bitonic sort sample

13_BitonicSort/CMakeLists.txt

@@ -0,0 +1,24 @@
+include(common RESULT_VARIABLE RES)
+if(NOT RES)
+	message(FATAL_ERROR "common.cmake not found. Should be in {repo_root}/cmake directory")
+endif()
+
+nbl_create_executable_project("" "" "" "" "${NBL_EXECUTABLE_PROJECT_CREATION_PCH_TARGET}")
+
+if(NBL_EMBED_BUILTIN_RESOURCES)
+	set(_BR_TARGET_ ${EXECUTABLE_NAME}_builtinResourceData)
+	set(RESOURCE_DIR "app_resources")
+
+	get_filename_component(_SEARCH_DIRECTORIES_ "${CMAKE_CURRENT_SOURCE_DIR}" ABSOLUTE)
+	get_filename_component(_OUTPUT_DIRECTORY_SOURCE_ "${CMAKE_CURRENT_BINARY_DIR}/src" ABSOLUTE)
+	get_filename_component(_OUTPUT_DIRECTORY_HEADER_ "${CMAKE_CURRENT_BINARY_DIR}/include" ABSOLUTE)
+
+    file(GLOB_RECURSE BUILTIN_RESOURCE_FILES RELATIVE "${CMAKE_CURRENT_SOURCE_DIR}/${RESOURCE_DIR}" "${CMAKE_CURRENT_SOURCE_DIR}/${RESOURCE_DIR}/*")
+    foreach(RES_FILE ${BUILTIN_RESOURCE_FILES})
+      LIST_BUILTIN_RESOURCE(RESOURCES_TO_EMBED "${RES_FILE}")
+    endforeach()
+
+	ADD_CUSTOM_BUILTIN_RESOURCES(${_BR_TARGET_} RESOURCES_TO_EMBED "${_SEARCH_DIRECTORIES_}" "${RESOURCE_DIR}" "nbl::this_example::builtin" "${_OUTPUT_DIRECTORY_HEADER_}" "${_OUTPUT_DIRECTORY_SOURCE_}")
+
+	LINK_BUILTIN_RESOURCES_TO_TARGET(${EXECUTABLE_NAME} ${_BR_TARGET_})
+endif()

13_BitonicSort/app_resources/bitonic_sort_shader.comp.hlsl

@@ -0,0 +1,87 @@
+#include "common.hlsl"
+#include "nbl/builtin/hlsl/workgroup/basic.hlsl"
+#include "nbl/builtin/hlsl/workgroup/bitonic_sort.hlsl"
+
+[[vk::push_constant]] PushConstantData pushConstants;
+
+using namespace nbl::hlsl;
+
+using BitonicSortConfig = workgroup::bitonic_sort::bitonic_sort_config<ElementsPerThreadLog2, WorkgroupSizeLog2, uint32_t, uint32_t, less<uint32_t> >;
+
+NBL_CONSTEXPR uint32_t WorkgroupSize = BitonicSortConfig::WorkgroupSize;
+
+groupshared uint32_t sharedmem[2 * WorkgroupSize];
+
+uint32_t3 glsl::gl_WorkGroupSize() { return uint32_t3(uint32_t(BitonicSortConfig::WorkgroupSize), 1, 1); }
+
+struct SharedMemoryAccessor
+{
+	template <typename AccessType, typename IndexType>
+	void set(IndexType idx, AccessType value)
+	{
+		sharedmem[idx] = value;
+	}
+
+	template <typename AccessType, typename IndexType>
+	void get(IndexType idx, NBL_REF_ARG(AccessType) value)
+	{
+		value = sharedmem[idx];
+	}
+
+	void workgroupExecutionAndMemoryBarrier()
+	{
+		glsl::barrier();
+	}
+
+};
+
+struct Accessor
+{
+	static Accessor create(const uint64_t address)
+	{
+		Accessor accessor;
+		accessor.address = address;
+		return accessor;
+	}
+
+	template <typename AccessType, typename IndexType>
+	void get(const IndexType index, NBL_REF_ARG(AccessType) value)
+	{
+		value = vk::RawBufferLoad<AccessType>(address + index * sizeof(AccessType));
+	}
+
+	template <typename AccessType, typename IndexType>
+	void set(const IndexType index, const AccessType value)
+	{
+		vk::RawBufferStore<AccessType>(address + index * sizeof(AccessType), value);
+	}
+
+	uint64_t address;
+};
+
+[numthreads(BitonicSortConfig::WorkgroupSize, 1, 1)]
+[shader("compute")]
+void main()
+{
+	Accessor accessor = Accessor::create(pushConstants.deviceBufferAddress);
+	SharedMemoryAccessor sharedmemAccessor;
+
+	const uint32_t threadID = glsl::gl_LocalInvocationID().x;
+
+	// Each thread handles 2 ADJACENT elements: lo and hi
+	// Following bitonic sort pattern: thread i handles elements [2*i] and [2*i + 1]
+	const uint32_t loIdx = threadID * 2;
+	const uint32_t hiIdx = threadID * 2 + 1;
+
+	uint32_t loKey, hiKey;
+	accessor.get(loIdx, loKey);
+	accessor.get(hiIdx, hiKey);
+
+	uint32_t loVal = loIdx;
+	uint32_t hiVal = hiIdx;
+
+	workgroup::BitonicSort<BitonicSortConfig>::template __call<Accessor, SharedMemoryAccessor>(accessor, sharedmemAccessor, loKey, hiKey, loVal, hiVal);
+
+	accessor.set(loIdx, loKey);
+	accessor.set(hiIdx, hiKey);
+}

13_BitonicSort/app_resources/common.hlsl

@@ -0,0 +1,13 @@
+#ifndef _BITONIC_SORT_COMMON_INCLUDED_
+#define _BITONIC_SORT_COMMON_INCLUDED_
+#include "nbl/builtin/hlsl/cpp_compat.hlsl"
+
+struct PushConstantData
+{
+	uint64_t deviceBufferAddress;
+};
+
+NBL_CONSTEXPR uint32_t WorkgroupSizeLog2 = 7;  // 128 threads
+NBL_CONSTEXPR uint32_t ElementsPerThreadLog2 = 1;  // 2 elements per thread
+NBL_CONSTEXPR uint32_t elementCount = uint32_t(1) << (WorkgroupSizeLog2 + ElementsPerThreadLog2);
+#endif

13_BitonicSort/config.json.template

@@ -0,0 +1,28 @@
+{
+  "enableParallelBuild": true,
+  "threadsPerBuildProcess" : 2,
+  "isExecuted": false,
+  "scriptPath": "",
+  "cmake": {
+    "configurations": [ "Release", "Debug", "RelWithDebInfo" ],
+    "buildModes": [],
+    "requiredOptions": []
+  }, 
+  "profiles": [
+    {
+      "backend": "vulkan", // should be none
+      "platform": "windows",
+      "buildModes": [],
+      "runConfiguration": "Release", // we also need to run in Debug nad RWDI because foundational example
+      "gpuArchitectures": []
+    }
+  ],
+  "dependencies": [],
+  "data": [
+    {
+      "dependencies": [],
+      "command": [""],
+      "outputs": []
+    }
+  ]
+}