-
Notifications
You must be signed in to change notification settings - Fork 0
/
myriadgroestl.cpp
117 lines (91 loc) · 2.9 KB
/
myriadgroestl.cpp
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
#include <string.h>
#include <stdint.h>
#include <cuda_runtime.h>
#include <openssl/sha.h>
#include "sph/sph_groestl.h"
#include "miner.h"
void myriadgroestl_cpu_init(int thr_id, uint32_t threads);
void myriadgroestl_cpu_free(int thr_id);
void myriadgroestl_cpu_setBlock(int thr_id, void *data, void *pTargetIn);
void myriadgroestl_cpu_hash(int thr_id, uint32_t threads, uint32_t startNounce, void *outputHashes, uint32_t *nounce);
void myriadhash(void *state, const void *input)
{
uint32_t _ALIGN(64) hash[16];
sph_groestl512_context ctx_groestl;
SHA256_CTX sha256;
sph_groestl512_init(&ctx_groestl);
sph_groestl512(&ctx_groestl, input, 80);
sph_groestl512_close(&ctx_groestl, hash);
SHA256_Init(&sha256);
SHA256_Update(&sha256,(unsigned char *)hash, 64);
SHA256_Final((unsigned char *)hash, &sha256);
memcpy(state, hash, 32);
}
static bool init[MAX_GPUS] = { 0 };
int scanhash_myriad(int thr_id, struct work *work, uint32_t max_nonce, unsigned long *hashes_done)
{
uint32_t _ALIGN(64) endiandata[32];
uint32_t *pdata = work->data;
uint32_t *ptarget = work->target;
uint32_t start_nonce = pdata[19];
uint32_t throughput = cuda_default_throughput(thr_id, 1U << 17);
if (init[thr_id]) throughput = min(throughput, max_nonce - start_nonce);
uint32_t *outputHash = (uint32_t*)malloc(throughput * 64);
if (opt_benchmark)
ptarget[7] = 0x0000ff;
// init
if(!init[thr_id])
{
cudaSetDevice(device_map[thr_id]);
if (opt_cudaschedule == -1 && gpu_threads == 1) {
cudaDeviceReset();
// reduce cpu usage
cudaSetDeviceFlags(cudaDeviceScheduleBlockingSync);
CUDA_LOG_ERROR();
}
myriadgroestl_cpu_init(thr_id, throughput);
init[thr_id] = true;
}
for (int k=0; k < 20; k++)
be32enc(&endiandata[k], pdata[k]);
// Context mit dem Endian gedrehten Blockheader vorbereiten (Nonce wird später ersetzt)
myriadgroestl_cpu_setBlock(thr_id, endiandata, (void*)ptarget);
do {
// GPU
uint32_t foundNounce = UINT32_MAX;
myriadgroestl_cpu_hash(thr_id, throughput, pdata[19], outputHash, &foundNounce);
*hashes_done = pdata[19] - start_nonce + throughput;
if (foundNounce < UINT32_MAX && bench_algo < 0)
{
uint32_t _ALIGN(64) vhash[8];
endiandata[19] = swab32(foundNounce);
myriadhash(vhash, endiandata);
if (vhash[7] <= ptarget[7] && fulltest(vhash, ptarget)) {
work_set_target_ratio(work, vhash);
pdata[19] = foundNounce;
free(outputHash);
return 1;
} else {
gpulog(LOG_WARNING, thr_id, "result for %08x does not validate on CPU!", foundNounce);
}
}
if ((uint64_t) throughput + pdata[19] >= max_nonce) {
pdata[19] = max_nonce;
break;
}
pdata[19] += throughput;
} while (!work_restart[thr_id].restart);
*hashes_done = max_nonce - start_nonce;
free(outputHash);
return 0;
}
// cleanup
void free_myriad(int thr_id)
{
if (!init[thr_id])
return;
cudaThreadSynchronize();
myriadgroestl_cpu_free(thr_id);
init[thr_id] = false;
cudaDeviceSynchronize();
}