- BN_CTX *ctx)
- {
- int i, top = a->top;
- int carry = 0;
- register BN_ULONG *r_d, *a_d = a->d;
- BN_ULONG t_d[BN_NIST_384_TOP],
- buf[BN_NIST_384_TOP],
- c_d[BN_NIST_384_TOP],
- *res;
- size_t mask;
- union { bn_addsub_f f; size_t p; } u;
- static const BIGNUM _bignum_nist_p_384_sqr = {
- (BN_ULONG *)_nist_p_384_sqr,
- sizeof(_nist_p_384_sqr)/sizeof(_nist_p_384_sqr[0]),
- sizeof(_nist_p_384_sqr)/sizeof(_nist_p_384_sqr[0]),
- 0,BN_FLG_STATIC_DATA };
-
-
- field = &_bignum_nist_p_384; /* just to make sure */
-
- if (BN_is_negative(a) || BN_ucmp(a,&_bignum_nist_p_384_sqr)>=0)
- return BN_nnmod(r, a, field, ctx);
-
- i = BN_ucmp(field, a);
- if (i == 0)
- {
- BN_zero(r);
- return 1;
- }
- else if (i > 0)
- return (r == a)? 1 : (BN_copy(r ,a) != NULL);
-
- if (r != a)
- {
- if (!bn_wexpand(r, BN_NIST_384_TOP))
- return 0;
- r_d = r->d;
- nist_cp_bn(r_d, a_d, BN_NIST_384_TOP);
- }
- else
- r_d = a_d;
-
- nist_cp_bn_0(buf, a_d + BN_NIST_384_TOP, top - BN_NIST_384_TOP, BN_NIST_384_TOP);
-
- /*S1*/
- nist_set_256(t_d, buf, 0, 0, 0, 0, 0, 23-4, 22-4, 21-4);
- /* left shift */
- {
- register BN_ULONG *ap,t,c;
- ap = t_d;
- c=0;
- for (i = 3; i != 0; --i)
- {
- t= *ap;
- *(ap++)=((t<<1)|c)&BN_MASK2;
- c=(t & BN_TBIT)?1:0;
- }
- *ap=c;
- }
- carry = bn_add_words(r_d+(128/BN_BITS2), r_d+(128/BN_BITS2),
- t_d, BN_NIST_256_TOP);
- /*S2 */
- carry += bn_add_words(r_d, r_d, buf, BN_NIST_384_TOP);
- /*S3*/
- nist_set_384(t_d,buf,20,19,18,17,16,15,14,13,12,23,22,21);
- carry += bn_add_words(r_d, r_d, t_d, BN_NIST_384_TOP);
- /*S4*/
- nist_set_384(t_d,buf,19,18,17,16,15,14,13,12,20,0,23,0);
- carry += bn_add_words(r_d, r_d, t_d, BN_NIST_384_TOP);
- /*S5*/
- nist_set_384(t_d, buf,0,0,0,0,23,22,21,20,0,0,0,0);
- carry += bn_add_words(r_d, r_d, t_d, BN_NIST_384_TOP);
- /*S6*/
- nist_set_384(t_d,buf,0,0,0,0,0,0,23,22,21,0,0,20);
- carry += bn_add_words(r_d, r_d, t_d, BN_NIST_384_TOP);
- /*D1*/
- nist_set_384(t_d,buf,22,21,20,19,18,17,16,15,14,13,12,23);
- carry -= bn_sub_words(r_d, r_d, t_d, BN_NIST_384_TOP);
- /*D2*/
- nist_set_384(t_d,buf,0,0,0,0,0,0,0,23,22,21,20,0);
- carry -= bn_sub_words(r_d, r_d, t_d, BN_NIST_384_TOP);
- /*D3*/
- nist_set_384(t_d,buf,0,0,0,0,0,0,0,23,23,0,0,0);
- carry -= bn_sub_words(r_d, r_d, t_d, BN_NIST_384_TOP);
-
- /* see BN_nist_mod_224 for explanation */
- u.f = bn_sub_words;
- if (carry > 0)
- carry = bn_sub_words(r_d,r_d,_nist_p_384[carry-1],BN_NIST_384_TOP);
- else if (carry < 0)
- {
- carry = bn_add_words(r_d,r_d,_nist_p_384[-carry-1],BN_NIST_384_TOP);
- mask = 0-(size_t)carry;
- u.p = ((size_t)bn_sub_words&mask) | ((size_t)bn_add_words&~mask);
- }
- else
- carry = 1;
-
- mask = 0-(size_t)(*u.f)(c_d,r_d,_nist_p_384[0],BN_NIST_384_TOP);
- mask &= 0-(size_t)carry;
- res = (BN_ULONG *)(((size_t)c_d&~mask) | ((size_t)r_d&mask));
- nist_cp_bn(r_d, res, BN_NIST_384_TOP);
- r->top = BN_NIST_384_TOP;
- bn_correct_top(r);
-
- return 1;
- }
-
-#define BN_NIST_521_RSHIFT (521%BN_BITS2)
-#define BN_NIST_521_LSHIFT (BN_BITS2-BN_NIST_521_RSHIFT)
-#define BN_NIST_521_TOP_MASK ((BN_ULONG)BN_MASK2>>BN_NIST_521_LSHIFT)
+ BN_CTX *ctx)
+{
+ int i, top = a->top;
+ int carry = 0;
+ register BN_ULONG *r_d, *a_d = a->d;
+ union {
+ BN_ULONG bn[BN_NIST_384_TOP];
+ unsigned int ui[BN_NIST_384_TOP * sizeof(BN_ULONG) /
+ sizeof(unsigned int)];
+ } buf;
+ BN_ULONG c_d[BN_NIST_384_TOP], *res;
+ PTR_SIZE_INT mask;
+ union {
+ bn_addsub_f f;
+ PTR_SIZE_INT p;
+ } u;
+ static const BIGNUM _bignum_nist_p_384_sqr = {
+ (BN_ULONG *)_nist_p_384_sqr,
+ OSSL_NELEM(_nist_p_384_sqr),
+ OSSL_NELEM(_nist_p_384_sqr),
+ 0, BN_FLG_STATIC_DATA
+ };
+
+ field = &_bignum_nist_p_384; /* just to make sure */
+
+ if (BN_is_negative(a) || BN_ucmp(a, &_bignum_nist_p_384_sqr) >= 0)
+ return BN_nnmod(r, a, field, ctx);
+
+ i = BN_ucmp(field, a);
+ if (i == 0) {
+ BN_zero(r);
+ return 1;
+ } else if (i > 0)
+ return (r == a) ? 1 : (BN_copy(r, a) != NULL);
+
+ if (r != a) {
+ if (!bn_wexpand(r, BN_NIST_384_TOP))
+ return 0;
+ r_d = r->d;
+ nist_cp_bn(r_d, a_d, BN_NIST_384_TOP);
+ } else
+ r_d = a_d;
+
+ nist_cp_bn_0(buf.bn, a_d + BN_NIST_384_TOP, top - BN_NIST_384_TOP,
+ BN_NIST_384_TOP);
+
+#if defined(NIST_INT64)
+ {
+ NIST_INT64 acc; /* accumulator */
+ unsigned int *rp = (unsigned int *)r_d;
+ const unsigned int *bp = (const unsigned int *)buf.ui;
+
+ acc = rp[0];
+ acc += bp[12 - 12];
+ acc += bp[21 - 12];
+ acc += bp[20 - 12];
+ acc -= bp[23 - 12];
+ rp[0] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[1];
+ acc += bp[13 - 12];
+ acc += bp[22 - 12];
+ acc += bp[23 - 12];
+ acc -= bp[12 - 12];
+ acc -= bp[20 - 12];
+ rp[1] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[2];
+ acc += bp[14 - 12];
+ acc += bp[23 - 12];
+ acc -= bp[13 - 12];
+ acc -= bp[21 - 12];
+ rp[2] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[3];
+ acc += bp[15 - 12];
+ acc += bp[12 - 12];
+ acc += bp[20 - 12];
+ acc += bp[21 - 12];
+ acc -= bp[14 - 12];
+ acc -= bp[22 - 12];
+ acc -= bp[23 - 12];
+ rp[3] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[4];
+ acc += bp[21 - 12];
+ acc += bp[21 - 12];
+ acc += bp[16 - 12];
+ acc += bp[13 - 12];
+ acc += bp[12 - 12];
+ acc += bp[20 - 12];
+ acc += bp[22 - 12];
+ acc -= bp[15 - 12];
+ acc -= bp[23 - 12];
+ acc -= bp[23 - 12];
+ rp[4] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[5];
+ acc += bp[22 - 12];
+ acc += bp[22 - 12];
+ acc += bp[17 - 12];
+ acc += bp[14 - 12];
+ acc += bp[13 - 12];
+ acc += bp[21 - 12];
+ acc += bp[23 - 12];
+ acc -= bp[16 - 12];
+ rp[5] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[6];
+ acc += bp[23 - 12];
+ acc += bp[23 - 12];
+ acc += bp[18 - 12];
+ acc += bp[15 - 12];
+ acc += bp[14 - 12];
+ acc += bp[22 - 12];
+ acc -= bp[17 - 12];
+ rp[6] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[7];
+ acc += bp[19 - 12];
+ acc += bp[16 - 12];
+ acc += bp[15 - 12];
+ acc += bp[23 - 12];
+ acc -= bp[18 - 12];
+ rp[7] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[8];
+ acc += bp[20 - 12];
+ acc += bp[17 - 12];
+ acc += bp[16 - 12];
+ acc -= bp[19 - 12];
+ rp[8] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[9];
+ acc += bp[21 - 12];
+ acc += bp[18 - 12];
+ acc += bp[17 - 12];
+ acc -= bp[20 - 12];
+ rp[9] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[10];
+ acc += bp[22 - 12];
+ acc += bp[19 - 12];
+ acc += bp[18 - 12];
+ acc -= bp[21 - 12];
+ rp[10] = (unsigned int)acc;
+ acc >>= 32;
+
+ acc += rp[11];
+ acc += bp[23 - 12];
+ acc += bp[20 - 12];
+ acc += bp[19 - 12];
+ acc -= bp[22 - 12];
+ rp[11] = (unsigned int)acc;
+
+ carry = (int)(acc >> 32);
+ }
+#else
+ {
+ BN_ULONG t_d[BN_NIST_384_TOP];
+
+ /*
+ * S1
+ */
+ nist_set_256(t_d, buf.bn, 0, 0, 0, 0, 0, 23 - 4, 22 - 4, 21 - 4);
+ /* left shift */
+ {
+ register BN_ULONG *ap, t, c;
+ ap = t_d;
+ c = 0;
+ for (i = 3; i != 0; --i) {
+ t = *ap;
+ *(ap++) = ((t << 1) | c) & BN_MASK2;
+ c = (t & BN_TBIT) ? 1 : 0;
+ }
+ *ap = c;
+ }
+ carry =
+ (int)bn_add_words(r_d + (128 / BN_BITS2), r_d + (128 / BN_BITS2),
+ t_d, BN_NIST_256_TOP);
+ /*
+ * S2
+ */
+ carry += (int)bn_add_words(r_d, r_d, buf.bn, BN_NIST_384_TOP);
+ /*
+ * S3
+ */
+ nist_set_384(t_d, buf.bn, 20, 19, 18, 17, 16, 15, 14, 13, 12, 23, 22,
+ 21);
+ carry += (int)bn_add_words(r_d, r_d, t_d, BN_NIST_384_TOP);
+ /*
+ * S4
+ */
+ nist_set_384(t_d, buf.bn, 19, 18, 17, 16, 15, 14, 13, 12, 20, 0, 23,
+ 0);
+ carry += (int)bn_add_words(r_d, r_d, t_d, BN_NIST_384_TOP);
+ /*
+ * S5
+ */
+ nist_set_384(t_d, buf.bn, 0, 0, 0, 0, 23, 22, 21, 20, 0, 0, 0, 0);
+ carry += (int)bn_add_words(r_d, r_d, t_d, BN_NIST_384_TOP);
+ /*
+ * S6
+ */
+ nist_set_384(t_d, buf.bn, 0, 0, 0, 0, 0, 0, 23, 22, 21, 0, 0, 20);
+ carry += (int)bn_add_words(r_d, r_d, t_d, BN_NIST_384_TOP);
+ /*
+ * D1
+ */
+ nist_set_384(t_d, buf.bn, 22, 21, 20, 19, 18, 17, 16, 15, 14, 13, 12,
+ 23);
+ carry -= (int)bn_sub_words(r_d, r_d, t_d, BN_NIST_384_TOP);
+ /*
+ * D2
+ */
+ nist_set_384(t_d, buf.bn, 0, 0, 0, 0, 0, 0, 0, 23, 22, 21, 20, 0);
+ carry -= (int)bn_sub_words(r_d, r_d, t_d, BN_NIST_384_TOP);
+ /*
+ * D3
+ */
+ nist_set_384(t_d, buf.bn, 0, 0, 0, 0, 0, 0, 0, 23, 23, 0, 0, 0);
+ carry -= (int)bn_sub_words(r_d, r_d, t_d, BN_NIST_384_TOP);
+
+ }
+#endif
+ /* see BN_nist_mod_224 for explanation */
+ u.f = bn_sub_words;
+ if (carry > 0)
+ carry =
+ (int)bn_sub_words(r_d, r_d, _nist_p_384[carry - 1],
+ BN_NIST_384_TOP);
+ else if (carry < 0) {
+ carry =
+ (int)bn_add_words(r_d, r_d, _nist_p_384[-carry - 1],
+ BN_NIST_384_TOP);
+ mask = 0 - (PTR_SIZE_INT) carry;
+ u.p = ((PTR_SIZE_INT) bn_sub_words & mask) |
+ ((PTR_SIZE_INT) bn_add_words & ~mask);
+ } else
+ carry = 1;
+
+ mask =
+ 0 - (PTR_SIZE_INT) (*u.f) (c_d, r_d, _nist_p_384[0], BN_NIST_384_TOP);
+ mask &= 0 - (PTR_SIZE_INT) carry;
+ res = c_d;
+ res = (BN_ULONG *)(((PTR_SIZE_INT) res & ~mask) |
+ ((PTR_SIZE_INT) r_d & mask));
+ nist_cp_bn(r_d, res, BN_NIST_384_TOP);
+ r->top = BN_NIST_384_TOP;
+ bn_correct_top(r);
+
+ return 1;
+}
+
+#define BN_NIST_521_RSHIFT (521%BN_BITS2)
+#define BN_NIST_521_LSHIFT (BN_BITS2-BN_NIST_521_RSHIFT)
+#define BN_NIST_521_TOP_MASK ((BN_ULONG)BN_MASK2>>BN_NIST_521_LSHIFT)