-
Notifications
You must be signed in to change notification settings - Fork 8
/
Copy pathanimecoin.cu
291 lines (238 loc) · 10.7 KB
/
animecoin.cu
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
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
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_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash);
extern void quark_bmw512_cpu_init(int thr_id, uint32_t threads);
extern void quark_bmw512_cpu_setBlock_80(int thr_id, void *pdata);
extern void quark_bmw512_cpu_hash_80(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_outputHash);
extern void quark_bmw512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_outputHash);
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);
extern void quark_doublegroestl512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash);
extern void quark_skein512_cpu_init(int thr_id);
extern void quark_skein512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash);
extern void quark_keccak512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash);
extern void quark_keccak512_cpu_hash_64_final(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash);
extern void quark_jh512_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash);
extern void quark_jh512_cpu_hash_64_final(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_hash);
extern void quark_compactTest_cpu_init(int thr_id, uint32_t threads);
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);
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);
extern uint32_t cuda_check_hash_branch(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_inputHash);
extern void cuda_check_quarkcoin(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_nonceVector, uint32_t *d_inputHash, uint32_t *foundnonces);
/* CPU Hash */
extern "C" void animehash(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_bmw512_init(&ctx_bmw);
// ZBMW;
sph_bmw512 (&ctx_bmw, (const void*) input, 80);
sph_bmw512_close(&ctx_bmw, (void*) hash);
sph_blake512_init(&ctx_blake);
// ZBLAKE;
sph_blake512 (&ctx_blake, hash, 64);
sph_blake512_close(&ctx_blake, (void*) hash);
if (hash[0] & 0x8)
{
sph_groestl512_init(&ctx_groestl);
// ZGROESTL;
sph_groestl512 (&ctx_groestl, (const void*) hash, 64);
sph_groestl512_close(&ctx_groestl, (void*) hash);
}
else
{
sph_skein512_init(&ctx_skein);
// ZSKEIN;
sph_skein512 (&ctx_skein, (const void*) hash, 64);
sph_skein512_close(&ctx_skein, (void*) hash);
}
sph_groestl512_init(&ctx_groestl);
// ZGROESTL;
sph_groestl512 (&ctx_groestl, (const void*) hash, 64);
sph_groestl512_close(&ctx_groestl, (void*) hash);
sph_jh512_init(&ctx_jh);
// ZJH;
sph_jh512 (&ctx_jh, (const void*) hash, 64);
sph_jh512_close(&ctx_jh, (void*) hash);
if (hash[0] & 0x8)
{
sph_blake512_init(&ctx_blake);
// ZBLAKE;
sph_blake512 (&ctx_blake, (const void*) hash, 64);
sph_blake512_close(&ctx_blake, (void*) hash);
}
else
{
sph_bmw512_init(&ctx_bmw);
// ZBMW;
sph_bmw512 (&ctx_bmw, (const void*) hash, 64);
sph_bmw512_close(&ctx_bmw, (void*) hash);
}
sph_keccak512_init(&ctx_keccak);
// ZKECCAK;
sph_keccak512 (&ctx_keccak, (const void*) hash, 64);
sph_keccak512_close(&ctx_keccak, (void*) hash);
sph_skein512_init(&ctx_skein);
// 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);
// ZKECCAK;
sph_keccak512 (&ctx_keccak, (const void*) hash, 64);
sph_keccak512_close(&ctx_keccak, (void*) hash);
}
else
{
sph_jh512_init(&ctx_jh);
// ZJH;
sph_jh512 (&ctx_jh, (const void*) hash, 64);
sph_jh512_close(&ctx_jh, (void*) hash);
}
memcpy(state, hash, 32);
}
/*
struct HashPredicate
{
HashPredicate(uint32_t *hashes, uint32_t startNonce) :
m_hashes(hashes),
m_startNonce(startNonce)
{ }
__device__
bool operator()(const uint32_t x)
{
uint32_t *const Hash = &m_hashes[(x - m_startNonce)*16];
return hash[0] & 0x8;
}
uint32_t *m_hashes;
uint32_t m_startNonce;
};
*/
static volatile bool init[MAX_GPUS] = { false };
extern int scanhash_anime(int thr_id, uint32_t *pdata,
uint32_t *ptarget, uint32_t max_nonce,
uint32_t *hashes_done)
{
const uint32_t first_nonce = pdata[19];
uint32_t throughput = device_intensity(device_map[thr_id], __func__, 1 << 20); // 256*256*8
throughput = min(throughput, (max_nonce - first_nonce)) & 0xfffffc00;
if (opt_benchmark)
ptarget[7] = 0x00ff;
if (!init[thr_id])
{
CUDA_SAFE_CALL(cudaSetDevice(device_map[thr_id]));
cudaSetDeviceFlags(cudaDeviceScheduleBlockingSync);
cudaDeviceSetCacheConfig(cudaFuncCachePreferL1);
CUDA_SAFE_CALL(cudaStreamCreate(&gpustream[thr_id]));
get_cuda_arch(&cuda_arch[thr_id]);
CUDA_SAFE_CALL(cudaMalloc(&d_hash[thr_id], 16 * sizeof(uint32_t) * throughput));
quark_groestl512_cpu_init(thr_id, throughput);
quark_skein512_cpu_init(thr_id);
quark_bmw512_cpu_init(thr_id, throughput);
cuda_check_cpu_init(thr_id, throughput);
quark_compactTest_cpu_init(thr_id, throughput);
CUDA_SAFE_CALL(cudaMalloc(&d_branch1Nonces[thr_id], sizeof(uint32_t)*throughput));
CUDA_SAFE_CALL(cudaMalloc(&d_branch2Nonces[thr_id], sizeof(uint32_t)*throughput));
CUDA_SAFE_CALL(cudaMalloc(&d_branch3Nonces[thr_id], sizeof(uint32_t)*throughput));
init[thr_id] = true;
}
uint32_t endiandata[20];
for (int k=0; k < 20; k++)
be32enc(&endiandata[k], pdata[k]);
quark_bmw512_cpu_setBlock_80(thr_id, (void*)endiandata);
cuda_check_cpu_setTarget(ptarget, thr_id);
do {
uint32_t nrm1=0, nrm2=0, nrm3=0;
// erstes BMW512 Hash mit CUDA
quark_bmw512_cpu_hash_80(thr_id, throughput, pdata[19], d_hash[thr_id]);
// das ist der unbedingte Branch für Blake512
quark_blake512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id]);
quark_compactTest_single_false_cpu_hash_64(thr_id, throughput, pdata[19], d_hash[thr_id], NULL,
d_branch3Nonces[thr_id], &nrm3);
// nur den Skein Branch weiterverfolgen
quark_skein512_cpu_hash_64(thr_id, nrm3, pdata[19], d_branch3Nonces[thr_id], d_hash[thr_id]);
// 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]);
// 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]);
// 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);
// 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]);
// 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]);
// 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]);
// 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]);
// 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);
quark_keccak512_cpu_hash_64_final(thr_id, nrm1, pdata[19], d_branch1Nonces[thr_id], d_hash[thr_id]);
quark_jh512_cpu_hash_64_final(thr_id, nrm2, pdata[19], d_branch2Nonces[thr_id], d_hash[thr_id]);
uint32_t foundnonces[2];
cuda_check_quarkcoin(thr_id, nrm3, pdata[19], d_branch3Nonces[thr_id], d_hash[thr_id], foundnonces);
if(stop_mining) {mining_has_stopped[thr_id] = true; cudaStreamDestroy(gpustream[thr_id]); pthread_exit(nullptr);}
if(foundnonces[0] != 0xffffffff)
{
const uint32_t Htarg = ptarget[7];
uint32_t vhash64[8];
be32enc(&endiandata[19], foundnonces[0]);
animehash(vhash64, endiandata);
if (vhash64[7] <= Htarg && fulltest(vhash64, ptarget))
{
int res = 1;
*hashes_done = pdata[19] - first_nonce + throughput;
// check if there was some other ones...
if (foundnonces[1] != 0xffffffff)
{
be32enc(&endiandata[19], foundnonces[1]);
animehash(vhash64, endiandata);
if (vhash64[7] <= Htarg && fulltest(vhash64, ptarget))
{
pdata[21] = foundnonces[1];
res++;
if (opt_benchmark) applog(LOG_INFO, "GPU #%d: Found second nonce $%08X", device_map[thr_id], foundnonces[1]);
}
}
pdata[19] = foundnonces[0];
if (opt_benchmark) applog(LOG_INFO, "GPU #%d: Found nonce $%08X", device_map[thr_id], foundnonces[0]);
return res;
}
else
{
if (vhash64[7] != Htarg) // don't show message if it is equal but fails fulltest
applog(LOG_INFO, "GPU #%d: result for nonce $%08X does not validate on CPU!", device_map[thr_id], foundnonces[0]);
}
}
pdata[19] += throughput; CUDA_SAFE_CALL(cudaGetLastError());
} while (!work_restart[thr_id].restart && ((uint64_t)max_nonce > ((uint64_t)(pdata[19]) + (uint64_t)throughput)));
*hashes_done = pdata[19] - first_nonce + 1;
return 0;
}