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

v3.8.4.1

Browse Source
This commit is contained in:
Jay D Dee
2018-03-22 14:28:03 -04:00
parent 20fe05054c
commit 3363d61524
8 changed files with 305 additions and 666 deletions
Download Patch File Download Diff File Expand all files Collapse all files

32
algo/lyra2/sponge.h
View File

@@ -55,23 +55,23 @@ static inline uint64_t rotr64( const uint64_t w, const unsigned c ){
// returns void, updates all args
#define G_4X64(a,b,c,d) \
a = _mm256_add_epi64( a, b ); \
d = mm256_rotr_64( _mm256_xor_si256( d, a), 32 ); \
d = mm256_ror_64( _mm256_xor_si256( d, a), 32 ); \
c = _mm256_add_epi64( c, d ); \
b = mm256_rotr_64( _mm256_xor_si256( b, c ), 24 ); \
b = mm256_ror_64( _mm256_xor_si256( b, c ), 24 ); \
a = _mm256_add_epi64( a, b ); \
d = mm256_rotr_64( _mm256_xor_si256( d, a ), 16 ); \
d = mm256_ror_64( _mm256_xor_si256( d, a ), 16 ); \
c = _mm256_add_epi64( c, d ); \
b = mm256_rotr_64( _mm256_xor_si256( b, c ), 63 );
b = mm256_ror_64( _mm256_xor_si256( b, c ), 63 );
#define LYRA_ROUND_AVX2( s0, s1, s2, s3 ) \
G_4X64( s0, s1, s2, s3 ); \
s1 = mm256_rotr256_1x64( s1); \
s1 = mm256_ror256_1x64( s1); \
s2 = mm256_swap_128( s2 ); \
s3 = mm256_rotl256_1x64( s3 ); \
s3 = mm256_rol256_1x64( s3 ); \
G_4X64( s0, s1, s2, s3 ); \
s1 = mm256_rotl256_1x64( s1 ); \
s1 = mm256_rol256_1x64( s1 ); \
s2 = mm256_swap_128( s2 ); \
s3 = mm256_rotr256_1x64( s3 );
s3 = mm256_ror256_1x64( s3 );
#define LYRA_12_ROUNDS_AVX2( s0, s1, s2, s3 ) \
LYRA_ROUND_AVX2( s0, s1, s2, s3 ) \
@@ -94,25 +94,25 @@ static inline uint64_t rotr64( const uint64_t w, const unsigned c ){
// returns void, all args updated
#define G_2X64(a,b,c,d) \
a = _mm_add_epi64( a, b ); \
d = mm_rotr_64( _mm_xor_si128( d, a), 32 ); \
d = mm_ror_64( _mm_xor_si128( d, a), 32 ); \
c = _mm_add_epi64( c, d ); \
b = mm_rotr_64( _mm_xor_si128( b, c ), 24 ); \
b = mm_ror_64( _mm_xor_si128( b, c ), 24 ); \
a = _mm_add_epi64( a, b ); \
d = mm_rotr_64( _mm_xor_si128( d, a ), 16 ); \
d = mm_ror_64( _mm_xor_si128( d, a ), 16 ); \
c = _mm_add_epi64( c, d ); \
b = mm_rotr_64( _mm_xor_si128( b, c ), 63 );
b = mm_ror_64( _mm_xor_si128( b, c ), 63 );
#define LYRA_ROUND_AVX(s0,s1,s2,s3,s4,s5,s6,s7) \
G_2X64( s0, s2, s4, s6 ); \
G_2X64( s1, s3, s5, s7 ); \
mm_rotr256_1x64( s2, s3 ); \
mm_ror256_1x64( s2, s3 ); \
mm_swap_128( s4, s5 ); \
mm_rotl256_1x64( s6, s7 ); \
mm_rol256_1x64( s6, s7 ); \
G_2X64( s0, s2, s4, s6 ); \
G_2X64( s1, s3, s5, s7 ); \
mm_rotl256_1x64( s2, s3 ); \
mm_rol256_1x64( s2, s3 ); \
mm_swap_128( s4, s5 ); \
mm_rotr256_1x64( s6, s7 );
mm_ror256_1x64( s6, s7 );
#define LYRA_12_ROUNDS_AVX(s0,s1,s2,s3,s4,s5,s6,s7) \
LYRA_ROUND_AVX(s0,s1,s2,s3,s4,s5,s6,s7) \