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

v3.8.1

Browse Source
This commit is contained in:
Jay D Dee
2018-02-07 16:38:45 -05:00
parent 54b8fd7362
commit a28daca3ce
83 changed files with 5153 additions and 3924 deletions
Download Patch File Download Diff File Expand all files Collapse all files

98
algo/x17/xevan-4way.c
View File

@@ -13,9 +13,9 @@
#include "algo/keccak/keccak-hash-4way.h"
#include "algo/skein/skein-hash-4way.h"
#include "algo/shavite/sph_shavite.h"
#include "algo/luffa/sse2/luffa_for_sse2.h"
#include "algo/luffa/luffa-hash-2way.h"
#include "algo/cubehash/sse2/cubehash_sse2.h"
#include "algo/simd/sse2/nist.h"
#include "algo/simd/simd-hash-2way.h"
#include "algo/echo/aes_ni/hash_api.h"
#include "algo/hamsi/hamsi-hash-4way.h"
#include "algo/fugue/sph_fugue.h"
@@ -31,10 +31,10 @@ typedef struct {
skein512_4way_context skein;
jh512_4way_context jh;
keccak512_4way_context keccak;
hashState_luffa luffa;
luffa_2way_context luffa;
cubehashParam cube;
sph_shavite512_context shavite;
hashState_sd simd;
simd_2way_context simd;
hashState_echo echo;
hamsi512_4way_context hamsi;
sph_fugue512_context fugue;
@@ -56,10 +56,10 @@ void init_xevan_4way_ctx()
skein512_4way_init(&xevan_4way_ctx.skein);
jh512_4way_init(&xevan_4way_ctx.jh);
keccak512_4way_init(&xevan_4way_ctx.keccak);
init_luffa( &xevan_4way_ctx.luffa, 512 );
luffa_2way_init( &xevan_4way_ctx.luffa, 512 );
cubehashInit( &xevan_4way_ctx.cube, 512, 16, 32 );
sph_shavite512_init( &xevan_4way_ctx.shavite );
init_sd( &xevan_4way_ctx.simd, 512 );
simd_2way_init( &xevan_4way_ctx.simd, 512 );
init_echo( &xevan_4way_ctx.echo, 512 );
hamsi512_4way_init( &xevan_4way_ctx.hamsi );
sph_fugue512_init( &xevan_4way_ctx.fugue );
@@ -127,20 +127,14 @@ void xevan_4way_hash( void *output, const void *input )
keccak512_4way( &ctx.keccak, vhash, dataLen );
keccak512_4way_close( &ctx.keccak, vhash );
// Serial
mm256_deinterleave_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
update_and_final_luffa( &ctx.luffa, (BitSequence*)hash0,
(const BitSequence*)hash0, dataLen );
memcpy( &ctx.luffa, &xevan_4way_ctx.luffa, sizeof(hashState_luffa) );
update_and_final_luffa( &ctx.luffa, (BitSequence*)hash1,
(const BitSequence*)hash1, dataLen );
memcpy( &ctx.luffa, &xevan_4way_ctx.luffa, sizeof(hashState_luffa) );
update_and_final_luffa( &ctx.luffa, (BitSequence*)hash2,
(const BitSequence*)hash2, dataLen );
memcpy( &ctx.luffa, &xevan_4way_ctx.luffa, sizeof(hashState_luffa) );
update_and_final_luffa( &ctx.luffa, (BitSequence*)hash3,
(const BitSequence*)hash3, dataLen );
mm256_interleave_2x128( vhash, hash0, hash1, dataLen<<3 );
luffa_2way_update_close( &ctx.luffa, vhash, vhash, dataLen );
mm256_deinterleave_2x128( hash0, hash1, vhash, dataLen<<3 );
mm256_interleave_2x128( vhash, hash2, hash3, dataLen<<3 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhash, vhash, dataLen );
mm256_deinterleave_2x128( hash2, hash3, vhash, dataLen<<3 );
cubehashUpdateDigest( &ctx.cube, (byte*)hash0, (const byte*) hash0,
dataLen );
@@ -169,17 +163,13 @@ void xevan_4way_hash( void *output, const void *input )
sph_shavite512( &ctx.shavite, hash3, dataLen );
sph_shavite512_close( &ctx.shavite, hash3 );
update_final_sd( &ctx.simd, (BitSequence *)hash0,
(const BitSequence *)hash0, dataLen<<3 );
memcpy( &ctx.simd, &xevan_4way_ctx.simd, sizeof(hashState_sd) );
update_final_sd( &ctx.simd, (BitSequence *)hash1,
(const BitSequence *)hash1, dataLen<<3 );
memcpy( &ctx.simd, &xevan_4way_ctx.simd, sizeof(hashState_sd) );
update_final_sd( &ctx.simd, (BitSequence *)hash2,
(const BitSequence *)hash2, dataLen<<3 );
memcpy( &ctx.simd, &xevan_4way_ctx.simd, sizeof(hashState_sd) );
update_final_sd( &ctx.simd, (BitSequence *)hash3,
(const BitSequence *)hash3, dataLen<<3 );
mm256_interleave_2x128( vhash, hash0, hash1, dataLen<<3 );
simd_2way_update_close( &ctx.simd, vhash, vhash, dataLen<<3 );
mm256_deinterleave_2x128( hash0, hash1, vhash, dataLen<<3 );
mm256_interleave_2x128( vhash, hash2, hash3, dataLen<<3 );
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhash, vhash, dataLen<<3 );
mm256_deinterleave_2x128( hash2, hash3, vhash, dataLen<<3 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
(const BitSequence *) hash0, dataLen<<3 );
@@ -192,12 +182,11 @@ void xevan_4way_hash( void *output, const void *input )
memcpy( &ctx.echo, &xevan_4way_ctx.echo, sizeof(hashState_echo) );
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, dataLen<<3 );
// Parallel 32 bit
mm_interleave_4x32( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
// Parallel
mm256_interleave_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
hamsi512_4way( &ctx.hamsi, vhash, dataLen );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm_deinterleave_4x32( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
mm256_deinterleave_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
sph_fugue512( &ctx.fugue, hash0, dataLen );
sph_fugue512_close( &ctx.fugue, hash0 );
@@ -278,18 +267,13 @@ void xevan_4way_hash( void *output, const void *input )
keccak512_4way_close( &ctx.keccak, vhash );
mm256_deinterleave_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
update_and_final_luffa( &ctx.luffa, (BitSequence*)hash0,
(const BitSequence*)hash0, dataLen );
memcpy( &ctx.luffa, &xevan_4way_ctx.luffa, sizeof(hashState_luffa) );
update_and_final_luffa( &ctx.luffa, (BitSequence*)hash1,
(const BitSequence*)hash1, dataLen );
memcpy( &ctx.luffa, &xevan_4way_ctx.luffa, sizeof(hashState_luffa) );
update_and_final_luffa( &ctx.luffa, (BitSequence*)hash2,
(const BitSequence*)hash2, dataLen );
memcpy( &ctx.luffa, &xevan_4way_ctx.luffa, sizeof(hashState_luffa) );
update_and_final_luffa( &ctx.luffa, (BitSequence*)hash3,
(const BitSequence*)hash3, dataLen );
mm256_interleave_2x128( vhash, hash0, hash1, dataLen<<3 );
luffa_2way_update_close( &ctx.luffa, vhash, vhash, dataLen );
mm256_deinterleave_2x128( hash0, hash1, vhash, dataLen<<3 );
mm256_interleave_2x128( vhash, hash2, hash3, dataLen<<3 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhash, vhash, dataLen );
mm256_deinterleave_2x128( hash2, hash3, vhash, dataLen<<3 );
cubehashUpdateDigest( &ctx.cube, (byte*)hash0, (const byte*) hash0,
dataLen );
@@ -318,17 +302,13 @@ void xevan_4way_hash( void *output, const void *input )
sph_shavite512( &ctx.shavite, hash3, dataLen );
sph_shavite512_close( &ctx.shavite, hash3 );
update_final_sd( &ctx.simd, (BitSequence *)hash0,
(const BitSequence *)hash0, dataLen<<3 );
memcpy( &ctx.simd, &xevan_4way_ctx.simd, sizeof(hashState_sd) );
update_final_sd( &ctx.simd, (BitSequence *)hash1,
(const BitSequence *)hash1, dataLen<<3 );
memcpy( &ctx.simd, &xevan_4way_ctx.simd, sizeof(hashState_sd) );
update_final_sd( &ctx.simd, (BitSequence *)hash2,
(const BitSequence *)hash2, dataLen<<3 );
memcpy( &ctx.simd, &xevan_4way_ctx.simd, sizeof(hashState_sd) );
update_final_sd( &ctx.simd, (BitSequence *)hash3,
(const BitSequence *)hash3, dataLen<<3 );
mm256_interleave_2x128( vhash, hash0, hash1, dataLen<<3 );
simd_2way_update_close( &ctx.simd, vhash, vhash, dataLen<<3 );
mm256_deinterleave_2x128( hash0, hash1, vhash, dataLen<<3 );
mm256_interleave_2x128( vhash, hash2, hash3, dataLen<<3 );
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhash, vhash, dataLen<<3 );
mm256_deinterleave_2x128( hash2, hash3, vhash, dataLen<<3 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
(const BitSequence *) hash0, dataLen<<3 );
@@ -342,10 +322,10 @@ void xevan_4way_hash( void *output, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, dataLen<<3 );
mm_interleave_4x32( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
mm256_interleave_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
hamsi512_4way( &ctx.hamsi, vhash, dataLen );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm_deinterleave_4x32( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
mm256_deinterleave_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
sph_fugue512( &ctx.fugue, hash0, dataLen );
sph_fugue512_close( &ctx.fugue, hash0 );