Tanguy Pruvot
10 years ago
11 changed files with 327 additions and 184 deletions
@ -1,165 +1,171 @@
@@ -1,165 +1,171 @@
|
||||
#include <stdio.h> |
||||
#include <memory.h> |
||||
|
||||
#include "cuda_helper.h" |
||||
|
||||
// aus heavy.cu |
||||
extern cudaError_t MyStreamSynchronize(cudaStream_t stream, int situation, int thr_id); |
||||
|
||||
#define U32TO64_LE(p) \ |
||||
(((uint64_t)(*p)) | (((uint64_t)(*(p + 1))) << 32)) |
||||
|
||||
#define U64TO32_LE(p, v) \ |
||||
*p = (uint32_t)((v)); *(p+1) = (uint32_t)((v) >> 32); |
||||
|
||||
__device__ __constant__ |
||||
static const uint64_t c_keccak_round_constants[24] = { |
||||
0x0000000000000001ull, 0x0000000000008082ull, |
||||
0x800000000000808aull, 0x8000000080008000ull, |
||||
0x000000000000808bull, 0x0000000080000001ull, |
||||
0x8000000080008081ull, 0x8000000000008009ull, |
||||
0x000000000000008aull, 0x0000000000000088ull, |
||||
0x0000000080008009ull, 0x000000008000000aull, |
||||
0x000000008000808bull, 0x800000000000008bull, |
||||
0x8000000000008089ull, 0x8000000000008003ull, |
||||
0x8000000000008002ull, 0x8000000000000080ull, |
||||
0x000000000000800aull, 0x800000008000000aull, |
||||
0x8000000080008081ull, 0x8000000000008080ull, |
||||
0x0000000080000001ull, 0x8000000080008008ull |
||||
}; |
||||
|
||||
static __device__ __forceinline__ void |
||||
keccak_block(uint64_t *s, const uint32_t *in, const uint64_t *keccak_round_constants) { |
||||
size_t i; |
||||
uint64_t t[5], u[5], v, w; |
||||
|
||||
/* absorb input */ |
||||
#pragma unroll 9 |
||||
for (i = 0; i < 72 / 8; i++, in += 2) |
||||
s[i] ^= U32TO64_LE(in); |
||||
|
||||
for (i = 0; i < 24; i++) { |
||||
/* theta: c = a[0,i] ^ a[1,i] ^ .. a[4,i] */ |
||||
t[0] = s[0] ^ s[5] ^ s[10] ^ s[15] ^ s[20]; |
||||
t[1] = s[1] ^ s[6] ^ s[11] ^ s[16] ^ s[21]; |
||||
t[2] = s[2] ^ s[7] ^ s[12] ^ s[17] ^ s[22]; |
||||
t[3] = s[3] ^ s[8] ^ s[13] ^ s[18] ^ s[23]; |
||||
t[4] = s[4] ^ s[9] ^ s[14] ^ s[19] ^ s[24]; |
||||
|
||||
/* theta: d[i] = c[i+4] ^ rotl(c[i+1],1) */ |
||||
u[0] = t[4] ^ ROTL64(t[1], 1); |
||||
u[1] = t[0] ^ ROTL64(t[2], 1); |
||||
u[2] = t[1] ^ ROTL64(t[3], 1); |
||||
u[3] = t[2] ^ ROTL64(t[4], 1); |
||||
u[4] = t[3] ^ ROTL64(t[0], 1); |
||||
|
||||
/* theta: a[0,i], a[1,i], .. a[4,i] ^= d[i] */ |
||||
s[0] ^= u[0]; s[5] ^= u[0]; s[10] ^= u[0]; s[15] ^= u[0]; s[20] ^= u[0]; |
||||
s[1] ^= u[1]; s[6] ^= u[1]; s[11] ^= u[1]; s[16] ^= u[1]; s[21] ^= u[1]; |
||||
s[2] ^= u[2]; s[7] ^= u[2]; s[12] ^= u[2]; s[17] ^= u[2]; s[22] ^= u[2]; |
||||
s[3] ^= u[3]; s[8] ^= u[3]; s[13] ^= u[3]; s[18] ^= u[3]; s[23] ^= u[3]; |
||||
s[4] ^= u[4]; s[9] ^= u[4]; s[14] ^= u[4]; s[19] ^= u[4]; s[24] ^= u[4]; |
||||
|
||||
/* rho pi: b[..] = rotl(a[..], ..) */ |
||||
v = s[ 1]; |
||||
s[ 1] = ROTL64(s[ 6], 44); |
||||
s[ 6] = ROTL64(s[ 9], 20); |
||||
s[ 9] = ROTL64(s[22], 61); |
||||
s[22] = ROTL64(s[14], 39); |
||||
s[14] = ROTL64(s[20], 18); |
||||
s[20] = ROTL64(s[ 2], 62); |
||||
s[ 2] = ROTL64(s[12], 43); |
||||
s[12] = ROTL64(s[13], 25); |
||||
s[13] = ROTL64(s[19], 8); |
||||
s[19] = ROTL64(s[23], 56); |
||||
s[23] = ROTL64(s[15], 41); |
||||
s[15] = ROTL64(s[ 4], 27); |
||||
s[ 4] = ROTL64(s[24], 14); |
||||
s[24] = ROTL64(s[21], 2); |
||||
s[21] = ROTL64(s[ 8], 55); |
||||
s[ 8] = ROTL64(s[16], 45); |
||||
s[16] = ROTL64(s[ 5], 36); |
||||
s[ 5] = ROTL64(s[ 3], 28); |
||||
s[ 3] = ROTL64(s[18], 21); |
||||
s[18] = ROTL64(s[17], 15); |
||||
s[17] = ROTL64(s[11], 10); |
||||
s[11] = ROTL64(s[ 7], 6); |
||||
s[ 7] = ROTL64(s[10], 3); |
||||
s[10] = ROTL64( v, 1); |
||||
|
||||
/* chi: a[i,j] ^= ~b[i,j+1] & b[i,j+2] */ |
||||
v = s[ 0]; w = s[ 1]; s[ 0] ^= (~w) & s[ 2]; s[ 1] ^= (~s[ 2]) & s[ 3]; s[ 2] ^= (~s[ 3]) & s[ 4]; s[ 3] ^= (~s[ 4]) & v; s[ 4] ^= (~v) & w; |
||||
v = s[ 5]; w = s[ 6]; s[ 5] ^= (~w) & s[ 7]; s[ 6] ^= (~s[ 7]) & s[ 8]; s[ 7] ^= (~s[ 8]) & s[ 9]; s[ 8] ^= (~s[ 9]) & v; s[ 9] ^= (~v) & w; |
||||
v = s[10]; w = s[11]; s[10] ^= (~w) & s[12]; s[11] ^= (~s[12]) & s[13]; s[12] ^= (~s[13]) & s[14]; s[13] ^= (~s[14]) & v; s[14] ^= (~v) & w; |
||||
v = s[15]; w = s[16]; s[15] ^= (~w) & s[17]; s[16] ^= (~s[17]) & s[18]; s[17] ^= (~s[18]) & s[19]; s[18] ^= (~s[19]) & v; s[19] ^= (~v) & w; |
||||
v = s[20]; w = s[21]; s[20] ^= (~w) & s[22]; s[21] ^= (~s[22]) & s[23]; s[22] ^= (~s[23]) & s[24]; s[23] ^= (~s[24]) & v; s[24] ^= (~v) & w; |
||||
|
||||
/* iota: a[0,0] ^= round constant */ |
||||
s[0] ^= keccak_round_constants[i]; |
||||
} |
||||
} |
||||
|
||||
__global__ void quark_keccak512_gpu_hash_64(int threads, uint32_t startNounce, uint64_t *g_hash, uint32_t *g_nonceVector) |
||||
{ |
||||
int thread = (blockDim.x * blockIdx.x + threadIdx.x); |
||||
if (thread < threads) |
||||
{ |
||||
uint32_t nounce = (g_nonceVector != NULL) ? g_nonceVector[thread] : (startNounce + thread); |
||||
|
||||
int hashPosition = nounce - startNounce; |
||||
uint32_t *inpHash = (uint32_t*)&g_hash[8 * hashPosition]; |
||||
|
||||
// Nachricht kopieren |
||||
uint32_t message[18]; |
||||
#pragma unroll 16 |
||||
for(int i=0;i<16;i++) |
||||
message[i] = inpHash[i]; |
||||
|
||||
message[16] = 0x01; |
||||
message[17] = 0x80000000; |
||||
|
||||
// State initialisieren |
||||
uint64_t keccak_gpu_state[25]; |
||||
#pragma unroll 25 |
||||
for (int i=0; i<25; i++) |
||||
keccak_gpu_state[i] = 0; |
||||
|
||||
// den Block einmal gut durchschütteln |
||||
keccak_block(keccak_gpu_state, message, c_keccak_round_constants); |
||||
|
||||
// das Hash erzeugen |
||||
uint32_t hash[16]; |
||||
|
||||
#pragma unroll 8 |
||||
for (size_t i = 0; i < 64; i += 8) { |
||||
U64TO32_LE((&hash[i/4]), keccak_gpu_state[i / 8]); |
||||
} |
||||
|
||||
// fertig |
||||
uint32_t *outpHash = (uint32_t*)&g_hash[8 * hashPosition]; |
||||
|
||||
#pragma unroll 16 |
||||
for(int i=0;i<16;i++) |
||||
outpHash[i] = hash[i]; |
||||
} |
||||
} |
||||
|
||||
// Setup-Funktionen |
||||
__host__ void quark_keccak512_cpu_init(int thr_id, int threads) |
||||
{ |
||||
} |
||||
|
||||
__host__ void quark_keccak512_cpu_hash_64(int thr_id, int threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order) |
||||
{ |
||||
const int threadsperblock = 256; |
||||
|
||||
// berechne wie viele Thread Blocks wir brauchen |
||||
dim3 grid((threads + threadsperblock-1)/threadsperblock); |
||||
dim3 block(threadsperblock); |
||||
|
||||
// Größe des dynamischen Shared Memory Bereichs |
||||
size_t shared_size = 0; |
||||
|
||||
quark_keccak512_gpu_hash_64<<<grid, block, shared_size>>>(threads, startNounce, (uint64_t*)d_hash, d_nonceVector); |
||||
MyStreamSynchronize(NULL, order, thr_id); |
||||
} |
||||
#include <stdio.h> |
||||
#include <memory.h> |
||||
|
||||
#include "cuda_helper.h" |
||||
|
||||
// heavy.cu |
||||
extern cudaError_t MyStreamSynchronize(cudaStream_t stream, int situation, int thr_id); |
||||
|
||||
#define U32TO64_LE(p) \ |
||||
(((uint64_t)(*p)) | (((uint64_t)(*(p + 1))) << 32)) |
||||
|
||||
#define U64TO32_LE(p, v) \ |
||||
*p = (uint32_t)((v)); *(p+1) = (uint32_t)((v) >> 32); |
||||
|
||||
static const uint64_t host_keccak_round_constants[24] = { |
||||
0x0000000000000001ull, 0x0000000000008082ull, |
||||
0x800000000000808aull, 0x8000000080008000ull, |
||||
0x000000000000808bull, 0x0000000080000001ull, |
||||
0x8000000080008081ull, 0x8000000000008009ull, |
||||
0x000000000000008aull, 0x0000000000000088ull, |
||||
0x0000000080008009ull, 0x000000008000000aull, |
||||
0x000000008000808bull, 0x800000000000008bull, |
||||
0x8000000000008089ull, 0x8000000000008003ull, |
||||
0x8000000000008002ull, 0x8000000000000080ull, |
||||
0x000000000000800aull, 0x800000008000000aull, |
||||
0x8000000080008081ull, 0x8000000000008080ull, |
||||
0x0000000080000001ull, 0x8000000080008008ull |
||||
}; |
||||
|
||||
__constant__ uint64_t c_keccak_round_constants[24]; |
||||
|
||||
static __device__ __forceinline__ void |
||||
keccak_block(uint64_t *s, const uint32_t *in, const uint64_t *keccak_round_constants) { |
||||
size_t i; |
||||
uint64_t t[5], u[5], v, w; |
||||
|
||||
/* absorb input */ |
||||
#pragma unroll 9 |
||||
for (i = 0; i < 72 / 8; i++, in += 2) |
||||
s[i] ^= U32TO64_LE(in); |
||||
|
||||
for (i = 0; i < 24; i++) { |
||||
/* theta: c = a[0,i] ^ a[1,i] ^ .. a[4,i] */ |
||||
t[0] = s[0] ^ s[5] ^ s[10] ^ s[15] ^ s[20]; |
||||
t[1] = s[1] ^ s[6] ^ s[11] ^ s[16] ^ s[21]; |
||||
t[2] = s[2] ^ s[7] ^ s[12] ^ s[17] ^ s[22]; |
||||
t[3] = s[3] ^ s[8] ^ s[13] ^ s[18] ^ s[23]; |
||||
t[4] = s[4] ^ s[9] ^ s[14] ^ s[19] ^ s[24]; |
||||
|
||||
/* theta: d[i] = c[i+4] ^ rotl(c[i+1],1) */ |
||||
u[0] = t[4] ^ ROTL64(t[1], 1); |
||||
u[1] = t[0] ^ ROTL64(t[2], 1); |
||||
u[2] = t[1] ^ ROTL64(t[3], 1); |
||||
u[3] = t[2] ^ ROTL64(t[4], 1); |
||||
u[4] = t[3] ^ ROTL64(t[0], 1); |
||||
|
||||
/* theta: a[0,i], a[1,i], .. a[4,i] ^= d[i] */ |
||||
s[0] ^= u[0]; s[5] ^= u[0]; s[10] ^= u[0]; s[15] ^= u[0]; s[20] ^= u[0]; |
||||
s[1] ^= u[1]; s[6] ^= u[1]; s[11] ^= u[1]; s[16] ^= u[1]; s[21] ^= u[1]; |
||||
s[2] ^= u[2]; s[7] ^= u[2]; s[12] ^= u[2]; s[17] ^= u[2]; s[22] ^= u[2]; |
||||
s[3] ^= u[3]; s[8] ^= u[3]; s[13] ^= u[3]; s[18] ^= u[3]; s[23] ^= u[3]; |
||||
s[4] ^= u[4]; s[9] ^= u[4]; s[14] ^= u[4]; s[19] ^= u[4]; s[24] ^= u[4]; |
||||
|
||||
/* rho pi: b[..] = rotl(a[..], ..) */ |
||||
v = s[ 1]; |
||||
s[ 1] = ROTL64(s[ 6], 44); |
||||
s[ 6] = ROTL64(s[ 9], 20); |
||||
s[ 9] = ROTL64(s[22], 61); |
||||
s[22] = ROTL64(s[14], 39); |
||||
s[14] = ROTL64(s[20], 18); |
||||
s[20] = ROTL64(s[ 2], 62); |
||||
s[ 2] = ROTL64(s[12], 43); |
||||
s[12] = ROTL64(s[13], 25); |
||||
s[13] = ROTL64(s[19], 8); |
||||
s[19] = ROTL64(s[23], 56); |
||||
s[23] = ROTL64(s[15], 41); |
||||
s[15] = ROTL64(s[ 4], 27); |
||||
s[ 4] = ROTL64(s[24], 14); |
||||
s[24] = ROTL64(s[21], 2); |
||||
s[21] = ROTL64(s[ 8], 55); |
||||
s[ 8] = ROTL64(s[16], 45); |
||||
s[16] = ROTL64(s[ 5], 36); |
||||
s[ 5] = ROTL64(s[ 3], 28); |
||||
s[ 3] = ROTL64(s[18], 21); |
||||
s[18] = ROTL64(s[17], 15); |
||||
s[17] = ROTL64(s[11], 10); |
||||
s[11] = ROTL64(s[ 7], 6); |
||||
s[ 7] = ROTL64(s[10], 3); |
||||
s[10] = ROTL64( v, 1); |
||||
|
||||
/* chi: a[i,j] ^= ~b[i,j+1] & b[i,j+2] */ |
||||
v = s[ 0]; w = s[ 1]; s[ 0] ^= (~w) & s[ 2]; s[ 1] ^= (~s[ 2]) & s[ 3]; s[ 2] ^= (~s[ 3]) & s[ 4]; s[ 3] ^= (~s[ 4]) & v; s[ 4] ^= (~v) & w; |
||||
v = s[ 5]; w = s[ 6]; s[ 5] ^= (~w) & s[ 7]; s[ 6] ^= (~s[ 7]) & s[ 8]; s[ 7] ^= (~s[ 8]) & s[ 9]; s[ 8] ^= (~s[ 9]) & v; s[ 9] ^= (~v) & w; |
||||
v = s[10]; w = s[11]; s[10] ^= (~w) & s[12]; s[11] ^= (~s[12]) & s[13]; s[12] ^= (~s[13]) & s[14]; s[13] ^= (~s[14]) & v; s[14] ^= (~v) & w; |
||||
v = s[15]; w = s[16]; s[15] ^= (~w) & s[17]; s[16] ^= (~s[17]) & s[18]; s[17] ^= (~s[18]) & s[19]; s[18] ^= (~s[19]) & v; s[19] ^= (~v) & w; |
||||
v = s[20]; w = s[21]; s[20] ^= (~w) & s[22]; s[21] ^= (~s[22]) & s[23]; s[22] ^= (~s[23]) & s[24]; s[23] ^= (~s[24]) & v; s[24] ^= (~v) & w; |
||||
|
||||
/* iota: a[0,0] ^= round constant */ |
||||
s[0] ^= keccak_round_constants[i]; |
||||
} |
||||
} |
||||
|
||||
__global__ void quark_keccak512_gpu_hash_64(int threads, uint32_t startNounce, uint64_t *g_hash, uint32_t *g_nonceVector) |
||||
{ |
||||
int thread = (blockDim.x * blockIdx.x + threadIdx.x); |
||||
if (thread < threads) |
||||
{ |
||||
uint32_t nounce = (g_nonceVector != NULL) ? g_nonceVector[thread] : (startNounce + thread); |
||||
|
||||
int hashPosition = nounce - startNounce; |
||||
uint32_t *inpHash = (uint32_t*)&g_hash[8 * hashPosition]; |
||||
|
||||
// Nachricht kopieren |
||||
uint32_t message[18]; |
||||
#pragma unroll 16 |
||||
for(int i=0;i<16;i++) |
||||
message[i] = inpHash[i]; |
||||
|
||||
message[16] = 0x01; |
||||
message[17] = 0x80000000; |
||||
|
||||
// State initialisieren |
||||
uint64_t keccak_gpu_state[25]; |
||||
#pragma unroll 25 |
||||
for (int i=0; i<25; i++) |
||||
keccak_gpu_state[i] = 0; |
||||
|
||||
// den Block einmal gut durchschütteln |
||||
keccak_block(keccak_gpu_state, message, c_keccak_round_constants); |
||||
|
||||
// das Hash erzeugen |
||||
uint32_t hash[16]; |
||||
|
||||
#pragma unroll 8 |
||||
for (size_t i = 0; i < 64; i += 8) { |
||||
U64TO32_LE((&hash[i/4]), keccak_gpu_state[i / 8]); |
||||
} |
||||
|
||||
// fertig |
||||
uint32_t *outpHash = (uint32_t*)&g_hash[8 * hashPosition]; |
||||
|
||||
#pragma unroll 16 |
||||
for(int i=0;i<16;i++) |
||||
outpHash[i] = hash[i]; |
||||
} |
||||
} |
||||
|
||||
// Setup-Funktionen |
||||
__host__ void quark_keccak512_cpu_init(int thr_id, int threads) |
||||
{ |
||||
// Kopiere die Hash-Tabellen in den GPU-Speicher |
||||
cudaMemcpyToSymbol( c_keccak_round_constants, |
||||
host_keccak_round_constants, |
||||
sizeof(host_keccak_round_constants), |
||||
0, cudaMemcpyHostToDevice); |
||||
} |
||||
|
||||
__host__ void quark_keccak512_cpu_hash_64(int thr_id, int threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order) |
||||
{ |
||||
const int threadsperblock = 256; |
||||
|
||||
// berechne wie viele Thread Blocks wir brauchen |
||||
dim3 grid((threads + threadsperblock-1)/threadsperblock); |
||||
dim3 block(threadsperblock); |
||||
|
||||
// Größe des dynamischen Shared Memory Bereichs |
||||
size_t shared_size = 0; |
||||
|
||||
quark_keccak512_gpu_hash_64<<<grid, block, shared_size>>>(threads, startNounce, (uint64_t*)d_hash, d_nonceVector); |
||||
MyStreamSynchronize(NULL, order, thr_id); |
||||
} |
||||
|
||||
@ -0,0 +1,117 @@
@@ -0,0 +1,117 @@
|
||||
/* |
||||
* whirlpool routine (djm) |
||||
*/ |
||||
extern "C" |
||||
{ |
||||
#include "sph/sph_whirlpool.h" |
||||
#include "miner.h" |
||||
} |
||||
|
||||
// from cpu-miner.c |
||||
extern int device_map[8]; |
||||
extern bool opt_benchmark; |
||||
|
||||
// Speicher für Input/Output der verketteten Hashfunktionen |
||||
static uint32_t *d_hash[8]; |
||||
|
||||
extern void x15_whirlpool_cpu_init(int thr_id, int threads, int mode); |
||||
extern void whirlpool512_setBlock_80(void *pdata, const void *ptarget); |
||||
extern void whirlpool512_cpu_hash_80(int thr_id, int threads, uint32_t startNounce, uint32_t *d_hash, int order); |
||||
extern void x15_whirlpool_cpu_hash_64(int thr_id, int threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order); |
||||
extern uint32_t whirlpool512_cpu_finalhash_64(int thr_id, int threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order); |
||||
|
||||
extern void cuda_check_cpu_init(int thr_id, int threads); |
||||
extern void cuda_check_cpu_setTarget(const void *ptarget); |
||||
extern uint32_t cuda_check_cpu_hash_64(int thr_id, int threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_inputHash, int order); |
||||
|
||||
// CPU Hash function |
||||
extern "C" void wcoinhash(void *state, const void *input) |
||||
{ |
||||
sph_whirlpool_context ctx_whirlpool; |
||||
|
||||
uint32_t hash[16]; |
||||
|
||||
// shavite 1 |
||||
sph_whirlpool1_init(&ctx_whirlpool); |
||||
sph_whirlpool1(&ctx_whirlpool, input, 80); |
||||
sph_whirlpool1_close(&ctx_whirlpool, (void*) hash); |
||||
|
||||
sph_whirlpool1_init(&ctx_whirlpool); |
||||
sph_whirlpool1(&ctx_whirlpool, (const void*) hash, 64); |
||||
sph_whirlpool1_close(&ctx_whirlpool, (void*) hash); |
||||
|
||||
sph_whirlpool1_init(&ctx_whirlpool); |
||||
sph_whirlpool1(&ctx_whirlpool, (const void*) hash, 64); |
||||
sph_whirlpool1_close(&ctx_whirlpool, (void*) hash); |
||||
|
||||
sph_whirlpool1_init(&ctx_whirlpool); |
||||
sph_whirlpool1(&ctx_whirlpool, (const void*) hash, 64); |
||||
sph_whirlpool1_close(&ctx_whirlpool, (void*) hash); |
||||
|
||||
memcpy(state, hash, 32); |
||||
} |
||||
|
||||
extern "C" int scanhash_whc(int thr_id, uint32_t *pdata, |
||||
const uint32_t *ptarget, uint32_t max_nonce, |
||||
unsigned long *hashes_done) |
||||
{ |
||||
const uint32_t first_nonce = pdata[19]; |
||||
const int throughput = 256*256*8; |
||||
static bool init[8] = {0,0,0,0,0,0,0,0}; |
||||
uint32_t endiandata[20]; |
||||
uint32_t Htarg = ptarget[7]; |
||||
|
||||
if (opt_benchmark) |
||||
((uint32_t*)ptarget)[7] = Htarg = 0x0000ff; |
||||
|
||||
if (!init[thr_id]) { |
||||
cudaSetDevice(device_map[thr_id]); |
||||
// Konstanten kopieren, Speicher belegen |
||||
cudaMalloc(&d_hash[thr_id], 16 * sizeof(uint32_t) * throughput); |
||||
x15_whirlpool_cpu_init(thr_id, throughput,1); |
||||
|
||||
init[thr_id] = true; |
||||
} |
||||
|
||||
for (int k=0; k < 20; k++) { |
||||
be32enc(&endiandata[k], ((uint32_t*)pdata)[k]); |
||||
} |
||||
|
||||
whirlpool512_setBlock_80((void*)endiandata, ptarget); |
||||
|
||||
do { |
||||
uint32_t foundNonce; |
||||
int order = 0; |
||||
|
||||
whirlpool512_cpu_hash_80(thr_id, throughput, pdata[19], d_hash[thr_id], order++); |
||||
x15_whirlpool_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++); |
||||
x15_whirlpool_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++); |
||||
|
||||
foundNonce = whirlpool512_cpu_finalhash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++); |
||||
if (foundNonce != 0xffffffff) |
||||
{ |
||||
uint32_t vhash64[8]; |
||||
be32enc(&endiandata[19], foundNonce); |
||||
|
||||
wcoinhash(vhash64, endiandata); |
||||
|
||||
if (vhash64[7] <= Htarg && fulltest(vhash64, ptarget)) |
||||
{ |
||||
pdata[19] = foundNonce; |
||||
*hashes_done = foundNonce - first_nonce + 1; |
||||
return 1; |
||||
} |
||||
else if (vhash64[7] > Htarg) { |
||||
applog(LOG_INFO, "GPU #%d: result for %08x is not in range: %x > %x", thr_id, foundNonce, vhash64[7], Htarg); |
||||
} |
||||
else { |
||||
applog(LOG_INFO, "GPU #%d: result for %08x does not validate on CPU!", thr_id, foundNonce); |
||||
} |
||||
} |
||||
pdata[19] += throughput; |
||||
|
||||
} while (pdata[19] < max_nonce && !work_restart[thr_id].restart); |
||||
|
||||
*hashes_done = pdata[19] - first_nonce + 1; |
||||
return 0; |
||||
} |
||||
Loading…
Reference in new issue