2 #ifdef BN_FAST_S_MP_SQR_C
3 /* LibTomMath, multiple-precision integer library -- Tom St Denis
5 * LibTomMath is a library that provides multiple-precision
6 * integer arithmetic as well as number theoretic functionality.
8 * The library was designed directly after the MPI library by
9 * Michael Fromberger but has been written from scratch with
10 * additional optimizations in place.
12 * The library is free for all purposes without any express
15 * Tom St Denis, tomstdenis@gmail.com, http://math.libtomcrypt.com
18 /* the jist of squaring...
19 * you do like mult except the offset of the tmpx [one that
20 * starts closer to zero] can't equal the offset of tmpy.
21 * So basically you set up iy like before then you min it with
22 * (ty-tx) so that it never happens. You double all those
23 * you add in the inner loop
25 After that loop you do the squares and add them in.
28 int fast_s_mp_sqr (mp_int * a, mp_int * b)
30 int olduse, res, pa, ix, iz;
31 mp_digit W[MP_WARRAY], *tmpx;
34 /* grow the destination as required */
35 pa = a->used + a->used;
37 if ((res = mp_grow (b, pa)) != MP_OKAY) {
42 /* number of output digits to produce */
44 for (ix = 0; ix < pa; ix++) {
52 /* get offsets into the two bignums */
53 ty = MIN(a->used-1, ix);
56 /* setup temp aliases */
60 /* this is the number of times the loop will iterrate, essentially
61 while (tx++ < a->used && ty-- >= 0) { ... }
63 iy = MIN(a->used-tx, ty+1);
65 /* now for squaring tx can never equal ty
66 * we halve the distance since they approach at a rate of 2x
67 * and we have to round because odd cases need to be executed
69 iy = MIN(iy, (ty-tx+1)>>1);
72 for (iz = 0; iz < iy; iz++) {
73 _W += ((mp_word)*tmpx++)*((mp_word)*tmpy--);
76 /* double the inner product and add carry */
79 /* even columns have the square term in them */
81 _W += ((mp_word)a->dp[ix>>1])*((mp_word)a->dp[ix>>1]);
85 W[ix] = (mp_digit)(_W & MP_MASK);
88 W1 = _W >> ((mp_word)DIGIT_BIT);
93 b->used = a->used+a->used;
98 for (ix = 0; ix < pa; ix++) {
99 *tmpb++ = W[ix] & MP_MASK;
102 /* clear unused digits [that existed in the old copy of c] */
103 for (; ix < olduse; ix++) {
112 /* $Source: /cvs/libtom/libtommath/bn_fast_s_mp_sqr.c,v $ */
113 /* $Revision: 1.3 $ */
114 /* $Date: 2006/03/31 14:18:44 $ */