popov/cpuminer-opt-gpu
1
0
Fork 0
mirror of https://github.com/JayDDee/cpuminer-opt.git synced 2025-09-17 23:44:27 +00:00
Code Issues Packages Projects Releases Wiki Activity

v3.10.7

Browse Source
This commit is contained in:
Jay D Dee
2019-12-28 15:00:29 -05:00
parent 241bc26767
commit b47cfaa720
75 changed files with 1884 additions and 773 deletions
Download Patch File Download Diff File Expand all files Collapse all files

38
algo/sha/sha-hash-4way.h
View File

@@ -41,13 +41,9 @@
#define SHA2_HASH_4WAY_H__ 1
#include <stddef.h>
#include "sph_types.h"
#include "simd-utils.h"
#if defined(__SSE2__)
//#if defined(__SSE4_2__)
//#define SPH_SIZE_sha256 256
// SHA-256 4 way
@@ -59,9 +55,12 @@ typedef struct {
} sha256_4way_context __attribute__ ((aligned (64)));
void sha256_4way_init( sha256_4way_context *sc );
void sha256_4way( sha256_4way_context *sc, const void *data, size_t len );
void sha256_4way_update( sha256_4way_context *sc, const void *data,
size_t len );
void sha256_4way_close( sha256_4way_context *sc, void *dst );
#endif // SSE2
#if defined (__AVX2__)
// SHA-256 8 way
@@ -75,10 +74,28 @@ typedef struct {
void sha256_8way_init( sha256_8way_context *sc );
void sha256_8way_update( sha256_8way_context *sc, const void *data, size_t len );
#define sha256_8way sha256_8way_update
void sha256_8way_close( sha256_8way_context *sc, void *dst );
//#define SPH_SIZE_sha512 512
#endif // AVX2
#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
// SHA-256 16 way
typedef struct {
__m512i buf[64>>2];
__m512i val[8];
uint32_t count_high, count_low;
bool initialized;
} sha256_16way_context __attribute__ ((aligned (128)));
void sha256_16way_init( sha256_16way_context *sc );
void sha256_16way_update( sha256_16way_context *sc, const void *data, size_t len );
void sha256_16way_close( sha256_16way_context *sc, void *dst );
#endif // AVX512
#if defined (__AVX2__)
// SHA-512 4 way
@@ -92,9 +109,10 @@ typedef struct {
void sha512_4way_init( sha512_4way_context *sc);
void sha512_4way_update( sha512_4way_context *sc, const void *data,
size_t len );
#define sha512_4way sha512_4way_update
void sha512_4way_close( sha512_4way_context *sc, void *dst );
#endif // AVX2
#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
// SHA-512 8 way
@@ -111,8 +129,6 @@ void sha512_8way_update( sha512_8way_context *sc, const void *data,
size_t len );
void sha512_8way_close( sha512_8way_context *sc, void *dst );
#endif // AVX512
#endif // __AVX2__
#endif // __SSE2__
#endif // SHA256_4WAY_H__

298
algo/sha/sha256-hash-4way.c
View File

@@ -39,47 +39,31 @@
// SHA-256 32 bit
/*
static const sph_u32 H256[8] = {
SPH_C32(0x6A09E667), SPH_C32(0xBB67AE85),
SPH_C32(0x3C6EF372), SPH_C32(0xA54FF53A),
SPH_C32(0x510E527F), SPH_C32(0x9B05688C),
SPH_C32(0x1F83D9AB), SPH_C32(0x5BE0CD19)
static const uint32_t H256[8] =
{
0x6A09E667, 0xBB67AE85, 0x3C6EF372, 0xA54FF53A,
0x510E527F, 0x9B05688C, 0x1F83D9AB, 0x5BE0CD19
};
*/
static const sph_u32 K256[64] = {
SPH_C32(0x428A2F98), SPH_C32(0x71374491),
SPH_C32(0xB5C0FBCF), SPH_C32(0xE9B5DBA5),
SPH_C32(0x3956C25B), SPH_C32(0x59F111F1),
SPH_C32(0x923F82A4), SPH_C32(0xAB1C5ED5),
SPH_C32(0xD807AA98), SPH_C32(0x12835B01),
SPH_C32(0x243185BE), SPH_C32(0x550C7DC3),
SPH_C32(0x72BE5D74), SPH_C32(0x80DEB1FE),
SPH_C32(0x9BDC06A7), SPH_C32(0xC19BF174),
SPH_C32(0xE49B69C1), SPH_C32(0xEFBE4786),
SPH_C32(0x0FC19DC6), SPH_C32(0x240CA1CC),
SPH_C32(0x2DE92C6F), SPH_C32(0x4A7484AA),
SPH_C32(0x5CB0A9DC), SPH_C32(0x76F988DA),
SPH_C32(0x983E5152), SPH_C32(0xA831C66D),
SPH_C32(0xB00327C8), SPH_C32(0xBF597FC7),
SPH_C32(0xC6E00BF3), SPH_C32(0xD5A79147),
SPH_C32(0x06CA6351), SPH_C32(0x14292967),
SPH_C32(0x27B70A85), SPH_C32(0x2E1B2138),
SPH_C32(0x4D2C6DFC), SPH_C32(0x53380D13),
SPH_C32(0x650A7354), SPH_C32(0x766A0ABB),
SPH_C32(0x81C2C92E), SPH_C32(0x92722C85),
SPH_C32(0xA2BFE8A1), SPH_C32(0xA81A664B),
SPH_C32(0xC24B8B70), SPH_C32(0xC76C51A3),
SPH_C32(0xD192E819), SPH_C32(0xD6990624),
SPH_C32(0xF40E3585), SPH_C32(0x106AA070),
SPH_C32(0x19A4C116), SPH_C32(0x1E376C08),
SPH_C32(0x2748774C), SPH_C32(0x34B0BCB5),
SPH_C32(0x391C0CB3), SPH_C32(0x4ED8AA4A),
SPH_C32(0x5B9CCA4F), SPH_C32(0x682E6FF3),
SPH_C32(0x748F82EE), SPH_C32(0x78A5636F),
SPH_C32(0x84C87814), SPH_C32(0x8CC70208),
SPH_C32(0x90BEFFFA), SPH_C32(0xA4506CEB),
SPH_C32(0xBEF9A3F7), SPH_C32(0xC67178F2)
static const uint32_t K256[64] =
{
0x428A2F98, 0x71374491, 0xB5C0FBCF, 0xE9B5DBA5,
0x3956C25B, 0x59F111F1, 0x923F82A4, 0xAB1C5ED5,
0xD807AA98, 0x12835B01, 0x243185BE, 0x550C7DC3,
0x72BE5D74, 0x80DEB1FE, 0x9BDC06A7, 0xC19BF174,
0xE49B69C1, 0xEFBE4786, 0x0FC19DC6, 0x240CA1CC,
0x2DE92C6F, 0x4A7484AA, 0x5CB0A9DC, 0x76F988DA,
0x983E5152, 0xA831C66D, 0xB00327C8, 0xBF597FC7,
0xC6E00BF3, 0xD5A79147, 0x06CA6351, 0x14292967,
0x27B70A85, 0x2E1B2138, 0x4D2C6DFC, 0x53380D13,
0x650A7354, 0x766A0ABB, 0x81C2C92E, 0x92722C85,
0xA2BFE8A1, 0xA81A664B, 0xC24B8B70, 0xC76C51A3,
0xD192E819, 0xD6990624, 0xF40E3585, 0x106AA070,
0x19A4C116, 0x1E376C08, 0x2748774C, 0x34B0BCB5,
0x391C0CB3, 0x4ED8AA4A, 0x5B9CCA4F, 0x682E6FF3,
0x748F82EE, 0x78A5636F, 0x84C87814, 0x8CC70208,
0x90BEFFFA, 0xA4506CEB, 0xBEF9A3F7, 0xC67178F2
};
// SHA-256 4 way
@@ -248,7 +232,7 @@ void sha256_4way_init( sha256_4way_context *sc )
*/
}
void sha256_4way( sha256_4way_context *sc, const void *data, size_t len )
void sha256_4way_update( sha256_4way_context *sc, const void *data, size_t len )
{
__m128i *vdata = (__m128i*)data;
size_t ptr;
@@ -273,7 +257,7 @@ void sha256_4way( sha256_4way_context *sc, const void *data, size_t len )
ptr = 0;
}
clow = sc->count_low;
clow2 = SPH_T32( clow + clen );
clow2 = clow + clen;
sc->count_low = clow2;
if ( clow2 < clow )
sc->count_high++;
@@ -306,10 +290,8 @@ void sha256_4way_close( sha256_4way_context *sc, void *dst )
sc->buf[ pad >> 2 ] =
mm128_bswap_32( m128_const1_32( high ) );
// mm128_bswap_32( _mm_set1_epi32( high ) );
sc->buf[ ( pad+4 ) >> 2 ] =
mm128_bswap_32( m128_const1_32( low ) );
// mm128_bswap_32( _mm_set1_epi32( low ) );
sha256_4way_round( sc, sc->buf, sc->val );
mm128_block_bswap_32( dst, sc->val );
@@ -483,7 +465,7 @@ void sha256_8way_init( sha256_8way_context *sc )
*/
}
void sha256_8way( sha256_8way_context *sc, const void *data, size_t len )
void sha256_8way_update( sha256_8way_context *sc, const void *data, size_t len )
{
__m256i *vdata = (__m256i*)data;
size_t ptr;
@@ -508,7 +490,7 @@ void sha256_8way( sha256_8way_context *sc, const void *data, size_t len )
ptr = 0;
}
clow = sc->count_low;
clow2 = SPH_T32( clow + clen );
clow2 = clow + clen;
sc->count_low = clow2;
if ( clow2 < clow )
sc->count_high++;
@@ -549,5 +531,233 @@ void sha256_8way_close( sha256_8way_context *sc, void *dst )
mm256_block_bswap_32( dst, sc->val );
}
#if defined(__AVX512F__) && defined(__AVX512VL__) && defined(__AVX512DQ__) && defined(__AVX512BW__)
// SHA-256 16 way
#define CHx16(X, Y, Z) \
_mm512_xor_si512( _mm512_and_si512( _mm512_xor_si512( Y, Z ), X ), Z )
#define MAJx16(X, Y, Z) \
_mm512_or_si512( _mm512_and_si512( X, Y ), \
_mm512_and_si512( _mm512_or_si512( X, Y ), Z ) )
#define BSG2_0x16(x) \
_mm512_xor_si512( _mm512_xor_si512( \
mm512_ror_32(x, 2), mm512_ror_32(x, 13) ), mm512_ror_32( x, 22) )
#define BSG2_1x16(x) \
_mm512_xor_si512( _mm512_xor_si512( \
mm512_ror_32(x, 6), mm512_ror_32(x, 11) ), mm512_ror_32( x, 25) )
#define SSG2_0x16(x) \
_mm512_xor_si512( _mm512_xor_si512( \
mm512_ror_32(x, 7), mm512_ror_32(x, 18) ), _mm512_srli_epi32(x, 3) )
#define SSG2_1x16(x) \
_mm512_xor_si512( _mm512_xor_si512( \
mm512_ror_32(x, 17), mm512_ror_32(x, 19) ), _mm512_srli_epi32(x, 10) )
#define SHA2x16_MEXP( a, b, c, d ) \
mm512_add4_32( SSG2_1x16( W[a] ), W[b], SSG2_0x16( W[c] ), W[d] );
#define SHA2s_16WAY_STEP(A, B, C, D, E, F, G, H, i, j) \
do { \
__m512i T1, T2; \
__m512i K = _mm512_set1_epi32( K256[( (j)+(i) )] ); \
T1 = _mm512_add_epi32( H, mm512_add4_32( BSG2_1x16(E), CHx16(E, F, G), \
K, W[i] ) ); \
T2 = _mm512_add_epi32( BSG2_0x16(A), MAJx16(A, B, C) ); \
D = _mm512_add_epi32( D, T1 ); \
H = _mm512_add_epi32( T1, T2 ); \
} while (0)
static void
sha256_16way_round( sha256_16way_context *ctx, __m512i *in, __m512i r[8] )
{
register __m512i A, B, C, D, E, F, G, H;
__m512i W[16];
mm512_block_bswap_32( W , in );
mm512_block_bswap_32( W+8, in+8 );
if ( ctx->initialized )
{
A = r[0];
B = r[1];
C = r[2];
D = r[3];
E = r[4];
F = r[5];
G = r[6];
H = r[7];
}
else
{
A = m512_const1_64( 0x6A09E6676A09E667 );
B = m512_const1_64( 0xBB67AE85BB67AE85 );
C = m512_const1_64( 0x3C6EF3723C6EF372 );
D = m512_const1_64( 0xA54FF53AA54FF53A );
E = m512_const1_64( 0x510E527F510E527F );
F = m512_const1_64( 0x9B05688C9B05688C );
G = m512_const1_64( 0x1F83D9AB1F83D9AB );
H = m512_const1_64( 0x5BE0CD195BE0CD19 );
}
SHA2s_16WAY_STEP( A, B, C, D, E, F, G, H, 0, 0 );
SHA2s_16WAY_STEP( H, A, B, C, D, E, F, G, 1, 0 );
SHA2s_16WAY_STEP( G, H, A, B, C, D, E, F, 2, 0 );
SHA2s_16WAY_STEP( F, G, H, A, B, C, D, E, 3, 0 );
SHA2s_16WAY_STEP( E, F, G, H, A, B, C, D, 4, 0 );
SHA2s_16WAY_STEP( D, E, F, G, H, A, B, C, 5, 0 );
SHA2s_16WAY_STEP( C, D, E, F, G, H, A, B, 6, 0 );
SHA2s_16WAY_STEP( B, C, D, E, F, G, H, A, 7, 0 );
SHA2s_16WAY_STEP( A, B, C, D, E, F, G, H, 8, 0 );
SHA2s_16WAY_STEP( H, A, B, C, D, E, F, G, 9, 0 );
SHA2s_16WAY_STEP( G, H, A, B, C, D, E, F, 10, 0 );
SHA2s_16WAY_STEP( F, G, H, A, B, C, D, E, 11, 0 );
SHA2s_16WAY_STEP( E, F, G, H, A, B, C, D, 12, 0 );
SHA2s_16WAY_STEP( D, E, F, G, H, A, B, C, 13, 0 );
SHA2s_16WAY_STEP( C, D, E, F, G, H, A, B, 14, 0 );
SHA2s_16WAY_STEP( B, C, D, E, F, G, H, A, 15, 0 );
for ( int j = 16; j < 64; j += 16 )
{
W[ 0] = SHA2x16_MEXP( 14, 9, 1, 0 );
W[ 1] = SHA2x16_MEXP( 15, 10, 2, 1 );
W[ 2] = SHA2x16_MEXP( 0, 11, 3, 2 );
W[ 3] = SHA2x16_MEXP( 1, 12, 4, 3 );
W[ 4] = SHA2x16_MEXP( 2, 13, 5, 4 );
W[ 5] = SHA2x16_MEXP( 3, 14, 6, 5 );
W[ 6] = SHA2x16_MEXP( 4, 15, 7, 6 );
W[ 7] = SHA2x16_MEXP( 5, 0, 8, 7 );
W[ 8] = SHA2x16_MEXP( 6, 1, 9, 8 );
W[ 9] = SHA2x16_MEXP( 7, 2, 10, 9 );
W[10] = SHA2x16_MEXP( 8, 3, 11, 10 );
W[11] = SHA2x16_MEXP( 9, 4, 12, 11 );
W[12] = SHA2x16_MEXP( 10, 5, 13, 12 );
W[13] = SHA2x16_MEXP( 11, 6, 14, 13 );
W[14] = SHA2x16_MEXP( 12, 7, 15, 14 );
W[15] = SHA2x16_MEXP( 13, 8, 0, 15 );
SHA2s_16WAY_STEP( A, B, C, D, E, F, G, H, 0, j );
SHA2s_16WAY_STEP( H, A, B, C, D, E, F, G, 1, j );
SHA2s_16WAY_STEP( G, H, A, B, C, D, E, F, 2, j );
SHA2s_16WAY_STEP( F, G, H, A, B, C, D, E, 3, j );
SHA2s_16WAY_STEP( E, F, G, H, A, B, C, D, 4, j );
SHA2s_16WAY_STEP( D, E, F, G, H, A, B, C, 5, j );
SHA2s_16WAY_STEP( C, D, E, F, G, H, A, B, 6, j );
SHA2s_16WAY_STEP( B, C, D, E, F, G, H, A, 7, j );
SHA2s_16WAY_STEP( A, B, C, D, E, F, G, H, 8, j );
SHA2s_16WAY_STEP( H, A, B, C, D, E, F, G, 9, j );
SHA2s_16WAY_STEP( G, H, A, B, C, D, E, F, 10, j );
SHA2s_16WAY_STEP( F, G, H, A, B, C, D, E, 11, j );
SHA2s_16WAY_STEP( E, F, G, H, A, B, C, D, 12, j );
SHA2s_16WAY_STEP( D, E, F, G, H, A, B, C, 13, j );
SHA2s_16WAY_STEP( C, D, E, F, G, H, A, B, 14, j );
SHA2s_16WAY_STEP( B, C, D, E, F, G, H, A, 15, j );
}
if ( ctx->initialized )
{
r[0] = _mm512_add_epi32( r[0], A );
r[1] = _mm512_add_epi32( r[1], B );
r[2] = _mm512_add_epi32( r[2], C );
r[3] = _mm512_add_epi32( r[3], D );
r[4] = _mm512_add_epi32( r[4], E );
r[5] = _mm512_add_epi32( r[5], F );
r[6] = _mm512_add_epi32( r[6], G );
r[7] = _mm512_add_epi32( r[7], H );
}
else
{
ctx->initialized = true;
r[0] = _mm512_add_epi32( A, m512_const1_64( 0x6A09E6676A09E667 ) );
r[1] = _mm512_add_epi32( B, m512_const1_64( 0xBB67AE85BB67AE85 ) );
r[2] = _mm512_add_epi32( C, m512_const1_64( 0x3C6EF3723C6EF372 ) );
r[3] = _mm512_add_epi32( D, m512_const1_64( 0xA54FF53AA54FF53A ) );
r[4] = _mm512_add_epi32( E, m512_const1_64( 0x510E527F510E527F ) );
r[5] = _mm512_add_epi32( F, m512_const1_64( 0x9B05688C9B05688C ) );
r[6] = _mm512_add_epi32( G, m512_const1_64( 0x1F83D9AB1F83D9AB ) );
r[7] = _mm512_add_epi32( H, m512_const1_64( 0x5BE0CD195BE0CD19 ) );
}
}
void sha256_16way_init( sha256_16way_context *sc )
{
sc->initialized = false;
sc->count_high = sc->count_low = 0;
}
void sha256_16way_update( sha256_16way_context *sc, const void *data,
size_t len )
{
__m512i *vdata = (__m512i*)data;
size_t ptr;
const int buf_size = 64;
ptr = (unsigned)sc->count_low & (buf_size - 1U);
while ( len > 0 )
{
size_t clen;
uint32_t clow, clow2;
clen = buf_size - ptr;
if ( clen > len )
clen = len;
memcpy_512( sc->buf + (ptr>>2), vdata, clen>>2 );
vdata = vdata + (clen>>2);
ptr += clen;
len -= clen;
if ( ptr == buf_size )
{
sha256_16way_round( sc, sc->buf, sc->val );
ptr = 0;
}
clow = sc->count_low;
clow2 = clow + clen;
sc->count_low = clow2;
if ( clow2 < clow )
sc->count_high++;
}
}
void sha256_16way_close( sha256_16way_context *sc, void *dst )
{
unsigned ptr;
uint32_t low, high;
const int buf_size = 64;
const int pad = buf_size - 8;
ptr = (unsigned)sc->count_low & (buf_size - 1U);
sc->buf[ ptr>>2 ] = m512_const1_64( 0x0000008000000080 );
ptr += 4;
if ( ptr > pad )
{
memset_zero_512( sc->buf + (ptr>>2), (buf_size - ptr) >> 2 );
sha256_16way_round( sc, sc->buf, sc->val );
memset_zero_512( sc->buf, pad >> 2 );
}
else
memset_zero_512( sc->buf + (ptr>>2), (pad - ptr) >> 2 );
low = sc->count_low;
high = (sc->count_high << 3) | (low >> 29);
low = low << 3;
sc->buf[ pad >> 2 ] =
mm512_bswap_32( m512_const1_32( high ) );
sc->buf[ ( pad+4 ) >> 2 ] =
mm512_bswap_32( m512_const1_32( low ) );
sha256_16way_round( sc, sc->buf, sc->val );
mm512_block_bswap_32( dst, sc->val );
}
#endif // AVX512
#endif // __AVX2__
#endif // __SSE2__

20
algo/sha/sha256q-4way.c
View File

@@ -15,19 +15,19 @@ void sha256q_8way_hash( void* output, const void* input )
sha256_8way_context ctx;
memcpy( &ctx, &sha256_ctx8, sizeof ctx );
sha256_8way( &ctx, input + (64<<3), 16 );
sha256_8way_update( &ctx, input + (64<<3), 16 );
sha256_8way_close( &ctx, vhash );
sha256_8way_init( &ctx );
sha256_8way( &ctx, vhash, 32 );
sha256_8way_update( &ctx, vhash, 32 );
sha256_8way_close( &ctx, vhash );
sha256_8way_init( &ctx );
sha256_8way( &ctx, vhash, 32 );
sha256_8way_update( &ctx, vhash, 32 );
sha256_8way_close( &ctx, vhash );
sha256_8way_init( &ctx );
sha256_8way( &ctx, vhash, 32 );
sha256_8way_update( &ctx, vhash, 32 );
sha256_8way_close( &ctx, output );
}
@@ -61,7 +61,7 @@ int scanhash_sha256q_8way( struct work *work, uint32_t max_nonce,
// Need big endian data
mm256_bswap32_intrlv80_8x32( vdata, pdata );
sha256_8way_init( &sha256_ctx8 );
sha256_8way( &sha256_ctx8, vdata, 64 );
sha256_8way_update( &sha256_ctx8, vdata, 64 );
for ( int m = 0; m < 6; m++ ) if ( Htarg <= htmax[m] )
{
@@ -108,19 +108,19 @@ void sha256q_4way_hash( void* output, const void* input )
sha256_4way_context ctx;
memcpy( &ctx, &sha256_ctx4, sizeof ctx );
sha256_4way( &ctx, input + (64<<2), 16 );
sha256_4way_update( &ctx, input + (64<<2), 16 );
sha256_4way_close( &ctx, vhash );
sha256_4way_init( &ctx );
sha256_4way( &ctx, vhash, 32 );
sha256_4way_update( &ctx, vhash, 32 );
sha256_4way_close( &ctx, vhash );
sha256_4way_init( &ctx );
sha256_4way( &ctx, vhash, 32 );
sha256_4way_update( &ctx, vhash, 32 );
sha256_4way_close( &ctx, vhash );
sha256_4way_init( &ctx );
sha256_4way( &ctx, vhash, 32 );
sha256_4way_update( &ctx, vhash, 32 );
sha256_4way_close( &ctx, output );
}
@@ -154,7 +154,7 @@ int scanhash_sha256q_4way( struct work *work, uint32_t max_nonce,
mm128_bswap32_intrlv80_4x32( vdata, pdata );
sha256_4way_init( &sha256_ctx4 );
sha256_4way( &sha256_ctx4, vdata, 64 );
sha256_4way_update( &sha256_ctx4, vdata, 64 );
for ( int m = 0; m < 6; m++ ) if ( Htarg <= htmax[m] )
{

16
algo/sha/sha256t-4way.c
View File

@@ -15,15 +15,15 @@ void sha256t_8way_hash( void* output, const void* input )
sha256_8way_context ctx;
memcpy( &ctx, &sha256_ctx8, sizeof ctx );
sha256_8way( &ctx, input + (64<<3), 16 );
sha256_8way_update( &ctx, input + (64<<3), 16 );
sha256_8way_close( &ctx, vhash );
sha256_8way_init( &ctx );
sha256_8way( &ctx, vhash, 32 );
sha256_8way_update( &ctx, vhash, 32 );
sha256_8way_close( &ctx, vhash );
sha256_8way_init( &ctx );
sha256_8way( &ctx, vhash, 32 );
sha256_8way_update( &ctx, vhash, 32 );
sha256_8way_close( &ctx, output );
}
@@ -59,7 +59,7 @@ int scanhash_sha256t_8way( struct work *work, const uint32_t max_nonce,
// Need big endian data
mm256_bswap32_intrlv80_8x32( vdata, pdata );
sha256_8way_init( &sha256_ctx8 );
sha256_8way( &sha256_ctx8, vdata, 64 );
sha256_8way_update( &sha256_ctx8, vdata, 64 );
for ( int m = 0; m < 6; m++ ) if ( Htarg <= htmax[m] )
{
@@ -101,15 +101,15 @@ void sha256t_4way_hash( void* output, const void* input )
sha256_4way_context ctx;
memcpy( &ctx, &sha256_ctx4, sizeof ctx );
sha256_4way( &ctx, input + (64<<2), 16 );
sha256_4way_update( &ctx, input + (64<<2), 16 );
sha256_4way_close( &ctx, vhash );
sha256_4way_init( &ctx );
sha256_4way( &ctx, vhash, 32 );
sha256_4way_update( &ctx, vhash, 32 );
sha256_4way_close( &ctx, vhash );
sha256_4way_init( &ctx );
sha256_4way( &ctx, vhash, 32 );
sha256_4way_update( &ctx, vhash, 32 );
sha256_4way_close( &ctx, output );
}
@@ -143,7 +143,7 @@ int scanhash_sha256t_4way( struct work *work, const uint32_t max_nonce,
mm128_bswap32_intrlv80_4x32( vdata, pdata );
sha256_4way_init( &sha256_ctx4 );
sha256_4way( &sha256_ctx4, vdata, 64 );
sha256_4way_update( &sha256_ctx4, vdata, 64 );
for ( int m = 0; m < 6; m++ ) if ( Htarg <= htmax[m] )
{

94
algo/sha/sha512-hash-4way.c
View File

@@ -37,55 +37,57 @@
#include "sha-hash-4way.h"
/*
static const sph_u64 H512[8] = {
SPH_C64(0x6A09E667F3BCC908), SPH_C64(0xBB67AE8584CAA73B),
SPH_C64(0x3C6EF372FE94F82B), SPH_C64(0xA54FF53A5F1D36F1),
SPH_C64(0x510E527FADE682D1), SPH_C64(0x9B05688C2B3E6C1F),
SPH_C64(0x1F83D9ABFB41BD6B), SPH_C64(0x5BE0CD19137E2179)
static const uit64_t H512[8] =
{
0x6A09E667F3BCC908, 0xBB67AE8584CAA73B,
0x3C6EF372FE94F82B, 0xA54FF53A5F1D36F1,
0x510E527FADE682D1, 0x9B05688C2B3E6C1F,
0x1F83D9ABFB41BD6B, 0x5BE0CD19137E2179
};
*/
static const sph_u64 K512[80] = {
SPH_C64(0x428A2F98D728AE22), SPH_C64(0x7137449123EF65CD),
SPH_C64(0xB5C0FBCFEC4D3B2F), SPH_C64(0xE9B5DBA58189DBBC),
SPH_C64(0x3956C25BF348B538), SPH_C64(0x59F111F1B605D019),
SPH_C64(0x923F82A4AF194F9B), SPH_C64(0xAB1C5ED5DA6D8118),
SPH_C64(0xD807AA98A3030242), SPH_C64(0x12835B0145706FBE),
SPH_C64(0x243185BE4EE4B28C), SPH_C64(0x550C7DC3D5FFB4E2),
SPH_C64(0x72BE5D74F27B896F), SPH_C64(0x80DEB1FE3B1696B1),
SPH_C64(0x9BDC06A725C71235), SPH_C64(0xC19BF174CF692694),
SPH_C64(0xE49B69C19EF14AD2), SPH_C64(0xEFBE4786384F25E3),
SPH_C64(0x0FC19DC68B8CD5B5), SPH_C64(0x240CA1CC77AC9C65),
SPH_C64(0x2DE92C6F592B0275), SPH_C64(0x4A7484AA6EA6E483),
SPH_C64(0x5CB0A9DCBD41FBD4), SPH_C64(0x76F988DA831153B5),
SPH_C64(0x983E5152EE66DFAB), SPH_C64(0xA831C66D2DB43210),
SPH_C64(0xB00327C898FB213F), SPH_C64(0xBF597FC7BEEF0EE4),
SPH_C64(0xC6E00BF33DA88FC2), SPH_C64(0xD5A79147930AA725),
SPH_C64(0x06CA6351E003826F), SPH_C64(0x142929670A0E6E70),
SPH_C64(0x27B70A8546D22FFC), SPH_C64(0x2E1B21385C26C926),
SPH_C64(0x4D2C6DFC5AC42AED), SPH_C64(0x53380D139D95B3DF),
SPH_C64(0x650A73548BAF63DE), SPH_C64(0x766A0ABB3C77B2A8),
SPH_C64(0x81C2C92E47EDAEE6), SPH_C64(0x92722C851482353B),
SPH_C64(0xA2BFE8A14CF10364), SPH_C64(0xA81A664BBC423001),
SPH_C64(0xC24B8B70D0F89791), SPH_C64(0xC76C51A30654BE30),
SPH_C64(0xD192E819D6EF5218), SPH_C64(0xD69906245565A910),
SPH_C64(0xF40E35855771202A), SPH_C64(0x106AA07032BBD1B8),
SPH_C64(0x19A4C116B8D2D0C8), SPH_C64(0x1E376C085141AB53),
SPH_C64(0x2748774CDF8EEB99), SPH_C64(0x34B0BCB5E19B48A8),
SPH_C64(0x391C0CB3C5C95A63), SPH_C64(0x4ED8AA4AE3418ACB),
SPH_C64(0x5B9CCA4F7763E373), SPH_C64(0x682E6FF3D6B2B8A3),
SPH_C64(0x748F82EE5DEFB2FC), SPH_C64(0x78A5636F43172F60),
SPH_C64(0x84C87814A1F0AB72), SPH_C64(0x8CC702081A6439EC),
SPH_C64(0x90BEFFFA23631E28), SPH_C64(0xA4506CEBDE82BDE9),
SPH_C64(0xBEF9A3F7B2C67915), SPH_C64(0xC67178F2E372532B),
SPH_C64(0xCA273ECEEA26619C), SPH_C64(0xD186B8C721C0C207),
SPH_C64(0xEADA7DD6CDE0EB1E), SPH_C64(0xF57D4F7FEE6ED178),
SPH_C64(0x06F067AA72176FBA), SPH_C64(0x0A637DC5A2C898A6),
SPH_C64(0x113F9804BEF90DAE), SPH_C64(0x1B710B35131C471B),
SPH_C64(0x28DB77F523047D84), SPH_C64(0x32CAAB7B40C72493),
SPH_C64(0x3C9EBE0A15C9BEBC), SPH_C64(0x431D67C49C100D4C),
SPH_C64(0x4CC5D4BECB3E42B6), SPH_C64(0x597F299CFC657E2A),
SPH_C64(0x5FCB6FAB3AD6FAEC), SPH_C64(0x6C44198C4A475817)
static const uint64_t K512[80] =
{
0x428A2F98D728AE22, 0x7137449123EF65CD,
0xB5C0FBCFEC4D3B2F, 0xE9B5DBA58189DBBC,
0x3956C25BF348B538, 0x59F111F1B605D019,
0x923F82A4AF194F9B, 0xAB1C5ED5DA6D8118,
0xD807AA98A3030242, 0x12835B0145706FBE,
0x243185BE4EE4B28C, 0x550C7DC3D5FFB4E2,
0x72BE5D74F27B896F, 0x80DEB1FE3B1696B1,
0x9BDC06A725C71235, 0xC19BF174CF692694,
0xE49B69C19EF14AD2, 0xEFBE4786384F25E3,
0x0FC19DC68B8CD5B5, 0x240CA1CC77AC9C65,
0x2DE92C6F592B0275, 0x4A7484AA6EA6E483,
0x5CB0A9DCBD41FBD4, 0x76F988DA831153B5,
0x983E5152EE66DFAB, 0xA831C66D2DB43210,
0xB00327C898FB213F, 0xBF597FC7BEEF0EE4,
0xC6E00BF33DA88FC2, 0xD5A79147930AA725,
0x06CA6351E003826F, 0x142929670A0E6E70,
0x27B70A8546D22FFC, 0x2E1B21385C26C926,
0x4D2C6DFC5AC42AED, 0x53380D139D95B3DF,
0x650A73548BAF63DE, 0x766A0ABB3C77B2A8,
0x81C2C92E47EDAEE6, 0x92722C851482353B,
0xA2BFE8A14CF10364, 0xA81A664BBC423001,
0xC24B8B70D0F89791, 0xC76C51A30654BE30,
0xD192E819D6EF5218, 0xD69906245565A910,
0xF40E35855771202A, 0x106AA07032BBD1B8,
0x19A4C116B8D2D0C8, 0x1E376C085141AB53,
0x2748774CDF8EEB99, 0x34B0BCB5E19B48A8,
0x391C0CB3C5C95A63, 0x4ED8AA4AE3418ACB,
0x5B9CCA4F7763E373, 0x682E6FF3D6B2B8A3,
0x748F82EE5DEFB2FC, 0x78A5636F43172F60,
0x84C87814A1F0AB72, 0x8CC702081A6439EC,
0x90BEFFFA23631E28, 0xA4506CEBDE82BDE9,
0xBEF9A3F7B2C67915, 0xC67178F2E372532B,
0xCA273ECEEA26619C, 0xD186B8C721C0C207,
0xEADA7DD6CDE0EB1E, 0xF57D4F7FEE6ED178,
0x06F067AA72176FBA, 0x0A637DC5A2C898A6,
0x113F9804BEF90DAE, 0x1B710B35131C471B,
0x28DB77F523047D84, 0x32CAAB7B40C72493,
0x3C9EBE0A15C9BEBC, 0x431D67C49C100D4C,
0x4CC5D4BECB3E42B6, 0x597F299CFC657E2A,
0x5FCB6FAB3AD6FAEC, 0x6C44198C4A475817
};