Browse Source

lyra2v2: increase default intensity

to be able to say, like sp, that its faster :p
master
Tanguy Pruvot 9 years ago
parent
commit
87edf84bf3
  1. 4
      lyra2/cuda_lyra2v2.cu
  2. 26
      lyra2/lyra2REv2.cu

4
lyra2/cuda_lyra2v2.cu

@ -350,10 +350,10 @@ __global__ void lyra2v2_gpu_hash_32(uint32_t threads, uint32_t startNounce, uint @@ -350,10 +350,10 @@ __global__ void lyra2v2_gpu_hash_32(uint32_t threads, uint32_t startNounce, uint
#endif
__host__
void lyra2v2_cpu_init(int thr_id, uint32_t threads, uint64_t *d_hash2)
void lyra2v2_cpu_init(int thr_id, uint32_t threads, uint64_t *d_matrix)
{
// just assign the device pointer allocated in main loop
cudaMemcpyToSymbol(DMatrix, &d_hash2, sizeof(uint64_t*), 0, cudaMemcpyHostToDevice);
cudaMemcpyToSymbol(DMatrix, &d_matrix, sizeof(uint64_t*), 0, cudaMemcpyHostToDevice);
}
__host__

26
lyra2/lyra2REv2.cu

@ -12,7 +12,7 @@ extern "C" { @@ -12,7 +12,7 @@ extern "C" {
static _ALIGN(64) uint64_t *d_hash[MAX_GPUS];
static uint64_t *d_hash2[MAX_GPUS];
static uint64_t* d_matrix[MAX_GPUS];
extern void blake256_cpu_init(int thr_id, uint32_t threads);
extern void blake256_cpu_hash_80(const int thr_id, const uint32_t threads, const uint32_t startNonce, uint64_t *Hash, int order);
@ -24,7 +24,7 @@ extern void skein256_cpu_init(int thr_id, uint32_t threads); @@ -24,7 +24,7 @@ extern void skein256_cpu_init(int thr_id, uint32_t threads);
extern void cubehash256_cpu_hash_32(int thr_id, uint32_t threads, uint32_t startNounce, uint64_t *d_hash, int order);
extern void lyra2v2_cpu_hash_32(int thr_id, uint32_t threads, uint32_t startNonce, uint64_t *d_outputHash, int order);
extern void lyra2v2_cpu_init(int thr_id, uint32_t threads, uint64_t* matrix);
extern void lyra2v2_cpu_init(int thr_id, uint32_t threads, uint64_t* d_matrix);
extern void bmw256_setTarget(const void *ptarget);
extern void bmw256_cpu_init(int thr_id, uint32_t threads);
@ -77,9 +77,10 @@ extern "C" int scanhash_lyra2v2(int thr_id, struct work* work, uint32_t max_nonc @@ -77,9 +77,10 @@ extern "C" int scanhash_lyra2v2(int thr_id, struct work* work, uint32_t max_nonc
uint32_t *pdata = work->data;
uint32_t *ptarget = work->target;
const uint32_t first_nonce = pdata[19];
int intensity = (device_sm[device_map[thr_id]] > 500 && !is_windows()) ? 18 : 17;
int dev_id = device_map[thr_id];
int intensity = (device_sm[dev_id] > 500 && !is_windows()) ? 20 : 18;
unsigned int defthr = 1U << intensity;
uint32_t throughput = device_intensity(device_map[thr_id], __func__, defthr);
uint32_t throughput = device_intensity(dev_id, __func__, defthr);
if (opt_benchmark)
((uint32_t*)ptarget)[7] = 0x00ff;
@ -102,10 +103,10 @@ extern "C" int scanhash_lyra2v2(int thr_id, struct work* work, uint32_t max_nonc @@ -102,10 +103,10 @@ extern "C" int scanhash_lyra2v2(int thr_id, struct work* work, uint32_t max_nonc
}
// DMatrix
CUDA_SAFE_CALL(cudaMalloc(&d_hash2[thr_id], 16 * 4 * 4 * sizeof(uint64_t) * throughput));
lyra2v2_cpu_init(thr_id, throughput, d_hash2[thr_id]);
CUDA_SAFE_CALL(cudaMalloc(&d_matrix[thr_id], (size_t)16 * sizeof(uint64_t) * 4 * 3 * throughput));
lyra2v2_cpu_init(thr_id, throughput, d_matrix[thr_id]);
CUDA_SAFE_CALL(cudaMalloc(&d_hash[thr_id], (size_t)throughput * 32));
CUDA_SAFE_CALL(cudaMalloc(&d_hash[thr_id], (size_t)32 * throughput));
init[thr_id] = true;
}
@ -130,18 +131,18 @@ extern "C" int scanhash_lyra2v2(int thr_id, struct work* work, uint32_t max_nonc @@ -130,18 +131,18 @@ extern "C" int scanhash_lyra2v2(int thr_id, struct work* work, uint32_t max_nonc
bmw256_cpu_hash_32(thr_id, throughput, pdata[19], d_hash[thr_id], foundNonces);
*hashes_done = pdata[19] - first_nonce + throughput;
if (foundNonces[0] != 0)
{
const uint32_t Htarg = ptarget[7];
uint32_t vhash64[8];
be32enc(&endiandata[19], foundNonces[0]);
lyra2v2_hash(vhash64, endiandata);
if (vhash64[7] <= Htarg && fulltest(vhash64, ptarget))
if (vhash64[7] <= ptarget[7] && fulltest(vhash64, ptarget))
{
int res = 1;
bn_store_hash_target_ratio(vhash64, ptarget, work);
// check if there was another one...
*hashes_done = pdata[19] - first_nonce + throughput;
if (foundNonces[1] != 0)
{
be32enc(&endiandata[19], foundNonces[1]);
@ -157,8 +158,7 @@ extern "C" int scanhash_lyra2v2(int thr_id, struct work* work, uint32_t max_nonc @@ -157,8 +158,7 @@ extern "C" int scanhash_lyra2v2(int thr_id, struct work* work, uint32_t max_nonc
}
else
{
if (vhash64[7] > Htarg) // don't show message if it is equal but fails fulltest
applog(LOG_WARNING, "GPU #%d: result does not validate on CPU!", thr_id);
applog(LOG_WARNING, "GPU #%d: result does not validate on CPU!", dev_id);
}
}
@ -180,7 +180,7 @@ extern "C" void free_lyra2v2(int thr_id) @@ -180,7 +180,7 @@ extern "C" void free_lyra2v2(int thr_id)
cudaSetDevice(device_map[thr_id]);
cudaFree(d_hash[thr_id]);
cudaFree(d_hash2[thr_id]);
cudaFree(d_matrix[thr_id]);
bmw256_cpu_free(thr_id);
init[thr_id] = false;

Loading…
Cancel
Save