38_ray_tracing.cpp

#include <algorithm>
#include <array>
#include <chrono>
#include <cstdlib>
#include <cstring>
#include <fstream>
#include <iostream>
#include <limits>
#include <memory>
#include <stdexcept>
#include <vector>

#if defined(__INTELLISENSE__) || !defined(USE_CPP20_MODULES)
#	include <vulkan/vulkan_raii.hpp>
#else
import vulkan_hpp;
#endif

#define GLFW_INCLUDE_VULKAN        // REQUIRED only for GLFW CreateWindowSurface.
#include <GLFW/glfw3.h>

#define GLM_FORCE_RADIANS
#define GLM_FORCE_DEPTH_ZERO_TO_ONE
#define GLM_ENABLE_EXPERIMENTAL
#include <glm/glm.hpp>
#include <glm/gtc/matrix_transform.hpp>
#include <glm/gtx/hash.hpp>

#define STB_IMAGE_IMPLEMENTATION
#include <stb_image.h>

#define TINYOBJLOADER_IMPLEMENTATION
#include <tiny_obj_loader.h>

#ifndef LAB_TASK_LEVEL
#	define LAB_TASK_LEVEL 1
#endif

#define LAB_TASK_AS_BUILD_AND_BIND 4
#define LAB_TASK_AS_ANIMATION 6
#define LAB_TASK_AS_OPAQUE_FLAG 7
#define LAB_TASK_INSTANCE_LUT 9
#define LAB_TASK_REFLECTIONS 11

constexpr uint32_t WIDTH                = 800;
constexpr uint32_t HEIGHT               = 600;
const std::string  MODEL_PATH           = "models/plant_on_table.obj";
constexpr int      MAX_FRAMES_IN_FLIGHT = 2;

const std::vector<char const *> validationLayers = {
    "VK_LAYER_KHRONOS_validation"};

#ifdef NDEBUG
constexpr bool enableValidationLayers = false;
#else
constexpr bool enableValidationLayers = true;
#endif

struct Vertex
{
	glm::vec3 pos;
	glm::vec3 color;
	glm::vec2 texCoord;
	glm::vec3 normal;

	static vk::VertexInputBindingDescription getBindingDescription()
	{
		return {0, sizeof(Vertex), vk::VertexInputRate::eVertex};
	}

	static std::array<vk::VertexInputAttributeDescription, 4> getAttributeDescriptions()
	{
		return {
		    vk::VertexInputAttributeDescription(0, 0, vk::Format::eR32G32B32Sfloat, offsetof(Vertex, pos)),
		    vk::VertexInputAttributeDescription(1, 0, vk::Format::eR32G32B32Sfloat, offsetof(Vertex, color)),
		    vk::VertexInputAttributeDescription(2, 0, vk::Format::eR32G32Sfloat, offsetof(Vertex, texCoord)),
		    vk::VertexInputAttributeDescription(3, 0, vk::Format::eR32G32B32Sfloat, offsetof(Vertex, normal))};
	}

	bool operator==(const Vertex &other) const
	{
		return pos == other.pos && color == other.color && texCoord == other.texCoord && normal == other.normal;
	}
};

template <>
struct std::hash<Vertex>
{
	size_t operator()(Vertex const &vertex) const noexcept
	{
		auto h = std::hash<glm::vec3>()(vertex.pos) ^ (std::hash<glm::vec3>()(vertex.color) << 1);
		h      = (h >> 1) ^ (std::hash<glm::vec2>()(vertex.texCoord) << 1);
		h      = (h >> 1) ^ (std::hash<glm::vec3>()(vertex.normal) << 1);
		return h;
	}
};

struct UniformBufferObject
{
	alignas(16) glm::mat4 model;
	alignas(16) glm::mat4 view;
	alignas(16) glm::mat4 proj;
	alignas(16) glm::vec3 cameraPos;
};

struct PushConstant
{
	uint32_t materialIndex;
#if LAB_TASK_LEVEL >= LAB_TASK_REFLECTIONS
	// TASK11
	uint32_t reflective;
#endif        // LAB_TASK_LEVEL >= LAB_TASK_REFLECTIONS
};

class VulkanRaytracingApplication
{
  public:
	void run()
	{
		initWindow();
		initVulkan();
		mainLoop();
		cleanup();
	}

  private:
	GLFWwindow *window = nullptr;

	vk::raii::Context                context;
	vk::raii::Instance               instance       = nullptr;
	vk::raii::DebugUtilsMessengerEXT debugMessenger = nullptr;
	vk::raii::SurfaceKHR             surface        = nullptr;

	vk::raii::PhysicalDevice physicalDevice = nullptr;
	vk::raii::Device         device         = nullptr;

	vk::raii::Queue graphicsQueue = nullptr;
	vk::raii::Queue presentQueue  = nullptr;

	vk::raii::SwapchainKHR           swapChain = nullptr;
	std::vector<vk::Image>           swapChainImages;
	vk::SurfaceFormatKHR             swapChainSurfaceFormat;
	vk::Extent2D                     swapChainExtent;
	std::vector<vk::raii::ImageView> swapChainImageViews;

	vk::raii::DescriptorSetLayout descriptorSetLayoutGlobal   = nullptr;
	vk::raii::DescriptorSetLayout descriptorSetLayoutMaterial = nullptr;
	vk::raii::PipelineLayout      pipelineLayout              = nullptr;
	vk::raii::Pipeline            graphicsPipeline            = nullptr;

	vk::raii::Image        depthImage       = nullptr;
	vk::raii::DeviceMemory depthImageMemory = nullptr;
	vk::raii::ImageView    depthImageView   = nullptr;

	std::vector<vk::raii::Image>        textureImages;
	std::vector<vk::raii::DeviceMemory> textureImageMemories;
	std::vector<vk::raii::ImageView>    textureImageViews;
	vk::raii::Sampler                   textureSampler = nullptr;

	std::vector<Vertex>    vertices;
	std::vector<uint32_t>  indices;
	vk::raii::Buffer       vertexBuffer       = nullptr;
	vk::raii::DeviceMemory vertexBufferMemory = nullptr;
	vk::raii::Buffer       indexBuffer        = nullptr;
	vk::raii::DeviceMemory indexBufferMemory  = nullptr;
	vk::raii::Buffer       uvBuffer           = nullptr;
	vk::raii::DeviceMemory uvBufferMemory     = nullptr;

	std::vector<vk::raii::Buffer>                   blasBuffers;
	std::vector<vk::raii::DeviceMemory>             blasMemories;
	std::vector<vk::raii::AccelerationStructureKHR> blasHandles;

	std::vector<vk::AccelerationStructureInstanceKHR> instances;
	vk::raii::Buffer                                  instanceBuffer = nullptr;
	vk::raii::DeviceMemory                            instanceMemory = nullptr;

	vk::raii::Buffer                   tlasBuffer        = nullptr;
	vk::raii::DeviceMemory             tlasMemory        = nullptr;
	vk::raii::Buffer                   tlasScratchBuffer = nullptr;
	vk::raii::DeviceMemory             tlasScratchMemory = nullptr;
	vk::raii::AccelerationStructureKHR tlas              = nullptr;

	struct InstanceLUT
	{
		uint32_t materialID;
		uint32_t indexBufferOffset;
	};
	std::vector<InstanceLUT> instanceLUTs;
	vk::raii::Buffer         instanceLUTBuffer       = nullptr;
	vk::raii::DeviceMemory   instanceLUTBufferMemory = nullptr;

	UniformBufferObject ubo{};

	std::vector<vk::raii::Buffer>       uniformBuffers;
	std::vector<vk::raii::DeviceMemory> uniformBuffersMemory;
	std::vector<void *>                 uniformBuffersMapped;

	struct SubMesh
	{
		uint32_t indexOffset;
		uint32_t indexCount;
		int      materialID;
		uint32_t firstVertex;
		uint32_t maxVertex;
		bool     alphaCut;
		bool     reflective;
	};
	std::vector<SubMesh>             submeshes;
	std::vector<tinyobj::material_t> materials;

	vk::raii::DescriptorPool             descriptorPool = nullptr;
	std::vector<vk::raii::DescriptorSet> globalDescriptorSets;
	std::vector<vk::raii::DescriptorSet> materialDescriptorSets;

	vk::raii::CommandPool                commandPool = nullptr;
	std::vector<vk::raii::CommandBuffer> commandBuffers;
	uint32_t                             graphicsIndex = 0;

	std::vector<vk::raii::Semaphore> presentCompleteSemaphores;
	std::vector<vk::raii::Semaphore> renderFinishedSemaphores;
	std::vector<vk::raii::Fence>     inFlightFences;
	uint32_t                         frameIndex = 0;

	bool framebufferResized = false;

	std::vector<const char *> requiredDeviceExtension = {
	    vk::KHRSwapchainExtensionName,
	    vk::KHRSpirv14ExtensionName,
	    vk::KHRSynchronization2ExtensionName,
	    vk::KHRCreateRenderpass2ExtensionName,
	    vk::KHRAccelerationStructureExtensionName,
	    vk::KHRBufferDeviceAddressExtensionName,
	    vk::KHRDeferredHostOperationsExtensionName,
	    vk::KHRRayQueryExtensionName};

	void initWindow()
	{
		glfwInit();

		glfwWindowHint(GLFW_CLIENT_API, GLFW_NO_API);

		window = glfwCreateWindow(WIDTH, HEIGHT, "Vulkan", nullptr, nullptr);
		glfwSetWindowUserPointer(window, this);
		glfwSetFramebufferSizeCallback(window, framebufferResizeCallback);
	}

	static void framebufferResizeCallback(GLFWwindow *window, int width, int height)
	{
		auto app                = static_cast<VulkanRaytracingApplication *>(glfwGetWindowUserPointer(window));
		app->framebufferResized = true;
	}

	void initVulkan()
	{
		createInstance();
		setupDebugMessenger();
		createSurface();
		pickPhysicalDevice();
		createLogicalDevice();
		createSwapChain();
		createImageViews();
		createCommandPool();
		loadModel();
		createDescriptorSetLayout();
		createGraphicsPipeline();
		createDepthResources();
		createTextureSampler();
		createVertexBuffer();
		createIndexBuffer();
		createUVBuffer();
		createAccelerationStructures();
		createInstanceLUTBuffer();
		createUniformBuffers();
		createDescriptorPool();
		createDescriptorSets();
		createCommandBuffers();
		createSyncObjects();
	}

	void mainLoop()
	{
		while (!glfwWindowShouldClose(window))
		{
			glfwPollEvents();
			drawFrame();
		}

		device.waitIdle();
	}

	void cleanupSwapChain()
	{
		swapChainImageViews.clear();
		swapChain = nullptr;
	}

	void cleanup() const
	{
		glfwDestroyWindow(window);

		glfwTerminate();
	}

	void recreateSwapChain()
	{
		int width = 0, height = 0;
		glfwGetFramebufferSize(window, &width, &height);
		while (width == 0 || height == 0)
		{
			glfwGetFramebufferSize(window, &width, &height);
			glfwWaitEvents();
		}

		device.waitIdle();

		cleanupSwapChain();
		createSwapChain();
		createImageViews();
		createDepthResources();
	}

	void createInstance()
	{
		constexpr vk::ApplicationInfo appInfo{.pApplicationName   = "Vulkan Tutorial Ray Tracing",
		                                      .applicationVersion = VK_MAKE_VERSION(1, 0, 0),
		                                      .pEngineName        = "No Engine",
		                                      .engineVersion      = VK_MAKE_VERSION(1, 0, 0),
		                                      .apiVersion         = vk::ApiVersion14};

		// Get the required layers
		std::vector<char const *> requiredLayers;
		if (enableValidationLayers)
		{
			requiredLayers.assign(validationLayers.begin(), validationLayers.end());
		}

		// Check if the required layers are supported by the Vulkan implementation.
		auto layerProperties    = context.enumerateInstanceLayerProperties();
		auto unsupportedLayerIt = std::ranges::find_if(requiredLayers,
		                                               [&layerProperties](auto const &requiredLayer) {
			                                               return std::ranges::none_of(layerProperties,
			                                                                           [requiredLayer](auto const &layerProperty) { return strcmp(layerProperty.layerName, requiredLayer) == 0; });
		                                               });
		if (unsupportedLayerIt != requiredLayers.end())
		{
			throw std::runtime_error("Required layer not supported: " + std::string(*unsupportedLayerIt));
		}

		// Get the required extensions.
		auto requiredExtensions = getRequiredInstanceExtensions();

		// Check if the required extensions are supported by the Vulkan implementation.
		auto extensionProperties = context.enumerateInstanceExtensionProperties();
		auto unsupportedPropertyIt =
		    std::ranges::find_if(requiredExtensions,
		                         [&extensionProperties](auto const &requiredExtension) {
			                         return std::ranges::none_of(extensionProperties,
			                                                     [requiredExtension](auto const &extensionProperty) { return strcmp(extensionProperty.extensionName, requiredExtension) == 0; });
		                         });
		if (unsupportedPropertyIt != requiredExtensions.end())
		{
			throw std::runtime_error("Required extension not supported: " + std::string(*unsupportedPropertyIt));
		}

		vk::InstanceCreateInfo createInfo{.pApplicationInfo        = &appInfo,
		                                  .enabledLayerCount       = static_cast<uint32_t>(requiredLayers.size()),
		                                  .ppEnabledLayerNames     = requiredLayers.data(),
		                                  .enabledExtensionCount   = static_cast<uint32_t>(requiredExtensions.size()),
		                                  .ppEnabledExtensionNames = requiredExtensions.data()};
		instance = vk::raii::Instance(context, createInfo);
	}

	void setupDebugMessenger()
	{
		if (!enableValidationLayers)
			return;

		vk::DebugUtilsMessageSeverityFlagsEXT severityFlags(vk::DebugUtilsMessageSeverityFlagBitsEXT::eWarning |
		                                                    vk::DebugUtilsMessageSeverityFlagBitsEXT::eError);
		vk::DebugUtilsMessageTypeFlagsEXT     messageTypeFlags(
            vk::DebugUtilsMessageTypeFlagBitsEXT::eGeneral | vk::DebugUtilsMessageTypeFlagBitsEXT::ePerformance | vk::DebugUtilsMessageTypeFlagBitsEXT::eValidation);
		vk::DebugUtilsMessengerCreateInfoEXT debugUtilsMessengerCreateInfoEXT{.messageSeverity = severityFlags,
		                                                                      .messageType     = messageTypeFlags,
		                                                                      .pfnUserCallback = &debugCallback};
		debugMessenger = instance.createDebugUtilsMessengerEXT(debugUtilsMessengerCreateInfoEXT);
	}

	void createSurface()
	{
		VkSurfaceKHR _surface;
		if (glfwCreateWindowSurface(*instance, window, nullptr, &_surface) != 0)
		{
			throw std::runtime_error("failed to create window surface!");
		}
		surface = vk::raii::SurfaceKHR(instance, _surface);
	}

	bool isDeviceSuitable(vk::raii::PhysicalDevice const &physicalDevice)
	{
		// Check if the physicalDevice supports the Vulkan 1.3 API version
		bool supportsVulkan1_3 = physicalDevice.getProperties().apiVersion >= VK_API_VERSION_1_3;

		// Check if any of the queue families support graphics operations
		auto queueFamilies    = physicalDevice.getQueueFamilyProperties();
		bool supportsGraphics = std::ranges::any_of(queueFamilies, [](auto const &qfp) { return !!(qfp.queueFlags & vk::QueueFlagBits::eGraphics); });

		// Check if all required physicalDevice extensions are available
		auto availableDeviceExtensions = physicalDevice.enumerateDeviceExtensionProperties();
		bool supportsAllRequiredExtensions =
		    std::ranges::all_of(requiredDeviceExtension,
		                        [&availableDeviceExtensions](auto const &requiredDeviceExtension) {
			                        return std::ranges::any_of(availableDeviceExtensions,
			                                                   [requiredDeviceExtension](auto const &availableDeviceExtension) { return strcmp(availableDeviceExtension.extensionName, requiredDeviceExtension) == 0; });
		                        });

		// Check if the physicalDevice supports the required features
		auto features                 = physicalDevice.template getFeatures2<vk::PhysicalDeviceFeatures2,
		                                                                     vk::PhysicalDeviceVulkan12Features,
		                                                                     vk::PhysicalDeviceVulkan13Features,
		                                                                     vk::PhysicalDeviceExtendedDynamicStateFeaturesEXT,
		                                                                     vk::PhysicalDeviceAccelerationStructureFeaturesKHR,
		                                                                     vk::PhysicalDeviceRayQueryFeaturesKHR>();
		bool supportsRequiredFeatures = features.template get<vk::PhysicalDeviceFeatures2>().features.samplerAnisotropy &&
		                                features.template get<vk::PhysicalDeviceVulkan13Features>().dynamicRendering &&
		                                features.template get<vk::PhysicalDeviceExtendedDynamicStateFeaturesEXT>().extendedDynamicState &&
		                                features.template get<vk::PhysicalDeviceVulkan12Features>().descriptorBindingSampledImageUpdateAfterBind &&
		                                features.template get<vk::PhysicalDeviceVulkan12Features>().descriptorBindingPartiallyBound &&
		                                features.template get<vk::PhysicalDeviceVulkan12Features>().descriptorBindingVariableDescriptorCount &&
		                                features.template get<vk::PhysicalDeviceVulkan12Features>().runtimeDescriptorArray &&
		                                features.template get<vk::PhysicalDeviceVulkan12Features>().shaderSampledImageArrayNonUniformIndexing &&
		                                features.template get<vk::PhysicalDeviceVulkan12Features>().bufferDeviceAddress &&
		                                features.template get<vk::PhysicalDeviceAccelerationStructureFeaturesKHR>().accelerationStructure &&
		                                features.template get<vk::PhysicalDeviceRayQueryFeaturesKHR>().rayQuery;

		// Return true if the physicalDevice meets all the criteria
		return supportsVulkan1_3 && supportsGraphics && supportsAllRequiredExtensions && supportsRequiredFeatures;
	}

	void pickPhysicalDevice()
	{
		std::vector<vk::raii::PhysicalDevice> physicalDevices = instance.enumeratePhysicalDevices();
		auto const                            devIter         = std::ranges::find_if(physicalDevices, [&](auto const &physicalDevice) { return isDeviceSuitable(physicalDevice); });
		if (devIter == physicalDevices.end())
		{
			throw std::runtime_error("failed to find a suitable GPU!");
		}
		physicalDevice = *devIter;
	}

	void createLogicalDevice()
	{
		// find the index of the first queue family that supports graphics
		std::vector<vk::QueueFamilyProperties> queueFamilyProperties = physicalDevice.getQueueFamilyProperties();

		// get the first index into queueFamilyProperties which supports graphics
		auto graphicsQueueFamilyProperty = std::ranges::find_if(queueFamilyProperties, [](auto const &qfp) { return (qfp.queueFlags & vk::QueueFlagBits::eGraphics) != static_cast<vk::QueueFlags>(0); });

		graphicsIndex = static_cast<uint32_t>(std::distance(queueFamilyProperties.begin(), graphicsQueueFamilyProperty));

		// determine a queueFamilyIndex that supports present
		// first check if the graphicsIndex is good enough
		auto presentIndex = physicalDevice.getSurfaceSupportKHR(graphicsIndex, *surface) ? graphicsIndex : ~0;
		if (presentIndex == queueFamilyProperties.size())
		{
			// the graphicsIndex doesn't support present -> look for another family index that supports both
			// graphics and present
			for (size_t i = 0; i < queueFamilyProperties.size(); i++)
			{
				if ((queueFamilyProperties[i].queueFlags & vk::QueueFlagBits::eGraphics) &&
				    physicalDevice.getSurfaceSupportKHR(static_cast<uint32_t>(i), *surface))
				{
					graphicsIndex = static_cast<uint32_t>(i);
					presentIndex  = graphicsIndex;
					break;
				}
			}
			if (presentIndex == queueFamilyProperties.size())
			{
				// there's nothing like a single family index that supports both graphics and present -> look for another
				// family index that supports present
				for (size_t i = 0; i < queueFamilyProperties.size(); i++)
				{
					if (physicalDevice.getSurfaceSupportKHR(static_cast<uint32_t>(i), *surface))
					{
						presentIndex = static_cast<uint32_t>(i);
						break;
					}
				}
			}
		}
		if ((graphicsIndex == queueFamilyProperties.size()) || (presentIndex == queueFamilyProperties.size()))
		{
			throw std::runtime_error("Could not find a queue for graphics or present -> terminating");
		}

		// query for Vulkan 1.3 features
		vk::StructureChain<vk::PhysicalDeviceFeatures2, vk::PhysicalDeviceVulkan12Features,
		                   vk::PhysicalDeviceVulkan13Features, vk::PhysicalDeviceExtendedDynamicStateFeaturesEXT,
		                   vk::PhysicalDeviceAccelerationStructureFeaturesKHR, vk::PhysicalDeviceRayQueryFeaturesKHR>
		    featureChain = {
		        {.features = {.samplerAnisotropy = true}},                                                                                                                                                                                                                                // vk::PhysicalDeviceFeatures2
		        {.shaderSampledImageArrayNonUniformIndexing = true, .descriptorBindingSampledImageUpdateAfterBind = true, .descriptorBindingPartiallyBound = true, .descriptorBindingVariableDescriptorCount = true, .runtimeDescriptorArray = true, .bufferDeviceAddress = true},        // vk::PhysicalDeviceVulkan12Features
		        {.synchronization2 = true, .dynamicRendering = true},                                                                                                                                                                                                                     // vk::PhysicalDeviceVulkan13Features
		        {.extendedDynamicState = true},                                                                                                                                                                                                                                           // vk::PhysicalDeviceExtendedDynamicStateFeaturesEXT
		        {.accelerationStructure = true},                                                                                                                                                                                                                                          // vk::PhysicalDeviceAccelerationStructureFeaturesKHR
		        {.rayQuery = true}                                                                                                                                                                                                                                                        // vk::PhysicalDeviceRayQueryFeaturesKHR
		    };

		// create a Device
		float                     queuePriority = 0.5f;
		vk::DeviceQueueCreateInfo deviceQueueCreateInfo{.queueFamilyIndex = graphicsIndex, .queueCount = 1, .pQueuePriorities = &queuePriority};
		vk::DeviceCreateInfo      deviceCreateInfo{.pNext                   = &featureChain.get<vk::PhysicalDeviceFeatures2>(),
		                                           .queueCreateInfoCount    = 1,
		                                           .pQueueCreateInfos       = &deviceQueueCreateInfo,
		                                           .enabledExtensionCount   = static_cast<uint32_t>(requiredDeviceExtension.size()),
		                                           .ppEnabledExtensionNames = requiredDeviceExtension.data()};

		device        = vk::raii::Device(physicalDevice, deviceCreateInfo);
		graphicsQueue = vk::raii::Queue(device, graphicsIndex, 0);
		presentQueue  = vk::raii::Queue(device, presentIndex, 0);
	}

	void createSwapChain()
	{
		vk::SurfaceCapabilitiesKHR surfaceCapabilities = physicalDevice.getSurfaceCapabilitiesKHR(*surface);
		swapChainExtent                                = chooseSwapExtent(surfaceCapabilities);
		uint32_t minImageCount                         = chooseSwapMinImageCount(surfaceCapabilities);

		std::vector<vk::SurfaceFormatKHR> availableFormats = physicalDevice.getSurfaceFormatsKHR(*surface);
		swapChainSurfaceFormat                             = chooseSwapSurfaceFormat(availableFormats);

		std::vector<vk::PresentModeKHR> availablePresentModes = physicalDevice.getSurfacePresentModesKHR(*surface);
		vk::PresentModeKHR              presentMode           = chooseSwapPresentMode(availablePresentModes);

		vk::SwapchainCreateInfoKHR swapChainCreateInfo{.surface          = *surface,
		                                               .minImageCount    = minImageCount,
		                                               .imageFormat      = swapChainSurfaceFormat.format,
		                                               .imageColorSpace  = swapChainSurfaceFormat.colorSpace,
		                                               .imageExtent      = swapChainExtent,
		                                               .imageArrayLayers = 1,
		                                               .imageUsage       = vk::ImageUsageFlagBits::eColorAttachment,
		                                               .imageSharingMode = vk::SharingMode::eExclusive,
		                                               .preTransform     = surfaceCapabilities.currentTransform,
		                                               .compositeAlpha   = vk::CompositeAlphaFlagBitsKHR::eOpaque,
		                                               .presentMode      = presentMode,
		                                               .clipped          = true};

		swapChain       = vk::raii::SwapchainKHR(device, swapChainCreateInfo);
		swapChainImages = swapChain.getImages();
	}

	void createImageViews()
	{
		vk::ImageViewCreateInfo imageViewCreateInfo{.viewType         = vk::ImageViewType::e2D,
		                                            .format           = swapChainSurfaceFormat.format,
		                                            .subresourceRange = {vk::ImageAspectFlagBits::eColor, 0, 1, 0, 1}};
		for (auto &image : swapChainImages)
		{
			imageViewCreateInfo.image = image;
			swapChainImageViews.emplace_back(device, imageViewCreateInfo);
		}
	}

	void createDescriptorSetLayout()
	{
		// Use descriptor set 0 for global data
		// TASK04: The acceleration structure uses binding 1
		std::array global_bindings = {
		    vk::DescriptorSetLayoutBinding(0, vk::DescriptorType::eUniformBuffer, 1, vk::ShaderStageFlagBits::eVertex | vk::ShaderStageFlagBits::eFragment, nullptr),
		    vk::DescriptorSetLayoutBinding(1, vk::DescriptorType::eAccelerationStructureKHR, 1, vk::ShaderStageFlagBits::eFragment, nullptr),
		    vk::DescriptorSetLayoutBinding(2, vk::DescriptorType::eStorageBuffer, 1, vk::ShaderStageFlagBits::eFragment, nullptr),
		    vk::DescriptorSetLayoutBinding(3, vk::DescriptorType::eStorageBuffer, 1, vk::ShaderStageFlagBits::eFragment, nullptr),
		    vk::DescriptorSetLayoutBinding(4, vk::DescriptorType::eStorageBuffer, 1, vk::ShaderStageFlagBits::eFragment, nullptr)};

		vk::DescriptorSetLayoutCreateInfo globalLayoutInfo{.bindingCount = static_cast<uint32_t>(global_bindings.size()), .pBindings = global_bindings.data()};

		descriptorSetLayoutGlobal = vk::raii::DescriptorSetLayout(device, globalLayoutInfo);

		// Use descriptor set 1 for bindless material data
		uint32_t textureCount = static_cast<uint32_t>(textureImageViews.size());

		std::array material_bindings = {
		    vk::DescriptorSetLayoutBinding(0, vk::DescriptorType::eSampler, 1, vk::ShaderStageFlagBits::eFragment, nullptr),
		    vk::DescriptorSetLayoutBinding(1, vk::DescriptorType::eSampledImage, static_cast<uint32_t>(textureCount), vk::ShaderStageFlagBits::eFragment, nullptr)};

		std::vector<vk::DescriptorBindingFlags> bindingFlags = {
		    vk::DescriptorBindingFlagBits::eUpdateAfterBind,
		    vk::DescriptorBindingFlagBits::ePartiallyBound | vk::DescriptorBindingFlagBits::eVariableDescriptorCount | vk::DescriptorBindingFlagBits::eUpdateAfterBind};

		vk::DescriptorSetLayoutBindingFlagsCreateInfo flagsCreateInfo{
		    .bindingCount  = static_cast<uint32_t>(bindingFlags.size()),
		    .pBindingFlags = bindingFlags.data()};

		vk::DescriptorSetLayoutCreateInfo materialLayoutInfo{
		    .pNext        = &flagsCreateInfo,
		    .flags        = vk::DescriptorSetLayoutCreateFlagBits::eUpdateAfterBindPool,
		    .bindingCount = static_cast<uint32_t>(material_bindings.size()),
		    .pBindings    = material_bindings.data(),
		};

		descriptorSetLayoutMaterial = vk::raii::DescriptorSetLayout(device, materialLayoutInfo);
	}

	void createGraphicsPipeline()
	{
		vk::raii::ShaderModule shaderModule = createShaderModule(readFile("shaders/slang.spv"));

		vk::PipelineShaderStageCreateInfo vertShaderStageInfo{.stage = vk::ShaderStageFlagBits::eVertex, .module = shaderModule, .pName = "vertMain"};
		vk::PipelineShaderStageCreateInfo fragShaderStageInfo{.stage = vk::ShaderStageFlagBits::eFragment, .module = shaderModule, .pName = "fragMain"};
		vk::PipelineShaderStageCreateInfo shaderStages[] = {vertShaderStageInfo, fragShaderStageInfo};

		auto                                   bindingDescription    = Vertex::getBindingDescription();
		auto                                   attributeDescriptions = Vertex::getAttributeDescriptions();
		vk::PipelineVertexInputStateCreateInfo vertexInputInfo{
		    .vertexBindingDescriptionCount   = 1,
		    .pVertexBindingDescriptions      = &bindingDescription,
		    .vertexAttributeDescriptionCount = static_cast<uint32_t>(attributeDescriptions.size()),
		    .pVertexAttributeDescriptions    = attributeDescriptions.data()};
		vk::PipelineInputAssemblyStateCreateInfo inputAssembly{
		    .topology               = vk::PrimitiveTopology::eTriangleList,
		    .primitiveRestartEnable = vk::False};
		vk::PipelineViewportStateCreateInfo viewportState{
		    .viewportCount = 1,
		    .scissorCount  = 1};
		vk::PipelineRasterizationStateCreateInfo rasterizer{
		    .depthClampEnable        = vk::False,
		    .rasterizerDiscardEnable = vk::False,
		    .polygonMode             = vk::PolygonMode::eFill,
		    .cullMode                = vk::CullModeFlagBits::eBack,
		    .frontFace               = vk::FrontFace::eCounterClockwise,
		    .depthBiasEnable         = vk::False,
		    .lineWidth               = 1.0f};
		vk::PipelineMultisampleStateCreateInfo multisampling{
		    .rasterizationSamples = vk::SampleCountFlagBits::e1,
		    .sampleShadingEnable  = vk::False};
		vk::PipelineDepthStencilStateCreateInfo depthStencil{
		    .depthTestEnable       = vk::True,
		    .depthWriteEnable      = vk::True,
		    .depthCompareOp        = vk::CompareOp::eLess,
		    .depthBoundsTestEnable = vk::False,
		    .stencilTestEnable     = vk::False};
		vk::PipelineColorBlendAttachmentState colorBlendAttachment{
		    .blendEnable    = vk::False,
		    .colorWriteMask = vk::ColorComponentFlagBits::eR | vk::ColorComponentFlagBits::eG | vk::ColorComponentFlagBits::eB | vk::ColorComponentFlagBits::eA};
		vk::PipelineColorBlendStateCreateInfo colorBlending{
		    .logicOpEnable   = vk::False,
		    .logicOp         = vk::LogicOp::eCopy,
		    .attachmentCount = 1,
		    .pAttachments    = &colorBlendAttachment};
		std::vector dynamicStates = {
		    vk::DynamicState::eViewport,
		    vk::DynamicState::eScissor};
		vk::PipelineDynamicStateCreateInfo dynamicState{.dynamicStateCount = static_cast<uint32_t>(dynamicStates.size()), .pDynamicStates = dynamicStates.data()};

		vk::DescriptorSetLayout setLayouts[] = {*descriptorSetLayoutGlobal, *descriptorSetLayoutMaterial};

		vk::PushConstantRange pushConstantRange{
		    .stageFlags = vk::ShaderStageFlagBits::eFragment,
		    .offset     = 0,
		    .size       = sizeof(PushConstant)};

		vk::PipelineLayoutCreateInfo pipelineLayoutInfo{.setLayoutCount = 2, .pSetLayouts = setLayouts, .pushConstantRangeCount = 1, .pPushConstantRanges = &pushConstantRange};

		pipelineLayout = vk::raii::PipelineLayout(device, pipelineLayoutInfo);

		vk::Format depthFormat = findDepthFormat();

		/* TASK01: Check the setup for dynamic rendering
		 *
		 * This new struct replaces what previously was the render pass in the pipeline creation.
		 * Note how this structure is now linked in .pNext below, and .renderPass is not used.
		 */
		vk::StructureChain<vk::GraphicsPipelineCreateInfo, vk::PipelineRenderingCreateInfo> pipelineCreateInfoChain = {
		    {.stageCount          = 2,
		     .pStages             = shaderStages,
		     .pVertexInputState   = &vertexInputInfo,
		     .pInputAssemblyState = &inputAssembly,
		     .pViewportState      = &viewportState,
		     .pRasterizationState = &rasterizer,
		     .pMultisampleState   = &multisampling,
		     .pDepthStencilState  = &depthStencil,
		     .pColorBlendState    = &colorBlending,
		     .pDynamicState       = &dynamicState,
		     .layout              = *pipelineLayout,
		     .renderPass          = nullptr},
		    {.colorAttachmentCount = 1, .pColorAttachmentFormats = &swapChainSurfaceFormat.format, .depthAttachmentFormat = depthFormat}};

		graphicsPipeline = vk::raii::Pipeline(device, nullptr, pipelineCreateInfoChain.get<vk::GraphicsPipelineCreateInfo>());
	}

	void createCommandPool()
	{
		vk::CommandPoolCreateInfo poolInfo{
		    .flags            = vk::CommandPoolCreateFlagBits::eResetCommandBuffer,
		    .queueFamilyIndex = graphicsIndex};
		commandPool = vk::raii::CommandPool(device, poolInfo);
	}

	void createDepthResources()
	{
		vk::Format depthFormat = findDepthFormat();

		createImage(swapChainExtent.width, swapChainExtent.height, depthFormat, vk::ImageTiling::eOptimal, vk::ImageUsageFlagBits::eDepthStencilAttachment, vk::MemoryPropertyFlagBits::eDeviceLocal, depthImage, depthImageMemory);
		depthImageView = createImageView(depthImage, depthFormat, vk::ImageAspectFlagBits::eDepth);
	}

	vk::Format findSupportedFormat(const std::vector<vk::Format> &candidates, vk::ImageTiling tiling, vk::FormatFeatureFlags features) const
	{
		for (const auto format : candidates)
		{
			vk::FormatProperties props = physicalDevice.getFormatProperties(format);

			if (tiling == vk::ImageTiling::eLinear && (props.linearTilingFeatures & features) == features)
			{
				return format;
			}
			if (tiling == vk::ImageTiling::eOptimal && (props.optimalTilingFeatures & features) == features)
			{
				return format;
			}
		}

		throw std::runtime_error("failed to find supported format!");
	}

	[[nodiscard]] vk::Format findDepthFormat() const
	{
		return findSupportedFormat(
		    {vk::Format::eD32Sfloat, vk::Format::eD32SfloatS8Uint, vk::Format::eD24UnormS8Uint},
		    vk::ImageTiling::eOptimal,
		    vk::FormatFeatureFlagBits::eDepthStencilAttachment);
	}

	static bool hasStencilComponent(vk::Format format)
	{
		return format == vk::Format::eD32SfloatS8Uint || format == vk::Format::eD24UnormS8Uint;
	}

	std::pair<vk::raii::Image, vk::raii::DeviceMemory> createTextureImage(const std::string &path)
	{
		int            texWidth, texHeight, texChannels;
		stbi_uc       *pixels    = stbi_load(path.c_str(), &texWidth, &texHeight, &texChannels, STBI_rgb_alpha);
		vk::DeviceSize imageSize = texWidth * texHeight * 4;

		if (!pixels)
		{
			throw std::runtime_error("failed to load texture image!");
		}

		vk::raii::Buffer       stagingBuffer({});
		vk::raii::DeviceMemory stagingBufferMemory({});
		createBuffer(imageSize, vk::BufferUsageFlagBits::eTransferSrc, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent, stagingBuffer, stagingBufferMemory);

		void *data = stagingBufferMemory.mapMemory(0, imageSize);
		memcpy(data, pixels, imageSize);
		stagingBufferMemory.unmapMemory();

		stbi_image_free(pixels);

		vk::raii::Image        textureImage       = nullptr;
		vk::raii::DeviceMemory textureImageMemory = nullptr;

		createImage(texWidth, texHeight, vk::Format::eR8G8B8A8Srgb, vk::ImageTiling::eOptimal, vk::ImageUsageFlagBits::eTransferDst | vk::ImageUsageFlagBits::eSampled, vk::MemoryPropertyFlagBits::eDeviceLocal, textureImage, textureImageMemory);

		transitionImageLayout(textureImage, vk::ImageLayout::eUndefined, vk::ImageLayout::eTransferDstOptimal);
		copyBufferToImage(stagingBuffer, textureImage, static_cast<uint32_t>(texWidth), static_cast<uint32_t>(texHeight));
		transitionImageLayout(textureImage, vk::ImageLayout::eTransferDstOptimal, vk::ImageLayout::eShaderReadOnlyOptimal);

		return std::make_pair(std::move(textureImage), std::move(textureImageMemory));
	}

	vk::raii::ImageView createTextureImageView(vk::raii::Image &textureImage)
	{
		return createImageView(textureImage, vk::Format::eR8G8B8A8Srgb, vk::ImageAspectFlagBits::eColor);
	}

	void createTextureSampler()
	{
		vk::PhysicalDeviceProperties properties = physicalDevice.getProperties();
		vk::SamplerCreateInfo        samplerInfo{
		           .magFilter        = vk::Filter::eLinear,
		           .minFilter        = vk::Filter::eLinear,
		           .mipmapMode       = vk::SamplerMipmapMode::eLinear,
		           .addressModeU     = vk::SamplerAddressMode::eRepeat,
		           .addressModeV     = vk::SamplerAddressMode::eRepeat,
		           .addressModeW     = vk::SamplerAddressMode::eRepeat,
		           .mipLodBias       = 0.0f,
		           .anisotropyEnable = vk::True,
		           .maxAnisotropy    = properties.limits.maxSamplerAnisotropy,
		           .compareEnable    = vk::False,
		           .compareOp        = vk::CompareOp::eAlways};
		textureSampler = vk::raii::Sampler(device, samplerInfo);
	}

	vk::raii::ImageView createImageView(vk::raii::Image &image, vk::Format format, vk::ImageAspectFlags aspectFlags)
	{
		vk::ImageViewCreateInfo viewInfo{
		    .image            = image,
		    .viewType         = vk::ImageViewType::e2D,
		    .format           = format,
		    .subresourceRange = {aspectFlags, 0, 1, 0, 1}};
		return vk::raii::ImageView(device, viewInfo);
	}

	void createImage(uint32_t width, uint32_t height, vk::Format format, vk::ImageTiling tiling, vk::ImageUsageFlags usage, vk::MemoryPropertyFlags properties, vk::raii::Image &image, vk::raii::DeviceMemory &imageMemory)
	{
		vk::ImageCreateInfo imageInfo{
		    .imageType     = vk::ImageType::e2D,
		    .format        = format,
		    .extent        = {width, height, 1},
		    .mipLevels     = 1,
		    .arrayLayers   = 1,
		    .samples       = vk::SampleCountFlagBits::e1,
		    .tiling        = tiling,
		    .usage         = usage,
		    .sharingMode   = vk::SharingMode::eExclusive,
		    .initialLayout = vk::ImageLayout::eUndefined};
		image = vk::raii::Image(device, imageInfo);

		vk::MemoryRequirements memRequirements = image.getMemoryRequirements();
		vk::MemoryAllocateInfo allocInfo{
		    .allocationSize  = memRequirements.size,
		    .memoryTypeIndex = findMemoryType(memRequirements.memoryTypeBits, properties)};
		imageMemory = vk::raii::DeviceMemory(device, allocInfo);
		image.bindMemory(imageMemory, 0);
	}

	void transitionImageLayout(const vk::raii::Image &image, vk::ImageLayout oldLayout, vk::ImageLayout newLayout)
	{
		auto commandBuffer = beginSingleTimeCommands();

		vk::ImageMemoryBarrier barrier{
		    .oldLayout        = oldLayout,
		    .newLayout        = newLayout,
		    .image            = image,
		    .subresourceRange = {vk::ImageAspectFlagBits::eColor, 0, 1, 0, 1}};

		vk::PipelineStageFlags sourceStage;
		vk::PipelineStageFlags destinationStage;

		if (oldLayout == vk::ImageLayout::eUndefined && newLayout == vk::ImageLayout::eTransferDstOptimal)
		{
			barrier.srcAccessMask = {};
			barrier.dstAccessMask = vk::AccessFlagBits::eTransferWrite;

			sourceStage      = vk::PipelineStageFlagBits::eTopOfPipe;
			destinationStage = vk::PipelineStageFlagBits::eTransfer;
		}
		else if (oldLayout == vk::ImageLayout::eTransferDstOptimal && newLayout == vk::ImageLayout::eShaderReadOnlyOptimal)
		{
			barrier.srcAccessMask = vk::AccessFlagBits::eTransferWrite;
			barrier.dstAccessMask = vk::AccessFlagBits::eShaderRead;

			sourceStage      = vk::PipelineStageFlagBits::eTransfer;
			destinationStage = vk::PipelineStageFlagBits::eFragmentShader;
		}
		else
		{
			throw std::invalid_argument("unsupported layout transition!");
		}
		commandBuffer->pipelineBarrier(sourceStage, destinationStage, {}, {}, nullptr, barrier);
		endSingleTimeCommands(*commandBuffer);
	}

	void copyBufferToImage(const vk::raii::Buffer &buffer, vk::raii::Image &image, uint32_t width, uint32_t height)
	{
		std::unique_ptr<vk::raii::CommandBuffer> commandBuffer = beginSingleTimeCommands();
		vk::BufferImageCopy                      region{
		                         .bufferOffset      = 0,
		                         .bufferRowLength   = 0,
		                         .bufferImageHeight = 0,
		                         .imageSubresource  = {vk::ImageAspectFlagBits::eColor, 0, 0, 1},
		                         .imageOffset       = {0, 0, 0},
		                         .imageExtent       = {width, height, 1}};
		commandBuffer->copyBufferToImage(buffer, image, vk::ImageLayout::eTransferDstOptimal, {region});
		endSingleTimeCommands(*commandBuffer);
	}

	void loadModel()
	{
		tinyobj::attrib_t                attrib;
		std::vector<tinyobj::shape_t>    shapes;
		std::vector<tinyobj::material_t> localMaterials;
		std::string                      warn, err;

		if (!LoadObj(&attrib, &shapes, &localMaterials, &warn, &err, MODEL_PATH.c_str(), MODEL_PATH.substr(0, MODEL_PATH.find_last_of("/\\")).c_str()))
		{
			throw std::runtime_error(warn + err);
		}

		size_t materialOffset  = materials.size();
		size_t oldTextureCount = textureImageViews.size();

		materials.insert(materials.end(), localMaterials.begin(), localMaterials.end());

		std::unordered_map<Vertex, uint32_t> uniqueVertices{};
		uint32_t                             indexOffset = 0;

		for (const auto &shape : shapes)
		{
			std::cout << "Loading mesh: " << shape.name << ": " << shape.mesh.indices.size() / 3 << " triangles\n";

			uint32_t startOffset = indexOffset;
			uint32_t localMaxV   = 0;

			for (const auto &index : shape.mesh.indices)
			{
				Vertex vertex{};

				vertex.pos = {
				    attrib.vertices[3 * index.vertex_index + 0],
				    attrib.vertices[3 * index.vertex_index + 1],
				    attrib.vertices[3 * index.vertex_index + 2]};

				vertex.texCoord = {
				    attrib.texcoords[2 * index.texcoord_index + 0],
				    1.0f - attrib.texcoords[2 * index.texcoord_index + 1]};

				vertex.color = {1.0f, 1.0f, 1.0f};

				if (index.normal_index >= 0)
				{
					vertex.normal = {
					    attrib.normals[3 * index.normal_index + 0],
					    attrib.normals[3 * index.normal_index + 1],
					    attrib.normals[3 * index.normal_index + 2]};
				}
				else
				{
					vertex.normal = {0.0f, 0.0f, 0.0f};
				}

				if (!uniqueVertices.contains(vertex))
				{
					uniqueVertices[vertex] = static_cast<uint32_t>(vertices.size());
					vertices.push_back(vertex);
				}

				indices.push_back(uniqueVertices[vertex]);

				indexOffset++;

				uint32_t vi;
				auto     it = uniqueVertices.find(vertex);
				if (it != uniqueVertices.end())
				{
					vi = it->second;
				}
				else
				{
					vi                     = static_cast<uint32_t>(vertices.size());
					uniqueVertices[vertex] = vi;
					vertices.push_back(vertex);
				}

				localMaxV = std::max(localMaxV, vi);
			}

			int localMaterialID  = shape.mesh.material_ids.empty() ? -1 : shape.mesh.material_ids[0];
			int globalMaterialID = (localMaterialID < 0) ? -1 : static_cast<int>(materialOffset + localMaterialID);

			uint32_t indexCount = indexOffset - startOffset;

			// Note that this is only valid for this particular MODEL_PATH
			bool alphaCut   = (shape.name.find("nettle_plant") != std::string::npos);
			bool reflective = (shape.name.find("table") != std::string::npos);

			submeshes.push_back({.indexOffset = startOffset,
			                     .indexCount  = indexCount,
			                     .materialID  = globalMaterialID,
			                     .firstVertex = 0u,
			                     .maxVertex   = localMaxV + 1,
			                     .alphaCut    = alphaCut,
			                     .reflective  = reflective});
		}

		for (size_t i = 0; i < localMaterials.size(); ++i)
		{
			const auto &material = localMaterials[i];

			if (!material.diffuse_texname.empty())
			{
				std::string texturePath = MODEL_PATH.substr(0, MODEL_PATH.find_last_of("/\\")) + "/" + material.diffuse_texname;
				auto [img, mem]         = createTextureImage(texturePath);
				textureImages.push_back(std::move(img));
				textureImageMemories.push_back(std::move(mem));
				textureImageViews.emplace_back(createTextureImageView(textureImages.back()));
			}
			else
			{
				std::cout << "No texture for material: " << material.name << std::endl;
			}
		}
	}

	void createVertexBuffer()
	{
		vk::DeviceSize         bufferSize = sizeof(vertices[0]) * vertices.size();
		vk::raii::Buffer       stagingBuffer({});
		vk::raii::DeviceMemory stagingBufferMemory({});
		createBuffer(bufferSize, vk::BufferUsageFlagBits::eTransferSrc, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent, stagingBuffer, stagingBufferMemory);

		void *dataStaging = stagingBufferMemory.mapMemory(0, bufferSize);
		memcpy(dataStaging, vertices.data(), bufferSize);
		stagingBufferMemory.unmapMemory();

		createBuffer(bufferSize, vk::BufferUsageFlagBits::eTransferDst | vk::BufferUsageFlagBits::eVertexBuffer | vk::BufferUsageFlagBits::eShaderDeviceAddress | vk::BufferUsageFlagBits::eAccelerationStructureBuildInputReadOnlyKHR, vk::MemoryPropertyFlagBits::eDeviceLocal, vertexBuffer, vertexBufferMemory);

		copyBuffer(stagingBuffer, vertexBuffer, bufferSize);
	}

	void createIndexBuffer()
	{
		vk::DeviceSize bufferSize = sizeof(indices[0]) * indices.size();

		vk::raii::Buffer       stagingBuffer({});
		vk::raii::DeviceMemory stagingBufferMemory({});
		createBuffer(bufferSize, vk::BufferUsageFlagBits::eTransferSrc, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent, stagingBuffer, stagingBufferMemory);

		void *data = stagingBufferMemory.mapMemory(0, bufferSize);
		memcpy(data, indices.data(), bufferSize);
		stagingBufferMemory.unmapMemory();

		createBuffer(bufferSize, vk::BufferUsageFlagBits::eTransferDst | vk::BufferUsageFlagBits::eIndexBuffer | vk::BufferUsageFlagBits::eShaderDeviceAddress | vk::BufferUsageFlagBits::eAccelerationStructureBuildInputReadOnlyKHR | vk::BufferUsageFlagBits::eStorageBuffer, vk::MemoryPropertyFlagBits::eDeviceLocal, indexBuffer, indexBufferMemory);

		copyBuffer(stagingBuffer, indexBuffer, bufferSize);
	}

	void createUVBuffer()
	{
		// Extract all texCoords into a separate vector
		std::vector<glm::vec2> uvs;
		uvs.reserve(vertices.size());
		for (auto &v : vertices)
		{
			uvs.push_back(v.texCoord);
		}

		vk::DeviceSize bufferSize = sizeof(uvs[0]) * uvs.size();

		vk::raii::Buffer       stagingBuffer({});
		vk::raii::DeviceMemory stagingBufferMemory({});
		createBuffer(bufferSize, vk::BufferUsageFlagBits::eTransferSrc, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent, stagingBuffer, stagingBufferMemory);

		void *dataStaging = stagingBufferMemory.mapMemory(0, bufferSize);
		memcpy(dataStaging, uvs.data(), bufferSize);
		stagingBufferMemory.unmapMemory();

		createBuffer(bufferSize, vk::BufferUsageFlagBits::eTransferDst | vk::BufferUsageFlagBits::eStorageBuffer,
		             vk::MemoryPropertyFlagBits::eDeviceLocal, uvBuffer, uvBufferMemory);

		copyBuffer(stagingBuffer, uvBuffer, bufferSize);
	}

	void createInstanceLUTBuffer()
	{
#if LAB_TASK_LEVEL >= LAB_TASK_INSTANCE_LUT
		// TASK09: build a buffer to store the instance look-up table
		vk::DeviceSize bufferSize = sizeof(InstanceLUT) * instanceLUTs.size();

		vk::raii::Buffer       stagingBuffer({});
		vk::raii::DeviceMemory stagingBufferMemory({});
		createBuffer(bufferSize, vk::BufferUsageFlagBits::eTransferSrc, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent, stagingBuffer, stagingBufferMemory);

		void *dataStaging = stagingBufferMemory.mapMemory(0, bufferSize);
		memcpy(dataStaging, instanceLUTs.data(), bufferSize);
		stagingBufferMemory.unmapMemory();

		createBuffer(bufferSize, vk::BufferUsageFlagBits::eTransferDst | vk::BufferUsageFlagBits::eStorageBuffer,
		             vk::MemoryPropertyFlagBits::eDeviceLocal, instanceLUTBuffer, instanceLUTBufferMemory);

		copyBuffer(stagingBuffer, instanceLUTBuffer, bufferSize);
#endif        // LAB_TASK_LEVEL >= LAB_TASK_INSTANCE_LUT
	}

	void createUniformBuffers()
	{
		uniformBuffers.clear();
		uniformBuffersMemory.clear();
		uniformBuffersMapped.clear();

		for (size_t i = 0; i < MAX_FRAMES_IN_FLIGHT; i++)
		{
			vk::DeviceSize         bufferSize = sizeof(UniformBufferObject);
			vk::raii::Buffer       buffer({});
			vk::raii::DeviceMemory bufferMem({});
			createBuffer(bufferSize, vk::BufferUsageFlagBits::eUniformBuffer, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent, buffer, bufferMem);
			uniformBuffers.emplace_back(std::move(buffer));
			uniformBuffersMemory.emplace_back(std::move(bufferMem));
			uniformBuffersMapped.emplace_back(uniformBuffersMemory[i].mapMemory(0, bufferSize));
		}
	}

	void createAccelerationStructures()
	{
#if LAB_TASK_LEVEL >= LAB_TASK_AS_BUILD_AND_BIND
		vk::BufferDeviceAddressInfo vai{.buffer = *vertexBuffer};
		vk::DeviceAddress           vertexAddr = device.getBufferAddressKHR(vai);
		vk::BufferDeviceAddressInfo iai{.buffer = *indexBuffer};
		vk::DeviceAddress           indexAddr = device.getBufferAddressKHR(iai);

		instances.reserve(submeshes.size());
		blasBuffers.reserve(submeshes.size());
		blasMemories.reserve(submeshes.size());
		blasHandles.reserve(submeshes.size());

		vk::TransformMatrixKHR identity{};
		identity.matrix = std::array<std::array<float, 4>, 3>{{std::array<float, 4>{1.f, 0.f, 0.f, 0.f},
		                                                       std::array<float, 4>{0.f, 1.f, 0.f, 0.f},
		                                                       std::array<float, 4>{0.f, 0.f, 1.f, 0.f}}};

		// TASK02: Build a bottom level acceleration structure for each submesh
		for (size_t i = 0; i < submeshes.size(); ++i)
		{
			const auto &submesh = submeshes[i];

			// Prepare the geometry data
			auto trianglesData = vk::AccelerationStructureGeometryTrianglesDataKHR{
			    .vertexFormat = vk::Format::eR32G32B32Sfloat,
			    .vertexData   = vertexAddr,
			    .vertexStride = sizeof(Vertex),
			    .maxVertex    = submesh.maxVertex,
			    .indexType    = vk::IndexType::eUint32,
			    .indexData    = indexAddr + submesh.indexOffset * sizeof(uint32_t)};

			vk::AccelerationStructureGeometryDataKHR geometryData(trianglesData);

			vk::AccelerationStructureGeometryKHR blasGeometry{
			    .geometryType = vk::GeometryTypeKHR::eTriangles,
			    .geometry     = geometryData,
			    .flags        = vk::GeometryFlagBitsKHR::eOpaque};
#	if LAB_TASK_LEVEL >= LAB_TASK_AS_OPAQUE_FLAG
			// TASK07
			blasGeometry.flags = (submesh.alphaCut) ? vk::GeometryFlagsKHR(0) : vk::GeometryFlagBitsKHR::eOpaque;
#	endif        // LAB_TASK_LEVEL >= LAB_TASK_AS_OPAQUE_FLAG

			vk::AccelerationStructureBuildGeometryInfoKHR blasBuildGeometryInfo{
			    .type          = vk::AccelerationStructureTypeKHR::eBottomLevel,
			    .mode          = vk::BuildAccelerationStructureModeKHR::eBuild,
			    .geometryCount = 1,
			    .pGeometries   = &blasGeometry,
			};

			// Query the memory sizes that will be needed for this BLAS
			auto primitiveCount = static_cast<uint32_t>(submesh.indexCount / 3);

			vk::AccelerationStructureBuildSizesInfoKHR blasBuildSizes =
			    device.getAccelerationStructureBuildSizesKHR(
			        vk::AccelerationStructureBuildTypeKHR::eDevice,
			        blasBuildGeometryInfo,
			        {primitiveCount});

			// Create a scratch buffer for the BLAS, this will hold temporary data
			// during the build process
			vk::raii::Buffer       scratchBuffer = nullptr;
			vk::raii::DeviceMemory scratchMemory = nullptr;
			createBuffer(blasBuildSizes.buildScratchSize,
			             vk::BufferUsageFlagBits::eStorageBuffer |
			                 vk::BufferUsageFlagBits::eShaderDeviceAddress,
			             vk::MemoryPropertyFlagBits::eDeviceLocal,
			             scratchBuffer, scratchMemory);

			// Save the scratch buffer address in the build info structure
			vk::BufferDeviceAddressInfo scratchAddressInfo{.buffer = *scratchBuffer};
			vk::DeviceAddress           scratchAddr         = device.getBufferAddressKHR(scratchAddressInfo);
			blasBuildGeometryInfo.scratchData.deviceAddress = scratchAddr;

			// Create a buffer for the BLAS itself now that we now the required size
			vk::raii::Buffer       blasBuffer = nullptr;
			vk::raii::DeviceMemory blasMemory = nullptr;
			blasBuffers.emplace_back(std::move(blasBuffer));
			blasMemories.emplace_back(std::move(blasMemory));
			createBuffer(blasBuildSizes.accelerationStructureSize,
			             vk::BufferUsageFlagBits::eAccelerationStructureStorageKHR |
			                 vk::BufferUsageFlagBits::eShaderDeviceAddress |
			                 vk::BufferUsageFlagBits::eAccelerationStructureBuildInputReadOnlyKHR,
			             vk::MemoryPropertyFlagBits::eDeviceLocal,
			             blasBuffers[i], blasMemories[i]);

			// Create and store the BLAS handle
			vk::AccelerationStructureCreateInfoKHR blasCreateInfo{
			    .buffer = blasBuffers[i],
			    .offset = 0,
			    .size   = blasBuildSizes.accelerationStructureSize,
			    .type   = vk::AccelerationStructureTypeKHR::eBottomLevel,
			};

			blasHandles.emplace_back(device.createAccelerationStructureKHR(blasCreateInfo));

			// Save the BLAS handle in the build info structure
			blasBuildGeometryInfo.dstAccelerationStructure = blasHandles[i];

			// Prepare the build range for the BLAS
			vk::AccelerationStructureBuildRangeInfoKHR blasRangeInfo{
			    .primitiveCount  = primitiveCount,
			    .primitiveOffset = 0,
			    .firstVertex     = submesh.firstVertex,
			    .transformOffset = 0};

			// Build the BLAS
			auto cmd = beginSingleTimeCommands();
			cmd->buildAccelerationStructuresKHR({blasBuildGeometryInfo}, {&blasRangeInfo});
			endSingleTimeCommands(*cmd);

			// TASK03: Create a BLAS instance for the TLAS
			vk::AccelerationStructureDeviceAddressInfoKHR addrInfo{
			    .accelerationStructure = *blasHandles[i]};
			vk::DeviceAddress blasDeviceAddr = device.getAccelerationStructureAddressKHR(addrInfo);

			vk::AccelerationStructureInstanceKHR instance{
			    .transform                      = identity,
			    .mask                           = 0xFF,
			    .accelerationStructureReference = blasDeviceAddr};

			instances.push_back(instance);

#	if LAB_TASK_LEVEL >= LAB_TASK_INSTANCE_LUT
			// TASK09: store the instance look-up table entry
			instances[i].instanceCustomIndex = static_cast<uint32_t>(i);

			instanceLUTs.push_back({static_cast<uint32_t>(submesh.materialID), submesh.indexOffset});
#	endif        // LAB_TASK_LEVEL >= LAB_TASK_INSTANCE_LUT
		}

		// TASK03: Prepare the instance data buffer
		vk::DeviceSize instBufferSize = sizeof(instances[0]) * instances.size();
		createBuffer(instBufferSize,
		             vk::BufferUsageFlagBits::eShaderDeviceAddress |
		                 vk::BufferUsageFlagBits::eTransferDst |
		                 vk::BufferUsageFlagBits::eAccelerationStructureBuildInputReadOnlyKHR,
		             vk::MemoryPropertyFlagBits::eHostVisible |
		                 vk::MemoryPropertyFlagBits::eHostCoherent,
		             instanceBuffer, instanceMemory);

		void *ptr = instanceMemory.mapMemory(0, instBufferSize);
		memcpy(ptr, instances.data(), instBufferSize);
		instanceMemory.unmapMemory();

		vk::BufferDeviceAddressInfo instanceAddrInfo{.buffer = instanceBuffer};
		vk::DeviceAddress           instanceAddr = device.getBufferAddressKHR(instanceAddrInfo);

		// Prepare the geometry (instance) data
		auto instancesData = vk::AccelerationStructureGeometryInstancesDataKHR{
		    .arrayOfPointers = vk::False,
		    .data            = instanceAddr};

		vk::AccelerationStructureGeometryDataKHR geometryData(instancesData);

		vk::AccelerationStructureGeometryKHR tlasGeometry{
		    .geometryType = vk::GeometryTypeKHR::eInstances,
		    .geometry     = geometryData};

		vk::AccelerationStructureBuildGeometryInfoKHR tlasBuildGeometryInfo{
		    .type          = vk::AccelerationStructureTypeKHR::eTopLevel,
		    .mode          = vk::BuildAccelerationStructureModeKHR::eBuild,
		    .geometryCount = 1,
		    .pGeometries   = &tlasGeometry};

#	if LAB_TASK_LEVEL >= LAB_TASK_AS_ANIMATION
		tlasBuildGeometryInfo.flags = vk::BuildAccelerationStructureFlagBitsKHR::eAllowUpdate;
#	endif        // LAB_TASK_LEVEL >= LAB_TASK_AS_ANIMATION

		// Query the memory sizes that will be needed for this TLAS
		auto primitiveCount = static_cast<uint32_t>(instances.size());

		vk::AccelerationStructureBuildSizesInfoKHR tlasBuildSizes =
		    device.getAccelerationStructureBuildSizesKHR(
		        vk::AccelerationStructureBuildTypeKHR::eDevice,
		        tlasBuildGeometryInfo,
		        {primitiveCount});

		// Create a scratch buffer for the TLAS, this will hold temporary data
		// during the build process
		createBuffer(
		    tlasBuildSizes.buildScratchSize,
		    vk::BufferUsageFlagBits::eStorageBuffer |
		        vk::BufferUsageFlagBits::eShaderDeviceAddress,
		    vk::MemoryPropertyFlagBits::eDeviceLocal,
		    tlasScratchBuffer, tlasScratchMemory);

		// Save the scratch buffer address in the build info structure
		vk::BufferDeviceAddressInfo scratchAddressInfo{.buffer = *tlasScratchBuffer};
		vk::DeviceAddress           scratchAddr         = device.getBufferAddressKHR(scratchAddressInfo);
		tlasBuildGeometryInfo.scratchData.deviceAddress = scratchAddr;

		// Create a buffer for the TLAS itself now that we now the required size
		createBuffer(
		    tlasBuildSizes.accelerationStructureSize,
		    vk::BufferUsageFlagBits::eAccelerationStructureStorageKHR |
		        vk::BufferUsageFlagBits::eShaderDeviceAddress |
		        vk::BufferUsageFlagBits::eAccelerationStructureBuildInputReadOnlyKHR,
		    vk::MemoryPropertyFlagBits::eDeviceLocal,
		    tlasBuffer, tlasMemory);

		// Create and store the TLAS handle
		vk::AccelerationStructureCreateInfoKHR tlasCreateInfo{
		    .buffer = tlasBuffer,
		    .offset = 0,
		    .size   = tlasBuildSizes.accelerationStructureSize,
		    .type   = vk::AccelerationStructureTypeKHR::eTopLevel,
		};

		tlas = device.createAccelerationStructureKHR(tlasCreateInfo);

		// Save the TLAS handle in the build info structure
		tlasBuildGeometryInfo.dstAccelerationStructure = tlas;

		// Prepare the build range for the TLAS
		vk::AccelerationStructureBuildRangeInfoKHR tlasRangeInfo{
		    .primitiveCount  = primitiveCount,
		    .primitiveOffset = 0,
		    .firstVertex     = 0,
		    .transformOffset = 0};

		// Build the TLAS
		auto cmd = beginSingleTimeCommands();

		cmd->buildAccelerationStructuresKHR({tlasBuildGeometryInfo}, {&tlasRangeInfo});

		endSingleTimeCommands(*cmd);
#endif        // LAB_TASK_LEVEL >= LAB_TASK_AS_BUILD_AND_BIND
	}

	void createDescriptorPool()
	{
		std::array poolSize{
		    vk::DescriptorPoolSize(vk::DescriptorType::eUniformBuffer, MAX_FRAMES_IN_FLIGHT),
		    vk::DescriptorPoolSize(vk::DescriptorType::eAccelerationStructureKHR, MAX_FRAMES_IN_FLIGHT),
		    vk::DescriptorPoolSize(vk::DescriptorType::eStorageBuffer, MAX_FRAMES_IN_FLIGHT * 3),        // indices, UVs, instance LUT
		    vk::DescriptorPoolSize(vk::DescriptorType::eSampler, MAX_FRAMES_IN_FLIGHT),
		    vk::DescriptorPoolSize(vk::DescriptorType::eSampledImage, (uint32_t) materials.size())};
		vk::DescriptorPoolCreateInfo poolInfo{
		    .flags = vk::DescriptorPoolCreateFlagBits::eFreeDescriptorSet |
		             vk::DescriptorPoolCreateFlagBits::eUpdateAfterBind,
		    .maxSets       = MAX_FRAMES_IN_FLIGHT + 1,        // + 1 for bindless materials
		    .poolSizeCount = static_cast<uint32_t>(poolSize.size()),
		    .pPoolSizes    = poolSize.data()};
		descriptorPool = vk::raii::DescriptorPool(device, poolInfo);
	}

	void createDescriptorSets()
	{
		// Global descriptor sets (per frame)
		std::vector<vk::DescriptorSetLayout> globalLayouts(MAX_FRAMES_IN_FLIGHT, descriptorSetLayoutGlobal);

		vk::DescriptorSetAllocateInfo allocInfoGlobal{
		    .descriptorPool     = descriptorPool,
		    .descriptorSetCount = static_cast<uint32_t>(globalLayouts.size()),
		    .pSetLayouts        = globalLayouts.data()};

		globalDescriptorSets.clear();
		globalDescriptorSets = device.allocateDescriptorSets(allocInfoGlobal);

		for (size_t i = 0; i < MAX_FRAMES_IN_FLIGHT; i++)
		{
			// Uniform buffer
			vk::DescriptorBufferInfo bufferInfo{
			    .buffer = uniformBuffers[i],
			    .offset = 0,
			    .range  = sizeof(UniformBufferObject)};

			vk::WriteDescriptorSet bufferWrite{
			    .dstSet          = globalDescriptorSets[i],
			    .dstBinding      = 0,
			    .dstArrayElement = 0,
			    .descriptorCount = 1,
			    .descriptorType  = vk::DescriptorType::eUniformBuffer,
			    .pBufferInfo     = &bufferInfo};

#if LAB_TASK_LEVEL >= LAB_TASK_AS_BUILD_AND_BIND
			// TASK04: define the acceleration structure descriptor.
			vk::WriteDescriptorSetAccelerationStructureKHR asInfo{
			    .accelerationStructureCount = 1,
			    .pAccelerationStructures    = {&*tlas}};

			vk::WriteDescriptorSet asWrite{
			    .pNext           = &asInfo,
			    .dstSet          = globalDescriptorSets[i],
			    .dstBinding      = 1,
			    .dstArrayElement = 0,
			    .descriptorCount = 1,
			    .descriptorType  = vk::DescriptorType::eAccelerationStructureKHR};
#endif        // LAB_TASK_LEVEL >= LAB_TASK_AS_BUILD_AND_BIND

			// Indices SSBO
			vk::DescriptorBufferInfo indexBufferInfo{
			    .buffer = indexBuffer,
			    .offset = 0,
			    .range  = sizeof(uint32_t) * indices.size()};

			vk::WriteDescriptorSet indexBufferWrite{
			    .dstSet          = globalDescriptorSets[i],
			    .dstBinding      = 2,
			    .dstArrayElement = 0,
			    .descriptorCount = 1,
			    .descriptorType  = vk::DescriptorType::eStorageBuffer,
			    .pBufferInfo     = &indexBufferInfo};

			// UVs SSBO
			vk::DescriptorBufferInfo uvBufferInfo{
			    .buffer = uvBuffer,
			    .offset = 0,
			    .range  = sizeof(glm::vec2) * vertices.size()};

			vk::WriteDescriptorSet uvBufferWrite{
			    .dstSet          = globalDescriptorSets[i],
			    .dstBinding      = 3,
			    .dstArrayElement = 0,
			    .descriptorCount = 1,
			    .descriptorType  = vk::DescriptorType::eStorageBuffer,
			    .pBufferInfo     = &uvBufferInfo};

#if LAB_TASK_LEVEL >= LAB_TASK_INSTANCE_LUT
			// TASK09: Instance LUT SSBO
			vk::DescriptorBufferInfo instanceLUTBufferInfo{
			    .buffer = instanceLUTBuffer,
			    .offset = 0,
			    .range  = sizeof(InstanceLUT) * instanceLUTs.size()};

			vk::WriteDescriptorSet instanceLUTBufferWrite{
			    .dstSet          = globalDescriptorSets[i],
			    .dstBinding      = 4,
			    .dstArrayElement = 0,
			    .descriptorCount = 1,
			    .descriptorType  = vk::DescriptorType::eStorageBuffer,
			    .pBufferInfo     = &instanceLUTBufferInfo};
#endif        // LAB_TASK_LEVEL >= LAB_TASK_INSTANCE_LUT

#if LAB_TASK_LEVEL >= LAB_TASK_INSTANCE_LUT
			// TASK09: Include the instance look-up table descriptor
			std::array<vk::WriteDescriptorSet, 5> descriptorWrites{bufferWrite, asWrite, indexBufferWrite, uvBufferWrite, instanceLUTBufferWrite};
#elif LAB_TASK_LEVEL >= LAB_TASK_AS_BUILD_AND_BIND
			// TASK04: Include the acceleration structure descriptor
			std::array<vk::WriteDescriptorSet, 4> descriptorWrites{bufferWrite, asWrite, indexBufferWrite, uvBufferWrite};
#else
			std::array<vk::WriteDescriptorSet, 3> descriptorWrites{bufferWrite, indexBufferWrite, uvBufferWrite};
#endif

			device.updateDescriptorSets(descriptorWrites, {});
		}

		// Material descriptor sets (per material)
		std::vector<uint32_t>                                variableCounts = {static_cast<uint32_t>(textureImageViews.size())};
		vk::DescriptorSetVariableDescriptorCountAllocateInfo variableCountInfo{
		    .descriptorSetCount = 1,
		    .pDescriptorCounts  = variableCounts.data()};

		std::vector<vk::DescriptorSetLayout> layouts{*descriptorSetLayoutMaterial};

		vk::DescriptorSetAllocateInfo allocInfo{
		    .pNext              = &variableCountInfo,
		    .descriptorPool     = descriptorPool,
		    .descriptorSetCount = static_cast<uint32_t>(layouts.size()),
		    .pSetLayouts        = layouts.data()};

		materialDescriptorSets = device.allocateDescriptorSets(allocInfo);

		// Sampler
		vk::DescriptorImageInfo samplerInfo{
		    .sampler = textureSampler};

		vk::WriteDescriptorSet samplerWrite{
		    .dstSet          = materialDescriptorSets[0],
		    .dstBinding      = 0,
		    .dstArrayElement = 0,
		    .descriptorCount = 1,
		    .descriptorType  = vk::DescriptorType::eSampler,
		    .pImageInfo      = &samplerInfo};

		device.updateDescriptorSets({samplerWrite}, {});

		// Textures
		std::vector<vk::DescriptorImageInfo> imageInfos;
		imageInfos.reserve(textureImageViews.size());
		for (auto &iv : textureImageViews)
		{
			vk::DescriptorImageInfo imageInfo{
			    .imageView   = iv,
			    .imageLayout = vk::ImageLayout::eShaderReadOnlyOptimal};
			imageInfos.push_back(imageInfo);
		}

		vk::WriteDescriptorSet materialWrite{
		    .dstSet          = materialDescriptorSets[0],
		    .dstBinding      = 1,
		    .dstArrayElement = 0,
		    .descriptorCount = static_cast<uint32_t>(imageInfos.size()),
		    .descriptorType  = vk::DescriptorType::eSampledImage,
		    .pImageInfo      = imageInfos.data()};

		device.updateDescriptorSets({materialWrite}, {});
	}

	void createBuffer(vk::DeviceSize size, vk::BufferUsageFlags usage, vk::MemoryPropertyFlags properties, vk::raii::Buffer &buffer, vk::raii::DeviceMemory &bufferMemory)
	{
		vk::BufferCreateInfo bufferInfo{
		    .size        = size,
		    .usage       = usage,
		    .sharingMode = vk::SharingMode::eExclusive};
		buffer                                 = vk::raii::Buffer(device, bufferInfo);
		vk::MemoryRequirements memRequirements = buffer.getMemoryRequirements();
		vk::MemoryAllocateInfo allocInfo{
		    .allocationSize  = memRequirements.size,
		    .memoryTypeIndex = findMemoryType(memRequirements.memoryTypeBits, properties)};
		vk::MemoryAllocateFlagsInfo allocFlagsInfo{};
		if (usage & vk::BufferUsageFlagBits::eShaderDeviceAddress)
		{
			allocFlagsInfo.flags = vk::MemoryAllocateFlagBits::eDeviceAddress;
			allocInfo.pNext      = &allocFlagsInfo;
		}
		bufferMemory = vk::raii::DeviceMemory(device, allocInfo);
		buffer.bindMemory(bufferMemory, 0);
	}

	std::unique_ptr<vk::raii::CommandBuffer> beginSingleTimeCommands()
	{
		vk::CommandBufferAllocateInfo allocInfo{
		    .commandPool        = commandPool,
		    .level              = vk::CommandBufferLevel::ePrimary,
		    .commandBufferCount = 1};
		std::unique_ptr<vk::raii::CommandBuffer> commandBuffer = std::make_unique<vk::raii::CommandBuffer>(std::move(vk::raii::CommandBuffers(device, allocInfo).front()));

		vk::CommandBufferBeginInfo beginInfo{
		    .flags = vk::CommandBufferUsageFlagBits::eOneTimeSubmit};
		commandBuffer->begin(beginInfo);

		return commandBuffer;
	}

	void endSingleTimeCommands(const vk::raii::CommandBuffer &commandBuffer) const
	{
		commandBuffer.end();

		vk::SubmitInfo submitInfo{.commandBufferCount = 1, .pCommandBuffers = &*commandBuffer};
		graphicsQueue.submit(submitInfo, nullptr);
		graphicsQueue.waitIdle();
	}

	void copyBuffer(vk::raii::Buffer &srcBuffer, vk::raii::Buffer &dstBuffer, vk::DeviceSize size)
	{
		vk::CommandBufferAllocateInfo allocInfo{.commandPool = commandPool, .level = vk::CommandBufferLevel::ePrimary, .commandBufferCount = 1};
		vk::raii::CommandBuffer       commandCopyBuffer = std::move(device.allocateCommandBuffers(allocInfo).front());
		commandCopyBuffer.begin(vk::CommandBufferBeginInfo{.flags = vk::CommandBufferUsageFlagBits::eOneTimeSubmit});
		commandCopyBuffer.copyBuffer(*srcBuffer, *dstBuffer, vk::BufferCopy{.size = size});
		commandCopyBuffer.end();
		graphicsQueue.submit(vk::SubmitInfo{.commandBufferCount = 1, .pCommandBuffers = &*commandCopyBuffer}, nullptr);
		graphicsQueue.waitIdle();
	}

	uint32_t findMemoryType(uint32_t typeFilter, vk::MemoryPropertyFlags properties)
	{
		vk::PhysicalDeviceMemoryProperties memProperties = physicalDevice.getMemoryProperties();

		for (uint32_t i = 0; i < memProperties.memoryTypeCount; i++)
		{
			if ((typeFilter & (1 << i)) && (memProperties.memoryTypes[i].propertyFlags & properties) == properties)
			{
				return i;
			}
		}

		throw std::runtime_error("failed to find suitable memory type!");
	}

	void createCommandBuffers()
	{
		commandBuffers.clear();
		vk::CommandBufferAllocateInfo allocInfo{.commandPool = commandPool, .level = vk::CommandBufferLevel::ePrimary, .commandBufferCount = MAX_FRAMES_IN_FLIGHT};
		commandBuffers = vk::raii::CommandBuffers(device, allocInfo);
	}

	void recordCommandBuffer(uint32_t imageIndex)
	{
		auto &commandBuffer = commandBuffers[frameIndex];
		commandBuffer.begin({});
		// Before starting rendering, transition the swapchain image to COLOR_ATTACHMENT_OPTIMAL
		transition_image_layout(
		    swapChainImages[imageIndex],
		    vk::ImageLayout::eUndefined,
		    vk::ImageLayout::eColorAttachmentOptimal,
		    {},                                                        // srcAccessMask (no need to wait for previous operations)
		    vk::AccessFlagBits2::eColorAttachmentWrite,                // dstAccessMask
		    vk::PipelineStageFlagBits2::eColorAttachmentOutput,        // srcStage
		    vk::PipelineStageFlagBits2::eColorAttachmentOutput,        // dstStage
		    vk::ImageAspectFlagBits::eColor);
		// Transition depth image to depth attachment optimal layout
		transition_image_layout(
		    *depthImage,
		    vk::ImageLayout::eUndefined,
		    vk::ImageLayout::eDepthAttachmentOptimal,
		    vk::AccessFlagBits2::eDepthStencilAttachmentWrite,
		    vk::AccessFlagBits2::eDepthStencilAttachmentWrite,
		    vk::PipelineStageFlagBits2::eEarlyFragmentTests | vk::PipelineStageFlagBits2::eLateFragmentTests,
		    vk::PipelineStageFlagBits2::eEarlyFragmentTests | vk::PipelineStageFlagBits2::eLateFragmentTests,
		    vk::ImageAspectFlagBits::eDepth);

		vk::ClearValue clearColor = vk::ClearColorValue(0.0f, 0.0f, 0.0f, 1.0f);
		vk::ClearValue clearDepth = vk::ClearDepthStencilValue(1.0f, 0);

		/* TASK01: Check the setup for dynamic rendering
		 *
		 * With dynamic rendering, we specify the image view and load/store operations directly
		 * in the vk::RenderingAttachmentInfo structure.
		 * This approach eliminates the need for explicit render pass and framebuffer objects,
		 * simplifying the code and providing flexibility to change attachments at runtime.
		 */

		vk::RenderingAttachmentInfo colorAttachmentInfo = {
		    .imageView   = swapChainImageViews[imageIndex],
		    .imageLayout = vk::ImageLayout::eColorAttachmentOptimal,
		    .loadOp      = vk::AttachmentLoadOp::eClear,
		    .storeOp     = vk::AttachmentStoreOp::eStore,
		    .clearValue  = clearColor};

		vk::RenderingAttachmentInfo depthAttachmentInfo = {
		    .imageView   = depthImageView,
		    .imageLayout = vk::ImageLayout::eDepthStencilAttachmentOptimal,
		    .loadOp      = vk::AttachmentLoadOp::eClear,
		    .storeOp     = vk::AttachmentStoreOp::eDontCare,
		    .clearValue  = clearDepth};

		// The vk::RenderingInfo structure combines these attachments with other rendering parameters.
		vk::RenderingInfo renderingInfo = {
		    .renderArea           = {.offset = {0, 0}, .extent = swapChainExtent},
		    .layerCount           = 1,
		    .colorAttachmentCount = 1,
		    .pColorAttachments    = &colorAttachmentInfo,
		    .pDepthAttachment     = &depthAttachmentInfo};

		// Note: .beginRendering replaces the previous .beginRenderPass call.
		commandBuffer.beginRendering(renderingInfo);

		commandBuffer.bindPipeline(vk::PipelineBindPoint::eGraphics, *graphicsPipeline);
		commandBuffer.setViewport(0, vk::Viewport(0.0f, 0.0f, static_cast<float>(swapChainExtent.width), static_cast<float>(swapChainExtent.height), 0.0f, 1.0f));
		commandBuffer.setScissor(0, vk::Rect2D(vk::Offset2D(0, 0), swapChainExtent));
		commandBuffer.bindVertexBuffers(0, *vertexBuffer, {0});
		commandBuffer.bindIndexBuffer(*indexBuffer, 0, vk::IndexType::eUint32);
		commandBuffer.bindDescriptorSets(vk::PipelineBindPoint::eGraphics, pipelineLayout, 0, *globalDescriptorSets[frameIndex], nullptr);
		commandBuffer.bindDescriptorSets(vk::PipelineBindPoint::eGraphics, pipelineLayout, 1, *materialDescriptorSets[0], nullptr);

		for (auto &sub : submeshes)
		{
			// TASK09: Bindless resources
			PushConstant pushConstant = {
			    .materialIndex = sub.materialID < 0 ? 0u : static_cast<uint32_t>(sub.materialID),
#if LAB_TASK_LEVEL >= LAB_TASK_REFLECTIONS
			    .reflective = sub.reflective
#endif        // LAB_TASK_LEVEL >= LAB_TASK_REFLECTIONS
			};
			commandBuffer.pushConstants<PushConstant>(pipelineLayout, vk::ShaderStageFlagBits::eFragment, 0, pushConstant);

			commandBuffer.drawIndexed(sub.indexCount, 1, sub.indexOffset, 0, 0);
		}

		commandBuffer.endRendering();

		// After rendering, transition the swapchain image to PRESENT_SRC
		transition_image_layout(
		    swapChainImages[imageIndex],
		    vk::ImageLayout::eColorAttachmentOptimal,
		    vk::ImageLayout::ePresentSrcKHR,
		    vk::AccessFlagBits2::eColorAttachmentWrite,                // srcAccessMask
		    {},                                                        // dstAccessMask
		    vk::PipelineStageFlagBits2::eColorAttachmentOutput,        // srcStage
		    vk::PipelineStageFlagBits2::eBottomOfPipe,                 // dstStage
		    vk::ImageAspectFlagBits::eColor);

		commandBuffer.end();
	}

	void transition_image_layout(
	    vk::Image               image,
	    vk::ImageLayout         old_layout,
	    vk::ImageLayout         new_layout,
	    vk::AccessFlags2        src_access_mask,
	    vk::AccessFlags2        dst_access_mask,
	    vk::PipelineStageFlags2 src_stage_mask,
	    vk::PipelineStageFlags2 dst_stage_mask,
	    vk::ImageAspectFlags    image_aspect_flags)
	{
		vk::ImageMemoryBarrier2 barrier = {
		    .srcStageMask        = src_stage_mask,
		    .srcAccessMask       = src_access_mask,
		    .dstStageMask        = dst_stage_mask,
		    .dstAccessMask       = dst_access_mask,
		    .oldLayout           = old_layout,
		    .newLayout           = new_layout,
		    .srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
		    .dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED,
		    .image               = image,
		    .subresourceRange    = {
		           .aspectMask     = image_aspect_flags,
		           .baseMipLevel   = 0,
		           .levelCount     = 1,
		           .baseArrayLayer = 0,
		           .layerCount     = 1}};
		vk::DependencyInfo dependency_info = {
		    .dependencyFlags         = {},
		    .imageMemoryBarrierCount = 1,
		    .pImageMemoryBarriers    = &barrier};
		commandBuffers[frameIndex].pipelineBarrier2(dependency_info);
	}

	void createSyncObjects()
	{
		assert(presentCompleteSemaphores.empty() && renderFinishedSemaphores.empty() && inFlightFences.empty());

		for (size_t i = 0; i < swapChainImages.size(); i++)
		{
			renderFinishedSemaphores.emplace_back(device, vk::SemaphoreCreateInfo());
		}

		for (size_t i = 0; i < MAX_FRAMES_IN_FLIGHT; i++)
		{
			presentCompleteSemaphores.emplace_back(device, vk::SemaphoreCreateInfo());
			inFlightFences.emplace_back(device, vk::FenceCreateInfo{.flags = vk::FenceCreateFlagBits::eSignaled});
		}
	}

	void updateUniformBuffer(uint32_t currentImage)
	{
		static auto startTime = std::chrono::high_resolution_clock::now();

		auto  currentTime = std::chrono::high_resolution_clock::now();
		float time        = std::chrono::duration<float>(currentTime - startTime).count();

		auto eye = glm::vec3(2.0f, 2.0f, 2.0f);

		ubo.model = rotate(glm::mat4(1.0f), time * 0.1f * glm::radians(90.0f), glm::vec3(0.0f, 0.0f, 1.0f));
		ubo.view  = lookAt(eye, glm::vec3(0.0f, 0.0f, 0.0f), glm::vec3(0.0f, 0.0f, 1.0f));
		ubo.proj  = glm::perspective(glm::radians(45.0f), static_cast<float>(swapChainExtent.width) / static_cast<float>(swapChainExtent.height), 0.1f, 10.0f);
		ubo.proj[1][1] *= -1;
		ubo.cameraPos = eye;

		memcpy(uniformBuffersMapped[currentImage], &ubo, sizeof(ubo));
	}

#if LAB_TASK_LEVEL >= LAB_TASK_AS_ANIMATION
	void updateTopLevelAS(const glm::mat4 &model)
	{
		vk::TransformMatrixKHR tm{};
		auto                  &M = model;
		tm.matrix                = std::array<std::array<float, 4>, 3>{{std::array<float, 4>{M[0][0], M[1][0], M[2][0], M[3][0]},
		                                                                std::array<float, 4>{M[0][1], M[1][1], M[2][1], M[3][1]},
		                                                                std::array<float, 4>{M[0][2], M[1][2], M[2][2], M[3][2]}}};

		// TASK06: update the instances to use the new transform matrix.
		for (auto &instance : instances)
		{
			instance.setTransform(tm);
		}

		auto           primitiveCount = static_cast<uint32_t>(instances.size());
		vk::DeviceSize instBufferSize = sizeof(instances[0]) * primitiveCount;

		vk::raii::Buffer       stagingBuffer({});
		vk::raii::DeviceMemory stagingBufferMemory({});
		createBuffer(instBufferSize, vk::BufferUsageFlagBits::eTransferSrc, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent, stagingBuffer, stagingBufferMemory);

		void *dataStaging = stagingBufferMemory.mapMemory(0, instBufferSize);
		memcpy(dataStaging, instances.data(), instBufferSize);
		stagingBufferMemory.unmapMemory();

		copyBuffer(stagingBuffer, instanceBuffer, instBufferSize);

		vk::BufferDeviceAddressInfo instanceAddrInfo{.buffer = instanceBuffer};
		vk::DeviceAddress           instanceAddr = device.getBufferAddressKHR(instanceAddrInfo);

		// Prepare the geometry (instance) data
		auto instancesData = vk::AccelerationStructureGeometryInstancesDataKHR{
		    .arrayOfPointers = vk::False,
		    .data            = instanceAddr};

		vk::AccelerationStructureGeometryDataKHR geometryData(instancesData);

		vk::AccelerationStructureGeometryKHR tlasGeometry{
		    .geometryType = vk::GeometryTypeKHR::eInstances,
		    .geometry     = geometryData};

		// TASK06: Note the new parameters to re-build the TLAS in-place
		vk::AccelerationStructureBuildGeometryInfoKHR tlasBuildGeometryInfo{
		    .type                     = vk::AccelerationStructureTypeKHR::eTopLevel,
		    .flags                    = vk::BuildAccelerationStructureFlagBitsKHR::eAllowUpdate,
		    .mode                     = vk::BuildAccelerationStructureModeKHR::eUpdate,
		    .srcAccelerationStructure = tlas,
		    .dstAccelerationStructure = tlas,
		    .geometryCount            = 1,
		    .pGeometries              = &tlasGeometry};

		vk::BufferDeviceAddressInfo scratchAddressInfo{.buffer = *tlasScratchBuffer};
		vk::DeviceAddress           scratchAddr         = device.getBufferAddressKHR(scratchAddressInfo);
		tlasBuildGeometryInfo.scratchData.deviceAddress = scratchAddr;

		// Prepare the build range for the TLAS
		vk::AccelerationStructureBuildRangeInfoKHR tlasRangeInfo{
		    .primitiveCount  = primitiveCount,
		    .primitiveOffset = 0,
		    .firstVertex     = 0,
		    .transformOffset = 0};

		// Re-build the TLAS
		auto cmd = beginSingleTimeCommands();

		// Pre-build barrier
		vk::MemoryBarrier preBarrier{
		    .srcAccessMask = vk::AccessFlagBits::eAccelerationStructureWriteKHR | vk::AccessFlagBits::eTransferWrite | vk::AccessFlagBits::eShaderRead,
		    .dstAccessMask = vk::AccessFlagBits::eAccelerationStructureReadKHR | vk::AccessFlagBits::eAccelerationStructureWriteKHR};

		cmd->pipelineBarrier(
		    vk::PipelineStageFlagBits::eAccelerationStructureBuildKHR | vk::PipelineStageFlagBits::eTransfer | vk::PipelineStageFlagBits::eFragmentShader,        // srcStageMask
		    vk::PipelineStageFlagBits::eAccelerationStructureBuildKHR,                                                                                            // dstStageMask
		    {},                                                                                                                                                   // dependencyFlags
		    preBarrier,                                                                                                                                           // memoryBarriers
		    {},                                                                                                                                                   // bufferMemoryBarriers
		    {}                                                                                                                                                    // imageMemoryBarriers
		);

		cmd->buildAccelerationStructuresKHR({tlasBuildGeometryInfo}, {&tlasRangeInfo});

		// Post-build barrier
		vk::MemoryBarrier postBarrier{
		    .srcAccessMask = vk::AccessFlagBits::eAccelerationStructureWriteKHR,
		    .dstAccessMask = vk::AccessFlagBits::eAccelerationStructureReadKHR | vk::AccessFlagBits::eShaderRead};

		cmd->pipelineBarrier(
		    vk::PipelineStageFlagBits::eAccelerationStructureBuildKHR,                                                     // srcStageMask
		    vk::PipelineStageFlagBits::eAccelerationStructureBuildKHR | vk::PipelineStageFlagBits::eFragmentShader,        // dstStageMask
		    {},                                                                                                            // dependencyFlags
		    postBarrier,                                                                                                   // memoryBarriers
		    {},                                                                                                            // bufferMemoryBarriers
		    {}                                                                                                             // imageMemoryBarriers
		);

		endSingleTimeCommands(*cmd);
	}
#endif        // LAB_TASK_LEVEL >= LAB_TASK_AS_ANIMATION

	void drawFrame()
	{
		// Note: inFlightFences, presentCompleteSemaphores, and commandBuffers are indexed by frameIndex,
		//       while renderFinishedSemaphores is indexed by imageIndex
		auto fenceResult = device.waitForFences(*inFlightFences[frameIndex], vk::True, UINT64_MAX);
		if (fenceResult != vk::Result::eSuccess)
		{
			throw std::runtime_error("failed to wait for fence!");
		}

		auto [result, imageIndex] = swapChain.acquireNextImage(UINT64_MAX, *presentCompleteSemaphores[frameIndex], nullptr);

		// Due to VULKAN_HPP_HANDLE_ERROR_OUT_OF_DATE_AS_SUCCESS being defined, eErrorOutOfDateKHR can be checked as a result
		// here and does not need to be caught by an exception.
		if (result == vk::Result::eErrorOutOfDateKHR)
		{
			recreateSwapChain();
			return;
		}
		// On other success codes than eSuccess and eSuboptimalKHR we just throw an exception.
		// On any error code, aquireNextImage already threw an exception.
		if (result != vk::Result::eSuccess && result != vk::Result::eSuboptimalKHR)
		{
			assert(result == vk::Result::eTimeout || result == vk::Result::eNotReady);
			throw std::runtime_error("failed to acquire swap chain image!");
		}
		updateUniformBuffer(frameIndex);
#if LAB_TASK_LEVEL >= LAB_TASK_AS_ANIMATION
		// TASK06: Update the TLAS with the current model matrix
		updateTopLevelAS(ubo.model);
#endif        // LAB_TASK_LEVEL >= LAB_TASK_AS_ANIMATION

		// Only reset the fence if we are submitting work
		device.resetFences(*inFlightFences[frameIndex]);

		commandBuffers[frameIndex].reset();
		recordCommandBuffer(imageIndex);

		vk::PipelineStageFlags waitDestinationStageMask(vk::PipelineStageFlagBits::eColorAttachmentOutput);
		const vk::SubmitInfo   submitInfo{.waitSemaphoreCount   = 1,
		                                  .pWaitSemaphores      = &*presentCompleteSemaphores[frameIndex],
		                                  .pWaitDstStageMask    = &waitDestinationStageMask,
		                                  .commandBufferCount   = 1,
		                                  .pCommandBuffers      = &*commandBuffers[frameIndex],
		                                  .signalSemaphoreCount = 1,
		                                  .pSignalSemaphores    = &*renderFinishedSemaphores[imageIndex]};
		graphicsQueue.submit(submitInfo, *inFlightFences[frameIndex]);

		const vk::PresentInfoKHR presentInfoKHR{.waitSemaphoreCount = 1,
		                                        .pWaitSemaphores    = &*renderFinishedSemaphores[imageIndex],
		                                        .swapchainCount     = 1,
		                                        .pSwapchains        = &*swapChain,
		                                        .pImageIndices      = &imageIndex};
		result = presentQueue.presentKHR(presentInfoKHR);
		// Due to VULKAN_HPP_HANDLE_ERROR_OUT_OF_DATE_AS_SUCCESS being defined, eErrorOutOfDateKHR can be checked as a result
		// here and does not need to be caught by an exception.
		if ((result == vk::Result::eSuboptimalKHR) || (result == vk::Result::eErrorOutOfDateKHR) || framebufferResized)
		{
			framebufferResized = false;
			recreateSwapChain();
		}
		else
		{
			// There are no other success codes than eSuccess; on any error code, presentKHR already threw an exception.
			assert(result == vk::Result::eSuccess);
		}
		frameIndex = (frameIndex + 1) % MAX_FRAMES_IN_FLIGHT;
	}

	[[nodiscard]] vk::raii::ShaderModule createShaderModule(const std::vector<char> &code) const
	{
		vk::ShaderModuleCreateInfo createInfo{.codeSize = code.size(), .pCode = reinterpret_cast<const uint32_t *>(code.data())};
		vk::raii::ShaderModule     shaderModule{device, createInfo};

		return shaderModule;
	}

	static uint32_t chooseSwapMinImageCount(vk::SurfaceCapabilitiesKHR const &surfaceCapabilities)
	{
		auto minImageCount = std::max(3u, surfaceCapabilities.minImageCount);
		if ((0 < surfaceCapabilities.maxImageCount) && (surfaceCapabilities.maxImageCount < minImageCount))
		{
			minImageCount = surfaceCapabilities.maxImageCount;
		}
		return minImageCount;
	}

	static vk::SurfaceFormatKHR chooseSwapSurfaceFormat(const std::vector<vk::SurfaceFormatKHR> &availableFormats)
	{
		assert(!availableFormats.empty());
		const auto formatIt = std::ranges::find_if(
		    availableFormats,
		    [](const auto &format) { return format.format == vk::Format::eB8G8R8A8Srgb && format.colorSpace == vk::ColorSpaceKHR::eSrgbNonlinear; });
		return formatIt != availableFormats.end() ? *formatIt : availableFormats[0];
	}

	static vk::PresentModeKHR chooseSwapPresentMode(std::vector<vk::PresentModeKHR> const &availablePresentModes)
	{
		return std::ranges::any_of(availablePresentModes,
		                           [](const vk::PresentModeKHR value) { return vk::PresentModeKHR::eMailbox == value; }) ?
		           vk::PresentModeKHR::eMailbox :
		           vk::PresentModeKHR::eFifo;
	}

	vk::Extent2D chooseSwapExtent(vk::SurfaceCapabilitiesKHR const &capabilities)
	{
		if (capabilities.currentExtent.width != std::numeric_limits<uint32_t>::max())
		{
			return capabilities.currentExtent;
		}
		int width, height;
		glfwGetFramebufferSize(window, &width, &height);

		return {
		    std::clamp<uint32_t>(width, capabilities.minImageExtent.width, capabilities.maxImageExtent.width),
		    std::clamp<uint32_t>(height, capabilities.minImageExtent.height, capabilities.maxImageExtent.height)};
	}

	[[nodiscard]] std::vector<const char *> getRequiredInstanceExtensions()
	{
		uint32_t glfwExtensionCount = 0;
		auto     glfwExtensions     = glfwGetRequiredInstanceExtensions(&glfwExtensionCount);

		std::vector extensions(glfwExtensions, glfwExtensions + glfwExtensionCount);
		if (enableValidationLayers)
		{
			extensions.push_back(vk::EXTDebugUtilsExtensionName);
		}

		return extensions;
	}

	static VKAPI_ATTR vk::Bool32 VKAPI_CALL debugCallback(vk::DebugUtilsMessageSeverityFlagBitsEXT severity, vk::DebugUtilsMessageTypeFlagsEXT type, const vk::DebugUtilsMessengerCallbackDataEXT *pCallbackData, void *)
	{
		if (severity == vk::DebugUtilsMessageSeverityFlagBitsEXT::eError || severity == vk::DebugUtilsMessageSeverityFlagBitsEXT::eWarning)
		{
			std::cerr << "validation layer: type " << to_string(type) << " msg: " << pCallbackData->pMessage << std::endl;
		}

		return vk::False;
	}

	static std::vector<char> readFile(const std::string &filename)
	{
		std::ifstream file(filename, std::ios::ate | std::ios::binary);
		if (!file.is_open())
		{
			throw std::runtime_error("failed to open file!");
		}
		std::vector<char> buffer(file.tellg());
		file.seekg(0, std::ios::beg);
		file.read(buffer.data(), static_cast<std::streamsize>(buffer.size()));
		file.close();
		return buffer;
	}
};

int main()
{
	try
	{
		VulkanRaytracingApplication app;
		app.run();
	}
	catch (const std::exception &e)
	{
		std::cerr << e.what() << std::endl;
		return EXIT_FAILURE;
	}

	return EXIT_SUCCESS;
}