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

v3.8.1.1

Browse Source
This commit is contained in:
Jay D Dee
2018-02-09 23:30:14 -05:00
parent a28daca3ce
commit e4265a6f11
10 changed files with 320 additions and 108 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
RELEASE_NOTES
View File

@@ -159,6 +159,10 @@ Support for even older x86_64 without AES_NI or SSE2 is not availble.
Change Log
----------
v3.8.1.1
Fixed Windows AVX2 crash.
v3.8.1
Fixes x16r on CPUs with only SSE2.

4
algo/bmw/bmw-hash-4way.c
View File

@@ -51,7 +51,9 @@ extern "C"{
// BMW small has a bug not present in big. Lanes 0 & 2 produce valid hash
// while lanes 1 & 3 produce invalid hash. The cause is not known.
// Some things that could cause it are: using epi64 instead of epi32,
// a memory write that is the wrong size, an attempt to index a vector
// like an array (only works for 64 bit elements).
static const sph_u32 IV256[] = {

26
algo/groestl/myr-groestl.c
View File

@@ -20,11 +20,11 @@ typedef struct {
#else
hashState_groestl groestl;
#endif
#ifndef USE_SPH_SHA
SHA256_CTX sha;
#else
//#ifndef USE_SPH_SHA
// SHA256_CTX sha;
//#else
sph_sha256_context sha;
#endif
//#endif
} myrgr_ctx_holder;
myrgr_ctx_holder myrgr_ctx;
@@ -36,11 +36,11 @@ void init_myrgr_ctx()
#else
init_groestl (&myrgr_ctx.groestl, 64 );
#endif
#ifndef USE_SPH_SHA
SHA256_Init( &myrgr_ctx.sha );
#else
//#ifndef USE_SPH_SHA
// SHA256_Init( &myrgr_ctx.sha );
//#else
sph_sha256_init( &myrgr_ctx.sha );
#endif
//#endif
}
void myriadhash( void *output, const void *input )
@@ -57,13 +57,13 @@ void myriadhash( void *output, const void *input )
(const char*)input, 640 );
#endif
#ifndef USE_SPH_SHA
SHA256_Update( &ctx.sha, hash, 64 );
SHA256_Final( (unsigned char*) hash, &ctx.sha );
#else
//#ifndef USE_SPH_SHA
// SHA256_Update( &ctx.sha, hash, 64 );
// SHA256_Final( (unsigned char*) hash, &ctx.sha );
//#else
sph_sha256(&ctx.sha, hash, 64);
sph_sha256_close(&ctx.sha, hash);
#endif
//#endif
memcpy(output, hash, 32);
}

70
algo/luffa/luffa-hash-2way.c
View File

@@ -251,10 +251,11 @@ __m256i CNS[32];
/* Round function */
/* state: hash context */
static void rnd512_2way( luffa_2way_context *state, __m256i msg1, __m256i msg0 )
void rnd512_2way( luffa_2way_context *state, __m256i *msg )
{
__m256i t[2];
__m256i *chainv = state->chainv;
__m256i msg0, msg1;
__m256i tmp[2];
__m256i x[8];
@@ -272,8 +273,8 @@ static void rnd512_2way( luffa_2way_context *state, __m256i msg1, __m256i msg0 )
MULT2( t[0], t[1] );
msg0 = _mm256_shuffle_epi32( msg0, 27 );
msg1 = _mm256_shuffle_epi32( msg1, 27 );
msg0 = _mm256_shuffle_epi32( msg[0], 27 );
msg1 = _mm256_shuffle_epi32( msg[1], 27 );
chainv[0] = _mm256_xor_si256( chainv[0], t[0] );
chainv[1] = _mm256_xor_si256( chainv[1], t[1] );
@@ -359,7 +360,6 @@ static void rnd512_2way( luffa_2way_context *state, __m256i msg1, __m256i msg0 )
chainv[9] = _mm256_or_si256( _mm256_slli_epi32( chainv[9], 4 ),
_mm256_srli_epi32( chainv[9], 28 ) );
NMLTOM1024( chainv[0], chainv[2], chainv[4], chainv[6],
x[0], x[1], x[2], x[3],
chainv[1],chainv[3],chainv[5],chainv[7],
@@ -382,6 +382,7 @@ static void rnd512_2way( luffa_2way_context *state, __m256i msg1, __m256i msg0 )
/* Process last 256-bit block */
STEP_PART2( chainv[8], chainv[9], t[0], t[1], CNS[16], CNS[17],
tmp[0], tmp[1] );
STEP_PART2( chainv[8], chainv[9], t[0], t[1], CNS[18], CNS[19],
tmp[0], tmp[1] );
STEP_PART2( chainv[8], chainv[9], t[0], t[1], CNS[20], CNS[21],
@@ -404,14 +405,16 @@ static void rnd512_2way( luffa_2way_context *state, __m256i msg1, __m256i msg0 )
/* state: hash context */
/* b[8]: hash values */
static void finalization512_2way( luffa_2way_context *state, uint32 *b )
void finalization512_2way( luffa_2way_context *state, uint32 *b )
{
uint32 hash[8] __attribute((aligned(64)));
__m256i* chainv = state->chainv;
__m256i t[2];
__m256i zero[2];
zero[0] = zero[1] = _mm256_setzero_si256();
/*---- blank round with m=0 ----*/
rnd512_2way( state, m256_zero, m256_zero );
rnd512_2way( state, zero );
t[0] = chainv[0];
t[1] = chainv[1];
@@ -434,7 +437,7 @@ static void finalization512_2way( luffa_2way_context *state, uint32 *b )
casti_m256i( b, 0 ) = mm256_bswap_32( casti_m256i( hash, 0 ) );
casti_m256i( b, 1 ) = mm256_bswap_32( casti_m256i( hash, 1 ) );
rnd512_2way( state, m256_zero, m256_zero );
rnd512_2way( state, zero );
t[0] = chainv[0];
t[1] = chainv[1];
@@ -487,6 +490,7 @@ int luffa_2way_update( luffa_2way_context *state, const void *data,
{
__m256i *vdata = (__m256i*)data;
__m256i *buffer = (__m256i*)state->buffer;
__m256i msg[2];
int i;
int blocks = (int)len / 32;
state-> rembytes = (int)len % 32;
@@ -494,8 +498,9 @@ int luffa_2way_update( luffa_2way_context *state, const void *data,
// full blocks
for ( i = 0; i < blocks; i++, vdata+=2 )
{
rnd512_2way( state, mm256_bswap_32( vdata[1] ) ,
mm256_bswap_32( vdata[0] ) );
msg[0] = mm256_bswap_32( vdata[ i ] );
msg[1] = mm256_bswap_32( vdata[ i+1 ] );
rnd512_2way( state, msg );
}
// 16 byte partial block exists for 80 byte len
@@ -513,17 +518,19 @@ int luffa_2way_update( luffa_2way_context *state, const void *data,
int luffa_2way_close( luffa_2way_context *state, void *hashval )
{
__m256i *buffer = (__m256i*)state->buffer;
__m256i msg[2];
// transform pad block
if ( state->rembytes )
// not empty, data is in buffer
rnd512_2way( state, buffer[1], buffer[0] );
rnd512_2way( state, buffer );
else
// empty pad block, constant data
rnd512_2way( state, m256_zero,
_mm256_set_epi8( 0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0,
0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0 ) );
{ // empty pad block, constant data
msg[0] = _mm256_set_epi8( 0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0,
0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0 );
msg[1] = m256_zero;
rnd512_2way( state, msg );
}
finalization512_2way( state, (uint32*)hashval );
if ( state->hashbitlen > 512 )
@@ -535,28 +542,37 @@ int luffa_2way_update_close( luffa_2way_context *state,
void *output, const void *data, size_t inlen )
{
// Optimized for integrals of 16 bytes, good for 64 and 80 byte len
__m256i *vdata = (__m256i*)data;
const __m256i *vdata = (__m256i*)data;
__m256i msg[2];
int i;
int blocks = (int)( inlen / 32 );
state->rembytes = inlen % 32;
const int blocks = (int)( inlen >> 5 );
state->rembytes = inlen & 0x1F;
// full blocks
for ( i = 0; i < blocks; i++, vdata+=2 )
rnd512_2way( state, mm256_bswap_32( vdata[1] ),
mm256_bswap_32( vdata[0] ) );
{
msg[0] = mm256_bswap_32( vdata[ 0 ] );
msg[1] = mm256_bswap_32( vdata[ 1 ] );
rnd512_2way( state, msg );
}
// 16 byte partial block exists for 80 byte len
if ( state->rembytes )
{
// padding of partial block
rnd512_2way( state,
_mm256_set_epi8( 0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0,
0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0 ),
mm256_bswap_32( vdata[0] ) );
msg[0] = mm256_bswap_32( vdata[0] );
msg[1] = _mm256_set_epi8( 0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0,
0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0 );
rnd512_2way( state, msg );
}
else
{
// empty pad block
rnd512_2way( state, m256_zero,
_mm256_set_epi8( 0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0,
0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0 ) );
msg[0] = _mm256_set_epi8( 0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0,
0,0,0,0, 0,0,0,0, 0,0,0,0, 0x80,0,0,0 );
msg[1] = m256_zero;
rnd512_2way( state, msg );
}
finalization512_2way( state, (uint32*)output );
if ( state->hashbitlen > 512 )

2
algo/qubit/deep-2way.c
View File

@@ -116,7 +116,7 @@ int scanhash_deep_2way( int thr_id, struct work *work,uint32_t max_nonce,
found[1] = true;
num_found++;
nonces[1] = n+1;
work_set_target_ratio( work, hash+64 );
work_set_target_ratio( work, hash+8 );
}
n += 2;
} while ( ( num_found == 0 ) && ( n < max_nonce )

238
algo/sha/sha2-hash-4way.c
View File

@@ -30,13 +30,235 @@
* @author Thomas Pornin <thomas.pornin@cryptolog.com>
*/
#if defined(__AVX__)
#include <stddef.h>
#include <string.h>
#include "sha2-hash-4way.h"
// SHA256 4 way 32 bit
static const sph_u32 H256[8] = {
SPH_C32(0x6A09E667), SPH_C32(0xBB67AE85),
SPH_C32(0x3C6EF372), SPH_C32(0xA54FF53A),
SPH_C32(0x510E527F), SPH_C32(0x9B05688C),
SPH_C32(0x1F83D9AB), SPH_C32(0x5BE0CD19)
};
static const sph_u32 K256[80] = {
SPH_C32(0x428A2F98), SPH_C32(0x71374491),
SPH_C32(0xB5C0FBCF), SPH_C32(0xE9B5DBA5),
SPH_C32(0x3956C25B), SPH_C32(0x59F111F1),
SPH_C32(0x923F82A4), SPH_C32(0xAB1C5ED5),
SPH_C32(0xD807AA98), SPH_C32(0x12835B01),
SPH_C32(0x243185BE), SPH_C32(0x550C7DC3),
SPH_C32(0x72BE5D74), SPH_C32(0x80DEB1FE),
SPH_C32(0x9BDC06A7), SPH_C32(0xC19BF174),
SPH_C32(0xE49B69C1), SPH_C32(0xEFBE4786),
SPH_C32(0x0FC19DC6), SPH_C32(0x240CA1CC),
SPH_C32(0x2DE92C6F), SPH_C32(0x4A7484AA),
SPH_C32(0x5CB0A9DC), SPH_C32(0x76F988DA),
SPH_C32(0x983E5152), SPH_C32(0xA831C66D),
SPH_C32(0xB00327C8), SPH_C32(0xBF597FC7),
SPH_C32(0xC6E00BF3), SPH_C32(0xD5A79147),
SPH_C32(0x06CA6351), SPH_C32(0x14292967),
SPH_C32(0x27B70A85), SPH_C32(0x2E1B2138),
SPH_C32(0x4D2C6DFC), SPH_C32(0x53380D13),
SPH_C32(0x650A7354), SPH_C32(0x766A0ABB),
SPH_C32(0x81C2C92E), SPH_C32(0x92722C85),
SPH_C32(0xA2BFE8A1), SPH_C32(0xA81A664B),
SPH_C32(0xC24B8B70), SPH_C32(0xC76C51A3),
SPH_C32(0xD192E819), SPH_C32(0xD6990624),
SPH_C32(0xF40E3585), SPH_C32(0x106AA070),
SPH_C32(0x19A4C116), SPH_C32(0x1E376C08),
SPH_C32(0x2748774C), SPH_C32(0x34B0BCB5),
SPH_C32(0x391C0CB3), SPH_C32(0x4ED8AA4A),
SPH_C32(0x5B9CCA4F), SPH_C32(0x682E6FF3),
SPH_C32(0x748F82EE), SPH_C32(0x78A5636F),
SPH_C32(0x84C87814), SPH_C32(0x8CC70208),
SPH_C32(0x90BEFFFA), SPH_C32(0xA4506CEB),
SPH_C32(0xBEF9A3F7), SPH_C32(0xC67178F2),
SPH_C32(0xCA273ECE), SPH_C32(0xD186B8C7),
SPH_C32(0xEADA7DD6), SPH_C32(0xF57D4F7F),
SPH_C32(0x06F067AA), SPH_C32(0x0A637DC5),
SPH_C32(0x113F9804), SPH_C32(0x1B710B35),
SPH_C32(0x28DB77F5), SPH_C32(0x32CAAB7B),
SPH_C32(0x3C9EBE0A), SPH_C32(0x431D67C4),
SPH_C32(0x4CC5D4BE), SPH_C32(0x597F299C),
SPH_C32(0x5FCB6FAB), SPH_C32(0x6C44198C)
};
#define CHs(X, Y, Z) \
_mm_xor_si128( _mm_and_si128( _mm_xor_si128( Y, Z ), X ), Z )
#define MAJs(X, Y, Z) \
_mm_or_si128( _mm_and_si128( X, Y ), \
_mm_and_si128( _mm_or_si128( X, Y ), Z ) )
#define BSG2_0(x) \
_mm_xor_si128( _mm_xor_si128( \
mm_rotr_32(x, 2), mm_rotr_32(x, 13) ), mm_rotr_32( x, 22) )
#define BSG2_1(x) \
_mm_xor_si128( _mm_xor_si128( \
mm_rotr_32(x, 6), mm_rotr_32(x, 11) ), mm_rotr_32( x, 25) )
#define SSG2_0(x) \
_mm_xor_si128( _mm_xor_si128( \
mm_rotr_32(x, 7), mm_rotr_32(x, 18) ), _mm_srli_epi32(x, 3) )
#define SSG2_1(x) \
_mm_xor_si128( _mm_xor_si128( \
mm_rotr_32(x, 17), mm_rotr_32(x, 19) ), _mm_srli_epi32(x, 10) )
#define SHA256_4WAY_STEP(A, B, C, D, E, F, G, H, i) \
do { \
__m128i T1, T2; \
T1 = _mm_add_epi32( _mm_add_epi32( _mm_add_epi32( \
_mm_add_epi32( H, BSG2_1(E) ), CHs(E, F, G) ), \
_mm_set1_epi32( K256[i] ) ), W[i] ); \
T2 = _mm_add_epi32( BSG2_0(A), MAJs(A, B, C) ); \
D = _mm_add_epi32( D, T1 ); \
H = _mm_add_epi32( T1, T2 ); \
} while (0)
static void
sha256_4way_round( __m128i *in, __m128i r[8] )
{
int i;
__m128i A, B, C, D, E, F, G, H;
__m128i W[80];
for ( i = 0; i < 16; i++ )
W[i] = mm_bswap_32( in[i] );
for ( i = 16; i < 80; i++ )
W[i] = _mm_add_epi32( _mm_add_epi32( _mm_add_epi32(
SSG2_1( W[ i-2 ] ), W[ i-7 ] ), SSG2_0( W[ i-15 ] ) ), W[ i-16 ] );
A = r[0];
B = r[1];
C = r[2];
D = r[3];
E = r[4];
F = r[5];
G = r[6];
H = r[7];
for ( i = 0; i < 80; i += 8 )
{
SHA256_4WAY_STEP( A, B, C, D, E, F, G, H, i + 0 );
SHA256_4WAY_STEP( H, A, B, C, D, E, F, G, i + 1 );
SHA256_4WAY_STEP( G, H, A, B, C, D, E, F, i + 2 );
SHA256_4WAY_STEP( F, G, H, A, B, C, D, E, i + 3 );
SHA256_4WAY_STEP( E, F, G, H, A, B, C, D, i + 4 );
SHA256_4WAY_STEP( D, E, F, G, H, A, B, C, i + 5 );
SHA256_4WAY_STEP( C, D, E, F, G, H, A, B, i + 6 );
SHA256_4WAY_STEP( B, C, D, E, F, G, H, A, i + 7 );
}
r[0] = _mm_add_epi32( r[0], A );
r[1] = _mm_add_epi32( r[1], B );
r[2] = _mm_add_epi32( r[2], C );
r[3] = _mm_add_epi32( r[3], D );
r[4] = _mm_add_epi32( r[4], E );
r[5] = _mm_add_epi32( r[5], F );
r[6] = _mm_add_epi32( r[6], G );
r[7] = _mm_add_epi32( r[7], H );
}
void sha256_4way_init( sha256_4way_context *sc )
{
sc->count_high = sc->count_low = 0;
sc->val[0] = _mm_set1_epi32( H256[0] );
sc->val[1] = _mm_set1_epi32( H256[1] );
sc->val[2] = _mm_set1_epi32( H256[2] );
sc->val[3] = _mm_set1_epi32( H256[3] );
sc->val[4] = _mm_set1_epi32( H256[4] );
sc->val[5] = _mm_set1_epi32( H256[5] );
sc->val[6] = _mm_set1_epi32( H256[6] );
sc->val[7] = _mm_set1_epi32( H256[7] );
}
void sha256_4way( sha256_4way_context *sc, const void *data, size_t len )
{
__m128i *vdata = (__m128i*)data;
size_t ptr;
const int buf_size = 64;
ptr = (unsigned)sc->count_low & (buf_size - 1U);
while ( len > 0 )
{
size_t clen;
uint32_t clow, clow2;
clen = buf_size - ptr;
if ( clen > len )
clen = len;
memcpy_128( sc->buf + (ptr>>2), vdata, clen>>2 );
vdata = vdata + (clen>>2);
ptr += clen;
len -= clen;
if ( ptr == buf_size )
{
sha256_4way_round( sc->buf, sc->val );
ptr = 0;
}
clow = sc->count_low;
clow2 = SPH_T32( clow + clen );
sc->count_low = clow2;
if ( clow2 < clow )
sc->count_high++;
}
}
void sha256_4way_close( sha256_4way_context *sc, void *dst )
{
unsigned ptr, u;
uint32_t low, high;
const int buf_size = 64;
const int pad = buf_size - 8;
ptr = (unsigned)sc->count_low & (buf_size - 1U);
sc->buf[ ptr>>2 ] = _mm_set1_epi32( 0x80 );
ptr += 4;
if ( ptr > pad )
{
memset_zero_128( sc->buf + (ptr>>2), (buf_size - ptr) >> 2 );
sha256_4way_round( sc->buf, sc->val );
memset_zero_128( sc->buf, pad >> 2 );
}
else
memset_zero_128( sc->buf + (ptr>>2), (pad - ptr) >> 2 );
low = sc->count_low;
high = (sc->count_high << 3) | (low >> 29);
low = low << 3;
sc->buf[ pad >> 2 ] =
mm_bswap_32( _mm_set1_epi32( high ) );
sc->buf[ ( pad+4 ) >> 2 ] =
mm_bswap_32( _mm_set1_epi32( low ) );
sha256_4way_round( sc->buf, sc->val );
for ( u = 0; u < 8; u ++ )
((__m128i*)dst)[u] = mm_bswap_32( sc->val[u] );
}
#if defined(__AVX2__)
// SHA512 4 way 64 bit
static const sph_u64 H512[8] = {
SPH_C64(0x6A09E667F3BCC908), SPH_C64(0xBB67AE8584CAA73B),
SPH_C64(0x3C6EF372FE94F82B), SPH_C64(0xA54FF53A5F1D36F1),
SPH_C64(0x510E527FADE682D1), SPH_C64(0x9B05688C2B3E6C1F),
SPH_C64(0x1F83D9ABFB41BD6B), SPH_C64(0x5BE0CD19137E2179)
};
static const sph_u64 K512[80] = {
SPH_C64(0x428A2F98D728AE22), SPH_C64(0x7137449123EF65CD),
SPH_C64(0xB5C0FBCFEC4D3B2F), SPH_C64(0xE9B5DBA58189DBBC),
@@ -80,13 +302,6 @@ static const sph_u64 K512[80] = {
SPH_C64(0x5FCB6FAB3AD6FAEC), SPH_C64(0x6C44198C4A475817)
};
static const sph_u64 H512[8] = {
SPH_C64(0x6A09E667F3BCC908), SPH_C64(0xBB67AE8584CAA73B),
SPH_C64(0x3C6EF372FE94F82B), SPH_C64(0xA54FF53A5F1D36F1),
SPH_C64(0x510E527FADE682D1), SPH_C64(0x9B05688C2B3E6C1F),
SPH_C64(0x1F83D9ABFB41BD6B), SPH_C64(0x5BE0CD19137E2179)
};
#define CH(X, Y, Z) \
_mm256_xor_si256( _mm256_and_si256( _mm256_xor_si256( Y, Z ), X ), Z )
@@ -182,7 +397,7 @@ void sha512_4way( sha512_4way_context *sc, const void *data, size_t len )
{
__m256i *vdata = (__m256i*)data;
size_t ptr;
int buf_size = 128;
const int buf_size = 128;
ptr = (unsigned)sc->count & (buf_size - 1U);
while ( len > 0 )
@@ -207,8 +422,8 @@ void sha512_4way( sha512_4way_context *sc, const void *data, size_t len )
void sha512_4way_close( sha512_4way_context *sc, void *dst )
{
unsigned ptr, u;
int buf_size = 128;
int pad = buf_size - 16;
const int buf_size = 128;
const int pad = buf_size - 16;
ptr = (unsigned)sc->count & (buf_size - 1U);
sc->buf[ ptr>>3 ] = _mm256_set1_epi64x( 0x80 );
@@ -233,4 +448,5 @@ void sha512_4way_close( sha512_4way_context *sc, void *dst )
((__m256i*)dst)[u] = mm256_bswap_64( sc->val[u] );
}
#endif
#endif // __AVX2__
#endif // __AVX__

45
algo/sha/sha2-hash-4way.h
View File

@@ -44,47 +44,19 @@
#include "sph_types.h"
#include "avxdefs.h"
#if 0
#define SPH_SIZE_sha224 224
#if defined(__AVX__)
#define SPH_SIZE_sha256 256
typedef struct {
#ifndef DOXYGEN_IGNORE
unsigned char buf[64]; /* first field, for alignment */
sph_u32 val[8];
#if SPH_64
sph_u64 count;
#else
sph_u32 count_high, count_low;
#endif
#endif
} sph_sha224_context;
__m128i buf[64>>2];
__m128i val[8];
uint32_t count_high, count_low;
} sha256_4way_context;
typedef sph_sha224_context sph_sha256_context;
void sph_sha224_init(void *cc);
void sph_sha224(void *cc, const void *data, size_t len);
void sph_sha224_close(void *cc, void *dst);
void sph_sha224_addbits_and_close(void *cc, unsigned ub, unsigned n, void *dst);
void sph_sha224_comp(const sph_u32 msg[16], sph_u32 val[8]);
void sph_sha256_init(void *cc);
void sph_sha256(void *cc, const void *data, size_t len);
void sph_sha256_close(void *cc, void *dst);
void sph_sha256_addbits_and_close(void *cc, unsigned ub, unsigned n, void *dst);
void sph_sha256_comp(const sph_u32 msg[16], sph_u32 val[8]);
#endif
void sha256_4way_init( sha256_4way_context *sc );
void sha256_4way( sha256_4way_context *sc, const void *data, size_t len );
void sha256_4way_close( sha256_4way_context *sc, void *dst );
#if defined (__AVX2__)
@@ -102,3 +74,4 @@ void sha512_4way_close( sha512_4way_context *sc, void *dst );
#endif
#endif
#endif

17
avxdefs.h
View File

@@ -29,7 +29,7 @@
//
// operation;
// data: variable/constant name
// function: dexcription of operation
// function: deription of operation
//
// size: size of element if applicable
//
@@ -99,17 +99,17 @@ typedef union m128_v8 m128_v8;
#define mm_setc1_64( x ) {{ x, x }}
#define mm_setc_32( x3, x2, x1, x0 ) {{ x3, x2, x1, x0 }}
#define mm_setc1_32( x ) {{ [0 ... 3] = x }}
#define mm_setc1_32( x ) {{ x,x,x,x }}
#define mm_setc_16( x7, x6, x5, x4, x3, x2, x1, x0 ) \
{{ x7, x6, x5, x4, x3, x2, x1, x0 }}
#define mm_setc1_16( x ) {{ [0 ... 7] = x }}
#define mm_setc1_16( x ) {{ x,x,x,x, x,x,x,x }}
#define mm_setc_8( x15, x14, x13, x12, x11, x10, x09, x08, \
x07, x06, x05, x04, x03, x02, x01, x00 ) \
{{ x15, x14, x13, x12, x11, x10, x09, x08, \
x07, x06, x05, x04, x03, x02, x01, x00 }}
#define mm_setc1_8( x ) {{ [0 ... 15] = x }}
#define mm_setc1_8( x ) {{ x,x,x,x, x,x,x,x, x,x,x,x, x,x,x,x }}
// Compile time constants, use only for initializing.
#define c128_zero mm_setc1_64( 0ULL )
@@ -582,17 +582,17 @@ typedef union m256_v8 m256_v8;
// simple constant vectors.
#define mm256_setc_64( x3, x2, x1, x0 ) {{ x3, x2, x1, x0 }}
#define mm256_setc1_64( x ) {{ [0 ... 3] = x }}
#define mm256_setc1_64( x ) {{ x,x,x,x }}
#define mm256_setc_32( x7, x6, x5, x4, x3, x2, x1, x0 ) \
{{ x7, x6, x5, x4, x3, x2, x1, x0 }}
#define mm256_setc1_32( x ) {{ [0 ... 7] = x }}
#define mm256_setc1_32( x ) {{ x,x,x,x, x,x,x,x }}
#define mm256_setc_16( x15, x14, x13, x12, x11, x10, x09, x08, \
x07, x06, x05, x04, x03, x02, x01, x00 ) \
{{ x15, x14, x13, x12, x11, x10, x09, x08, \
x07, x06, x05, x04, x03, x02, x01, x00 }}
#define mm256_setc1_16( x ) {{ [0 ... 15] = x }}
#define mm256_setc1_16( x ) {{ x,x,x,x, x,x,x,x, x,x,x,x, x,x,x,x }}
#define mm256_setc_8( x31, x30, x29, x28, x27, x26, x25, x24, \
x23, x22, x21, x20, x19, x18, x17, x16, \
@@ -602,7 +602,8 @@ typedef union m256_v8 m256_v8;
x23, x22, x21, x20, x19, x18, x17, x16, \
x15, x14, x13, x12, x11, x10, x09, x08, \
x07, x06, x05, x04, x03, x02, x01, x00 }}
#define mm256_setc1_8( x ) {{ [0 ... 31] = x }}
#define mm256_setc1_8( x ) {{ x,x,x,x, x,x,x,x, x,x,x,x, x,x,x,x, \
x,x,x,x, x,x,x,x, x,x,x,x, x,x,x,x }}
// Predefined compile time constant vectors.
// Use Pseudo constants at run time for all simple constant vectors.

20
configure vendored
View File

@@ -1,6 +1,6 @@
#! /bin/sh
# Guess values for system-dependent variables and create Makefiles.
# Generated by GNU Autoconf 2.69 for cpuminer-opt 3.8.1.
# Generated by GNU Autoconf 2.69 for cpuminer-opt 3.8.1.1.
#
#
# Copyright (C) 1992-1996, 1998-2012 Free Software Foundation, Inc.
@@ -577,8 +577,8 @@ MAKEFLAGS=
# Identity of this package.
PACKAGE_NAME='cpuminer-opt'
PACKAGE_TARNAME='cpuminer-opt'
PACKAGE_VERSION='3.8.1'
PACKAGE_STRING='cpuminer-opt 3.8.1'
PACKAGE_VERSION='3.8.1.1'
PACKAGE_STRING='cpuminer-opt 3.8.1.1'
PACKAGE_BUGREPORT=''
PACKAGE_URL=''
@@ -1321,7 +1321,7 @@ if test "$ac_init_help" = "long"; then
# Omit some internal or obsolete options to make the list less imposing.
# This message is too long to be a string in the A/UX 3.1 sh.
cat <<_ACEOF
\`configure' configures cpuminer-opt 3.8.1 to adapt to many kinds of systems.
\`configure' configures cpuminer-opt 3.8.1.1 to adapt to many kinds of systems.
Usage: $0 [OPTION]... [VAR=VALUE]...
@@ -1392,7 +1392,7 @@ fi
if test -n "$ac_init_help"; then
case $ac_init_help in
short | recursive ) echo "Configuration of cpuminer-opt 3.8.1:";;
short | recursive ) echo "Configuration of cpuminer-opt 3.8.1.1:";;
esac
cat <<\_ACEOF
@@ -1497,7 +1497,7 @@ fi
test -n "$ac_init_help" && exit $ac_status
if $ac_init_version; then
cat <<\_ACEOF
cpuminer-opt configure 3.8.1
cpuminer-opt configure 3.8.1.1
generated by GNU Autoconf 2.69
Copyright (C) 2012 Free Software Foundation, Inc.
@@ -2000,7 +2000,7 @@ cat >config.log <<_ACEOF
This file contains any messages produced by compilers while
running configure, to aid debugging if configure makes a mistake.
It was created by cpuminer-opt $as_me 3.8.1, which was
It was created by cpuminer-opt $as_me 3.8.1.1, which was
generated by GNU Autoconf 2.69. Invocation command line was
$ $0 $@
@@ -2981,7 +2981,7 @@ fi
# Define the identity of the package.
PACKAGE='cpuminer-opt'
VERSION='3.8.1'
VERSION='3.8.1.1'
cat >>confdefs.h <<_ACEOF
@@ -6677,7 +6677,7 @@ cat >>$CONFIG_STATUS <<\_ACEOF || ac_write_fail=1
# report actual input values of CONFIG_FILES etc. instead of their
# values after options handling.
ac_log="
This file was extended by cpuminer-opt $as_me 3.8.1, which was
This file was extended by cpuminer-opt $as_me 3.8.1.1, which was
generated by GNU Autoconf 2.69. Invocation command line was
CONFIG_FILES = $CONFIG_FILES
@@ -6743,7 +6743,7 @@ _ACEOF
cat >>$CONFIG_STATUS <<_ACEOF || ac_write_fail=1
ac_cs_config="`$as_echo "$ac_configure_args" | sed 's/^ //; s/[\\""\`\$]/\\\\&/g'`"
ac_cs_version="\\
cpuminer-opt config.status 3.8.1
cpuminer-opt config.status 3.8.1.1
configured by $0, generated by GNU Autoconf 2.69,
with options \\"\$ac_cs_config\\"

2
configure.ac
View File

@@ -1,4 +1,4 @@
AC_INIT([cpuminer-opt], [3.8.1])
AC_INIT([cpuminer-opt], [3.8.1.1])
AC_PREREQ([2.59c])
AC_CANONICAL_SYSTEM