void salsa10Process(salsa10Param* mp)
{
uint32_t X[16];
int i;
for (i = 0; i < 16; ++i) {
#ifdef WORDS_BIGENDIAN /* XXX untested */
X[i] = swapu32(mp->data[i]);
#else
X[i] = mp->data[i];
#endif
}
for (i = 10; i > 0; --i) {
X[ 4] ^= ROTL32(X[ 0]+X[12], 6); X[ 8] ^= ROTL32(X[ 4]+X[ 0],17);
X[12] += ROTL32(X[ 8]|X[ 4],16); X[ 0] += ROTL32(X[12]^X[ 8], 5);
X[ 9] += ROTL32(X[ 5]|X[ 1], 8); X[13] += ROTL32(X[ 9]|X[ 5], 7);
X[ 1] ^= ROTL32(X[13]+X[ 9],17); X[ 5] += ROTL32(X[ 1]^X[13],12);
X[14] ^= ROTL32(X[10]+X[ 6], 7); X[ 2] += ROTL32(X[14]^X[10],15);
X[ 6] ^= ROTL32(X[ 2]+X[14],13); X[10] ^= ROTL32(X[ 6]+X[ 2],15);
X[ 3] += ROTL32(X[15]|X[11],20); X[ 7] ^= ROTL32(X[ 3]+X[15],16);
X[11] += ROTL32(X[ 7]^X[ 3], 7); X[15] += ROTL32(X[11]^X[ 7], 8);
X[ 1] += ROTL32(X[ 0]|X[ 3], 8)^i;X[ 2] ^= ROTL32(X[ 1]+X[ 0],14);
X[ 3] ^= ROTL32(X[ 2]+X[ 1], 6); X[ 0] += ROTL32(X[ 3]^X[ 2],18);
X[ 6] += ROTL32(X[ 5]^X[ 4], 8); X[ 7] += ROTL32(X[ 6]^X[ 5],12);
X[ 4] += ROTL32(X[ 7]|X[ 6],13); X[ 5] ^= ROTL32(X[ 4]+X[ 7],15);
X[11] ^= ROTL32(X[10]+X[ 9],18); X[ 8] += ROTL32(X[11]^X[10],11);
X[ 9] ^= ROTL32(X[ 8]+X[11], 8); X[10] += ROTL32(X[ 9]|X[ 8], 6);
X[12] += ROTL32(X[15]^X[14],17); X[13] ^= ROTL32(X[12]+X[15],15);
X[14] += ROTL32(X[13]|X[12], 9); X[15] += ROTL32(X[14]^X[13], 7);
}
for (i = 0; i < 16; ++i)
X[i] += mp->data[i];
/* CBC chaining on stream cipher blocks. */
for (i = 0; i < 16; i++)
mp->h[i] += X[i];
}
void sha1Process(sha1Param* sp)
{
register uint32_t a, b, c, d, e;
register uint32_t *w;
register byte t;
#if WORDS_BIGENDIAN
w = sp->data + 16;
#else
w = sp->data;
t = 16;
while (t--)
{
register uint32_t temp = swapu32(*w);
*(w++) = temp;
}
#endif
t = 64;
while (t--)
{
register uint32_t temp = w[-3] ^ w[-8] ^ w[-14] ^ w[-16];
*(w++) = ROTL32(temp, 1);
}
w = sp->data;
a = sp->h[0]; b = sp->h[1]; c = sp->h[2]; d = sp->h[3]; e = sp->h[4];
SUBROUND1(a,b,c,d,e,w[ 0],k[0]);
SUBROUND1(e,a,b,c,d,w[ 1],k[0]);
SUBROUND1(d,e,a,b,c,w[ 2],k[0]);
SUBROUND1(c,d,e,a,b,w[ 3],k[0]);
SUBROUND1(b,c,d,e,a,w[ 4],k[0]);
SUBROUND1(a,b,c,d,e,w[ 5],k[0]);
SUBROUND1(e,a,b,c,d,w[ 6],k[0]);
SUBROUND1(d,e,a,b,c,w[ 7],k[0]);
SUBROUND1(c,d,e,a,b,w[ 8],k[0]);
SUBROUND1(b,c,d,e,a,w[ 9],k[0]);
SUBROUND1(a,b,c,d,e,w[10],k[0]);
SUBROUND1(e,a,b,c,d,w[11],k[0]);
SUBROUND1(d,e,a,b,c,w[12],k[0]);
SUBROUND1(c,d,e,a,b,w[13],k[0]);
SUBROUND1(b,c,d,e,a,w[14],k[0]);
SUBROUND1(a,b,c,d,e,w[15],k[0]);
SUBROUND1(e,a,b,c,d,w[16],k[0]);
SUBROUND1(d,e,a,b,c,w[17],k[0]);
SUBROUND1(c,d,e,a,b,w[18],k[0]);
SUBROUND1(b,c,d,e,a,w[19],k[0]);
SUBROUND2(a,b,c,d,e,w[20],k[1]);
SUBROUND2(e,a,b,c,d,w[21],k[1]);
SUBROUND2(d,e,a,b,c,w[22],k[1]);
SUBROUND2(c,d,e,a,b,w[23],k[1]);
SUBROUND2(b,c,d,e,a,w[24],k[1]);
SUBROUND2(a,b,c,d,e,w[25],k[1]);
SUBROUND2(e,a,b,c,d,w[26],k[1]);
SUBROUND2(d,e,a,b,c,w[27],k[1]);
SUBROUND2(c,d,e,a,b,w[28],k[1]);
SUBROUND2(b,c,d,e,a,w[29],k[1]);
SUBROUND2(a,b,c,d,e,w[30],k[1]);
SUBROUND2(e,a,b,c,d,w[31],k[1]);
SUBROUND2(d,e,a,b,c,w[32],k[1]);
SUBROUND2(c,d,e,a,b,w[33],k[1]);
SUBROUND2(b,c,d,e,a,w[34],k[1]);
SUBROUND2(a,b,c,d,e,w[35],k[1]);
SUBROUND2(e,a,b,c,d,w[36],k[1]);
SUBROUND2(d,e,a,b,c,w[37],k[1]);
SUBROUND2(c,d,e,a,b,w[38],k[1]);
SUBROUND2(b,c,d,e,a,w[39],k[1]);
SUBROUND3(a,b,c,d,e,w[40],k[2]);
SUBROUND3(e,a,b,c,d,w[41],k[2]);
SUBROUND3(d,e,a,b,c,w[42],k[2]);
SUBROUND3(c,d,e,a,b,w[43],k[2]);
SUBROUND3(b,c,d,e,a,w[44],k[2]);
SUBROUND3(a,b,c,d,e,w[45],k[2]);
SUBROUND3(e,a,b,c,d,w[46],k[2]);
SUBROUND3(d,e,a,b,c,w[47],k[2]);
SUBROUND3(c,d,e,a,b,w[48],k[2]);
SUBROUND3(b,c,d,e,a,w[49],k[2]);
SUBROUND3(a,b,c,d,e,w[50],k[2]);
SUBROUND3(e,a,b,c,d,w[51],k[2]);
SUBROUND3(d,e,a,b,c,w[52],k[2]);
SUBROUND3(c,d,e,a,b,w[53],k[2]);
SUBROUND3(b,c,d,e,a,w[54],k[2]);
SUBROUND3(a,b,c,d,e,w[55],k[2]);
SUBROUND3(e,a,b,c,d,w[56],k[2]);
SUBROUND3(d,e,a,b,c,w[57],k[2]);
SUBROUND3(c,d,e,a,b,w[58],k[2]);
SUBROUND3(b,c,d,e,a,w[59],k[2]);
SUBROUND4(a,b,c,d,e,w[60],k[3]);
SUBROUND4(e,a,b,c,d,w[61],k[3]);
SUBROUND4(d,e,a,b,c,w[62],k[3]);
SUBROUND4(c,d,e,a,b,w[63],k[3]);
SUBROUND4(b,c,d,e,a,w[64],k[3]);
SUBROUND4(a,b,c,d,e,w[65],k[3]);
SUBROUND4(e,a,b,c,d,w[66],k[3]);
SUBROUND4(d,e,a,b,c,w[67],k[3]);
SUBROUND4(c,d,e,a,b,w[68],k[3]);
SUBROUND4(b,c,d,e,a,w[69],k[3]);
SUBROUND4(a,b,c,d,e,w[70],k[3]);
SUBROUND4(e,a,b,c,d,w[71],k[3]);
SUBROUND4(d,e,a,b,c,w[72],k[3]);
SUBROUND4(c,d,e,a,b,w[73],k[3]);
SUBROUND4(b,c,d,e,a,w[74],k[3]);
SUBROUND4(a,b,c,d,e,w[75],k[3]);
SUBROUND4(e,a,b,c,d,w[76],k[3]);
SUBROUND4(d,e,a,b,c,w[77],k[3]);
SUBROUND4(c,d,e,a,b,w[78],k[3]);
SUBROUND4(b,c,d,e,a,w[79],k[3]);
sp->h[0] += a;
sp->h[1] += b;
sp->h[2] += c;
sp->h[3] += d;
sp->h[4] += e;
}
void ripemd256Process(ripemd256Param* mp)
{
register uint32_t la, lb, lc, ld;
register uint32_t ra, rb, rc, rd;
register uint32_t temp;
register uint32_t* x;
#ifdef WORDS_BIGENDIAN
register byte t;
#endif
x = mp->data;
#ifdef WORDS_BIGENDIAN
t = 16;
while (t--)
{
temp = swapu32(*x);
*(x++) = temp;
}
x = mp->data;
#endif
la = mp->h[0]; lb = mp->h[1]; lc = mp->h[2]; ld = mp->h[3];
ra = mp->h[4]; rb = mp->h[5]; rc = mp->h[6]; rd = mp->h[7];
/* In theory OpenMP would allows us to do the 'left' and 'right' sections in parallel,
* but in practice the overhead make the code much slower
*/
/* left round 1 */
LSR1(la, lb, lc, ld, x[ 0], 11);
LSR1(ld, la, lb, lc, x[ 1], 14);
LSR1(lc, ld, la, lb, x[ 2], 15);
LSR1(lb, lc, ld, la, x[ 3], 12);
LSR1(la, lb, lc, ld, x[ 4], 5);
LSR1(ld, la, lb, lc, x[ 5], 8);
LSR1(lc, ld, la, lb, x[ 6], 7);
LSR1(lb, lc, ld, la, x[ 7], 9);
LSR1(la, lb, lc, ld, x[ 8], 11);
LSR1(ld, la, lb, lc, x[ 9], 13);
LSR1(lc, ld, la, lb, x[10], 14);
LSR1(lb, lc, ld, la, x[11], 15);
LSR1(la, lb, lc, ld, x[12], 6);
LSR1(ld, la, lb, lc, x[13], 7);
LSR1(lc, ld, la, lb, x[14], 9);
LSR1(lb, lc, ld, la, x[15], 8);
/* right round 1 */
RSR1(ra, rb, rc, rd, x[ 5], 8);
RSR1(rd, ra, rb, rc, x[14], 9);
RSR1(rc, rd, ra, rb, x[ 7], 9);
RSR1(rb, rc, rd, ra, x[ 0], 11);
RSR1(ra, rb, rc, rd, x[ 9], 13);
RSR1(rd, ra, rb, rc, x[ 2], 15);
RSR1(rc, rd, ra, rb, x[11], 15);
RSR1(rb, rc, rd, ra, x[ 4], 5);
RSR1(ra, rb, rc, rd, x[13], 7);
RSR1(rd, ra, rb, rc, x[ 6], 7);
RSR1(rc, rd, ra, rb, x[15], 8);
RSR1(rb, rc, rd, ra, x[ 8], 11);
RSR1(ra, rb, rc, rd, x[ 1], 14);
RSR1(rd, ra, rb, rc, x[10], 14);
RSR1(rc, rd, ra, rb, x[ 3], 12);
RSR1(rb, rc, rd, ra, x[12], 6);
temp = la; la = ra; ra = temp;
/* left round 2 */
LSR2(la, lb, lc, ld, x[ 7], 7);
LSR2(ld, la, lb, lc, x[ 4], 6);
LSR2(lc, ld, la, lb, x[13], 8);
LSR2(lb, lc, ld, la, x[ 1], 13);
LSR2(la, lb, lc, ld, x[10], 11);
LSR2(ld, la, lb, lc, x[ 6], 9);
LSR2(lc, ld, la, lb, x[15], 7);
LSR2(lb, lc, ld, la, x[ 3], 15);
LSR2(la, lb, lc, ld, x[12], 7);
LSR2(ld, la, lb, lc, x[ 0], 12);
LSR2(lc, ld, la, lb, x[ 9], 15);
LSR2(lb, lc, ld, la, x[ 5], 9);
LSR2(la, lb, lc, ld, x[ 2], 11);
LSR2(ld, la, lb, lc, x[14], 7);
LSR2(lc, ld, la, lb, x[11], 13);
LSR2(lb, lc, ld, la, x[ 8], 12);
/* right round 2 */
RSR2(ra, rb, rc, rd, x[ 6], 9);
RSR2(rd, ra, rb, rc, x[11], 13);
RSR2(rc, rd, ra, rb, x[ 3], 15);
RSR2(rb, rc, rd, ra, x[ 7], 7);
RSR2(ra, rb, rc, rd, x[ 0], 12);
RSR2(rd, ra, rb, rc, x[13], 8);
RSR2(rc, rd, ra, rb, x[ 5], 9);
RSR2(rb, rc, rd, ra, x[10], 11);
RSR2(ra, rb, rc, rd, x[14], 7);
RSR2(rd, ra, rb, rc, x[15], 7);
RSR2(rc, rd, ra, rb, x[ 8], 12);
RSR2(rb, rc, rd, ra, x[12], 7);
RSR2(ra, rb, rc, rd, x[ 4], 6);
RSR2(rd, ra, rb, rc, x[ 9], 15);
RSR2(rc, rd, ra, rb, x[ 1], 13);
RSR2(rb, rc, rd, ra, x[ 2], 11);
temp = lb; lb = rb; rb = temp;
/* left round 3 */
LSR3(la, lb, lc, ld, x[ 3], 11);
LSR3(ld, la, lb, lc, x[10], 13);
LSR3(lc, ld, la, lb, x[14], 6);
LSR3(lb, lc, ld, la, x[ 4], 7);
LSR3(la, lb, lc, ld, x[ 9], 14);
LSR3(ld, la, lb, lc, x[15], 9);
LSR3(lc, ld, la, lb, x[ 8], 13);
LSR3(lb, lc, ld, la, x[ 1], 15);
LSR3(la, lb, lc, ld, x[ 2], 14);
LSR3(ld, la, lb, lc, x[ 7], 8);
LSR3(lc, ld, la, lb, x[ 0], 13);
LSR3(lb, lc, ld, la, x[ 6], 6);
LSR3(la, lb, lc, ld, x[13], 5);
LSR3(ld, la, lb, lc, x[11], 12);
LSR3(lc, ld, la, lb, x[ 5], 7);
LSR3(lb, lc, ld, la, x[12], 5);
/* right round 3 */
RSR3(ra, rb, rc, rd, x[15], 9);
RSR3(rd, ra, rb, rc, x[ 5], 7);
RSR3(rc, rd, ra, rb, x[ 1], 15);
RSR3(rb, rc, rd, ra, x[ 3], 11);
RSR3(ra, rb, rc, rd, x[ 7], 8);
RSR3(rd, ra, rb, rc, x[14], 6);
RSR3(rc, rd, ra, rb, x[ 6], 6);
RSR3(rb, rc, rd, ra, x[ 9], 14);
RSR3(ra, rb, rc, rd, x[11], 12);
RSR3(rd, ra, rb, rc, x[ 8], 13);
RSR3(rc, rd, ra, rb, x[12], 5);
RSR3(rb, rc, rd, ra, x[ 2], 14);
RSR3(ra, rb, rc, rd, x[10], 13);
RSR3(rd, ra, rb, rc, x[ 0], 13);
RSR3(rc, rd, ra, rb, x[ 4], 7);
RSR3(rb, rc, rd, ra, x[13], 5);
temp = lc; lc = rc; rc = temp;
/* left round 4 */
LSR4(la, lb, lc, ld, x[ 1], 11);
LSR4(ld, la, lb, lc, x[ 9], 12);
LSR4(lc, ld, la, lb, x[11], 14);
LSR4(lb, lc, ld, la, x[10], 15);
LSR4(la, lb, lc, ld, x[ 0], 14);
LSR4(ld, la, lb, lc, x[ 8], 15);
LSR4(lc, ld, la, lb, x[12], 9);
LSR4(lb, lc, ld, la, x[ 4], 8);
LSR4(la, lb, lc, ld, x[13], 9);
LSR4(ld, la, lb, lc, x[ 3], 14);
LSR4(lc, ld, la, lb, x[ 7], 5);
LSR4(lb, lc, ld, la, x[15], 6);
LSR4(la, lb, lc, ld, x[14], 8);
LSR4(ld, la, lb, lc, x[ 5], 6);
LSR4(lc, ld, la, lb, x[ 6], 5);
LSR4(lb, lc, ld, la, x[ 2], 12);
/* right round 4 */
RSR4(ra, rb, rc, rd, x[ 8], 15);
RSR4(rd, ra, rb, rc, x[ 6], 5);
RSR4(rc, rd, ra, rb, x[ 4], 8);
RSR4(rb, rc, rd, ra, x[ 1], 11);
RSR4(ra, rb, rc, rd, x[ 3], 14);
RSR4(rd, ra, rb, rc, x[11], 14);
RSR4(rc, rd, ra, rb, x[15], 6);
RSR4(rb, rc, rd, ra, x[ 0], 14);
RSR4(ra, rb, rc, rd, x[ 5], 6);
RSR4(rd, ra, rb, rc, x[12], 9);
RSR4(rc, rd, ra, rb, x[ 2], 12);
RSR4(rb, rc, rd, ra, x[13], 9);
RSR4(ra, rb, rc, rd, x[ 9], 12);
RSR4(rd, ra, rb, rc, x[ 7], 5);
RSR4(rc, rd, ra, rb, x[10], 15);
RSR4(rb, rc, rd, ra, x[14], 8);
temp = ld; ld = rd; rd = temp;
/* combine results */
mp->h[0] += la;
mp->h[1] += lb;
mp->h[2] += lc;
mp->h[3] += ld;
mp->h[4] += ra;
mp->h[5] += rb;
mp->h[6] += rc;
mp->h[7] += rd;
}
void md4Process(md4Param *mp)
/*@modifies mp @*/
{
uint32_t a, b, c, d;
register uint32_t* w;
#ifdef WORDS_BIGENDIAN
register byte t;
#endif
w = mp->data;
#ifdef WORDS_BIGENDIAN
t = 16;
while (t--)
{
register uint32_t temp = swapu32(*w);
*(w++) = temp;
}
w = mp->data;
#endif
a = mp->h[0];
b = mp->h[1];
c = mp->h[2];
d = mp->h[3];
FF (a, b, c, d, w[ 0], 3);
FF (d, a, b, c, w[ 1], 7);
FF (c, d, a, b, w[ 2], 11);
FF (b, c, d, a, w[ 3], 19);
FF (a, b, c, d, w[ 4], 3);
FF (d, a, b, c, w[ 5], 7);
FF (c, d, a, b, w[ 6], 11);
FF (b, c, d, a, w[ 7], 19);
FF (a, b, c, d, w[ 8], 3);
FF (d, a, b, c, w[ 9], 7);
FF (c, d, a, b, w[10], 11);
FF (b, c, d, a, w[11], 19);
FF (a, b, c, d, w[12], 3);
FF (d, a, b, c, w[13], 7);
FF (c, d, a, b, w[14], 11);
FF (b, c, d, a, w[15], 19);
GG (a, b, c, d, w[ 0], 3);
GG (d, a, b, c, w[ 4], 5);
GG (c, d, a, b, w[ 8], 9);
GG (b, c, d, a, w[12], 13);
GG (a, b, c, d, w[ 1], 3);
GG (d, a, b, c, w[ 5], 5);
GG (c, d, a, b, w[ 9], 9);
GG (b, c, d, a, w[13], 13);
GG (a, b, c, d, w[ 2], 3);
GG (d, a, b, c, w[ 6], 5);
GG (c, d, a, b, w[10], 9);
GG (b, c, d, a, w[14], 13);
GG (a, b, c, d, w[ 3], 3);
GG (d, a, b, c, w[ 7], 5);
GG (c, d, a, b, w[11], 9);
GG (b, c, d, a, w[15], 13);
HH (a, b, c, d, w[ 0], 3);
HH (d, a, b, c, w[ 8], 9);
HH (c, d, a, b, w[ 4], 11);
HH (b, c, d, a, w[12], 15);
HH (a, b, c, d, w[ 2], 3);
HH (d, a, b, c, w[10], 9);
HH (c, d, a, b, w[ 6], 11);
HH (b, c, d, a, w[14], 15);
HH (a, b, c, d, w[ 1], 3);
HH (d, a, b, c, w[ 9], 9);
HH (c, d, a, b, w[ 5], 11);
HH (b, c, d, a, w[13], 15);
HH (a, b, c, d, w[ 3], 3);
HH (d, a, b, c, w[11], 9);
HH (c, d, a, b, w[ 7], 11);
HH (b, c, d, a, w[15], 15);
mp->h[0] += a;
mp->h[1] += b;
mp->h[2] += c;
mp->h[3] += d;
}
