X16s implem

Algo512/cuda_echo512_shortMsg.cu

@@ -2,8 +2,7 @@
 * Echo512 cuda kernel for short messages
 * 
 * 80 & 64 bytes messages implemented with precalculations
-* Code ready for message length from 8 to 
-* 
+* Code ready for message length from 8 to 96 bytes
 *
 * DumaxFr@github 2018
 * Based on tpruvot 2018 - GPL code

Makefile.am

@@ -45,6 +45,7 @@ ccminer_SOURCES	= elist.h miner.h compat.h \
 			  Algo256/blake256.cu Algo256/decred.cu Algo256/vanilla.cu Algo256/keccak256.cu \
 			  Algo256/blake2s.cu sph/blake2s.c \
 			  Algo256/bmw.cu Algo256/cuda_bmw.cu \
+			  Algo512/cuda_echo512_shortMsg.cu \
 			  crypto/xmr-rpc.cpp crypto/wildkeccak-cpu.cpp crypto/wildkeccak.cu \
 			  crypto/cryptolight.cu crypto/cryptolight-core.cu crypto/cryptolight-cpu.cpp \
 			  crypto/cryptonight.cu crypto/cryptonight-core.cu crypto/cryptonight-extra.cu \
@@ -76,7 +77,7 @@ ccminer_SOURCES	= elist.h miner.h compat.h \
 			  x13/hsr.cu x13/cuda_hsr_sm3.cu x13/sm3.c \
 			  x15/x14.cu x15/x15.cu x15/cuda_x14_shabal512.cu x15/cuda_x15_whirlpool.cu \
 			  x15/whirlpool.cu x15/cuda_x15_whirlpool_sm3.cu \
-			  x16r/x16r.cu x16r/cuda_x16_echo512.cu x16r/cuda_x16_fugue512.cu \
+			  x16r/x16r.cu x16r/cuda_x16_fugue512.cu \
 			  x16r/cuda_x16_shabal512.cu x16r/cuda_x16_simd512_80.cu \
 			  x17/x17.cu x17/hmq17.cu x17/cuda_x17_haval256.cu x17/cuda_x17_sha512.cu \
 			  x11/phi.cu x11/cuda_streebog_maxwell.cu \
@@ -114,7 +115,7 @@ endif
 ccminer_LDADD += -lcuda
 
 nvcc_ARCH :=
-#nvcc_ARCH += -gencode=arch=compute_61,code=\"sm_61,compute_61\"
+nvcc_ARCH += -gencode=arch=compute_61,code=\"sm_61,compute_61\"
 nvcc_ARCH += -gencode=arch=compute_52,code=\"sm_52,compute_52\"
 nvcc_ARCH += -gencode=arch=compute_50,code=\"sm_50,compute_50\"
 #nvcc_ARCH += -gencode=arch=compute_35,code=\"sm_35,compute_35\"

algos.h

@@ -59,8 +59,9 @@ enum sha_algos {
 	ALGO_X13,
 	ALGO_X14,
 	ALGO_X15,
-	ALGO_X16R,
-	ALGO_X17,
+    ALGO_X16R,
+    ALGO_X16S,
+    ALGO_X17,
 	ALGO_VANILLA,
 	ALGO_VELTOR,
 	ALGO_WHIRLCOIN,
@@ -129,8 +130,9 @@ static const char *algo_names[] = {
 	"x13",
 	"x14",
 	"x15",
-	"x16r",
-	"x17",
+    "x16r",
+    "x16s",
+    "x17",
 	"vanilla",
 	"veltor",
 	"whirlcoin",

ccminer.cpp

@@ -294,6 +294,7 @@ Options:\n\
 			x14         X14\n\
 			x15         X15\n\
 			x16r        X16R (Raven)\n\
+			x16s        X16S (Pigeon)\n\
 			x17         X17\n\
 			wildkeccak  Boolberry\n\
 			zr5         ZR5 (ZiftrCoin)\n\
@@ -1706,8 +1707,9 @@ static bool stratum_gen_work(struct stratum_ctx *sctx, struct work *work)
 		case ALGO_LYRA2Z:
 		case ALGO_TIMETRAVEL:
 		case ALGO_BITCORE:
-		case ALGO_X16R:
-			work_set_target(work, sctx->job.diff / (256.0 * opt_difficulty));
+        case ALGO_X16R:
+        case ALGO_X16S:
+            work_set_target(work, sctx->job.diff / (256.0 * opt_difficulty));
 			break;
 		case ALGO_KECCAK:
 		case ALGO_LYRA2:
@@ -2501,10 +2503,13 @@ static void *miner_thread(void *userdata)
 		case ALGO_X15:
 			rc = scanhash_x15(thr_id, &work, max_nonce, &hashes_done);
 			break;
-		case ALGO_X16R:
-			rc = scanhash_x16r(thr_id, &work, max_nonce, &hashes_done);
-			break;
-		case ALGO_X17:
+        case ALGO_X16R:
+            rc = scanhash_x16x(thr_id, &work, max_nonce, &hashes_done, 'r');
+            break;
+        case ALGO_X16S:
+            rc = scanhash_x16x(thr_id, &work, max_nonce, &hashes_done, 's');
+            break;
+        case ALGO_X17:
 			rc = scanhash_x17(thr_id, &work, max_nonce, &hashes_done);
 			break;
 		case ALGO_ZR5:
@@ -3866,7 +3871,7 @@ int main(int argc, char *argv[])
 	// get opt_quiet early
 	parse_single_opt('q', argc, argv);
 
-	printf("*** ccminer " PACKAGE_VERSION " for nVidia GPUs by tpruvot@github ***\n");
+	printf("*** " PACKAGE_NAME " " PACKAGE_VERSION " for nVidia GPUs by DumaxFr@github ***\n");
 	if (!opt_quiet) {
 		const char* arch = is_x64() ? "64-bits" : "32-bits";
 #ifdef _MSC_VER
@@ -3875,10 +3880,13 @@ int main(int argc, char *argv[])
 		printf("    Built with the nVidia CUDA Toolkit %d.%d %s\n\n",
 #endif
 			CUDART_VERSION/1000, (CUDART_VERSION % 1000)/10, arch);
-		printf("  Originally based on Christian Buchner and Christian H. project\n");
-		printf("  Include some kernels from alexis78, djm34, djEzo, tsiv and krnlx.\n\n");
-		printf("BTC donation address: 1AJdfCpLWPNoAMDfHF1wD5y8VgKSSTHxPo (tpruvot)\n\n");
-	}
+        printf("  Based on a tpruvot/ccminer fork\n");
+        printf("  Originally based on Christian Buchner and Christian H. project\n");
+        printf("  Include some kernels from alexis78, djm34, djEzo, tsiv and krnlx.\n\n");
+        printf("BTC donation address: 1AJdfCpLWPNoAMDfHF1wD5y8VgKSSTHxPo (tpruvot)\n\n");
+        printf("RVN donation address: RQXpsvSaVrGYo4tyGityWDNBQMFcnqANyj (DumaxFr)\n");
+        printf("BTC donation address: 1AtQXFbnzYTsjzy2bzSH6nPGxqZ32NG42T (DumaxFr)\n\n");
+    }
 
 	rpc_user = strdup("");
 	rpc_pass = strdup("");

compat/ccminer-config.h

@@ -161,10 +161,10 @@
 #define PACKAGE_NAME "ccminer"
 
 /* Define to the home page for this package. */
-#define PACKAGE_URL "http://github.com/tpruvot/ccminer"
+#define PACKAGE_URL "https://github.com/DumaxFr/ccminer"
 
 /* Define to the version of this package. */
-#define PACKAGE_VERSION "2.2.5"
+#define PACKAGE_VERSION "dumax-0.9.0"
 
 /* If using the C implementation of alloca, define if you know the
    direction of stack growth for your system; otherwise it will be

configure.ac

@@ -1,4 +1,4 @@
-AC_INIT([ccminer], [2.2.5], [], [ccminer], [http://github.com/tpruvot/ccminer])
+AC_INIT([ccminer], [dumax-0.9.0], [], [ccminer], [https://github.com/DumaxFr/ccminer])
 
 AC_PREREQ([2.59c])
 AC_CANONICAL_SYSTEM

miner.h

@@ -315,7 +315,7 @@ extern "C" {
     extern int scanhash_x13(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
     extern int scanhash_x14(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
     extern int scanhash_x15(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
-    extern int scanhash_x16r(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
+    extern int scanhash_x16x(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done, const char variation);
     extern int scanhash_x17(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done);
     extern int scanhash_zr5(int thr_id, struct work *work, uint32_t max_nonce, unsigned long *hashes_done);
 
@@ -925,7 +925,7 @@ extern "C" {
     void x13hash(void *output, const void *input);
     void x14hash(void *output, const void *input);
     void x15hash(void *output, const void *input);
-    void x16r_hash(void *output, const void *input);
+    void x16x_hash(void *output, const void *input, const char variation);
     void x17hash(void *output, const void *input);
     void wildkeccak_hash(void *output, const void *input, uint64_t* scratchpad, uint64_t ssize);
     void zr5hash(void *output, const void *input);

util.cpp

@@ -2325,7 +2325,7 @@ void print_hash_tests(void)
 	x15hash(&hash[0], &buf[0]);
 	printpfx("X15", hash);
 
-	x16r_hash(&hash[0], &buf[0]);
+	x16x_hash(&hash[0], &buf[0], 'r');
 	printpfx("X16r", hash);
 
 	x17hash(&hash[0], &buf[0]);

x16r/x16r.cu

@@ -79,7 +79,7 @@ static __thread uint32_t s_ntime = UINT32_MAX;
 static __thread bool s_implemented = false;
 static __thread char hashOrder[HASH_FUNC_COUNT + 1] = { 0 };
 
-static void getAlgoString(const uint32_t* prevblock, char *output) {
+static void getAlgoString16r(const uint32_t* prevblock, char *output) {
     char *sptr = output;
     uint8_t* data = (uint8_t*)prevblock;
 
@@ -95,8 +95,36 @@ static void getAlgoString(const uint32_t* prevblock, char *output) {
     *sptr = '\0';
 }
 
-// X16R CPU Hash (Validation)
-extern "C" void x16r_hash(void *output, const void *input) {
+static void getAlgoString16s(const uint32_t* prevblock, char *output) {
+
+    char *ptrHash = output;
+    uint8_t* data = (uint8_t*)prevblock;
+
+    strcpy(ptrHash, "0123456789ABCDEF");
+
+    for (int j = 0; j < HASH_FUNC_COUNT; j++) {
+        uint8_t b = (15 - j) >> 1;
+        uint8_t algoDigit = (j & 1) ? data[b] & 0xF : data[b] >> 4;
+        if (algoDigit > 0) {
+            char pull = ptrHash[algoDigit];
+            for (int i = algoDigit; i > 0; i--) {
+                ptrHash[i] = ptrHash[i-1];
+            }
+            ptrHash[0] = pull;
+        }
+    }
+
+}
+
+static void getAlgoString(const uint32_t* prevblock, char *output, const char baseAlgo) {
+    if (baseAlgo == 'r')
+        getAlgoString16r(prevblock, output);
+    else
+        getAlgoString16s(prevblock, output);
+}
+
+// X16x CPU Hash (Validation)
+extern "C" void x16x_hash(void *output, const void *input, const char variation) {
     unsigned char _ALIGN(64) hash[128];
 
     sph_blake512_context ctx_blake;
@@ -120,7 +148,7 @@ extern "C" void x16r_hash(void *output, const void *input) {
     int size = 80;
 
     uint32_t *in32 = (uint32_t*)input;
-    getAlgoString(&in32[1], hashOrder);
+    getAlgoString(&in32[1], hashOrder, variation);
 
     for (int i = 0; i < 16; i++) {
         const char elem = hashOrder[i];
@@ -247,7 +275,7 @@ static cudaEvent_t x16r_kernel_stop[MAX_GPUS];
 
 
 
-extern "C" int scanhash_x16r(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done) {
+extern "C" int scanhash_x16x(int thr_id, struct work* work, uint32_t max_nonce, unsigned long *hashes_done, const char variation) {
     uint32_t *pdata = work->data;
     uint32_t *ptarget = work->target;
     const uint32_t first_nonce = pdata[19];
@@ -308,7 +336,7 @@ extern "C" int scanhash_x16r(int thr_id, struct work* work, uint32_t max_nonce,
 
     uint32_t ntime = swab32(pdata[17]);
     if (s_ntime != ntime) {
-        getAlgoString(&endiandata[1], hashOrder);
+        getAlgoString(&endiandata[1], hashOrder, variation);
         s_ntime = ntime;
         s_implemented = true;
         if (opt_debug && !thr_id) {
@@ -592,15 +620,15 @@ extern "C" int scanhash_x16r(int thr_id, struct work* work, uint32_t max_nonce,
             const uint32_t Htarg = ptarget[7];
             uint32_t _ALIGN(64) vhash[8];
             be32enc(&endiandata[19], work->nonces[0]);
-            x16r_hash(vhash, endiandata);
+            x16x_hash(vhash, endiandata, variation);
 
             if (vhash[7] <= Htarg && fulltest(vhash, ptarget)) {
                 work->valid_nonces = 1;
                 work->nonces[1] = cuda_check_hash_suppl(thr_id, throughput, pdata[19], d_hash[thr_id], 1);
                 work_set_target_ratio(work, vhash);
                 if (work->nonces[1] != 0) {
                     be32enc(&endiandata[19], work->nonces[1]);
-                    x16r_hash(vhash, endiandata);
+                    x16x_hash(vhash, endiandata, variation);
                     bn_set_target_ratio(work, vhash, 1);
                     work->valid_nonces++;
                     pdata[19] = max(work->nonces[0], work->nonces[1]) + 1;