static void
hamsi_small(sph_hamsi_small_context *sc, const unsigned char *buf, size_t num)
{
DECL_STATE_SMALL
#if !SPH_64
sph_u32 tmp;
#endif
#if SPH_64
sc->count += (sph_u64)num << 5;
#else
tmp = SPH_T32((sph_u32)num << 5);
sc->count_low = SPH_T32(sc->count_low + tmp);
sc->count_high += (sph_u32)((num >> 13) >> 14);
if (sc->count_low < tmp)
sc->count_high ++;
#endif
READ_STATE_SMALL(sc);
while (num -- > 0) {
sph_u32 m0, m1, m2, m3, m4, m5, m6, m7;
INPUT_SMALL;
P_SMALL;
T_SMALL;
buf += 4;
}
WRITE_STATE_SMALL(sc);
}
static void
fugue2_core(sph_fugue_context *sc, const void *data, size_t len)
{
DECL_STATE_SMALL
CORE_ENTRY
READ_STATE_SMALL(sc);
rshift = sc->round_shift;
switch (rshift) {
for (;;) {
sph_u32 q;
case 0:
q = p;
TIX2(q, S00, S01, S08, S10, S24);
CMIX30(S27, S28, S29, S01, S02, S03, S12, S13, S14);
SMIX(S27, S28, S29, S00);
CMIX30(S24, S25, S26, S28, S29, S00, S09, S10, S11);
SMIX(S24, S25, S26, S27);
NEXT(1);
/* fall through */
case 1:
q = p;
TIX2(q, S24, S25, S02, S04, S18);
CMIX30(S21, S22, S23, S25, S26, S27, S06, S07, S08);
SMIX(S21, S22, S23, S24);
CMIX30(S18, S19, S20, S22, S23, S24, S03, S04, S05);
SMIX(S18, S19, S20, S21);
NEXT(2);
/* fall through */
case 2:
q = p;
TIX2(q, S18, S19, S26, S28, S12);
CMIX30(S15, S16, S17, S19, S20, S21, S00, S01, S02);
SMIX(S15, S16, S17, S18);
CMIX30(S12, S13, S14, S16, S17, S18, S27, S28, S29);
SMIX(S12, S13, S14, S15);
NEXT(3);
/* fall through */
case 3:
q = p;
TIX2(q, S12, S13, S20, S22, S06);
CMIX30(S09, S10, S11, S13, S14, S15, S24, S25, S26);
SMIX(S09, S10, S11, S12);
CMIX30(S06, S07, S08, S10, S11, S12, S21, S22, S23);
SMIX(S06, S07, S08, S09);
NEXT(4);
/* fall through */
case 4:
q = p;
TIX2(q, S06, S07, S14, S16, S00);
CMIX30(S03, S04, S05, S07, S08, S09, S18, S19, S20);
SMIX(S03, S04, S05, S06);
CMIX30(S00, S01, S02, S04, S05, S06, S15, S16, S17);
SMIX(S00, S01, S02, S03);
NEXT(0);
}
}
CORE_EXIT
WRITE_STATE_SMALL(sc);
}
static void
hamsi_small_final(sph_hamsi_small_context *sc, const unsigned char *buf)
{
sph_u32 m0, m1, m2, m3, m4, m5, m6, m7;
DECL_STATE_SMALL
READ_STATE_SMALL(sc);
INPUT_SMALL;
PF_SMALL;
T_SMALL;
WRITE_STATE_SMALL(sc);
}
/* obsolete */
static void
skein_small_core(sph_skein_small_context *sc, const void *data, size_t len)
{
unsigned char *buf;
size_t ptr, clen;
unsigned first;
DECL_STATE_SMALL
buf = sc->buf;
ptr = sc->ptr;
clen = (sizeof sc->buf) - ptr;
if (len <= clen) {
memcpy(buf + ptr, data, len);
sc->ptr = ptr + len;
return;
}
if (clen != 0) {
memcpy(buf + ptr, data, clen);
data = (const unsigned char *)data + clen;
len -= clen;
}
#if SPH_SMALL_FOOTPRINT_SKEIN
READ_STATE_SMALL(sc);
first = (bcount == 0) << 7;
for (;;) {
bcount ++;
UBI_SMALL(96 + first, 0);
if (len <= sizeof sc->buf)
break;
first = 0;
memcpy(buf, data, sizeof sc->buf);
data = (const unsigned char *)data + sizeof sc->buf;
len -= sizeof sc->buf;
}
WRITE_STATE_SMALL(sc);
sc->ptr = len;
memcpy(buf, data, len);
#else
/*
* Unrolling the loop yields a slight performance boost, while
* keeping the code size aorund 24 kB on 32-bit x86.
*/
READ_STATE_SMALL(sc);
first = (bcount == 0) << 7;
for (;;) {
bcount ++;
UBI_SMALL(96 + first, 0);
if (len <= sizeof sc->buf)
break;
buf = (unsigned char *)data;
bcount ++;
UBI_SMALL(96, 0);
if (len <= 2 * sizeof sc->buf) {
data = buf + sizeof sc->buf;
len -= sizeof sc->buf;
break;
}
buf += sizeof sc->buf;
data = buf + sizeof sc->buf;
first = 0;
len -= 2 * sizeof sc->buf;
}
WRITE_STATE_SMALL(sc);
sc->ptr = len;
memcpy(sc->buf, data, len);
#endif
}