Add code to check the results from the metal backend. The t=0 time is not showing data correctly. (8466d10a) · Commits · Cianciosa, Mark / graph_framework

graph_framework/jit.hpp

+4 −2

Original line number	Diff line number	Diff line
		@@ -207,7 +207,8 @@ namespace jit {
		const size_t num_steps,
		const size_t num_rays) {
		GPU_CONTEXT context;
		context.create_pipeline(source_buffer.str(), name, inputs, num_rays);
		context.create_pipeline(source_buffer.str(), name, inputs, num_rays,
		num_steps, 0);

		const timeing::measure_diagnostic gpu_time("GPU Time");

		@@ -215,8 +216,9 @@ namespace jit {
		context.step();
		}
		context.wait();

		gpu_time.stop();

		context.print_results<BACKEND> (num_steps);
		}
		};
		}

graph_framework/metal_context.hpp

+69 −3

Original line number	Diff line number	Diff line
		@@ -34,6 +34,14 @@ namespace gpu {
		NSUInteger thread_groups;
		/// Number of threads in a group.
		NSUInteger threads_per_group;
		/// Result buffers.
		std::vector<id<MTLBuffer>> result_buffers;
		/// Index offset.
		size_t buffer_offset;
		/// Buffer element size.
		size_t buffer_element_size;
		/// Time offset.
		size_t time_offset;

		public:
		//------------------------------------------------------------------------------
		@@ -50,12 +58,16 @@ namespace gpu {
		/// @param[in] kernel_name Name of the kernel for later reference.
		/// @param[in] inputs Input nodes of the kernel.
		/// @param[in] num_rays Number of rays to trace.
		/// @param[in] num_times Number of times to record.
		/// @param[in] ray_index Index of the ray to save.
		//------------------------------------------------------------------------------
		template<class BACKEND>
		void create_pipeline(const std::string kernel_source,
		const std::string kernel_name,
		graph::input_nodes<BACKEND> inputs,
		const size_t num_rays) {
		const size_t num_rays,
		const size_t num_times,
		const size_t ray_index) {
		@autoreleasepool {
		MTLCompileOptions *options = [MTLCompileOptions new];
		options.fastMathEnabled = NO;
		@@ -86,10 +98,15 @@ namespace gpu {
		NSLog(@"%@", error);
		}

		buffer_element_size = sizeof(typename BACKEND::base);
		buffer_offset = ray_index*buffer_element_size;
		time_offset = 0;
		for (graph::shared_variable<BACKEND> &input : inputs) {
		const BACKEND backend = input->evaluate();
		buffers.push_back([device newBufferWithBytes:&backend[0]
		length:backend.size()*sizeof(typename BACKEND::base)
		length:backend.size()*buffer_element_size
		options:MTLResourceStorageModeManaged]);
		result_buffers.push_back([device newBufferWithLength:num_times*buffer_element_size
		options:MTLResourceStorageModeManaged]);
		}

		@@ -99,9 +116,29 @@ namespace gpu {
		std::cout << " Threads per group : " << threads_per_group << std::endl;
		std::cout << " Number of groups : " << thread_groups << std::endl;
		std::cout << " Total problem size : " << threads_per_group*thread_groups << std::endl;

		encode_blit();
		[command_buffer commit];
		}
		}

		//------------------------------------------------------------------------------
		/// @brief Encode a blit command.
		//------------------------------------------------------------------------------
		void encode_blit() {
		id<MTLBlitCommandEncoder> blit = [command_buffer blitCommandEncoder];
		for (size_t i = 0, ie = buffers.size(); i < ie; i++) {
		[blit copyFromBuffer:buffers[i]
		sourceOffset:buffer_offset
		toBuffer:result_buffers[i]
		destinationOffset:time_offset
		size:buffer_element_size];
		}
		[blit endEncoding];

		time_offset += buffer_element_size;
		}

		//------------------------------------------------------------------------------
		/// @brief Perform a time step.
		///
		@@ -122,6 +159,9 @@ namespace gpu {
		[encoder dispatchThreadgroups:MTLSizeMake(thread_groups, 1, 1)
		threadsPerThreadgroup:MTLSizeMake(threads_per_group, 1, 1)];
		[encoder endEncoding];

		encode_blit();

		[command_buffer commit];
		}
		}
		@@ -130,8 +170,34 @@ namespace gpu {
		/// @brief Hold the current thread until the current command buffer has complete.
		//------------------------------------------------------------------------------
		void wait() {
		command_buffer = [queue commandBuffer];
		id<MTLBlitCommandEncoder> blit = [command_buffer blitCommandEncoder];
		for (size_t i = 0, ie = buffers.size(); i < ie; i++) {
		[blit synchronizeResource:result_buffers[i]];
		}
		[blit endEncoding];

		[command_buffer commit];
		[command_buffer waitUntilCompleted];
		}

		//------------------------------------------------------------------------------
		/// @brief Print out the results.
		///
		/// @param[in] num_times Number of times to record.
		//------------------------------------------------------------------------------
		template<class BACKEND>
		void print_results(const size_t num_times) {
		for (size_t i = 0, ie = num_times + 1; i < ie; i++) {
		std::cout << i << " ";
		for (id<MTLBuffer> buffer : result_buffers) {
		const typename BACKEND::base contents = static_cast<typename BACKEND::base > ([buffer contents]);
		std::cout << contents[i] << " ";
		}
		std::cout << std::endl;
		}
		std::cout << std::endl;
		}
		};
		}

Admin message