     r[9] = bn_mul_add_words(&(r[1]), a, 8, b[1]);
989     r[10] = bn_mul_add_words(&(r[2]), a, 8, b[2]);
990     r[11] = bn_mul_add_words(&(r[3]), a, 8, b[3]);
991     r[12] = bn_mul_add_words(&(r[4]), a, 8, b[4]);
992     r[13] = bn_mul_add_words(&(r[5]), a, 8, b[5]);
993     r[14] = bn_mul_add_words(&(r[6]), a, 8, b[6]);
994     r[15] = bn_mul_add_words(&(r[7]), a, 8, b[7]);
995 }
996 
997 # ifdef OPENSSL_NO_ASM
998 #  ifdef OPENSSL_BN_ASM_MONT
999 #   include <alloca.h>
bn_mul_mont(BN_ULONG * rp,const BN_ULONG * ap,const BN_ULONG * bp,const BN_ULONG * np,const BN_ULONG * n0p,int num)1000 int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp,
1001                 const BN_ULONG *np, const BN_ULONG *n0p, int num)
1002 {
1003     BN_ULONG c0, c1, *tp, n0 = *n0p;
1004     volatile BN_ULONG *vp;
1005     int i = 0, j;
1006 
1007     vp = tp = alloca((num + 2) * sizeof(BN_ULONG));
1008 
1009     for (i = 0; i <= num; i++)
1010         tp[i] = 0;
1011 
1012     for (i = 0; i < num; i++) {
1013         c0 = bn_mul_add_words(tp, ap, num, bp[i]);
1014         c1 = (tp[num] + c0) & BN_MASK2;
1015         tp[num] = c1;
1016         tp[num + 1] = (c1 < c0 ? 1 : 0);
1017 
1018         c0 = bn_mul_add_words(tp, np, num, tp[0] * n0);
1019         c1 = (tp[num] + c0) & BN_MASK2;
1020         tp[num] = c1;
1021         tp[num + 1] += (c1 < c0 ? 1 : 0);
1022         for (j = 0; j <= num; j++)
1023             tp[j] = tp[j + 1];
1024     }
1025 
1026     if (tp[num] != 0 || tp[num - 1] >= np[num - 1]) {
1027         c0 = bn_sub_words(rp, tp, np, num);
1028         if (tp[num] != 0 || c0 == 0) {
1029             for (i = 0; i < num + 2; i++)
1030                 vp[i] = 0;
1031             return 1;
1032         }
1033     }
1034     for (i = 0; i < num; i++)
1035         rp[i] = tp[i], vp[i] = 0;
1036     vp[num] = 0;
1037     vp[num + 1] = 0;
1038     return 1;
1039 }
1040 #  else
bn_mul_mont(BN_ULONG * rp,const BN_ULONG * ap,const BN_ULONG * bp,const BN_ULONG * np,const BN_ULONG * n0,int num)1041 int bn_mul_mont(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp,
1042                 const BN_ULONG *np, const BN_ULONG *n0, int num)
1043 {
1044     return 0;
1045 }
1046 #  endif                        /* OPENSSL_BN_ASM_MONT */
1047 # endif
1048 
1049 #endif                          /* !BN_MUL_COMBA */
1050