Add improved Darkcoin (X11-mod) algorithm

algorithm.c

@@ -62,6 +62,21 @@ static cl_int queue_scrypt_kernel(struct __clState *clState, struct _dev_blk_ctx
     return status;
 }
 
+static cl_int queue_maxcoin_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
+{
+    cl_kernel *kernel = &clState->kernel;
+    unsigned int num = 0;
+    cl_int status = 0;
+
+    flip80(clState->cldata, blk->work->data);
+    status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL,NULL);
+
+    CL_SET_ARG(clState->CLbuffer0);
+    CL_SET_ARG(clState->outputBuffer);
+
+    return status;
+}
+
 static cl_int queue_sph_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
 {
     cl_kernel *kernel = &clState->kernel;
@@ -80,19 +95,66 @@ static cl_int queue_sph_kernel(struct __clState *clState, struct _dev_blk_ctx *b
     return status;
 }
 
-static cl_int queue_maxcoin_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
+static cl_int queue_darkcoin_mod_kernel(struct __clState *clState, struct _dev_blk_ctx *blk, __maybe_unused cl_uint threads)
 {
-    cl_kernel *kernel = &clState->kernel;
+  cl_kernel *kernel;
-    unsigned int num = 0;
+  unsigned int num;
-    cl_int status = 0;
+  cl_ulong le_target;
+  cl_int status = 0;
 
-    flip80(clState->cldata, blk->work->data);
+  le_target = *(cl_ulong *)(blk->work->device_target + 24);
-    status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL,NULL);
+  flip80(clState->cldata, blk->work->data);
+  status = clEnqueueWriteBuffer(clState->commandQueue, clState->CLbuffer0, true, 0, 80, clState->cldata, 0, NULL,NULL);
 
-    CL_SET_ARG(clState->CLbuffer0);
+  // blake - search
-    CL_SET_ARG(clState->outputBuffer);
+  kernel = &clState->kernel;
+  num = 0;
+  CL_SET_ARG(clState->CLbuffer0);
+  CL_SET_ARG(clState->padbuffer8);
+  // bmw - search1
+  kernel = clState->extra_kernels;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  // groestl - search2
+  kernel++;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  // skein - search3
+  kernel++;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  // jh - search4
+  kernel++;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  // keccak - search5
+  kernel++;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  // luffa - search6
+  kernel++;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  // cubehash - search7
+  kernel++;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  // shavite - search8
+  kernel++;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  // simd - search9
+  kernel++;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  // echo - search10
+  kernel++;
+  num = 0;
+  CL_SET_ARG(clState->padbuffer8);
+  CL_SET_ARG(clState->outputBuffer);
+  CL_SET_ARG(le_target);
 
-    return status;
+  return status;
 }
 
 typedef struct _algorithm_settings_t {
@@ -106,6 +168,9 @@ typedef struct _algorithm_settings_t {
     unsigned long long   diff_nonce;
     unsigned long long   diff_numerator;
     uint32_t diff1targ;
+    size_t n_extra_kernels;
+    long rw_buffer_size;
+    cl_command_queue_properties cq_properties;
     void     (*regenhash)(struct work *);
     cl_int   (*queue_kernel)(struct __clState *, struct _dev_blk_ctx *, cl_uint);
     void     (*gen_hash)(const unsigned char *, unsigned int, unsigned char *);
@@ -114,7 +179,7 @@ typedef struct _algorithm_settings_t {
 static algorithm_settings_t algos[] = {
     // kernels starting from this will have difficulty calculated by using litecoin algorithm
 #define A_SCRYPT(a) \
-    { a, 1, 65536, 65536, 0, 0, 0xFF, 0x0000ffff00000000ULL, 0xFFFFFFFFULL, 0x0000ffffUL, scrypt_regenhash, queue_scrypt_kernel, gen_hash}
+    { a, 1, 65536, 65536, 0, 0, 0xFF, 0x0000ffff00000000ULL, 0xFFFFFFFFULL, 0x0000ffffUL, 0, -1, CL_QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE, scrypt_regenhash, queue_scrypt_kernel, gen_hash}
     A_SCRYPT( "ckolivas" ),
     A_SCRYPT( "alexkarnew" ),
     A_SCRYPT( "alexkarnold" ),
@@ -125,7 +190,7 @@ static algorithm_settings_t algos[] = {
 
     // kernels starting from this will have difficulty calculated by using quarkcoin algorithm
 #define A_QUARK(a, b) \
-    { a, 256, 256, 256, 0, 0, 0xFF, 0x000000ffff000000ULL, 0xFFFFFFULL, 0x0000ffffUL, b, queue_sph_kernel, gen_hash}
+    { a, 256, 256, 256, 0, 0, 0xFF, 0x000000ffff000000ULL, 0xFFFFFFULL, 0x0000ffffUL, 0, 0, CL_QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE, b, queue_sph_kernel, gen_hash}
     A_QUARK( "quarkcoin", quarkcoin_regenhash),
     A_QUARK( "qubitcoin", qubitcoin_regenhash),
     A_QUARK( "animecoin", animecoin_regenhash),
@@ -134,25 +199,26 @@ static algorithm_settings_t algos[] = {
 
     // kernels starting from this will have difficulty calculated by using bitcoin algorithm
 #define A_DARK(a, b) \
-    { a, 1, 1, 1, 0, 0, 0xFF, 0x00000000ffff0000ULL, 0xFFFFULL, 0x0000ffffUL, b, queue_sph_kernel, gen_hash}
+    { a, 1, 1, 1, 0, 0, 0xFF, 0x00000000ffff0000ULL, 0xFFFFULL, 0x0000ffffUL, 0, 0, CL_QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE, b, queue_sph_kernel, gen_hash}
     A_DARK( "darkcoin",           darkcoin_regenhash),
     A_DARK( "inkcoin",            inkcoin_regenhash),
     A_DARK( "myriadcoin-groestl", myriadcoin_groestl_regenhash),
     A_DARK( "marucoin",           marucoin_regenhash),
 #undef A_DARK
 
-    { "twecoin", 1, 1, 1, 0, 0, 0xFF, 0x00000000ffff0000ULL, 0xFFFFULL, 0x0000ffffUL, twecoin_regenhash, queue_sph_kernel, sha256},
+    { "twecoin", 1, 1, 1, 0, 0, 0xFF, 0x00000000ffff0000ULL, 0xFFFFULL, 0x0000ffffUL, 0, 0, CL_QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE, twecoin_regenhash, queue_sph_kernel, sha256},
-    { "maxcoin", 1, 256, 1, 4, 15, 0x0F, 0x00000000ffff0000ULL, 0xFFFFULL, 0x000000ffUL, maxcoin_regenhash, queue_maxcoin_kernel, sha256},
+    { "maxcoin", 1, 256, 1, 4, 15, 0x0F, 0x00000000ffff0000ULL, 0xFFFFULL, 0x000000ffUL, 0, 0, CL_QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE, maxcoin_regenhash, queue_maxcoin_kernel, sha256},
+    { "darkcoin-mod", 1, 1, 1, 0, 0, 0xFF, 0x00000000ffff0000ULL, 0xFFFFULL, 0x0000ffffUL, 10, 8 * 16 * 4194304, 0, darkcoin_regenhash, queue_darkcoin_mod_kernel, gen_hash},
 
     // kernels starting from this will have difficulty calculated by using fuguecoin algorithm
 #define A_FUGUE(a, b) \
-    { a, 1, 256, 256, 0, 0, 0xFF, 0x00000000ffff0000ULL, 0xFFFFULL, 0x0000ffffUL, b, queue_sph_kernel, sha256}
+    { a, 1, 256, 256, 0, 0, 0xFF, 0x00000000ffff0000ULL, 0xFFFFULL, 0x0000ffffUL, 0, 0, CL_QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE, b, queue_sph_kernel, sha256}
     A_FUGUE( "fuguecoin",   fuguecoin_regenhash),
     A_FUGUE( "groestlcoin", groestlcoin_regenhash),
 #undef A_FUGUE
 
     // Terminator (do not remove)
-    { NULL, 0, 0, 0, 0, 0, 0, 0, 0, 0, NULL, NULL, NULL}
+    { NULL, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, NULL, NULL, NULL}
 };
 
 void copy_algorithm_settings(algorithm_t* dest, const char* algo) {
@@ -172,6 +238,9 @@ void copy_algorithm_settings(algorithm_t* dest, const char* algo) {
             dest->diff_nonce = src->diff_nonce;
             dest->diff_numerator = src->diff_numerator;
             dest->diff1targ = src->diff1targ;
+            dest->n_extra_kernels = src->n_extra_kernels;
+            dest->rw_buffer_size = src->rw_buffer_size;
+            dest->cq_properties = src->cq_properties;
             dest->regenhash = src->regenhash;
             dest->queue_kernel = src->queue_kernel;
             dest->gen_hash = src->gen_hash;

algorithm.h

@@ -32,6 +32,9 @@ typedef struct _algorithm_t {
     unsigned long long   diff_nonce;
     unsigned long long   diff_numerator;
     uint32_t diff1targ;
+    size_t n_extra_kernels;
+    long rw_buffer_size;
+    cl_command_queue_properties cq_properties;
     void     (*regenhash)(struct work *);
     cl_int   (*queue_kernel)(struct __clState *, struct _dev_blk_ctx *, cl_uint);
     void     (*gen_hash)(const unsigned char *, unsigned int, unsigned char *);

driver-opencl.c

@@ -1304,7 +1304,6 @@ static int64_t opencl_scanhash(struct thr_info *thr, struct work *work,
 	struct opencl_thread_data *thrdata = (struct opencl_thread_data *)thr->cgpu_data;
 	struct cgpu_info *gpu = thr->cgpu;
 	_clState *clState = clStates[thr_id];
-	const cl_kernel *kernel = &clState->kernel;
 	const int dynamic_us = opt_dynamic_interval * 1000;
 
 	cl_int status;
@@ -1314,6 +1313,7 @@ static int64_t opencl_scanhash(struct thr_info *thr, struct work *work,
 	int64_t hashes;
 	int found = gpu->algorithm.found_idx;
 	int buffersize = BUFFERSIZE;
+    unsigned int i;
 
 	/* Windows' timer resolution is only 15ms so oversample 5x */
 	if (gpu->dynamic && (++gpu->intervals * dynamic_us) > 70000) {
@@ -1346,13 +1346,23 @@ static int64_t opencl_scanhash(struct thr_info *thr, struct work *work,
 
     if (clState->goffset)
         p_global_work_offset = &work->blk.nonce;
-    status = clEnqueueNDRangeKernel(clState->commandQueue, *kernel, 1, p_global_work_offset,
+
+    status = clEnqueueNDRangeKernel(clState->commandQueue, clState->kernel, 1, p_global_work_offset,
                     globalThreads, localThreads, 0,  NULL, NULL);
 	if (unlikely(status != CL_SUCCESS)) {
 		applog(LOG_ERR, "Error %d: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel)", status);
 		return -1;
 	}
 
+  for (i = 0; i < clState->n_extra_kernels; i++) {
+      status = clEnqueueNDRangeKernel(clState->commandQueue, clState->extra_kernels[i], 1, p_global_work_offset,
+                      globalThreads, localThreads, 0,  NULL, NULL);
+      if (unlikely(status != CL_SUCCESS)) {
+          applog(LOG_ERR, "Error %d: Enqueueing kernel onto command queue. (clEnqueueNDRangeKernel)", status);
+          return -1;
+      }
+  }
+
 	status = clEnqueueReadBuffer(clState->commandQueue, clState->outputBuffer, CL_FALSE, 0,
 				     buffersize, thrdata->res, 0, NULL, NULL);
 	if (unlikely(status != CL_SUCCESS)) {
@@ -1393,14 +1403,19 @@ static void opencl_thread_shutdown(struct thr_info *thr)
 {
 	const int thr_id = thr->id;
 	_clState *clState = clStates[thr_id];
+    cl_kernel *kernel = clState->extra_kernels;
 	clStates[thr_id] = NULL;
+    unsigned int i;
 
 	if (clState) {
 		clFinish(clState->commandQueue);
 		clReleaseMemObject(clState->outputBuffer);
 		clReleaseMemObject(clState->CLbuffer0);
-		clReleaseMemObject(clState->padbuffer8);
+		if (clState->padbuffer8)
+      clReleaseMemObject(clState->padbuffer8);
 		clReleaseKernel(clState->kernel);
+        for (i = 0; i < clState->n_extra_kernels; i++)
+            clReleaseKernel(clState->extra_kernels[i]);
 		clReleaseProgram(clState->program);
 		clReleaseCommandQueue(clState->commandQueue);
 		clReleaseContext(clState->context);

ocl.c

@@ -329,7 +329,7 @@ _clState *initCl(unsigned int gpu, char *name, size_t nameSize, algorithm_t *alg
 	// Create an OpenCL command queue
 	/////////////////////////////////////////////////////////////////
 	clState->commandQueue = clCreateCommandQueue(clState->context, devices[gpu],
-						     CL_QUEUE_OUT_OF_ORDER_EXEC_MODE_ENABLE, &status);
+						     cgpu->algorithm.cq_properties, &status);
 	if (status != CL_SUCCESS) /* Try again without OOE enable */
 		clState->commandQueue = clCreateCommandQueue(clState->context, devices[gpu], 0 , &status);
 	if (status != CL_SUCCESS) {
@@ -760,29 +760,56 @@ built:
 		return NULL;
 	}
 
-	size_t ipt = (algorithm->n / cgpu->lookup_gap +
-		      (algorithm->n % cgpu->lookup_gap > 0));
-	size_t bufsize = 128 * ipt * cgpu->thread_concurrency;
 
-	/* Use the max alloc value which has been rounded to a power of
+  clState->n_extra_kernels = algorithm->n_extra_kernels;
-	 * 2 greater >= required amount earlier */
+  if (clState->n_extra_kernels > 0) {
-	if (bufsize > cgpu->max_alloc) {
+    unsigned int i;
-		applog(LOG_WARNING, "Maximum buffer memory device %d supports says %lu",
+    char *kernel_name = (char *)malloc(9); // max: search99 + 0x0
-			   gpu, (unsigned long)(cgpu->max_alloc));
-		applog(LOG_WARNING, "Your scrypt settings come to %lu", (unsigned long)bufsize);
-	}
-	applog(LOG_DEBUG, "Creating scrypt buffer sized %lu", (unsigned long)bufsize);
-	clState->padbufsize = bufsize;
 
-	/* This buffer is weird and might work to some degree even if
+    clState->extra_kernels = (cl_kernel *)malloc(sizeof(cl_kernel) * clState->n_extra_kernels);
-	 * the create buffer call has apparently failed, so check if we
+
-	 * get anything back before we call it a failure. */
+    for (i = 0; i < clState->n_extra_kernels; i++) {
-	clState->padbuffer8 = NULL;
+      snprintf(kernel_name, 9, "%s%d", "search", i + 1);
-	clState->padbuffer8 = clCreateBuffer(clState->context, CL_MEM_READ_WRITE, bufsize, NULL, &status);
+      clState->extra_kernels[i] = clCreateKernel(clState->program, kernel_name, &status);
-	if (status != CL_SUCCESS && !clState->padbuffer8) {
+      if (status != CL_SUCCESS) {
-		applog(LOG_ERR, "Error %d: clCreateBuffer (padbuffer8), decrease TC or increase LG", status);
+        applog(LOG_ERR, "Error %d: Creating ExtraKernel #%d from program. (clCreateKernel)", status, i);
-		return NULL;
+        return NULL;
-	}
+      }
+    }
+
+    free(kernel_name);
+  }
+
+	size_t bufsize;
+
+  if (algorithm->rw_buffer_size < 0) {
+    size_t ipt = (algorithm->n / cgpu->lookup_gap +
+            (algorithm->n % cgpu->lookup_gap > 0));
+    bufsize = 128 * ipt * cgpu->thread_concurrency;
+  } else
+    bufsize = (size_t) algorithm->rw_buffer_size;
+
+  clState->padbuffer8 = NULL;
+
+  if (bufsize > 0) {
+    /* Use the max alloc value which has been rounded to a power of
+     * 2 greater >= required amount earlier */
+    if (bufsize > cgpu->max_alloc) {
+      applog(LOG_WARNING, "Maximum buffer memory device %d supports says %lu",
+           gpu, (unsigned long)(cgpu->max_alloc));
+      applog(LOG_WARNING, "Your settings come to %lu", (unsigned long)bufsize);
+    }
+    applog(LOG_DEBUG, "Creating buffer sized %lu", (unsigned long)bufsize);
+
+    /* This buffer is weird and might work to some degree even if
+     * the create buffer call has apparently failed, so check if we
+     * get anything back before we call it a failure. */
+    clState->padbuffer8 = clCreateBuffer(clState->context, CL_MEM_READ_WRITE, bufsize, NULL, &status);
+    if (status != CL_SUCCESS && !clState->padbuffer8) {
+      applog(LOG_ERR, "Error %d: clCreateBuffer (padbuffer8), decrease TC or increase LG", status);
+      return NULL;
+    }
+  }
 
 	clState->CLbuffer0 = clCreateBuffer(clState->context, CL_MEM_READ_ONLY, 128, NULL, &status);
 	if (status != CL_SUCCESS) {

ocl.h

@@ -15,12 +15,13 @@
 typedef struct __clState {
 	cl_context context;
 	cl_kernel kernel;
+	cl_kernel *extra_kernels;
+	size_t n_extra_kernels;
 	cl_command_queue commandQueue;
 	cl_program program;
 	cl_mem outputBuffer;
 	cl_mem CLbuffer0;
 	cl_mem padbuffer8;
-	size_t padbufsize;
 	unsigned char cldata[80];
 	bool hasBitAlign;
 	bool hasOpenCL11plus;