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

417
algo/x17/sonoa-4way.c
View File

@@ -21,6 +21,11 @@
#include "algo/whirlpool/sph_whirlpool.h"
#include "algo/haval/haval-hash-4way.h"
#include "algo/sha/sha-hash-4way.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(SONOA_8WAY)
@@ -28,21 +33,27 @@ union _sonoa_8way_context_overlay
{
blake512_8way_context blake;
bmw512_8way_context bmw;
hashState_groestl groestl;
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;
hashState_echo echo;
hamsi512_8way_context hamsi;
sph_fugue512_context fugue;
shabal512_8way_context shabal;
sph_whirlpool_context whirlpool;
sha512_8way_context sha512;
haval256_5_8way_context haval;
#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
} __attribute__ ((aligned (64)));
typedef union _sonoa_8way_context_overlay sonoa_8way_context_overlay;
@@ -72,6 +83,19 @@ void sonoa_8way_hash( void *state, 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 );
@@ -95,6 +119,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -119,6 +145,15 @@ void sonoa_8way_hash( void *state, 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 );
@@ -150,11 +185,24 @@ void sonoa_8way_hash( void *state, 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 );
@@ -182,16 +230,31 @@ void sonoa_8way_hash( void *state, const void *input )
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash7,
(const BitSequence *) hash7, 512 );
// 2
intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
hash7 );
#endif
// 2
bmw512_8way_init( &ctx.bmw );
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 );
@@ -215,6 +278,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -239,6 +304,15 @@ void sonoa_8way_hash( void *state, 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 );
@@ -270,11 +344,24 @@ void sonoa_8way_hash( void *state, 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 );
@@ -306,6 +393,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -316,6 +405,19 @@ void sonoa_8way_hash( void *state, 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 );
@@ -339,6 +441,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -363,6 +467,15 @@ void sonoa_8way_hash( void *state, 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 );
@@ -394,11 +507,24 @@ void sonoa_8way_hash( void *state, 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 );
@@ -430,6 +556,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -471,6 +599,19 @@ void sonoa_8way_hash( void *state, 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 );
@@ -494,6 +635,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -518,6 +661,15 @@ void sonoa_8way_hash( void *state, 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 );
@@ -549,11 +701,24 @@ void sonoa_8way_hash( void *state, 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 );
@@ -585,6 +750,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -630,6 +797,17 @@ void sonoa_8way_hash( void *state, const void *input )
hamsi512_8way_update( &ctx.hamsi, vhashA, 64 );
hamsi512_8way_close( &ctx.hamsi, vhash );
#if defined(__VAES__)
rintrlv_8x64_4x128( vhashA, vhashB, vhash, 512 );
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 );
#else
dintrlv_8x64_512( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
vhash );
@@ -658,6 +836,18 @@ void sonoa_8way_hash( void *state, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash7,
(const BitSequence *) hash7, 512 );
#endif
#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 );
rintrlv_4x128_8x64( vhash, vhashA, vhashB, 512 );
#else
sph_shavite512_init( &ctx.shavite );
sph_shavite512( &ctx.shavite, hash0, 64 );
@@ -684,11 +874,13 @@ void sonoa_8way_hash( void *state, const void *input )
sph_shavite512( &ctx.shavite, hash7, 64 );
sph_shavite512_close( &ctx.shavite, hash7 );
// 5
intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
hash7 );
#endif
// 5
bmw512_8way_init( &ctx.bmw );
bmw512_8way_update( &ctx.bmw, vhash, 64 );
bmw512_8way_close( &ctx.bmw, vhash );
@@ -699,6 +891,19 @@ void sonoa_8way_hash( void *state, const void *input )
shabal512_8way_update( &ctx.shabal, vhashA, 64 );
shabal512_8way_close( &ctx.shabal, vhash );
#if defined(__VAES__)
rintrlv_8x32_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_8x32_512( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
vhash );
@@ -722,6 +927,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -746,6 +953,15 @@ void sonoa_8way_hash( void *state, 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 );
@@ -777,14 +993,27 @@ void sonoa_8way_hash( void *state, 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 );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
(const BitSequence *) hash0, 512 );
@@ -813,6 +1042,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -889,6 +1120,19 @@ void sonoa_8way_hash( void *state, 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 );
@@ -912,6 +1156,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -936,6 +1182,15 @@ void sonoa_8way_hash( void *state, 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 );
@@ -967,11 +1222,24 @@ void sonoa_8way_hash( void *state, 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 );
@@ -1003,6 +1271,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -1114,6 +1384,19 @@ void sonoa_8way_hash( void *state, 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 );
@@ -1137,6 +1420,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -1161,6 +1446,15 @@ void sonoa_8way_hash( void *state, 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 );
@@ -1192,11 +1486,24 @@ void sonoa_8way_hash( void *state, 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 );
@@ -1228,6 +1535,8 @@ void sonoa_8way_hash( void *state, 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 );
@@ -1319,7 +1628,7 @@ int scanhash_sonoa_8way( struct work *work, uint32_t max_nonce,
uint32_t *pdata = work->data;
const uint32_t *ptarget = work->target;
const uint32_t first_nonce = pdata[19];
const uint32_t last_nonce = max_nonce - 8;
const uint32_t last_nonce = max_nonce - 8;
__m512i *noncev = (__m512i*)vdata + 9; // aligned
uint32_t n = first_nonce;
const int thr_id = mythr->id;
@@ -1350,8 +1659,6 @@ int scanhash_sonoa_8way( struct work *work, uint32_t max_nonce,
return 0;
}
#elif defined(SONOA_4WAY)
union _sonoa_4way_context_overlay
@@ -1391,11 +1698,11 @@ void sonoa_4way_hash( void *state, const void *input )
// 1
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 );
@@ -1412,15 +1719,15 @@ void sonoa_4way_hash( void *state, 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 );
@@ -1466,7 +1773,7 @@ void sonoa_4way_hash( void *state, const void *input )
intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
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 );
@@ -1483,15 +1790,15 @@ void sonoa_4way_hash( void *state, 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 );
@@ -1535,13 +1842,13 @@ void sonoa_4way_hash( void *state, const void *input )
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 );
// 3
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 );
@@ -1558,15 +1865,15 @@ void sonoa_4way_hash( void *state, 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 );
@@ -1610,7 +1917,7 @@ void sonoa_4way_hash( void *state, const void *input )
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 );
@@ -1632,7 +1939,7 @@ void sonoa_4way_hash( void *state, const void *input )
intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
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 );
@@ -1649,15 +1956,15 @@ void sonoa_4way_hash( void *state, 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 );
@@ -1701,7 +2008,7 @@ void sonoa_4way_hash( void *state, const void *input )
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 );
@@ -1722,13 +2029,13 @@ void sonoa_4way_hash( void *state, 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 );
rintrlv_4x32_4x64( vhashB, vhash, 512 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhashB, 64 );
hamsi512_4way_update( &ctx.hamsi, vhashB, 64 );
hamsi512_4way_close( &ctx.hamsi, vhash );
dintrlv_4x64_512( hash0, hash1, hash2, hash3, vhash );
@@ -1758,13 +2065,13 @@ void sonoa_4way_hash( void *state, const void *input )
rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_update( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
rintrlv_4x64_4x32( vhashB, vhash, 512 );
shabal512_4way_init( &ctx.shabal );
shabal512_4way( &ctx.shabal, vhashB, 64 );
shabal512_4way_update( &ctx.shabal, vhashB, 64 );
shabal512_4way_close( &ctx.shabal, vhash );
dintrlv_4x32_512( hash0, hash1, hash2, hash3, vhash );
@@ -1781,15 +2088,15 @@ void sonoa_4way_hash( void *state, 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 );
@@ -1833,7 +2140,7 @@ void sonoa_4way_hash( void *state, const void *input )
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 );
@@ -1854,7 +2161,7 @@ void sonoa_4way_hash( void *state, 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, hash1, hash2, hash3, vhash );
@@ -1877,7 +2184,7 @@ void sonoa_4way_hash( void *state, const void *input )
intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
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 );
@@ -1894,15 +2201,15 @@ void sonoa_4way_hash( void *state, 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 );
@@ -1946,7 +2253,7 @@ void sonoa_4way_hash( void *state, const void *input )
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 );
@@ -1967,7 +2274,7 @@ void sonoa_4way_hash( void *state, 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, hash1, hash2, hash3, vhash );
@@ -1988,7 +2295,7 @@ void sonoa_4way_hash( void *state, const void *input )
intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
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, hash1, hash2, hash3, vhash );
@@ -2011,7 +2318,7 @@ void sonoa_4way_hash( void *state, const void *input )
intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
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 );
@@ -2028,15 +2335,15 @@ void sonoa_4way_hash( void *state, 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 );
@@ -2080,7 +2387,7 @@ void sonoa_4way_hash( void *state, const void *input )
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 );
@@ -2101,7 +2408,7 @@ void sonoa_4way_hash( void *state, 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, hash1, hash2, hash3, vhash );
@@ -2122,13 +2429,13 @@ void sonoa_4way_hash( void *state, const void *input )
intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, 64 );
sha512_4way_update( &ctx.sha512, vhash, 64 );
sha512_4way_close( &ctx.sha512, vhash );
rintrlv_4x64_4x32( vhashB, vhash, 512 );
haval256_5_4way_init( &ctx.haval );
haval256_5_4way( &ctx.haval, vhashB, 64 );
haval256_5_4way_update( &ctx.haval, vhashB, 64 );
haval256_5_4way_close( &ctx.haval, state );
}

2
algo/x17/sonoa-gate.c
View File

@@ -13,7 +13,7 @@ bool register_sonoa_algo( algo_gate_t* gate )
gate->scanhash = (void*)&scanhash_sonoa;
gate->hash = (void*)&sonoa_hash;
#endif
gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT;
gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
return true;
};

142
algo/x17/x17-4way.c
View File

@@ -13,6 +13,11 @@
#include "algo/cubehash/cube-hash-2way.h"
#include "algo/shavite/sph_shavite.h"
#include "algo/shavite/shavite-hash-2way.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
#include "algo/simd/simd-hash-2way.h"
#include "algo/echo/aes_ni/hash_api.h"
#include "algo/hamsi/hamsi-hash-4way.h"
@@ -28,15 +33,21 @@ union _x17_8way_context_overlay
{
blake512_8way_context blake;
bmw512_8way_context bmw;
hashState_groestl groestl;
skein512_8way_context skein;
jh512_8way_context jh;
keccak512_8way_context keccak;
luffa_4way_context luffa;
cube_4way_context cube;
#if defined(__VAES__)
groestl512_4way_context groestl;
shavite512_4way_context shavite;
echo_4way_context echo;
#else
hashState_groestl groestl;
sph_shavite512_context shavite;
simd_4way_context simd;
hashState_echo echo;
#endif
simd_4way_context simd;
hamsi512_8way_context hamsi;
sph_fugue512_context fugue;
shabal512_8way_context shabal;
@@ -49,8 +60,8 @@ typedef union _x17_8way_context_overlay x17_8way_context_overlay;
void x17_8way_hash( void *state, const void *input )
{
uint64_t vhash[8*8] __attribute__ ((aligned (128)));
uint64_t vhash0[8*8] __attribute__ ((aligned (64)));
uint64_t vhash1[8*8] __attribute__ ((aligned (64)));
uint64_t vhashA[8*8] __attribute__ ((aligned (64)));
uint64_t vhashB[8*8] __attribute__ ((aligned (64)));
uint64_t hash0[8] __attribute__ ((aligned (64)));
uint64_t hash1[8] __attribute__ ((aligned (64)));
uint64_t hash2[8] __attribute__ ((aligned (64)));
@@ -61,7 +72,7 @@ void x17_8way_hash( void *state, const void *input )
uint64_t hash7[8] __attribute__ ((aligned (64)));
x17_8way_context_overlay ctx;
// 1 Blake parallel 4 way 64 bit
// 1 Blake
blake512_8way_init( &ctx.blake );
blake512_8way_update( &ctx.blake, input, 80 );
blake512_8way_close( &ctx.blake, vhash );
@@ -71,11 +82,24 @@ void x17_8way_hash( void *state, const void *input )
bmw512_8way_update( &ctx.bmw, vhash, 64 );
bmw512_8way_close( &ctx.bmw, vhash );
// Serialize
// 3 Groestl
#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 );
// 3 Groestl
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
init_groestl( &ctx.groestl, 64 );
@@ -93,10 +117,11 @@ void x17_8way_hash( void *state, const void *input )
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash7, (char*)hash7, 512 );
// Parallellize
intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
hash7 );
#endif
// 4 Skein parallel 4 way 64 bit
skein512_8way_init( &ctx.skein );
skein512_8way_update( &ctx.skein, vhash, 64 );
@@ -112,24 +137,34 @@ void x17_8way_hash( void *state, const void *input )
keccak512_8way_update( &ctx.keccak, vhash, 64 );
keccak512_8way_close( &ctx.keccak, vhash );
rintrlv_8x64_4x128( vhash0, vhash1, vhash, 512 );
rintrlv_8x64_4x128( vhashA, vhashB, vhash, 512 );
// 7 Luffa
luffa_4way_init( &ctx.luffa, 512 );
luffa_4way_update_close( &ctx.luffa, vhash0, vhash0, 64 );
luffa_4way_update_close( &ctx.luffa, vhashA, vhashA, 64 );
luffa_4way_init( &ctx.luffa, 512 );
luffa_4way_update_close( &ctx.luffa, vhash1, vhash1, 64 );
luffa_4way_update_close( &ctx.luffa, vhashB, vhashB, 64 );
// 8 Cubehash
cube_4way_init( &ctx.cube, 512, 16, 32 );
cube_4way_update_close( &ctx.cube, vhash0, vhash0, 64 );
cube_4way_update_close( &ctx.cube, vhashA, vhashA, 64 );
cube_4way_init( &ctx.cube, 512, 16, 32 );
cube_4way_update_close( &ctx.cube, vhash1, vhash1, 64 );
dintrlv_4x128_512( hash0, hash1, hash2, hash3, vhash0 );
dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhash1 );
cube_4way_update_close( &ctx.cube, vhashB, vhashB, 64 );
// 9 Shavite
#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 );
sph_shavite512_init( &ctx.shavite );
sph_shavite512( &ctx.shavite, hash0, 64 );
sph_shavite512_close( &ctx.shavite, hash0 );
@@ -155,18 +190,33 @@ void x17_8way_hash( void *state, const void *input )
sph_shavite512( &ctx.shavite, hash7, 64 );
sph_shavite512_close( &ctx.shavite, hash7 );
intrlv_4x128_512( vhashA, hash0, hash1, hash2, hash3 );
intrlv_4x128_512( vhashB, hash4, hash5, hash6, hash7 );
#endif
// 10 Simd
intrlv_4x128_512( vhash, hash0, hash1, hash2, hash3 );
simd_4way_init( &ctx.simd, 512 );
simd_4way_update_close( &ctx.simd, vhash, vhash, 512 );
dintrlv_4x128_512( hash0, hash1, hash2, hash3, vhash );
intrlv_4x128_512( vhash, hash4, hash5, hash6, hash7 );
simd_4way_update_close( &ctx.simd, vhashA, vhashA, 512 );
simd_4way_init( &ctx.simd, 512 );
simd_4way_update_close( &ctx.simd, vhash, vhash, 512 );
dintrlv_4x128_512( hash4, hash5, hash6, hash7, vhash );
simd_4way_update_close( &ctx.simd, vhashB, vhashB, 512 );
// 11 Echo
#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 );
// 11 Echo serial
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
(const BitSequence *) hash0, 512 );
@@ -192,9 +242,11 @@ void x17_8way_hash( void *state, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash7,
(const BitSequence *) hash7, 512 );
// 12 Hamsi parallel 4 way 64 bit
intrlv_8x64_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
hash7 );
#endif
// 12 Hamsi
hamsi512_8way_init( &ctx.hamsi );
hamsi512_8way_update( &ctx.hamsi, vhash, 64 );
@@ -229,7 +281,7 @@ void x17_8way_hash( void *state, const void *input )
sph_fugue512( &ctx.fugue, hash7, 64 );
sph_fugue512_close( &ctx.fugue, hash7 );
// 14 Shabal, parallel 4 way 32 bit
// 14 Shabal, parallel 8 way 32 bit
intrlv_8x32_512( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
hash7 );
@@ -275,10 +327,10 @@ void x17_8way_hash( void *state, const void *input )
sha512_8way_close( &ctx.sha512, vhash );
// 17 Haval parallel 32 bit
rintrlv_8x64_8x32( vhash0, vhash, 512 );
rintrlv_8x64_8x32( vhashA, vhash, 512 );
haval256_5_8way_init( &ctx.haval );
haval256_5_8way_update( &ctx.haval, vhash0, 64 );
haval256_5_8way_update( &ctx.haval, vhashA, 64 );
haval256_5_8way_close( &ctx.haval, state );
}
@@ -292,7 +344,7 @@ int scanhash_x17_8way( struct work *work, uint32_t max_nonce,
uint32_t *pdata = work->data;
const uint32_t *ptarget = work->target;
const uint32_t first_nonce = pdata[19];
const uint32_t last_nonce = max_nonce - 8;
const uint32_t last_nonce = max_nonce - 8;
__m512i *noncev = (__m512i*)vdata + 9; // aligned
uint32_t n = first_nonce;
const int thr_id = mythr->id;
@@ -349,23 +401,23 @@ typedef union _x17_4way_context_overlay x17_4way_context_overlay;
void x17_4way_hash( void *state, const void *input )
{
uint64_t vhash[8*4] __attribute__ ((aligned (128)));
uint64_t vhashA[8*4] __attribute__ ((aligned (64)));
uint64_t vhashB[8*4] __attribute__ ((aligned (64)));
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)));
uint64_t vhashA[8*4] __attribute__ ((aligned (64)));
uint64_t vhashB[8*4] __attribute__ ((aligned (64)));
x17_4way_context_overlay ctx;
// 1 Blake parallel 4 way 64 bit
blake512_4way_init( &ctx.blake );
blake512_4way( &ctx.blake, input, 80 );
blake512_4way_update( &ctx.blake, input, 80 );
blake512_4way_close( &ctx.blake, vhash );
// 2 Bmw
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_update( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
// Serialize
@@ -386,17 +438,17 @@ void x17_4way_hash( void *state, const void *input )
// 4 Skein parallel 4 way 64 bit
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, 64 );
skein512_4way_update( &ctx.skein, vhash, 64 );
skein512_4way_close( &ctx.skein, vhash );
// 5 JH
jh512_4way_init( &ctx.jh );
jh512_4way( &ctx.jh, vhash, 64 );
jh512_4way_update( &ctx.jh, vhash, 64 );
jh512_4way_close( &ctx.jh, vhash );
// 6 Keccak
keccak512_4way_init( &ctx.keccak );
keccak512_4way( &ctx.keccak, vhash, 64 );
keccak512_4way_update( &ctx.keccak, vhash, 64 );
keccak512_4way_close( &ctx.keccak, vhash );
// 7 Luffa parallel 2 way 128 bit
@@ -428,7 +480,6 @@ void x17_4way_hash( void *state, const void *input )
dintrlv_2x128_512( hash0, hash1, vhashA );
dintrlv_2x128_512( hash2, hash3, vhashB );
// 11 Echo serial
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -447,7 +498,7 @@ void x17_4way_hash( void *state, const void *input )
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 );
@@ -470,7 +521,7 @@ void x17_4way_hash( void *state, 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, hash1, hash2, hash3, vhash );
@@ -493,27 +544,28 @@ void x17_4way_hash( void *state, const void *input )
intrlv_4x64_512( vhash, hash0, hash1, hash2, hash3 );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, 64 );
sha512_4way_update( &ctx.sha512, vhash, 64 );
sha512_4way_close( &ctx.sha512, vhash );
// 17 Haval parallel 32 bit
rintrlv_4x64_4x32( vhashB, vhash, 512 );
haval256_5_4way_init( &ctx.haval );
haval256_5_4way( &ctx.haval, vhashB, 64 );
haval256_5_4way_update( &ctx.haval, vhashB, 64 );
haval256_5_4way_close( &ctx.haval, state );
}
int scanhash_x17_4way( struct work *work, uint32_t max_nonce,
uint64_t *hashes_done, struct thr_info *mythr )
{
uint32_t hash[4*16] __attribute__ ((aligned (64)));
uint32_t hash[16*4] __attribute__ ((aligned (128)));
uint32_t vdata[24*4] __attribute__ ((aligned (64)));
uint32_t lane_hash[8] __attribute__ ((aligned (32)));
uint32_t lane_hash[8] __attribute__ ((aligned (64)));
uint32_t *hash7 = &(hash[7<<2]);
uint32_t *pdata = work->data;
const uint32_t *ptarget = work->target;
const uint32_t first_nonce = pdata[19];
const uint32_t last_nonce = max_nonce -4;
__m256i *noncev = (__m256i*)vdata + 9; // aligned
uint32_t n = first_nonce;
const int thr_id = mythr->id;
@@ -537,9 +589,9 @@ int scanhash_x17_4way( struct work *work, uint32_t max_nonce,
}
}
n += 4;
} while ( likely( ( n < max_nonce - 4 ) && !work_restart[thr_id].restart ) );
} while ( likely( ( n < last_nonce ) && !work_restart[thr_id].restart ) );
*hashes_done = n - first_nonce + 1;
*hashes_done = n - first_nonce;
return 0;
}

2
algo/x17/x17-gate.c
View File

@@ -12,7 +12,7 @@ bool register_x17_algo( algo_gate_t* gate )
gate->scanhash = (void*)&scanhash_x17;
gate->hash = (void*)&x17_hash;
#endif
gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT;
gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
return true;
};

135
algo/x17/xevan-4way.c
View File

@@ -22,6 +22,11 @@
#include "algo/whirlpool/sph_whirlpool.h"
#include "algo/sha/sha-hash-4way.h"
#include "algo/haval/haval-hash-4way.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(XEVAN_8WAY)
@@ -29,13 +34,11 @@ union _xevan_8way_context_overlay
{
blake512_8way_context blake;
bmw512_8way_context bmw;
hashState_groestl groestl;
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;
hashState_echo echo;
hamsi512_8way_context hamsi;
@@ -44,6 +47,15 @@ union _xevan_8way_context_overlay
sph_whirlpool_context whirlpool;
sha512_8way_context sha512;
haval256_5_8way_context haval;
#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
} __attribute__ ((aligned (64)));
typedef union _xevan_8way_context_overlay xevan_8way_context_overlay;
@@ -72,6 +84,19 @@ void xevan_8way_hash( void *output, const void *input )
bmw512_8way_update( &ctx.bmw, vhash, dataLen );
bmw512_8way_close( &ctx.bmw, vhash );
#if defined(__VAES__)
rintrlv_8x64_4x128( vhashA, vhashB, vhash, dataLen<<3 );
groestl512_4way_init( &ctx.groestl, 64 );
groestl512_4way_update_close( &ctx.groestl, vhashA, vhashA, dataLen<<3 );
groestl512_4way_init( &ctx.groestl, 64 );
groestl512_4way_update_close( &ctx.groestl, vhashB, vhashB, dataLen<<3 );
rintrlv_4x128_8x64( vhash, vhashA, vhashB, dataLen<<3 );
#else
dintrlv_8x64( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
vhash, dataLen<<3 );
@@ -103,6 +128,8 @@ void xevan_8way_hash( void *output, const void *input )
intrlv_8x64( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
hash7, dataLen<<3 );
#endif
skein512_8way_init( &ctx.skein );
skein512_8way_update( &ctx.skein, vhash, dataLen );
skein512_8way_close( &ctx.skein, vhash );
@@ -127,6 +154,15 @@ void xevan_8way_hash( void *output, const void *input )
cube_4way_init( &ctx.cube, 512, 16, 32 );
cube_4way_update_close( &ctx.cube, vhashB, vhashB, dataLen );
#if defined(__VAES__)
shavite512_4way_init( &ctx.shavite );
shavite512_4way_update_close( &ctx.shavite, vhashA, vhashA, dataLen );
shavite512_4way_init( &ctx.shavite );
shavite512_4way_update_close( &ctx.shavite, vhashB, vhashB, dataLen );
#else
dintrlv_4x128( hash0, hash1, hash2, hash3, vhashA, dataLen<<3 );
dintrlv_4x128( hash4, hash5, hash6, hash7, vhashB, dataLen<<3 );
@@ -158,11 +194,26 @@ void xevan_8way_hash( void *output, const void *input )
intrlv_4x128( vhashA, hash0, hash1, hash2, hash3, dataLen<<3 );
intrlv_4x128( vhashB, hash4, hash5, hash6, hash7, dataLen<<3 );
#endif
simd_4way_init( &ctx.simd, 512 );
simd_4way_update_close( &ctx.simd, vhashA, vhashA, dataLen<<3 );
simd_4way_init( &ctx.simd, 512 );
simd_4way_update_close( &ctx.simd, vhashB, vhashB, dataLen<<3 );
/*
#if defined(__VAES__)
echo_4way_init( &ctx.echo, 512 );
echo_4way_update_close( &ctx.echo, vhashA, vhashA, dataLen<<3 );
echo_4way_init( &ctx.echo, 512 );
echo_4way_update_close( &ctx.echo, vhashB, vhashB, dataLen<<3 );
rintrlv_4x128_8x64( vhash, vhashA, vhashB, dataLen<<3 );
#else
*/
dintrlv_4x128( hash0, hash1, hash2, hash3, vhashA, dataLen<<3 );
dintrlv_4x128( hash4, hash5, hash6, hash7, vhashB, dataLen<<3 );
@@ -194,6 +245,8 @@ void xevan_8way_hash( void *output, const void *input )
intrlv_8x64( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
hash7, dataLen<<3 );
//#endif
hamsi512_8way_init( &ctx.hamsi );
hamsi512_8way_update( &ctx.hamsi, vhash, dataLen );
hamsi512_8way_close( &ctx.hamsi, vhash );
@@ -286,6 +339,19 @@ void xevan_8way_hash( void *output, const void *input )
bmw512_8way_update( &ctx.bmw, vhash, dataLen );
bmw512_8way_close( &ctx.bmw, vhash );
#if defined(__VAES__)
rintrlv_8x64_4x128( vhashA, vhashB, vhash, dataLen<<3 );
groestl512_4way_init( &ctx.groestl, 64 );
groestl512_4way_update_close( &ctx.groestl, vhashA, vhashA, dataLen<<3 );
groestl512_4way_init( &ctx.groestl, 64 );
groestl512_4way_update_close( &ctx.groestl, vhashB, vhashB, dataLen<<3 );
rintrlv_4x128_8x64( vhash, vhashA, vhashB, dataLen<<3 );
#else
dintrlv_8x64( hash0, hash1, hash2, hash3, hash4, hash5, hash6, hash7,
vhash, dataLen<<3 );
@@ -317,6 +383,8 @@ void xevan_8way_hash( void *output, const void *input )
intrlv_8x64( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
hash7, dataLen<<3 );
#endif
skein512_8way_init( &ctx.skein );
skein512_8way_update( &ctx.skein, vhash, dataLen );
skein512_8way_close( &ctx.skein, vhash );
@@ -341,6 +409,15 @@ void xevan_8way_hash( void *output, const void *input )
cube_4way_init( &ctx.cube, 512, 16, 32 );
cube_4way_update_close( &ctx.cube, vhashB, vhashB, dataLen );
#if defined(__VAES__)
shavite512_4way_init( &ctx.shavite );
shavite512_4way_update_close( &ctx.shavite, vhashA, vhashA, dataLen );
shavite512_4way_init( &ctx.shavite );
shavite512_4way_update_close( &ctx.shavite, vhashB, vhashB, dataLen );
#else
dintrlv_4x128( hash0, hash1, hash2, hash3, vhashA, dataLen<<3 );
dintrlv_4x128( hash4, hash5, hash6, hash7, vhashB, dataLen<<3 );
@@ -372,11 +449,27 @@ void xevan_8way_hash( void *output, const void *input )
intrlv_4x128( vhashA, hash0, hash1, hash2, hash3, dataLen<<3 );
intrlv_4x128( vhashB, hash4, hash5, hash6, hash7, dataLen<<3 );
#endif
simd_4way_init( &ctx.simd, 512 );
simd_4way_update_close( &ctx.simd, vhashA, vhashA, dataLen<<3 );
simd_4way_init( &ctx.simd, 512 );
simd_4way_update_close( &ctx.simd, vhashB, vhashB, dataLen<<3 );
/*
#if defined(__VAES__)
echo_4way_init( &ctx.echo, 512 );
echo_4way_update_close( &ctx.echo, vhashA, vhashA, dataLen<<3 );
echo_4way_init( &ctx.echo, 512 );
echo_4way_update_close( &ctx.echo, vhashB, vhashB, dataLen<<3 );
rintrlv_4x128_8x64( vhash, vhashA, vhashB, dataLen<<3 );
#else
*/
dintrlv_4x128( hash0, hash1, hash2, hash3, vhashA, dataLen<<3 );
dintrlv_4x128( hash4, hash5, hash6, hash7, vhashB, dataLen<<3 );
@@ -408,6 +501,8 @@ void xevan_8way_hash( void *output, const void *input )
intrlv_8x64( vhash, hash0, hash1, hash2, hash3, hash4, hash5, hash6,
hash7, dataLen<<3 );
//#endif
hamsi512_8way_init( &ctx.hamsi );
hamsi512_8way_update( &ctx.hamsi, vhash, dataLen );
hamsi512_8way_close( &ctx.hamsi, vhash );
@@ -569,12 +664,12 @@ void xevan_4way_hash( void *output, const void *input )
// parallel 4 way
blake512_4way_init( &ctx.blake );
blake512_4way( &ctx.blake, input, 80 );
blake512_4way_update( &ctx.blake, input, 80 );
blake512_4way_close(&ctx.blake, vhash);
memset( &vhash[8<<2], 0, 64<<2 );
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, dataLen );
bmw512_4way_update( &ctx.bmw, vhash, dataLen );
bmw512_4way_close( &ctx.bmw, vhash );
// Serial
@@ -597,15 +692,15 @@ void xevan_4way_hash( void *output, const void *input )
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, dataLen );
skein512_4way_update( &ctx.skein, vhash, dataLen );
skein512_4way_close( &ctx.skein, vhash );
jh512_4way_init( &ctx.jh );
jh512_4way( &ctx.jh, vhash, dataLen );
jh512_4way_update( &ctx.jh, vhash, dataLen );
jh512_4way_close( &ctx.jh, vhash );
keccak512_4way_init( &ctx.keccak );
keccak512_4way( &ctx.keccak, vhash, dataLen );
keccak512_4way_update( &ctx.keccak, vhash, dataLen );
keccak512_4way_close( &ctx.keccak, vhash );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, dataLen<<3 );
@@ -649,7 +744,7 @@ void xevan_4way_hash( void *output, const void *input )
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, dataLen );
hamsi512_4way_update( &ctx.hamsi, vhash, dataLen );
hamsi512_4way_close( &ctx.hamsi, vhash );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
@@ -671,7 +766,7 @@ void xevan_4way_hash( void *output, const void *input )
intrlv_4x32( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
shabal512_4way_init( &ctx.shabal );
shabal512_4way( &ctx.shabal, vhash, dataLen );
shabal512_4way_update( &ctx.shabal, vhash, dataLen );
shabal512_4way_close( &ctx.shabal, vhash );
dintrlv_4x32( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
@@ -693,13 +788,13 @@ void xevan_4way_hash( void *output, const void *input )
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, dataLen );
sha512_4way_update( &ctx.sha512, vhash, dataLen );
sha512_4way_close( &ctx.sha512, vhash );
rintrlv_4x64_4x32( vhashA, vhash, dataLen<<3 );
haval256_5_4way_init( &ctx.haval );
haval256_5_4way( &ctx.haval, vhashA, dataLen );
haval256_5_4way_update( &ctx.haval, vhashA, dataLen );
haval256_5_4way_close( &ctx.haval, vhashA );
rintrlv_4x32_4x64( vhash, vhashA, dataLen<<3 );
@@ -707,11 +802,11 @@ void xevan_4way_hash( void *output, const void *input )
memset( &vhash[ 4<<2 ], 0, (dataLen-32) << 2 );
blake512_4way_init( &ctx.blake );
blake512_4way( &ctx.blake, vhash, dataLen );
blake512_4way_update( &ctx.blake, vhash, dataLen );
blake512_4way_close(&ctx.blake, vhash);
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, dataLen );
bmw512_4way_update( &ctx.bmw, vhash, dataLen );
bmw512_4way_close( &ctx.bmw, vhash );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
@@ -732,15 +827,15 @@ void xevan_4way_hash( void *output, const void *input )
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, dataLen );
skein512_4way_update( &ctx.skein, vhash, dataLen );
skein512_4way_close( &ctx.skein, vhash );
jh512_4way_init( &ctx.jh );
jh512_4way( &ctx.jh, vhash, dataLen );
jh512_4way_update( &ctx.jh, vhash, dataLen );
jh512_4way_close( &ctx.jh, vhash );
keccak512_4way_init( &ctx.keccak );
keccak512_4way( &ctx.keccak, vhash, dataLen );
keccak512_4way_update( &ctx.keccak, vhash, dataLen );
keccak512_4way_close( &ctx.keccak, vhash );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, dataLen<<3 );
@@ -784,7 +879,7 @@ void xevan_4way_hash( void *output, const void *input )
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, dataLen );
hamsi512_4way_update( &ctx.hamsi, vhash, dataLen );
hamsi512_4way_close( &ctx.hamsi, vhash );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
@@ -805,7 +900,7 @@ void xevan_4way_hash( void *output, const void *input )
intrlv_4x32( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
shabal512_4way_init( &ctx.shabal );
shabal512_4way( &ctx.shabal, vhash, dataLen );
shabal512_4way_update( &ctx.shabal, vhash, dataLen );
shabal512_4way_close( &ctx.shabal, vhash );
dintrlv_4x32( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
@@ -826,13 +921,13 @@ void xevan_4way_hash( void *output, const void *input )
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, dataLen );
sha512_4way_update( &ctx.sha512, vhash, dataLen );
sha512_4way_close( &ctx.sha512, vhash );
rintrlv_4x64_4x32( vhashA, vhash, dataLen<<3 );
haval256_5_4way_init( &ctx.haval );
haval256_5_4way( &ctx.haval, vhashA, dataLen );
haval256_5_4way_update( &ctx.haval, vhashA, dataLen );
haval256_5_4way_close( &ctx.haval, output );
}

2
algo/x17/xevan-gate.c
View File

@@ -13,7 +13,7 @@ bool register_xevan_algo( algo_gate_t* gate )
gate->scanhash = (void*)&scanhash_xevan;
gate->hash = (void*)&xevan_hash;
#endif
gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT;
gate->optimizations = SSE2_OPT | AES_OPT | AVX2_OPT | AVX512_OPT | VAES_OPT;
opt_target_factor = 256.0;
return true;
};