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

v3.9.0

Browse Source
This commit is contained in:
Jay D Dee
2019-05-19 13:39:45 -04:00
parent bfd1c002f9
commit e1aead3c76
139 changed files with 10907 additions and 4218 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
algo/x14/polytimos-4way.c
View File

@@ -52,7 +52,7 @@ void polytimos_4way_hash( void *output, const void *input )
mm256_reinterleave_4x32( vhash32, vhash, 512 );
shabal512_4way( &ctx.shabal, vhash32, 64 );
shabal512_4way_close( &ctx.shabal, vhash32 );
mm_deinterleave_4x32( hash0, hash1, hash2, hash3, vhash32, 512 );
mm128_deinterleave_4x32( hash0, hash1, hash2, hash3, vhash32, 512 );
update_final_echo ( &ctx.echo, (BitSequence *)hash0,
(const BitSequence *)hash0, 512 );

4
algo/x14/veltor-4way.c
View File

@@ -54,10 +54,10 @@ void veltor_4way_hash( void *output, const void *input )
sph_shavite512( &ctx.shavite, hash3, 64 );
sph_shavite512_close( &ctx.shavite, hash3 );
mm_interleave_4x32( vhash, hash0, hash1, hash2, hash3, 512 );
mm128_interleave_4x32( vhash, hash0, hash1, hash2, hash3, 512 );
shabal512_4way( &ctx.shabal, vhash, 64 );
shabal512_4way_close( &ctx.shabal, vhash );
mm_deinterleave_4x32( hash0, hash1, hash2, hash3, vhash, 512 );
mm128_deinterleave_4x32( hash0, hash1, hash2, hash3, vhash, 512 );
sph_gost512( &ctx.gost, hash0, 64 );
sph_gost512_close( &ctx.gost, hash0 );

35
algo/x14/x14-4way.c
View File

@@ -13,7 +13,7 @@
#include "algo/jh/jh-hash-4way.h"
#include "algo/keccak/keccak-hash-4way.h"
#include "algo/luffa/luffa-hash-2way.h"
#include "algo/cubehash/sse2/cubehash_sse2.h"
#include "algo/cubehash/cubehash_sse2.h"
#include "algo/shavite/sph_shavite.h"
#include "algo/simd/simd-hash-2way.h"
#include "algo/echo/aes_ni/hash_api.h"
@@ -183,21 +183,16 @@ void x14_4way_hash( void *state, const void *input )
sph_fugue512_close( &ctx.fugue, hash3 );
// 14 Shabal, parallel 32 bit
mm_interleave_4x32( vhash, hash0, hash1, hash2, hash3, 512 );
mm128_interleave_4x32( vhash, hash0, hash1, hash2, hash3, 512 );
shabal512_4way( &ctx.shabal, vhash, 64 );
shabal512_4way_close( &ctx.shabal, vhash );
mm_deinterleave_4x32( hash0, hash1, hash2, hash3, vhash, 512 );
memcpy( state, hash0, 32 );
memcpy( state+32, hash1, 32 );
memcpy( state+64, hash2, 32 );
memcpy( state+96, hash3, 32 );
shabal512_4way_close( &ctx.shabal, state );
}
int scanhash_x14_4way( int thr_id, struct work *work, uint32_t max_nonce,
uint64_t *hashes_done )
{
uint32_t hash[4*8] __attribute__ ((aligned (64)));
uint32_t hash[4*16] __attribute__ ((aligned (64)));
uint32_t vdata[24*4] __attribute__ ((aligned (64)));
uint32_t endiandata[20] __attribute__((aligned(64)));
uint32_t *pdata = work->data;
@@ -233,13 +228,21 @@ int scanhash_x14_4way( int thr_id, struct work *work, uint32_t max_nonce,
x14_4way_hash( hash, vdata );
pdata[19] = n;
for ( int i = 0; i < 4; i++ )
if ( ( ( (hash+(i<<3))[7] & mask ) == 0 )
&& fulltest( hash+(i<<3), ptarget ) )
uint32_t *hash7 = &(hash[7<<2]);
for ( int lane = 0; lane < 4; lane++ )
if ( ( hash7[ lane ] & mask ) == 0 )
{
pdata[19] = n+i;
nonces[ num_found++ ] = n+i;
work_set_target_ratio( work, hash+(i<<3) );
// deinterleave hash for lane
uint32_t lane_hash[8];
mm128_extract_lane_4x32( lane_hash, hash, lane, 256 );
if ( fulltest( lane_hash, ptarget ) )
{
pdata[19] = n + lane;
nonces[ num_found++ ] = n + lane;
work_set_target_ratio( work, lane_hash );
}
}
n += 4;
} while ( ( num_found == 0 ) && ( n < max_nonce )

3
algo/x14/x14.c
View File

@@ -21,9 +21,8 @@
#include "algo/shabal/sph_shabal.h"
#include "algo/luffa/luffa_for_sse2.h"
#include "algo/cubehash/sse2/cubehash_sse2.h"
#include "algo/cubehash/cubehash_sse2.h"
#include "algo/simd/nist.h"
#include "algo/echo/sse2/sph_echo.h"
#include "algo/blake/sse2/blake.c"
#include "algo/bmw/sse2/bmw.c"
#include "algo/keccak/sse2/keccak.c"