mirror of
https://github.com/JayDDee/cpuminer-opt.git
synced 2025-09-17 23:44:27 +00:00
v3.10.0
This commit is contained in:
Jay D Dee
@@ -36,7 +36,6 @@
|
||||
#include <string.h>
|
||||
#include "skein-hash-4way.h"
|
||||
|
||||
|
||||
#ifdef __cplusplus
|
||||
extern "C"{
|
||||
#endif
|
||||
@@ -45,6 +44,22 @@ extern "C"{
|
||||
#pragma warning (disable: 4146)
|
||||
#endif
|
||||
|
||||
/*
|
||||
static const sph_u64 IV256[] = {
|
||||
SPH_C64(0xCCD044A12FDB3E13), SPH_C64(0xE83590301A79A9EB),
|
||||
SPH_C64(0x55AEA0614F816E6F), SPH_C64(0x2A2767A4AE9B94DB),
|
||||
SPH_C64(0xEC06025E74DD7683), SPH_C64(0xE7A436CDC4746251),
|
||||
SPH_C64(0xC36FBAF9393AD185), SPH_C64(0x3EEDBA1833EDFC13)
|
||||
};
|
||||
|
||||
static const sph_u64 IV512[] = {
|
||||
SPH_C64(0x4903ADFF749C51CE), SPH_C64(0x0D95DE399746DF03),
|
||||
SPH_C64(0x8FD1934127C79BCE), SPH_C64(0x9A255629FF352CB1),
|
||||
SPH_C64(0x5DB62599DF6CA7B0), SPH_C64(0xEABE394CA9D5C3F4),
|
||||
SPH_C64(0x991112C71A75B523), SPH_C64(0xAE18A40B660FCC33)
|
||||
};
|
||||
*/
|
||||
|
||||
/*
|
||||
* M9_ ## s ## _ ## i evaluates to s+i mod 9 (0 <= s <= 18, 0 <= i <= 7).
|
||||
*/
|
||||
@@ -270,8 +285,151 @@ extern "C"{
|
||||
#define SKBI(k, s, i) XCAT(k, XCAT(XCAT(XCAT(M9_, s), _), i))
|
||||
#define SKBT(t, s, v) XCAT(t, XCAT(XCAT(XCAT(M3_, s), _), v))
|
||||
|
||||
#define READ_STATE_BIG(sc) do { \
|
||||
h0 = (sc)->h0; \
|
||||
h1 = (sc)->h1; \
|
||||
h2 = (sc)->h2; \
|
||||
h3 = (sc)->h3; \
|
||||
h4 = (sc)->h4; \
|
||||
h5 = (sc)->h5; \
|
||||
h6 = (sc)->h6; \
|
||||
h7 = (sc)->h7; \
|
||||
bcount = sc->bcount; \
|
||||
} while (0)
|
||||
|
||||
#define WRITE_STATE_BIG(sc) do { \
|
||||
(sc)->h0 = h0; \
|
||||
(sc)->h1 = h1; \
|
||||
(sc)->h2 = h2; \
|
||||
(sc)->h3 = h3; \
|
||||
(sc)->h4 = h4; \
|
||||
(sc)->h5 = h5; \
|
||||
(sc)->h6 = h6; \
|
||||
(sc)->h7 = h7; \
|
||||
sc->bcount = bcount; \
|
||||
} while (0)
|
||||
|
||||
// AVX2 all scalar vars are now vectors representing 4 nonces in parallel
|
||||
|
||||
|
||||
#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
|
||||
|
||||
#define TFBIG_KINIT_8WAY( k0, k1, k2, k3, k4, k5, k6, k7, k8, t0, t1, t2 ) \
|
||||
do { \
|
||||
k8 = _mm512_xor_si512( _mm512_xor_si512( \
|
||||
_mm512_xor_si512( _mm512_xor_si512( k0, k1 ), \
|
||||
_mm512_xor_si512( k2, k3 ) ), \
|
||||
_mm512_xor_si512( _mm512_xor_si512( k4, k5 ), \
|
||||
_mm512_xor_si512( k6, k7 ) ) ), \
|
||||
m512_const1_64( 0x1BD11BDAA9FC1A22) ); \
|
||||
t2 = t0 ^ t1; \
|
||||
} while (0)
|
||||
|
||||
#define TFBIG_ADDKEY_8WAY(w0, w1, w2, w3, w4, w5, w6, w7, k, t, s) \
|
||||
do { \
|
||||
w0 = _mm512_add_epi64( w0, SKBI(k,s,0) ); \
|
||||
w1 = _mm512_add_epi64( w1, SKBI(k,s,1) ); \
|
||||
w2 = _mm512_add_epi64( w2, SKBI(k,s,2) ); \
|
||||
w3 = _mm512_add_epi64( w3, SKBI(k,s,3) ); \
|
||||
w4 = _mm512_add_epi64( w4, SKBI(k,s,4) ); \
|
||||
w5 = _mm512_add_epi64( w5, _mm512_add_epi64( SKBI(k,s,5), \
|
||||
m512_const1_64( SKBT(t,s,0) ) ) ); \
|
||||
w6 = _mm512_add_epi64( w6, _mm512_add_epi64( SKBI(k,s,6), \
|
||||
m512_const1_64( SKBT(t,s,1) ) ) ); \
|
||||
w7 = _mm512_add_epi64( w7, _mm512_add_epi64( SKBI(k,s,7), \
|
||||
m512_const1_64( s ) ) ); \
|
||||
} while (0)
|
||||
|
||||
|
||||
#define TFBIG_MIX_8WAY(x0, x1, rc) \
|
||||
do { \
|
||||
x0 = _mm512_add_epi64( x0, x1 ); \
|
||||
x1 = _mm512_xor_si512( mm512_rol_64( x1, rc ), x0 ); \
|
||||
} while (0)
|
||||
|
||||
#define TFBIG_MIX8_8WAY(w0, w1, w2, w3, w4, w5, w6, w7, rc0, rc1, rc2, rc3) do { \
|
||||
TFBIG_MIX_8WAY(w0, w1, rc0); \
|
||||
TFBIG_MIX_8WAY(w2, w3, rc1); \
|
||||
TFBIG_MIX_8WAY(w4, w5, rc2); \
|
||||
TFBIG_MIX_8WAY(w6, w7, rc3); \
|
||||
} while (0)
|
||||
|
||||
#define TFBIG_8WAY_4e(s) do { \
|
||||
TFBIG_ADDKEY_8WAY(p0, p1, p2, p3, p4, p5, p6, p7, h, t, s); \
|
||||
TFBIG_MIX8_8WAY(p0, p1, p2, p3, p4, p5, p6, p7, 46, 36, 19, 37); \
|
||||
TFBIG_MIX8_8WAY(p2, p1, p4, p7, p6, p5, p0, p3, 33, 27, 14, 42); \
|
||||
TFBIG_MIX8_8WAY(p4, p1, p6, p3, p0, p5, p2, p7, 17, 49, 36, 39); \
|
||||
TFBIG_MIX8_8WAY(p6, p1, p0, p7, p2, p5, p4, p3, 44, 9, 54, 56); \
|
||||
} while (0)
|
||||
|
||||
#define TFBIG_8WAY_4o(s) do { \
|
||||
TFBIG_ADDKEY_8WAY(p0, p1, p2, p3, p4, p5, p6, p7, h, t, s); \
|
||||
TFBIG_MIX8_8WAY(p0, p1, p2, p3, p4, p5, p6, p7, 39, 30, 34, 24); \
|
||||
TFBIG_MIX8_8WAY(p2, p1, p4, p7, p6, p5, p0, p3, 13, 50, 10, 17); \
|
||||
TFBIG_MIX8_8WAY(p4, p1, p6, p3, p0, p5, p2, p7, 25, 29, 39, 43); \
|
||||
TFBIG_MIX8_8WAY(p6, p1, p0, p7, p2, p5, p4, p3, 8, 35, 56, 22); \
|
||||
} while (0)
|
||||
|
||||
#define UBI_BIG_8WAY(etype, extra) \
|
||||
do { \
|
||||
sph_u64 t0, t1, t2; \
|
||||
__m512i h8; \
|
||||
__m512i m0 = buf[0]; \
|
||||
__m512i m1 = buf[1]; \
|
||||
__m512i m2 = buf[2]; \
|
||||
__m512i m3 = buf[3]; \
|
||||
__m512i m4 = buf[4]; \
|
||||
__m512i m5 = buf[5]; \
|
||||
__m512i m6 = buf[6]; \
|
||||
__m512i m7 = buf[7]; \
|
||||
\
|
||||
__m512i p0 = m0; \
|
||||
__m512i p1 = m1; \
|
||||
__m512i p2 = m2; \
|
||||
__m512i p3 = m3; \
|
||||
__m512i p4 = m4; \
|
||||
__m512i p5 = m5; \
|
||||
__m512i p6 = m6; \
|
||||
__m512i p7 = m7; \
|
||||
t0 = SPH_T64(bcount << 6) + (sph_u64)(extra); \
|
||||
t1 = (bcount >> 58) + ((sph_u64)(etype) << 55); \
|
||||
TFBIG_KINIT_8WAY(h0, h1, h2, h3, h4, h5, h6, h7, h8, t0, t1, t2); \
|
||||
TFBIG_8WAY_4e(0); \
|
||||
TFBIG_8WAY_4o(1); \
|
||||
TFBIG_8WAY_4e(2); \
|
||||
TFBIG_8WAY_4o(3); \
|
||||
TFBIG_8WAY_4e(4); \
|
||||
TFBIG_8WAY_4o(5); \
|
||||
TFBIG_8WAY_4e(6); \
|
||||
TFBIG_8WAY_4o(7); \
|
||||
TFBIG_8WAY_4e(8); \
|
||||
TFBIG_8WAY_4o(9); \
|
||||
TFBIG_8WAY_4e(10); \
|
||||
TFBIG_8WAY_4o(11); \
|
||||
TFBIG_8WAY_4e(12); \
|
||||
TFBIG_8WAY_4o(13); \
|
||||
TFBIG_8WAY_4e(14); \
|
||||
TFBIG_8WAY_4o(15); \
|
||||
TFBIG_8WAY_4e(16); \
|
||||
TFBIG_8WAY_4o(17); \
|
||||
TFBIG_ADDKEY_8WAY(p0, p1, p2, p3, p4, p5, p6, p7, h, t, 18); \
|
||||
h0 = _mm512_xor_si512( m0, p0 );\
|
||||
h1 = _mm512_xor_si512( m1, p1 );\
|
||||
h2 = _mm512_xor_si512( m2, p2 );\
|
||||
h3 = _mm512_xor_si512( m3, p3 );\
|
||||
h4 = _mm512_xor_si512( m4, p4 );\
|
||||
h5 = _mm512_xor_si512( m5, p5 );\
|
||||
h6 = _mm512_xor_si512( m6, p6 );\
|
||||
h7 = _mm512_xor_si512( m7, p7 );\
|
||||
} while (0)
|
||||
|
||||
#define DECL_STATE_BIG_8WAY \
|
||||
__m512i h0, h1, h2, h3, h4, h5, h6, h7; \
|
||||
sph_u64 bcount;
|
||||
|
||||
|
||||
#endif // AVX512
|
||||
|
||||
#define TFBIG_KINIT_4WAY( k0, k1, k2, k3, k4, k5, k6, k7, k8, t0, t1, t2 ) \
|
||||
do { \
|
||||
k8 = _mm256_xor_si256( _mm256_xor_si256( \
|
||||
@@ -298,39 +456,34 @@ do { \
|
||||
m256_const1_64( s ) ) ); \
|
||||
} while (0)
|
||||
|
||||
|
||||
#define TFBIG_MIX_4WAY(x0, x1, rc) \
|
||||
do { \
|
||||
x0 = _mm256_add_epi64( x0, x1 ); \
|
||||
x1 = _mm256_xor_si256( mm256_rol_64( x1, rc ), x0 ); \
|
||||
} while (0)
|
||||
|
||||
|
||||
// typeless
|
||||
#define TFBIG_MIX8(w0, w1, w2, w3, w4, w5, w6, w7, rc0, rc1, rc2, rc3) do { \
|
||||
TFBIG_MIX_4WAY(w0, w1, rc0); \
|
||||
TFBIG_MIX_4WAY(w2, w3, rc1); \
|
||||
TFBIG_MIX_4WAY(w4, w5, rc2); \
|
||||
TFBIG_MIX_4WAY(w6, w7, rc3); \
|
||||
} while (0)
|
||||
#define TFBIG_MIX8_4WAY(w0, w1, w2, w3, w4, w5, w6, w7, rc0, rc1, rc2, rc3) do { \
|
||||
TFBIG_MIX_4WAY(w0, w1, rc0); \
|
||||
TFBIG_MIX_4WAY(w2, w3, rc1); \
|
||||
TFBIG_MIX_4WAY(w4, w5, rc2); \
|
||||
TFBIG_MIX_4WAY(w6, w7, rc3); \
|
||||
} while (0)
|
||||
|
||||
#define TFBIG_4WAY_4e(s) do { \
|
||||
TFBIG_ADDKEY_4WAY(p0, p1, p2, p3, p4, p5, p6, p7, h, t, s); \
|
||||
TFBIG_MIX8_4WAY(p0, p1, p2, p3, p4, p5, p6, p7, 46, 36, 19, 37); \
|
||||
TFBIG_MIX8_4WAY(p2, p1, p4, p7, p6, p5, p0, p3, 33, 27, 14, 42); \
|
||||
TFBIG_MIX8_4WAY(p4, p1, p6, p3, p0, p5, p2, p7, 17, 49, 36, 39); \
|
||||
TFBIG_MIX8_4WAY(p6, p1, p0, p7, p2, p5, p4, p3, 44, 9, 54, 56); \
|
||||
} while (0)
|
||||
|
||||
#define TFBIG_4e(s) do { \
|
||||
TFBIG_ADDKEY_4WAY(p0, p1, p2, p3, p4, p5, p6, p7, h, t, s); \
|
||||
TFBIG_MIX8(p0, p1, p2, p3, p4, p5, p6, p7, 46, 36, 19, 37); \
|
||||
TFBIG_MIX8(p2, p1, p4, p7, p6, p5, p0, p3, 33, 27, 14, 42); \
|
||||
TFBIG_MIX8(p4, p1, p6, p3, p0, p5, p2, p7, 17, 49, 36, 39); \
|
||||
TFBIG_MIX8(p6, p1, p0, p7, p2, p5, p4, p3, 44, 9, 54, 56); \
|
||||
} while (0)
|
||||
|
||||
#define TFBIG_4o(s) do { \
|
||||
TFBIG_ADDKEY_4WAY(p0, p1, p2, p3, p4, p5, p6, p7, h, t, s); \
|
||||
TFBIG_MIX8(p0, p1, p2, p3, p4, p5, p6, p7, 39, 30, 34, 24); \
|
||||
TFBIG_MIX8(p2, p1, p4, p7, p6, p5, p0, p3, 13, 50, 10, 17); \
|
||||
TFBIG_MIX8(p4, p1, p6, p3, p0, p5, p2, p7, 25, 29, 39, 43); \
|
||||
TFBIG_MIX8(p6, p1, p0, p7, p2, p5, p4, p3, 8, 35, 56, 22); \
|
||||
} while (0)
|
||||
|
||||
#define TFBIG_4WAY_4o(s) do { \
|
||||
TFBIG_ADDKEY_4WAY(p0, p1, p2, p3, p4, p5, p6, p7, h, t, s); \
|
||||
TFBIG_MIX8_4WAY(p0, p1, p2, p3, p4, p5, p6, p7, 39, 30, 34, 24); \
|
||||
TFBIG_MIX8_4WAY(p2, p1, p4, p7, p6, p5, p0, p3, 13, 50, 10, 17); \
|
||||
TFBIG_MIX8_4WAY(p4, p1, p6, p3, p0, p5, p2, p7, 25, 29, 39, 43); \
|
||||
TFBIG_MIX8_4WAY(p6, p1, p0, p7, p2, p5, p4, p3, 8, 35, 56, 22); \
|
||||
} while (0)
|
||||
|
||||
// scale buf offset by 4
|
||||
#define UBI_BIG_4WAY(etype, extra) \
|
||||
@@ -357,24 +510,24 @@ do { \
|
||||
t0 = SPH_T64(bcount << 6) + (sph_u64)(extra); \
|
||||
t1 = (bcount >> 58) + ((sph_u64)(etype) << 55); \
|
||||
TFBIG_KINIT_4WAY(h0, h1, h2, h3, h4, h5, h6, h7, h8, t0, t1, t2); \
|
||||
TFBIG_4e(0); \
|
||||
TFBIG_4o(1); \
|
||||
TFBIG_4e(2); \
|
||||
TFBIG_4o(3); \
|
||||
TFBIG_4e(4); \
|
||||
TFBIG_4o(5); \
|
||||
TFBIG_4e(6); \
|
||||
TFBIG_4o(7); \
|
||||
TFBIG_4e(8); \
|
||||
TFBIG_4o(9); \
|
||||
TFBIG_4e(10); \
|
||||
TFBIG_4o(11); \
|
||||
TFBIG_4e(12); \
|
||||
TFBIG_4o(13); \
|
||||
TFBIG_4e(14); \
|
||||
TFBIG_4o(15); \
|
||||
TFBIG_4e(16); \
|
||||
TFBIG_4o(17); \
|
||||
TFBIG_4WAY_4e(0); \
|
||||
TFBIG_4WAY_4o(1); \
|
||||
TFBIG_4WAY_4e(2); \
|
||||
TFBIG_4WAY_4o(3); \
|
||||
TFBIG_4WAY_4e(4); \
|
||||
TFBIG_4WAY_4o(5); \
|
||||
TFBIG_4WAY_4e(6); \
|
||||
TFBIG_4WAY_4o(7); \
|
||||
TFBIG_4WAY_4e(8); \
|
||||
TFBIG_4WAY_4o(9); \
|
||||
TFBIG_4WAY_4e(10); \
|
||||
TFBIG_4WAY_4o(11); \
|
||||
TFBIG_4WAY_4e(12); \
|
||||
TFBIG_4WAY_4o(13); \
|
||||
TFBIG_4WAY_4e(14); \
|
||||
TFBIG_4WAY_4o(15); \
|
||||
TFBIG_4WAY_4e(16); \
|
||||
TFBIG_4WAY_4o(17); \
|
||||
TFBIG_ADDKEY_4WAY(p0, p1, p2, p3, p4, p5, p6, p7, h, t, 18); \
|
||||
h0 = _mm256_xor_si256( m0, p0 );\
|
||||
h1 = _mm256_xor_si256( m1, p1 );\
|
||||
@@ -391,45 +544,142 @@ do { \
|
||||
__m256i h0, h1, h2, h3, h4, h5, h6, h7; \
|
||||
sph_u64 bcount;
|
||||
|
||||
#define READ_STATE_BIG(sc) do { \
|
||||
h0 = (sc)->h0; \
|
||||
h1 = (sc)->h1; \
|
||||
h2 = (sc)->h2; \
|
||||
h3 = (sc)->h3; \
|
||||
h4 = (sc)->h4; \
|
||||
h5 = (sc)->h5; \
|
||||
h6 = (sc)->h6; \
|
||||
h7 = (sc)->h7; \
|
||||
bcount = sc->bcount; \
|
||||
} while (0)
|
||||
#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
|
||||
|
||||
#define WRITE_STATE_BIG(sc) do { \
|
||||
(sc)->h0 = h0; \
|
||||
(sc)->h1 = h1; \
|
||||
(sc)->h2 = h2; \
|
||||
(sc)->h3 = h3; \
|
||||
(sc)->h4 = h4; \
|
||||
(sc)->h5 = h5; \
|
||||
(sc)->h6 = h6; \
|
||||
(sc)->h7 = h7; \
|
||||
sc->bcount = bcount; \
|
||||
} while (0)
|
||||
void skein256_8way_init( skein256_8way_context *sc )
|
||||
{
|
||||
sc->h0 = m512_const1_64( 0xCCD044A12FDB3E13 );
|
||||
sc->h1 = m512_const1_64( 0xE83590301A79A9EB );
|
||||
sc->h2 = m512_const1_64( 0x55AEA0614F816E6F );
|
||||
sc->h3 = m512_const1_64( 0x2A2767A4AE9B94DB );
|
||||
sc->h4 = m512_const1_64( 0xEC06025E74DD7683 );
|
||||
sc->h5 = m512_const1_64( 0xE7A436CDC4746251 );
|
||||
sc->h6 = m512_const1_64( 0xC36FBAF9393AD185 );
|
||||
sc->h7 = m512_const1_64( 0x3EEDBA1833EDFC13 );
|
||||
sc->bcount = 0;
|
||||
sc->ptr = 0;
|
||||
}
|
||||
|
||||
/*
|
||||
static const sph_u64 IV256[] = {
|
||||
SPH_C64(0xCCD044A12FDB3E13), SPH_C64(0xE83590301A79A9EB),
|
||||
SPH_C64(0x55AEA0614F816E6F), SPH_C64(0x2A2767A4AE9B94DB),
|
||||
SPH_C64(0xEC06025E74DD7683), SPH_C64(0xE7A436CDC4746251),
|
||||
SPH_C64(0xC36FBAF9393AD185), SPH_C64(0x3EEDBA1833EDFC13)
|
||||
};
|
||||
void skein512_8way_init( skein512_8way_context *sc )
|
||||
{
|
||||
sc->h0 = m512_const1_64( 0x4903ADFF749C51CE );
|
||||
sc->h1 = m512_const1_64( 0x0D95DE399746DF03 );
|
||||
sc->h2 = m512_const1_64( 0x8FD1934127C79BCE );
|
||||
sc->h3 = m512_const1_64( 0x9A255629FF352CB1 );
|
||||
sc->h4 = m512_const1_64( 0x5DB62599DF6CA7B0 );
|
||||
sc->h5 = m512_const1_64( 0xEABE394CA9D5C3F4 );
|
||||
sc->h6 = m512_const1_64( 0x991112C71A75B523 );
|
||||
sc->h7 = m512_const1_64( 0xAE18A40B660FCC33 );
|
||||
sc->bcount = 0;
|
||||
sc->ptr = 0;
|
||||
}
|
||||
|
||||
static void
|
||||
skein_big_core_8way( skein512_8way_context *sc, const void *data,
|
||||
size_t len )
|
||||
{
|
||||
__m512i *vdata = (__m512i*)data;
|
||||
__m512i *buf;
|
||||
size_t ptr;
|
||||
unsigned first;
|
||||
DECL_STATE_BIG_8WAY
|
||||
|
||||
buf = sc->buf;
|
||||
ptr = sc->ptr;
|
||||
const int buf_size = 64; // 64 * _m256i
|
||||
|
||||
if ( len <= buf_size - ptr )
|
||||
{
|
||||
memcpy_512( buf + (ptr>>3), vdata, len>>3 );
|
||||
sc->ptr = ptr + len;
|
||||
return;
|
||||
}
|
||||
|
||||
READ_STATE_BIG( sc );
|
||||
first = ( bcount == 0 ) << 7;
|
||||
do {
|
||||
size_t clen;
|
||||
|
||||
if ( ptr == buf_size )
|
||||
{
|
||||
bcount ++;
|
||||
UBI_BIG_8WAY( 96 + first, 0 );
|
||||
first = 0;
|
||||
ptr = 0;
|
||||
}
|
||||
clen = buf_size - ptr;
|
||||
if ( clen > len )
|
||||
clen = len;
|
||||
memcpy_512( buf + (ptr>>3), vdata, clen>>3 );
|
||||
ptr += clen;
|
||||
vdata += (clen>>3);
|
||||
len -= clen;
|
||||
} while ( len > 0 );
|
||||
WRITE_STATE_BIG( sc );
|
||||
sc->ptr = ptr;
|
||||
}
|
||||
|
||||
static void
|
||||
skein_big_close_8way( skein512_8way_context *sc, unsigned ub, unsigned n,
|
||||
void *dst, size_t out_len )
|
||||
{
|
||||
__m512i *buf;
|
||||
size_t ptr;
|
||||
unsigned et;
|
||||
DECL_STATE_BIG_8WAY
|
||||
|
||||
buf = sc->buf;
|
||||
ptr = sc->ptr;
|
||||
const int buf_size = 64;
|
||||
|
||||
READ_STATE_BIG(sc);
|
||||
|
||||
memset_zero_512( buf + (ptr>>3), (buf_size - ptr) >> 3 );
|
||||
et = 352 + ((bcount == 0) << 7);
|
||||
UBI_BIG_8WAY( et, ptr );
|
||||
|
||||
memset_zero_512( buf, buf_size >> 3 );
|
||||
bcount = 0;
|
||||
UBI_BIG_8WAY( 510, 8 );
|
||||
|
||||
buf[0] = h0;
|
||||
buf[1] = h1;
|
||||
buf[2] = h2;
|
||||
buf[3] = h3;
|
||||
buf[4] = h4;
|
||||
buf[5] = h5;
|
||||
buf[6] = h6;
|
||||
buf[7] = h7;
|
||||
|
||||
memcpy_512( dst, buf, out_len >> 3 );
|
||||
}
|
||||
|
||||
void
|
||||
skein256_8way_update(void *cc, const void *data, size_t len)
|
||||
{
|
||||
skein_big_core_8way(cc, data, len);
|
||||
}
|
||||
|
||||
void
|
||||
skein256_8way_close(void *cc, void *dst)
|
||||
{
|
||||
skein_big_close_8way(cc, 0, 0, dst, 32);
|
||||
}
|
||||
|
||||
void
|
||||
skein512_8way_update(void *cc, const void *data, size_t len)
|
||||
{
|
||||
skein_big_core_8way(cc, data, len);
|
||||
}
|
||||
|
||||
void
|
||||
skein512_8way_close(void *cc, void *dst)
|
||||
{
|
||||
skein_big_close_8way(cc, 0, 0, dst, 64);
|
||||
}
|
||||
|
||||
#endif // AVX512
|
||||
|
||||
static const sph_u64 IV512[] = {
|
||||
SPH_C64(0x4903ADFF749C51CE), SPH_C64(0x0D95DE399746DF03),
|
||||
SPH_C64(0x8FD1934127C79BCE), SPH_C64(0x9A255629FF352CB1),
|
||||
SPH_C64(0x5DB62599DF6CA7B0), SPH_C64(0xEABE394CA9D5C3F4),
|
||||
SPH_C64(0x991112C71A75B523), SPH_C64(0xAE18A40B660FCC33)
|
||||
};
|
||||
*/
|
||||
|
||||
void skein256_4way_init( skein256_4way_context *sc )
|
||||
{
|
||||
@@ -517,66 +767,30 @@ skein_big_close_4way( skein512_4way_context *sc, unsigned ub, unsigned n,
|
||||
ptr = sc->ptr;
|
||||
const int buf_size = 64;
|
||||
|
||||
/*
|
||||
* At that point, if ptr == 0, then the message was empty;
|
||||
* otherwise, there is between 1 and 64 bytes (inclusive) which
|
||||
* are yet to be processed. Either way, we complete the buffer
|
||||
* to a full block with zeros (the Skein specification mandates
|
||||
* that an empty message is padded so that there is at least
|
||||
* one block to process).
|
||||
*
|
||||
* Once this block has been processed, we do it again, with
|
||||
* a block full of zeros, for the output (that block contains
|
||||
* the encoding of "0", over 8 bytes, then padded with zeros).
|
||||
*/
|
||||
|
||||
READ_STATE_BIG(sc);
|
||||
|
||||
memset_zero_256( buf + (ptr>>3), (buf_size - ptr) >> 3 );
|
||||
memset_zero_256( buf + (ptr>>3), (buf_size - ptr) >> 3 );
|
||||
et = 352 + ((bcount == 0) << 7);
|
||||
UBI_BIG_4WAY( et, ptr );
|
||||
UBI_BIG_4WAY( et, ptr );
|
||||
|
||||
memset_zero_256( buf, buf_size >> 3 );
|
||||
bcount = 0;
|
||||
UBI_BIG_4WAY( 510, 8 );
|
||||
memset_zero_256( buf, buf_size >> 3 );
|
||||
bcount = 0;
|
||||
UBI_BIG_4WAY( 510, 8 );
|
||||
|
||||
buf[0] = h0;
|
||||
buf[1] = h1;
|
||||
buf[2] = h2;
|
||||
buf[3] = h3;
|
||||
buf[4] = h4;
|
||||
buf[5] = h5;
|
||||
buf[6] = h6;
|
||||
buf[7] = h7;
|
||||
buf[0] = h0;
|
||||
buf[1] = h1;
|
||||
buf[2] = h2;
|
||||
buf[3] = h3;
|
||||
buf[4] = h4;
|
||||
buf[5] = h5;
|
||||
buf[6] = h6;
|
||||
buf[7] = h7;
|
||||
|
||||
memcpy_256( dst, buf, out_len >> 3 );
|
||||
memcpy_256( dst, buf, out_len >> 3 );
|
||||
}
|
||||
|
||||
/*
|
||||
static const sph_u64 IV256[] = {
|
||||
SPH_C64(0xCCD044A12FDB3E13), SPH_C64(0xE83590301A79A9EB),
|
||||
SPH_C64(0x55AEA0614F816E6F), SPH_C64(0x2A2767A4AE9B94DB),
|
||||
SPH_C64(0xEC06025E74DD7683), SPH_C64(0xE7A436CDC4746251),
|
||||
SPH_C64(0xC36FBAF9393AD185), SPH_C64(0x3EEDBA1833EDFC13)
|
||||
};
|
||||
|
||||
static const sph_u64 IV512[] = {
|
||||
SPH_C64(0x4903ADFF749C51CE), SPH_C64(0x0D95DE399746DF03),
|
||||
SPH_C64(0x8FD1934127C79BCE), SPH_C64(0x9A255629FF352CB1),
|
||||
SPH_C64(0x5DB62599DF6CA7B0), SPH_C64(0xEABE394CA9D5C3F4),
|
||||
SPH_C64(0x991112C71A75B523), SPH_C64(0xAE18A40B660FCC33)
|
||||
};
|
||||
*/
|
||||
/*
|
||||
void
|
||||
skein256_4way_init(void *cc)
|
||||
{
|
||||
skein_big_init_4way(cc, IV256);
|
||||
}
|
||||
*/
|
||||
|
||||
void
|
||||
skein256_4way(void *cc, const void *data, size_t len)
|
||||
skein256_4way_update(void *cc, const void *data, size_t len)
|
||||
{
|
||||
skein_big_core_4way(cc, data, len);
|
||||
}
|
||||
@@ -587,16 +801,8 @@ skein256_4way_close(void *cc, void *dst)
|
||||
skein_big_close_4way(cc, 0, 0, dst, 32);
|
||||
}
|
||||
|
||||
/*
|
||||
void
|
||||
skein512_4way_init(void *cc)
|
||||
{
|
||||
skein_big_init_4way(cc, IV512);
|
||||
}
|
||||
*/
|
||||
|
||||
void
|
||||
skein512_4way(void *cc, const void *data, size_t len)
|
||||
skein512_4way_update(void *cc, const void *data, size_t len)
|
||||
{
|
||||
skein_big_core_4way(cc, data, len);
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user