1 #include "tommath_private.h"
2 #ifdef BN_S_MP_SQR_C
3 /* LibTomMath, multiple-precision integer library -- Tom St Denis */
4 /* SPDX-License-Identifier: Unlicense */
5 
6 /* low level squaring, b = a*a, HAC pp.596-597, Algorithm 14.16 */
s_mp_sqr(const mp_int * a,mp_int * b)7 mp_err s_mp_sqr(const mp_int *a, mp_int *b)
8 {
9    mp_int   t;
10    int      ix, iy, pa;
11    mp_err   err;
12    mp_word  r;
13    mp_digit u, tmpx, *tmpt;
14 
15    pa = a->used;
16    if ((err = mp_init_size(&t, (2 * pa) + 1)) != MP_OKAY) {
17       return err;
18    }
19 
20    /* default used is maximum possible size */
21    t.used = (2 * pa) + 1;
22 
23    for (ix = 0; ix < pa; ix++) {
24       /* first calculate the digit at 2*ix */
25       /* calculate double precision result */
26       r = (mp_word)t.dp[2*ix] +
27           ((mp_word)a->dp[ix] * (mp_word)a->dp[ix]);
28 
29       /* store lower part in result */
30       t.dp[ix+ix] = (mp_digit)(r & (mp_word)MP_MASK);
31 
32       /* get the carry */
33       u           = (mp_digit)(r >> (mp_word)MP_DIGIT_BIT);
34 
35       /* left hand side of A[ix] * A[iy] */
36       tmpx        = a->dp[ix];
37 
38       /* alias for where to store the results */
39       tmpt        = t.dp + ((2 * ix) + 1);
40 
41       for (iy = ix + 1; iy < pa; iy++) {
42          /* first calculate the product */
43          r       = (mp_word)tmpx * (mp_word)a->dp[iy];
44 
45          /* now calculate the double precision result, note we use
46           * addition instead of *2 since it's easier to optimize
47           */
48          r       = (mp_word)*tmpt + r + r + (mp_word)u;
49 
50          /* store lower part */
51          *tmpt++ = (mp_digit)(r & (mp_word)MP_MASK);
52 
53          /* get carry */
54          u       = (mp_digit)(r >> (mp_word)MP_DIGIT_BIT);
55       }
56       /* propagate upwards */
57       while (u != 0uL) {
58          r       = (mp_word)*tmpt + (mp_word)u;
59          *tmpt++ = (mp_digit)(r & (mp_word)MP_MASK);
60          u       = (mp_digit)(r >> (mp_word)MP_DIGIT_BIT);
61       }
62    }
63 
64    mp_clamp(&t);
65    mp_exch(&t, b);
66    mp_clear(&t);
67    return MP_OKAY;
68 }
69 #endif
70