sib coin algo (X11 + Streebog)

Signed-off-by: Tanguy Pruvot <tanguy.pruvot@gmail.com>

Makefile.am

@@ -47,7 +47,7 @@ ccminer_SOURCES	= elist.h miner.h compat.h \
 			  cuda_nist5.cu pentablake.cu skein.cu cuda_skeincoin.cu skein2.cpp zr5.cu \
 			  sph/bmw.c sph/blake.c sph/groestl.c sph/jh.c sph/keccak.c sph/skein.c \
 			  sph/cubehash.c sph/echo.c sph/luffa.c sph/sha2.c sph/shavite.c sph/simd.c \
-			  sph/hamsi.c sph/hamsi_helper.c sph/sph_hamsi.h \
+			  sph/hamsi.c sph/hamsi_helper.c sph/streebog.c \
 			  sph/shabal.c sph/whirlpool.c sph/sha2big.c sph/haval.c \
 			  qubit/qubit.cu qubit/qubit_luffa512.cu qubit/deep.cu qubit/luffa.cu \
 			  x11/x11.cu x11/fresh.cu x11/cuda_x11_luffa512.cu x11/cuda_x11_cubehash512.cu \
@@ -57,7 +57,7 @@ ccminer_SOURCES	= elist.h miner.h compat.h \
 			  x15/x14.cu x15/x15.cu x15/cuda_x14_shabal512.cu x15/cuda_x15_whirlpool.cu \
 			  x15/whirlpool.cu x15/whirlpoolx.cu x15/cuda_whirlpoolx.cu \
 			  x17/x17.cu x17/cuda_x17_haval512.cu x17/cuda_x17_sha512.cu \
-			  x11/c11.cu x11/s3.cu
+			  x11/c11.cu x11/s3.cu x11/sib.cu x11/cuda_streebog.cu
 
 # scrypt
 ccminer_SOURCES += scrypt.cpp scrypt-jane.cpp \

algos.h

@@ -29,6 +29,7 @@ enum sha_algos {
 	ALGO_QUBIT,
 	ALGO_SCRYPT,
 	ALGO_SCRYPT_JANE,
+	ALGO_SIB,
 	ALGO_SKEIN,
 	ALGO_SKEIN2,
 	ALGO_S3,
@@ -72,6 +73,7 @@ static const char *algo_names[] = {
 	"qubit",
 	"scrypt",
 	"scrypt-jane",
+	"sib",
 	"skein",
 	"skein2",
 	"s3",

ccminer.cpp

@@ -229,6 +229,7 @@ Options:\n\
 			penta       Pentablake hash (5x Blake 512)\n\
 			quark       Quark\n\
 			qubit       Qubit\n\
+			sib         Sibcoin (X11+Streebog)\n\
 			scrypt      Scrypt\n\
 			scrypt-jane Scrypt-jane Chacha\n\
 			skein       Skein SHA2 (Skeincoin)\n\
@@ -1774,6 +1775,7 @@ static void *miner_thread(void *userdata)
 				break;
 			case ALGO_LYRA2:
 			case ALGO_NEOSCRYPT:
+			case ALGO_SIB:
 			case ALGO_SCRYPT:
 				minmax = 0x80000;
 				break;
@@ -1901,6 +1903,9 @@ static void *miner_thread(void *userdata)
 		case ALGO_SKEIN2:
 			rc = scanhash_skein2(thr_id, &work, max_nonce, &hashes_done);
 			break;
+		case ALGO_SIB:
+			rc = scanhash_sib(thr_id, &work, max_nonce, &hashes_done);
+			break;
 		case ALGO_S3:
 			rc = scanhash_s3(thr_id, &work, max_nonce, &hashes_done);
 			break;

miner.h

@@ -280,6 +280,7 @@ extern int scanhash_nist5(int thr_id, struct work *work, uint32_t max_nonce, uns
 extern int scanhash_pentablake(int thr_id, struct work *work, uint32_t max_nonce, unsigned long *hashes_done);
 extern int scanhash_quark(int thr_id, struct work *work, uint32_t max_nonce, unsigned long *hashes_done);
 extern int scanhash_qubit(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
+extern int scanhash_sib(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
 extern int scanhash_skeincoin(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
 extern int scanhash_skein2(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
 extern int scanhash_s3(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
@@ -320,6 +321,7 @@ extern void free_nist5(int thr_id);
 extern void free_pentablake(int thr_id);
 extern void free_quark(int thr_id);
 extern void free_qubit(int thr_id);
+extern void free_sib(int thr_id);
 extern void free_skeincoin(int thr_id);
 extern void free_skein2(int thr_id);
 extern void free_s3(int thr_id);
@@ -784,6 +786,7 @@ void quarkhash(void *state, const void *input);
 void qubithash(void *state, const void *input);
 void scrypthash(void* output, const void* input);
 void scryptjane_hash(void* output, const void* input);
+void sibhash(void *output, const void *input);
 void skeincoinhash(void *output, const void *input);
 void skein2hash(void *output, const void *input);
 void s3hash(void *output, const void *input);

sph/sph_streebog.h

@@ -0,0 +1,185 @@
+/* $Id: sph_gost.h 216 2010-06-08 09:46:57Z tp $ */
+/**
+ * GOST interface. This is the interface for GOST R 12 with the
+ * recommended parameters for SHA-3, with output lengths 256
+ * and 512 bits.
+ *
+ * ==========================(LICENSE BEGIN)============================
+ *
+ * Copyright (c) 2007-2010  Projet RNRT SAPHIR
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining
+ * a copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sublicense, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be
+ * included in all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
+ * IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
+ * CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
+ * TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
+ * SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * ===========================(LICENSE END)=============================
+ *
+ * @file     sph_gost.h
+ * @author   Mish <mish@btchouse.com>
+ */
+
+#ifndef SPH_GOST_H__
+#define SPH_GOST_H__
+
+#ifdef __cplusplus
+extern "C"{
+#endif
+
+#include <stddef.h>
+#include "sph_types.h"
+
+/**
+ * Output size (in bits) for GOST-256.
+ */
+#define SPH_SIZE_gost256   256
+
+/**
+ * Output size (in bits) for GOST-512.
+ */
+#define SPH_SIZE_gost512   512
+
+/**
+ * This structure is a context for Keccak computations: it contains the
+ * intermediate values and some data from the last entered block. Once a
+ * GOST computation has been performed, the context can be reused for
+ * another computation.
+ *
+ * The contents of this structure are private. A running GOST computation
+ * can be cloned by copying the context (e.g. with a simple
+ * <code>memcpy()</code>).
+ */
+
+/**
+ * This structure is a context for Gost-256 computations.
+ */
+
+typedef struct {
+#ifndef DOXYGEN_IGNORE
+	unsigned char buf[32];    /* first field, for alignment */
+	size_t ptr;
+	sph_u32 V[3][8];
+#endif
+} sph_gost256_context;
+
+/**
+ * This structure is a context for Gost-512 computations.
+ */
+typedef struct {
+#ifndef DOXYGEN_IGNORE
+	unsigned char buf[64];    /* first field, for alignment */
+	size_t ptr;
+	sph_u32 V[5][8];
+#endif
+} sph_gost512_context;
+
+
+/**
+ * Initialize a GOST-256 context. This process performs no memory allocation.
+ *
+ * @param cc   the GOST-256 context (pointer to a
+ *             <code>sph_gost256_context</code>)
+ */
+void sph_gost256_init(void *cc);
+
+/**
+ * Process some data bytes. It is acceptable that <code>len</code> is zero
+ * (in which case this function does nothing).
+ *
+ * @param cc     the Gost-256 context
+ * @param data   the input data
+ * @param len    the input data length (in bytes)
+ */
+void sph_gost256(void *cc, const void *data, size_t len);
+
+/**
+ * Terminate the current GOST-256 computation and output the result into
+ * the provided buffer. The destination buffer must be wide enough to
+ * accomodate the result (32 bytes). The context is automatically
+ * reinitialized.
+ *
+ * @param cc    the GOST-256 context
+ * @param dst   the destination buffer
+ */
+void sph_gost256_close(void *cc, void *dst);
+
+/**
+ * Add a few additional bits (0 to 7) to the current computation, then
+ * terminate it and output the result in the provided buffer, which must
+ * be wide enough to accomodate the result (32 bytes). If bit number i
+ * in <code>ub</code> has value 2^i, then the extra bits are those
+ * numbered 7 downto 8-n (this is the big-endian convention at the byte
+ * level). The context is automatically reinitialized.
+ *
+ * @param cc    the GOST-256 context
+ * @param ub    the extra bits
+ * @param n     the number of extra bits (0 to 7)
+ * @param dst   the destination buffer
+ */
+void sph_gost256_addbits_and_close(
+	void *cc, unsigned ub, unsigned n, void *dst);
+
+/**
+ * Initialize a Gost-512 context. This process performs no memory allocation.
+ *
+ * @param cc   the GOST-512 context (pointer to a
+ *             <code>sph_gost512_context</code>)
+ */
+void sph_gost512_init(void *cc);
+
+/**
+ * Process some data bytes. It is acceptable that <code>len</code> is zero
+ * (in which case this function does nothing).
+ *
+ * @param cc     the GOST-512 context
+ * @param data   the input data
+ * @param len    the input data length (in bytes)
+ */
+void sph_gost512(void *cc, const void *data, size_t len);
+
+/**
+ * Terminate the current GOST-512 computation and output the result into
+ * the provided buffer. The destination buffer must be wide enough to
+ * accomodate the result (64 bytes). The context is automatically
+ * reinitialized.
+ *
+ * @param cc    the GOST-512 context
+ * @param dst   the destination buffer
+ */
+void sph_gost512_close(void *cc, void *dst);
+
+/**
+ * Add a few additional bits (0 to 7) to the current computation, then
+ * terminate it and output the result in the provided buffer, which must
+ * be wide enough to accomodate the result (64 bytes). If bit number i
+ * in <code>ub</code> has value 2^i, then the extra bits are those
+ * numbered 7 downto 8-n (this is the big-endian convention at the byte
+ * level). The context is automatically reinitialized.
+ *
+ * @param cc    the GOST-512 context
+ * @param ub    the extra bits
+ * @param n     the number of extra bits (0 to 7)
+ * @param dst   the destination buffer
+ */
+void sph_gost512_addbits_and_close(
+	void *cc, unsigned ub, unsigned n, void *dst);
+
+#ifdef __cplusplus
+}
+#endif
+
+#endif

util.cpp

@@ -1865,7 +1865,7 @@ void do_gpu_tests(void)
 	//scanhash_scrypt_jane(0, &work, NULL, 1, &done, &tv, &tv);
 
 	memset(work.data, 0, sizeof(work.data));
-	scanhash_quark(0, &work, 1, &done);
+	scanhash_sib(0, &work, 1, &done);
 
 	free(work_restart);
 	work_restart = NULL;
@@ -1954,6 +1954,9 @@ void print_hash_tests(void)
 	scryptjane_hash(&hash[0], &buf[0]);
 	printpfx("scrypt-jane", hash);
 
+	sibhash(&hash[0], &buf[0]);
+	printpfx("sib", hash);
+
 	skeincoinhash(&hash[0], &buf[0]);
 	printpfx("skein", hash);
 

x11/sib.cu

@@ -0,0 +1,257 @@
+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 "sph/sph_luffa.h"
+#include "sph/sph_cubehash.h"
+#include "sph/sph_shavite.h"
+#include "sph/sph_simd.h"
+#include "sph/sph_echo.h"
+#include "sph/sph_streebog.h"
+}
+
+#include "miner.h"
+#include "cuda_helper.h"
+#include "cuda_x11.h"
+
+extern void streebog_cpu_hash_64(int thr_id, uint32_t threads, uint32_t startNounce, uint32_t *d_hash);
+
+#include <stdio.h>
+#include <memory.h>
+
+//#define _DEBUG
+
+static uint32_t *d_hash[MAX_GPUS];
+
+// Sibcoin CPU Hash
+extern "C" void sibhash(void *output, const void *input)
+{
+	unsigned char _ALIGN(128) hash[128] = { 0 };
+
+	sph_blake512_context ctx_blake;
+	sph_bmw512_context ctx_bmw;
+	sph_groestl512_context ctx_groestl;
+	sph_skein512_context ctx_skein;
+	sph_jh512_context ctx_jh;
+	sph_keccak512_context ctx_keccak;
+	sph_gost512_context ctx_gost;
+	sph_luffa512_context ctx_luffa;
+	sph_cubehash512_context ctx_cubehash;
+	sph_shavite512_context ctx_shavite;
+	sph_simd512_context ctx_simd;
+	sph_echo512_context ctx_echo;
+
+	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);
+
+	sph_groestl512_init(&ctx_groestl);
+	sph_groestl512 (&ctx_groestl, (const void*) hash, 64);
+	sph_groestl512_close(&ctx_groestl, (void*) hash);
+
+	sph_skein512_init(&ctx_skein);
+	sph_skein512 (&ctx_skein, (const void*) hash, 64);
+	sph_skein512_close(&ctx_skein, (void*) hash);
+
+	sph_jh512_init(&ctx_jh);
+	sph_jh512 (&ctx_jh, (const void*) hash, 64);
+	sph_jh512_close(&ctx_jh, (void*) hash);
+
+	sph_keccak512_init(&ctx_keccak);
+	sph_keccak512 (&ctx_keccak, (const void*) hash, 64);
+	sph_keccak512_close(&ctx_keccak, (void*) hash);
+
+	sph_gost512_init(&ctx_gost);
+	sph_gost512(&ctx_gost, (const void*) hash, 64);
+	sph_gost512_close(&ctx_gost, (void*) hash);
+	//applog_hash64(hash);
+
+	sph_luffa512_init(&ctx_luffa);
+	sph_luffa512 (&ctx_luffa, (const void*) hash, 64);
+	sph_luffa512_close (&ctx_luffa, (void*) hash);
+
+	sph_cubehash512_init(&ctx_cubehash);
+	sph_cubehash512 (&ctx_cubehash, (const void*) hash, 64);
+	sph_cubehash512_close(&ctx_cubehash, (void*) hash);
+
+	sph_shavite512_init(&ctx_shavite);
+	sph_shavite512 (&ctx_shavite, (const void*) hash, 64);
+	sph_shavite512_close(&ctx_shavite, (void*) hash);
+
+	sph_simd512_init(&ctx_simd);
+	sph_simd512 (&ctx_simd, (const void*) hash, 64);
+	sph_simd512_close(&ctx_simd, (void*) hash);
+
+	sph_echo512_init(&ctx_echo);
+	sph_echo512 (&ctx_echo, (const void*) hash, 64);
+	sph_echo512_close(&ctx_echo, (void*) hash);
+
+	memcpy(output, hash, 32);
+}
+
+#ifdef _DEBUG
+#define TRACE(algo) { \
+	if (max_nonce == 1 && pdata[19] <= 1 && !opt_benchmark) { \
+		uint32_t oft = 0; \
+		uint32_t* debugbuf = NULL; \
+		cudaMallocHost(&debugbuf, 16*sizeof(uint32_t)); \
+		cudaMemcpy(debugbuf, d_hash[thr_id] + oft, 16*sizeof(uint32_t), cudaMemcpyDeviceToHost); \
+		printf("SIB %s %08x %08x %08x %08x %08x %08x %08x %08x  %08x %08x %08x %08x %08x %08x %08x %08x\n", algo, \
+			swab32(debugbuf[0]), swab32(debugbuf[1]), swab32(debugbuf[2]), swab32(debugbuf[3]), \
+			swab32(debugbuf[4]), swab32(debugbuf[5]), swab32(debugbuf[6]), swab32(debugbuf[7]), \
+			swab32(debugbuf[8]), swab32(debugbuf[9]), swab32(debugbuf[10]),swab32(debugbuf[11]), \
+			swab32(debugbuf[12]),swab32(debugbuf[13]),swab32(debugbuf[14]),swab32(debugbuf[15])); \
+		cudaFreeHost(debugbuf); \
+	} \
+}
+#else
+#define TRACE(algo) {}
+#endif
+
+static bool init[MAX_GPUS] = { 0 };
+
+extern "C" int scanhash_sib(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done)
+{
+	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()) ? 19 : 18;
+	uint32_t throughput = cuda_default_throughput(thr_id, 1U << intensity); // 19=256*256*8;
+	if (init[thr_id]) throughput = min(throughput, max_nonce - first_nonce);
+
+	if (opt_benchmark)
+		ptarget[7] = 0xf;
+
+	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();
+		}
+
+		quark_blake512_cpu_init(thr_id, throughput);
+		quark_bmw512_cpu_init(thr_id, throughput);
+		quark_groestl512_cpu_init(thr_id, throughput);
+		quark_skein512_cpu_init(thr_id, throughput);
+		quark_keccak512_cpu_init(thr_id, throughput);
+		quark_jh512_cpu_init(thr_id, throughput);
+		x11_luffaCubehash512_cpu_init(thr_id, throughput);
+		x11_shavite512_cpu_init(thr_id, throughput);
+		x11_echo512_cpu_init(thr_id, throughput);
+		if (x11_simd512_cpu_init(thr_id, throughput) != 0) {
+			return 0;
+		}
+		CUDA_CALL_OR_RET_X(cudaMalloc(&d_hash[thr_id], (size_t) 64 * throughput), 0);
+
+		cuda_check_cpu_init(thr_id, throughput);
+
+		init[thr_id] = true;
+	}
+
+	uint32_t endiandata[20];
+	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 foundNonce;
+
+		// Hash with CUDA
+		quark_blake512_cpu_hash_80(thr_id, throughput, pdata[19], d_hash[thr_id]); order++;
+		TRACE("blake  :");
+		quark_bmw512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++);
+		TRACE("bmw    :");
+		quark_groestl512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++);
+		TRACE("groestl:");
+		quark_skein512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++);
+		TRACE("skein  :");
+		quark_jh512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++);
+		TRACE("jh512  :");
+		quark_keccak512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++);
+		TRACE("keccak :");
+		streebog_cpu_hash_64(thr_id, throughput, pdata[19], d_hash[thr_id]);
+		TRACE("gost   :");
+		x11_luffaCubehash512_cpu_hash_64(thr_id, throughput, d_hash[thr_id], order++);
+		TRACE("luffa+c:");
+		x11_shavite512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++);
+		TRACE("shavite:");
+		x11_simd512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++);
+		TRACE("simd   :");
+		x11_echo512_cpu_hash_64(thr_id, throughput, pdata[19], NULL, d_hash[thr_id], order++);
+		TRACE("echo => ");
+
+		foundNonce = cuda_check_hash(thr_id, throughput, pdata[19], d_hash[thr_id]);
+		if (foundNonce != UINT32_MAX)
+		{
+			const uint32_t Htarg = ptarget[7];
+			uint32_t vhash64[8];
+			be32enc(&endiandata[19], foundNonce);
+			sibhash(vhash64, endiandata);
+
+			if (vhash64[7] <= Htarg && fulltest(vhash64, ptarget)) {
+				int res = 1;
+				// check if there was some other ones...
+				uint32_t secNonce = cuda_check_hash_suppl(thr_id, throughput, pdata[19], d_hash[thr_id], 1);
+				work_set_target_ratio(work, vhash64);
+				*hashes_done = pdata[19] - first_nonce + throughput;
+				if (secNonce != 0) {
+					be32enc(&endiandata[19], secNonce);
+					sibhash(vhash64, endiandata);
+					if (bn_hash_target_ratio(vhash64, ptarget) > work->shareratio)
+						work_set_target_ratio(work, vhash64);
+					pdata[21] = secNonce;
+					res++;
+				}
+				pdata[19] = foundNonce;
+				return res;
+			} else {
+				gpulog(LOG_WARNING, thr_id, "result for %08x does not validate on CPU!", foundNonce);
+				pdata[19] = foundNonce + 1;
+				continue;
+			}
+		}
+
+		if ((uint64_t) throughput + pdata[19] >= max_nonce) {
+			pdata[19] = max_nonce;
+			break;
+		}
+		pdata[19] += throughput;
+
+	} while (!work_restart[thr_id].restart);
+
+	*hashes_done = pdata[19] - first_nonce;
+	return 0;
+}
+
+// cleanup
+extern "C" void free_sib(int thr_id)
+{
+	if (!init[thr_id])
+		return;
+
+	cudaThreadSynchronize();
+
+	cudaFree(d_hash[thr_id]);
+
+	quark_blake512_cpu_free(thr_id);
+	quark_groestl512_cpu_free(thr_id);
+	x11_simd512_cpu_free(thr_id);
+
+	cuda_check_cpu_free(thr_id);
+	init[thr_id] = false;
+
+	cudaDeviceSynchronize();
+}