// Copyright(c) 2019, NVIDIA CORPORATION. All rights reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
//
// VulkanHpp Samples : DynamicUniform
//                     Draw 2 Cubes using dynamic uniform buffer

#include "../utils/geometries.hpp"
#include "../utils/math.hpp"
#include "../utils/shaders.hpp"
#include "../utils/utils.hpp"
#include "vulkan/vulkan.hpp"
#include "SPIRV/GlslangToSpv.h"
#include <iostream>
#include <thread>

static char const* AppName = "DynamicUniform";
static char const* EngineName = "Vulkan.hpp";

int main(int /*argc*/, char ** /*argv*/)
{
  try
  {
    vk::UniqueInstance instance = vk::su::createInstance(AppName, EngineName, {}, vk::su::getInstanceExtensions());
#if !defined(NDEBUG)
    vk::UniqueDebugUtilsMessengerEXT debugUtilsMessenger = vk::su::createDebugUtilsMessenger(instance);
#endif

    vk::PhysicalDevice physicalDevice = instance->enumeratePhysicalDevices().front();

    vk::su::SurfaceData surfaceData(instance, AppName, vk::Extent2D(500, 500));

    std::pair<uint32_t, uint32_t> graphicsAndPresentQueueFamilyIndex = vk::su::findGraphicsAndPresentQueueFamilyIndex(physicalDevice, *surfaceData.surface);
    vk::UniqueDevice device = vk::su::createDevice(physicalDevice, graphicsAndPresentQueueFamilyIndex.first, vk::su::getDeviceExtensions());

    vk::UniqueCommandPool commandPool = vk::su::createCommandPool(device, graphicsAndPresentQueueFamilyIndex.first);
    vk::UniqueCommandBuffer commandBuffer = std::move(device->allocateCommandBuffersUnique(vk::CommandBufferAllocateInfo(commandPool.get(), vk::CommandBufferLevel::ePrimary, 1)).front());

    vk::Queue graphicsQueue = device->getQueue(graphicsAndPresentQueueFamilyIndex.first, 0);
    vk::Queue presentQueue = device->getQueue(graphicsAndPresentQueueFamilyIndex.second, 0);

    vk::su::SwapChainData swapChainData(physicalDevice, device, *surfaceData.surface, surfaceData.extent, vk::ImageUsageFlagBits::eColorAttachment | vk::ImageUsageFlagBits::eTransferSrc,
                                        vk::UniqueSwapchainKHR(), graphicsAndPresentQueueFamilyIndex.first, graphicsAndPresentQueueFamilyIndex.second);

    vk::su::DepthBufferData depthBufferData(physicalDevice, device, vk::Format::eD16Unorm, surfaceData.extent);

    vk::UniqueRenderPass renderPass = vk::su::createRenderPass(device, vk::su::pickSurfaceFormat(physicalDevice.getSurfaceFormatsKHR(surfaceData.surface.get())).format, depthBufferData.format);

    glslang::InitializeProcess();
    vk::UniqueShaderModule vertexShaderModule = vk::su::createShaderModule(device, vk::ShaderStageFlagBits::eVertex, vertexShaderText_PC_C);
    vk::UniqueShaderModule fragmentShaderModule = vk::su::createShaderModule(device, vk::ShaderStageFlagBits::eFragment, fragmentShaderText_C_C);
    glslang::FinalizeProcess();

    std::vector<vk::UniqueFramebuffer> framebuffers = vk::su::createFramebuffers(device, renderPass, swapChainData.imageViews, depthBufferData.imageView, surfaceData.extent);

    vk::su::BufferData vertexBufferData(physicalDevice, device, sizeof(coloredCubeData), vk::BufferUsageFlagBits::eVertexBuffer);
    vk::su::copyToDevice(device, vertexBufferData.deviceMemory, coloredCubeData, sizeof(coloredCubeData) / sizeof(coloredCubeData[0]));

    /* VULKAN_KEY_START */

    vk::PhysicalDeviceLimits limits = physicalDevice.getProperties().limits;
    if (limits.maxDescriptorSetUniformBuffersDynamic < 1)
    {
      std::cout << "No dynamic uniform buffers supported\n";
      exit(-1);
    }

    /* Set up uniform buffer with 2 transform matrices in it */
    glm::mat4x4 mvpcs[2];
    glm::mat4x4 model = glm::mat4x4(1.0f);
    glm::mat4x4 view = glm::lookAt(glm::vec3(0.0f, 3.0f, -10.0f), glm::vec3(0.0f, 0.0f, 0.0f), glm::vec3(0.0f, -1.0f, 0.0f));
    glm::mat4x4 projection = glm::perspective(glm::radians(45.0f), 1.0f, 0.1f, 100.0f);
    glm::mat4x4 clip = glm::mat4x4(1.0f, 0.0f, 0.0f, 0.0f, 0.0f, -1.0f, 0.0f, 0.0f, 0.0f, 0.0f, 0.5f, 0.0f, 0.0f, 0.0f, 0.5f, 1.0f);   // vulkan clip space has inverted y and half z !
    mvpcs[0] = clip * projection * view * model;

    model = glm::translate(model, glm::vec3(-1.5f, 1.5f, -1.5f));
    mvpcs[1] = clip * projection * view * model;

    VkDeviceSize bufferSize = sizeof(glm::mat4x4);
    if (limits.minUniformBufferOffsetAlignment)
    {
      bufferSize = (bufferSize + limits.minUniformBufferOffsetAlignment - 1) & ~(limits.minUniformBufferOffsetAlignment - 1);
    }

    vk::su::BufferData uniformBufferData(physicalDevice, device, 2 * bufferSize, vk::BufferUsageFlagBits::eUniformBuffer);
    vk::su::copyToDevice(device, uniformBufferData.deviceMemory, mvpcs, 2, bufferSize);

    // create a DescriptorSetLayout with vk::DescriptorType::eUniformBufferDynamic
    vk::UniqueDescriptorSetLayout descriptorSetLayout = vk::su::createDescriptorSetLayout(device, { {vk::DescriptorType::eUniformBufferDynamic, 1, vk::ShaderStageFlagBits::eVertex} });
    vk::UniquePipelineLayout pipelineLayout = device->createPipelineLayoutUnique(vk::PipelineLayoutCreateInfo(vk::PipelineLayoutCreateFlags(), 1, &descriptorSetLayout.get()));

    // create a DescriptorPool with vk::DescriptorType::eUniformBufferDynamic
    vk::UniqueDescriptorPool descriptorPool = vk::su::createDescriptorPool(device, { { vk::DescriptorType::eUniformBufferDynamic, 1 } });
    vk::UniqueDescriptorSet descriptorSet = std::move(device->allocateDescriptorSetsUnique(vk::DescriptorSetAllocateInfo(*descriptorPool, 1, &*descriptorSetLayout)).front());

    vk::su::updateDescriptorSets(device, descriptorSet, {{vk::DescriptorType::eUniformBufferDynamic, uniformBufferData.buffer, vk::UniqueBufferView()}}, {});

    vk::UniquePipelineCache pipelineCache = device->createPipelineCacheUnique(vk::PipelineCacheCreateInfo());
    vk::UniquePipeline graphicsPipeline = vk::su::createGraphicsPipeline(device, pipelineCache, std::make_pair(*vertexShaderModule, nullptr), std::make_pair(*fragmentShaderModule, nullptr),
                                                                         sizeof(coloredCubeData[0]), { { vk::Format::eR32G32B32A32Sfloat, 0 }, { vk::Format::eR32G32B32A32Sfloat, 16 } },
                                                                         vk::FrontFace::eClockwise, true, pipelineLayout, renderPass);
    // Get the index of the next available swapchain image:
    vk::UniqueSemaphore imageAcquiredSemaphore = device->createSemaphoreUnique(vk::SemaphoreCreateInfo());
    vk::ResultValue<uint32_t> currentBuffer = device->acquireNextImageKHR(swapChainData.swapChain.get(), vk::su::FenceTimeout, imageAcquiredSemaphore.get(), nullptr);
    assert(currentBuffer.result == vk::Result::eSuccess);
    assert(currentBuffer.value < framebuffers.size());

    commandBuffer->begin(vk::CommandBufferBeginInfo(vk::CommandBufferUsageFlags()));

    vk::ClearValue clearValues[2];
    clearValues[0].color = vk::ClearColorValue(std::array<float, 4>({ 0.2f, 0.2f, 0.2f, 0.2f }));
    clearValues[1].depthStencil = vk::ClearDepthStencilValue(1.0f, 0);
    vk::RenderPassBeginInfo renderPassBeginInfo(renderPass.get(), framebuffers[currentBuffer.value].get(), vk::Rect2D(vk::Offset2D(0, 0), surfaceData.extent), 2, clearValues);
    commandBuffer->beginRenderPass(renderPassBeginInfo, vk::SubpassContents::eInline);
    commandBuffer->bindPipeline(vk::PipelineBindPoint::eGraphics, graphicsPipeline.get());

    commandBuffer->setViewport(0, vk::Viewport(0.0f, 0.0f, static_cast<float>(surfaceData.extent.width), static_cast<float>(surfaceData.extent.height), 0.0f, 1.0f));
    commandBuffer->setScissor(0, vk::Rect2D(vk::Offset2D(0, 0), surfaceData.extent));

    /* The first draw should use the first matrix in the buffer */
    uint32_t dynamicOffset = 0;
    commandBuffer->bindDescriptorSets(vk::PipelineBindPoint::eGraphics, pipelineLayout.get(), 0, descriptorSet.get(), dynamicOffset);

    commandBuffer->bindVertexBuffers(0, *vertexBufferData.buffer, {0});
    commandBuffer->draw(12 * 3, 1, 0, 0);

    // the second draw should use the second matrix in the buffer;
    dynamicOffset = (uint32_t)bufferSize;
    commandBuffer->bindDescriptorSets(vk::PipelineBindPoint::eGraphics, pipelineLayout.get(), 0, descriptorSet.get(), dynamicOffset);
    commandBuffer->draw(12 * 3, 1, 0, 0);

    commandBuffer->endRenderPass();
    commandBuffer->end();

    vk::UniqueFence drawFence = device->createFenceUnique(vk::FenceCreateInfo());

    vk::PipelineStageFlags waitDestinationStageMask(vk::PipelineStageFlagBits::eColorAttachmentOutput);
    vk::SubmitInfo submitInfo(1, &imageAcquiredSemaphore.get(), &waitDestinationStageMask, 1, &commandBuffer.get());
    graphicsQueue.submit(submitInfo, drawFence.get());

    while (vk::Result::eTimeout == device->waitForFences(drawFence.get(), VK_TRUE, vk::su::FenceTimeout))
      ;

    presentQueue.presentKHR(vk::PresentInfoKHR(0, nullptr, 1, &swapChainData.swapChain.get(), &currentBuffer.value));
    std::this_thread::sleep_for(std::chrono::milliseconds(1000));

    /* VULKAN_KEY_END */

    device->waitIdle();
  }
  catch (vk::SystemError& err)
  {
    std::cout << "vk::SystemError: " << err.what() << std::endl;
    exit(-1);
  }
  catch (std::runtime_error& err)
  {
    std::cout << "std::runtime_error: " << err.what() << std::endl;
    exit(-1);
  }
  catch (...)
  {
    std::cout << "unknown error\n";
    exit(-1);
  }
  return 0;
}