mirror of
https://github.com/GOSTSec/ccminer
synced 2025-01-09 14:28:15 +00:00
5bf1f98200
X11+ algos and quark are not compatible for the moment
but these ones are :
Benchmark results for Gigabyte GTX 460 (SM 2.1 / 1 GB):
blakecoin : 159090.5 kH/s, 1 MB, 1048576 thr.
blake : 70208.9 kH/s, 1 MB, 1048576 thr.
bmw : 122802.6 kH/s, 65 MB, 2097152 thr.
deep : 3533.6 kH/s, 33 MB, 524288 thr.
fugue256 : 43177.9 kH/s, 17 MB, 524288 thr.
heavy : 4118.2 kH/s, 147 MB, 524032 thr.
keccak : 18673.1 kH/s, 129 MB, 2097152 thr.
luffa : 28816.0 kH/s, 257 MB, 4194304 thr.
lyra2 : 213.7 kH/s, 570 MB, 65536 thr.
mjollnir : 3895.6 kH/s, 147 MB, 524032 thr.
nist5 : 1101.4 kH/s, 67 MB, 1048576 thr.
penta : 501.6 kH/s, 21 MB, 327680 thr.
skein : 5432.4 kH/s, 65 MB, 1048576 thr.
skein2 : 6788.9 kH/s, 33 MB, 524288 thr.
whirlpool : 688.5 kH/s, 33 MB, 524288 thr.
zr5 : 122.5 kH/s, 86 MB, 262144 thr.
281 lines
10 KiB
Plaintext
281 lines
10 KiB
Plaintext
extern "C"
|
|
{
|
|
#include "sph/sph_blake.h"
|
|
#include "sph/sph_bmw.h"
|
|
#include "sph/sph_groestl.h"
|
|
#include "sph/sph_skein.h"
|
|
#include "sph/sph_jh.h"
|
|
#include "sph/sph_keccak.h"
|
|
}
|
|
|
|
#include "miner.h"
|
|
|
|
#include "cuda_helper.h"
|
|
|
|
static uint32_t *d_hash[MAX_GPUS];
|
|
|
|
// Speicher zur Generierung der Noncevektoren für die bedingten Hashes
|
|
static uint32_t *d_branch1Nonces[MAX_GPUS];
|
|
static uint32_t *d_branch2Nonces[MAX_GPUS];
|
|
static uint32_t *d_branch3Nonces[MAX_GPUS];
|
|
|
|
extern void quark_blake512_cpu_init(int thr_id, uint32_t threads);
|
|
extern void quark_blake512_cpu_setBlock_80(int thr_id, uint32_t *pdata);
|
|
extern void quark_blake512_cpu_hash_80(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_hash);
|
|
extern void quark_blake512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order);
|
|
|
|
extern void quark_bmw512_cpu_init(int thr_id, uint32_t threads);
|
|
extern void quark_bmw512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order);
|
|
|
|
extern void quark_groestl512_cpu_init(int thr_id, uint32_t threads);
|
|
extern void quark_groestl512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order);
|
|
extern void quark_doublegroestl512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order);
|
|
extern void quark_groestl512_cpu_free(int thr_id);
|
|
|
|
extern void quark_skein512_cpu_init(int thr_id, uint32_t threads);
|
|
extern void quark_skein512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order);
|
|
|
|
extern void quark_keccak512_cpu_init(int thr_id, uint32_t threads);
|
|
extern void quark_keccak512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order);
|
|
|
|
extern void quark_jh512_cpu_init(int thr_id, uint32_t threads);
|
|
extern void quark_jh512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash, int order);
|
|
|
|
extern void quark_compactTest_cpu_init(int thr_id, uint32_t threads);
|
|
extern void quark_compactTest_cpu_free(int thr_id);
|
|
extern void quark_compactTest_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *inpHashes, uint32_t *d_validNonceTable,
|
|
uint32_t *d_nonces1, uint32_t *nrm1,
|
|
uint32_t *d_nonces2, uint32_t *nrm2,
|
|
int order);
|
|
extern void quark_compactTest_single_false_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *inpHashes, uint32_t *d_validNonceTable,
|
|
uint32_t *d_nonces1, uint32_t *nrm1,
|
|
int order);
|
|
|
|
extern uint32_t cuda_check_hash_branch(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_inputHash, int order);
|
|
|
|
|
|
// Original Quarkhash Funktion aus einem miner Quelltext
|
|
extern "C" void quarkhash(void *state, const void *input)
|
|
{
|
|
sph_blake512_context ctx_blake;
|
|
sph_bmw512_context ctx_bmw;
|
|
sph_groestl512_context ctx_groestl;
|
|
sph_jh512_context ctx_jh;
|
|
sph_keccak512_context ctx_keccak;
|
|
sph_skein512_context ctx_skein;
|
|
|
|
unsigned char hash[64];
|
|
|
|
sph_blake512_init(&ctx_blake);
|
|
sph_blake512 (&ctx_blake, input, 80);
|
|
sph_blake512_close(&ctx_blake, (void*) hash);
|
|
|
|
sph_bmw512_init(&ctx_bmw);
|
|
sph_bmw512 (&ctx_bmw, (const void*) hash, 64);
|
|
sph_bmw512_close(&ctx_bmw, (void*) hash);
|
|
|
|
if (hash[0] & 0x8)
|
|
{
|
|
sph_groestl512_init(&ctx_groestl);
|
|
sph_groestl512 (&ctx_groestl, (const void*) hash, 64);
|
|
sph_groestl512_close(&ctx_groestl, (void*) hash);
|
|
}
|
|
else
|
|
{
|
|
sph_skein512_init(&ctx_skein);
|
|
sph_skein512 (&ctx_skein, (const void*) hash, 64);
|
|
sph_skein512_close(&ctx_skein, (void*) hash);
|
|
}
|
|
|
|
sph_groestl512_init(&ctx_groestl);
|
|
sph_groestl512 (&ctx_groestl, (const void*) hash, 64);
|
|
sph_groestl512_close(&ctx_groestl, (void*) hash);
|
|
|
|
sph_jh512_init(&ctx_jh);
|
|
sph_jh512 (&ctx_jh, (const void*) hash, 64);
|
|
sph_jh512_close(&ctx_jh, (void*) hash);
|
|
|
|
if (hash[0] & 0x8)
|
|
{
|
|
sph_blake512_init(&ctx_blake);
|
|
sph_blake512 (&ctx_blake, (const void*) hash, 64);
|
|
sph_blake512_close(&ctx_blake, (void*) hash);
|
|
}
|
|
else
|
|
{
|
|
sph_bmw512_init(&ctx_bmw);
|
|
sph_bmw512 (&ctx_bmw, (const void*) hash, 64);
|
|
sph_bmw512_close(&ctx_bmw, (void*) hash);
|
|
}
|
|
|
|
sph_keccak512_init(&ctx_keccak);
|
|
sph_keccak512 (&ctx_keccak, (const void*) hash, 64);
|
|
sph_keccak512_close(&ctx_keccak, (void*) hash);
|
|
|
|
sph_skein512_init(&ctx_skein);
|
|
sph_skein512 (&ctx_skein, (const void*) hash, 64);
|
|
sph_skein512_close(&ctx_skein, (void*) hash);
|
|
|
|
if (hash[0] & 0x8)
|
|
{
|
|
sph_keccak512_init(&ctx_keccak);
|
|
sph_keccak512 (&ctx_keccak, (const void*) hash, 64);
|
|
sph_keccak512_close(&ctx_keccak, (void*) hash);
|
|
}
|
|
else
|
|
{
|
|
sph_jh512_init(&ctx_jh);
|
|
sph_jh512 (&ctx_jh, (const void*) hash, 64);
|
|
sph_jh512_close(&ctx_jh, (void*) hash);
|
|
}
|
|
|
|
memcpy(state, hash, 32);
|
|
}
|
|
|
|
static bool init[MAX_GPUS] = { 0 };
|
|
|
|
extern "C" int scanhash_quark(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done)
|
|
{
|
|
uint32_t _ALIGN(64) endiandata[20];
|
|
uint32_t *pdata = work->data;
|
|
uint32_t *ptarget = work->target;
|
|
const uint32_t first_nonce = pdata[19];
|
|
|
|
uint32_t throughput = cuda_default_throughput(thr_id, 1 << 20); // 256*4096
|
|
if (init[thr_id]) throughput = min(throughput, max_nonce - first_nonce);
|
|
|
|
if (opt_benchmark)
|
|
((uint32_t*)ptarget)[7] = 0x00F;
|
|
|
|
if (!init[thr_id])
|
|
{
|
|
cudaSetDevice(device_map[thr_id]);
|
|
|
|
// Konstanten kopieren, Speicher belegen
|
|
CUDA_SAFE_CALL(cudaMalloc(&d_hash[thr_id], 16 * sizeof(uint32_t) * throughput));
|
|
|
|
quark_blake512_cpu_init(thr_id, throughput);
|
|
quark_groestl512_cpu_init(thr_id, throughput);
|
|
quark_skein512_cpu_init(thr_id, throughput);
|
|
quark_bmw512_cpu_init(thr_id, throughput);
|
|
quark_keccak512_cpu_init(thr_id, throughput);
|
|
quark_jh512_cpu_init(thr_id, throughput);
|
|
cuda_check_cpu_init(thr_id, throughput);
|
|
quark_compactTest_cpu_init(thr_id, throughput);
|
|
|
|
cudaMalloc(&d_branch1Nonces[thr_id], sizeof(uint32_t)*throughput);
|
|
cudaMalloc(&d_branch2Nonces[thr_id], sizeof(uint32_t)*throughput);
|
|
cudaMalloc(&d_branch3Nonces[thr_id], sizeof(uint32_t)*throughput);
|
|
|
|
init[thr_id] = true;
|
|
}
|
|
|
|
for (int k=0; k < 20; k++)
|
|
be32enc(&endiandata[k], pdata[k]);
|
|
|
|
quark_blake512_cpu_setBlock_80(thr_id, endiandata);
|
|
cuda_check_cpu_setTarget(ptarget);
|
|
|
|
do {
|
|
int order = 0;
|
|
uint32_t nrm1=0, nrm2=0, nrm3=0;
|
|
|
|
// erstes Blake512 Hash mit CUDA
|
|
quark_blake512_cpu_hash_80(thr_id, throughput, pdata[19], d_hash[thr_id]); order++;
|
|
|
|
// das ist der unbedingte Branch für BMW512
|
|
quark_bmw512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++);
|
|
|
|
quark_compactTest_single_false_cpu_hash_64(thr_id, throughput, pdata[19], d_hash[thr_id], NULL,
|
|
d_branch3Nonces[thr_id], &nrm3,
|
|
order++);
|
|
|
|
// nur den Skein Branch weiterverfolgen
|
|
quark_skein512_cpu_hash_64(thr_id, nrm3, pdata[19], d_branch3Nonces[thr_id], d_hash[thr_id], order++);
|
|
|
|
// das ist der unbedingte Branch für Groestl512
|
|
quark_groestl512_cpu_hash_64(thr_id, nrm3, pdata[19], d_branch3Nonces[thr_id], d_hash[thr_id], order++);
|
|
|
|
// das ist der unbedingte Branch für JH512
|
|
quark_jh512_cpu_hash_64(thr_id, nrm3, pdata[19], d_branch3Nonces[thr_id], d_hash[thr_id], order++);
|
|
|
|
// quarkNonces in branch1 und branch2 aufsplitten gemäss if (hash[0] & 0x8)
|
|
quark_compactTest_cpu_hash_64(thr_id, nrm3, pdata[19], d_hash[thr_id], d_branch3Nonces[thr_id],
|
|
d_branch1Nonces[thr_id], &nrm1,
|
|
d_branch2Nonces[thr_id], &nrm2,
|
|
order++);
|
|
|
|
// das ist der bedingte Branch für Blake512
|
|
quark_blake512_cpu_hash_64(thr_id, nrm1, pdata[19], d_branch1Nonces[thr_id], d_hash[thr_id], order++);
|
|
|
|
// das ist der bedingte Branch für Bmw512
|
|
quark_bmw512_cpu_hash_64(thr_id, nrm2, pdata[19], d_branch2Nonces[thr_id], d_hash[thr_id], order++);
|
|
|
|
// das ist der unbedingte Branch für Keccak512
|
|
quark_keccak512_cpu_hash_64(thr_id, nrm3, pdata[19], d_branch3Nonces[thr_id], d_hash[thr_id], order++);
|
|
|
|
// das ist der unbedingte Branch für Skein512
|
|
quark_skein512_cpu_hash_64(thr_id, nrm3, pdata[19], d_branch3Nonces[thr_id], d_hash[thr_id], order++);
|
|
|
|
// quarkNonces in branch1 und branch2 aufsplitten gemäss if (hash[0] & 0x8)
|
|
quark_compactTest_cpu_hash_64(thr_id, nrm3, pdata[19], d_hash[thr_id], d_branch3Nonces[thr_id],
|
|
d_branch1Nonces[thr_id], &nrm1,
|
|
d_branch2Nonces[thr_id], &nrm2,
|
|
order++);
|
|
|
|
// das ist der bedingte Branch für Keccak512
|
|
quark_keccak512_cpu_hash_64(thr_id, nrm1, pdata[19], d_branch1Nonces[thr_id], d_hash[thr_id], order++);
|
|
|
|
// das ist der bedingte Branch für JH512
|
|
quark_jh512_cpu_hash_64(thr_id, nrm2, pdata[19], d_branch2Nonces[thr_id], d_hash[thr_id], order++);
|
|
|
|
*hashes_done = pdata[19] - first_nonce + 1;
|
|
|
|
// Scan nach Gewinner Hashes auf der GPU
|
|
uint32_t foundNonce = cuda_check_hash_branch(thr_id, nrm3, pdata[19], d_branch3Nonces[thr_id], d_hash[thr_id], order++);
|
|
if (foundNonce != 0xffffffff)
|
|
{
|
|
uint32_t vhash64[8];
|
|
be32enc(&endiandata[19], foundNonce);
|
|
quarkhash(vhash64, endiandata);
|
|
|
|
if (vhash64[7] <= ptarget[7] && fulltest(vhash64, ptarget)) {
|
|
work_set_target_ratio(work, vhash64);
|
|
pdata[19] = foundNonce;
|
|
return 1;
|
|
} else {
|
|
applog(LOG_WARNING, "GPU #%d: result for nonce %08x does not validate on CPU!", device_map[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;
|
|
}
|
|
|
|
// cleanup
|
|
extern "C" void free_quark(int thr_id)
|
|
{
|
|
if (!init[thr_id])
|
|
return;
|
|
|
|
cudaThreadSynchronize();
|
|
|
|
cudaFree(d_hash[thr_id]);
|
|
|
|
cudaFree(d_branch1Nonces[thr_id]);
|
|
cudaFree(d_branch2Nonces[thr_id]);
|
|
cudaFree(d_branch3Nonces[thr_id]);
|
|
|
|
quark_groestl512_cpu_free(thr_id);
|
|
quark_compactTest_cpu_free(thr_id);
|
|
|
|
cuda_check_cpu_free(thr_id);
|
|
init[thr_id] = false;
|
|
|
|
cudaDeviceSynchronize();
|
|
}
|