Browse Source

neoscrypt: cleanup...

My SM 3.0 functions are ok but djm34 implementation uses too much registers
for this arch...
2upstream
Tanguy Pruvot 10 years ago
parent
commit
1b31f11252
  1. 43
      neoscrypt/cuda_neoscrypt.cu
  2. 19
      neoscrypt/cuda_vectors.h
  3. 30
      neoscrypt/neoscrypt.cpp

43
neoscrypt/cuda_neoscrypt.cu

@ -4,9 +4,9 @@
#include "cuda_helper.h" #include "cuda_helper.h"
#include "cuda_vectors.h" /* NOT COMPATIBLE WITH SM 3.0 !!! */ #include "cuda_vectors.h" /* NOT COMPATIBLE WITH SM 3.0 !!! */
__device__ uint4* W; static uint32_t *d_buffer[MAX_GPUS];
uint32_t *d_NNonce[MAX_GPUS]; static uint32_t *d_NNonce[MAX_GPUS];
uint32_t *d_nnounce[MAX_GPUS]; __constant__ uint4* W;
__constant__ uint32_t pTarget[8]; __constant__ uint32_t pTarget[8];
__constant__ uint32_t key_init[16]; __constant__ uint32_t key_init[16];
__constant__ uint32_t input_init[16]; __constant__ uint32_t input_init[16];
@ -423,14 +423,14 @@ static __device__ __forceinline__ void neoscrypt_salsa(uint16 *XV)
#define SHIFT 130 #define SHIFT 130
__global__ __launch_bounds__(128, 1) __global__ __launch_bounds__(128, 1)
void neoscrypt_gpu_hash_k0(int stratum, uint32_t threads, uint32_t startNonce) void neoscrypt_gpu_hash_k0(uint32_t threads, uint32_t startNonce, bool stratum)
{ {
uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x); uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x);
uint32_t shift = SHIFT * 16 * thread;
// if (thread < threads) // if (thread < threads)
{ {
uint32_t data[80]; uint32_t data[80];
uint16 X[4]; uint16 X[4];
uint32_t shift = thread * SHIFT * 16;
const uint32_t nonce = startNonce + thread; const uint32_t nonce = startNonce + thread;
for (int i = 0; i<20; i++) { for (int i = 0; i<20; i++) {
@ -451,10 +451,10 @@ __global__ __launch_bounds__(128, 1)
void neoscrypt_gpu_hash_k01(uint32_t threads, uint32_t startNonce) void neoscrypt_gpu_hash_k01(uint32_t threads, uint32_t startNonce)
{ {
uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x); uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x);
uint32_t shift = SHIFT * 16 * thread;
// if (thread < threads) // if (thread < threads)
{ {
uint16 X[4]; uint16 X[4];
uint32_t shift = thread * SHIFT * 16;
((uintx64 *)X)[0]= ldg256(&(W + shift)[0]); ((uintx64 *)X)[0]= ldg256(&(W + shift)[0]);
//#pragma unroll //#pragma unroll
@ -471,10 +471,10 @@ __global__ __launch_bounds__(128, 1)
void neoscrypt_gpu_hash_k2(uint32_t threads, uint32_t startNonce) void neoscrypt_gpu_hash_k2(uint32_t threads, uint32_t startNonce)
{ {
uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x); uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x);
uint32_t shift = SHIFT * 16 * thread;
// if (thread < threads) // if (thread < threads)
{ {
uint16 X[4]; uint16 X[4];
uint32_t shift = thread * SHIFT * 16;
((uintx64 *)X)[0] = ldg256(&(W + shift)[2048]); ((uintx64 *)X)[0] = ldg256(&(W + shift)[2048]);
for (int t = 0; t < 128; t++) for (int t = 0; t < 128; t++)
@ -495,7 +495,7 @@ void neoscrypt_gpu_hash_k3(uint32_t threads, uint32_t startNonce)
uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x); uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x);
// if (thread < threads) // if (thread < threads)
{ {
uint32_t shift = SHIFT * 16 * thread; uint32_t shift = thread * SHIFT * 16;
uint16 Z[4]; uint16 Z[4];
((uintx64*)Z)[0] = ldg256(&(W + shift)[0]); ((uintx64*)Z)[0] = ldg256(&(W + shift)[0]);
@ -510,14 +510,14 @@ void neoscrypt_gpu_hash_k3(uint32_t threads, uint32_t startNonce)
} }
__global__ __launch_bounds__(128, 1) __global__ __launch_bounds__(128, 1)
void neoscrypt_gpu_hash_k4(int stratum, uint32_t threads, uint32_t startNonce, uint32_t *nonceVector) void neoscrypt_gpu_hash_k4(uint32_t threads, uint32_t startNonce, uint32_t *nonceRes, bool stratum)
{ {
uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x); uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x);
// if (thread < threads) if (thread < threads)
{ {
const uint32_t nonce = startNonce + thread; const uint32_t nonce = startNonce + thread;
uint32_t shift = SHIFT * 16 * thread; uint32_t shift = thread * SHIFT * 16;
uint16 Z[4]; uint16 Z[4];
uint32_t outbuf[8]; uint32_t outbuf[8];
uint32_t data[80]; uint32_t data[80];
@ -539,33 +539,37 @@ void neoscrypt_gpu_hash_k4(int stratum, uint32_t threads, uint32_t startNonce, u
((uintx64 *)Z)[0] ^= ldg256(&(W + shift)[2064]); ((uintx64 *)Z)[0] ^= ldg256(&(W + shift)[2064]);
fastkdf32(data, (uint32_t*)Z, outbuf); fastkdf32(data, (uint32_t*)Z, outbuf);
if (outbuf[7] <= pTarget[7]) { if (outbuf[7] <= pTarget[7]) {
uint32_t tmp = atomicExch(&nonceVector[0], nonce); atomicMin(nonceRes, nonce); // init val is UINT32_MAX
} }
} }
} }
void neoscrypt_cpu_init(int thr_id, uint32_t threads, uint32_t *hash) __host__
void neoscrypt_cpu_init(int thr_id, uint32_t threads)
{ {
cudaMemcpyToSymbol(BLAKE2S_SIGMA, BLAKE2S_SIGMA_host, sizeof(BLAKE2S_SIGMA_host), 0, cudaMemcpyHostToDevice); cuda_get_arch(thr_id);
cudaMemcpyToSymbol(W, &hash, sizeof(hash), 0, cudaMemcpyHostToDevice);
cudaMalloc(&d_NNonce[thr_id], sizeof(uint32_t)); cudaMalloc(&d_NNonce[thr_id], sizeof(uint32_t));
CUDA_SAFE_CALL(cudaMalloc(&d_buffer[thr_id], threads * 256 * SHIFT));
cudaMemcpyToSymbol(W, &d_buffer[thr_id], sizeof(uint4*), 0, cudaMemcpyHostToDevice);
cudaMemcpyToSymbol(BLAKE2S_SIGMA, BLAKE2S_SIGMA_host, sizeof(BLAKE2S_SIGMA_host), 0, cudaMemcpyHostToDevice);
} }
__host__ __host__
uint32_t neoscrypt_cpu_hash_k4(int stratum, int thr_id, uint32_t threads, uint32_t startNounce, int order) uint32_t neoscrypt_cpu_hash_k4(int thr_id, uint32_t threads, uint32_t startNounce, bool have_stratum, int order)
{ {
uint32_t result[MAX_GPUS] = { 0xffffffff }; uint32_t result[MAX_GPUS];
memset(result, 0xff, sizeof(result));
cudaMemset(d_NNonce[thr_id], 0xff, sizeof(uint32_t)); cudaMemset(d_NNonce[thr_id], 0xff, sizeof(uint32_t));
const uint32_t threadsperblock = 128; const uint32_t threadsperblock = 128;
dim3 grid((threads + threadsperblock - 1) / threadsperblock); dim3 grid((threads + threadsperblock - 1) / threadsperblock);
dim3 block(threadsperblock); dim3 block(threadsperblock);
neoscrypt_gpu_hash_k0 <<< grid, block >>>(stratum, threads, startNounce); neoscrypt_gpu_hash_k0 <<< grid, block >>>(threads, startNounce, have_stratum);
neoscrypt_gpu_hash_k01 <<< grid, block >>>(threads, startNounce); neoscrypt_gpu_hash_k01 <<< grid, block >>>(threads, startNounce);
neoscrypt_gpu_hash_k2 <<< grid, block >>>(threads, startNounce); neoscrypt_gpu_hash_k2 <<< grid, block >>>(threads, startNounce);
neoscrypt_gpu_hash_k3 <<< grid, block >>>(threads, startNounce); neoscrypt_gpu_hash_k3 <<< grid, block >>>(threads, startNounce);
neoscrypt_gpu_hash_k4 <<< grid, block >>>(stratum, threads, startNounce, d_NNonce[thr_id]); neoscrypt_gpu_hash_k4 <<< grid, block >>>(threads, startNounce, d_NNonce[thr_id], have_stratum);
MyStreamSynchronize(NULL, order, thr_id); MyStreamSynchronize(NULL, order, thr_id);
cudaMemcpy(&result[thr_id], d_NNonce[thr_id], sizeof(uint32_t), cudaMemcpyDeviceToHost); cudaMemcpy(&result[thr_id], d_NNonce[thr_id], sizeof(uint32_t), cudaMemcpyDeviceToHost);
@ -578,6 +582,7 @@ void neoscrypt_setBlockTarget(uint32_t* pdata, const void *target)
{ {
unsigned char PaddedMessage[80*4]; //bring balance to the force unsigned char PaddedMessage[80*4]; //bring balance to the force
uint32_t input[16], key[16] = { 0 }; uint32_t input[16], key[16] = { 0 };
memcpy(PaddedMessage, pdata, 80); memcpy(PaddedMessage, pdata, 80);
memcpy(PaddedMessage + 80, pdata, 80); memcpy(PaddedMessage + 80, pdata, 80);
memcpy(PaddedMessage + 160, pdata, 80); memcpy(PaddedMessage + 160, pdata, 80);

19
neoscrypt/cuda_vectors.h

@ -478,23 +478,25 @@ static __forceinline__ __device__ uint32_t rotateR(uint32_t vec4, uint32_t shift
#if __CUDA_ARCH__ < 320 #if __CUDA_ARCH__ < 320
// right shift a 64 bytes input (256-bits integer) by 0 8 16 24 bits // right shift a 64-bytes integer (256-bits) by 0 8 16 24 bits
static __forceinline__ __device__ void shift256R(uint32_t* ret, const uint8 &vec4, uint32_t shift) // require a uint32_t[9] ret array
// note: djm neoscrypt implementation is near the limits of gpu capabilities
// and weird behaviors can happen when tuning device functions code...
__device__ void shift256R(uint32_t* ret, const uint8 &vec4, uint32_t shift)
{ {
uint8_t *v = (uint8_t*) &vec4.s0; uint8_t *v = (uint8_t*) &vec4.s0;
uint8_t *r = (uint8_t*) ret; uint8_t *r = (uint8_t*) ret;
uint8_t bytes = (uint8_t) (shift >> 3); uint8_t bytes = (uint8_t) (shift >> 3);
for (uint8_t i=0; i<bytes; i++) ret[0] = 0;
r[i] = 0;
for (uint8_t i=bytes; i<32; i++) for (uint8_t i=bytes; i<32; i++)
r[i] = v[i-bytes]; r[i] = v[i-bytes];
ret[8] = vec4.s7 >> (32 - shift); // shuffled part required ? ret[8] = vec4.s7 >> (32 - shift); // shuffled part required
//printf("A %02u %08x %08x > %08x %08x\n", shift, vec4.s6, vec4.s7, ret[7], ret[8]);
} }
#else #else
// right shift a 32 bytes input (256-bits integer) by 0 8 16 24 bits // same for SM 3.5+, really faster ?
static __forceinline__ __device__ void shift256R(uint32_t* ret, const uint8 &vec4, uint32_t shift) __device__ void shift256R(uint32_t* ret, const uint8 &vec4, uint32_t shift)
{ {
uint32_t truc = 0, truc2 = cuda_swab32(vec4.s7), truc3 = 0; uint32_t truc = 0, truc2 = cuda_swab32(vec4.s7), truc3 = 0;
asm("shf.r.clamp.b32 %0, %1, %2, %3;" : "=r"(truc) : "r"(truc3), "r"(truc2), "r"(shift)); asm("shf.r.clamp.b32 %0, %1, %2, %3;" : "=r"(truc) : "r"(truc3), "r"(truc2), "r"(shift));
@ -522,7 +524,6 @@ static __forceinline__ __device__ void shift256R(uint32_t* ret, const uint8 &vec
ret[1] = cuda_swab32(truc); ret[1] = cuda_swab32(truc);
asm("shr.b32 %0, %1, %2;" : "=r"(truc) : "r"(truc3), "r"(shift)); asm("shr.b32 %0, %1, %2;" : "=r"(truc) : "r"(truc3), "r"(shift));
ret[0] = cuda_swab32(truc); ret[0] = cuda_swab32(truc);
//printf("B %02u %08x %08x > %08x %08x\n", shift, vec4.s6, vec4.s7, ret[7], ret[8]);
} }
#endif #endif

30
neoscrypt/neoscrypt.cpp

@ -2,13 +2,9 @@
#include "miner.h" #include "miner.h"
#include "neoscrypt/neoscrypt.h" #include "neoscrypt/neoscrypt.h"
static uint32_t *d_hash[MAX_GPUS] ;
extern void neoscrypt_setBlockTarget(uint32_t * data, const void *ptarget); extern void neoscrypt_setBlockTarget(uint32_t * data, const void *ptarget);
extern void neoscrypt_cpu_init(int thr_id, uint32_t threads, uint32_t* hash); extern void neoscrypt_cpu_init(int thr_id, uint32_t threads);
extern uint32_t neoscrypt_cpu_hash_k4(int stratum, int thr_id, uint32_t threads, uint32_t startNounce, int order); extern uint32_t neoscrypt_cpu_hash_k4(int thr_id, uint32_t threads, uint32_t startNounce, bool have_stratum, int order);
extern int cuda_get_arch(int thr_id);
#define SHIFT 130
static bool init[MAX_GPUS] = { 0 }; static bool init[MAX_GPUS] = { 0 };
@ -16,43 +12,35 @@ int scanhash_neoscrypt(int thr_id, uint32_t *pdata, const uint32_t *ptarget, uin
{ {
const uint32_t first_nonce = pdata[19]; const uint32_t first_nonce = pdata[19];
if (opt_benchmark)
((uint32_t*)ptarget)[7] = 0x0000ff;
int intensity = is_windows() ? 18 : 19; int intensity = is_windows() ? 18 : 19;
uint32_t throughput = device_intensity(thr_id, __func__, 1U << intensity); uint32_t throughput = device_intensity(thr_id, __func__, 1U << intensity);
throughput = throughput / 32; /* set for max intensity ~= 20 */ throughput = throughput / 32; /* set for max intensity ~= 20 */
throughput = min(throughput, max_nonce - first_nonce + 1); throughput = min(throughput, max_nonce - first_nonce + 1);
if (opt_benchmark)
((uint32_t*)ptarget)[7] = 0x0000ff;
if (!init[thr_id]) if (!init[thr_id])
{ {
int dev_id = device_map[thr_id]; int dev_id = device_map[thr_id];
cudaSetDevice(dev_id); cudaSetDevice(dev_id);
cudaDeviceSetCacheConfig(cudaFuncCachePreferL1); cudaDeviceSetCacheConfig(cudaFuncCachePreferL1);
cuda_get_arch(thr_id);
if (device_sm[dev_id] <= 300) { if (device_sm[dev_id] <= 300) {
applog(LOG_ERR, "Sorry neoscrypt is not supported on SM 3.0 devices"); applog(LOG_ERR, "Sorry neoscrypt is not supported on SM 3.0 devices");
proper_exit(EXIT_CODE_CUDA_ERROR); proper_exit(EXIT_CODE_CUDA_ERROR);
} }
cudaMalloc(&d_hash[thr_id], 32 * SHIFT * sizeof(uint64_t) * throughput);
neoscrypt_cpu_init(thr_id, throughput, d_hash[thr_id]);
applog(LOG_INFO, "Using %d cuda threads", throughput); applog(LOG_INFO, "Using %d cuda threads", throughput);
if (cudaGetLastError() != cudaSuccess) { neoscrypt_cpu_init(thr_id, throughput);
cudaError_t err = cudaGetLastError();
fprintf(stderr, "Cuda error in func '%s' at line %i : %s.\n",
__FUNCTION__, __LINE__, cudaGetErrorString(err) );
proper_exit(EXIT_FAILURE);
}
init[thr_id] = true; init[thr_id] = true;
} }
uint32_t endiandata[20]; uint32_t endiandata[20];
if (have_stratum) { if (have_stratum) {
for (int k = 0; k < 20; k++) for (int k = 0; k < 20; k++)
be32enc(&endiandata[k], ((uint32_t*)pdata)[k]); be32enc(&endiandata[k], pdata[k]);
} else { } else {
for (int k = 0; k < 20; k++) for (int k = 0; k < 20; k++)
endiandata[k] = pdata[k]; endiandata[k] = pdata[k];
@ -61,7 +49,7 @@ int scanhash_neoscrypt(int thr_id, uint32_t *pdata, const uint32_t *ptarget, uin
neoscrypt_setBlockTarget(endiandata,ptarget); neoscrypt_setBlockTarget(endiandata,ptarget);
do { do {
uint32_t foundNonce = neoscrypt_cpu_hash_k4((int)have_stratum, thr_id, throughput, pdata[19], 0); uint32_t foundNonce = neoscrypt_cpu_hash_k4(thr_id, throughput, pdata[19], have_stratum, 0);
if (foundNonce != UINT32_MAX) if (foundNonce != UINT32_MAX)
{ {
uint32_t _ALIGN(64) vhash64[8]; uint32_t _ALIGN(64) vhash64[8];

Loading…
Cancel
Save