mirror of https://github.com/GOSTSec/ccminer
Tanguy Pruvot
10 years ago
19 changed files with 977 additions and 292 deletions
@ -0,0 +1,119 @@
@@ -0,0 +1,119 @@
|
||||
/*-
|
||||
* COPYRIGHT (C) 1986 Gary S. Brown. You may use this program, or |
||||
* code or tables extracted from it, as desired without restriction. |
||||
* |
||||
* First, the polynomial itself and its table of feedback terms. The |
||||
* polynomial is |
||||
* X^32+X^26+X^23+X^22+X^16+X^12+X^11+X^10+X^8+X^7+X^5+X^4+X^2+X^1+X^0 |
||||
* |
||||
* Note that we take it "backwards" and put the highest-order term in |
||||
* the lowest-order bit. The X^32 term is "implied"; the LSB is the |
||||
* X^31 term, etc. The X^0 term (usually shown as "+1") results in |
||||
* the MSB being 1 |
||||
* |
||||
* Note that the usual hardware shift register implementation, which |
||||
* is what we're using (we're merely optimizing it by doing eight-bit |
||||
* chunks at a time) shifts bits into the lowest-order term. In our |
||||
* implementation, that means shifting towards the right. Why do we |
||||
* do it this way? Because the calculated CRC must be transmitted in |
||||
* order from highest-order term to lowest-order term. UARTs transmit |
||||
* characters in order from LSB to MSB. By storing the CRC this way |
||||
* we hand it to the UART in the order low-byte to high-byte; the UART |
||||
* sends each low-bit to hight-bit; and the result is transmission bit |
||||
* by bit from highest- to lowest-order term without requiring any bit |
||||
* shuffling on our part. Reception works similarly |
||||
* |
||||
* The feedback terms table consists of 256, 32-bit entries. Notes |
||||
* |
||||
* The table can be generated at runtime if desired; code to do so |
||||
* is shown later. It might not be obvious, but the feedback |
||||
* terms simply represent the results of eight shift/xor opera |
||||
* tions for all combinations of data and CRC register values |
||||
* |
||||
* The values must be right-shifted by eight bits by the "updcrc |
||||
* logic; the shift must be unsigned (bring in zeroes). On some |
||||
* hardware you could probably optimize the shift in assembler by |
||||
* using byte-swap instructions |
||||
* polynomial $edb88320 |
||||
* |
||||
* |
||||
* CRC32 code derived from work by Gary S. Brown. |
||||
*/ |
||||
|
||||
#include <stdlib.h> |
||||
#include <stdint.h> |
||||
|
||||
static uint32_t crc32_tab[] = { |
||||
0x00000000, 0x77073096, 0xee0e612c, 0x990951ba, 0x076dc419, 0x706af48f, |
||||
0xe963a535, 0x9e6495a3, 0x0edb8832, 0x79dcb8a4, 0xe0d5e91e, 0x97d2d988, |
||||
0x09b64c2b, 0x7eb17cbd, 0xe7b82d07, 0x90bf1d91, 0x1db71064, 0x6ab020f2, |
||||
0xf3b97148, 0x84be41de, 0x1adad47d, 0x6ddde4eb, 0xf4d4b551, 0x83d385c7, |
||||
0x136c9856, 0x646ba8c0, 0xfd62f97a, 0x8a65c9ec, 0x14015c4f, 0x63066cd9, |
||||
0xfa0f3d63, 0x8d080df5, 0x3b6e20c8, 0x4c69105e, 0xd56041e4, 0xa2677172, |
||||
0x3c03e4d1, 0x4b04d447, 0xd20d85fd, 0xa50ab56b, 0x35b5a8fa, 0x42b2986c, |
||||
0xdbbbc9d6, 0xacbcf940, 0x32d86ce3, 0x45df5c75, 0xdcd60dcf, 0xabd13d59, |
||||
0x26d930ac, 0x51de003a, 0xc8d75180, 0xbfd06116, 0x21b4f4b5, 0x56b3c423, |
||||
0xcfba9599, 0xb8bda50f, 0x2802b89e, 0x5f058808, 0xc60cd9b2, 0xb10be924, |
||||
0x2f6f7c87, 0x58684c11, 0xc1611dab, 0xb6662d3d, 0x76dc4190, 0x01db7106, |
||||
0x98d220bc, 0xefd5102a, 0x71b18589, 0x06b6b51f, 0x9fbfe4a5, 0xe8b8d433, |
||||
0x7807c9a2, 0x0f00f934, 0x9609a88e, 0xe10e9818, 0x7f6a0dbb, 0x086d3d2d, |
||||
0x91646c97, 0xe6635c01, 0x6b6b51f4, 0x1c6c6162, 0x856530d8, 0xf262004e, |
||||
0x6c0695ed, 0x1b01a57b, 0x8208f4c1, 0xf50fc457, 0x65b0d9c6, 0x12b7e950, |
||||
0x8bbeb8ea, 0xfcb9887c, 0x62dd1ddf, 0x15da2d49, 0x8cd37cf3, 0xfbd44c65, |
||||
0x4db26158, 0x3ab551ce, 0xa3bc0074, 0xd4bb30e2, 0x4adfa541, 0x3dd895d7, |
||||
0xa4d1c46d, 0xd3d6f4fb, 0x4369e96a, 0x346ed9fc, 0xad678846, 0xda60b8d0, |
||||
0x44042d73, 0x33031de5, 0xaa0a4c5f, 0xdd0d7cc9, 0x5005713c, 0x270241aa, |
||||
0xbe0b1010, 0xc90c2086, 0x5768b525, 0x206f85b3, 0xb966d409, 0xce61e49f, |
||||
0x5edef90e, 0x29d9c998, 0xb0d09822, 0xc7d7a8b4, 0x59b33d17, 0x2eb40d81, |
||||
0xb7bd5c3b, 0xc0ba6cad, 0xedb88320, 0x9abfb3b6, 0x03b6e20c, 0x74b1d29a, |
||||
0xead54739, 0x9dd277af, 0x04db2615, 0x73dc1683, 0xe3630b12, 0x94643b84, |
||||
0x0d6d6a3e, 0x7a6a5aa8, 0xe40ecf0b, 0x9309ff9d, 0x0a00ae27, 0x7d079eb1, |
||||
0xf00f9344, 0x8708a3d2, 0x1e01f268, 0x6906c2fe, 0xf762575d, 0x806567cb, |
||||
0x196c3671, 0x6e6b06e7, 0xfed41b76, 0x89d32be0, 0x10da7a5a, 0x67dd4acc, |
||||
0xf9b9df6f, 0x8ebeeff9, 0x17b7be43, 0x60b08ed5, 0xd6d6a3e8, 0xa1d1937e, |
||||
0x38d8c2c4, 0x4fdff252, 0xd1bb67f1, 0xa6bc5767, 0x3fb506dd, 0x48b2364b, |
||||
0xd80d2bda, 0xaf0a1b4c, 0x36034af6, 0x41047a60, 0xdf60efc3, 0xa867df55, |
||||
0x316e8eef, 0x4669be79, 0xcb61b38c, 0xbc66831a, 0x256fd2a0, 0x5268e236, |
||||
0xcc0c7795, 0xbb0b4703, 0x220216b9, 0x5505262f, 0xc5ba3bbe, 0xb2bd0b28, |
||||
0x2bb45a92, 0x5cb36a04, 0xc2d7ffa7, 0xb5d0cf31, 0x2cd99e8b, 0x5bdeae1d, |
||||
0x9b64c2b0, 0xec63f226, 0x756aa39c, 0x026d930a, 0x9c0906a9, 0xeb0e363f, |
||||
0x72076785, 0x05005713, 0x95bf4a82, 0xe2b87a14, 0x7bb12bae, 0x0cb61b38, |
||||
0x92d28e9b, 0xe5d5be0d, 0x7cdcefb7, 0x0bdbdf21, 0x86d3d2d4, 0xf1d4e242, |
||||
0x68ddb3f8, 0x1fda836e, 0x81be16cd, 0xf6b9265b, 0x6fb077e1, 0x18b74777, |
||||
0x88085ae6, 0xff0f6a70, 0x66063bca, 0x11010b5c, 0x8f659eff, 0xf862ae69, |
||||
0x616bffd3, 0x166ccf45, 0xa00ae278, 0xd70dd2ee, 0x4e048354, 0x3903b3c2, |
||||
0xa7672661, 0xd06016f7, 0x4969474d, 0x3e6e77db, 0xaed16a4a, 0xd9d65adc, |
||||
0x40df0b66, 0x37d83bf0, 0xa9bcae53, 0xdebb9ec5, 0x47b2cf7f, 0x30b5ffe9, |
||||
0xbdbdf21c, 0xcabac28a, 0x53b39330, 0x24b4a3a6, 0xbad03605, 0xcdd70693, |
||||
0x54de5729, 0x23d967bf, 0xb3667a2e, 0xc4614ab8, 0x5d681b02, 0x2a6f2b94, |
||||
0xb40bbe37, 0xc30c8ea1, 0x5a05df1b, 0x2d02ef8d |
||||
}; |
||||
|
||||
/* Real CRC32 Function */ |
||||
extern uint32_t crc32(uint32_t crc, const void *buf, size_t size) |
||||
{ |
||||
const uint8_t *p; |
||||
|
||||
p = buf; |
||||
crc = crc ^ ~0U; |
||||
|
||||
while (size--) |
||||
crc = crc32_tab[(crc ^ *p++) & 0xFF] ^ (crc >> 8); |
||||
|
||||
return crc ^ ~0U; |
||||
} |
||||
|
||||
/* CRC32 Function simplified for ccminer */ |
||||
extern uint32_t crc32_u32t(const uint32_t *buf, size_t size) |
||||
{ |
||||
const uint8_t *p; |
||||
uint32_t crc = 0; |
||||
|
||||
p = (uint8_t *) buf; |
||||
crc = crc ^ ~0U; |
||||
|
||||
while (size--) |
||||
crc = crc32_tab[(crc ^ *p++) & 0xFF] ^ (crc >> 8); |
||||
|
||||
return crc ^ ~0U; |
||||
} |
||||
@ -0,0 +1,600 @@
@@ -0,0 +1,600 @@
|
||||
/** |
||||
* Penta Blake-512 Cuda Kernel (Tested on SM 5.0) |
||||
* |
||||
* Tanguy Pruvot - Aug. 2014 |
||||
*/ |
||||
|
||||
#include "miner.h" |
||||
|
||||
extern "C" { |
||||
#include "sph/sph_blake.h" |
||||
#include <stdint.h> |
||||
#include <memory.h> |
||||
} |
||||
|
||||
/* threads per block */ |
||||
#define TPB 192 |
||||
|
||||
/* hash by cpu with blake 256 */ |
||||
extern "C" void pentablakehash(void *output, const void *input) |
||||
{ |
||||
unsigned char hash[128]; |
||||
#define hashB hash + 64 |
||||
sph_blake512_context ctx; |
||||
|
||||
sph_blake512_init(&ctx); |
||||
sph_blake512(&ctx, input, 80); |
||||
sph_blake512_close(&ctx, hash); |
||||
|
||||
sph_blake512(&ctx, hash, 64); |
||||
sph_blake512_close(&ctx, hashB); |
||||
|
||||
sph_blake512(&ctx, hashB, 64); |
||||
sph_blake512_close(&ctx, hash); |
||||
|
||||
sph_blake512(&ctx, hash, 64); |
||||
sph_blake512_close(&ctx, hashB); |
||||
|
||||
sph_blake512(&ctx, hashB, 64); |
||||
sph_blake512_close(&ctx, hash); |
||||
|
||||
memcpy(output, hash, 32); |
||||
} |
||||
|
||||
#include "cuda_helper.h" |
||||
|
||||
#define MAXU 0xffffffffU |
||||
|
||||
// in cpu-miner.c |
||||
extern bool opt_n_threads; |
||||
extern bool opt_benchmark; |
||||
extern int device_map[8]; |
||||
|
||||
__constant__ |
||||
static uint32_t __align__(32) c_Target[8]; |
||||
|
||||
__constant__ |
||||
static uint64_t __align__(32) c_data[32]; |
||||
|
||||
static uint32_t *d_hash[8]; |
||||
static uint32_t *d_resNounce[8]; |
||||
static uint32_t *h_resNounce[8]; |
||||
static uint32_t extra_results[2] = { MAXU, MAXU }; |
||||
|
||||
/* prefer uint32_t to prevent size conversions = speed +5/10 % */ |
||||
__constant__ |
||||
static uint32_t __align__(32) c_sigma[16][16]; |
||||
const uint32_t host_sigma[16][16] = { |
||||
{ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 }, |
||||
{14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 }, |
||||
{11, 8, 12, 0, 5, 2, 15, 13, 10, 14, 3, 6, 7, 1, 9, 4 }, |
||||
{ 7, 9, 3, 1, 13, 12, 11, 14, 2, 6, 5, 10, 4, 0, 15, 8 }, |
||||
{ 9, 0, 5, 7, 2, 4, 10, 15, 14, 1, 11, 12, 6, 8, 3, 13 }, |
||||
{ 2, 12, 6, 10, 0, 11, 8, 3, 4, 13, 7, 5, 15, 14, 1, 9 }, |
||||
{12, 5, 1, 15, 14, 13, 4, 10, 0, 7, 6, 3, 9, 2, 8, 11 }, |
||||
{13, 11, 7, 14, 12, 1, 3, 9, 5, 0, 15, 4, 8, 6, 2, 10 }, |
||||
{ 6, 15, 14, 9, 11, 3, 0, 8, 12, 2, 13, 7, 1, 4, 10, 5 }, |
||||
{10, 2, 8, 4, 7, 6, 1, 5, 15, 11, 9, 14, 3, 12, 13 , 0 }, |
||||
{ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 }, |
||||
{14, 10, 4, 8, 9, 15, 13, 6, 1, 12, 0, 2, 11, 7, 5, 3 }, |
||||
{11, 8, 12, 0, 5, 2, 15, 13, 10, 14, 3, 6, 7, 1, 9, 4 }, |
||||
{ 7, 9, 3, 1, 13, 12, 11, 14, 2, 6, 5, 10, 4, 0, 15, 8 }, |
||||
{ 9, 0, 5, 7, 2, 4, 10, 15, 14, 1, 11, 12, 6, 8, 3, 13 }, |
||||
{ 2, 12, 6, 10, 0, 11, 8, 3, 4, 13, 7, 5, 15, 14, 1, 9 } |
||||
}; |
||||
|
||||
__device__ __constant__ |
||||
static const uint64_t __align__(32) c_IV512[8] = { |
||||
0x6a09e667f3bcc908ULL, |
||||
0xbb67ae8584caa73bULL, |
||||
0x3c6ef372fe94f82bULL, |
||||
0xa54ff53a5f1d36f1ULL, |
||||
0x510e527fade682d1ULL, |
||||
0x9b05688c2b3e6c1fULL, |
||||
0x1f83d9abfb41bd6bULL, |
||||
0x5be0cd19137e2179ULL |
||||
}; |
||||
|
||||
__device__ __constant__ |
||||
const uint64_t c_u512[16] = |
||||
{ |
||||
0x243f6a8885a308d3ULL, 0x13198a2e03707344ULL, |
||||
0xa4093822299f31d0ULL, 0x082efa98ec4e6c89ULL, |
||||
0x452821e638d01377ULL, 0xbe5466cf34e90c6cULL, |
||||
0xc0ac29b7c97c50ddULL, 0x3f84d5b5b5470917ULL, |
||||
0x9216d5d98979fb1bULL, 0xd1310ba698dfb5acULL, |
||||
0x2ffd72dbd01adfb7ULL, 0xb8e1afed6a267e96ULL, |
||||
0xba7c9045f12c7f99ULL, 0x24a19947b3916cf7ULL, |
||||
0x0801f2e2858efc16ULL, 0x636920d871574e69ULL |
||||
}; |
||||
|
||||
#define G(a,b,c,d,x) { \ |
||||
uint32_t idx1 = c_sigma[i][x]; \ |
||||
uint32_t idx2 = c_sigma[i][x+1]; \ |
||||
v[a] += (m[idx1] ^ c_u512[idx2]) + v[b]; \ |
||||
v[d] = ROTR64(v[d] ^ v[a], 32); \ |
||||
v[c] += v[d]; \ |
||||
v[b] = ROTR64(v[b] ^ v[c], 25); \ |
||||
v[a] += (m[idx2] ^ c_u512[idx1]) + v[b]; \ |
||||
v[d] = ROTR64(v[d] ^ v[a], 16); \ |
||||
v[c] += v[d]; \ |
||||
v[b] = ROTR64(v[b] ^ v[c], 11); \ |
||||
} |
||||
|
||||
// Hash-Padding |
||||
__device__ __constant__ |
||||
static const uint64_t d_constHashPadding[8] = { |
||||
0x0000000000000080ull, |
||||
0, |
||||
0, |
||||
0, |
||||
0, |
||||
0x0100000000000000ull, |
||||
0, |
||||
0x0002000000000000ull |
||||
}; |
||||
|
||||
#if 0 |
||||
|
||||
__device__ __constant__ |
||||
static const uint64_t __align__(32) c_Padding[16] = { |
||||
0, 0, 0, 0, |
||||
0x80000000ULL, 0, 0, 0, |
||||
0, 0, 0, 0, |
||||
0, 1, 0, 640, |
||||
}; |
||||
|
||||
__device__ static |
||||
void pentablake_compress(uint64_t *h, const uint64_t *block, const uint32_t T0) |
||||
{ |
||||
uint64_t v[16], m[16]; |
||||
|
||||
m[0] = block[0]; |
||||
m[1] = block[1]; |
||||
m[2] = block[2]; |
||||
m[3] = block[3]; |
||||
|
||||
for (uint32_t i = 4; i < 16; i++) { |
||||
m[i] = (T0 == 0x200) ? block[i] : c_Padding[i]; |
||||
} |
||||
|
||||
//#pragma unroll 8 |
||||
for(uint32_t i = 0; i < 8; i++) |
||||
v[i] = h[i]; |
||||
|
||||
v[ 8] = c_u512[0]; |
||||
v[ 9] = c_u512[1]; |
||||
v[10] = c_u512[2]; |
||||
v[11] = c_u512[3]; |
||||
|
||||
v[12] = xor1(c_u512[4], T0); |
||||
v[13] = xor1(c_u512[5], T0); |
||||
v[14] = c_u512[6]; |
||||
v[15] = c_u512[7]; |
||||
|
||||
for (uint32_t i = 0; i < 16; i++) { |
||||
/* column step */ |
||||
G(0, 4, 0x8, 0xC, 0x0); |
||||
G(1, 5, 0x9, 0xD, 0x2); |
||||
G(2, 6, 0xA, 0xE, 0x4); |
||||
G(3, 7, 0xB, 0xF, 0x6); |
||||
/* diagonal step */ |
||||
G(0, 5, 0xA, 0xF, 0x8); |
||||
G(1, 6, 0xB, 0xC, 0xA); |
||||
G(2, 7, 0x8, 0xD, 0xC); |
||||
G(3, 4, 0x9, 0xE, 0xE); |
||||
} |
||||
|
||||
//#pragma unroll 16 |
||||
for (uint32_t i = 0; i < 16; i++) { |
||||
uint32_t j = i % 8; |
||||
h[j] ^= v[i]; |
||||
} |
||||
} |
||||
|
||||
__global__ |
||||
void pentablake_gpu_hash_80(uint32_t threads, uint32_t startNounce, uint32_t *resNounce) |
||||
{ |
||||
uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x); |
||||
if (thread < threads) |
||||
{ |
||||
const uint32_t nounce = startNounce + thread; |
||||
uint64_t h[8]; |
||||
|
||||
#pragma unroll |
||||
for(int i=0; i<8; i++) { |
||||
h[i] = c_IV512[i]; |
||||
} |
||||
|
||||
uint64_t ending[4]; |
||||
ending[0] = c_data[16]; |
||||
ending[1] = c_data[17]; |
||||
ending[2] = c_data[18]; |
||||
ending[3] = nounce; /* our tested value */ |
||||
|
||||
pentablake_compress(h, ending, 640); |
||||
|
||||
// ----------------------------------- |
||||
|
||||
for (int r = 0; r < 4; r++) { |
||||
uint64_t data[8]; |
||||
for (int i = 0; i < 7; i++) { |
||||
data[i] = h[i]; |
||||
} |
||||
pentablake_compress(h, data, 512); /* todo: use h,h when ok*/ |
||||
} |
||||
} |
||||
} |
||||
#endif |
||||
|
||||
__device__ static |
||||
void pentablake_compress(uint64_t *h, const uint64_t *block, const uint64_t T0) |
||||
{ |
||||
uint64_t v[16], m[16], i; |
||||
|
||||
#pragma unroll 16 |
||||
for(i = 0; i < 16; i++) { |
||||
m[i] = cuda_swab64(block[i]); |
||||
} |
||||
|
||||
#pragma unroll 8 |
||||
for (i = 0; i < 8; i++) |
||||
v[i] = h[i]; |
||||
|
||||
v[ 8] = c_u512[0]; |
||||
v[ 9] = c_u512[1]; |
||||
v[10] = c_u512[2]; |
||||
v[11] = c_u512[3]; |
||||
v[12] = c_u512[4] ^ T0; |
||||
v[13] = c_u512[5] ^ T0; |
||||
v[14] = c_u512[6]; |
||||
v[15] = c_u512[7]; |
||||
|
||||
//#pragma unroll 16 |
||||
for( i = 0; i < 16; i++) |
||||
{ |
||||
/* column step */ |
||||
G(0, 4, 0x8, 0xC, 0x0); |
||||
G(1, 5, 0x9, 0xD, 0x2); |
||||
G(2, 6, 0xA, 0xE, 0x4); |
||||
G(3, 7, 0xB, 0xF, 0x6); |
||||
/* diagonal step */ |
||||
G(0, 5, 0xA, 0xF, 0x8); |
||||
G(1, 6, 0xB, 0xC, 0xA); |
||||
G(2, 7, 0x8, 0xD, 0xC); |
||||
G(3, 4, 0x9, 0xE, 0xE); |
||||
} |
||||
|
||||
//#pragma unroll 16 |
||||
for (i = 0; i < 16; i++) { |
||||
uint32_t idx = i % 8; |
||||
h[idx] ^= v[i]; |
||||
} |
||||
} |
||||
|
||||
__global__ |
||||
void pentablake_gpu_hash_80(int threads, const uint32_t startNounce, void *outputHash) |
||||
{ |
||||
int thread = (blockDim.x * blockIdx.x + threadIdx.x); |
||||
if (thread < threads) |
||||
{ |
||||
uint64_t h[8]; |
||||
uint64_t buf[16]; |
||||
uint32_t nounce = startNounce + thread; |
||||
|
||||
//#pragma unroll 8 |
||||
for(int i=0; i<8; i++) |
||||
h[i] = c_IV512[i]; |
||||
|
||||
//#pragma unroll 16 |
||||
for (int i=0; i < 16; i++) |
||||
buf[i] = c_data[i]; |
||||
|
||||
// The test Nonce |
||||
((uint32_t*)buf)[19] = cuda_swab32(nounce); |
||||
|
||||
pentablake_compress(h, buf, 640ULL); |
||||
|
||||
#if __CUDA_ARCH__ < 300 |
||||
uint32_t *outHash = (uint32_t *)outputHash + 16 * thread; |
||||
#pragma unroll 8 |
||||
for (uint32_t i=0; i < 8; i++) { |
||||
outHash[2*i] = cuda_swab32( _HIWORD(h[i]) ); |
||||
outHash[2*i+1] = cuda_swab32( _LOWORD(h[i]) ); |
||||
} |
||||
#else |
||||
uint64_t *outHash = (uint64_t *)outputHash + 8 * thread; |
||||
for (uint32_t i=0; i < 8; i++) { |
||||
outHash[i] = cuda_swab64( h[i] ); |
||||
} |
||||
#endif |
||||
|
||||
} |
||||
} |
||||
|
||||
__host__ |
||||
void pentablake_cpu_hash_80(int thr_id, int threads, const uint32_t startNounce, uint32_t *d_outputHash, int order) |
||||
{ |
||||
const int threadsperblock = TPB; |
||||
|
||||
dim3 grid((threads + threadsperblock-1)/threadsperblock); |
||||
dim3 block(threadsperblock); |
||||
size_t shared_size = 0; |
||||
|
||||
pentablake_gpu_hash_80 <<<grid, block, shared_size>>> (threads, startNounce, d_outputHash); |
||||
|
||||
//MyStreamSynchronize(NULL, order, thr_id); |
||||
cudaDeviceSynchronize(); |
||||
} |
||||
|
||||
|
||||
__global__ |
||||
void pentablake_gpu_hash_64(int threads, uint32_t startNounce, uint64_t *g_hash) |
||||
{ |
||||
int thread = (blockDim.x * blockIdx.x + threadIdx.x); |
||||
|
||||
if (thread < threads) |
||||
{ |
||||
uint64_t *inpHash = &g_hash[thread<<3]; // hashPosition * 8 |
||||
uint64_t buf[16]; // 128 Bytes |
||||
uint64_t h[8]; // State |
||||
|
||||
#pragma unroll 8 |
||||
for (int i=0; i<8; i++) |
||||
h[i] = c_IV512[i]; |
||||
|
||||
// Message for first round |
||||
#pragma unroll 8 |
||||
for (int i=0; i < 8; ++i) |
||||
buf[i] = inpHash[i]; |
||||
|
||||
#pragma unroll 8 |
||||
for (int i=0; i < 8; i++) |
||||
buf[i+8] = d_constHashPadding[i]; |
||||
|
||||
// Ending round |
||||
pentablake_compress(h, buf, 512); |
||||
|
||||
#if __CUDA_ARCH__ < 300 |
||||
uint32_t *outHash = (uint32_t*)&g_hash[thread<<3]; |
||||
#pragma unroll 8 |
||||
for (int i=0; i < 8; i++) { |
||||
outHash[2*i+0] = cuda_swab32( _HIWORD(h[i]) ); |
||||
outHash[2*i+1] = cuda_swab32( _LOWORD(h[i]) ); |
||||
} |
||||
#else |
||||
uint64_t *outHash = &g_hash[thread<<3]; |
||||
for (int i=0; i < 8; i++) { |
||||
outHash[i] = cuda_swab64(h[i]); |
||||
} |
||||
#endif |
||||
} |
||||
} |
||||
|
||||
__host__ |
||||
void pentablake_cpu_hash_64(int thr_id, int threads, uint32_t startNounce, uint32_t *d_outputHash, int order) |
||||
{ |
||||
const int threadsperblock = TPB; |
||||
|
||||
dim3 grid((threads + threadsperblock-1)/threadsperblock); |
||||
dim3 block(threadsperblock); |
||||
size_t shared_size = 0; |
||||
|
||||
pentablake_gpu_hash_64 <<<grid, block, shared_size>>> (threads, startNounce, (uint64_t*)d_outputHash); |
||||
|
||||
//MyStreamSynchronize(NULL, order, thr_id); |
||||
cudaDeviceSynchronize(); |
||||
} |
||||
|
||||
#if 0 |
||||
|
||||
__host__ |
||||
uint32_t pentablake_cpu_hash_80(int thr_id, uint32_t threads, uint32_t startNounce) |
||||
{ |
||||
const int threadsperblock = TPB; |
||||
uint32_t result = MAXU; |
||||
|
||||
dim3 grid((threads + threadsperblock-1)/threadsperblock); |
||||
dim3 block(threadsperblock); |
||||
size_t shared_size = 0; |
||||
|
||||
/* Check error on Ctrl+C or kill to prevent segfaults on exit */ |
||||
if (cudaMemset(d_resNounce[thr_id], 0xff, 2*sizeof(uint32_t)) != cudaSuccess) |
||||
return result; |
||||
|
||||
pentablake_gpu_hash_80<<<grid, block, shared_size>>>(threads, startNounce, d_resNounce[thr_id]); |
||||
cudaDeviceSynchronize(); |
||||
if (cudaSuccess == cudaMemcpy(h_resNounce[thr_id], d_resNounce[thr_id], 2*sizeof(uint32_t), cudaMemcpyDeviceToHost)) { |
||||
cudaThreadSynchronize(); |
||||
result = h_resNounce[thr_id][0]; |
||||
extra_results[0] = h_resNounce[thr_id][1]; |
||||
} |
||||
return result; |
||||
} |
||||
#endif |
||||
|
||||
__global__ |
||||
void pentablake_gpu_check_hash(uint32_t threads, uint32_t startNounce, uint32_t *g_hash, uint32_t *resNounce) |
||||
{ |
||||
uint32_t thread = (blockDim.x * blockIdx.x + threadIdx.x); |
||||
if (thread < threads) |
||||
{ |
||||
uint32_t nounce = startNounce + thread; |
||||
uint32_t *inpHash = &g_hash[thread<<4]; |
||||
uint32_t h[8]; |
||||
|
||||
#pragma unroll 8 |
||||
for (int i=0; i < 8; i++) |
||||
h[i] = inpHash[i]; |
||||
|
||||
for (int i = 7; i >= 0; i--) { |
||||
uint32_t hash = h[i]; // cuda_swab32(h[i]); |
||||
if (hash > c_Target[i]) { |
||||
return; |
||||
} |
||||
if (hash < c_Target[i]) { |
||||
break; |
||||
} |
||||
} |
||||
|
||||
/* keep the smallest nounce, + extra one if found */ |
||||
if (resNounce[0] > nounce) { |
||||
resNounce[1] = resNounce[0]; |
||||
resNounce[0] = nounce; |
||||
} |
||||
else |
||||
resNounce[1] = nounce; |
||||
} |
||||
} |
||||
|
||||
__host__ static |
||||
uint32_t pentablake_check_hash(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_inputHash, int order) |
||||
{ |
||||
const int threadsperblock = TPB; |
||||
uint32_t result = MAXU; |
||||
|
||||
dim3 grid((threads + threadsperblock-1)/threadsperblock); |
||||
dim3 block(threadsperblock); |
||||
size_t shared_size = 0; |
||||
|
||||
/* Check error on Ctrl+C or kill to prevent segfaults on exit */ |
||||
if (cudaMemset(d_resNounce[thr_id], 0xff, 2*sizeof(uint32_t)) != cudaSuccess) |
||||
return result; |
||||
|
||||
pentablake_gpu_check_hash <<<grid, block, shared_size>>> (threads, startNounce, d_inputHash, d_resNounce[thr_id]); |
||||
|
||||
CUDA_SAFE_CALL(cudaDeviceSynchronize()); |
||||
if (cudaSuccess == cudaMemcpy(h_resNounce[thr_id], d_resNounce[thr_id], 2*sizeof(uint32_t), cudaMemcpyDeviceToHost)) { |
||||
cudaThreadSynchronize(); |
||||
result = h_resNounce[thr_id][0]; |
||||
extra_results[0] = h_resNounce[thr_id][1]; |
||||
} |
||||
return result; |
||||
} |
||||
|
||||
|
||||
__host__ |
||||
void pentablake_cpu_setBlock_80(uint32_t *pdata, const uint32_t *ptarget) |
||||
{ |
||||
uint8_t data[128]; |
||||
memcpy((void*) data, (void*) pdata, 80); |
||||
memset(data+80, 0, 48); |
||||
|
||||
// to swab... |
||||
data[80] = 0x80; |
||||
data[111] = 1; |
||||
data[126] = 0x02; |
||||
data[127] = 0x80; |
||||
|
||||
CUDA_SAFE_CALL(cudaMemcpyToSymbol(c_data, data, sizeof(data), 0, cudaMemcpyHostToDevice)); |
||||
CUDA_SAFE_CALL(cudaMemcpyToSymbol(c_sigma, host_sigma, sizeof(host_sigma), 0, cudaMemcpyHostToDevice)); |
||||
CUDA_SAFE_CALL(cudaMemcpyToSymbol(c_Target, ptarget, 32, 0, cudaMemcpyHostToDevice)); |
||||
} |
||||
|
||||
extern "C" int scanhash_pentablake(int thr_id, uint32_t *pdata, const uint32_t *ptarget, |
||||
uint32_t max_nonce, unsigned long *hashes_done) |
||||
{ |
||||
const uint32_t first_nonce = pdata[19]; |
||||
static bool init[8] = { 0, 0, 0, 0, 0, 0, 0, 0 }; |
||||
uint32_t throughput = min(128 * 2560, max_nonce - first_nonce); |
||||
uint32_t endiandata[20]; |
||||
int rc = 0; |
||||
|
||||
if (extra_results[0] != MAXU) { |
||||
// possible extra result found in previous call |
||||
if (first_nonce <= extra_results[0] && max_nonce >= extra_results[0]) { |
||||
pdata[19] = extra_results[0]; |
||||
*hashes_done = pdata[19] - first_nonce + 1; |
||||
extra_results[0] = MAXU; |
||||
rc = 1; |
||||
goto exit_scan; |
||||
} |
||||
} |
||||
|
||||
if (opt_benchmark) |
||||
((uint32_t*)ptarget)[7] = 0x000F; |
||||
|
||||
if (!init[thr_id]) { |
||||
if (opt_n_threads > 1) { |
||||
CUDA_SAFE_CALL(cudaSetDevice(device_map[thr_id])); |
||||
} |
||||
CUDA_SAFE_CALL(cudaMalloc(&d_hash[thr_id], 64 * throughput)); |
||||
CUDA_SAFE_CALL(cudaMallocHost(&h_resNounce[thr_id], 2*sizeof(uint32_t))); |
||||
CUDA_SAFE_CALL(cudaMalloc(&d_resNounce[thr_id], 2*sizeof(uint32_t))); |
||||
|
||||
init[thr_id] = true; |
||||
} |
||||
|
||||
for (int k=0; k < 20; k++) |
||||
be32enc(&endiandata[k], pdata[k]); |
||||
|
||||
pentablake_cpu_setBlock_80(endiandata, ptarget); |
||||
|
||||
do { |
||||
int order = 0; |
||||
|
||||
// GPU HASH |
||||
pentablake_cpu_hash_80(thr_id, throughput, pdata[19], d_hash[thr_id], order++); |
||||
|
||||
pentablake_cpu_hash_64(thr_id, throughput, pdata[19], d_hash[thr_id], order++); |
||||
pentablake_cpu_hash_64(thr_id, throughput, pdata[19], d_hash[thr_id], order++); |
||||
pentablake_cpu_hash_64(thr_id, throughput, pdata[19], d_hash[thr_id], order++); |
||||
pentablake_cpu_hash_64(thr_id, throughput, pdata[19], d_hash[thr_id], order++); |
||||
|
||||
uint32_t foundNonce = pentablake_check_hash(thr_id, throughput, pdata[19], d_hash[thr_id], order++); |
||||
|
||||
if (foundNonce != MAXU) |
||||
{ |
||||
uint32_t vhashcpu[8]; |
||||
uint32_t Htarg = ptarget[7]; |
||||
|
||||
be32enc(&endiandata[19], foundNonce); |
||||
|
||||
pentablakehash(vhashcpu, endiandata); |
||||
|
||||
if (vhashcpu[7] <= Htarg && fulltest(vhashcpu, ptarget)) |
||||
{ |
||||
pdata[19] = foundNonce; |
||||
rc = 1; |
||||
|
||||
// Rare but possible if the throughput is big |
||||
be32enc(&endiandata[19], extra_results[0]); |
||||
pentablakehash(vhashcpu, endiandata); |
||||
if (vhashcpu[7] <= Htarg && fulltest(vhashcpu, ptarget)) { |
||||
applog(LOG_NOTICE, "GPU found more than one result yippee!"); |
||||
rc = 2; |
||||
} else { |
||||
extra_results[0] = MAXU; |
||||
} |
||||
|
||||
goto exit_scan; |
||||
} |
||||
else if (vhashcpu[7] > Htarg) { |
||||
applog(LOG_WARNING, "GPU #%d: result for nounce %08x is not in range: %x > %x", thr_id, foundNonce, vhashcpu[7], Htarg); |
||||
} |
||||
else if (vhashcpu[6] > ptarget[6]) { |
||||
applog(LOG_WARNING, "GPU #%d: hash[6] for nounce %08x is not in range: %x > %x", thr_id, foundNonce, vhashcpu[6], ptarget[6]); |
||||
} |
||||
else { |
||||
applog(LOG_WARNING, "GPU #%d: result for nounce %08x does not validate on CPU!", thr_id, foundNonce); |
||||
} |
||||
} |
||||
|
||||
pdata[19] += throughput; |
||||
|
||||
} while (pdata[19] < max_nonce && !work_restart[thr_id].restart); |
||||
|
||||
exit_scan: |
||||
*hashes_done = pdata[19] - first_nonce + 1; |
||||
#if 0 |
||||
/* reset the device to allow multiple instances |
||||
* could be made in cpu-miner... check later if required */ |
||||
if (opt_n_threads == 1) { |
||||
CUDA_SAFE_CALL(cudaDeviceReset()); |
||||
init[thr_id] = false; |
||||
} |
||||
#endif |
||||
|
||||
cudaDeviceSynchronize(); |
||||
return rc; |
||||
} |
||||
Loading…
Reference in new issue