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

v3.9.5.4

Browse Source
This commit is contained in:
Jay D Dee
2019-07-15 17:00:26 -04:00
parent e625ed5420
commit e2d5762ef2
63 changed files with 1973 additions and 2980 deletions
Download Patch File Download Diff File Expand all files Collapse all files

123
algo/x17/sonoa-4way.c
View File

@@ -69,7 +69,7 @@ void sonoa_4way_hash( void *state, const void *input )
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -80,7 +80,7 @@ void sonoa_4way_hash( void *state, const void *input )
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, 64 );
@@ -94,7 +94,7 @@ void sonoa_4way_hash( void *state, const void *input )
keccak512_4way( &ctx.keccak, vhash, 64 );
keccak512_4way_close( &ctx.keccak, vhash );
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, 64 );
@@ -116,8 +116,8 @@ void sonoa_4way_hash( void *state, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, 512 );
mm256_dintrlv_2x128( hash0, hash1, vhashA, 512 );
mm256_dintrlv_2x128( hash2, hash3, vhashB, 512 );
dintrlv_2x128_512( hash0, hash1, vhashA );
dintrlv_2x128_512( hash2, hash3, vhashB );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -134,13 +134,13 @@ void sonoa_4way_hash( void *state, const void *input )
// 2
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -151,7 +151,7 @@ void sonoa_4way_hash( void *state, const void *input )
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, 64 );
@@ -165,7 +165,7 @@ void sonoa_4way_hash( void *state, const void *input )
keccak512_4way( &ctx.keccak, vhash, 64 );
keccak512_4way_close( &ctx.keccak, vhash );
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, 64 );
@@ -187,8 +187,8 @@ void sonoa_4way_hash( void *state, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, 512 );
mm256_dintrlv_2x128( hash0, hash1, vhashA, 512 );
mm256_dintrlv_2x128( hash2, hash3, vhashB, 512 );
dintrlv_2x128_512( hash0, hash1, vhashA );
dintrlv_2x128_512( hash2, hash3, vhashB );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -203,7 +203,7 @@ void sonoa_4way_hash( void *state, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, 64 );
@@ -215,7 +215,7 @@ void sonoa_4way_hash( void *state, const void *input )
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -226,7 +226,7 @@ void sonoa_4way_hash( void *state, const void *input )
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, 64 );
@@ -240,7 +240,7 @@ void sonoa_4way_hash( void *state, const void *input )
keccak512_4way( &ctx.keccak, vhash, 64 );
keccak512_4way_close( &ctx.keccak, vhash );
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, 64 );
@@ -262,8 +262,8 @@ void sonoa_4way_hash( void *state, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, 512 );
mm256_dintrlv_2x128( hash0, hash1, vhashA, 512 );
mm256_dintrlv_2x128( hash2, hash3, vhashB, 512 );
dintrlv_2x128_512( hash0, hash1, vhashA );
dintrlv_2x128_512( hash2, hash3, vhashB );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -278,13 +278,13 @@ void sonoa_4way_hash( void *state, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, 64 );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
sph_fugue512_init( &ctx.fugue );
sph_fugue512( &ctx.fugue, hash0, 64 );
@@ -300,13 +300,13 @@ void sonoa_4way_hash( void *state, const void *input )
sph_fugue512_close( &ctx.fugue, hash3 );
// 4
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -317,7 +317,7 @@ void sonoa_4way_hash( void *state, const void *input )
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, 64 );
@@ -331,7 +331,7 @@ void sonoa_4way_hash( void *state, const void *input )
keccak512_4way( &ctx.keccak, vhash, 64 );
keccak512_4way_close( &ctx.keccak, vhash );
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, 64 );
@@ -353,8 +353,8 @@ void sonoa_4way_hash( void *state, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, 512 );
mm256_dintrlv_2x128( hash0, hash1, vhashA, 512 );
mm256_dintrlv_2x128( hash2, hash3, vhashB, 512 );
dintrlv_2x128_512( hash0, hash1, vhashA );
dintrlv_2x128_512( hash2, hash3, vhashB );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -369,13 +369,13 @@ void sonoa_4way_hash( void *state, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, 64 );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
sph_fugue512_init( &ctx.fugue );
sph_fugue512( &ctx.fugue, hash0, 64 );
@@ -396,13 +396,13 @@ void sonoa_4way_hash( void *state, const void *input )
shabal512_4way( &ctx.shabal, vhash, 64 );
shabal512_4way_close( &ctx.shabal, vhash );
mm256_rintrlv_4x32_4x64( vhashB, vhash, 512 );
rintrlv_4x32_4x64( vhashB, vhash, 512 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhashB, 64 );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -417,8 +417,8 @@ void sonoa_4way_hash( void *state, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, 512 );
mm256_intrlv_2x128( vhashA, hash0, hash1, 512 );
mm256_intrlv_2x128( vhashB, hash2, hash3, 512 );
intrlv_2x128_512( vhashA, hash0, hash1 );
intrlv_2x128_512( vhashB, hash2, hash3 );
shavite512_2way_init( &ctx.shavite );
shavite512_2way_update_close( &ctx.shavite, vhashA, vhashA, 64 );
@@ -426,13 +426,13 @@ void sonoa_4way_hash( void *state, const void *input )
shavite512_2way_update_close( &ctx.shavite, vhashB, vhashB, 64 );
// 5
mm256_rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
rintrlv_2x128_4x64( vhash, vhashA, vhashB, 512 );
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
mm256_rintrlv_4x64_4x32( vhashB, vhash, 512 );
rintrlv_4x64_4x32( vhashB, vhash, 512 );
shabal512_4way_init( &ctx.shabal );
shabal512_4way( &ctx.shabal, vhashB, 64 );
@@ -449,7 +449,7 @@ void sonoa_4way_hash( void *state, const void *input )
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, 64 );
@@ -463,7 +463,7 @@ void sonoa_4way_hash( void *state, const void *input )
keccak512_4way( &ctx.keccak, vhash, 64 );
keccak512_4way_close( &ctx.keccak, vhash );
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, 64 );
@@ -485,8 +485,8 @@ void sonoa_4way_hash( void *state, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, 512 );
mm256_dintrlv_2x128( hash0, hash1, vhashA, 512 );
mm256_dintrlv_2x128( hash2, hash3, vhashB, 512 );
dintrlv_2x128_512( hash0, hash1, vhashA );
dintrlv_2x128_512( hash2, hash3, vhashB );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -501,13 +501,13 @@ void sonoa_4way_hash( void *state, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, 64 );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
sph_fugue512_init( &ctx.fugue );
sph_fugue512( &ctx.fugue, hash0, 64 );
@@ -545,13 +545,13 @@ void sonoa_4way_hash( void *state, const void *input )
// 6
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -562,7 +562,7 @@ void sonoa_4way_hash( void *state, const void *input )
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, 64 );
@@ -576,7 +576,7 @@ void sonoa_4way_hash( void *state, const void *input )
keccak512_4way( &ctx.keccak, vhash, 64 );
keccak512_4way_close( &ctx.keccak, vhash );
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, 64 );
@@ -598,8 +598,8 @@ void sonoa_4way_hash( void *state, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, 512 );
mm256_dintrlv_2x128( hash0, hash1, vhashA, 512 );
mm256_dintrlv_2x128( hash2, hash3, vhashB, 512 );
dintrlv_2x128_512( hash0, hash1, vhashA );
dintrlv_2x128_512( hash2, hash3, vhashB );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -614,13 +614,13 @@ void sonoa_4way_hash( void *state, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, 64 );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
sph_fugue512_init( &ctx.fugue );
sph_fugue512( &ctx.fugue, hash0, 64 );
@@ -656,13 +656,13 @@ void sonoa_4way_hash( void *state, const void *input )
sph_whirlpool( &ctx.whirlpool, hash3, 64 );
sph_whirlpool_close( &ctx.whirlpool, hash3 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, 64 );
sha512_4way_close( &ctx.sha512, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
sph_whirlpool_init( &ctx.whirlpool );
sph_whirlpool( &ctx.whirlpool, hash0, 64 );
@@ -679,13 +679,13 @@ void sonoa_4way_hash( void *state, const void *input )
// 7
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
bmw512_4way_init( &ctx.bmw );
bmw512_4way( &ctx.bmw, vhash, 64 );
bmw512_4way_close( &ctx.bmw, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0, 512 );
@@ -696,7 +696,7 @@ void sonoa_4way_hash( void *state, const void *input )
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, 64 );
@@ -710,7 +710,7 @@ void sonoa_4way_hash( void *state, const void *input )
keccak512_4way( &ctx.keccak, vhash, 64 );
keccak512_4way_close( &ctx.keccak, vhash );
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, 64 );
@@ -732,8 +732,8 @@ void sonoa_4way_hash( void *state, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, 512 );
mm256_dintrlv_2x128( hash0, hash1, vhashA, 512 );
mm256_dintrlv_2x128( hash2, hash3, vhashB, 512 );
dintrlv_2x128_512( hash0, hash1, vhashA );
dintrlv_2x128_512( hash2, hash3, vhashB );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -748,13 +748,13 @@ void sonoa_4way_hash( void *state, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, 512 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, 64 );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
sph_fugue512_init( &ctx.fugue );
sph_fugue512( &ctx.fugue, hash0, 64 );
@@ -790,13 +790,13 @@ void sonoa_4way_hash( void *state, const void *input )
sph_whirlpool( &ctx.whirlpool, hash3, 64 );
sph_whirlpool_close( &ctx.whirlpool, hash3 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, 64 );
sha512_4way_close( &ctx.sha512, vhash );
mm256_rintrlv_4x64_4x32( vhashB, vhash, 512 );
rintrlv_4x64_4x32( vhashB, vhash, 512 );
haval256_5_4way_init( &ctx.haval );
haval256_5_4way( &ctx.haval, vhashB, 64 );
@@ -808,7 +808,6 @@ int scanhash_sonoa_4way( struct work *work, uint32_t max_nonce,
{
uint32_t hash[4*8] __attribute__ ((aligned (64)));
uint32_t vdata[24*4] __attribute__ ((aligned (64)));
uint32_t edata[20] __attribute__ ((aligned (64)));
uint32_t lane_hash[8] __attribute__ ((aligned (32)));
uint32_t *hash7 = &(hash[7<<2]);
uint32_t *pdata = work->data;
@@ -824,9 +823,7 @@ int scanhash_sonoa_4way( struct work *work, uint32_t max_nonce,
0xFFFFF000, 0xFFFF0000, 0 };
// Need big endian data
swab32_array( edata, pdata, 20 );
mm256_intrlv_4x64( vdata, edata, edata, edata, edata, 640 );
// mm256_bswap_intrlv80_4x64( vdata, pdata );
mm256_bswap32_intrlv80_4x64( vdata, pdata );
for ( int m=0; m < 6; m++ ) if ( Htarg <= htmax[m] )
{
uint32_t mask = masks[m];

25
algo/x17/x17-4way.c
View File

@@ -68,7 +68,7 @@ void x17_4way_hash( void *state, const void *input )
bmw512_4way_close( &ctx.bmw, vhash );
// Serialize
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
// 3 Groestl
init_groestl( &ctx.groestl, 64 );
@@ -81,7 +81,7 @@ void x17_4way_hash( void *state, const void *input )
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3, 512 );
// Parallellize
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
// 4 Skein parallel 4 way 64 bit
skein512_4way_init( &ctx.skein );
@@ -99,7 +99,7 @@ void x17_4way_hash( void *state, const void *input )
keccak512_4way_close( &ctx.keccak, vhash );
// 7 Luffa parallel 2 way 128 bit
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, 512 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, 64 );
@@ -124,8 +124,8 @@ void x17_4way_hash( void *state, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, 512 );
mm256_dintrlv_2x128_512( hash0, hash1, vhashA );
mm256_dintrlv_2x128_512( hash2, hash3, vhashB );
dintrlv_2x128_512( hash0, hash1, vhashA );
dintrlv_2x128_512( hash2, hash3, vhashB );
// 11 Echo serial
init_echo( &ctx.echo, 512 );
@@ -142,13 +142,13 @@ void x17_4way_hash( void *state, const void *input )
(const BitSequence *) hash3, 512 );
// 12 Hamsi parallel 4 way 64 bit
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, 64 );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, 512 );
// 13 Fugue serial
sph_fugue512_init( &ctx.fugue );
@@ -188,14 +188,14 @@ void x17_4way_hash( void *state, const void *input )
sph_whirlpool_close( &ctx.whirlpool, hash3 );
// 16 SHA512 parallel 64 bit
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, 512 );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, 64 );
sha512_4way_close( &ctx.sha512, vhash );
// 17 Haval parallel 32 bit
mm256_rintrlv_4x64_4x32( vhashB, vhash, 512 );
rintrlv_4x64_4x32( vhashB, vhash, 512 );
haval256_5_4way_init( &ctx.haval );
haval256_5_4way( &ctx.haval, vhashB, 64 );
@@ -207,7 +207,6 @@ int scanhash_x17_4way( struct work *work, uint32_t max_nonce,
{
uint32_t hash[4*8] __attribute__ ((aligned (64)));
uint32_t vdata[24*4] __attribute__ ((aligned (64)));
uint32_t edata[20] __attribute__ ((aligned (64)));
uint32_t lane_hash[8] __attribute__ ((aligned (32)));
uint32_t *hash7 = &(hash[7<<2]);
uint32_t *pdata = work->data;
@@ -223,9 +222,7 @@ int scanhash_x17_4way( struct work *work, uint32_t max_nonce,
0xFFFFF000, 0xFFFF0000, 0 };
// Need big endian data
swab32_array( edata, pdata, 20 );
mm256_intrlv_4x64( vdata, edata, edata, edata, edata, 640 );
// mm256_bswap_intrlv80_4x64( vdata, pdata );
mm256_bswap32_intrlv80_4x64( vdata, pdata );
for ( int m = 0; m < 6; m++ ) if ( Htarg <= htmax[m] )
{
uint32_t mask = masks[ m ];
@@ -235,7 +232,7 @@ int scanhash_x17_4way( struct work *work, uint32_t max_nonce,
_mm256_set_epi32( n+3, 0, n+2, 0, n+1, 0, n, 0 ) ), *noncev );
x17_4way_hash( hash, vdata );
for ( int lane = 0; lane < 4; lane++ )
for ( int lane = 0; lane < 4; lane++ )
if ( ( hash7[ lane ] & mask ) == 0 )
{
extr_lane_4x32( lane_hash, hash, lane, 256 );

43
algo/x17/xevan-4way.c
View File

@@ -71,7 +71,7 @@ void xevan_4way_hash( void *output, const void *input )
bmw512_4way_close( &ctx.bmw, vhash );
// Serial
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0,
@@ -87,7 +87,7 @@ void xevan_4way_hash( void *output, const void *input )
dataLen<<3 );
// Parallel 4way
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, dataLen );
@@ -101,7 +101,7 @@ void xevan_4way_hash( void *output, const void *input )
keccak512_4way( &ctx.keccak, vhash, dataLen );
keccak512_4way_close( &ctx.keccak, vhash );
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, dataLen<<3 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, dataLen<<3 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, dataLen );
@@ -123,8 +123,8 @@ void xevan_4way_hash( void *output, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, dataLen<<3 );
mm256_dintrlv_2x128( hash0, hash1, vhashA, dataLen<<3 );
mm256_dintrlv_2x128( hash2, hash3, vhashB, dataLen<<3 );
dintrlv_2x128( hash0, hash1, vhashA, dataLen<<3 );
dintrlv_2x128( hash2, hash3, vhashB, dataLen<<3 );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -139,13 +139,13 @@ void xevan_4way_hash( void *output, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, dataLen<<3 );
// Parallel
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, dataLen );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
sph_fugue512_init( &ctx.fugue );
sph_fugue512( &ctx.fugue, hash0, dataLen );
@@ -183,19 +183,19 @@ void xevan_4way_hash( void *output, const void *input )
sph_whirlpool( &ctx.whirlpool, hash3, dataLen );
sph_whirlpool_close( &ctx.whirlpool, hash3 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, dataLen );
sha512_4way_close( &ctx.sha512, vhash );
mm256_rintrlv_4x64_4x32( vhashA, vhash, dataLen<<3 );
rintrlv_4x64_4x32( vhashA, vhash, dataLen<<3 );
haval256_5_4way_init( &ctx.haval );
haval256_5_4way( &ctx.haval, vhashA, dataLen );
haval256_5_4way_close( &ctx.haval, vhashA );
mm256_rintrlv_4x32_4x64( vhash, vhashA, dataLen<<3 );
rintrlv_4x32_4x64( vhash, vhashA, dataLen<<3 );
memset( &vhash[ 4<<2 ], 0, (dataLen-32) << 2 );
@@ -207,7 +207,7 @@ void xevan_4way_hash( void *output, const void *input )
bmw512_4way( &ctx.bmw, vhash, dataLen );
bmw512_4way_close( &ctx.bmw, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
init_groestl( &ctx.groestl, 64 );
update_and_final_groestl( &ctx.groestl, (char*)hash0, (char*)hash0,
@@ -222,7 +222,7 @@ void xevan_4way_hash( void *output, const void *input )
update_and_final_groestl( &ctx.groestl, (char*)hash3, (char*)hash3,
dataLen<<3 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
skein512_4way_init( &ctx.skein );
skein512_4way( &ctx.skein, vhash, dataLen );
@@ -236,7 +236,7 @@ void xevan_4way_hash( void *output, const void *input )
keccak512_4way( &ctx.keccak, vhash, dataLen );
keccak512_4way_close( &ctx.keccak, vhash );
mm256_rintrlv_4x64_2x128( vhashA, vhashB, vhash, dataLen<<3 );
rintrlv_4x64_2x128( vhashA, vhashB, vhash, dataLen<<3 );
luffa_2way_init( &ctx.luffa, 512 );
luffa_2way_update_close( &ctx.luffa, vhashA, vhashA, dataLen );
@@ -258,8 +258,8 @@ void xevan_4way_hash( void *output, const void *input )
simd_2way_init( &ctx.simd, 512 );
simd_2way_update_close( &ctx.simd, vhashB, vhashB, dataLen<<3 );
mm256_dintrlv_2x128( hash0, hash1, vhashA, dataLen<<3 );
mm256_dintrlv_2x128( hash2, hash3, vhashB, dataLen<<3 );
dintrlv_2x128( hash0, hash1, vhashA, dataLen<<3 );
dintrlv_2x128( hash2, hash3, vhashB, dataLen<<3 );
init_echo( &ctx.echo, 512 );
update_final_echo( &ctx.echo, (BitSequence *)hash0,
@@ -274,13 +274,13 @@ void xevan_4way_hash( void *output, const void *input )
update_final_echo( &ctx.echo, (BitSequence *)hash3,
(const BitSequence *) hash3, dataLen<<3 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
hamsi512_4way_init( &ctx.hamsi );
hamsi512_4way( &ctx.hamsi, vhash, dataLen );
hamsi512_4way_close( &ctx.hamsi, vhash );
mm256_dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
dintrlv_4x64( hash0, hash1, hash2, hash3, vhash, dataLen<<3 );
sph_fugue512_init( &ctx.fugue );
sph_fugue512( &ctx.fugue, hash0, dataLen );
@@ -316,13 +316,13 @@ void xevan_4way_hash( void *output, const void *input )
sph_whirlpool( &ctx.whirlpool, hash3, dataLen );
sph_whirlpool_close( &ctx.whirlpool, hash3 );
mm256_intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
intrlv_4x64( vhash, hash0, hash1, hash2, hash3, dataLen<<3 );
sha512_4way_init( &ctx.sha512 );
sha512_4way( &ctx.sha512, vhash, dataLen );
sha512_4way_close( &ctx.sha512, vhash );
mm256_rintrlv_4x64_4x32( vhashA, vhash, dataLen<<3 );
rintrlv_4x64_4x32( vhashA, vhash, dataLen<<3 );
haval256_5_4way_init( &ctx.haval );
haval256_5_4way( &ctx.haval, vhashA, dataLen );
@@ -334,7 +334,6 @@ int scanhash_xevan_4way( struct work *work, uint32_t max_nonce,
{
uint32_t hash[4*8] __attribute__ ((aligned (64)));
uint32_t vdata[24*4] __attribute__ ((aligned (64)));
uint32_t edata[20] __attribute__ ((aligned (64)));
uint32_t lane_hash[8] __attribute__ ((aligned (32)));
uint32_t *hash7 = &(hash[7<<2]);
uint32_t *pdata = work->data;
@@ -349,9 +348,7 @@ int scanhash_xevan_4way( struct work *work, uint32_t max_nonce,
if ( opt_benchmark )
ptarget[7] = 0x0cff;
swab32_array( edata, pdata, 20 );
mm256_intrlv_4x64( vdata, edata, edata, edata, edata, 640 );
// mm256_bswap_intrlv80_4x64( vdata, pdata );
mm256_bswap32_intrlv80_4x64( vdata, pdata );
do {
*noncev = mm256_intrlv_blend_32( mm256_bswap_32(
_mm256_set_epi32( n+3, 0,n+2, 0,n+1, 0, n, 0 ) ), *noncev );