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

v3.10.5

Browse Source
This commit is contained in:
Jay D Dee
2019-12-21 13:19:29 -05:00
parent a17ff6f189
commit c65b0ff7a6
72 changed files with 9090 additions and 1336 deletions
Download Patch File Download Diff File Expand all files Collapse all files

52
algo/bmw/bmw256-hash-4way.c
View File

@@ -874,6 +874,57 @@ void compress_small_8way( const __m256i *M, const __m256i H[16],
mm256_xor4( qt[24], qt[25], qt[26], qt[27] ),
mm256_xor4( qt[28], qt[29], qt[30], qt[31] ) ) );
#define DH1L( m, sl, sr, a, b, c ) \
_mm256_add_epi32( \
_mm256_xor_si256( M[m], \
_mm256_xor_si256( _mm256_slli_epi32( xh, sl ), \
_mm256_srli_epi32( qt[a], sr ) ) ), \
_mm256_xor_si256( _mm256_xor_si256( xl, qt[b] ), qt[c] ) )
#define DH1R( m, sl, sr, a, b, c ) \
_mm256_add_epi32( \
_mm256_xor_si256( M[m], \
_mm256_xor_si256( _mm256_srli_epi32( xh, sl ), \
_mm256_slli_epi32( qt[a], sr ) ) ), \
_mm256_xor_si256( _mm256_xor_si256( xl, qt[b] ), qt[c] ) )
#define DH2L( m, rl, sl, h, a, b, c ) \
_mm256_add_epi32( _mm256_add_epi32( \
mm256_rol_32( dH[h], rl ), \
_mm256_xor_si256( _mm256_xor_si256( xh, qt[a] ), M[m] )), \
_mm256_xor_si256( _mm256_slli_epi32( xl, sl ), \
_mm256_xor_si256( qt[b], qt[c] ) ) );
#define DH2R( m, rl, sr, h, a, b, c ) \
_mm256_add_epi32( _mm256_add_epi32( \
mm256_rol_32( dH[h], rl ), \
_mm256_xor_si256( _mm256_xor_si256( xh, qt[a] ), M[m] )), \
_mm256_xor_si256( _mm256_srli_epi32( xl, sr ), \
_mm256_xor_si256( qt[b], qt[c] ) ) );
dH[ 0] = DH1L( 0, 5, 5, 16, 24, 0 );
dH[ 1] = DH1R( 1, 7, 8, 17, 25, 1 );
dH[ 2] = DH1R( 2, 5, 5, 18, 26, 2 );
dH[ 3] = DH1R( 3, 1, 5, 19, 27, 3 );
dH[ 4] = DH1R( 4, 3, 0, 20, 28, 4 );
dH[ 5] = DH1L( 5, 6, 6, 21, 29, 5 );
dH[ 6] = DH1R( 6, 4, 6, 22, 30, 6 );
dH[ 7] = DH1R( 7, 11, 2, 23, 31, 7 );
dH[ 8] = DH2L( 8, 9, 8, 4, 24, 23, 8 );
dH[ 9] = DH2R( 9, 10, 6, 5, 25, 16, 9 );
dH[10] = DH2L( 10, 11, 6, 6, 26, 17, 10 );
dH[11] = DH2L( 11, 12, 4, 7, 27, 18, 11 );
dH[12] = DH2R( 12, 13, 3, 0, 28, 19, 12 );
dH[13] = DH2R( 13, 14, 4, 1, 29, 20, 13 );
dH[14] = DH2R( 14, 15, 7, 2, 30, 21, 14 );
dH[15] = DH2R( 15, 16, 2, 3, 31, 22, 15 );
#undef DH1L
#undef DH1R
#undef DH2L
#undef DH2R
/*
dH[ 0] = _mm256_add_epi32(
_mm256_xor_si256( M[0],
_mm256_xor_si256( _mm256_slli_epi32( xh, 5 ),
@@ -954,6 +1005,7 @@ void compress_small_8way( const __m256i *M, const __m256i H[16],
_mm256_xor_si256( _mm256_xor_si256( xh, qt[31] ), M[15] )),
_mm256_xor_si256( _mm256_srli_epi32( xl, 2 ),
_mm256_xor_si256( qt[22], qt[15] ) ) );
*/
}
static const __m256i final_s8[16] =