testing Monero POW v2 fireice-uk#1831 (as of 20180919)

CMakeLists.txt

@@ -445,6 +445,26 @@ if(CMAKE_LINK_STATIC)
     endif()
 endif()
 
+if(CMAKE_C_COMPILER_ID MATCHES "MSVC")
+    # asm optimized monero v8 code
+    enable_language(ASM_MASM)
+    set_property(SOURCE "xmrstak/backend/cpu/crypto/asm/cryptonight_v8_main_loop.asm" PROPERTY ASM_MASM)
+    add_library(xmr-stak-asm
+        STATIC
+        "xmrstak/backend/cpu/crypto/asm/cryptonight_v8_main_loop.asm"
+    )
+else()
+    # asm optimized monero v8 code
+    enable_language(ASM)
+    set_property(SOURCE "xmrstak/backend/cpu/crypto/asm/cryptonight_v8_main_loop.S" PROPERTY C)
+    add_library(xmr-stak-asm
+        STATIC
+        "xmrstak/backend/cpu/crypto/asm/cryptonight_v8_main_loop.S"
+    )
+endif()
+
+set_property(TARGET xmr-stak-asm PROPERTY LINKER_LANGUAGE C)
+
 # compile C files
 file(GLOB SRCFILES_C "xmrstak/backend/cpu/crypto/*.c")
 
@@ -456,7 +476,7 @@ set_property(TARGET xmr-stak-c PROPERTY C_STANDARD 99)
 if(MICROHTTPD_ENABLE)
     target_link_libraries(xmr-stak-c ${MHTD})
 endif()
-target_link_libraries(xmr-stak-c ${LIBS})
+target_link_libraries(xmr-stak-c ${LIBS} xmr-stak-asm)
 
 # compile generic backend files
 file(GLOB BACKEND_CPP
@@ -472,7 +492,7 @@ add_library(xmr-stak-backend
     STATIC
     ${BACKEND_CPP}
 )
-target_link_libraries(xmr-stak-backend xmr-stak-c ${CMAKE_DL_LIBS})
+target_link_libraries(xmr-stak-backend xmr-stak-c ${CMAKE_DL_LIBS} xmr-stak-asm)
 
 # compile CUDA backend
 if(CUDA_FOUND)
@@ -499,7 +519,7 @@ if(CUDA_FOUND)
         )
     endif()
     target_link_libraries(xmrstak_cuda_backend ${CUDA_LIBRARIES})
-    target_link_libraries(xmrstak_cuda_backend xmr-stak-backend)
+    target_link_libraries(xmrstak_cuda_backend xmr-stak-backend xmr-stak-asm)
 endif()
 
 # compile AMD backend
@@ -512,7 +532,7 @@ if(OpenCL_FOUND)
         ${OPENCLSRCFILES}
     )
     target_link_libraries(xmrstak_opencl_backend ${OpenCL_LIBRARY} )
-    target_link_libraries(xmrstak_opencl_backend xmr-stak-backend)
+    target_link_libraries(xmrstak_opencl_backend xmr-stak-backend xmr-stak-asm)
 endif()
 
 # compile final binary
@@ -528,7 +548,7 @@ endif()
 set(EXECUTABLE_OUTPUT_PATH "bin" CACHE STRING "Path to place executables relative to ${CMAKE_INSTALL_PREFIX}")
 set(LIBRARY_OUTPUT_PATH "bin" CACHE STRING "Path to place libraries relative to ${CMAKE_INSTALL_PREFIX}")
 
-target_link_libraries(xmr-stak ${LIBS} xmr-stak-c xmr-stak-backend)
+target_link_libraries(xmr-stak ${LIBS} xmr-stak-c xmr-stak-backend xmr-stak-asm)
 
 ################################################################################
 # Install
@@ -559,4 +579,4 @@ if( NOT CMAKE_INSTALL_PREFIX STREQUAL PROJECT_BINARY_DIR )
 else()
     # this rule is used if the install prefix is the build directory
     install(CODE "MESSAGE(\"xmr-stak installed to folder 'bin'\")")
-endif()
+endif()

README.md

@@ -60,6 +60,7 @@ If your prefered coin is not listed, you can choose one of the following algorit
     - cryptonight_masari
     - cryptonight_v7
     - cryptonight_v7_stellite
+    - cryptonight_v8
 - 4MiB scratchpad memory
     - cryptonight_haven
     - cryptonight_heavy

xmrstak/backend/amd/amd_gpu/gpu.cpp

@@ -405,6 +405,7 @@ size_t InitOpenCLGpu(cl_context opencl_ctx, GpuContext* ctx, const char* source_
 		options += " -DCOMP_MODE=" + std::to_string(ctx->compMode ? 1u : 0u);
 		options += " -DMEMORY=" + std::to_string(hashMemSize);
 		options += " -DALGO=" + std::to_string(miner_algo[ii]);
+		options += " -DCN_UNROLL=" + std::to_string(ctx->unroll);
 
 		/* create a hash for the compile time cache
 		 * used data:
@@ -901,6 +902,9 @@ size_t InitOpenCL(GpuContext* ctx, size_t num_gpus, size_t platform_idx)
 
 	//char* source_code = LoadTextFile(sSourcePath);
 
+	const char *fastIntMathV2CL =
+			#include "./opencl/fast_int_math_v2.cl"
+	;
 	const char *cryptonightCL =
 			#include "./opencl/cryptonight.cl"
 	;
@@ -921,6 +925,7 @@ size_t InitOpenCL(GpuContext* ctx, size_t num_gpus, size_t platform_idx)
 	;
 
 	std::string source_code(cryptonightCL);
+	source_code = std::regex_replace(source_code, std::regex("XMRSTAK_INCLUDE_FAST_INT_MATH_V2"), fastIntMathV2CL);
 	source_code = std::regex_replace(source_code, std::regex("XMRSTAK_INCLUDE_WOLF_AES"), wolfAesCL);
 	source_code = std::regex_replace(source_code, std::regex("XMRSTAK_INCLUDE_WOLF_SKEIN"), wolfSkeinCL);
 	source_code = std::regex_replace(source_code, std::regex("XMRSTAK_INCLUDE_JH"), jhCL);
@@ -930,16 +935,37 @@ size_t InitOpenCL(GpuContext* ctx, size_t num_gpus, size_t platform_idx)
 	// create a directory  for the OpenCL compile cache
 	create_directory(get_home() + "/.openclcache");
 
+	// check if cryptonight_monero_v8 is selected for the user or dev pool
+	bool useCryptonight_v8 =
+		::jconf::inst()->GetCurrentCoinSelection().GetDescription(1).GetMiningAlgo() == cryptonight_monero_v8 ||
+		::jconf::inst()->GetCurrentCoinSelection().GetDescription(1).GetMiningAlgoRoot() == cryptonight_monero_v8 ||
+		::jconf::inst()->GetCurrentCoinSelection().GetDescription(0).GetMiningAlgo() == cryptonight_monero_v8 ||
+		::jconf::inst()->GetCurrentCoinSelection().GetDescription(0).GetMiningAlgoRoot() == cryptonight_monero_v8;
+
 	for(int i = 0; i < num_gpus; ++i)
 	{
+		const std::string backendName = xmrstak::params::inst().openCLVendor;
 		if(ctx[i].stridedIndex == 2 && (ctx[i].rawIntensity % ctx[i].workSize) != 0)
 		{
 			size_t reduced_intensity = (ctx[i].rawIntensity / ctx[i].workSize) * ctx[i].workSize;
 			ctx[i].rawIntensity = reduced_intensity;
-			const std::string backendName = xmrstak::params::inst().openCLVendor;
 			printer::inst()->print_msg(L0, "WARNING %s: gpu %d intensity is not a multiple of 'worksize', auto reduce intensity to %d", backendName.c_str(), ctx[i].deviceIdx, int(reduced_intensity));
 		}
 
+		if(useCryptonight_v8)
+		{
+			if(ctx[i].stridedIndex == 1)
+			{
+				printer::inst()->print_msg(L0, "ERROR %s: gpu %d stridedIndex is not allowed to be `true` or `1` for the selected currency", backendName.c_str(), ctx[i].deviceIdx);
+				return ERR_STUPID_PARAMS;
+			}
+			if(ctx[i].stridedIndex == 2 && ctx[i].memChunk < 2)
+			{
+				printer::inst()->print_msg(L0, "ERROR %s: gpu %d memChunk bust be >= 2 for the selected currency", backendName.c_str(), ctx[i].deviceIdx);
+				return ERR_STUPID_PARAMS;
+			}
+		}
+
 		if((ret = InitOpenCLGpu(opencl_ctx, &ctx[i], source_code.c_str())) != ERR_SUCCESS)
 		{
 			return ret;

xmrstak/backend/amd/amd_gpu/gpu.hpp

@@ -27,6 +27,7 @@ struct GpuContext
 	size_t workSize;
 	int stridedIndex;
 	int memChunk;
+	int unroll = 0;
 	bool isNVIDIA = false;
 	int compMode;