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

v3.11.0

Browse Source
This commit is contained in:
Jay D Dee
2020-01-02 23:54:08 -05:00
parent 241bc26767
commit 3572cb53c4
118 changed files with 7030 additions and 1575 deletions
Download Patch File Download Diff File Expand all files Collapse all files

89
algo/x22/x22i-4way.c
View File

@@ -22,6 +22,11 @@
#include "algo/lyra2/lyra2.h"
#include "algo/gost/sph_gost.h"
#include "algo/swifftx/swifftx.h"
#if defined(__VAES__)
#include "algo/groestl/groestl512-hash-4way.h"
#include "algo/shavite/shavite-hash-4way.h"
#include "algo/echo/echo-hash-4way.h"
#endif
#if defined(X22I_8WAY)
@@ -30,14 +35,11 @@ union _x22i_8way_ctx_overlay
{
blake512_8way_context blake;
bmw512_8way_context bmw;
hashState_groestl groestl;
hashState_echo echo;
skein512_8way_context skein;
jh512_8way_context jh;
keccak512_8way_context keccak;
luffa_4way_context luffa;
cube_4way_context cube;
sph_shavite512_context shavite;
simd_4way_context simd;
hamsi512_8way_context hamsi;
sph_fugue512_context fugue;
@@ -48,6 +50,15 @@ union _x22i_8way_ctx_overlay
sph_tiger_context tiger;
sph_gost512_context gost;
sha256_8way_context sha256;
#if defined(__VAES__)
groestl512_4way_context groestl;
shavite512_4way_context shavite;
echo_4way_context echo;
#else
hashState_groestl groestl;
sph_shavite512_context shavite;
hashState_echo echo;
#endif
};
typedef union _x22i_8way_ctx_overlay x22i_8way_ctx_overlay;
@@ -84,6 +95,19 @@ void x22i_8way_hash( void *output, const void *input )
bmw512_8way_update( &ctx.bmw, vhash, 64 );
bmw512_8way_close( &ctx.bmw, vhash );
#if defined(__VAES__)
rintrlv_8x64_4x128( vhashA, vhashB, vhash, 512 );
groestl512_4way_init( &ctx.groestl, 64 );
groestl512_4way_update_close( &ctx.groestl, vhashA, vhashA, 512 );
groestl512_4way_init( &ctx.groestl, 64 );
groestl512_4way_update_close( &ctx.groestl, vhashB, vhashB, 512 );
rintrlv_4x128_8x64( vhash, vhashA, vhashB, 512 );
#else
dintrlv_8x64_512( hash0, hash1, hash2, hash3,
hash4, hash5, hash6, hash7, vhash );
@@ -115,6 +139,8 @@ void x22i_8way_hash( void *output, const void *input )
intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3,
hash4, hash5, hash6, hash7 );
#endif
skein512_8way_init( &ctx.skein );
skein512_8way_update( &ctx.skein, vhash, 64 );
skein512_8way_close( &ctx.skein, vhash );
@@ -139,6 +165,15 @@ void x22i_8way_hash( void *output, const void *input )
cube_4way_init( &ctx.cube, 512, 16, 32 );
cube_4way_update_close( &ctx.cube, vhashB, vhashB, 64 );
#if defined(__VAES__)
shavite512_4way_init( &ctx.shavite );
shavite512_4way_update_close( &ctx.shavite, vhashA, vhashA, 64 );
shavite512_4way_init( &ctx.shavite );
shavite512_4way_update_close( &ctx.shavite, vhashB, vhashB, 64 );
#else
dintrlv_4x128_512( hash0, hash1, hash2, hash3, vhashA );
dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhashB );
@@ -170,11 +205,24 @@ void x22i_8way_hash( void *output, const void *input )
intrlv_4x128_512( vhashA, hash0, hash1, hash2, hash3 );
intrlv_4x128_512( vhashB, hash4, hash5, hash6, hash7 );
#endif
simd_4way_init( &ctx.simd, 512 );
simd_4way_update_close( &ctx.simd, vhashA, vhashA, 512 );
simd_4way_init( &ctx.simd, 512 );
simd_4way_update_close( &ctx.simd, vhashB, vhashB, 512 );
#if defined(__VAES__)
echo_4way_init( &ctx.echo, 512 );
echo_4way_update_close( &ctx.echo, vhashA, vhashA, 512 );
echo_4way_init( &ctx.echo, 512 );
echo_4way_update_close( &ctx.echo, vhashB, vhashB, 512 );
rintrlv_4x128_8x64( vhash, vhashA, vhashB, 512 );
#else
dintrlv_4x128_512( hash0, hash1, hash2, hash3, vhashA );
dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhashB );
@@ -206,6 +254,8 @@ void x22i_8way_hash( void *output, const void *input )
intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3,
hash4, hash5, hash6, hash7 );
#endif
hamsi512_8way_init( &ctx.hamsi );
hamsi512_8way_update( &ctx.hamsi, vhash, 64 );
hamsi512_8way_close( &ctx.hamsi, vhash );
@@ -439,10 +489,8 @@ int scanhash_x22i_8way( struct work* work, uint32_t max_nonce,
return 0;
}
#elif defined(X22I_4WAY)
union _x22i_4way_ctx_overlay
{
blake512_4way_context blake;
@@ -477,8 +525,6 @@ void x22i_4way_hash( void *output, const void *input )
uint64_t vhash[8*4] __attribute__ ((aligned (64)));
uint64_t vhashA[8*4] __attribute__ ((aligned (64)));
uint64_t vhashB[8*4] __attribute__ ((aligned (64)));
// unsigned char hash[64 * 4] __attribute__((aligned(64))) = {0};
unsigned char hashA0[64] __attribute__((aligned(64))) = {0};
unsigned char hashA1[64] __attribute__((aligned(32))) = {0};
unsigned char hashA2[64] __attribute__((aligned(32))) = {0};
@@ -486,13 +532,12 @@ void x22i_4way_hash( void *output, const void *input )
x22i_ctx_overlay ctx;
blake512_4way_init( &ctx.blake );
blake512_4way( &ctx.blake, input, 80 );
blake512_4way_update( &ctx.blake, input, 80 );
blake512_4way_close( &ctx.blake, vhash );
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_update( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );
init_groestl( &ctx.groestl, 64 );
@@ -511,15 +556,15 @@ void x22i_4way_hash( void *output, const void *input )
intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, 64 );
skein512_4way_update( &ctx.skein, vhash, 64 );
skein512_4way_close( &ctx.skein, vhash );
jh512_4way_init( &ctx.jh );
jh512_4way( &ctx.jh, vhash, 64 );
jh512_4way_update( &ctx.jh, vhash, 64 );
jh512_4way_close( &ctx.jh, vhash );
keccak512_4way_init( &ctx.keccak );
keccak512_4way( &ctx.keccak, vhash, 64 );
keccak512_4way_update( &ctx.keccak, vhash, 64 );
keccak512_4way_close( &ctx.keccak, vhash );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
@@ -560,13 +605,11 @@ void x22i_4way_hash( void *output, const void *input )
update_final_echo ( &ctx.echo, (BitSequence*)hash3,
(const BitSequence*)hash3, 512 );
intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, 64 );
hamsi512_4way_update( &ctx.hamsi, vhash, 64 );
hamsi512_4way_close( &ctx.hamsi, vhash );
dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );
sph_fugue512_init( &ctx.fugue );
@@ -585,9 +628,8 @@ void x22i_4way_hash( void *output, const void *input )
intrlv_4x32_512( vhash, hash0, hash1, hash2, hash3 );
shabal512_4way_init( &ctx.shabal );
shabal512_4way( &ctx.shabal, vhash, 64 );
shabal512_4way_update( &ctx.shabal, vhash, 64 );
shabal512_4way_close( &ctx.shabal, vhash );
dintrlv_4x32_512( &hash0[8], &hash1[8], &hash2[8], &hash3[8], vhash );
sph_whirlpool_init( &ctx.whirlpool );
@@ -606,12 +648,10 @@ void x22i_4way_hash( void *output, const void *input )
intrlv_4x64_512( vhash, &hash0[16], &hash1[16], &hash2[16], &hash3[16] );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, 64 );
sha512_4way_update( &ctx.sha512, vhash, 64 );
sha512_4way_close( &ctx.sha512, vhash );
dintrlv_4x64_512( &hash0[24], &hash1[24], &hash2[24], &hash3[24], vhash );
// InitializeSWIFFTX();
ComputeSingleSWIFFTX((unsigned char*)hash0, (unsigned char*)hashA0);
ComputeSingleSWIFFTX((unsigned char*)hash1, (unsigned char*)hashA1);
ComputeSingleSWIFFTX((unsigned char*)hash2, (unsigned char*)hashA2);
@@ -622,9 +662,8 @@ void x22i_4way_hash( void *output, const void *input )
memset( vhash, 0, 64*4 );
haval256_5_4way_init( &ctx.haval );
haval256_5_4way( &ctx.haval, vhashA, 64 );
haval256_5_4way_update( &ctx.haval, vhashA, 64 );
haval256_5_4way_close( &ctx.haval, vhash );
dintrlv_4x32_512( hash0, hash1, hash2, hash3, vhash );
memset( hashA0, 0, 64 );
@@ -675,10 +714,8 @@ void x22i_4way_hash( void *output, const void *input )
intrlv_4x32_512( vhash, hash0, hash1, hash2, hash3 );
sha256_4way_init( &ctx.sha256 );
sha256_4way( &ctx.sha256, vhash, 64 );
sha256_4way_update( &ctx.sha256, vhash, 64 );
sha256_4way_close( &ctx.sha256, output );
// memcpy(output, hash, 32);
}