Merge pull request #646 from rushstrike/master

split CUDA/CL localWorkSize,globalWorkSize vs cudaBlockSize,cudaGridSize

Author: jean-m-cyr

ethminer/MinerAux.h

@@ -281,7 +281,7 @@ class MinerCLI
 				if(m_openclThreadsPerHash != 1 && m_openclThreadsPerHash != 2 &&
 				   m_openclThreadsPerHash != 4 && m_openclThreadsPerHash != 8) {
 					BOOST_THROW_EXCEPTION(BadArgument());
-				} 
+				}
 			}
 			catch(...) {
 				cerr << "Bad " << arg << " option: " << argv[i] << endl;
@@ -300,30 +300,55 @@ class MinerCLI
 				BOOST_THROW_EXCEPTION(BadArgument());
 			}
 		}
-#endif
-#if ETH_ETHASHCL || ETH_ETHASHCUDA
-		else if ((arg == "--cl-global-work" || arg == "--cuda-grid-size")  && i + 1 < argc)
-			try {
+		else if ( arg == "--cl-global-work"  && i + 1 < argc)
+		{
+			try
+			{
 				m_globalWorkSizeMultiplier = stol(argv[++i]);
+
 			}
 			catch (...)
 			{
 				cerr << "Bad " << arg << " option: " << argv[i] << endl;
 				BOOST_THROW_EXCEPTION(BadArgument());
 			}
-		else if ((arg == "--cl-local-work" || arg == "--cuda-block-size") && i + 1 < argc)
-			try {
-				m_localWorkSize = stol(argv[++i]);
+		}
+		else if ( arg == "--cl-local-work" && i + 1 < argc)
+			try
+			{
+					m_localWorkSize = stol(argv[++i]);
 			}
 			catch (...)
 			{
 				cerr << "Bad " << arg << " option: " << argv[i] << endl;
 				BOOST_THROW_EXCEPTION(BadArgument());
 			}
+#endif
+#if ETH_ETHASHCL || ETH_ETHASHCUDA
 		else if (arg == "--list-devices")
 			m_shouldListDevices = true;
 #endif
 #if ETH_ETHASHCUDA
+		else if ( arg == "--cuda-grid-size" && i + 1 < argc)
+			try
+			{
+				m_cudaGridSize = stol(argv[++i]);
+			}
+			catch (...)
+			{
+				cerr << "Bad " << arg << " option: " << argv[i] << endl;
+				BOOST_THROW_EXCEPTION(BadArgument());
+			}
+		else if ( arg == "--cuda-block-size" && i + 1 < argc)
+			try
+			{
+				m_cudaBlockSize= stol(argv[++i]);
+			}
+			catch (...)
+			{
+				cerr << "Bad " << arg << " option: " << argv[i] << endl;
+				BOOST_THROW_EXCEPTION(BadArgument());
+			}
 		else if (arg == "--cuda-devices")
 		{
 			while (m_cudaDeviceCount < 16 && i + 1 < argc)
@@ -508,7 +533,7 @@ class MinerCLI
 				CLMiner::setDevices(m_openclDevices, m_openclDeviceCount);
 				m_miningThreads = m_openclDeviceCount;
 			}
-			
+
 			CLMiner::setCLKernel(m_openclSelectedKernel);
 			CLMiner::setThreadsPerHash(m_openclThreadsPerHash);
 
@@ -538,8 +563,8 @@ class MinerCLI
 
 			CUDAMiner::setNumInstances(m_miningThreads);
 			if (!CUDAMiner::configureGPU(
-				m_localWorkSize,
-				m_globalWorkSizeMultiplier,
+				m_cudaBlockSize,
+				m_cudaGridSize,
 				m_numStreams,
 				m_cudaSchedule,
 				0,
@@ -803,11 +828,11 @@ class MinerCLI
 		h256 id = h256::random();
 		Farm f;
 		f.set_pool_addresses(m_farmURL, m_port, m_farmFailOverURL, m_fport);
-		
+
 #if API_CORE
 		Api api(this->m_api_port, f);
 #endif
-		
+
 		f.setSealers(sealers);
 
 		if (_m == MinerType::CL)
@@ -941,11 +966,11 @@ class MinerCLI
 
 		Farm f;
 		f.set_pool_addresses(m_farmURL, m_port, m_farmFailOverURL, m_fport);
-		
+
 #if API_CORE
 		Api api(this->m_api_port, f);
 #endif
-	
+
 		// this is very ugly, but if Stratum Client V2 tunrs out to be a success, V1 will be completely removed anyway
 		if (m_stratumClientVersion == 1) {
 			EthStratumClient client(&f, m_minerType, m_farmURL, m_port, m_user, m_pass, m_maxFarmRetries, m_worktimeout, m_stratumProtocol, m_email);
@@ -972,7 +997,7 @@ class MinerCLI
 				}
 				return false;
 			});
-			f.onMinerRestart([&](){ 
+			f.onMinerRestart([&](){
 				client.reconnect();
 			});
 
@@ -992,7 +1017,7 @@ class MinerCLI
 					{
 						minelog << "Waiting for work package...";
 					}
-					
+
 					if (this->m_report_stratum_hashrate) {
 						auto rate = mp.rate();
 						client.submitHashrate(toJS(rate));
@@ -1021,7 +1046,7 @@ class MinerCLI
 				client.submit(sol);
 				return false;
 			});
-			f.onMinerRestart([&](){ 
+			f.onMinerRestart([&](){
 				client.reconnect();
 			});
 
@@ -1041,7 +1066,7 @@ class MinerCLI
 					{
 						minelog << "Waiting for work package...";
 					}
-					
+
 					if (this->m_report_stratum_hashrate) {
 						auto rate = mp.rate();
 						client.submitHashrate(toJS(rate));
@@ -1067,18 +1092,16 @@ class MinerCLI
 	unsigned m_openclDeviceCount = 0;
 	unsigned m_openclDevices[16];
 	unsigned m_openclThreadsPerHash = 8;
-#if !ETH_ETHASHCUDA
 	unsigned m_globalWorkSizeMultiplier = CLMiner::c_defaultGlobalWorkSizeMultiplier;
 	unsigned m_localWorkSize = CLMiner::c_defaultLocalWorkSize;
 #endif
-#endif
 #if ETH_ETHASHCUDA
-	unsigned m_globalWorkSizeMultiplier = CUDAMiner::c_defaultGridSize;
-	unsigned m_localWorkSize = CUDAMiner::c_defaultBlockSize;
 	unsigned m_cudaDeviceCount = 0;
 	unsigned m_cudaDevices[16];
 	unsigned m_numStreams = CUDAMiner::c_defaultNumStreams;
 	unsigned m_cudaSchedule = 4; // sync
+	unsigned m_cudaGridSize = CUDAMiner::c_defaultGridSize;
+	unsigned m_cudaBlockSize = CUDAMiner::c_defaultBlockSize;
 #endif
 	unsigned m_dagLoadMode = 0; // parallel
 	unsigned m_dagCreateDevice = 0;
@@ -1103,7 +1126,7 @@ class MinerCLI
 	bool m_show_hwmonitors = false;
 #if API_CORE
 	int m_api_port = 0;
-#endif	
+#endif
 
 	bool m_report_stratum_hashrate = false;
 	int m_stratumClientVersion = 1;