整体思路：
GLM生产出的透视投影矩阵默认使用OpenGL的深度范围，收敛在 -1.0 到 1.0。我们需要使用GLM_FORCE_DEPTH_ZERO_TO_ONE定义将其配置为使用 0.0 到 1.0 的Vulkan深度范围 ； 使用 tinyobjloader 库来从OBJ文件中加载vertices和faces数据 ，并将 tiny_obj_loader.h 头文件（https://github.com/tinyobjloader/tinyobjloader） 下载到库目录中的文件夹中；模型使用的是viking_room.obj（https://github.com/Overv/VulkanTutorial/tree/master/resources）。
注意： 索引数据类型为uint32_t 。因为将会有超过65535个或者更多的顶点。
顶点着色器：
#version 450
#extension GL_ARB_separate_shader_objects : enable
layout(binding = 0) uniform UniformBufferObject {
mat4 ModelMatrix;
mat4 ViewMatrix;
mat4 ProjectiveMatrix;
} UBO;
layout(location = 0) in vec3 inPosition;
layout(location = 1) in vec3 inColor;
layout(location = 2) in vec2 inTexCoord;
layout(location = 0) out vec3 fragColor;
layout(location = 1) out vec2 fragTexCoord;
out gl_PerVertex {
vec4 gl_Position;
};
void main() {
gl_Position = UBO.ProjectiveMatrix * UBO.ViewMatrix * UBO.ModelMatrix * vec4(inPosition,1.0);
fragColor = inColor;
fragTexCoord = inTexCoord;
}片段着色器：
#version 450
#extension GL_ARB_separate_shader_objects : enable
layout(binding = 1) uniform sampler2D texSampler;
layout(location = 0) in vec3 fragColor;
layout(location = 1) in vec2 fragTexCoord;
layout(location = 0) out vec4 outColor;
void main() {
outColor = texture(texSampler, fragTexCoord);
}
主程序：
// LNK1107.cpp
// compile with: /clr /LD
#define GLFW_INCLUDE_VULKAN
#include <GLFW/glfw3.h>
#include <glm/glm.hpp>
#include <glm/gtx/hash.hpp>
#define GLM_FORCE_RADIANS //确保像 glm::rotate 这样的函数使用弧度制作为参数，以避免任何可能的混淆
#define GLM_FORCE_DEPTH_ZERO_TO_ONE //借助GLM生产出的透视投影矩阵默认使用OpenGL的深度范围，收敛在 -1.0 到 1.0。我们需要使用GLM_FORCE_DEPTH_ZERO_TO_ONE定义将其配置为使用 0.0 到 1.0 的Vulkan深度范围。
#define GLM_ENABLE_EXPERIMENTAL
#include <glm/gtc/matrix_transform.hpp> //对外提供用于生成模型变换矩阵
#include <chrono> //对外提供计时功能
#define STB_IMAGE_IMPLEMENTATION
#include <STB_IMAGE/stb_image.h>
#define TINYOBJLOADER_IMPLEMENTATION
#include <tinyobjloader/tiny_obj_loader.h>
#include <fstream>
#include <iostream>
#include <stdexcept>
#include <vector>
#include <cstring>
#include <cstdlib>
#include <optional>
#include <set>
#include <algorithm>
#include <array>
#include <unordered_map> //来跟踪唯一的顶点和相应的索引信息，删除重复的顶点数据
//总体功能
//1.实现纹理映射，当图像已经有了贴图，就需要图形管线采样，读取纹素
//2.图像不能直接访问，需要借助图像视图访问纹理图像
//3.1使用描述符访问图像资源，修改描述符布局，描述符对象池和描述符集合，以包括这样一个组合的图像采样器描述符。完成之后，我们会添加纹理贴图坐标到Vertex数据中，
//并修改片段着色器从纹理中读取颜色，而不是内插顶点颜色。
const int WIDTH = 800;
const int HEIGHT = 600;
const int MAX_FRAMES_IN_FLIGHT = 2;
const std::string MODEL_PATH = "viking_room.obj";
const std::string TEXTURE_PATH = "viking_room.png";
//SDK通过请求VK_LAYER_LUNARG_standard_validaction层，
//来隐式的开启有所关于诊断layers，
//从而避免明确的指定所有的明确的诊断层。
const std::vector<const char*> ValidationLayers = {
"VK_LAYER_KHRONOS_validation"
};
//定义所需的设备扩展列表
const std::vector<const char*> DeviceExtensions = {
VK_KHR_SWAPCHAIN_EXTENSION_NAME
};
//添加两个配置变量来指定要启用的layers以及是否开启它们
#ifdef NDEBUG
const bool enableValidationLayers = false;
#else
const bool enableValidationLayers = true;
#endif
//填写完结构体VkDebugUtilsMessengerCreateInfoEXT 信息后，将它作为一个参数调用vkCreateDebugUtilsMessengerEXT函数来创建VkDebugUtilsMessengerEXT对象。
//但由于vkCreateDebugUtilsMessengerEXT函数是一个扩展函数，不会被Vulkan库自动加载，需要使用vkGetInstanceProcAddr函数来加载
//FUNCTION:一个代理函数，主要是为了加载vkCreateDebugUtilsMessengerEXT函数
VkResult createDebugUtilsMessengerEXT(VkInstance instance, const VkDebugUtilsMessengerCreateInfoEXT* pCreateInfo, const VkAllocationCallbacks* pAllocator, VkDebugUtilsMessengerEXT* pDebugMessenger) {
auto func = (PFN_vkCreateDebugUtilsMessengerEXT)vkGetInstanceProcAddr(instance, "vkCreateDebugUtilsMessengerEXT");
if (func != nullptr) {
return func(instance, pCreateInfo, pAllocator, pDebugMessenger);
}
else {
return VK_ERROR_EXTENSION_NOT_PRESENT;
}
}
//清理vkDestroyDebugUtilsMessengerEXT对象
void destroyDebugUtilsMessengerEXT(VkInstance vInstance,
VkDebugUtilsMessengerEXT vCallback,
const VkAllocationCallbacks* pAllocator)
{
auto func = (PFN_vkDestroyDebugUtilsMessengerEXT)vkGetInstanceProcAddr(vInstance, "vkDestroyDebugUtilsMessengerEXT");
if (func != nullptr) {
func(vInstance, vCallback, pAllocator);
}
}
//具体的Vulkan实现可能对窗口系统进行了支持，但这并不意味着所有平台的Vulkan实现都支持同样的特性
//需要扩展isDeviceSuitable函数来确保设备可以在我们创建的表面上显示图像
struct EQueueFamilyIndices
{ //支持绘制指令的队列簇和支持表现的队列簇
int GraphicsFamily = -1;
int PresentFamily = -1;
bool isComplete()
{
return (GraphicsFamily >= 0 && PresentFamily >= 0);
}
};
//检查交换链的细节：1.surface特性 2.surface格式 3.可用的呈现模式
struct ESwapChainSupportDetails {
VkSurfaceCapabilitiesKHR Capablities;
std::vector<VkSurfaceFormatKHR> SurfaceFormats;
std::vector<VkPresentModeKHR> PresentModes;
};
//顶点数据
struct EVertex {
glm::vec3 Pos;
glm::vec3 Color;
glm::vec2 TexCoord;
//一旦数据被提交到GPU的显存中，就需要告诉Vulkan传递到顶点着色器中数据的格式。有两个结构体用于描述这部分信息
//1.顶点数据绑定:
static VkVertexInputBindingDescription getBindingDescription()
{
VkVertexInputBindingDescription BindingDescription = {};
//定数据条目之间的间隔字节数以及是否每个顶点之后或者每个instance之后移动到下一个条目
BindingDescription.binding = 0;
BindingDescription.stride = sizeof(EVertex);
BindingDescription.inputRate = VK_VERTEX_INPUT_RATE_VERTEX; // 移动到每个顶点后的下一个数据条目
return BindingDescription;
}
//2.如何处理顶点的输入
static std::array<VkVertexInputAttributeDescription, 3> getAttributeDescriptions()
{
std::array<VkVertexInputAttributeDescription, 3> AttributeDescriptions = {};
//一个属性描述结构体最终描述了顶点属性如何从对应的绑定描述过的顶点数据来解析数据
//|Pos属性
AttributeDescriptions[0].binding = 0;
AttributeDescriptions[0].location = 0; //与VertexShader相对应，Pos属性
AttributeDescriptions[0].format = VK_FORMAT_R32G32B32_SFLOAT; //32bit单精度数据
AttributeDescriptions[0].offset = offsetof(EVertex, Pos);
//Color属性
AttributeDescriptions[1].binding = 0;
AttributeDescriptions[1].location = 1; //Color属性
AttributeDescriptions[1].format = VK_FORMAT_R32G32B32_SFLOAT;
AttributeDescriptions[1].offset = offsetof(EVertex, Color);
//TexCoord属性
AttributeDescriptions[2].binding = 0;
AttributeDescriptions[2].location = 2;
AttributeDescriptions[2].format = VK_FORMAT_R32G32_SFLOAT;
AttributeDescriptions[2].offset = offsetof(EVertex, TexCoord);
return AttributeDescriptions;
}
bool operator==(const EVertex& other) const
{
return Pos == other.Pos && Color == other.Color && TexCoord == other.TexCoord;
}
};
namespace std {
template<> struct hash<EVertex> {
size_t operator()(EVertex const& vertex) const
{
return ((hash<glm::vec3>()(vertex.Pos) ^
(hash<glm::vec3>()(vertex.Color) << 1)) >> 1) ^
(hash<glm::vec2>()(vertex.TexCoord) << 1);
}
};
}
struct UniformBufferObject {
alignas(16) glm::mat4 ModelMatrix;
alignas(16) glm::mat4 ViewMatrix;
alignas(16) glm::mat4 ProjectiveMatrix;
};
class CTest
{
public:
void run()
{
__initWindow();
__initVulkan();
__mainLoop();
__cleanUp();
}
private:
GLFWwindow * m_pWindow;
VkInstance m_Instance;
VkPhysicalDevice m_PhysicalDevice = VK_NULL_HANDLE;
VkDevice m_Device;
VkQueue m_GraphicsQueue;
VkQueue m_PresentQueue;
VkDebugUtilsMessengerEXT m_CallBack;
VkSurfaceKHR m_WindowSurface;
//存储交换链的
VkSwapchainKHR m_SwapChain;
//获取交换链图像的图像句柄
//交换链图像由交换链自己负责创建，并在交换链清除时自动被清除，不需要我们自己进行创建和清除操作
std::vector<VkImage> m_SwapChainImages;
VkFormat m_SwapChainImageFormat;
VkExtent2D m_SwapChainExtent;
//存储图像视图的句柄集
std::vector<VkImageView> m_SwapChainImageViews;
VkRenderPass m_RenderPass;
std::vector<VkFramebuffer> m_SwapChainFrameBuffers;
//创建指令缓冲对象之前，需要先创建指令池对象
VkCommandPool m_CommandPool;
//创建指令缓冲对象,用来记录绘制命令，由于绘制操作是在帧缓冲上进行的，需要为交换链中每一个图像分配一个指令缓冲对象
//VkCommandBuffer m_CommandBuffer;
//指令缓冲对象会在指令池对象被清除时自动被清除，不需要我们自己显式地清除它
std::vector<VkCommandBuffer> m_CommandBuffers;
VkPipelineLayout m_PipelineLayout;
VkPipeline m_GraphicsPipeline;
//所有的描述符绑定都会被组合在一个单独的VkDescriptorSetLayout对象
VkDescriptorSetLayout m_DescriptorSetLayout;
//信号量
VkSemaphore ImageAvailableSemaphore; //准备渲染的信号量
VkSemaphore RenderFinishedSemaphore; //渲染结束信号量，准备进行呈现present
std::vector<VkSemaphore> m_ImageAvailableSemaphores;
std::vector<VkSemaphore> m_RenderFinishedSemaphores;
std::vector<VkFence> m_InFlightFences;
std::vector<VkFence> m_ImagesInFlight;
size_t m_CurrentFrame = 0;
//许多驱动程序会在窗口大小改变后触发VK_ERROR_OUT_OF_DATE_KHR信息，
//但这种触发并不可靠，所以最好添加一些代码来显式地在窗口大小改变时重建交换链。
//添加一个新的变量来标记窗口大小是否发生改变：
bool m_WindowResized = false;
//顶点缓冲
VkBuffer m_VertexBuffer;
//存储内存句柄
VkDeviceMemory m_VertexBufferMemory;
//索引缓冲
VkBuffer m_IndexBuffer;
VkDeviceMemory m_IndexBufferMemory;
//Uniform 缓冲区\
//在每一帧中我们需要拷贝新的数据到UBO缓冲区，所以存在一个暂存缓冲区是没有意义的。
//在这种情况下，它只会增加额外的开销，并且可能降低性能而不是提升性能。因此本次不使用临时缓冲区
std::vector<VkBuffer> m_UniformBuffers;
std::vector<VkDeviceMemory> m_UniformBuffersMemory;
//创建的描述符池对象
VkDescriptorPool m_DescriptorPool;
std::vector<VkDescriptorSet> m_DescriptorSets;
//图像中像素被称为纹理元素
VkImage m_TextureImage;
VkDeviceMemory m_TextureImageMemory;
//纹理图像
VkImageView m_TextureImageView;
//采样器对象
VkSampler m_TextureSampler;
//深度图像，深度附件是基于图像的，就像颜色附件，所不同的是交换链不会自动创建深度图像
//深度图像需要三种资源：图像， 内存， 图像视图
VkImage m_DepthImage;
VkDeviceMemory m_DepthImageMemory;
VkImageView m_DepthImageView;
std::vector<EVertex> Vertices;
std::vector<uint32_t> Indices;
//创建窗口
void __initWindow()
{
glfwInit();
//由于GLFW库最初是为了OpenGL设计，需要显式设置GLFW阻止它自动创建OpenGL上下文
glfwWindowHint(GLFW_CLIENT_API, GLFW_NO_API);
m_pWindow = glfwCreateWindow(WIDTH, HEIGHT, "VulKanTest", nullptr, nullptr);
//GLFW允许我们使用glfwSetWindowUserPointer将任意指针存储在窗体对象中，
//因此可以指定静态类成员调用glfwGetWindowUserPointer返回原始的实例对象
glfwSetWindowUserPointer(m_pWindow, this);
//会在窗体发生大小变化的时候被事件回调
glfwSetFramebufferSizeCallback(m_pWindow, CTest::__framebufferResizeCallback);
}
//回调函数,必须为静态函数，才能作为回调函数
static void __framebufferResizeCallback(GLFWwindow* pWindow, int vWidth, int vHeight) {
if (vWidth == 0 || vHeight == 0) return;
auto app = reinterpret_cast<CTest*>(glfwGetWindowUserPointer(pWindow));
app->m_WindowResized = true;
}
//创建vulkan实例
void __initVulkan()
{
__createInstance();
//__setupDebugCallBack();
__createSurface();
//创建实例之后，需要在系统中找到一个支持功能的显卡，查找第一个图像卡作为适合物理设备
__pickPhysicalDevice();
__createLogicalDevice();
__createSwapShain();
__createImageViews();
__createRenderPass();
//要在管线创建时，为着色器提供关于每个描述符绑定的详细信息,考虑到会在管线中使用，因此需要在管线创建之前调用
__createDescritorSetLayout();
__createGraphicsPipelines();
__createCommandPool();
__createDepthResources();
//确保在深度图像视图实际创建后调用帧缓冲
__createFrameBuffers();
__createTextureImage();
//有了纹理，需要纹理视图才能访问纹理
__createTextureImageView();
//配置采样器对象。稍后会使用它从着色器中读取颜色。样器没有任何地方引用VkImage。采样器是一个独特的对象，它提供了从纹理中提取颜色的接口。它可以应用在任何你期望的图像中，无论是1D，2D，或者是3D
__createTextureSampler();
//填充顶点和索引数据
__loadModel();
__createVertexBuffer();
__createIndexBuffer();
__createUniformBuffer();
//创建描述符池
__createDescritorPool();
__createDescriptorSet();
__createCommandBuffers();
__creatSemaphores(); //创造信号量
}
//重建交换链
void __recreateSwapChain()
{
int Width = 0, Height = 0;
//当窗口最小化时，窗口的帧缓冲实际大小也为0，设置应用程序在窗口最小化后停止渲染，直到窗口重新可见时重建交换链
if (Width == 0 || Height == 0)
{
glfwGetFramebufferSize(m_pWindow, &Width, &Height);
glfwWaitEvents();
}
vkDeviceWaitIdle(m_Device);
__createSwapShain();
__createImageViews(); //图像视图是建立在交换链图像基础上的
__createRenderPass(); //依赖于交换链图像的格式
__createGraphicsPipelines(); //创建图形管线期间指定Viewport和scissor 矩形大小,改变了图像的大小
__createDepthResources(); //如果渲染通道包含深度模版附件，则必须指定深度模版状态。
__createFrameBuffers(); //依赖于交换链图像
__createCommandBuffers(); //依赖于交换链图像
}
//循环渲染
void __mainLoop()
{
while (!glfwWindowShouldClose(m_pWindow))
{
glfwPollEvents();
//函数中所有的操作都是异步的。意味着当程序退出mainLoop，
//绘制和呈现操作可能仍然在执行。所以清理该部分的资源是不友好的
//为了解决这个问题，我们应该在退出mainLoop销毁窗体前等待逻辑设备的操作完成:
__drawFrame(); //异步操作
//更新UniformBuffer
}
vkDeviceWaitIdle(m_Device);
}
//为了确保重新创建相关的对象之前，老版本的对象被系统正确回收清理,需要重构Cleanup
void __cleanUpSwapChain()
{
vkDestroyImageView(m_Device, m_DepthImageView, nullptr);
vkDestroyImage(m_Device, m_DepthImage, nullptr);
vkFreeMemory(m_Device, m_DepthImageMemory, nullptr);
//删除帧缓冲
for (size_t i = 0; i < m_SwapChainFrameBuffers.size(); i++)
{
vkDestroyFramebuffer(m_Device, m_SwapChainFrameBuffers[i], nullptr);
}
//选择借助vkFreeCommandBuffers函数清理已经存在的命令缓冲区。这种方式可以重用对象池中已经分配的命令缓冲区。
vkFreeCommandBuffers(m_Device, m_CommandPool, static_cast<uint32_t>(m_CommandBuffers.size()), m_CommandBuffers.data());
vkDestroyPipeline(m_Device, m_GraphicsPipeline, nullptr);
vkDestroyPipelineLayout(m_Device, m_PipelineLayout, nullptr);
vkDestroyRenderPass(m_Device, m_RenderPass, nullptr);
//循环删除视图
for (size_t i = 0; i < m_SwapChainImageViews.size(); i++)
{
vkDestroyImageView(m_Device, m_SwapChainImageViews[i], nullptr);
}
vkDestroySwapchainKHR(m_Device, m_SwapChain, nullptr);
}
//退出
void __cleanUp()
{
__cleanUpSwapChain();
for (size_t i = 0; i < MAX_FRAMES_IN_FLIGHT; i++) {
vkDestroySemaphore(m_Device, m_RenderFinishedSemaphores[i], nullptr);
vkDestroySemaphore(m_Device, m_ImageAvailableSemaphores[i], nullptr);
vkDestroyFence(m_Device, m_InFlightFences[i], nullptr);
}
//删除指令池对象
vkDestroyCommandPool(m_Device, m_CommandPool, nullptr);
//缓冲区在程序退出之前为渲染指令提供支持，并不依赖交换链，因此在cleauo
vkDestroyBuffer(m_Device, m_VertexBuffer, nullptr);
vkFreeMemory(m_Device, m_VertexBufferMemory, nullptr);
vkDestroyBuffer(m_Device, m_IndexBuffer, nullptr);
vkFreeMemory(m_Device, m_IndexBufferMemory, nullptr);
vkDestroyDescriptorSetLayout(m_Device, m_DescriptorSetLayout, nullptr);
vkDestroyDescriptorPool(m_Device, m_DescriptorPool, nullptr);
vkDestroySampler(m_Device, m_TextureSampler, nullptr);
vkDestroyImageView(m_Device, m_TextureImageView, nullptr);
//UBO的数据将被用于所有的绘制使用，所以包含它的缓冲区只能在最后销毁：
for (size_t i = 0; i < m_SwapChainImages.size(); i++)
{
vkDestroyBuffer(m_Device, m_UniformBuffers[i], nullptr);
vkFreeMemory(m_Device, m_UniformBuffersMemory[i], nullptr);
vkDestroyFence(m_Device, m_ImagesInFlight[i], nullptr);
}
vkDestroyImage(m_Device, m_TextureImage, nullptr);
vkFreeMemory(m_Device, m_TextureImageMemory, nullptr);
vkDestroyDevice(m_Device, nullptr);
//使用Vulkan API创建的对象也需要被清除，且应该在Vulkan实例清除之前被清除。
if (enableValidationLayers)
{
destroyDebugUtilsMessengerEXT(m_Instance, m_CallBack, nullptr);
}
//注意删除顺序，先删除surface 后删除Instance
vkDestroySurfaceKHR(m_Instance, m_WindowSurface, nullptr);
vkDestroyInstance(m_Instance, nullptr);
glfwDestroyWindow(m_pWindow);
glfwTerminate();
}
//创建实例对象
void __createInstance()
{
if (enableValidationLayers && !__checkValidationLayerSupport())
{
throw std::runtime_error("validation layers requested, but not available!");
}
//1.1应用程序信息
VkApplicationInfo AppInfo = { };
AppInfo.sType = VK_STRUCTURE_TYPE_APPLICATION_INFO;
AppInfo.pApplicationName = "Hello World!";
AppInfo.applicationVersion = VK_MAKE_VERSION(1, 0, 0);
AppInfo.pEngineName = "No Engine";
AppInfo.engineVersion = VK_MAKE_VERSION(1, 0, 0);
AppInfo.apiVersion = VK_API_VERSION_1_0;
//1.2Vulkan实例信息：设置Vulkan的驱动程序需要使用的全局扩展和校验层
VkInstanceCreateInfo CreateInfo = { };
CreateInfo.sType = VK_STRUCTURE_TYPE_INSTANCE_CREATE_INFO;
CreateInfo.pApplicationInfo = &AppInfo;
//需要一个扩展才能与不同平台的窗体系统进行交互
//GLFW有一个方便的内置函数，返回它有关的扩展信息
//传递给struct:
auto Extensions = __getRequiredExtensions();
CreateInfo.enabledExtensionCount = static_cast<uint32_t>(Extensions.size());;
CreateInfo.ppEnabledExtensionNames = Extensions.data();
//指定全局校验层
//CreateInfo.enabledExtensionCount = 0;
//DebugInfo
VkDebugUtilsMessengerCreateInfoEXT CreateDebugInfo;
//修改VkInstanceCreateInfo结构体，填充当前上下文已经开启的validation layers名称集合。
if (enableValidationLayers)
{
CreateInfo.enabledLayerCount = static_cast<uint32_t>(ValidationLayers.size());
CreateInfo.ppEnabledLayerNames = ValidationLayers.data();
__populateDebugMessengerCreateInfo(CreateDebugInfo);
CreateInfo.pNext = (VkDebugUtilsMessengerCreateInfoEXT*)&CreateDebugInfo;
}
else {
CreateInfo.enabledLayerCount = 0;
CreateInfo.pNext = nullptr;
}
//1.3创建Instance
VkResult InstanceResult = vkCreateInstance(&CreateInfo, nullptr, &m_Instance);
//检查是否创建成功
if (InstanceResult != VK_SUCCESS) throw std::runtime_error("faild to create Instance!");
}
//选择支持我们需要的特性的设备
void __pickPhysicalDevice()
{
//1.使用VkPhysicalDevice对象储存显卡信息，这个对象可以在Instance对象清除时，自动清除自己，所有不用人工清除
//2.获取图像卡列表
uint32_t DeviceCount = 0;
vkEnumeratePhysicalDevices(m_Instance, &DeviceCount, nullptr);
if (DeviceCount == 0) throw std::runtime_error("failed to find GPUS with Vulkan support!");
//3.获取完设备数量后，就可以分配数组来存储VkPhysicalDevice对象
std::vector<VkPhysicalDevice> Devices(DeviceCount);
vkEnumeratePhysicalDevices(m_Instance, &DeviceCount, Devices.data());
//4.检查它们是否适合我们要执行的操作
for (const auto& device : Devices)
{
if (__isDeviceSuitable(device))
{
m_PhysicalDevice = device;
break;
}
}
if (m_PhysicalDevice == VK_NULL_HANDLE)
{
throw std::runtime_error("failed to find a suitable GPU!");
}
}
//创建逻辑设备：
//FUNC:主要与物理设备进行交互
void __createLogicalDevice()
{
//创建队列
EQueueFamilyIndices Indices = __findQueueFamilies(m_PhysicalDevice);
//描述队列蔟中预定申请的队列个数,创建支持表现和显示功能
std::vector<VkDeviceQueueCreateInfo> QueueCreateInfos;
//需要多个VkDeviceQueueCreateInfo结构来创建不同功能的队列。一个方式是针对不同功能的队列簇创建一个set集合确保队列簇的唯一性
std::set<int> UniqueQueueFamilies = { Indices.GraphicsFamily, Indices.PresentFamily };
//优先级
float QueuePriority = 1.0f;
for (int queueFamily : UniqueQueueFamilies) {
VkDeviceQueueCreateInfo queueCreateInfo = {};
queueCreateInfo.sType = VK_STRUCTURE_TYPE_DEVICE_QUEUE_CREATE_INFO;
queueCreateInfo.queueFamilyIndex = queueFamily;
queueCreateInfo.queueCount = 1;
queueCreateInfo.pQueuePriorities = &QueuePriority;
QueueCreateInfos.push_back(queueCreateInfo);
}
////指定使用的设备特性
VkPhysicalDeviceFeatures PhysicalDeviceFeatures = {};
PhysicalDeviceFeatures.samplerAnisotropy = VK_TRUE;
////创建逻辑设备
VkDeviceCreateInfo DeviceCreateInfo = {};
DeviceCreateInfo.sType = VK_STRUCTURE_TYPE_DEVICE_CREATE_INFO;
////添加指向队列创建信息的结构体和设备功能结构体:
DeviceCreateInfo.pQueueCreateInfos = QueueCreateInfos.data();
DeviceCreateInfo.queueCreateInfoCount = static_cast<uint32_t>(QueueCreateInfos.size());
DeviceCreateInfo.pEnabledFeatures = &PhysicalDeviceFeatures;
DeviceCreateInfo.enabledExtensionCount = static_cast<uint32_t>(DeviceExtensions.size());
DeviceCreateInfo.ppEnabledExtensionNames = DeviceExtensions.data();
////为设备开启validation layers
if (enableValidationLayers) {
DeviceCreateInfo.enabledLayerCount = static_cast<uint32_t>(ValidationLayers.size());
DeviceCreateInfo.ppEnabledLayerNames = ValidationLayers.data();
}
else {
DeviceCreateInfo.enabledLayerCount = 0;
}
if (vkCreateDevice(m_PhysicalDevice, &DeviceCreateInfo, nullptr, &m_Device) != VK_SUCCESS) {
throw std::runtime_error("failed to create logical device!");
}
//获取指定队列族的队列句柄
vkGetDeviceQueue(m_Device, Indices.GraphicsFamily, 0, &m_GraphicsQueue);
vkGetDeviceQueue(m_Device, Indices.PresentFamily, 0, &m_PresentQueue);
}
//判断是否支持gPU函数
bool __isDeviceSuitable(VkPhysicalDevice vDevice)
{
EQueueFamilyIndices Indices = __findQueueFamilies(vDevice);
//是否支持交换链支持
bool DeviceExtensionSupport = __checkDeviceExtensionSupport(vDevice);
//各向异性滤波器
VkPhysicalDeviceFeatures SupportedFeatures;
vkGetPhysicalDeviceFeatures(vDevice, &SupportedFeatures);
bool SwapChainAdequate = false;
if (DeviceExtensionSupport)
{
ESwapChainSupportDetails SwapChainSupport = __querySwapChainSupport(vDevice);
//返回真，说明交换链的能力满足需要
SwapChainAdequate = !SwapChainSupport.SurfaceFormats.empty() && !SwapChainSupport.PresentModes.empty();
}
return (Indices.isComplete() && DeviceExtensionSupport && SwapChainAdequate && SupportedFeatures.samplerAnisotropy);
}
//检测设备中支持的队列蔟
EQueueFamilyIndices __findQueueFamilies(VkPhysicalDevice vDevice)
{
EQueueFamilyIndices Indices;
//设备队列族的个数
uint32_t QueueFamilyCount = 0;
vkGetPhysicalDeviceQueueFamilyProperties(vDevice, &QueueFamilyCount, nullptr);
std::vector<VkQueueFamilyProperties> QueueFamilies(QueueFamilyCount);
vkGetPhysicalDeviceQueueFamilyProperties(vDevice, &QueueFamilyCount, QueueFamilies.data());
//找到一个支持VK_QUEUE_GRAPHICS_BIT的队列簇
int i = 0;
for (const auto& QueueFamily : QueueFamilies)
{
//查找带有表现图像到窗口表面能力的队列族
VkBool32 PresentSupport = false;
vkGetPhysicalDeviceSurfaceSupportKHR(vDevice, i, m_WindowSurface, &PresentSupport);
//根据队列簇中的队列数量和是否支持表现能力来确定使用的表现队列簇的索引
if (QueueFamily.queueCount > 0 && PresentSupport)
{
Indices.PresentFamily = i;
}
if (QueueFamily.queueCount > 0 && QueueFamily.queueFlags && VK_QUEUE_GRAPHICS_BIT)
{
Indices.GraphicsFamily = i;
}
if (Indices.isComplete()) break;
i++;
}
return Indices;
}
//检查所有请求的校验层layers是否可用
bool __checkValidationLayerSupport()
{
uint32_t LayerCount;
vkEnumerateInstanceLayerProperties(&LayerCount, nullptr);
std::vector<VkLayerProperties> AvailableLayers(LayerCount);
vkEnumerateInstanceLayerProperties(&LayerCount, AvailableLayers.data());
//检查是否所有validationLayers列表中的校验层都可以在availableLayers列表中找到：
for (const char* layName : ValidationLayers)
{
bool LayerFound = false;
for (const auto& layerProperties : AvailableLayers)
{
if (strcmp(layName, layerProperties.layerName) == 0)
{
LayerFound = true;
break;
}
}
if (!LayerFound)
{
return false;
}
}
return true;
}
//仅仅启用校验层并没有任何用处，我们不能得到任何有用的调试信息。
//为了获得调试信息，我们需要使用VK_EXT_debug_utils扩展，设置回调函数来接受调试信息。
std::vector<const char*> __getRequiredExtensions()
{
unsigned int GlfwExtensionCount = 0;
const char** GlfwExtensions;
GlfwExtensions = glfwGetRequiredInstanceExtensions(&GlfwExtensionCount);
std::vector<const char*> Extensions(GlfwExtensions, GlfwExtensions + GlfwExtensionCount);
for (unsigned int i = 0; i < GlfwExtensionCount; i++)
{
Extensions.push_back(GlfwExtensions[i]);
}
if (enableValidationLayers)
{
Extensions.push_back(VK_EXT_DEBUG_REPORT_EXTENSION_NAME);
}
return Extensions;
}
//回调函数：返回Bool值，表示引发校验层处理的Vulkan API调用是否中断
//返回True:对应Vulkan API调用就会返回VK_ERROR_VALIDATION_FAILED_EXT错误代码。通常，只在测试校验层本身时会返回true
//其余情况下，回调函数应该返回VK_FALSE
static VKAPI_ATTR VkBool32 VKAPI_CALL __debugCallBack(
VkDebugUtilsMessageSeverityFlagBitsEXT vMessageSeverity, //指定消息的级别
VkDebugUtilsMessageTypeFlagsEXT vMessageType, //错误行为
const VkDebugUtilsMessengerCallbackDataEXT* pCallbackData, //指向VkDebugUtilsMessengerCallbackDataEXT结构体指针，结构成员有：pMessage, pObject, objectCount
void* pUserData) //指向回调函数时，传递数据的指针
{
std::cerr << "validation layer:" << pCallbackData->pMessage << std::endl;
return VK_FALSE;
}
void __populateDebugMessengerCreateInfo(VkDebugUtilsMessengerCreateInfoEXT& VCreateInfo) {
VCreateInfo = {};
VCreateInfo.sType = VK_STRUCTURE_TYPE_DEBUG_UTILS_MESSENGER_CREATE_INFO_EXT;
VCreateInfo.messageSeverity = VK_DEBUG_UTILS_MESSAGE_SEVERITY_VERBOSE_BIT_EXT | VK_DEBUG_UTILS_MESSAGE_SEVERITY_WARNING_BIT_EXT | VK_DEBUG_UTILS_MESSAGE_SEVERITY_ERROR_BIT_EXT;
VCreateInfo.messageType = VK_DEBUG_UTILS_MESSAGE_TYPE_GENERAL_BIT_EXT | VK_DEBUG_UTILS_MESSAGE_TYPE_VALIDATION_BIT_EXT | VK_DEBUG_UTILS_MESSAGE_TYPE_PERFORMANCE_BIT_EXT;
VCreateInfo.pfnUserCallback = __debugCallBack;
}
//调用回调函数
void __setupDebugCallBack()
{
if (!enableValidationLayers) return;
VkDebugUtilsMessengerCreateInfoEXT createInfo;
__populateDebugMessengerCreateInfo(createInfo);
if (createDebugUtilsMessengerEXT(m_Instance, &createInfo, nullptr, &m_CallBack) != VK_SUCCESS) {
throw std::runtime_error("failed to set up debug messenger!");
}
}
//Surface
void __createSurface()
{
if (glfwCreateWindowSurface(m_Instance, m_pWindow, nullptr, &m_WindowSurface) != VK_SUCCESS) {
throw std::runtime_error("failed to create window surface!");
}
}
//检测交换链是否支持：使用交换链必须保证VK_KHR_swapchain设备启用
bool __checkDeviceExtensionSupport(VkPhysicalDevice vPhysicalDevice)
{
uint32_t ExtensionCount;
vkEnumerateDeviceExtensionProperties(vPhysicalDevice, nullptr, &ExtensionCount, nullptr);
//思想：将所需的扩展保存在一个集合，权举所有可用的扩展，将集合中扩展删除，若集合为空，则所需扩展均能得到满足
std::vector<VkExtensionProperties> AvailableExtensions(ExtensionCount);
vkEnumerateDeviceExtensionProperties(vPhysicalDevice, nullptr, &ExtensionCount, AvailableExtensions.data());
std::set<std::string> RequireExtensions(DeviceExtensions.begin(), DeviceExtensions.end());
for (const auto& entension : AvailableExtensions)
{
RequireExtensions.erase(entension.extensionName);
}
return RequireExtensions.empty();
}
//检查交换链细节
ESwapChainSupportDetails __querySwapChainSupport(VkPhysicalDevice vPhsicalDevice)
{
ESwapChainSupportDetails Details;
//查询基础表面特性
vkGetPhysicalDeviceSurfaceCapabilitiesKHR(vPhsicalDevice, m_WindowSurface, &Details.Capablities);
//查询表面支持的格式
uint32_t FormatCount;
vkGetPhysicalDeviceSurfaceFormatsKHR(vPhsicalDevice, m_WindowSurface, &FormatCount, nullptr);
if (FormatCount != 0)
{
Details.SurfaceFormats.resize(FormatCount);
vkGetPhysicalDeviceSurfaceFormatsKHR(vPhsicalDevice, m_WindowSurface, &FormatCount, Details.SurfaceFormats.data());
}
//查询支持的呈现模式
uint32_t PresentModeCount;
vkGetPhysicalDeviceSurfacePresentModesKHR(vPhsicalDevice, m_WindowSurface, &PresentModeCount, nullptr);
if (PresentModeCount != 0)
{
Details.PresentModes.resize(PresentModeCount);
vkGetPhysicalDeviceSurfacePresentModesKHR(vPhsicalDevice, m_WindowSurface, &FormatCount, Details.PresentModes.data());
}
return Details;
}
//为交换链选择合适的设置：1. 表面格式 2. 呈现模式 3.交换范围
//表面格式
VkSurfaceFormatKHR __chooseSwapSurfaceFormat(const std::vector<VkSurfaceFormatKHR> &vAvailablFormats)
{
//若是没有自己首选的模式，则直接返回设定的模式
if (vAvailablFormats.size() == 1 && vAvailablFormats[0].format == VK_FORMAT_UNDEFINED)
{
return { VK_FORMAT_B8G8R8A8_UNORM, VK_COLOR_SPACE_SRGB_NONLINEAR_KHR };
}
for (const auto& availableFormat : vAvailablFormats)
{
if (availableFormat.format == VK_FORMAT_B8G8R8A8_UNORM && availableFormat.colorSpace == VK_COLOR_SPACE_SRGB_NONLINEAR_KHR)
{
return availableFormat;
}
}
//大多数情况下，直接使用列表中的第一个格式也是非常不错的选择
return vAvailablFormats[0];
}
//呈现模式，选择最佳模式
VkPresentModeKHR __chooseSwapPresentMode(const std::vector<VkPresentModeKHR> vAvailablePresentModes)
{
VkPresentModeKHR BestMode = VK_PRESENT_MODE_FIFO_KHR;
//使用三缓冲
for (const auto& availablePresentMode : vAvailablePresentModes) {
if (availablePresentMode == VK_PRESENT_MODE_MAILBOX_KHR)
{
return availablePresentMode;
}
else if (availablePresentMode == VK_PRESENT_MODE_IMMEDIATE_KHR)
{
BestMode = availablePresentMode;
}
}
return BestMode;
}
//交换范围：具体表示图像的分辨率
VkExtent2D __chooseSwapExtent(const VkSurfaceCapabilitiesKHR &vSurfaceCapabilities)
{
int Width, Height;
glfwGetWindowSize(m_pWindow, &Width, &Height);
if (vSurfaceCapabilities.currentExtent.width != std::numeric_limits<uint32_t>::max())
{
return vSurfaceCapabilities.currentExtent;
}
else {
VkExtent2D ActualExtent = { Width, Height };
ActualExtent.width = std::max(vSurfaceCapabilities.minImageExtent.width, std::min(vSurfaceCapabilities.maxImageExtent.width, ActualExtent.width));
ActualExtent.height = std::max(vSurfaceCapabilities.minImageExtent.height, std::min(vSurfaceCapabilities.maxImageExtent.height, ActualExtent.height));
return ActualExtent;
}
}
//创造交换链
void __createSwapShain()
{
//查询可以支持交换链的
ESwapChainSupportDetails SwapChainSupport = __querySwapChainSupport(m_PhysicalDevice);
VkSurfaceFormatKHR SurfaceFormat = __chooseSwapSurfaceFormat(SwapChainSupport.SurfaceFormats);
VkPresentModeKHR PresentMode = __chooseSwapPresentMode(SwapChainSupport.PresentModes);
VkExtent2D Extent = __chooseSwapExtent(SwapChainSupport.Capablities);
//设置交换链中的图像个数，三倍缓冲
uint32_t ImageCount = SwapChainSupport.Capablities.minImageCount + 1;
if (SwapChainSupport.Capablities.maxImageCount > 0 && ImageCount > SwapChainSupport.Capablities.maxImageCount)
{
ImageCount = SwapChainSupport.Capablities.maxImageCount;
}
//创建交换链图像的信息
VkSwapchainCreateInfoKHR SwapChainCreateInfo = {};
SwapChainCreateInfo.sType = VK_STRUCTURE_TYPE_SWAPCHAIN_CREATE_INFO_KHR;
SwapChainCreateInfo.surface = m_WindowSurface;
SwapChainCreateInfo.minImageCount = ImageCount;
SwapChainCreateInfo.imageFormat = SurfaceFormat.format;
SwapChainCreateInfo.imageColorSpace = SurfaceFormat.colorSpace;
SwapChainCreateInfo.imageExtent = Extent;
SwapChainCreateInfo.imageArrayLayers = 1; //指定每个图像所包含的层次
SwapChainCreateInfo.imageUsage = VK_IMAGE_USAGE_COLOR_ATTACHMENT_BIT; //指定在图像上进行怎样的操作
//指定多个队列簇使用交换链图像的方式
//这里通过图形队列在交换链图像上进行绘制操作，然后将图像提交给呈现队列进行显示
EQueueFamilyIndices Indices = __findQueueFamilies(m_PhysicalDevice);
uint32_t QueueFamilyIndices[] = { (uint32_t)Indices.GraphicsFamily, (uint32_t)Indices.PresentFamily };
//图形和呈现不是同一个队列族，使用协同模式来避免处理图像所有权问题；否则就不能使用
if (Indices.GraphicsFamily != Indices.PresentFamily)
{
SwapChainCreateInfo.imageSharingMode = VK_SHARING_MODE_CONCURRENT;
SwapChainCreateInfo.queueFamilyIndexCount = 2;
SwapChainCreateInfo.pQueueFamilyIndices = QueueFamilyIndices;
}
else {
SwapChainCreateInfo.imageSharingMode = VK_SHARING_MODE_EXCLUSIVE;
SwapChainCreateInfo.queueFamilyIndexCount = 0;
SwapChainCreateInfo.pQueueFamilyIndices = nullptr;
}
//可以为交换链中的图像指定一个固定的变换操作
SwapChainCreateInfo.compositeAlpha = VK_COMPOSITE_ALPHA_OPAQUE_BIT_KHR; //指定alpha通道是否被用来和窗口系统中的其它窗口进行混合操作
SwapChainCreateInfo.preTransform = SwapChainSupport.Capablities.currentTransform;
//设置呈现模式
SwapChainCreateInfo.presentMode = PresentMode;
SwapChainCreateInfo.clipped = VK_TRUE; //clipped成员变量被设置为VK_TRUE表示我们不关心被窗口系统中的其它窗口遮挡的像素的颜色
SwapChainCreateInfo.oldSwapchain = VK_NULL_HANDLE; //需要指定它，因为应用程序在运行过程中可能会失效，由于之前还没有创建任何一个交换链，将设置为VK_NULL_HANDLE即可。
//创建交换链
if (vkCreateSwapchainKHR(m_Device, &SwapChainCreateInfo, nullptr, &m_SwapChain) != VK_SUCCESS)
{
throw std::runtime_error("failed to create SwapChain!");
}
//获取交换链图像句柄,先获取数量，然后分配空间，获取句柄
vkGetSwapchainImagesKHR(m_Device, m_SwapChain, &ImageCount, nullptr);
m_SwapChainImages.resize(ImageCount);
vkGetSwapchainImagesKHR(m_Device, m_SwapChain, &ImageCount, m_SwapChainImages.data());
m_SwapChainImageFormat = SurfaceFormat.format;
m_SwapChainExtent = Extent;
}
//将__createImageViews和_createTextureImagView抽象
VkImageView __createImageView(VkImage vImage, VkFormat vFormat, VkImageAspectFlags vAspectFlags)
{
VkImageViewCreateInfo ImageViewCreateInfo = {};
ImageViewCreateInfo.sType = VK_STRUCTURE_TYPE_IMAGE_VIEW_CREATE_INFO;
ImageViewCreateInfo.image = vImage;
ImageViewCreateInfo.viewType = VK_IMAGE_VIEW_TYPE_2D;
ImageViewCreateInfo.format = vFormat;
//components字段调整颜色通道的最终映射逻辑
ImageViewCreateInfo.components.r = VK_COMPONENT_SWIZZLE_IDENTITY;
ImageViewCreateInfo.components.g = VK_COMPONENT_SWIZZLE_IDENTITY;
ImageViewCreateInfo.components.b = VK_COMPONENT_SWIZZLE_IDENTITY;
ImageViewCreateInfo.components.a = VK_COMPONENT_SWIZZLE_IDENTITY;
//subresourceRangle字段用于描述图像的使用目标是什么，以及可以被访问的有效区域
ImageViewCreateInfo.subresourceRange.aspectMask = vAspectFlags;
ImageViewCreateInfo.subresourceRange.baseMipLevel = 0;
ImageViewCreateInfo.subresourceRange.levelCount = 1;
ImageViewCreateInfo.subresourceRange.baseArrayLayer = 0;
ImageViewCreateInfo.subresourceRange.layerCount = 1;
VkImageView ImageView;
if (vkCreateImageView(m_Device, &ImageViewCreateInfo, nullptr, &ImageView) != VK_SUCCESS)
{
throw std::runtime_error("failed to create ImageView!");
}
return ImageView;
}
//为每一个交换链中图像创建基本的图像
void __createImageViews()
{
//保存图像集大小
m_SwapChainImageViews.resize(m_SwapChainImages.size());
//循环迭代交换链图像
for (size_t i = 0; i < m_SwapChainImages.size(); i++)
{
m_SwapChainImageViews[i] = __createImageView(m_SwapChainImages[i], m_SwapChainImageFormat, VK_IMAGE_ASPECT_COLOR_BIT);
}
}
//读文件
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!");
}
size_t fileSize = (size_t)file.tellg();
std::vector<char> buffer(fileSize);
file.seekg(0);
file.read(buffer.data(), fileSize);
file.close();
return buffer;
}
//将shader代码传递给渲染管线之前，必须将其封装到VkShaderModule对象中
VkShaderModule __createShaderModule(const std::vector<char>& vShaderCode)
{
VkShaderModuleCreateInfo ShaderModuleCreateInfo = {};
ShaderModuleCreateInfo.sType = VK_STRUCTURE_TYPE_SHADER_MODULE_CREATE_INFO;
ShaderModuleCreateInfo.codeSize = vShaderCode.size();
//Shadercode字节码是以字节为指定，但字节码指针是uint32_t类型，因此需要转换类型
ShaderModuleCreateInfo.pCode = reinterpret_cast<const uint32_t*>(vShaderCode.data());
//创建vkShaderModule
VkShaderModule ShaderModule;
if (vkCreateShaderModule(m_Device, &ShaderModuleCreateInfo, nullptr, &ShaderModule) != VK_SUCCESS)
{
throw std::runtime_error("failed to create ShaderModule!");
}
return ShaderModule;
}
//创建渲染通道
void __createRenderPass()
{
//附件描述
VkAttachmentDescription ColorAttachment = {}; //这里只用一个颜色缓冲附件
ColorAttachment.format = m_SwapChainImageFormat; //指定颜色附件格式，这里与交换链中保持一致
ColorAttachment.samples = VK_SAMPLE_COUNT_1_BIT; //这里不做多重采样
//渲染前和渲染后数据在对应附件的操作行为，应用颜色和深度数据
ColorAttachment.loadOp = VK_ATTACHMENT_LOAD_OP_CLEAR; //起始阶段以一个常量清理附件内容
ColorAttachment.storeOp = VK_ATTACHMENT_STORE_OP_STORE; //渲染的内容会存储在内存，并在之后进行读取操作
//用于模板数
ColorAttachment.stencilLoadOp = VK_ATTACHMENT_LOAD_OP_DONT_CARE;
ColorAttachment.stencilStoreOp = VK_ATTACHMENT_STORE_OP_DONT_CARE; //图像在交换链中被呈现
ColorAttachment.initialLayout = VK_IMAGE_LAYOUT_UNDEFINED;
ColorAttachment.finalLayout = VK_IMAGE_LAYOUT_PRESENT_SRC_KHR;
//子通道：一个单独的渲染通道可以由多个子通道组成。子通道是渲染操作的一个序列
VkAttachmentReference ColorAttachmentRef = {};
ColorAttachmentRef.attachment = 0; //通过附件描述符集合中的索引来持有,这里只有一个颜色附件
ColorAttachmentRef.layout = VK_IMAGE_LAYOUT_COLOR_ATTACHMENT_OPTIMAL;
//深度附件
VkAttachmentDescription DepthAttachment = {};
DepthAttachment.format = __findDepthFormat();
DepthAttachment.samples = VK_SAMPLE_COUNT_1_BIT;
DepthAttachment.loadOp = VK_ATTACHMENT_LOAD_OP_CLEAR;
DepthAttachment.storeOp = VK_ATTACHMENT_STORE_OP_DONT_CARE;
DepthAttachment.stencilLoadOp = VK_ATTACHMENT_LOAD_OP_DONT_CARE;
DepthAttachment.stencilStoreOp = VK_ATTACHMENT_STORE_OP_DONT_CARE;
DepthAttachment.initialLayout = VK_IMAGE_LAYOUT_UNDEFINED;
DepthAttachment.finalLayout = VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL;
VkAttachmentReference DepthAttachmentRef = {};
DepthAttachmentRef.attachment = 1;
DepthAttachmentRef.layout = VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL;
//子通道
VkSubpassDescription SubpassDescription = {};
SubpassDescription.pipelineBindPoint = VK_PIPELINE_BIND_POINT_GRAPHICS; //指定graphics subpass图形子通道
SubpassDescription.colorAttachmentCount = 1;
SubpassDescription.pColorAttachments = &ColorAttachmentRef; //指定颜色附件引用。附件在数组中的索引直接从片段着色器引用，其layout(location = 0) out vec4 outColor 指令!
SubpassDescription.pDepthStencilAttachment = &DepthAttachmentRef;
//子通道依赖：渲染通道中的子通道会自动处理布局的变换。这些变换通过子通道的依赖关系进行控制
VkSubpassDependency SubpassDependency = {};
SubpassDependency.srcSubpass = VK_SUBPASS_EXTERNAL; //指定要依赖关系
SubpassDependency.dstSubpass = 0; //从属子通道的索引
SubpassDependency.srcStageMask = VK_PIPELINE_STAGE_COLOR_ATTACHMENT_OUTPUT_BIT; //指定等待的操作，这里等待交换链完成对应图像的读取操作。这可以通过等待颜色附件输出的阶段来实现
SubpassDependency.srcAccessMask = 0;
//在颜色附件阶段的操作及涉及颜色附件的读取和写入的操作应该等待。这些设置将阻止转换发生
SubpassDependency.dstStageMask = VK_PIPELINE_STAGE_COLOR_ATTACHMENT_OUTPUT_BIT;
SubpassDependency.dstAccessMask = VK_ACCESS_COLOR_ATTACHMENT_READ_BIT | VK_ACCESS_COLOR_ATTACHMENT_WRITE_BIT;
//创建渲染管道
std::array<VkAttachmentDescription, 2> Attachments = { ColorAttachment, DepthAttachment };
VkRenderPassCreateInfo RenderPassCreateInfo = {};
RenderPassCreateInfo.sType = VK_STRUCTURE_TYPE_RENDER_PASS_CREATE_INFO;
RenderPassCreateInfo.attachmentCount = static_cast<uint32_t>(Attachments.size());;
RenderPassCreateInfo.pAttachments = Attachments.data();
RenderPassCreateInfo.subpassCount = 1;
RenderPassCreateInfo.pSubpasses = &SubpassDescription;
RenderPassCreateInfo.dependencyCount = 1;
RenderPassCreateInfo.pDependencies = &SubpassDependency;
if (vkCreateRenderPass(m_Device, &RenderPassCreateInfo, nullptr, &m_RenderPass) != VK_SUCCESS) {
throw std::runtime_error("failed to create render pass!");
}
}
//创建图形管线
void __createGraphicsPipelines()
{
auto VertShaderCode = __readFile("vert.spv");
auto FragShaderCode = __readFile("frag.spv");
VkShaderModule VertShaderModule = __createShaderModule(VertShaderCode);
VkShaderModule FragShaderModule = __createShaderModule(FragShaderCode);
//着色器创建：VkShaderModule对象只是字节码缓冲区的一个包装容器。着色器并没有彼此链接
//通过VkPipelineShaderStageCreateInfo结构将着色器模块分配到管线中的顶点或者片段着色器阶段
//顶点
VkPipelineShaderStageCreateInfo VertShaderStageInfo = {};
VertShaderStageInfo.sType = VK_STRUCTURE_TYPE_PIPELINE_SHADER_STAGE_CREATE_INFO;
VertShaderStageInfo.stage = VK_SHADER_STAGE_VERTEX_BIT;
VertShaderStageInfo.module = VertShaderModule;
VertShaderStageInfo.pName = "main";
//片段
VkPipelineShaderStageCreateInfo FragShaderStageInfo = {};
FragShaderStageInfo.sType = VK_STRUCTURE_TYPE_PIPELINE_SHADER_STAGE_CREATE_INFO;
FragShaderStageInfo.stage = VK_SHADER_STAGE_FRAGMENT_BIT;
FragShaderStageInfo.module = FragShaderModule;
FragShaderStageInfo.pName = "main";
VkPipelineShaderStageCreateInfo ShaderStages[] = { VertShaderStageInfo, FragShaderStageInfo };
//顶点信息
auto BindingDescription = EVertex::getBindingDescription();
auto AttributeDescription = EVertex::getAttributeDescriptions();
VkPipelineVertexInputStateCreateInfo PipelineVertexInputStateCreateInfo = {};
PipelineVertexInputStateCreateInfo.sType = VK_STRUCTURE_TYPE_PIPELINE_VERTEX_INPUT_STATE_CREATE_INFO;
PipelineVertexInputStateCreateInfo.vertexBindingDescriptionCount = 1;
PipelineVertexInputStateCreateInfo.pVertexBindingDescriptions = &BindingDescription;
PipelineVertexInputStateCreateInfo.vertexAttributeDescriptionCount = static_cast<uint32_t> (AttributeDescription.size());
PipelineVertexInputStateCreateInfo.pVertexAttributeDescriptions = AttributeDescription.data();
//顶点数据的集合图元拓扑结构和是否启用顶点索重新开始图元
VkPipelineInputAssemblyStateCreateInfo PipelineInputAssemblyStateCreateInfo = {};
PipelineInputAssemblyStateCreateInfo.sType = VK_STRUCTURE_TYPE_PIPELINE_INPUT_ASSEMBLY_STATE_CREATE_INFO;
PipelineInputAssemblyStateCreateInfo.topology = VK_PRIMITIVE_TOPOLOGY_TRIANGLE_LIST; //三点共面，顶点不共用
PipelineInputAssemblyStateCreateInfo.primitiveRestartEnable = VK_FALSE;
//视图
VkViewport Viewport = {};
Viewport.x = 0.0f;
Viewport.y = 0.0f;
Viewport.width = (float)m_SwapChainExtent.width;
Viewport.height = (float)m_SwapChainExtent.height;
Viewport.minDepth = 0.0f;
Viewport.maxDepth = 1.0f;
//定义裁剪矩形覆盖到整体图像
VkRect2D Scissor = {};
Scissor.offset = { 0, 0 };
Scissor.extent = m_SwapChainExtent;
//将viewport和Scissor联合使用
VkPipelineViewportStateCreateInfo PipelineViewportStateCreateInfo = {};
PipelineViewportStateCreateInfo.sType = VK_STRUCTURE_TYPE_PIPELINE_VIEWPORT_STATE_CREATE_INFO;;
PipelineViewportStateCreateInfo.viewportCount = 1;
PipelineViewportStateCreateInfo.pViewports = &Viewport;
PipelineViewportStateCreateInfo.scissorCount = 1;
PipelineViewportStateCreateInfo.pScissors = &Scissor;
//光栅化:通过顶点着色器以及具体的几何算法将顶点塑形，并将图形传递到片段着色器进行着色
//执行深度测试，面剪切，裁剪测试
VkPipelineRasterizationStateCreateInfo PipelineRasterizationStateCreateInfo = {};
PipelineRasterizationStateCreateInfo.sType = VK_STRUCTURE_TYPE_PIPELINE_RASTERIZATION_STATE_CREATE_INFO;;
PipelineRasterizationStateCreateInfo.depthClampEnable = VK_FALSE; //vK_TRUE:超过远近裁剪面的片元进行收敛，而不是舍弃
PipelineRasterizationStateCreateInfo.rasterizerDiscardEnable = VK_FALSE; //VK_TRUE：几何图元永远不会传递到光栅化
PipelineRasterizationStateCreateInfo.polygonMode = VK_POLYGON_MODE_FILL; //决定几何产生图片的内容，多边形填充
PipelineRasterizationStateCreateInfo.lineWidth = 1.0f; //根据片元的数量描述线的宽度。最大的线宽支持取决于硬件，任何大于1.0的线宽需要开启GPU的wideLines特性支持
PipelineRasterizationStateCreateInfo.cullMode = VK_CULL_MODE_BACK_BIT; //指定剪裁面的类型方式
PipelineRasterizationStateCreateInfo.frontFace = VK_FRONT_FACE_COUNTER_CLOCKWISE; //front-facing面的顶点的顺序，可以是顺时针也可以是逆时针
PipelineRasterizationStateCreateInfo.depthBiasEnable = VK_FALSE;
PipelineRasterizationStateCreateInfo.depthBiasConstantFactor = 0.0f;
PipelineRasterizationStateCreateInfo.depthBiasClamp = 0.0f;
PipelineRasterizationStateCreateInfo.depthBiasSlopeFactor = 0.0f;
//重采样
VkPipelineMultisampleStateCreateInfo PipelineMultisampleStateCreateInfo = {};
PipelineMultisampleStateCreateInfo.sType = VK_STRUCTURE_TYPE_PIPELINE_MULTISAMPLE_STATE_CREATE_INFO;
PipelineMultisampleStateCreateInfo.sampleShadingEnable = VK_FALSE;
PipelineMultisampleStateCreateInfo.rasterizationSamples = VK_SAMPLE_COUNT_1_BIT;
PipelineMultisampleStateCreateInfo.minSampleShading = 1.0f; // Optional
PipelineMultisampleStateCreateInfo.pSampleMask = nullptr; // Optional
PipelineMultisampleStateCreateInfo.alphaToCoverageEnable = VK_FALSE; // Optional
PipelineMultisampleStateCreateInfo.alphaToOneEnable = VK_FALSE; // Optional
//深度测试
VkPipelineDepthStencilStateCreateInfo DepthStencil = {};
DepthStencil.sType = VK_STRUCTURE_TYPE_PIPELINE_DEPTH_STENCIL_STATE_CREATE_INFO;
DepthStencil.depthTestEnable = VK_TRUE; //指定是否应该将新的深度缓冲区与深度缓冲区进行比较，以确认是否应该被丢弃
DepthStencil.depthWriteEnable = VK_TRUE; //指定通过深度测试的新的片段深度是否应该被实际写入深度缓冲区
DepthStencil.depthCompareOp = VK_COMPARE_OP_LESS; //指定执行保留或者丢弃片段的比较细节
DepthStencil.depthBoundsTestEnable = VK_FALSE;
DepthStencil.stencilTestEnable = VK_FALSE;
//混色
VkPipelineColorBlendAttachmentState PipelineColorBlendAttachmentState = {};
PipelineColorBlendAttachmentState.colorWriteMask = VK_COLOR_COMPONENT_R_BIT | VK_COLOR_COMPONENT_G_BIT | VK_COLOR_COMPONENT_B_BIT | VK_COLOR_COMPONENT_A_BIT;
PipelineColorBlendAttachmentState.blendEnable = VK_FALSE;
PipelineColorBlendAttachmentState.srcColorBlendFactor = VK_BLEND_FACTOR_ONE; // Optional
PipelineColorBlendAttachmentState.dstColorBlendFactor = VK_BLEND_FACTOR_ZERO; // Optional
PipelineColorBlendAttachmentState.colorBlendOp = VK_BLEND_OP_ADD; // Optional
PipelineColorBlendAttachmentState.srcAlphaBlendFactor = VK_BLEND_FACTOR_ONE; // Optional
PipelineColorBlendAttachmentState.dstAlphaBlendFactor = VK_BLEND_FACTOR_ZERO; // Optional
PipelineColorBlendAttachmentState.alphaBlendOp = VK_BLEND_OP_ADD; // Optional
//所有帧缓冲区的引用
VkPipelineColorBlendStateCreateInfo PipelineColorBlendStateCreateInfo = {};
PipelineColorBlendStateCreateInfo.sType = VK_STRUCTURE_TYPE_PIPELINE_COLOR_BLEND_STATE_CREATE_INFO;
PipelineColorBlendStateCreateInfo.logicOpEnable = VK_FALSE;
PipelineColorBlendStateCreateInfo.logicOp = VK_LOGIC_OP_COPY; // Optional
PipelineColorBlendStateCreateInfo.attachmentCount = 1;
PipelineColorBlendStateCreateInfo.pAttachments = &PipelineColorBlendAttachmentState;
PipelineColorBlendStateCreateInfo.blendConstants[0] = 0.0f; // Optional
PipelineColorBlendStateCreateInfo.blendConstants[1] = 0.0f; // Optional
PipelineColorBlendStateCreateInfo.blendConstants[2] = 0.0f; // Optional
PipelineColorBlendStateCreateInfo.blendConstants[3] = 0.0f; // Optional
//需要在创建管线的时候指定描述符集合的布局，用以告知Vulkan着色器将要使用的描述符。描述符布局在管线布局对象中指定
VkPipelineLayoutCreateInfo PipelineLayoutCreateInfo = {};
PipelineLayoutCreateInfo.sType = VK_STRUCTURE_TYPE_PIPELINE_LAYOUT_CREATE_INFO;
PipelineLayoutCreateInfo.setLayoutCount = 1; // Optional
PipelineLayoutCreateInfo.pSetLayouts = &m_DescriptorSetLayout; // Optional
if (vkCreatePipelineLayout(m_Device, &PipelineLayoutCreateInfo, nullptr, &m_PipelineLayout) != VK_SUCCESS) {
throw std::runtime_error("failed to create pipeline layout!");
}
//创建管线
VkGraphicsPipelineCreateInfo GraphicsPipelineCreateInfo = {};
GraphicsPipelineCreateInfo.sType = VK_STRUCTURE_TYPE_GRAPHICS_PIPELINE_CREATE_INFO;
GraphicsPipelineCreateInfo.stageCount = 2;
GraphicsPipelineCreateInfo.pStages = ShaderStages;
GraphicsPipelineCreateInfo.pVertexInputState = &PipelineVertexInputStateCreateInfo;
GraphicsPipelineCreateInfo.pInputAssemblyState = &PipelineInputAssemblyStateCreateInfo;
GraphicsPipelineCreateInfo.pViewportState = &PipelineViewportStateCreateInfo;
GraphicsPipelineCreateInfo.pRasterizationState = &PipelineRasterizationStateCreateInfo;
GraphicsPipelineCreateInfo.pMultisampleState = &PipelineMultisampleStateCreateInfo;
GraphicsPipelineCreateInfo.pDepthStencilState = &DepthStencil; //如果渲染通道包含深度模版附件，则必须指定深度模版状态。
GraphicsPipelineCreateInfo.pColorBlendState = &PipelineColorBlendStateCreateInfo;
GraphicsPipelineCreateInfo.pDynamicState = nullptr; // Optional
GraphicsPipelineCreateInfo.layout = m_PipelineLayout; //是句柄不是结构体
GraphicsPipelineCreateInfo.renderPass = m_RenderPass;
GraphicsPipelineCreateInfo.subpass = 0;
//可以通过已经存在的管线创建新的图形管线（功能相同，节约内存消耗）
GraphicsPipelineCreateInfo.basePipelineHandle = VK_NULL_HANDLE;
GraphicsPipelineCreateInfo.basePipelineIndex = -1;
//创建图形管线
if (vkCreateGraphicsPipelines(m_Device, VK_NULL_HANDLE, 1, &GraphicsPipelineCreateInfo, nullptr, &m_GraphicsPipeline) != VK_SUCCESS)
{
throw std::runtime_error("failed to create graphics pipeline!");
}
//删除
vkDestroyShaderModule(m_Device, FragShaderModule, nullptr);
vkDestroyShaderModule(m_Device, VertShaderModule, nullptr);
}
//创建帧缓冲
void __createFrameBuffers()
{
//动态调整用于保存framebuffers的容器大小
m_SwapChainFrameBuffers.resize(m_SwapChainImageViews.size());
//迭代图像视图并通过他们建立对应FrameBuffer
for (size_t i = 0; i < m_SwapChainImageViews.size(); i++)
{
std::array<VkImageView, 2> Attachments = { m_SwapChainImageViews[i], m_DepthImageView };
VkFramebufferCreateInfo FrameBufferCreateInfo = {};
FrameBufferCreateInfo.sType = VK_STRUCTURE_TYPE_FRAMEBUFFER_CREATE_INFO;
FrameBufferCreateInfo.renderPass = m_RenderPass;
FrameBufferCreateInfo.attachmentCount = static_cast<uint32_t>(Attachments.size());;
FrameBufferCreateInfo.pAttachments = Attachments.data();
FrameBufferCreateInfo.width = m_SwapChainExtent.width;
FrameBufferCreateInfo.height = m_SwapChainExtent.height;
FrameBufferCreateInfo.layers = 1;
if (vkCreateFramebuffer(m_Device, &FrameBufferCreateInfo, nullptr, &m_SwapChainFrameBuffers[i]) != VK_SUCCESS)
{
throw std::runtime_error("failed to create Frame Buffer!");
}
}
}
//创建指令池
void __createCommandPool()
{
//每个指令池对象分配的指令缓冲对象只能提交给一个特定类型的队列，这里使用绘制指令
EQueueFamilyIndices QueueFamilyIndices = __findQueueFamilies(m_PhysicalDevice);
VkCommandPoolCreateInfo CommandPoolCreateInfo = {};
CommandPoolCreateInfo.sType = VK_STRUCTURE_TYPE_COMMAND_POOL_CREATE_INFO;
CommandPoolCreateInfo.queueFamilyIndex = QueueFamilyIndices.GraphicsFamily;
CommandPoolCreateInfo.flags = 0;
if (vkCreateCommandPool(m_Device, &CommandPoolCreateInfo, nullptr, &m_CommandPool) != VK_SUCCESS)
{
throw std::runtime_error("failed to create Command Pool!");
}
}
//创建指令缓冲对象
void __createCommandBuffers()
{
m_CommandBuffers.resize(m_SwapChainFrameBuffers.size());
VkCommandBufferAllocateInfo CommandBufferAllocateInfo = {};
CommandBufferAllocateInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_ALLOCATE_INFO;
CommandBufferAllocateInfo.commandPool = m_CommandPool;
CommandBufferAllocateInfo.level = VK_COMMAND_BUFFER_LEVEL_PRIMARY; //可以被提交到队列进行执行，但不能被其它指令缓冲对象调用。
CommandBufferAllocateInfo.commandBufferCount = (uint32_t)m_CommandBuffers.size();
if (vkAllocateCommandBuffers(m_Device, &CommandBufferAllocateInfo, m_CommandBuffers.data()) != VK_SUCCESS) {
throw std::runtime_error("failed to allocate command buffers!");
}
//记录指令到指令缓冲
for (size_t i = 0; i < m_CommandBuffers.size(); i++)
{
VkCommandBufferBeginInfo CommandBufferBeginInfo = {};
CommandBufferBeginInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_BEGIN_INFO;
CommandBufferBeginInfo.flags = VK_COMMAND_BUFFER_USAGE_SIMULTANEOUS_USE_BIT; //在指令缓冲等待执行时，仍然可以提交这一指令缓冲
CommandBufferBeginInfo.pInheritanceInfo = nullptr;
if (vkBeginCommandBuffer(m_CommandBuffers[i], &CommandBufferBeginInfo) != VK_SUCCESS) {
throw std::runtime_error("failed to begin recording command buffer!");
}
//开始渲染流程
VkRenderPassBeginInfo RenderPassBeginInfo = {};
RenderPassBeginInfo.sType = VK_STRUCTURE_TYPE_RENDER_PASS_BEGIN_INFO;
RenderPassBeginInfo.renderPass = m_RenderPass; //指定渲染流程对象
RenderPassBeginInfo.framebuffer = m_SwapChainFrameBuffers[i]; //指定帧缓冲对象
//渲染区域，这里与附着保持一致
RenderPassBeginInfo.renderArea.offset = { 0,0 };
RenderPassBeginInfo.renderArea.extent = m_SwapChainExtent;
//清除值
std::array<VkClearValue, 2> ClearValues = {};
ClearValues[0].color = { 0.0f, 0.0f, 0.0f, 1.0f };
ClearValues[1].depthStencil = { 1.0f, 0 };
RenderPassBeginInfo.clearValueCount = static_cast<uint32_t>(ClearValues.size());
RenderPassBeginInfo.pClearValues = ClearValues.data();
//参数：1.指令缓冲对象 2.渲染流程信息 3.所有要执行的指令都在主要指令缓冲中，没有辅助指令缓冲需要执行
vkCmdBeginRenderPass(m_CommandBuffers[i], &RenderPassBeginInfo, VK_SUBPASS_CONTENTS_INLINE);
//绑定图形管线:第二个参数用于指定管线对象是图形管线还是计算管线
vkCmdBindPipeline(m_CommandBuffers[i], VK_PIPELINE_BIND_POINT_GRAPHICS, m_GraphicsPipeline);
//绑定顶点缓冲区
VkBuffer VertexBuffers[] = { m_VertexBuffer };
VkDeviceSize OffSets[] = { 0 };
vkCmdBindVertexBuffers(m_CommandBuffers[i], 0, 1, VertexBuffers, OffSets);
//索引缓冲区
vkCmdBindIndexBuffer(m_CommandBuffers[i], m_IndexBuffer, 0, VK_INDEX_TYPE_UINT32);
//将描述符集合绑定到实际的着色器的描述符中
vkCmdBindDescriptorSets(m_CommandBuffers[i], VK_PIPELINE_BIND_POINT_GRAPHICS, m_PipelineLayout, 0, 1, &m_DescriptorSets[i], 0, nullptr);
//提交绘制操作到指令缓冲，
//参数：1.指定索引的数量 2.几何instanceing数量. 3：没有使用instancing，所以指定1。索引数表示被传递到顶点缓冲区中的顶点数量
//4：指定索引缓冲区的偏移量，使用1将会导致图形卡在第二个索引处开始读取 5.定索引缓冲区中添加的索引的偏移。 6.指定instancing偏移量，我们没有使用该特性
vkCmdDrawIndexed(m_CommandBuffers[i], static_cast<uint32_t>(Indices.size()), 1, 0, 0, 0);
//结束渲染流程
vkCmdEndRenderPass(m_CommandBuffers[i]);
//结束记录指令到指令缓冲
if (vkEndCommandBuffer(m_CommandBuffers[i]) != VK_SUCCESS) {
throw std::runtime_error("failed to record command buffer!");
}
}
}
//绘制步骤：需要同步操作，栅栏（用于应用程序与渲染操作进行同步）或者信号量（用于在命令队列内或者跨命令队列同步操作）实现同步
//1.首先，从交换链中获取一图像
//2.从帧缓冲中使用作为附件的图像执行命令缓冲区中的命令
//3.渲染后的图像返回交换链
void __drawFrame()
{
vkWaitForFences(m_Device, 1, &m_InFlightFences[m_CurrentFrame], VK_TRUE, UINT64_MAX);
uint32_t ImageIndex;
//重建交换链
VkResult Result = vkAcquireNextImageKHR(m_Device, m_SwapChain, std::numeric_limits<uint64_t>::max(), m_ImageAvailableSemaphores[m_CurrentFrame], VK_NULL_HANDLE, &ImageIndex);
if (Result == VK_ERROR_OUT_OF_DATE_KHR) //VK_ERROR_OUT_OF_DATE_KHR：交换链不能继续使用。通常发生在窗口大小改变后
{
__recreateSwapChain();
return;
}
else if (Result != VK_SUCCESS && Result != VK_SUBOPTIMAL_KHR)
{
throw std::runtime_error("failed to acquire SwapChain Image!");
}
__updateUniformBuffer(ImageIndex);
//让CPU在当前位置被阻塞掉，然后一直等待到它接受的Fence变为signaled的状态
//vkWaitForFences(m_Device, 1, &m_InFlightFences[m_CurrentFrame], VK_TRUE, UINT64_MAX);
//1.从交换链中获取一图像
//获取命令缓冲
//vkAcquireNextImageKHR(m_Device, m_SwapChain, UINT64_MAX, m_ImageAvailableSemaphores[m_CurrentFrame], VK_NULL_HANDLE, &ImageIndex);
if (m_ImagesInFlight[ImageIndex] != VK_NULL_HANDLE) {
vkWaitForFences(m_Device, 1, &m_ImagesInFlight[ImageIndex], VK_TRUE, UINT64_MAX);
}
m_ImagesInFlight[ImageIndex] = m_InFlightFences[m_CurrentFrame];
//2.提交命令缓冲区
VkSubmitInfo SubmitInfo = {};
SubmitInfo.sType = VK_STRUCTURE_TYPE_SUBMIT_INFO;
VkSemaphore WaitSemaphores[] = { m_ImageAvailableSemaphores[m_CurrentFrame] };
VkPipelineStageFlags WaitStages[] = { VK_PIPELINE_STAGE_COLOR_ATTACHMENT_OUTPUT_BIT };
SubmitInfo.waitSemaphoreCount = 1;
//执行开始之前要等待的哪个信号量及要等待的通道的哪个阶段
SubmitInfo.pWaitSemaphores = WaitSemaphores;
SubmitInfo.pWaitDstStageMask = WaitStages;
//指定哪个命令缓冲区被实际提交执行，这里为提交命令缓冲区，它将我们刚获取的交换链图像做为颜色附件进行绑定。
SubmitInfo.commandBufferCount = 1;
SubmitInfo.pCommandBuffers = &m_CommandBuffers[ImageIndex];
//指定了当命令缓冲区执行结束向哪些信号量发出信号。根据需要使用renderFinishedSemaphore
VkSemaphore SignalSemaphores[] = { m_RenderFinishedSemaphores[m_CurrentFrame] };
SubmitInfo.signalSemaphoreCount = 1;
SubmitInfo.pSignalSemaphores = SignalSemaphores;
vkResetFences(m_Device, 1, &m_InFlightFences[m_CurrentFrame]);
//向图像队列提交命令缓冲区
if (vkQueueSubmit(m_GraphicsQueue, 1, &SubmitInfo, VK_NULL_HANDLE) != VK_SUCCESS) {
throw std::runtime_error("failed to submit draw command buffer!");
}
//3.将结果提交到交换链，使其最终显示在屏幕上
VkPresentInfoKHR PresentInfo = {};
PresentInfo.sType = VK_STRUCTURE_TYPE_PRESENT_INFO_KHR;
PresentInfo.waitSemaphoreCount = 1;
//指定在进行presentation之前要等待的信号量
PresentInfo.pWaitSemaphores = SignalSemaphores;
//指定用于提交图像的交换链和每个交换链图像索引。大多数情况下仅一个
VkSwapchainKHR SwapChains[] = { m_SwapChain };
PresentInfo.swapchainCount = 1;
PresentInfo.pSwapchains = SwapChains;
PresentInfo.pImageIndices = &ImageIndex;
//PresentInfo.pResults = nullptr;
//不完全匹配时重建交换链
//提交请求呈现交换链中的图像
Result = vkQueuePresentKHR(m_PresentQueue, &PresentInfo);
if (Result == VK_ERROR_OUT_OF_DATE_KHR || Result == VK_SUBOPTIMAL_KHR || m_WindowResized) {
m_WindowResized = false;
__recreateSwapChain();
}
else if (Result != VK_SUCCESS) {
throw std::runtime_error("failed to present swap chain image!");
}
//开始绘制下一帧之前明确的等待presentation完成:
//vkQueueWaitIdle(m_PresentQueue);
m_CurrentFrame = (m_CurrentFrame + 1) % MAX_FRAMES_IN_FLIGHT;
}
//创建信号量
void __creatSemaphores()
{
m_ImageAvailableSemaphores.resize(MAX_FRAMES_IN_FLIGHT);
m_RenderFinishedSemaphores.resize(MAX_FRAMES_IN_FLIGHT);
m_InFlightFences.resize(MAX_FRAMES_IN_FLIGHT);
m_ImagesInFlight.resize(m_SwapChainImages.size(), VK_NULL_HANDLE);
VkSemaphoreCreateInfo semaphoreInfo = {};
semaphoreInfo.sType = VK_STRUCTURE_TYPE_SEMAPHORE_CREATE_INFO;
VkFenceCreateInfo fenceInfo = {};
fenceInfo.sType = VK_STRUCTURE_TYPE_FENCE_CREATE_INFO;
fenceInfo.flags = VK_FENCE_CREATE_SIGNALED_BIT;
for (size_t i = 0; i < MAX_FRAMES_IN_FLIGHT; i++) {
if (vkCreateSemaphore(m_Device, &semaphoreInfo, nullptr, &m_ImageAvailableSemaphores[i]) != VK_SUCCESS ||
vkCreateSemaphore(m_Device, &semaphoreInfo, nullptr, &m_RenderFinishedSemaphores[i]) != VK_SUCCESS ||
vkCreateFence(m_Device, &fenceInfo, nullptr, &m_InFlightFences[i]) != VK_SUCCESS) {
throw std::runtime_error("failed to create synchronization objects for a frame!");
}
}
}
//创建顶点缓冲,顶点着色器创建之后，使用CPU可见的缓冲作为临时缓冲，使用显卡读取较快的缓冲作为真正的顶点缓冲
//新的关联stagingBufferMemory作为内存的stagingBuffer缓冲对象来存放CPU加载的顶点数据。
void __createVertexBuffer()
{
VkDeviceSize bufferSize = sizeof(Vertices[0]) * Vertices.size();
VkBuffer stagingBuffer;
VkDeviceMemory stagingBufferMemory;
__createBuffer(bufferSize, VK_BUFFER_USAGE_TRANSFER_SRC_BIT,
VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT |
VK_MEMORY_PROPERTY_HOST_COHERENT_BIT, stagingBuffer,
stagingBufferMemory);
void* data;
vkMapMemory(m_Device, stagingBufferMemory, 0, bufferSize, 0, &data);
memcpy(data, Vertices.data(), (size_t)bufferSize);
vkUnmapMemory(m_Device, stagingBufferMemory);
//顶点缓冲区
__createBuffer(bufferSize, VK_BUFFER_USAGE_TRANSFER_DST_BIT | //缓冲可以被用作内存传输操作的目的缓冲
VK_BUFFER_USAGE_VERTEX_BUFFER_BIT, VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT,
m_VertexBuffer, m_VertexBufferMemory);
//vertexBuffer现在关联的内存是设备所有的，不能vkMapMemory函数对它关联的内存进行映射。
//只能通过stagingBuffer来向vertexBuffer复制数据
__copyBuffer(stagingBuffer, m_VertexBuffer, bufferSize);
//清除使用过的缓冲对象和关联的内存对象
vkDestroyBuffer(m_Device, stagingBuffer, nullptr);
vkFreeMemory(m_Device, stagingBufferMemory, nullptr);
//绑定顶点缓冲区，在__createCommandBuffer函数定义
}
//应根据缓冲区和应用程序来找到正确的内存类型
uint32_t __findMemoryType(uint32_t vTypeFilter, VkMemoryPropertyFlags vkMemoryPropertyFlags)
{
//遍历有效的内存类型
VkPhysicalDeviceMemoryProperties MemoryProperties;
vkGetPhysicalDeviceMemoryProperties(m_PhysicalDevice, &MemoryProperties);
for (uint32_t i = 0; i < MemoryProperties.memoryTypeCount; i++)
{
//并根据需要的类型与每个内存属性的类型进行AND操作，判断是否为1
//memoryTypes数组描述了堆以及每个内存类型的相关属性。属性定义了内存的特殊功能，可以从CPU向它写入数据。
if (vTypeFilter & (1 << i) && (MemoryProperties.memoryTypes[i].propertyFlags & vkMemoryPropertyFlags) == vkMemoryPropertyFlags) return i;
}
throw std::runtime_error("failed to find suitable memory type!");
}
//创建抽象缓冲
void __createBuffer(VkDeviceSize vSize, VkBufferUsageFlags vUsage,
VkMemoryPropertyFlags vProperties, VkBuffer& pBuffer, VkDeviceMemory& pBufferMemory)
{
VkBufferCreateInfo BufferCreateInfo = {};
BufferCreateInfo.sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO;
BufferCreateInfo.size = vSize;
BufferCreateInfo.usage = vUsage; //指定缓冲区数据如何被使用
BufferCreateInfo.sharingMode = VK_SHARING_MODE_EXCLUSIVE; //冲区也可以由特定的队列簇占有或者多个同时共享,在这里我们使用图形队列，使用独占访问模式
if (vkCreateBuffer(m_Device, &BufferCreateInfo, nullptr, &pBuffer) != VK_SUCCESS)
{
throw std::runtime_error("failed to create buffer!");
}
//虽然缓冲区已经创建好，但实际上并没有分配任何可用内存，因此需要申请内存
VkMemoryRequirements MemoryRequirements;
vkGetBufferMemoryRequirements(m_Device, pBuffer, &MemoryRequirements);
//有了内存之后，就分配内存
VkMemoryAllocateInfo MemoryAllocateInfo = {};
MemoryAllocateInfo.sType = VK_STRUCTURE_TYPE_MEMORY_ALLOCATE_INFO;
MemoryAllocateInfo.allocationSize = MemoryRequirements.size;
//显卡可以分配不同类型的内存，因此我们应根据缓冲区和应用程序来找到正确的内存类型
MemoryAllocateInfo.memoryTypeIndex = __findMemoryType(MemoryRequirements.memoryTypeBits, VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT | VK_MEMORY_PROPERTY_HOST_COHERENT_BIT);
if (vkAllocateMemory(m_Device, &MemoryAllocateInfo, nullptr, &pBufferMemory) != VK_SUCCESS)
{
throw std::runtime_error("failed to allocate buffer memory!");
}
//分配好内存，将内存关联到缓冲区,返回创建的缓冲对象和它关联的内存对象
vkBindBufferMemory(m_Device, pBuffer, pBufferMemory, 0);
}
//复制缓冲区数据
void __copyBuffer(VkBuffer vSrcBuffer, VkBuffer vDstBuffer, VkDeviceSize vSize)
{
VkCommandBuffer CommandBuffer;
CommandBuffer = __beginSingleTimeCommands();
//缓冲数据复制
VkBufferCopy BufferCopy = {}; //指定复制操作源缓冲位置偏移
BufferCopy.srcOffset = 0;
BufferCopy.dstOffset = 0;
BufferCopy.size = vSize;
vkCmdCopyBuffer(CommandBuffer, vSrcBuffer, vDstBuffer, 1, &BufferCopy);
__endSingleTimeCommands(CommandBuffer);
}
//布局变换
VkCommandBuffer __beginSingleTimeCommands()
{
VkCommandBufferAllocateInfo CommandBufferAllocateInfo = {};
CommandBufferAllocateInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_ALLOCATE_INFO;
CommandBufferAllocateInfo.level = VK_COMMAND_BUFFER_LEVEL_PRIMARY;
CommandBufferAllocateInfo.commandPool = m_CommandPool;
CommandBufferAllocateInfo.commandBufferCount = 1;
VkCommandBuffer CommandBuffer;
vkAllocateCommandBuffers(m_Device, &CommandBufferAllocateInfo, &CommandBuffer);
//记录内存传输指令
VkCommandBufferBeginInfo CommandBufferBeginInfo = {};
CommandBufferBeginInfo.sType = VK_STRUCTURE_TYPE_COMMAND_BUFFER_BEGIN_INFO;
CommandBufferBeginInfo.flags = VK_COMMAND_BUFFER_USAGE_ONE_TIME_SUBMIT_BIT; //标记告诉驱动程序我们如何使用这个指令缓冲
vkBeginCommandBuffer(CommandBuffer, &CommandBufferBeginInfo);
return CommandBuffer;
}
//布局变换
void __endSingleTimeCommands(VkCommandBuffer vCommandBuffer)
{
//此处内存传输操作使用的指令缓冲只有复制指令，记录完，就会结束指令缓冲的基础擦欧总，提交指令缓冲完成传输操作执行
vkEndCommandBuffer(vCommandBuffer);
//等待传输操作完成,有两种等待的方法
//1.使用栅栏(fence)，通过vkWaitForFences函数等待，使用栅栏(fence)可以同步多个不同的内存传输操作，给驱动程序的优化空间也更大
//2.通过vkQueueWaitIdle函数等待
VkSubmitInfo SubmitInfo = {};
SubmitInfo.sType = VK_STRUCTURE_TYPE_SUBMIT_INFO;
SubmitInfo.commandBufferCount = 1;
SubmitInfo.pCommandBuffers = &vCommandBuffer;
vkQueueSubmit(m_GraphicsQueue, 1, &SubmitInfo, VK_NULL_HANDLE);
vkQueueWaitIdle(m_GraphicsQueue);
//清除指令缓冲对象
vkFreeCommandBuffers(m_Device, m_CommandPool, 1, &vCommandBuffer);
}
//使用缓冲区读取图像，则命令要求图像在正确的布局中
void __transitionImageLayout(VkImage vImage, VkFormat vFormat, VkImageLayout vOldLayout, VkImageLayout VNewLayout)
{
VkCommandBuffer CommandBuffer = __beginSingleTimeCommands();
//主流的做法用于处理图像变换是使用 image memory barrier,主要用于访问资源的同步操作
VkImageMemoryBarrier ImageMemoryBarrier = {};
ImageMemoryBarrier.sType = VK_STRUCTURE_TYPE_IMAGE_MEMORY_BARRIER;
ImageMemoryBarrier.oldLayout = vOldLayout;
ImageMemoryBarrier.newLayout = VNewLayout;
//如果针对传输队列簇的宿主使用屏障，这两个参数需要设置队列簇的索引。如果不关心，则必须设置VK_QUEUE_FAMILY_IGNORED(不是默认值)。
ImageMemoryBarrier.srcQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED;
ImageMemoryBarrier.dstQueueFamilyIndex = VK_QUEUE_FAMILY_IGNORED;
//image和subresourceRange指定受到影响的图像和图像的特定区域。
//我们的图像不是数组，也没有使用mipmapping levels，所以只指定一级，并且一个层。
ImageMemoryBarrier.image = vImage;
ImageMemoryBarrier.subresourceRange.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT;
ImageMemoryBarrier.subresourceRange.baseMipLevel = 0;
ImageMemoryBarrier.subresourceRange.levelCount = 1;
ImageMemoryBarrier.subresourceRange.baseArrayLayer = 0;
ImageMemoryBarrier.subresourceRange.layerCount = 1;
ImageMemoryBarrier.srcAccessMask = 0;
ImageMemoryBarrier.dstAccessMask = 0;
//预屏障:色器读取操作应该等待传输写入，特别是 fragment shader进行读取，因为这是要使用纹理的地方
VkPipelineStageFlags SourceStage;
VkPipelineStageFlags DestinationStage;
//添加正确的访问掩码和管线阶段：
if (vOldLayout == VK_IMAGE_LAYOUT_UNDEFINED && VNewLayout == VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL) {
ImageMemoryBarrier.srcAccessMask = 0;
ImageMemoryBarrier.dstAccessMask = VK_ACCESS_TRANSFER_WRITE_BIT;
SourceStage = VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT;
DestinationStage = VK_PIPELINE_STAGE_TRANSFER_BIT;
}
else if (vOldLayout == VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL && VNewLayout == VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL) {
ImageMemoryBarrier.srcAccessMask = VK_ACCESS_TRANSFER_WRITE_BIT;
ImageMemoryBarrier.dstAccessMask = VK_ACCESS_SHADER_READ_BIT;
SourceStage = VK_PIPELINE_STAGE_TRANSFER_BIT;
DestinationStage = VK_PIPELINE_STAGE_FRAGMENT_SHADER_BIT;
}
else if (vOldLayout == VK_IMAGE_LAYOUT_UNDEFINED && VNewLayout == VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL) {
ImageMemoryBarrier.srcAccessMask = 0;
ImageMemoryBarrier.dstAccessMask = VK_ACCESS_DEPTH_STENCIL_ATTACHMENT_READ_BIT | VK_ACCESS_DEPTH_STENCIL_ATTACHMENT_WRITE_BIT;
SourceStage = VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT;
DestinationStage = VK_PIPELINE_STAGE_EARLY_FRAGMENT_TESTS_BIT;
}
else {
throw std::invalid_argument("unsupported layout transition!");
}
if (VNewLayout == VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL) {
ImageMemoryBarrier.subresourceRange.aspectMask = VK_IMAGE_ASPECT_DEPTH_BIT;
if (__hasStencilComponent(vFormat))
{
ImageMemoryBarrier.subresourceRange.aspectMask |= VK_IMAGE_ASPECT_STENCIL_BIT;
}
}
else {
ImageMemoryBarrier.subresourceRange.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT;
}
//屏障主要用于同步目的，所以必须在应用屏障前指定哪一种操作类型及涉及到的资源，同时要指定哪一种操作及资源必须等待屏障。
vkCmdPipelineBarrier(
CommandBuffer,
SourceStage, DestinationStage,
0,
0, nullptr,
0, nullptr,
1, &ImageMemoryBarrier
);
__endSingleTimeCommands(CommandBuffer);
}
//缓冲区拷贝到图像
void __copyBufferToImage(VkBuffer vBuffer, VkImage vImage, uint32_t vWidth, uint32_t vHeight)
{
VkCommandBuffer CommandBuffer = __beginSingleTimeCommands();
//指定拷贝具体哪一部分到图像的区域
VkBufferImageCopy Region = {};
Region.bufferOffset = 0; //指定缓冲区中byte偏移量
Region.bufferRowLength = 0; //指定像素在内存中的布局，0表示像素紧密排列
Region.bufferImageHeight = 0;
//imageSubresource，imageOffset 和 imageExtent字段指定我们将要拷贝图像的哪一部分像素
Region.imageSubresource.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT;
Region.imageSubresource.mipLevel = 0;
Region.imageSubresource.baseArrayLayer = 0;
Region.imageSubresource.layerCount = 1;
Region.imageOffset = { 0, 0, 0 };
Region.imageExtent = {
vWidth,
vHeight,
1
};
//缓冲区拷贝到图像的操作将会使用vkCmdCopyBufferToImage函数到队列中：
vkCmdCopyBufferToImage(
CommandBuffer,
vBuffer,
vImage,
VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL,
1,
&Region
);
__endSingleTimeCommands(CommandBuffer);
}
//索引缓冲
void __createIndexBuffer()
{
VkDeviceSize BufferSize = sizeof(Indices[0]) * Indices.size();
VkBuffer StagingBuffer;
VkDeviceMemory StagingBufferMemory;
__createBuffer(BufferSize, VK_BUFFER_USAGE_TRANSFER_SRC_BIT,
VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT | VK_MEMORY_PROPERTY_HOST_COHERENT_BIT,
StagingBuffer, StagingBufferMemory);
void* data;
vkMapMemory(m_Device, StagingBufferMemory, 0, BufferSize, 0, &data);
memcpy(data, Indices.data(), (size_t)BufferSize);
vkUnmapMemory(m_Device, StagingBufferMemory);
//顶点缓冲区
__createBuffer(BufferSize, VK_BUFFER_USAGE_TRANSFER_DST_BIT | //缓冲可以被用作内存传输操作的目的缓冲
VK_BUFFER_USAGE_INDEX_BUFFER_BIT, VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT,
m_IndexBuffer, m_IndexBufferMemory);
//vertexBuffer现在关联的内存是设备所有的，不能vkMapMemory函数对它关联的内存进行映射。
//只能通过stagingBuffer来向vertexBuffer复制数据
__copyBuffer(StagingBuffer, m_IndexBuffer, BufferSize);
//清除使用过的缓冲对象和关联的内存对象
vkDestroyBuffer(m_Device, StagingBuffer, nullptr);
vkFreeMemory(m_Device, StagingBufferMemory, nullptr);
//绑定顶点缓冲区，在__createCommandBuffer函数定义
}
//每个绑定都会通过VkDescriptorSetLayoutBinding结构体描述
void __createDescritorSetLayout()
{
//UBO描述符
VkDescriptorSetLayoutBinding UBOLayoutBinding = {};
UBOLayoutBinding.binding = 0; //着色器中使用的binding
UBOLayoutBinding.descriptorType = VK_DESCRIPTOR_TYPE_UNIFORM_BUFFER; //指定着色器中描述符类型
UBOLayoutBinding.descriptorCount = 1; //指定数组中数值，由于MVP变换为单UBO对象，因此为1
UBOLayoutBinding.stageFlags = VK_SHADER_STAGE_VERTEX_BIT; //指定描述符在着色器哪个阶段杯引用，这里仅在顶点着色器中使用描述符
UBOLayoutBinding.pImmutableSamplers = nullptr; //与图像采样的描述符有关
//另一个描述符：组合图像取样器(combined image sampler)
VkDescriptorSetLayoutBinding SamplerLayoutBinding{};
SamplerLayoutBinding.binding = 1;
SamplerLayoutBinding.descriptorCount = 1;
SamplerLayoutBinding.descriptorType = VK_DESCRIPTOR_TYPE_COMBINED_IMAGE_SAMPLER;
SamplerLayoutBinding.pImmutableSamplers = nullptr;
//指定在片段着色器中使用组合图像采样器描述符
SamplerLayoutBinding.stageFlags = VK_SHADER_STAGE_FRAGMENT_BIT;
std::array<VkDescriptorSetLayoutBinding, 2> bindings = { UBOLayoutBinding, SamplerLayoutBinding };
VkDescriptorSetLayoutCreateInfo layoutInfo{};
layoutInfo.sType = VK_STRUCTURE_TYPE_DESCRIPTOR_SET_LAYOUT_CREATE_INFO;
layoutInfo.bindingCount = static_cast<uint32_t>(bindings.size());
layoutInfo.pBindings = bindings.data();
if (vkCreateDescriptorSetLayout(m_Device, &layoutInfo, nullptr, &m_DescriptorSetLayout) != VK_SUCCESS) {
throw std::runtime_error("failed to create descriptor set layout!");
}
}
//uniform缓冲区
//在每一帧中我们需要拷贝新的数据到UBO缓冲区，所以存在一个暂存缓冲区是没有意义的。在这种情况下，它只会增加额外的开销，并且可能降低性能而不是提升性能。
void __createUniformBuffer()
{
VkDeviceSize BufferSize = sizeof(UniformBufferObject);
m_UniformBuffers.resize(m_SwapChainImages.size());
m_UniformBuffersMemory.resize(m_SwapChainImages.size());
for (size_t i = 0; i < m_SwapChainImages.size(); i++) {
__createBuffer(BufferSize, VK_BUFFER_USAGE_UNIFORM_BUFFER_BIT,
VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT |
VK_MEMORY_PROPERTY_HOST_COHERENT_BIT,
m_UniformBuffers[i], m_UniformBuffersMemory[i]);
}
}
//更新uniform数据
//该函数会在每一帧中创建新的变换矩阵以确保几何图形旋转。我们需要引入新的头文件使用该功能：
void __updateUniformBuffer(uint32_t vCurrentImage)
{
static auto StartTime = std::chrono::high_resolution_clock::now();
auto CurrentTime = std::chrono::high_resolution_clock::now();
float Time = std::chrono::duration<float, std::chrono::seconds::period>(CurrentTime - StartTime).count() / 1000.0f;
UniformBufferObject UBO = {};
UBO.ModelMatrix = glm::rotate(glm::mat4(1.0f), Time * glm::radians(90.0f), glm::vec3(0.0f, 0.0f, 1.0f));
//glm::lookAt 函数以眼睛位置，中心位置和上方向为参数。
UBO.ViewMatrix = glm::lookAt(glm::vec3(2.0f, 2.0f, 2.0f), glm::vec3(0.0f, 0.0f, 0.0f), glm::vec3(0.0f, 0.0f, 1.0f));
//选择使用FOV为45度的透视投影。其他参数是宽高比，近裁剪面和远裁剪面。重要的是使用当前的交换链扩展来计算宽高比，以便在窗体调整大小后参考最新的窗体宽度和高度。
UBO.ProjectiveMatrix = glm::perspective(glm::radians(45.0f), m_SwapChainExtent.width / (float)m_SwapChainExtent.height, 0.1f, 10.0f);
//GLM最初是为OpenGL设计的，它的裁剪坐标的Y是反转的。修正该问题的最简单的方法是在投影矩阵中Y轴的缩放因子反转
UBO.ProjectiveMatrix[1][1] *= -1;
//以将UBO中的数据复制到uniform缓冲区
void* Data;
vkMapMemory(m_Device, m_UniformBuffersMemory[vCurrentImage], 0, sizeof(UBO), 0, &Data);
memcpy(Data, &UBO, sizeof(UBO));
vkUnmapMemory(m_Device, m_UniformBuffersMemory[vCurrentImage]);
}
//创建描述符池
void __createDescritorPool()
{
std::array<VkDescriptorPoolSize, 2> PoolSizes = {};
PoolSizes[0].type = VK_DESCRIPTOR_TYPE_UNIFORM_BUFFER;
PoolSizes[0].descriptorCount = static_cast<uint32_t>(m_SwapChainImages.size());
PoolSizes[1].type = VK_DESCRIPTOR_TYPE_COMBINED_IMAGE_SAMPLER;
PoolSizes[1].descriptorCount = static_cast<uint32_t>(m_SwapChainImages.size());
//定义描述符池的大小
VkDescriptorPoolCreateInfo DescriptorPoolCreateInfo = {};
DescriptorPoolCreateInfo.sType = VK_STRUCTURE_TYPE_DESCRIPTOR_POOL_CREATE_INFO;
DescriptorPoolCreateInfo.pPoolSizes = PoolSizes.data();
DescriptorPoolCreateInfo.poolSizeCount = static_cast<uint32_t>(PoolSizes.size());
DescriptorPoolCreateInfo.maxSets = static_cast<uint32_t>(m_SwapChainImages.size()); //指定可以分配的最大描述符集个数：
if (vkCreateDescriptorPool(m_Device, &DescriptorPoolCreateInfo, nullptr, &m_DescriptorPool) != VK_SUCCESS)
{
throw std::runtime_error("failed to create descriptor pool!");
}
}
//有个描述符池就可以开始创建描述符集对象了
// 最后一步是将实际的图像和采样器资源绑定到描述符集合中的具体描述符
void __createDescriptorSet()
{
std::vector<VkDescriptorSetLayout> DescriptorSetLayouts(m_SwapChainImages.size(), m_DescriptorSetLayout);
VkDescriptorSetAllocateInfo AllocateInfo = {};
AllocateInfo.sType = VK_STRUCTURE_TYPE_DESCRIPTOR_SET_ALLOCATE_INFO;
AllocateInfo.descriptorPool = m_DescriptorPool;
AllocateInfo.descriptorSetCount = static_cast<uint32_t>(m_SwapChainImages.size());
AllocateInfo.pSetLayouts = DescriptorSetLayouts.data();
m_DescriptorSets.resize(m_SwapChainImages.size());
//不需要明确清理描述符集合，因为它们会在描述符对象池销毁的时候自动清
if (vkAllocateDescriptorSets(m_Device, &AllocateInfo, m_DescriptorSets.data()) != VK_SUCCESS)
{
throw std::runtime_error("failed to allocate descriptor set!");
}
for (size_t i = 0; i < m_SwapChainImages.size(); i++)
{
//描述符集合分配好之后，配置描述符
VkDescriptorBufferInfo BufferInfo{};
BufferInfo.buffer = m_UniformBuffers[i];
BufferInfo.offset = 0;
BufferInfo.range = sizeof(UniformBufferObject);
VkDescriptorImageInfo imageInfo{};
imageInfo.imageLayout = VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL;
imageInfo.imageView = m_TextureImageView;
imageInfo.sampler = m_TextureSampler;
std::array<VkWriteDescriptorSet, 2> descriptorWrites{};
descriptorWrites[0].sType = VK_STRUCTURE_TYPE_WRITE_DESCRIPTOR_SET;
descriptorWrites[0].dstSet = m_DescriptorSets[i];
descriptorWrites[0].dstBinding = 0;
descriptorWrites[0].dstArrayElement = 0;
descriptorWrites[0].descriptorType = VK_DESCRIPTOR_TYPE_UNIFORM_BUFFER;
descriptorWrites[0].descriptorCount = 1;
descriptorWrites[0].pBufferInfo = &BufferInfo;
descriptorWrites[1].sType = VK_STRUCTURE_TYPE_WRITE_DESCRIPTOR_SET;
descriptorWrites[1].dstSet = m_DescriptorSets[i];
descriptorWrites[1].dstBinding = 1; ////为 uniform buffer 绑定的索引设定为0
descriptorWrites[1].dstArrayElement = 0; ////描述符可以是数组，所以需要指定要更新的数组索引。在这里没有使用数组，所以简单的设置为0
descriptorWrites[1].descriptorType = VK_DESCRIPTOR_TYPE_COMBINED_IMAGE_SAMPLER; //指定描述符类型
descriptorWrites[1].descriptorCount = 1; //指定描述多少描述符需要被更新
descriptorWrites[1].pImageInfo = &imageInfo; //指定描述符引用的图像数据
vkUpdateDescriptorSets(m_Device, static_cast<uint32_t>(descriptorWrites.size()), descriptorWrites.data(), 0, nullptr);
////现在需要更新__createCommandBuffers函数，使用cmdBindDescriptorSets将描述符集合绑定到实际的着色器的描述符中：
}
}
//加载图片和提交到VulKan图像对象中，途中会用到命令缓冲区，因此在__createCommandPool函数之后
void __createTextureImage()
{
int TexWidth, TexHeight, TexChannels;
//STBI_rgb_alpha值强制加载图片的alpha通道，即使没有改通道值，像素逐行排列，每个像素有四个字节
stbi_uc* Pixels = stbi_load(TEXTURE_PATH.c_str(), &TexWidth, &TexHeight, &TexChannels, STBI_rgb_alpha);
VkDeviceSize ImageSize = TexWidth * TexHeight * 4;
if (!Pixels) {
throw std::runtime_error("failed to load texture image!");
}
VkBuffer StagingBuffer;
VkDeviceMemory StagingBufferMemory;
//创建临时缓冲区
__createBuffer(ImageSize, VK_BUFFER_USAGE_TRANSFER_SRC_BIT,
VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT | VK_MEMORY_PROPERTY_HOST_COHERENT_BIT,
StagingBuffer, StagingBufferMemory);
void* data;
vkMapMemory(m_Device, StagingBufferMemory, 0, ImageSize, 0, &data);
memcpy(data, Pixels, static_cast<size_t>(ImageSize));
vkUnmapMemory(m_Device, StagingBufferMemory);
//不要忘记清理原图像的像素数据
stbi_image_free(Pixels);
__createImage(TexWidth, TexHeight, VK_FORMAT_R8G8B8A8_UNORM, VK_IMAGE_TILING_OPTIMAL,
VK_IMAGE_USAGE_TRANSFER_DST_BIT | VK_IMAGE_USAGE_SAMPLED_BIT, VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT,
m_TextureImage, m_TextureImageMemory);
//创建贴图图像，下一步copy暂存缓冲区到贴图图像
//1.变换贴图图像到 VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL
__transitionImageLayout(m_TextureImage, VK_FORMAT_R8G8B8A8_SRGB, VK_IMAGE_LAYOUT_UNDEFINED, VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL);
//2.执行缓冲区到图像的拷贝操作
__copyBufferToImage(StagingBuffer, m_TextureImage, static_cast<uint32_t>(TexWidth), static_cast<uint32_t>(TexHeight));
//在shader着色器中开始从贴图图像的采样，我们需要最后一个变换来准备着色器访问
__transitionImageLayout(m_TextureImage, VK_FORMAT_R8G8B8A8_SRGB, VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL, VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL);
vkDestroyBuffer(m_Device, StagingBuffer, nullptr);
vkFreeMemory(m_Device, StagingBufferMemory, nullptr);
}
//创建图像对象和内存分配
void __createImage(uint32_t vWidth, uint32_t vHeight, VkFormat vFormat, VkImageTiling vTiling, VkImageUsageFlags vUsage,
VkMemoryPropertyFlags vProperties, VkImage& vImage, VkDeviceMemory& vImageMemory)
{
//本可以通过着色器访问缓冲区中像素值，但在VuklKam中最好使用image对象访问缓冲区，可以快速检索颜色
VkImageCreateInfo ImageCreateInfo = {};
ImageCreateInfo.sType = VK_STRUCTURE_TYPE_IMAGE_CREATE_INFO;
ImageCreateInfo.imageType = VK_IMAGE_TYPE_2D; //指定图像类型，告知Vulkan采用什么样的坐标系在图像中采集纹素。它可以是1D，2D和3D图像。1D图像用于存储数组数据或者灰度图，2D图像主要用于纹理贴图，3D图像用于存储立体纹素。
ImageCreateInfo.extent.width = static_cast<uint32_t>(vWidth); //图像尺寸
ImageCreateInfo.extent.height = static_cast<uint32_t>(vHeight);
ImageCreateInfo.extent.depth = 1;
ImageCreateInfo.mipLevels = 1;
ImageCreateInfo.arrayLayers = 1;
ImageCreateInfo.format = vFormat; //指定图像格式，但必须保证缓冲区纹素与像素一样的格式，否则copy失败
ImageCreateInfo.tiling = vTiling; //纹素基于具体的实现来定义布局，以实现最佳访问,其中，不能在以后修改，如何需要在内存图像中更直接访问纹素，必须使用VK_IMAGE_TILING_LINEAR（纹素基于行主序的布局，如pixels数组）
ImageCreateInfo.initialLayout = VK_IMAGE_LAYOUT_UNDEFINED; //GPU不能使用，第一个变换将丢弃纹素，在这里几乎灭有必要在第一次转换时保留纹素
ImageCreateInfo.usage = vUsage; //图像将会被用作缓冲区copy目标，这里图像作为传输目的地，此外希望从着色器中访问图像对mesh进行着色
ImageCreateInfo.sharingMode = VK_SHARING_MODE_EXCLUSIVE; //图像会在一个队列簇中使用：支持图形或者传输操作
ImageCreateInfo.samples = VK_SAMPLE_COUNT_1_BIT; //标志位与多重采样相关。这仅仅适用于作为附件的图像，所以我们坚持一个采样数值。与稀疏图像相关的图像有一些可选的标志
ImageCreateInfo.flags = 0;
if (vkCreateImage(m_Device, &ImageCreateInfo, nullptr, &vImage) != VK_SUCCESS) {
throw std::runtime_error("failed to create image!");
}
//为图像工作分配内存
VkMemoryRequirements MemoryRequirements;
vkGetImageMemoryRequirements(m_Device, vImage, &MemoryRequirements);
VkMemoryAllocateInfo allocInfo = {};
allocInfo.sType = VK_STRUCTURE_TYPE_MEMORY_ALLOCATE_INFO;
allocInfo.allocationSize = MemoryRequirements.size;
allocInfo.memoryTypeIndex = __findMemoryType(MemoryRequirements.memoryTypeBits, vProperties);
if(vkAllocateMemory(m_Device, &allocInfo, nullptr, &vImageMemory) != VK_SUCCESS) {
throw std::runtime_error("failed to allocate image memory!");
}
vkBindImageMemory(m_Device, vImage, vImageMemory, 0);
}
//创建纹理视图
void __createTextureImageView()
{
m_TextureImageView = __createImageView(m_TextureImage, VK_FORMAT_R8G8B8A8_UNORM, VK_IMAGE_ASPECT_COLOR_BIT);
}
//着色器直接从图像中读取纹素是可以的，但是当它们作为纹理图像的时候并不常见。纹理图像通常使用采样器访问
//创建采样器对象
void __createTextureSampler()
{
VkSamplerCreateInfo SamplerInfo = {};
//指定过滤器和变换
SamplerInfo.sType = VK_STRUCTURE_TYPE_SAMPLER_CREATE_INFO;
SamplerInfo.magFilter = VK_FILTER_LINEAR; //指定纹素放大内插值方式
SamplerInfo.minFilter = VK_FILTER_LINEAR; //指定纹素缩小内插值方式
//指定每个轴向使用的寻址模式，纹理空间坐标UVW对应着XYZ
SamplerInfo.addressModeU = VK_SAMPLER_ADDRESS_MODE_REPEAT; //当超过图像尺寸的时候采用循环填充。
SamplerInfo.addressModeV = VK_SAMPLER_ADDRESS_MODE_REPEAT;
SamplerInfo.addressModeW = VK_SAMPLER_ADDRESS_MODE_REPEAT;
SamplerInfo.anisotropyEnable = VK_TRUE; // 指定是否使用各向异性过滤器。没有理由不使用该特性，除非性能是一个问题
SamplerInfo.maxAnisotropy = 16; //限制可用于计算最终颜色的纹素采样的数量。低的数值会得到比较好的性能，但是会得到较差的质量
//定采样范围超过图像时候返回的颜色，与之对应的是边缘寻址模式。可以以float或者int格式返回黑色，白色或者透明度。但是不能指定任意颜色
SamplerInfo.borderColor = VK_BORDER_COLOR_INT_OPAQUE_BLACK;
//指定使用的坐标系统，用于访问图像的纹素。如果字段为VK_TRUE，意味着可以简单的使用坐标范围为 [ 0, texWidth ) 和 [ 0, texHeight )。
//如果使用VK_FALSE，意味着每个轴向纹素访问使用 [ 0, 1) 范围。真实的应用程序总是使用归一化的坐标。
SamplerInfo.unnormalizedCoordinates = VK_FALSE;
//如果开启比较功能，那么纹素首先和值进行比较，并且比较后的值用于过滤操作。主要用在阴影纹理映射的percentage-closer filtering 即百分比近似过滤器。
SamplerInfo.compareEnable = VK_FALSE;
SamplerInfo.compareOp = VK_COMPARE_OP_ALWAYS;
//mipmapping
SamplerInfo.mipmapMode = VK_SAMPLER_MIPMAP_MODE_LINEAR;
SamplerInfo.mipLodBias = 0.0f;
SamplerInfo.minLod = 0.0f;
SamplerInfo.maxLod = 0.0f;
if (vkCreateSampler(m_Device, &SamplerInfo, nullptr, &m_TextureSampler) != VK_SUCCESS) {
throw std::runtime_error("failed to create texture sampler!");
}
}
//s申请深度图像的资源
void __createDepthResources()
{
//深度格式
VkFormat DepthFormat = __findDepthFormat();
__createImage(m_SwapChainExtent.width, m_SwapChainExtent.height, DepthFormat, VK_IMAGE_TILING_OPTIMAL, VK_IMAGE_USAGE_DEPTH_STENCIL_ATTACHMENT_BIT, VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT, m_DepthImage, m_DepthImageMemory);
m_DepthImageView = __createImageView(m_DepthImage, DepthFormat, VK_IMAGE_ASPECT_DEPTH_BIT);
//变换为合适的深度附件使用的布局，在这里我们选择使用管线屏障，因为变换只会发生一次
//__transitionImageLayout(m_DepthImage, DepthFormat, VK_IMAGE_LAYOUT_UNDEFINED, VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL);
}
//择具有深度组件的格式，该深度组件支持使用深度附件：
VkFormat __findDepthFormat() {
return __findSupportedFormat({ VK_FORMAT_D32_SFLOAT, VK_FORMAT_D32_SFLOAT_S8_UINT, VK_FORMAT_D24_UNORM_S8_UINT },
VK_IMAGE_TILING_OPTIMAL, VK_FORMAT_FEATURE_DEPTH_STENCIL_ATTACHMENT_BIT );
}
//检测第一个所支持的格式
VkFormat __findSupportedFormat(const std::vector<VkFormat>& vCandidates, VkImageTiling vTiling, VkFormatFeatureFlags vFeatures)
{
for (VkFormat format : vCandidates) {
VkFormatProperties props;
vkGetPhysicalDeviceFormatProperties(m_PhysicalDevice, format, &props);
if (vTiling == VK_IMAGE_TILING_LINEAR && (props.linearTilingFeatures & vFeatures) == vFeatures) {
return format;
}
else if (vTiling == VK_IMAGE_TILING_OPTIMAL && (props.optimalTilingFeatures & vFeatures) == vFeatures) {
return format;
}
}
throw std::runtime_error("failed to find supported format!");
}
//所选择的深度格式是否包含模版组件
bool __hasStencilComponent(VkFormat vFormat) {
return vFormat == VK_FORMAT_D32_SFLOAT_S8_UINT || vFormat == VK_FORMAT_D24_UNORM_S8_UINT;
}
//加载模型
void __loadModel() {
tinyobj::attrib_t Attrib;
std::vector<tinyobj::shape_t> shapes;
std::vector<tinyobj::material_t> materials;
std::string warn, err;
if (!tinyobj::LoadObj(&Attrib, &shapes, &materials, &warn, &err, MODEL_PATH.c_str())) {
throw std::runtime_error(warn + err);
}
std::unordered_map<EVertex, uint32_t> uniqueVertices{};
for (const auto& shape : shapes) {
for (const auto& index : shape.mesh.indices) {
EVertex 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 (uniqueVertices.count(vertex) == 0) {
uniqueVertices[vertex] = static_cast<uint32_t>(Vertices.size());
Vertices.push_back(vertex);
}
Indices.push_back(uniqueVertices[vertex]);
}
}
}
};
}结果：
来源：oschina
链接：https://my.oschina.net/1024and1314/blog/4360367