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

v3.7.4

Browse Source
This commit is contained in:
Jay D Dee
2021-09-29 17:31:16 -04:00
parent 9b905fccc8
commit 2cd1507c2e
80 changed files with 8145 additions and 2097 deletions
Download Patch File Download Diff File Expand all files Collapse all files

143
algo/x16/x16r-4way.c
View File

@@ -52,10 +52,10 @@ void x16r_8way_prehash( void *vdata, void *pdata )
break;
case CUBEHASH:
mm128_bswap32_80( edata, pdata );
cubehashInit( &x16r_ctx.cube, 512, 16, 32 );
cubehashUpdate( &x16r_ctx.cube, (const byte*)edata, 64 );
intrlv_8x64( vdata, edata, edata, edata, edata,
edata, edata, edata, edata, 640 );
intrlv_4x128( vdata2, edata, edata, edata, edata, 640 );
cube_4way_init( &x16r_ctx.cube, 512, 16, 32 );
cube_4way_update( &x16r_ctx.cube, vdata2, 64 );
rintrlv_4x128_8x64( vdata, vdata2, vdata2, 640 );
break;
case HAMSI:
mm512_bswap32_intrlv80_8x64( vdata, pdata );
@@ -207,15 +207,15 @@ int x16r_8way_hash_generic( void* output, const void* input, int thrid )
case LUFFA:
if ( i == 0 )
{
intrlv_4x128( vhash, in0, in1, in2, in3, size<<3 );
luffa_4way_update_close( &ctx.luffa, vhash,
vhash + (16<<2), 16 );
dintrlv_4x128_512( hash0, hash1, hash2, hash3, vhash );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
intrlv_4x128( vhash, in4, in5, in6, in7, size<<3 );
luffa_4way_update_close( &ctx.luffa, vhash,
vhash + (16<<2), 16 );
dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhash );
intrlv_4x128( vhash, in0, in1, in2, in3, size<<3 );
luffa_4way_update_close( &ctx.luffa, vhash,
vhash + (16<<2), 16 );
dintrlv_4x128_512( hash0, hash1, hash2, hash3, vhash );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
intrlv_4x128( vhash, in4, in5, in6, in7, size<<3 );
luffa_4way_update_close( &ctx.luffa, vhash,
vhash + (16<<2), 16 );
dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhash );
}
else
{
@@ -230,56 +230,24 @@ int x16r_8way_hash_generic( void* output, const void* input, int thrid )
case CUBEHASH:
if ( i == 0 )
{
cubehashUpdateDigest( &ctx.cube, (byte*)hash0,
(const byte*)in0 + 64, 16 );
intrlv_4x128( vhash, in0, in1, in2, in3, size<<3 );
cube_4way_update_close( &ctx.cube, vhash,
vhash + (16<<2), 16 );
dintrlv_4x128_512( hash0, hash1, hash2, hash3, vhash );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*)hash1,
(const byte*)in1 + 64, 16 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*)hash2,
(const byte*)in2 + 64, 16 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*)hash3,
(const byte*)in3 + 64, 16 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*)hash4,
(const byte*)in4 + 64, 16 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*)hash5,
(const byte*)in5 + 64, 16 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*)hash6,
(const byte*)in6 + 64, 16 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*)hash7,
(const byte*)in7 + 64, 16 );
intrlv_4x128( vhash, in4, in5, in6, in7, size<<3 );
cube_4way_update_close( &ctx.cube, vhash,
vhash + (16<<2), 16 );
dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhash );
}
else
{
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash0,
(const byte*)in0, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash1,
(const byte*)in1, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash2,
(const byte*)in2, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash3,
(const byte*)in3, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash4,
(const byte*)in4, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash5,
(const byte*)in5, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash6,
(const byte*)in6, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash7,
(const byte*)in7, size );
intrlv_4x128( vhash, in0, in1, in2, in3, size<<3 );
cube_4way_full( &ctx.cube, vhash, 512, vhash, size );
dintrlv_4x128_512( hash0, hash1, hash2, hash3, vhash );
intrlv_4x128( vhash, in4, in5, in6, in7, size<<3 );
cube_4way_full( &ctx.cube, vhash, 512, vhash, size );
dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhash );
}
break;
case SHAVITE:
@@ -556,9 +524,10 @@ void x16r_4way_prehash( void *vdata, void *pdata )
break;
case CUBEHASH:
mm128_bswap32_80( edata, pdata );
cubehashInit( &x16r_ctx.cube, 512, 16, 32 );
cubehashUpdate( &x16r_ctx.cube, (const byte*)edata, 64 );
intrlv_4x64( vdata, edata, edata, edata, edata, 640 );
intrlv_2x128( vdata2, edata, edata, 640 );
cube_2way_init( &x16r_ctx.cube, 512, 16, 32 );
cube_2way_update( &x16r_ctx.cube, vdata2, 64 );
rintrlv_2x128_4x64( vdata, vdata2, vdata2, 640 );
break;
case HAMSI:
mm256_bswap32_intrlv80_4x64( vdata, pdata );
@@ -680,13 +649,13 @@ int x16r_4way_hash_generic( void* output, const void* input, int thrid )
case LUFFA:
if ( i == 0 )
{
intrlv_2x128( vhash, hash0, hash1, 640 );
luffa_2way_update_close( &ctx.luffa, vhash, vhash + (16<<1), 16 );
dintrlv_2x128_512( hash0, hash1, vhash );
intrlv_2x128( vhash, hash2, hash3, 640 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
luffa_2way_update_close( &ctx.luffa, vhash, vhash + (16<<1), 16 );
dintrlv_2x128_512( hash2, hash3, vhash );
intrlv_2x128( vhash, hash0, hash1, 640 );
luffa_2way_update_close( &ctx.luffa, vhash, vhash + (16<<1), 16 );
dintrlv_2x128_512( hash0, hash1, vhash );
intrlv_2x128( vhash, hash2, hash3, 640 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
luffa_2way_update_close( &ctx.luffa, vhash, vhash + (16<<1), 16 );
dintrlv_2x128_512( hash2, hash3, vhash );
}
else
{
@@ -701,32 +670,24 @@ int x16r_4way_hash_generic( void* output, const void* input, int thrid )
case CUBEHASH:
if ( i == 0 )
{
cubehashUpdateDigest( &ctx.cube, (byte*)hash0,
(const byte*)in0 + 64, 16 );
intrlv_2x128( vhash, in0, in1, size<<3 );
cube_2way_update_close( &ctx.cube, vhash,
vhash + (16<<1), 16 );
dintrlv_2x128_512( hash0, hash1, vhash );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*) hash1,
(const byte*)in1 + 64, 16 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*) hash2,
(const byte*)in2 + 64, 16 );
memcpy( &ctx, &x16r_ctx, sizeof(ctx) );
cubehashUpdateDigest( &ctx.cube, (byte*) hash3,
(const byte*)in3 + 64, 16 );
intrlv_2x128( vhash, in2, in3, size<<3 );
cube_2way_update_close( &ctx.cube, vhash,
vhash + (16<<1), 16 );
dintrlv_2x128_512( hash2, hash3, vhash );
}
else
{
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash0,
(const byte*)in0, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash1,
(const byte*)in1, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash2,
(const byte*)in2, size );
cubehashInit( &ctx.cube, 512, 16, 32 );
cubehashUpdateDigest( &ctx.cube, (byte*) hash3,
(const byte*)in3, size );
intrlv_2x128( vhash, in0, in1, size<<3 );
cube_2way_full( &ctx.cube, vhash, 512, vhash, size );
dintrlv_2x128_512( hash0, hash1, vhash );
intrlv_2x128( vhash, in2, in3, size<<3 );
cube_2way_full( &ctx.cube, vhash, 512, vhash, size );
dintrlv_2x128_512( hash2, hash3, vhash );
}
break;
case SHAVITE:

1
algo/x16/x16r-gate.c
View File

@@ -1,4 +1,5 @@
#include "x16r-gate.h"
#include "algo/sha/sha256d.h"
__thread char x16r_hash_order[ X16R_HASH_FUNC_COUNT + 1 ] = { 0 };

5
algo/x16/x16r-gate.h
View File

@@ -37,6 +37,7 @@
#include "algo/jh/jh-hash-4way.h"
#include "algo/keccak/keccak-hash-4way.h"
#include "algo/luffa/luffa-hash-2way.h"
#include "algo/cubehash/cube-hash-2way.h"
#include "algo/simd/simd-hash-2way.h"
#include "algo/echo/aes_ni/hash_api.h"
#include "algo/hamsi/hamsi-hash-4way.h"
@@ -115,7 +116,7 @@ union _x16r_8way_context_overlay
jh512_8way_context jh;
keccak512_8way_context keccak;
luffa_4way_context luffa;
cubehashParam cube;
cube_4way_context cube;
simd_4way_context simd;
hamsi512_8way_context hamsi;
hashState_fugue fugue;
@@ -164,8 +165,8 @@ union _x16r_4way_context_overlay
jh512_4way_context jh;
keccak512_4way_context keccak;
luffa_2way_context luffa;
cube_2way_context cube;
hashState_luffa luffa1;
cubehashParam cube;
simd_2way_context simd;
hamsi512_4way_context hamsi;
hashState_fugue fugue;

22
algo/x16/x21s-4way.c
View File

@@ -13,7 +13,7 @@
#include "algo/gost/sph_gost.h"
#include "algo/lyra2/lyra2.h"
#if defined(__SHA__)
#include "algo/sha/sph_sha2.h"
#include "algo/sha/sha256-hash.h"
#endif
#if defined (X21S_8WAY)
@@ -208,9 +208,7 @@ union _x21s_4way_context_overlay
haval256_5_4way_context haval;
sph_tiger_context tiger;
sph_gost512_context gost;
#if defined(__SHA__)
sph_sha256_context sha256;
#else
#if !defined(__SHA__)
sha256_4way_context sha256;
#endif
} __attribute__ ((aligned (64)));
@@ -275,18 +273,10 @@ int x21s_4way_hash( void* output, const void* input, int thrid )
#if defined(__SHA__)
sph_sha256_init( &ctx.sha256 );
sph_sha256( &ctx.sha256, hash0, 64 );
sph_sha256_close( &ctx.sha256, output );
sph_sha256_init( &ctx.sha256 );
sph_sha256( &ctx.sha256, hash1, 64 );
sph_sha256_close( &ctx.sha256, output+32 );
sph_sha256_init( &ctx.sha256 );
sph_sha256( &ctx.sha256, hash2, 64 );
sph_sha256_close( &ctx.sha256, output+64 );
sph_sha256_init( &ctx.sha256 );
sph_sha256( &ctx.sha256, hash3, 64 );
sph_sha256_close( &ctx.sha256, output+96 );
sha256_full( output, hash0, 64 );
sha256_full( output+32, hash1, 64 );
sha256_full( output+64, hash2, 64 );
sha256_full( output+96, hash3, 64 );
#else

8
algo/x16/x21s.c
View File

@@ -8,7 +8,7 @@
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "algo/sha/sph_sha2.h"
#include "algo/sha/sha256-hash.h"
#include "algo/haval/sph-haval.h"
#include "algo/tiger/sph_tiger.h"
#include "algo/gost/sph_gost.h"
@@ -23,7 +23,7 @@ union _x21s_context_overlay
sph_haval256_5_context haval;
sph_tiger_context tiger;
sph_gost512_context gost;
sph_sha256_context sha256;
sha256_context sha256;
};
typedef union _x21s_context_overlay x21s_context_overlay;
@@ -50,9 +50,7 @@ int x21s_hash( void* output, const void* input, int thrid )
sph_gost512 ( &ctx.gost, (const void*) hash, 64 );
sph_gost512_close( &ctx.gost, (void*) hash );
sph_sha256_init( &ctx.sha256 );
sph_sha256( &ctx.sha256, hash, 64 );
sph_sha256_close( &ctx.sha256, hash );
sha256_full( hash, hash, 64 );
memcpy( output, hash, 32 );