mirror of
https://github.com/JayDDee/cpuminer-opt.git
synced 2025-09-17 23:44:27 +00:00
v3.9.6
This commit is contained in:
Jay D Dee
283
algo/x13/x13bcd-4way.c
Normal file
283
algo/x13/x13bcd-4way.c
Normal file
@@ -0,0 +1,283 @@
|
||||
#include "x13sm3-gate.h"
|
||||
|
||||
#if defined(X13SM3_4WAY)
|
||||
|
||||
#include <stdlib.h>
|
||||
#include <stdint.h>
|
||||
#include <string.h>
|
||||
#include <stdio.h>
|
||||
#include "algo/blake/blake-hash-4way.h"
|
||||
#include "algo/bmw/bmw-hash-4way.h"
|
||||
#include "algo/groestl/aes_ni/hash-groestl.h"
|
||||
#include "algo/skein/skein-hash-4way.h"
|
||||
#include "algo/jh/jh-hash-4way.h"
|
||||
#include "algo/keccak/keccak-hash-4way.h"
|
||||
//#include "algo/luffa/luffa-hash-2way.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"
|
||||
#include "algo/sm3/sm3-hash-4way.h"
|
||||
#include "algo/hamsi/hamsi-hash-4way.h"
|
||||
#include "algo/fugue/sph_fugue.h"
|
||||
|
||||
typedef struct {
|
||||
blake512_4way_context blake;
|
||||
bmw512_4way_context bmw;
|
||||
hashState_groestl groestl;
|
||||
skein512_4way_context skein;
|
||||
jh512_4way_context jh;
|
||||
keccak512_4way_context keccak;
|
||||
// luffa_2way_context luffa;
|
||||
cubehashParam cube;
|
||||
sph_shavite512_context shavite;
|
||||
simd_2way_context simd;
|
||||
hashState_echo echo;
|
||||
sm3_4way_ctx_t sm3;
|
||||
hamsi512_4way_context hamsi;
|
||||
sph_fugue512_context fugue;
|
||||
} x13bcd_4way_ctx_holder;
|
||||
|
||||
x13bcd_4way_ctx_holder x13bcd_4way_ctx __attribute__ ((aligned (64)));
|
||||
static __thread blake512_4way_context x13bcd_ctx_mid;
|
||||
|
||||
void init_x13bcd_4way_ctx()
|
||||
{
|
||||
blake512_4way_init( &x13bcd_4way_ctx.blake );
|
||||
bmw512_4way_init( &x13bcd_4way_ctx.bmw );
|
||||
init_groestl( &x13bcd_4way_ctx.groestl, 64 );
|
||||
skein512_4way_init( &x13bcd_4way_ctx.skein );
|
||||
jh512_4way_init( &x13bcd_4way_ctx.jh );
|
||||
keccak512_4way_init( &x13bcd_4way_ctx.keccak );
|
||||
// luffa_2way_init( &x13bcd_4way_ctx.luffa, 512 );
|
||||
cubehashInit( &x13bcd_4way_ctx.cube, 512, 16, 32 );
|
||||
sph_shavite512_init( &x13bcd_4way_ctx.shavite );
|
||||
simd_2way_init( &x13bcd_4way_ctx.simd, 512 );
|
||||
init_echo( &x13bcd_4way_ctx.echo, 512 );
|
||||
sm3_4way_init( &x13bcd_4way_ctx.sm3 );
|
||||
hamsi512_4way_init( &x13bcd_4way_ctx.hamsi );
|
||||
sph_fugue512_init( &x13bcd_4way_ctx.fugue );
|
||||
};
|
||||
|
||||
void x13bcd_4way_hash( void *state, const void *input )
|
||||
{
|
||||
uint64_t hash0[8] __attribute__ ((aligned (64)));
|
||||
uint64_t hash1[8] __attribute__ ((aligned (64)));
|
||||
uint64_t hash2[8] __attribute__ ((aligned (64)));
|
||||
uint64_t hash3[8] __attribute__ ((aligned (64)));
|
||||
uint64_t vhash[8*4] __attribute__ ((aligned (64)));
|
||||
x13bcd_4way_ctx_holder ctx;
|
||||
memcpy( &ctx, &x13bcd_4way_ctx, sizeof(x13bcd_4way_ctx) );
|
||||
|
||||
// Blake
|
||||
memcpy( &ctx.blake, &x13bcd_ctx_mid, sizeof(x13bcd_ctx_mid) );
|
||||
blake512_4way( &ctx.blake, input + (64<<2), 16 );
|
||||
|
||||
// blake512_4way( &ctx.blake, input, 80 );
|
||||
blake512_4way_close( &ctx.blake, vhash );
|
||||
|
||||
// Bmw
|
||||
bmw512_4way( &ctx.bmw, vhash, 64 );
|
||||
bmw512_4way_close( &ctx.bmw, vhash );
|
||||
|
||||
// Serial
|
||||
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
|
||||
|
||||
// Groestl
|
||||
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
|
||||
reinit_groestl( &ctx.groestl );
|
||||
update_and_final_groestl( &ctx.groestl, (char*)hash1, (char*)hash1, 512 );
|
||||
reinit_groestl( &ctx.groestl );
|
||||
update_and_final_groestl( &ctx.groestl, (char*)hash2, (char*)hash2, 512 );
|
||||
reinit_groestl( &ctx.groestl );
|
||||
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );
|
||||
|
||||
// Parallel 4way
|
||||
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
|
||||
|
||||
// Skein
|
||||
skein512_4way( &ctx.skein, vhash, 64 );
|
||||
skein512_4way_close( &ctx.skein, vhash );
|
||||
|
||||
// JH
|
||||
jh512_4way( &ctx.jh, vhash, 64 );
|
||||
jh512_4way_close( &ctx.jh, vhash );
|
||||
|
||||
// Keccak
|
||||
keccak512_4way( &ctx.keccak, vhash, 64 );
|
||||
keccak512_4way_close( &ctx.keccak, vhash );
|
||||
|
||||
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
|
||||
|
||||
intrlv_4x32( vhash, hash0, hash1, hash2, hash3, 512 );
|
||||
|
||||
// SM3 parallel 32 bit
|
||||
uint32_t sm3_vhash[32*4] __attribute__ ((aligned (64)));
|
||||
memset( sm3_vhash, 0, sizeof sm3_vhash );
|
||||
uint32_t sm3_hash0[32] __attribute__ ((aligned (32)));
|
||||
memset( sm3_hash0, 0, sizeof sm3_hash0 );
|
||||
uint32_t sm3_hash1[32] __attribute__ ((aligned (32)));
|
||||
memset( sm3_hash1, 0, sizeof sm3_hash1 );
|
||||
uint32_t sm3_hash2[32] __attribute__ ((aligned (32)));
|
||||
memset( sm3_hash2, 0, sizeof sm3_hash2 );
|
||||
uint32_t sm3_hash3[32] __attribute__ ((aligned (32)));
|
||||
memset( sm3_hash3, 0, sizeof sm3_hash3 );
|
||||
|
||||
sm3_4way( &ctx.sm3, vhash, 64 );
|
||||
sm3_4way_close( &ctx.sm3, sm3_vhash );
|
||||
dintrlv_4x32( hash0, hash1, hash2, hash3, sm3_vhash, 512 );
|
||||
|
||||
/*
|
||||
// Luffa
|
||||
intrlv_2x128( vhash, hash0, hash1, 512 );
|
||||
luffa_2way_update_close( &ctx.luffa, vhash, vhash, 64 );
|
||||
dintrlv_2x128( hash0, hash1, vhash, 512 );
|
||||
intrlv_2x128( vhash, hash2, hash3, 512 );
|
||||
luffa_2way_init( &ctx.luffa, 512 );
|
||||
luffa_2way_update_close( &ctx.luffa, vhash, vhash, 64 );
|
||||
dintrlv_2x128( hash2, hash3, vhash, 512 );
|
||||
*/
|
||||
|
||||
// Cubehash
|
||||
cubehashUpdateDigest( &ctx.cube, (byte*)hash0, (const byte*) hash0, 64 );
|
||||
memcpy( &ctx.cube, &x13bcd_4way_ctx.cube, sizeof(cubehashParam) );
|
||||
cubehashUpdateDigest( &ctx.cube, (byte*)hash1, (const byte*) hash1, 64 );
|
||||
memcpy( &ctx.cube, &x13bcd_4way_ctx.cube, sizeof(cubehashParam) );
|
||||
cubehashUpdateDigest( &ctx.cube, (byte*)hash2, (const byte*) hash2, 64 );
|
||||
memcpy( &ctx.cube, &x13bcd_4way_ctx.cube, sizeof(cubehashParam) );
|
||||
cubehashUpdateDigest( &ctx.cube, (byte*)hash3, (const byte*) hash3, 64 );
|
||||
|
||||
// Shavite
|
||||
sph_shavite512( &ctx.shavite, hash0, 64 );
|
||||
sph_shavite512_close( &ctx.shavite, hash0 );
|
||||
memcpy( &ctx.shavite, &x13bcd_4way_ctx.shavite,
|
||||
sizeof(sph_shavite512_context) );
|
||||
sph_shavite512( &ctx.shavite, hash1, 64 );
|
||||
sph_shavite512_close( &ctx.shavite, hash1 );
|
||||
memcpy( &ctx.shavite, &x13bcd_4way_ctx.shavite,
|
||||
sizeof(sph_shavite512_context) );
|
||||
sph_shavite512( &ctx.shavite, hash2, 64 );
|
||||
sph_shavite512_close( &ctx.shavite, hash2 );
|
||||
memcpy( &ctx.shavite, &x13bcd_4way_ctx.shavite,
|
||||
sizeof(sph_shavite512_context) );
|
||||
sph_shavite512( &ctx.shavite, hash3, 64 );
|
||||
sph_shavite512_close( &ctx.shavite, hash3 );
|
||||
|
||||
// Simd
|
||||
intrlv_2x128( vhash, hash0, hash1, 512 );
|
||||
simd_2way_update_close( &ctx.simd, vhash, vhash, 512 );
|
||||
dintrlv_2x128( hash0, hash1, vhash, 512 );
|
||||
intrlv_2x128( vhash, hash2, hash3, 512 );
|
||||
simd_2way_init( &ctx.simd, 512 );
|
||||
simd_2way_update_close( &ctx.simd, vhash, vhash, 512 );
|
||||
dintrlv_2x128( hash2, hash3, vhash, 512 );
|
||||
|
||||
// Echo
|
||||
update_final_echo( &ctx.echo, (BitSequence *)hash0,
|
||||
(const BitSequence *) hash0, 512 );
|
||||
memcpy( &ctx.echo, &x13bcd_4way_ctx.echo, sizeof(hashState_echo) );
|
||||
update_final_echo( &ctx.echo, (BitSequence *)hash1,
|
||||
(const BitSequence *) hash1, 512 );
|
||||
memcpy( &ctx.echo, &x13bcd_4way_ctx.echo, sizeof(hashState_echo) );
|
||||
update_final_echo( &ctx.echo, (BitSequence *)hash2,
|
||||
(const BitSequence *) hash2, 512 );
|
||||
memcpy( &ctx.echo, &x13bcd_4way_ctx.echo, sizeof(hashState_echo) );
|
||||
update_final_echo( &ctx.echo, (BitSequence *)hash3,
|
||||
(const BitSequence *) hash3, 512 );
|
||||
|
||||
/*
|
||||
intrlv_4x32( vhash, hash0, hash1, hash2, hash3, 512 );
|
||||
|
||||
// SM3 parallel 32 bit
|
||||
uint32_t sm3_vhash[32*4] __attribute__ ((aligned (64)));
|
||||
memset( sm3_vhash, 0, sizeof sm3_vhash );
|
||||
uint32_t sm3_hash0[32] __attribute__ ((aligned (32)));
|
||||
memset( sm3_hash0, 0, sizeof sm3_hash0 );
|
||||
uint32_t sm3_hash1[32] __attribute__ ((aligned (32)));
|
||||
memset( sm3_hash1, 0, sizeof sm3_hash1 );
|
||||
uint32_t sm3_hash2[32] __attribute__ ((aligned (32)));
|
||||
memset( sm3_hash2, 0, sizeof sm3_hash2 );
|
||||
uint32_t sm3_hash3[32] __attribute__ ((aligned (32)));
|
||||
memset( sm3_hash3, 0, sizeof sm3_hash3 );
|
||||
|
||||
sm3_4way( &ctx.sm3, vhash, 64 );
|
||||
sm3_4way_close( &ctx.sm3, sm3_vhash );
|
||||
dintrlv_4x32( hash0, hash1, hash2, hash3, sm3_vhash, 512 );
|
||||
*/
|
||||
|
||||
// Hamsi parallel 4x32x2
|
||||
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
|
||||
hamsi512_4way( &ctx.hamsi, vhash, 64 );
|
||||
hamsi512_4way_close( &ctx.hamsi, vhash );
|
||||
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
|
||||
|
||||
// Fugue serial
|
||||
sph_fugue512( &ctx.fugue, hash0, 64 );
|
||||
sph_fugue512_close( &ctx.fugue, hash0 );
|
||||
memcpy( &ctx.fugue, &x13bcd_4way_ctx.fugue, sizeof(sph_fugue512_context) );
|
||||
sph_fugue512( &ctx.fugue, hash1, 64 );
|
||||
sph_fugue512_close( &ctx.fugue, hash1 );
|
||||
memcpy( &ctx.fugue, &x13bcd_4way_ctx.fugue, sizeof(sph_fugue512_context) );
|
||||
sph_fugue512( &ctx.fugue, hash2, 64 );
|
||||
sph_fugue512_close( &ctx.fugue, hash2 );
|
||||
memcpy( &ctx.fugue, &x13bcd_4way_ctx.fugue, sizeof(sph_fugue512_context) );
|
||||
sph_fugue512( &ctx.fugue, hash3, 64 );
|
||||
sph_fugue512_close( &ctx.fugue, hash3 );
|
||||
|
||||
memcpy( state, hash0, 32 );
|
||||
memcpy( state+32, hash1, 32 );
|
||||
memcpy( state+64, hash2, 32 );
|
||||
memcpy( state+96, hash3, 32 );
|
||||
}
|
||||
|
||||
int scanhash_x13bcd_4way( struct work *work, uint32_t max_nonce,
|
||||
uint64_t *hashes_done, struct thr_info *mythr )
|
||||
{
|
||||
uint32_t hash[4*8] __attribute__ ((aligned (64)));
|
||||
uint32_t vdata[24*4] __attribute__ ((aligned (64)));
|
||||
uint32_t *pdata = work->data;
|
||||
uint32_t *ptarget = work->target;
|
||||
uint32_t n = pdata[19];
|
||||
const uint32_t first_nonce = pdata[19];
|
||||
__m256i *noncev = (__m256i*)vdata + 9; // aligned
|
||||
int thr_id = mythr->id; // thr_id arg is deprecated
|
||||
const uint32_t Htarg = ptarget[7];
|
||||
uint64_t htmax[] = { 0, 0xF, 0xFF,
|
||||
0xFFF, 0xFFFF, 0x10000000 };
|
||||
uint32_t masks[] = { 0xFFFFFFFF, 0xFFFFFFF0, 0xFFFFFF00,
|
||||
0xFFFFF000, 0xFFFF0000, 0 };
|
||||
|
||||
mm256_bswap32_intrlv80_4x64( vdata, pdata );
|
||||
|
||||
blake512_4way_init( &x13bcd_ctx_mid );
|
||||
blake512_4way( &x13bcd_ctx_mid, vdata, 64 );
|
||||
|
||||
for ( int m=0; m < 6; m++ )
|
||||
if ( Htarg <= htmax[m] )
|
||||
{
|
||||
uint32_t mask = masks[m];
|
||||
do
|
||||
{
|
||||
*noncev = mm256_intrlv_blend_32( mm256_bswap_32(
|
||||
_mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ) ), *noncev );
|
||||
|
||||
x13bcd_4way_hash( hash, vdata );
|
||||
pdata[19] = n;
|
||||
|
||||
for ( int i = 0; i < 4; i++ )
|
||||
if ( ( ( (hash+(i<<3))[7] & mask ) == 0 ) )
|
||||
if ( fulltest( hash+(i<<3), ptarget ) && !opt_benchmark )
|
||||
{
|
||||
pdata[19] = n+i;
|
||||
submit_lane_solution( work, hash+(i<<3), mythr, i );
|
||||
}
|
||||
n += 4;
|
||||
} while ( ( n < max_nonce ) && !work_restart[thr_id].restart );
|
||||
break;
|
||||
}
|
||||
|
||||
*hashes_done = n - first_nonce + 1;
|
||||
return 0;
|
||||
}
|
||||
|
||||
#endif
|
||||
258
algo/x13/x13bcd.c
Normal file
258
algo/x13/x13bcd.c
Normal file
@@ -0,0 +1,258 @@
|
||||
#include "x13sm3-gate.h"
|
||||
|
||||
#include <stdlib.h>
|
||||
#include <stdint.h>
|
||||
#include <string.h>
|
||||
#include <stdio.h>
|
||||
|
||||
#include "algo/groestl/sph_groestl.h"
|
||||
#include "algo/shavite/sph_shavite.h"
|
||||
#include "algo/luffa/sph_luffa.h"
|
||||
#include "algo/cubehash/sph_cubehash.h"
|
||||
#include "algo/simd/sph_simd.h"
|
||||
#include "algo/echo/sph_echo.h"
|
||||
#include "algo/hamsi/sph_hamsi.h"
|
||||
#include "algo/fugue/sph_fugue.h"
|
||||
#include "algo/sm3/sph_sm3.h"
|
||||
|
||||
//#include "algo/luffa/luffa_for_sse2.h"
|
||||
#include "algo/cubehash/cubehash_sse2.h"
|
||||
#include "algo/simd/nist.h"
|
||||
#include "algo/blake/sse2/blake.c"
|
||||
#include "algo/bmw/sse2/bmw.c"
|
||||
#include "algo/keccak/sse2/keccak.c"
|
||||
#include "algo/skein/sse2/skein.c"
|
||||
#include "algo/jh/sse2/jh_sse2_opt64.h"
|
||||
|
||||
#ifndef NO_AES_NI
|
||||
#include "algo/groestl/aes_ni/hash-groestl.h"
|
||||
#include "algo/echo/aes_ni/hash_api.h"
|
||||
#endif
|
||||
|
||||
typedef struct {
|
||||
#ifdef NO_AES_NI
|
||||
sph_groestl512_context groestl;
|
||||
sph_echo512_context echo;
|
||||
#else
|
||||
hashState_echo echo;
|
||||
hashState_groestl groestl;
|
||||
#endif
|
||||
// hashState_luffa luffa;
|
||||
cubehashParam cube;
|
||||
sph_shavite512_context shavite;
|
||||
hashState_sd simd;
|
||||
sm3_ctx_t sm3;
|
||||
sph_hamsi512_context hamsi;
|
||||
sph_fugue512_context fugue;
|
||||
} x13bcd_ctx_holder;
|
||||
|
||||
x13bcd_ctx_holder x13bcd_ctx;
|
||||
|
||||
void init_x13bcd_ctx()
|
||||
{
|
||||
#ifdef NO_AES_NI
|
||||
sph_groestl512_init(&x13bcd_ctx.groestl);
|
||||
sph_echo512_init(&x13bcd_ctx.echo);
|
||||
#else
|
||||
init_echo(&x13bcd_ctx.echo, 512);
|
||||
init_groestl(&x13bcd_ctx.groestl, 64 );
|
||||
#endif
|
||||
// init_luffa(&x13bcd_ctx.luffa,512);
|
||||
cubehashInit(&x13bcd_ctx.cube,512,16,32);
|
||||
sph_shavite512_init(&x13bcd_ctx.shavite);
|
||||
init_sd(&x13bcd_ctx.simd,512);
|
||||
sm3_init( &x13bcd_ctx.sm3 );
|
||||
sph_hamsi512_init(&x13bcd_ctx.hamsi);
|
||||
sph_fugue512_init(&x13bcd_ctx.fugue);
|
||||
};
|
||||
|
||||
void x13bcd_hash(void *output, const void *input)
|
||||
{
|
||||
unsigned char hash[128] __attribute__ ((aligned (32)));
|
||||
|
||||
x13bcd_ctx_holder ctx;
|
||||
memcpy(&ctx, &x13bcd_ctx, sizeof(x13bcd_ctx));
|
||||
|
||||
unsigned char hashbuf[128];
|
||||
size_t hashptr;
|
||||
sph_u64 hashctA;
|
||||
sph_u64 hashctB;
|
||||
|
||||
//---blake1---
|
||||
|
||||
DECL_BLK;
|
||||
BLK_I;
|
||||
BLK_W;
|
||||
BLK_C;
|
||||
|
||||
//---bmw2---
|
||||
|
||||
DECL_BMW;
|
||||
BMW_I;
|
||||
BMW_U;
|
||||
|
||||
#define M(x) sph_dec64le_aligned(data + 8 * (x))
|
||||
#define H(x) (h[x])
|
||||
#define dH(x) (dh[x])
|
||||
|
||||
BMW_C;
|
||||
|
||||
#undef M
|
||||
#undef H
|
||||
#undef dH
|
||||
|
||||
//---groestl----
|
||||
|
||||
#ifdef NO_AES_NI
|
||||
sph_groestl512 (&ctx.groestl, hash, 64);
|
||||
sph_groestl512_close(&ctx.groestl, hash);
|
||||
#else
|
||||
update_and_final_groestl( &ctx.groestl, (char*)hash,
|
||||
(const char*)hash, 512 );
|
||||
#endif
|
||||
|
||||
//---skein4---
|
||||
|
||||
DECL_SKN;
|
||||
SKN_I;
|
||||
SKN_U;
|
||||
SKN_C;
|
||||
|
||||
//---jh5------
|
||||
|
||||
DECL_JH;
|
||||
JH_H;
|
||||
|
||||
//---keccak6---
|
||||
|
||||
DECL_KEC;
|
||||
KEC_I;
|
||||
KEC_U;
|
||||
KEC_C;
|
||||
|
||||
uint32_t sm3_hash[32] __attribute__ ((aligned (32)));
|
||||
memset(sm3_hash, 0, sizeof sm3_hash);
|
||||
|
||||
sph_sm3(&ctx.sm3, hash, 64);
|
||||
sph_sm3_close(&ctx.sm3, sm3_hash);
|
||||
|
||||
cubehashUpdateDigest( &ctx.cube, (byte*) hash,
|
||||
(const byte*)sm3_hash, 64 );
|
||||
|
||||
/*
|
||||
//--- luffa7
|
||||
update_and_final_luffa( &ctx.luffa, (BitSequence*)hash,
|
||||
(const BitSequence*)hash, 64 );
|
||||
|
||||
// 8 Cube
|
||||
cubehashUpdateDigest( &ctx.cube, (byte*) hash,
|
||||
(const byte*)hash, 64 );
|
||||
*/
|
||||
|
||||
// 9 Shavite
|
||||
sph_shavite512( &ctx.shavite, hash, 64);
|
||||
sph_shavite512_close( &ctx.shavite, hash);
|
||||
|
||||
// 10 Simd
|
||||
update_final_sd( &ctx.simd, (BitSequence *)hash,
|
||||
(const BitSequence *)hash, 512 );
|
||||
|
||||
//11---echo---
|
||||
#ifdef NO_AES_NI
|
||||
sph_echo512(&ctx.echo, hash, 64);
|
||||
sph_echo512_close(&ctx.echo, hash);
|
||||
#else
|
||||
update_final_echo ( &ctx.echo, (BitSequence *)hash,
|
||||
(const BitSequence *)hash, 512 );
|
||||
#endif
|
||||
|
||||
/*
|
||||
uint32_t sm3_hash[32] __attribute__ ((aligned (32)));
|
||||
memset(sm3_hash, 0, sizeof sm3_hash);
|
||||
|
||||
sph_sm3(&ctx.sm3, hash, 64);
|
||||
sph_sm3_close(&ctx.sm3, sm3_hash);
|
||||
|
||||
sph_hamsi512(&ctx.hamsi, sm3_hash, 64);
|
||||
*/
|
||||
|
||||
sph_hamsi512(&ctx.hamsi, hash, 64);
|
||||
sph_hamsi512_close(&ctx.hamsi, hash);
|
||||
|
||||
sph_fugue512(&ctx.fugue, hash, 64);
|
||||
sph_fugue512_close(&ctx.fugue, hash);
|
||||
|
||||
asm volatile ("emms");
|
||||
memcpy(output, hash, 32);
|
||||
}
|
||||
|
||||
int scanhash_x13bcd( struct work *work, uint32_t max_nonce,
|
||||
uint64_t *hashes_done, struct thr_info *mythr)
|
||||
{
|
||||
uint32_t endiandata[20] __attribute__((aligned(64)));
|
||||
uint32_t hash64[8] __attribute__((aligned(64)));
|
||||
uint32_t *pdata = work->data;
|
||||
uint32_t *ptarget = work->target;
|
||||
uint32_t n = pdata[19] - 1;
|
||||
const uint32_t first_nonce = pdata[19];
|
||||
int thr_id = mythr->id; // thr_id arg is deprecated
|
||||
const uint32_t Htarg = ptarget[7];
|
||||
|
||||
uint64_t htmax[] = {
|
||||
0,
|
||||
0xF,
|
||||
0xFF,
|
||||
0xFFF,
|
||||
0xFFFF,
|
||||
0x10000000
|
||||
};
|
||||
uint32_t masks[] = {
|
||||
0xFFFFFFFF,
|
||||
0xFFFFFFF0,
|
||||
0xFFFFFF00,
|
||||
0xFFFFF000,
|
||||
0xFFFF0000,
|
||||
0
|
||||
};
|
||||
|
||||
// we need bigendian data...
|
||||
swab32_array( endiandata, pdata, 20 );
|
||||
|
||||
#ifdef DEBUG_ALGO
|
||||
if (Htarg != 0)
|
||||
printf("[%d] Htarg=%X\n", thr_id, Htarg);
|
||||
#endif
|
||||
for (int m=0; m < 6; m++) {
|
||||
if (Htarg <= htmax[m]) {
|
||||
uint32_t mask = masks[m];
|
||||
do {
|
||||
pdata[19] = ++n;
|
||||
be32enc(&endiandata[19], n);
|
||||
x13bcd_hash(hash64, endiandata);
|
||||
#ifndef DEBUG_ALGO
|
||||
if ((!(hash64[7] & mask)) && fulltest(hash64, ptarget)) {
|
||||
*hashes_done = n - first_nonce + 1;
|
||||
return true;
|
||||
}
|
||||
#else
|
||||
if (!(n % 0x1000) && !thr_id) printf(".");
|
||||
if (!(hash64[7] & mask)) {
|
||||
printf("[%d]",thr_id);
|
||||
if (fulltest(hash64, ptarget)) {
|
||||
work_set_target_ratio( work, hash64 );
|
||||
*hashes_done = n - first_nonce + 1;
|
||||
return true;
|
||||
}
|
||||
}
|
||||
#endif
|
||||
} while (n < max_nonce && !work_restart[thr_id].restart);
|
||||
// see blake.c if else to understand the loop on htmax => mask
|
||||
break;
|
||||
}
|
||||
}
|
||||
|
||||
*hashes_done = n - first_nonce + 1;
|
||||
pdata[19] = n;
|
||||
return 0;
|
||||
}
|
||||
|
||||
@@ -16,3 +16,19 @@ bool register_x13sm3_algo( algo_gate_t* gate )
|
||||
return true;
|
||||
};
|
||||
|
||||
bool register_x13bcd_algo( algo_gate_t* gate )
|
||||
{
|
||||
#if defined (X13SM3_4WAY)
|
||||
init_x13bcd_4way_ctx();
|
||||
gate->scanhash = (void*)&scanhash_x13bcd_4way;
|
||||
gate->hash = (void*)&x13bcd_4way_hash;
|
||||
#else
|
||||
init_x13bcd_ctx();
|
||||
gate->scanhash = (void*)&scanhash_x13bcd;
|
||||
gate->hash = (void*)&x13bcd_hash;
|
||||
#endif
|
||||
gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT;
|
||||
gate->get_max64 = (void*)&get_max64_0x3ffff;
|
||||
return true;
|
||||
};
|
||||
|
||||
|
||||
@@ -10,23 +10,31 @@
|
||||
|
||||
bool register_x13sm3_algo( algo_gate_t* gate );
|
||||
|
||||
bool register_x13bcd_algo( algo_gate_t* gate );
|
||||
|
||||
#if defined(X13SM3_4WAY)
|
||||
|
||||
void x13sm3_4way_hash( void *state, const void *input );
|
||||
|
||||
int scanhash_x13sm3_4way( struct work *work, uint32_t max_nonce,
|
||||
uint64_t *hashes_done, struct thr_info *mythr );
|
||||
|
||||
void init_x13sm3_4way_ctx();
|
||||
|
||||
void x13bcd_4way_hash( void *state, const void *input );
|
||||
int scanhash_x13bcd_4way( struct work *work, uint32_t max_nonce,
|
||||
uint64_t *hashes_done, struct thr_info *mythr );
|
||||
void init_x13bcd_4way_ctx();
|
||||
|
||||
#endif
|
||||
|
||||
void x13sm3_hash( void *state, const void *input );
|
||||
|
||||
int scanhash_x13sm3( struct work *work, uint32_t max_nonce,
|
||||
uint64_t *hashes_done, struct thr_info *mythr );
|
||||
|
||||
void init_x13sm3_ctx();
|
||||
|
||||
void x13bcd_hash( void *state, const void *input );
|
||||
int scanhash_x13bcd( struct work *work, uint32_t max_nonce,
|
||||
uint64_t *hashes_done, struct thr_info *mythr );
|
||||
void init_x13bcd_ctx();
|
||||
|
||||
#endif
|
||||
|
||||
|
||||
Reference in New Issue
Block a user