/* * Copyright (c) 2015-2016 The Khronos Group Inc. * Copyright (c) 2015-2016 Valve Corporation * Copyright (c) 2015-2016 LunarG, Inc. * * 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. * * Author: Jeremy Hayes */ #if defined(VK_USE_PLATFORM_XLIB_KHR) || defined(VK_USE_PLATFORM_XCB_KHR) #include #endif #include #include #include #include #include #include #define VULKAN_HPP_NO_EXCEPTIONS #include #include #include "linmath.h" #ifndef NDEBUG #define VERIFY(x) assert(x) #else #define VERIFY(x) ((void)(x)) #endif #define APP_SHORT_NAME "cube" #ifdef _WIN32 #define APP_NAME_STR_LEN 80 #endif // Allow a maximum of two outstanding presentation operations. #define FRAME_LAG 2 #define ARRAY_SIZE(a) (sizeof(a) / sizeof(a[0])) #ifdef _WIN32 #define ERR_EXIT(err_msg, err_class) \ do { \ if (!suppress_popups) \ MessageBox(nullptr, err_msg, err_class, MB_OK); \ exit(1); \ } while (0) #else #define ERR_EXIT(err_msg, err_class) \ do { \ printf(err_msg); \ fflush(stdout); \ exit(1); \ } while (0) #endif struct texture_object { vk::Sampler sampler; vk::Image image; vk::ImageLayout imageLayout; vk::MemoryAllocateInfo mem_alloc; vk::DeviceMemory mem; vk::ImageView view; int32_t tex_width; int32_t tex_height; }; static char const *const tex_files[] = {"lunarg.ppm"}; static int validation_error = 0; struct vkcube_vs_uniform { // Must start with MVP float mvp[4][4]; float position[12 * 3][4]; float color[12 * 3][4]; }; struct vktexcube_vs_uniform { // Must start with MVP float mvp[4][4]; float position[12 * 3][4]; float attr[12 * 3][4]; }; //-------------------------------------------------------------------------------------- // Mesh and VertexFormat Data //-------------------------------------------------------------------------------------- // clang-format off static const float g_vertex_buffer_data[] = { -1.0f,-1.0f,-1.0f, // -X side -1.0f,-1.0f, 1.0f, -1.0f, 1.0f, 1.0f, -1.0f, 1.0f, 1.0f, -1.0f, 1.0f,-1.0f, -1.0f,-1.0f,-1.0f, -1.0f,-1.0f,-1.0f, // -Z side 1.0f, 1.0f,-1.0f, 1.0f,-1.0f,-1.0f, -1.0f,-1.0f,-1.0f, -1.0f, 1.0f,-1.0f, 1.0f, 1.0f,-1.0f, -1.0f,-1.0f,-1.0f, // -Y side 1.0f,-1.0f,-1.0f, 1.0f,-1.0f, 1.0f, -1.0f,-1.0f,-1.0f, 1.0f,-1.0f, 1.0f, -1.0f,-1.0f, 1.0f, -1.0f, 1.0f,-1.0f, // +Y side -1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f, -1.0f, 1.0f,-1.0f, 1.0f, 1.0f, 1.0f, 1.0f, 1.0f,-1.0f, 1.0f, 1.0f,-1.0f, // +X side 1.0f, 1.0f, 1.0f, 1.0f,-1.0f, 1.0f, 1.0f,-1.0f, 1.0f, 1.0f,-1.0f,-1.0f, 1.0f, 1.0f,-1.0f, -1.0f, 1.0f, 1.0f, // +Z side -1.0f,-1.0f, 1.0f, 1.0f, 1.0f, 1.0f, -1.0f,-1.0f, 1.0f, 1.0f,-1.0f, 1.0f, 1.0f, 1.0f, 1.0f, }; static const float g_uv_buffer_data[] = { 0.0f, 1.0f, // -X side 1.0f, 1.0f, 1.0f, 0.0f, 1.0f, 0.0f, 0.0f, 0.0f, 0.0f, 1.0f, 1.0f, 1.0f, // -Z side 0.0f, 0.0f, 0.0f, 1.0f, 1.0f, 1.0f, 1.0f, 0.0f, 0.0f, 0.0f, 1.0f, 0.0f, // -Y side 1.0f, 1.0f, 0.0f, 1.0f, 1.0f, 0.0f, 0.0f, 1.0f, 0.0f, 0.0f, 1.0f, 0.0f, // +Y side 0.0f, 0.0f, 0.0f, 1.0f, 1.0f, 0.0f, 0.0f, 1.0f, 1.0f, 1.0f, 1.0f, 0.0f, // +X side 0.0f, 0.0f, 0.0f, 1.0f, 0.0f, 1.0f, 1.0f, 1.0f, 1.0f, 0.0f, 0.0f, 0.0f, // +Z side 0.0f, 1.0f, 1.0f, 0.0f, 0.0f, 1.0f, 1.0f, 1.0f, 1.0f, 0.0f, }; // clang-format on typedef struct { vk::Image image; vk::CommandBuffer cmd; vk::CommandBuffer graphics_to_present_cmd; vk::ImageView view; } SwapchainBuffers; #ifdef _WIN32 // MS-Windows event handling function: LRESULT CALLBACK WndProc(HWND hWnd, UINT uMsg, WPARAM wParam, LPARAM lParam); #endif struct Demo { Demo() : #if defined(VK_USE_PLATFORM_WIN32_KHR) connection{nullptr}, window{nullptr}, minsize(POINT{ 0, 0}), // Use explicit construction to avoid MSVC error C2797. #elif defined(VK_USE_PLATFORM_XLIB_KHR) || defined(VK_USE_PLATFORM_XCB_KHR) display{nullptr}, xlib_window{0}, xlib_wm_delete_window{0}, connection{nullptr}, screen{nullptr}, xcb_window{0}, atom_wm_delete_window{nullptr}, #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) display{nullptr}, registry{nullptr}, compositor{nullptr}, window{nullptr}, shell{nullptr}, shell_surface{nullptr}, #endif prepared{false}, use_staging_buffer{false}, use_xlib{false}, graphics_queue_family_index{0}, present_queue_family_index{0}, enabled_extension_count{0}, enabled_layer_count{0}, width{0}, height{0}, swapchainImageCount{0}, frame_index{0}, spin_angle{0.0f}, spin_increment{0.0f}, pause{false}, quit{false}, curFrame{0}, frameCount{0}, validate{false}, use_break{false}, suppress_popups{false}, current_buffer{0}, queue_family_count{0} { #if defined(VK_USE_PLATFORM_WIN32_KHR) memset(name, '\0', APP_NAME_STR_LEN); #elif defined(VK_USE_PLATFORM_XLIB_KHR) || defined(VK_USE_PLATFORM_XCB_KHR) #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) #endif memset(projection_matrix, 0, sizeof(projection_matrix)); memset(view_matrix, 0, sizeof(view_matrix)); memset(model_matrix, 0, sizeof(model_matrix)); } void build_image_ownership_cmd(uint32_t const &i) { auto const cmd_buf_info = vk::CommandBufferBeginInfo().setFlags( vk::CommandBufferUsageFlagBits::eSimultaneousUse); auto result = buffers[i].graphics_to_present_cmd.begin(&cmd_buf_info); VERIFY(result == vk::Result::eSuccess); auto const image_ownership_barrier = vk::ImageMemoryBarrier() .setSrcAccessMask(vk::AccessFlags()) .setDstAccessMask(vk::AccessFlagBits::eColorAttachmentWrite) .setOldLayout(vk::ImageLayout::ePresentSrcKHR) .setNewLayout(vk::ImageLayout::ePresentSrcKHR) .setSrcQueueFamilyIndex(graphics_queue_family_index) .setDstQueueFamilyIndex(present_queue_family_index) .setImage(buffers[i].image) .setSubresourceRange(vk::ImageSubresourceRange( vk::ImageAspectFlagBits::eColor, 0, 1, 0, 1)); buffers[i].graphics_to_present_cmd.pipelineBarrier( vk::PipelineStageFlagBits::eColorAttachmentOutput, vk::PipelineStageFlagBits::eColorAttachmentOutput, vk::DependencyFlagBits(), 0, nullptr, 0, nullptr, 1, &image_ownership_barrier); result = buffers[i].graphics_to_present_cmd.end(); VERIFY(result == vk::Result::eSuccess); } vk::Bool32 check_layers(uint32_t check_count, char const *const *const check_names, uint32_t layer_count, vk::LayerProperties *layers) { for (uint32_t i = 0; i < check_count; i++) { vk::Bool32 found = VK_FALSE; for (uint32_t j = 0; j < layer_count; j++) { if (!strcmp(check_names[i], layers[j].layerName)) { found = VK_TRUE; break; } } if (!found) { fprintf(stderr, "Cannot find layer: %s\n", check_names[i]); return 0; } } return VK_TRUE; } void cleanup() { prepared = false; device.waitIdle(); // Wait for fences from present operations for (uint32_t i = 0; i < FRAME_LAG; i++) { device.waitForFences(1, &fences[i], VK_TRUE, UINT64_MAX); device.destroyFence(fences[i], nullptr); device.destroySemaphore(image_acquired_semaphores[i], nullptr); device.destroySemaphore(draw_complete_semaphores[i], nullptr); if (separate_present_queue) { device.destroySemaphore(image_ownership_semaphores[i], nullptr); } } for (uint32_t i = 0; i < swapchainImageCount; i++) { device.destroyFramebuffer(framebuffers[i], nullptr); } device.destroyDescriptorPool(desc_pool, nullptr); device.destroyPipeline(pipeline, nullptr); device.destroyPipelineCache(pipelineCache, nullptr); device.destroyRenderPass(render_pass, nullptr); device.destroyPipelineLayout(pipeline_layout, nullptr); device.destroyDescriptorSetLayout(desc_layout, nullptr); for (uint32_t i = 0; i < texture_count; i++) { device.destroyImageView(textures[i].view, nullptr); device.destroyImage(textures[i].image, nullptr); device.freeMemory(textures[i].mem, nullptr); device.destroySampler(textures[i].sampler, nullptr); } device.destroySwapchainKHR(swapchain, nullptr); device.destroyImageView(depth.view, nullptr); device.destroyImage(depth.image, nullptr); device.freeMemory(depth.mem, nullptr); device.destroyBuffer(uniform_data.buf, nullptr); device.freeMemory(uniform_data.mem, nullptr); for (uint32_t i = 0; i < swapchainImageCount; i++) { device.destroyImageView(buffers[i].view, nullptr); device.freeCommandBuffers(cmd_pool, 1, &buffers[i].cmd); } device.destroyCommandPool(cmd_pool, nullptr); if (separate_present_queue) { device.destroyCommandPool(present_cmd_pool, nullptr); } device.destroy(nullptr); inst.destroySurfaceKHR(surface, nullptr); inst.destroy(nullptr); #if defined(VK_USE_PLATFORM_XLIB_KHR) || defined(VK_USE_PLATFORM_XCB_KHR) if (use_xlib) { XDestroyWindow(display, xlib_window); XCloseDisplay(display); } else { xcb_destroy_window(connection, xcb_window); xcb_disconnect(connection); } free(atom_wm_delete_window); #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) wl_shell_surface_destroy(shell_surface); wl_surface_destroy(window); wl_shell_destroy(shell); wl_compositor_destroy(compositor); wl_registry_destroy(registry); wl_display_disconnect(display); #endif } void create_device() { float const priorities[1] = {0.0}; vk::DeviceQueueCreateInfo queues[2]; queues[0].setQueueFamilyIndex(graphics_queue_family_index); queues[0].setQueueCount(1); queues[0].setPQueuePriorities(priorities); auto deviceInfo = vk::DeviceCreateInfo() .setQueueCreateInfoCount(1) .setPQueueCreateInfos(queues) .setEnabledLayerCount(0) .setPpEnabledLayerNames(nullptr) .setEnabledExtensionCount(enabled_extension_count) .setPpEnabledExtensionNames( (const char *const *)extension_names) .setPEnabledFeatures(nullptr); if (separate_present_queue) { queues[1].setQueueFamilyIndex(present_queue_family_index); queues[1].setQueueCount(1); queues[1].setPQueuePriorities(priorities); deviceInfo.setQueueCreateInfoCount(2); } auto result = gpu.createDevice(&deviceInfo, nullptr, &device); VERIFY(result == vk::Result::eSuccess); } void destroy_texture_image(texture_object *tex_objs) { // clean up staging resources device.freeMemory(tex_objs->mem, nullptr); device.destroyImage(tex_objs->image, nullptr); } void draw() { // Ensure no more than FRAME_LAG presentations are outstanding device.waitForFences(1, &fences[frame_index], VK_TRUE, UINT64_MAX); device.resetFences(1, &fences[frame_index]); // Get the index of the next available swapchain image: auto result = device.acquireNextImageKHR( swapchain, UINT64_MAX, image_acquired_semaphores[frame_index], fences[frame_index], ¤t_buffer); if (result == vk::Result::eErrorOutOfDateKHR) { // swapchain is out of date (e.g. the window was resized) and // must be recreated: frame_index += 1; frame_index %= FRAME_LAG; resize(); draw(); return; } else if (result == vk::Result::eSuboptimalKHR) { // swapchain is not as optimal as it could be, but the platform's // presentation engine will still present the image correctly. } else { VERIFY(result == vk::Result::eSuccess); } // Wait for the image acquired semaphore to be signaled to ensure // that the image won't be rendered to until the presentation // engine has fully released ownership to the application, and it is // okay to render to the image. vk::PipelineStageFlags const pipe_stage_flags = vk::PipelineStageFlagBits::eColorAttachmentOutput; auto const submit_info = vk::SubmitInfo() .setPWaitDstStageMask(&pipe_stage_flags) .setWaitSemaphoreCount(1) .setPWaitSemaphores(&image_acquired_semaphores[frame_index]) .setCommandBufferCount(1) .setPCommandBuffers(&buffers[current_buffer].cmd) .setSignalSemaphoreCount(1) .setPSignalSemaphores(&draw_complete_semaphores[frame_index]); result = graphics_queue.submit(1, &submit_info, vk::Fence()); VERIFY(result == vk::Result::eSuccess); if (separate_present_queue) { // If we are using separate queues, change image ownership to the // present queue before presenting, waiting for the draw complete // semaphore and signalling the ownership released semaphore when // finished auto const submit_info = vk::SubmitInfo() .setPWaitDstStageMask(&pipe_stage_flags) .setWaitSemaphoreCount(1) .setPWaitSemaphores(&draw_complete_semaphores[frame_index]) .setCommandBufferCount(1) .setPCommandBuffers( &buffers[current_buffer].graphics_to_present_cmd) .setSignalSemaphoreCount(1) .setPSignalSemaphores( &image_ownership_semaphores[frame_index]); result = present_queue.submit(1, &submit_info, vk::Fence()); VERIFY(result == vk::Result::eSuccess); } // If we are using separate queues we have to wait for image ownership, // otherwise wait for draw complete auto const presentInfo = vk::PresentInfoKHR() .setWaitSemaphoreCount(1) .setPWaitSemaphores( separate_present_queue ? &image_ownership_semaphores[frame_index] : &draw_complete_semaphores[frame_index]) .setSwapchainCount(1) .setPSwapchains(&swapchain) .setPImageIndices(¤t_buffer); result = present_queue.presentKHR(&presentInfo); frame_index += 1; frame_index %= FRAME_LAG; if (result == vk::Result::eErrorOutOfDateKHR) { // swapchain is out of date (e.g. the window was resized) and // must be recreated: resize(); } else if (result == vk::Result::eSuboptimalKHR) { // swapchain is not as optimal as it could be, but the platform's // presentation engine will still present the image correctly. } else { VERIFY(result == vk::Result::eSuccess); } } void draw_build_cmd(vk::CommandBuffer commandBuffer) { auto const commandInfo = vk::CommandBufferBeginInfo().setFlags( vk::CommandBufferUsageFlagBits::eSimultaneousUse); vk::ClearValue const clearValues[2] = { vk::ClearColorValue(std::array({0.2f, 0.2f, 0.2f, 0.2f})), vk::ClearDepthStencilValue(1.0f, 0u)}; auto const passInfo = vk::RenderPassBeginInfo() .setRenderPass(render_pass) .setFramebuffer(framebuffers[current_buffer]) .setRenderArea( vk::Rect2D(vk::Offset2D(0, 0), vk::Extent2D((uint32_t)width, (uint32_t)height))) .setClearValueCount(2) .setPClearValues(clearValues); auto result = commandBuffer.begin(&commandInfo); VERIFY(result == vk::Result::eSuccess); commandBuffer.beginRenderPass(&passInfo, vk::SubpassContents::eInline); commandBuffer.bindPipeline(vk::PipelineBindPoint::eGraphics, pipeline); commandBuffer.bindDescriptorSets(vk::PipelineBindPoint::eGraphics, pipeline_layout, 0, 1, &desc_set, 0, nullptr); auto const viewport = vk::Viewport() .setWidth((float)width) .setHeight((float)height) .setMinDepth((float)0.0f) .setMaxDepth((float)1.0f); commandBuffer.setViewport(0, 1, &viewport); vk::Rect2D const scissor(vk::Offset2D(0, 0), vk::Extent2D(width, height)); commandBuffer.setScissor(0, 1, &scissor); commandBuffer.draw(12 * 3, 1, 0, 0); // Note that ending the renderpass changes the image's layout from // COLOR_ATTACHMENT_OPTIMAL to PRESENT_SRC_KHR commandBuffer.endRenderPass(); if (separate_present_queue) { // We have to transfer ownership from the graphics queue family to // the // present queue family to be able to present. Note that we don't // have // to transfer from present queue family back to graphics queue // family at // the start of the next frame because we don't care about the // image's // contents at that point. auto const image_ownership_barrier = vk::ImageMemoryBarrier() .setSrcAccessMask(vk::AccessFlags()) .setDstAccessMask(vk::AccessFlagBits::eColorAttachmentWrite) .setOldLayout(vk::ImageLayout::ePresentSrcKHR) .setNewLayout(vk::ImageLayout::ePresentSrcKHR) .setSrcQueueFamilyIndex(graphics_queue_family_index) .setDstQueueFamilyIndex(present_queue_family_index) .setImage(buffers[current_buffer].image) .setSubresourceRange(vk::ImageSubresourceRange( vk::ImageAspectFlagBits::eColor, 0, 1, 0, 1)); commandBuffer.pipelineBarrier( vk::PipelineStageFlagBits::eColorAttachmentOutput, vk::PipelineStageFlagBits::eBottomOfPipe, vk::DependencyFlagBits(), 0, nullptr, 0, nullptr, 1, &image_ownership_barrier); } result = commandBuffer.end(); VERIFY(result == vk::Result::eSuccess); } void flush_init_cmd() { // TODO: hmm. // This function could get called twice if the texture uses a staging // buffer // In that case the second call should be ignored if (!cmd) { return; } auto result = cmd.end(); VERIFY(result == vk::Result::eSuccess); auto const fenceInfo = vk::FenceCreateInfo().setFlags(vk::FenceCreateFlagBits(0)); vk::Fence fence; device.createFence(&fenceInfo, nullptr, &fence); vk::CommandBuffer const commandBuffers[] = {cmd}; auto const submitInfo = vk::SubmitInfo().setCommandBufferCount(1).setPCommandBuffers( commandBuffers); result = graphics_queue.submit(1, &submitInfo, fence); VERIFY(result == vk::Result::eSuccess); result = device.waitForFences(1, &fence, VK_TRUE, UINT64_MAX); VERIFY(result == vk::Result::eSuccess); device.freeCommandBuffers(cmd_pool, 1, commandBuffers); device.destroyFence(fence, nullptr); cmd = vk::CommandBuffer(); } void init(int argc, char **argv) { vec3 eye = {0.0f, 3.0f, 5.0f}; vec3 origin = {0, 0, 0}; vec3 up = {0.0f, 1.0f, 0.0}; frameCount = UINT32_MAX; use_xlib = false; for (int i = 1; i < argc; i++) { if (strcmp(argv[i], "--use_staging") == 0) { use_staging_buffer = true; continue; } if (strcmp(argv[i], "--break") == 0) { use_break = true; continue; } if (strcmp(argv[i], "--validate") == 0) { validate = true; continue; } #if defined(VK_USE_PLATFORM_XLIB_KHR) if (strcmp(argv[i], "--xlib") == 0) { use_xlib = true; continue; } #endif if (strcmp(argv[i], "--c") == 0 && frameCount == UINT32_MAX && i < argc - 1 && sscanf(argv[i + 1], "%d", &frameCount) == 1) { i++; continue; } if (strcmp(argv[i], "--suppress_popups") == 0) { suppress_popups = true; continue; } fprintf(stderr, "Usage:\n %s [--use_staging] [--validate] [--break] " #if defined(VK_USE_PLATFORM_XLIB_KHR) "[--xlib] " #endif "[--c ] [--suppress_popups]\n", APP_SHORT_NAME); fflush(stderr); exit(1); } if (!use_xlib) { init_connection(); } init_vk(); width = 500; height = 500; spin_angle = 4.0f; spin_increment = 0.2f; pause = false; mat4x4_perspective(projection_matrix, (float)degreesToRadians(45.0f), 1.0f, 0.1f, 100.0f); mat4x4_look_at(view_matrix, eye, origin, up); mat4x4_identity(model_matrix); projection_matrix[1][1] *= -1; // Flip projection matrix from GL to Vulkan orientation. } void init_connection() { #if defined(VK_USE_PLATFORM_XCB_KHR) const xcb_setup_t *setup; xcb_screen_iterator_t iter; int scr; connection = xcb_connect(nullptr, &scr); if (xcb_connection_has_error(connection) > 0) { printf("Cannot find a compatible Vulkan installable client driver " "(ICD).\nExiting ...\n"); fflush(stdout); exit(1); } setup = xcb_get_setup(connection); iter = xcb_setup_roots_iterator(setup); while (scr-- > 0) xcb_screen_next(&iter); screen = iter.data; #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) display = wl_display_connect(nullptr); if (display == nullptr) { printf("Cannot find a compatible Vulkan installable client driver " "(ICD).\nExiting ...\n"); fflush(stdout); exit(1); } registry = wl_display_get_registry(display); wl_registry_add_listener(registry, ®istry_listener, this); wl_display_dispatch(display); #endif } void init_vk() { uint32_t instance_extension_count = 0; uint32_t instance_layer_count = 0; uint32_t validation_layer_count = 0; char const *const *instance_validation_layers = nullptr; enabled_extension_count = 0; enabled_layer_count = 0; char const *const instance_validation_layers_alt1[] = { "VK_LAYER_LUNARG_standard_validation"}; char const *const instance_validation_layers_alt2[] = { "VK_LAYER_GOOGLE_threading", "VK_LAYER_LUNARG_parameter_validation", "VK_LAYER_LUNARG_object_tracker", "VK_LAYER_LUNARG_image", "VK_LAYER_LUNARG_core_validation", "VK_LAYER_LUNARG_swapchain", "VK_LAYER_GOOGLE_unique_objects"}; // Look for validation layers vk::Bool32 validation_found = VK_FALSE; if (validate) { auto result = vk::enumerateInstanceLayerProperties( &instance_layer_count, nullptr); VERIFY(result == vk::Result::eSuccess); instance_validation_layers = instance_validation_layers_alt1; if (instance_layer_count > 0) { std::unique_ptr instance_layers( new vk::LayerProperties[instance_layer_count]); result = vk::enumerateInstanceLayerProperties( &instance_layer_count, instance_layers.get()); VERIFY(result == vk::Result::eSuccess); validation_found = check_layers(ARRAY_SIZE(instance_validation_layers_alt1), instance_validation_layers, instance_layer_count, instance_layers.get()); if (validation_found) { enabled_layer_count = ARRAY_SIZE(instance_validation_layers_alt1); enabled_layers[0] = "VK_LAYER_LUNARG_standard_validation"; validation_layer_count = 1; } else { // use alternative set of validation layers instance_validation_layers = instance_validation_layers_alt2; enabled_layer_count = ARRAY_SIZE(instance_validation_layers_alt2); validation_found = check_layers( ARRAY_SIZE(instance_validation_layers_alt2), instance_validation_layers, instance_layer_count, instance_layers.get()); validation_layer_count = ARRAY_SIZE(instance_validation_layers_alt2); for (uint32_t i = 0; i < validation_layer_count; i++) { enabled_layers[i] = instance_validation_layers[i]; } } } if (!validation_found) { ERR_EXIT("vkEnumerateInstanceLayerProperties failed to find " "required validation layer.\n\n" "Please look at the Getting Started guide for " "additional information.\n", "vkCreateInstance Failure"); } } /* Look for instance extensions */ vk::Bool32 surfaceExtFound = VK_FALSE; vk::Bool32 platformSurfaceExtFound = VK_FALSE; #if defined(VK_USE_PLATFORM_XLIB_KHR) vk::Bool32 xlibSurfaceExtFound = VK_FALSE; #endif memset(extension_names, 0, sizeof(extension_names)); auto result = vk::enumerateInstanceExtensionProperties( nullptr, &instance_extension_count, nullptr); VERIFY(result == vk::Result::eSuccess); if (instance_extension_count > 0) { std::unique_ptr instance_extensions( new vk::ExtensionProperties[instance_extension_count]); result = vk::enumerateInstanceExtensionProperties( nullptr, &instance_extension_count, instance_extensions.get()); VERIFY(result == vk::Result::eSuccess); for (uint32_t i = 0; i < instance_extension_count; i++) { if (!strcmp(VK_KHR_SURFACE_EXTENSION_NAME, instance_extensions[i].extensionName)) { surfaceExtFound = 1; extension_names[enabled_extension_count++] = VK_KHR_SURFACE_EXTENSION_NAME; } #if defined(VK_USE_PLATFORM_WIN32_KHR) if (!strcmp(VK_KHR_WIN32_SURFACE_EXTENSION_NAME, instance_extensions[i].extensionName)) { platformSurfaceExtFound = 1; extension_names[enabled_extension_count++] = VK_KHR_WIN32_SURFACE_EXTENSION_NAME; } #endif #if defined(VK_USE_PLATFORM_XLIB_KHR) if (!strcmp(VK_KHR_XLIB_SURFACE_EXTENSION_NAME, instance_extensions[i].extensionName)) { platformSurfaceExtFound = 1; xlibSurfaceExtFound = 1; extension_names[enabled_extension_count++] = VK_KHR_XLIB_SURFACE_EXTENSION_NAME; } #endif #if defined(VK_USE_PLATFORM_XCB_KHR) if (!strcmp(VK_KHR_XCB_SURFACE_EXTENSION_NAME, instance_extensions[i].extensionName)) { platformSurfaceExtFound = 1; extension_names[enabled_extension_count++] = VK_KHR_XCB_SURFACE_EXTENSION_NAME; } #endif #if defined(VK_USE_PLATFORM_WAYLAND_KHR) if (!strcmp(VK_KHR_WAYLAND_SURFACE_EXTENSION_NAME, instance_extensions[i].extensionName)) { platformSurfaceExtFound = 1; extension_names[enabled_extension_count++] = VK_KHR_WAYLAND_SURFACE_EXTENSION_NAME; } #endif assert(enabled_extension_count < 64); } } if (!surfaceExtFound) { ERR_EXIT("vkEnumerateInstanceExtensionProperties failed to find " "the " VK_KHR_SURFACE_EXTENSION_NAME " extension.\n\n" "Do you have a compatible Vulkan installable client " "driver (ICD) installed?\n" "Please look at the Getting Started guide for additional " "information.\n", "vkCreateInstance Failure"); } if (!platformSurfaceExtFound) { #if defined(VK_USE_PLATFORM_WIN32_KHR) ERR_EXIT("vkEnumerateInstanceExtensionProperties failed to find " "the " VK_KHR_WIN32_SURFACE_EXTENSION_NAME " extension.\n\n" "Do you have a compatible Vulkan installable client " "driver (ICD) installed?\n" "Please look at the Getting Started guide for additional " "information.\n", "vkCreateInstance Failure"); #elif defined(VK_USE_PLATFORM_XCB_KHR) ERR_EXIT("vkEnumerateInstanceExtensionProperties failed to find " "the " VK_KHR_XCB_SURFACE_EXTENSION_NAME " extension.\n\n" "Do you have a compatible Vulkan installable client " "driver (ICD) installed?\n" "Please look at the Getting Started guide for additional " "information.\n", "vkCreateInstance Failure"); #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) ERR_EXIT("vkEnumerateInstanceExtensionProperties failed to find " "the " VK_KHR_WAYLAND_SURFACE_EXTENSION_NAME " extension.\n\n" "Do you have a compatible Vulkan installable client " "driver (ICD) installed?\n" "Please look at the Getting Started guide for additional " "information.\n", "vkCreateInstance Failure"); #endif } #if defined(VK_USE_PLATFORM_XLIB_KHR) if (use_xlib && !xlibSurfaceExtFound) { ERR_EXIT("vkEnumerateInstanceExtensionProperties failed to find " "the " VK_KHR_XLIB_SURFACE_EXTENSION_NAME " extension.\n\n" "Do you have a compatible Vulkan installable client " "driver (ICD) installed?\n" "Please look at the Getting Started guide for additional " "information.\n", "vkCreateInstance Failure"); } #endif auto const app = vk::ApplicationInfo() .setPApplicationName(APP_SHORT_NAME) .setApplicationVersion(0) .setPEngineName(APP_SHORT_NAME) .setEngineVersion(0) .setApiVersion(VK_API_VERSION_1_0); auto const inst_info = vk::InstanceCreateInfo() .setPApplicationInfo(&app) .setEnabledLayerCount(enabled_layer_count) .setPpEnabledLayerNames(instance_validation_layers) .setEnabledExtensionCount(enabled_extension_count) .setPpEnabledExtensionNames(extension_names); result = vk::createInstance(&inst_info, nullptr, &inst); if (result == vk::Result::eErrorIncompatibleDriver) { ERR_EXIT("Cannot find a compatible Vulkan installable client " "driver (ICD).\n\n" "Please look at the Getting Started guide for additional " "information.\n", "vkCreateInstance Failure"); } else if (result == vk::Result::eErrorExtensionNotPresent) { ERR_EXIT("Cannot find a specified extension library.\n" "Make sure your layers path is set appropriately.\n", "vkCreateInstance Failure"); } else if (result != vk::Result::eSuccess) { ERR_EXIT("vkCreateInstance failed.\n\n" "Do you have a compatible Vulkan installable client " "driver (ICD) installed?\n" "Please look at the Getting Started guide for additional " "information.\n", "vkCreateInstance Failure"); } /* Make initial call to query gpu_count, then second call for gpu info*/ uint32_t gpu_count; result = inst.enumeratePhysicalDevices(&gpu_count, nullptr); VERIFY(result == vk::Result::eSuccess); assert(gpu_count > 0); if (gpu_count > 0) { std::unique_ptr physical_devices( new vk::PhysicalDevice[gpu_count]); result = inst.enumeratePhysicalDevices(&gpu_count, physical_devices.get()); VERIFY(result == vk::Result::eSuccess); /* For cube demo we just grab the first physical device */ gpu = physical_devices[0]; } else { ERR_EXIT("vkEnumeratePhysicalDevices reported zero accessible " "devices.\n\n" "Do you have a compatible Vulkan installable client " "driver (ICD) installed?\n" "Please look at the Getting Started guide for additional " "information.\n", "vkEnumeratePhysicalDevices Failure"); } /* Look for device extensions */ uint32_t device_extension_count = 0; vk::Bool32 swapchainExtFound = VK_FALSE; enabled_extension_count = 0; memset(extension_names, 0, sizeof(extension_names)); result = gpu.enumerateDeviceExtensionProperties( nullptr, &device_extension_count, nullptr); VERIFY(result == vk::Result::eSuccess); if (device_extension_count > 0) { std::unique_ptr device_extensions( new vk::ExtensionProperties[device_extension_count]); result = gpu.enumerateDeviceExtensionProperties( nullptr, &device_extension_count, device_extensions.get()); VERIFY(result == vk::Result::eSuccess); for (uint32_t i = 0; i < device_extension_count; i++) { if (!strcmp(VK_KHR_SWAPCHAIN_EXTENSION_NAME, device_extensions[i].extensionName)) { swapchainExtFound = 1; extension_names[enabled_extension_count++] = VK_KHR_SWAPCHAIN_EXTENSION_NAME; } assert(enabled_extension_count < 64); } } if (!swapchainExtFound) { ERR_EXIT("vkEnumerateDeviceExtensionProperties failed to find " "the " VK_KHR_SWAPCHAIN_EXTENSION_NAME " extension.\n\n" "Do you have a compatible Vulkan installable client " "driver (ICD) installed?\n" "Please look at the Getting Started guide for additional " "information.\n", "vkCreateInstance Failure"); } gpu.getProperties(&gpu_props); /* Call with nullptr data to get count */ gpu.getQueueFamilyProperties(&queue_family_count, nullptr); assert(queue_family_count >= 1); queue_props.reset(new vk::QueueFamilyProperties[queue_family_count]); gpu.getQueueFamilyProperties(&queue_family_count, queue_props.get()); // Query fine-grained feature support for this device. // If app has specific feature requirements it should check supported // features based on this query vk::PhysicalDeviceFeatures physDevFeatures; gpu.getFeatures(&physDevFeatures); } void init_vk_swapchain() { // Create a WSI surface for the window: #if defined(VK_USE_PLATFORM_WIN32_KHR) { auto const createInfo = vk::Win32SurfaceCreateInfoKHR() .setHinstance(connection) .setHwnd(window); auto result = inst.createWin32SurfaceKHR(&createInfo, nullptr, &surface); VERIFY(result == vk::Result::eSuccess); } #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) && !defined(VK_USE_PLATFORM_XCB_KHR) { auto const createInfo = vk::WaylandSurfaceCreateInfoKHR() .setDisplay(display) .setSurface(window); auto result = inst.createWaylandSurfaceKHR(&createInfo, nullptr, &surface); VERIFY(result == vk::Result::eSuccess); } #endif if (use_xlib) { #if defined(VK_USE_PLATFORM_XLIB_KHR) auto const createInfo = vk::XlibSurfaceCreateInfoKHR().setDpy(display).setWindow( xlib_window); auto result = inst.createXlibSurfaceKHR(&createInfo, nullptr, &surface); VERIFY(result == vk::Result::eSuccess); #endif } else { #if defined(VK_USE_PLATFORM_XCB_KHR) auto const createInfo = vk::XcbSurfaceCreateInfoKHR() .setConnection(connection) .setWindow(xcb_window); auto result = inst.createXcbSurfaceKHR(&createInfo, nullptr, &surface); VERIFY(result == vk::Result::eSuccess); #endif } // Iterate over each queue to learn whether it supports presenting: std::unique_ptr supportsPresent( new vk::Bool32[queue_family_count]); for (uint32_t i = 0; i < queue_family_count; i++) { gpu.getSurfaceSupportKHR(i, surface, &supportsPresent[i]); } uint32_t graphicsQueueFamilyIndex = UINT32_MAX; uint32_t presentQueueFamilyIndex = UINT32_MAX; for (uint32_t i = 0; i < queue_family_count; i++) { if (queue_props[i].queueFlags & vk::QueueFlagBits::eGraphics) { if (graphicsQueueFamilyIndex == UINT32_MAX) { graphicsQueueFamilyIndex = i; } if (supportsPresent[i] == VK_TRUE) { graphicsQueueFamilyIndex = i; presentQueueFamilyIndex = i; break; } } } if (presentQueueFamilyIndex == UINT32_MAX) { // If didn't find a queue that supports both graphics and present, // then // find a separate present queue. for (uint32_t i = 0; i < queue_family_count; ++i) { if (supportsPresent[i] == VK_TRUE) { presentQueueFamilyIndex = i; break; } } } // Generate error if could not find both a graphics and a present queue if (graphicsQueueFamilyIndex == UINT32_MAX || presentQueueFamilyIndex == UINT32_MAX) { ERR_EXIT("Could not find both graphics and present queues\n", "Swapchain Initialization Failure"); } graphics_queue_family_index = graphicsQueueFamilyIndex; present_queue_family_index = presentQueueFamilyIndex; separate_present_queue = (graphics_queue_family_index != present_queue_family_index); create_device(); device.getQueue(graphics_queue_family_index, 0, &graphics_queue); if (!separate_present_queue) { present_queue = graphics_queue; } else { device.getQueue(present_queue_family_index, 0, &present_queue); } // Get the list of VkFormat's that are supported: uint32_t formatCount; auto result = gpu.getSurfaceFormatsKHR(surface, &formatCount, nullptr); VERIFY(result == vk::Result::eSuccess); std::unique_ptr surfFormats( new vk::SurfaceFormatKHR[formatCount]); result = gpu.getSurfaceFormatsKHR(surface, &formatCount, surfFormats.get()); VERIFY(result == vk::Result::eSuccess); // If the format list includes just one entry of VK_FORMAT_UNDEFINED, // the surface has no preferred format. Otherwise, at least one // supported format will be returned. if (formatCount == 1 && surfFormats[0].format == vk::Format::eUndefined) { format = vk::Format::eB8G8R8A8Unorm; } else { assert(formatCount >= 1); format = surfFormats[0].format; } color_space = surfFormats[0].colorSpace; quit = false; curFrame = 0; // Create semaphores to synchronize acquiring presentable buffers before // rendering and waiting for drawing to be complete before presenting auto const semaphoreCreateInfo = vk::SemaphoreCreateInfo(); // Create fences that we can use to throttle if we get too far // ahead of the image presents vk::FenceCreateInfo const fence_ci; for (uint32_t i = 0; i < FRAME_LAG; i++) { device.createFence(&fence_ci, nullptr, &fences[i]); result = device.createSemaphore(&semaphoreCreateInfo, nullptr, &image_acquired_semaphores[i]); VERIFY(result == vk::Result::eSuccess); result = device.createSemaphore(&semaphoreCreateInfo, nullptr, &draw_complete_semaphores[i]); VERIFY(result == vk::Result::eSuccess); if (separate_present_queue) { result = device.createSemaphore(&semaphoreCreateInfo, nullptr, &image_ownership_semaphores[i]); VERIFY(result == vk::Result::eSuccess); } } frame_index = 0; // Get Memory information and properties gpu.getMemoryProperties(&memory_properties); } void prepare() { auto const cmd_pool_info = vk::CommandPoolCreateInfo().setQueueFamilyIndex( graphics_queue_family_index); auto result = device.createCommandPool(&cmd_pool_info, nullptr, &cmd_pool); VERIFY(result == vk::Result::eSuccess); auto const cmd = vk::CommandBufferAllocateInfo() .setCommandPool(cmd_pool) .setLevel(vk::CommandBufferLevel::ePrimary) .setCommandBufferCount(1); prepare_buffers(); prepare_depth(); prepare_textures(); prepare_cube_data_buffer(); prepare_descriptor_layout(); prepare_render_pass(); prepare_pipeline(); for (uint32_t i = 0; i < swapchainImageCount; ++i) { result = device.allocateCommandBuffers(&cmd, &buffers[i].cmd); VERIFY(result == vk::Result::eSuccess); } if (separate_present_queue) { auto const cmd_pool_info = vk::CommandPoolCreateInfo().setQueueFamilyIndex( present_queue_family_index); result = device.createCommandPool(&cmd_pool_info, nullptr, &present_cmd_pool); VERIFY(result == vk::Result::eSuccess); auto const cmd = vk::CommandBufferAllocateInfo() .setCommandPool(present_cmd_pool) .setLevel(vk::CommandBufferLevel::ePrimary) .setCommandBufferCount(1); for (uint32_t i = 0; i < swapchainImageCount; i++) { result = device.allocateCommandBuffers( &cmd, &buffers[i].graphics_to_present_cmd); VERIFY(result == vk::Result::eSuccess); build_image_ownership_cmd(i); } } prepare_descriptor_pool(); prepare_descriptor_set(); prepare_framebuffers(); for (uint32_t i = 0; i < swapchainImageCount; ++i) { current_buffer = i; draw_build_cmd(buffers[i].cmd); } /* * Prepare functions above may generate pipeline commands * that need to be flushed before beginning the render loop. */ flush_init_cmd(); current_buffer = 0; prepared = true; } void prepare_buffers() { vk::SwapchainKHR oldSwapchain = swapchain; // Check the surface capabilities and formats vk::SurfaceCapabilitiesKHR surfCapabilities; auto result = gpu.getSurfaceCapabilitiesKHR(surface, &surfCapabilities); VERIFY(result == vk::Result::eSuccess); uint32_t presentModeCount; result = gpu.getSurfacePresentModesKHR(surface, &presentModeCount, nullptr); VERIFY(result == vk::Result::eSuccess); std::unique_ptr presentModes( new vk::PresentModeKHR[presentModeCount]); result = gpu.getSurfacePresentModesKHR(surface, &presentModeCount, presentModes.get()); VERIFY(result == vk::Result::eSuccess); vk::Extent2D swapchainExtent; // width and height are either both -1, or both not -1. if (surfCapabilities.currentExtent.width == (uint32_t)-1) { // If the surface size is undefined, the size is set to // the size of the images requested. swapchainExtent.width = width; swapchainExtent.height = height; } else { // If the surface size is defined, the swap chain size must match swapchainExtent = surfCapabilities.currentExtent; width = surfCapabilities.currentExtent.width; height = surfCapabilities.currentExtent.height; } // The FIFO present mode is guaranteed by the spec to be supported // and to have no tearing. It's a great default present mode to use. vk::PresentModeKHR swapchainPresentMode = vk::PresentModeKHR::eFifo; // There are times when you may wish to use another present mode. The // following code shows how to select them, and the comments provide some // reasons you may wish to use them. // // It should be noted that Vulkan 1.0 doesn't provide a method for // synchronizing rendering with the presentation engine's display. There // is a method provided for throttling rendering with the display, but // there are some presentation engines for which this method will not work. // If an application doesn't throttle its rendering, and if it renders much // faster than the refresh rate of the display, this can waste power on // mobile devices. That is because power is being spent rendering images // that may never be seen. //#define DESIRE_VK_PRESENT_MODE_IMMEDIATE_KHR //#define DESIRE_VK_PRESENT_MODE_MAILBOX_KHR //#define DESIRE_VK_PRESENT_MODE_FIFO_RELAXED_KHR #if defined(DESIRE_VK_PRESENT_MODE_IMMEDIATE_KHR) // VK_PRESENT_MODE_IMMEDIATE_KHR is for applications that don't care // about // tearing, or have some way of synchronizing their rendering with the // display. for (size_t i = 0; i < presentModeCount; ++i) { if (presentModes[i] == vk::PresentModeKHR::eImmediate) { swapchainPresentMode = vk::PresentModeKHR::eImmediate; break; } } #elif defined(DESIRE_VK_PRESENT_MODE_MAILBOX_KHR) // VK_PRESENT_MODE_MAILBOX_KHR may be useful for applications that // generally render a new presentable image every refresh cycle, but are // occasionally early. In this case, the application wants the new // image // to be displayed instead of the previously-queued-for-presentation // image // that has not yet been displayed. for (size_t i = 0; i < presentModeCount; ++i) { if (presentModes[i] == vk::PresentModeKHR::eMailbox) { swapchainPresentMode = vk::PresentModeKHR::eMailbox; break; } } #elif defined(DESIRE_VK_PRESENT_MODE_FIFO_RELAXED_KHR) // VK_PRESENT_MODE_FIFO_RELAXED_KHR is for applications that generally // render a new presentable image every refresh cycle, but are // occasionally // late. In this case (perhaps because of stuttering/latency concerns), // the application wants the late image to be immediately displayed, // even // though that may mean some tearing. for (size_t i = 0; i < presentModeCount; ++i) { if (presentModes[i] == vk::PresentModeKHR::eFifoRelaxed) { swapchainPresentMode = vk::PresentModeKHR::eFifoRelaxed; break; } } #endif // Determine the number of VkImage's to use in the swap chain (we desire // to // own only 1 image at a time, besides the images being displayed and // queued for display): uint32_t desiredNumberOfSwapchainImages = surfCapabilities.minImageCount + 1; // If maxImageCount is 0, we can ask for as many images as we want, // otherwise // we're limited to maxImageCount if ((surfCapabilities.maxImageCount > 0) && (desiredNumberOfSwapchainImages > surfCapabilities.maxImageCount)) { // Application must settle for fewer images than desired: desiredNumberOfSwapchainImages = surfCapabilities.maxImageCount; } vk::SurfaceTransformFlagBitsKHR preTransform; if (surfCapabilities.supportedTransforms & vk::SurfaceTransformFlagBitsKHR::eIdentity) { preTransform = vk::SurfaceTransformFlagBitsKHR::eIdentity; } else { preTransform = surfCapabilities.currentTransform; } auto const swapchain_ci = vk::SwapchainCreateInfoKHR() .setSurface(surface) .setMinImageCount(desiredNumberOfSwapchainImages) .setImageFormat(format) .setImageColorSpace(color_space) .setImageExtent({swapchainExtent.width, swapchainExtent.height}) .setImageArrayLayers(1) .setImageUsage(vk::ImageUsageFlagBits::eColorAttachment) .setImageSharingMode(vk::SharingMode::eExclusive) .setQueueFamilyIndexCount(0) .setPQueueFamilyIndices(nullptr) .setPreTransform(preTransform) .setCompositeAlpha(vk::CompositeAlphaFlagBitsKHR::eOpaque) .setPresentMode(swapchainPresentMode) .setClipped(true) .setOldSwapchain(oldSwapchain); result = device.createSwapchainKHR(&swapchain_ci, nullptr, &swapchain); VERIFY(result == vk::Result::eSuccess); // If we just re-created an existing swapchain, we should destroy the // old // swapchain at this point. // Note: destroying the swapchain also cleans up all its associated // presentable images once the platform is done with them. if (oldSwapchain) { device.destroySwapchainKHR(oldSwapchain, nullptr); } result = device.getSwapchainImagesKHR(swapchain, &swapchainImageCount, nullptr); VERIFY(result == vk::Result::eSuccess); std::unique_ptr swapchainImages( new vk::Image[swapchainImageCount]); result = device.getSwapchainImagesKHR(swapchain, &swapchainImageCount, swapchainImages.get()); VERIFY(result == vk::Result::eSuccess); buffers.reset(new SwapchainBuffers[swapchainImageCount]); for (uint32_t i = 0; i < swapchainImageCount; ++i) { auto const color_image_view = vk::ImageViewCreateInfo() .setImage(swapchainImages[i]) .setViewType(vk::ImageViewType::e2D) .setFormat(format) .setSubresourceRange(vk::ImageSubresourceRange( vk::ImageAspectFlagBits::eColor, 0, 1, 0, 1)); buffers[i].image = swapchainImages[i]; result = device.createImageView(&color_image_view, nullptr, &buffers[i].view); VERIFY(result == vk::Result::eSuccess); } } void prepare_cube_data_buffer() { mat4x4 VP; mat4x4_mul(VP, projection_matrix, view_matrix); mat4x4 MVP; mat4x4_mul(MVP, VP, model_matrix); vktexcube_vs_uniform data; memcpy(data.mvp, MVP, sizeof(MVP)); // dumpMatrix("MVP", MVP) for (int32_t i = 0; i < 12 * 3; i++) { data.position[i][0] = g_vertex_buffer_data[i * 3]; data.position[i][1] = g_vertex_buffer_data[i * 3 + 1]; data.position[i][2] = g_vertex_buffer_data[i * 3 + 2]; data.position[i][3] = 1.0f; data.attr[i][0] = g_uv_buffer_data[2 * i]; data.attr[i][1] = g_uv_buffer_data[2 * i + 1]; data.attr[i][2] = 0; data.attr[i][3] = 0; } auto const buf_info = vk::BufferCreateInfo() .setSize(sizeof(data)) .setUsage(vk::BufferUsageFlagBits::eUniformBuffer); auto result = device.createBuffer(&buf_info, nullptr, &uniform_data.buf); VERIFY(result == vk::Result::eSuccess); vk::MemoryRequirements mem_reqs; device.getBufferMemoryRequirements(uniform_data.buf, &mem_reqs); uniform_data.mem_alloc.setAllocationSize(mem_reqs.size); uniform_data.mem_alloc.setMemoryTypeIndex(0); bool const pass = memory_type_from_properties( mem_reqs.memoryTypeBits, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent, &uniform_data.mem_alloc.memoryTypeIndex); VERIFY(pass); result = device.allocateMemory(&uniform_data.mem_alloc, nullptr, &(uniform_data.mem)); VERIFY(result == vk::Result::eSuccess); auto pData = device.mapMemory(uniform_data.mem, 0, uniform_data.mem_alloc.allocationSize, vk::MemoryMapFlags()); VERIFY(pData.result == vk::Result::eSuccess); memcpy(pData.value, &data, sizeof data); device.unmapMemory(uniform_data.mem); result = device.bindBufferMemory(uniform_data.buf, uniform_data.mem, 0); VERIFY(result == vk::Result::eSuccess); uniform_data.buffer_info.buffer = uniform_data.buf; uniform_data.buffer_info.offset = 0; uniform_data.buffer_info.range = sizeof(data); } void prepare_depth() { depth.format = vk::Format::eD16Unorm; auto const image = vk::ImageCreateInfo() .setImageType(vk::ImageType::e2D) .setFormat(depth.format) .setExtent({(uint32_t)width, (uint32_t)height, 1}) .setMipLevels(1) .setArrayLayers(1) .setSamples(vk::SampleCountFlagBits::e1) .setTiling(vk::ImageTiling::eOptimal) .setUsage(vk::ImageUsageFlagBits::eDepthStencilAttachment) .setSharingMode(vk::SharingMode::eExclusive) .setQueueFamilyIndexCount(0) .setPQueueFamilyIndices(nullptr) .setInitialLayout(vk::ImageLayout::eUndefined); auto result = device.createImage(&image, nullptr, &depth.image); VERIFY(result == vk::Result::eSuccess); vk::MemoryRequirements mem_reqs; device.getImageMemoryRequirements(depth.image, &mem_reqs); depth.mem_alloc.setAllocationSize(mem_reqs.size); depth.mem_alloc.setMemoryTypeIndex(0); auto const pass = memory_type_from_properties( mem_reqs.memoryTypeBits, vk::MemoryPropertyFlagBits(0), &depth.mem_alloc.memoryTypeIndex); VERIFY(pass); result = device.allocateMemory(&depth.mem_alloc, nullptr, &depth.mem); VERIFY(result == vk::Result::eSuccess); result = device.bindImageMemory(depth.image, depth.mem, 0); VERIFY(result == vk::Result::eSuccess); auto const view = vk::ImageViewCreateInfo() .setImage(depth.image) .setViewType(vk::ImageViewType::e2D) .setFormat(depth.format) .setSubresourceRange(vk::ImageSubresourceRange( vk::ImageAspectFlagBits::eDepth, 0, 1, 0, 1)); result = device.createImageView(&view, nullptr, &depth.view); VERIFY(result == vk::Result::eSuccess); } void prepare_descriptor_layout() { vk::DescriptorSetLayoutBinding const layout_bindings[2] = { vk::DescriptorSetLayoutBinding() .setBinding(0) .setDescriptorType(vk::DescriptorType::eUniformBuffer) .setDescriptorCount(1) .setStageFlags(vk::ShaderStageFlagBits::eVertex) .setPImmutableSamplers(nullptr), vk::DescriptorSetLayoutBinding() .setBinding(1) .setDescriptorType(vk::DescriptorType::eCombinedImageSampler) .setDescriptorCount(texture_count) .setStageFlags(vk::ShaderStageFlagBits::eFragment) .setPImmutableSamplers(nullptr)}; auto const descriptor_layout = vk::DescriptorSetLayoutCreateInfo().setBindingCount(2).setPBindings( layout_bindings); auto result = device.createDescriptorSetLayout(&descriptor_layout, nullptr, &desc_layout); VERIFY(result == vk::Result::eSuccess); auto const pPipelineLayoutCreateInfo = vk::PipelineLayoutCreateInfo().setSetLayoutCount(1).setPSetLayouts( &desc_layout); result = device.createPipelineLayout(&pPipelineLayoutCreateInfo, nullptr, &pipeline_layout); VERIFY(result == vk::Result::eSuccess); } void prepare_descriptor_pool() { vk::DescriptorPoolSize const poolSizes[2] = { vk::DescriptorPoolSize() .setType(vk::DescriptorType::eUniformBuffer) .setDescriptorCount(1), vk::DescriptorPoolSize() .setType(vk::DescriptorType::eCombinedImageSampler) .setDescriptorCount(texture_count)}; auto const descriptor_pool = vk::DescriptorPoolCreateInfo() .setMaxSets(1) .setPoolSizeCount(2) .setPPoolSizes(poolSizes); auto result = device.createDescriptorPool(&descriptor_pool, nullptr, &desc_pool); VERIFY(result == vk::Result::eSuccess); } void prepare_descriptor_set() { auto const alloc_info = vk::DescriptorSetAllocateInfo() .setDescriptorPool(desc_pool) .setDescriptorSetCount(1) .setPSetLayouts(&desc_layout); auto result = device.allocateDescriptorSets(&alloc_info, &desc_set); VERIFY(result == vk::Result::eSuccess); vk::DescriptorImageInfo tex_descs[texture_count]; for (uint32_t i = 0; i < texture_count; i++) { tex_descs[i].setSampler(textures[i].sampler); tex_descs[i].setImageView(textures[i].view); tex_descs[i].setImageLayout(vk::ImageLayout::eGeneral); } vk::WriteDescriptorSet writes[2]; writes[0].setDstSet(desc_set); writes[0].setDescriptorCount(1); writes[0].setDescriptorType(vk::DescriptorType::eUniformBuffer); writes[0].setPBufferInfo(&uniform_data.buffer_info); writes[1].setDstSet(desc_set); writes[1].setDstBinding(1); writes[1].setDescriptorCount(texture_count); writes[1].setDescriptorType(vk::DescriptorType::eCombinedImageSampler); writes[1].setPImageInfo(tex_descs); device.updateDescriptorSets(2, writes, 0, nullptr); } void prepare_framebuffers() { vk::ImageView attachments[2]; attachments[1] = depth.view; auto const fb_info = vk::FramebufferCreateInfo() .setRenderPass(render_pass) .setAttachmentCount(2) .setPAttachments(attachments) .setWidth((uint32_t)width) .setHeight((uint32_t)height) .setLayers(1); framebuffers.reset(new vk::Framebuffer[swapchainImageCount]); for (uint32_t i = 0; i < swapchainImageCount; i++) { attachments[0] = buffers[i].view; auto const result = device.createFramebuffer(&fb_info, nullptr, &framebuffers[i]); VERIFY(result == vk::Result::eSuccess); } } vk::ShaderModule prepare_fs() { size_t size = 0; void *fragShaderCode = read_spv("cube-frag.spv", &size); frag_shader_module = prepare_shader_module(fragShaderCode, size); free(fragShaderCode); return frag_shader_module; } void prepare_pipeline() { vk::PipelineCacheCreateInfo const pipelineCacheInfo; auto result = device.createPipelineCache(&pipelineCacheInfo, nullptr, &pipelineCache); VERIFY(result == vk::Result::eSuccess); vk::PipelineShaderStageCreateInfo const shaderStageInfo[2] = { vk::PipelineShaderStageCreateInfo() .setStage(vk::ShaderStageFlagBits::eVertex) .setModule(prepare_vs()) .setPName("main"), vk::PipelineShaderStageCreateInfo() .setStage(vk::ShaderStageFlagBits::eFragment) .setModule(prepare_fs()) .setPName("main")}; vk::PipelineVertexInputStateCreateInfo const vertexInputInfo; auto const inputAssemblyInfo = vk::PipelineInputAssemblyStateCreateInfo().setTopology( vk::PrimitiveTopology::eTriangleList); // TODO: Where are pViewports and pScissors set? auto const viewportInfo = vk::PipelineViewportStateCreateInfo() .setViewportCount(1) .setScissorCount(1); auto const rasterizationInfo = vk::PipelineRasterizationStateCreateInfo() .setDepthClampEnable(VK_FALSE) .setRasterizerDiscardEnable(VK_FALSE) .setPolygonMode(vk::PolygonMode::eFill) .setCullMode(vk::CullModeFlagBits::eBack) .setFrontFace(vk::FrontFace::eCounterClockwise) .setDepthBiasEnable(VK_FALSE) .setLineWidth(1.0f); auto const multisampleInfo = vk::PipelineMultisampleStateCreateInfo(); auto const stencilOp = vk::StencilOpState() .setFailOp(vk::StencilOp::eKeep) .setPassOp(vk::StencilOp::eKeep) .setCompareOp(vk::CompareOp::eAlways); auto const depthStencilInfo = vk::PipelineDepthStencilStateCreateInfo() .setDepthTestEnable(VK_TRUE) .setDepthWriteEnable(VK_TRUE) .setDepthCompareOp(vk::CompareOp::eLessOrEqual) .setDepthBoundsTestEnable(VK_FALSE) .setStencilTestEnable(VK_FALSE) .setFront(stencilOp) .setBack(stencilOp); vk::PipelineColorBlendAttachmentState const colorBlendAttachments[1] = { vk::PipelineColorBlendAttachmentState().setColorWriteMask( vk::ColorComponentFlagBits::eR | vk::ColorComponentFlagBits::eG | vk::ColorComponentFlagBits::eB | vk::ColorComponentFlagBits::eA)}; auto const colorBlendInfo = vk::PipelineColorBlendStateCreateInfo() .setAttachmentCount(1) .setPAttachments(colorBlendAttachments); vk::DynamicState const dynamicStates[2] = {vk::DynamicState::eViewport, vk::DynamicState::eScissor}; auto const dynamicStateInfo = vk::PipelineDynamicStateCreateInfo() .setPDynamicStates(dynamicStates) .setDynamicStateCount(2); auto const pipeline = vk::GraphicsPipelineCreateInfo() .setStageCount(2) .setPStages(shaderStageInfo) .setPVertexInputState(&vertexInputInfo) .setPInputAssemblyState(&inputAssemblyInfo) .setPViewportState(&viewportInfo) .setPRasterizationState(&rasterizationInfo) .setPMultisampleState(&multisampleInfo) .setPDepthStencilState(&depthStencilInfo) .setPColorBlendState(&colorBlendInfo) .setPDynamicState(&dynamicStateInfo) .setLayout(pipeline_layout) .setRenderPass(render_pass); result = device.createGraphicsPipelines(pipelineCache, 1, &pipeline, nullptr, &this->pipeline); VERIFY(result == vk::Result::eSuccess); device.destroyShaderModule(frag_shader_module, nullptr); device.destroyShaderModule(vert_shader_module, nullptr); } void prepare_render_pass() { // The initial layout for the color and depth attachments will be LAYOUT_UNDEFINED // because at the start of the renderpass, we don't care about their contents. // At the start of the subpass, the color attachment's layout will be transitioned // to LAYOUT_COLOR_ATTACHMENT_OPTIMAL and the depth stencil attachment's layout // will be transitioned to LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL. At the end of // the renderpass, the color attachment's layout will be transitioned to // LAYOUT_PRESENT_SRC_KHR to be ready to present. This is all done as part of // the renderpass, no barriers are necessary. const vk::AttachmentDescription attachments[2] = { vk::AttachmentDescription() .setFlags(vk::AttachmentDescriptionFlagBits::eMayAlias) .setFormat(format) .setSamples(vk::SampleCountFlagBits::e1) .setLoadOp(vk::AttachmentLoadOp::eClear) .setStoreOp(vk::AttachmentStoreOp::eStore) .setStencilLoadOp(vk::AttachmentLoadOp::eDontCare) .setStencilStoreOp(vk::AttachmentStoreOp::eDontCare) .setInitialLayout(vk::ImageLayout::eUndefined) .setFinalLayout(vk::ImageLayout::ePresentSrcKHR), vk::AttachmentDescription() .setFlags(vk::AttachmentDescriptionFlagBits::eMayAlias) .setFormat(depth.format) .setSamples(vk::SampleCountFlagBits::e1) .setLoadOp(vk::AttachmentLoadOp::eClear) .setStoreOp(vk::AttachmentStoreOp::eDontCare) .setStencilLoadOp(vk::AttachmentLoadOp::eDontCare) .setStencilStoreOp(vk::AttachmentStoreOp::eDontCare) .setInitialLayout( vk::ImageLayout::eUndefined) .setFinalLayout( vk::ImageLayout::eDepthStencilAttachmentOptimal)}; auto const color_reference = vk::AttachmentReference().setAttachment(0).setLayout( vk::ImageLayout::eColorAttachmentOptimal); auto const depth_reference = vk::AttachmentReference().setAttachment(1).setLayout( vk::ImageLayout::eDepthStencilAttachmentOptimal); auto const subpass = vk::SubpassDescription() .setPipelineBindPoint(vk::PipelineBindPoint::eGraphics) .setInputAttachmentCount(0) .setPInputAttachments(nullptr) .setColorAttachmentCount(1) .setPColorAttachments(&color_reference) .setPResolveAttachments(nullptr) .setPDepthStencilAttachment(&depth_reference) .setPreserveAttachmentCount(0) .setPPreserveAttachments(nullptr); auto const rp_info = vk::RenderPassCreateInfo() .setAttachmentCount(2) .setPAttachments(attachments) .setSubpassCount(1) .setPSubpasses(&subpass) .setDependencyCount(0) .setPDependencies(nullptr); auto result = device.createRenderPass(&rp_info, nullptr, &render_pass); VERIFY(result == vk::Result::eSuccess); } vk::ShaderModule prepare_shader_module(const void *code, size_t size) { auto const moduleCreateInfo = vk::ShaderModuleCreateInfo().setCodeSize(size).setPCode( (uint32_t const *)code); vk::ShaderModule module; auto result = device.createShaderModule(&moduleCreateInfo, nullptr, &module); VERIFY(result == vk::Result::eSuccess); return module; } void prepare_texture_image(const char *filename, texture_object *tex_obj, vk::ImageTiling tiling, vk::ImageUsageFlags usage, vk::MemoryPropertyFlags required_props) { int32_t tex_width; int32_t tex_height; if (!loadTexture(filename, nullptr, nullptr, &tex_width, &tex_height)) { ERR_EXIT("Failed to load textures", "Load Texture Failure"); } tex_obj->tex_width = tex_width; tex_obj->tex_height = tex_height; auto const image_create_info = vk::ImageCreateInfo() .setImageType(vk::ImageType::e2D) .setFormat(vk::Format::eR8G8B8A8Unorm) .setExtent({(uint32_t)tex_width, (uint32_t)tex_height, 1}) .setMipLevels(1) .setArrayLayers(1) .setSamples(vk::SampleCountFlagBits::e1) .setTiling(tiling) .setUsage(usage) .setSharingMode(vk::SharingMode::eExclusive) .setQueueFamilyIndexCount(0) .setPQueueFamilyIndices(nullptr) .setInitialLayout(vk::ImageLayout::ePreinitialized); auto result = device.createImage(&image_create_info, nullptr, &tex_obj->image); VERIFY(result == vk::Result::eSuccess); vk::MemoryRequirements mem_reqs; device.getImageMemoryRequirements(tex_obj->image, &mem_reqs); tex_obj->mem_alloc.setAllocationSize(mem_reqs.size); tex_obj->mem_alloc.setMemoryTypeIndex(0); auto pass = memory_type_from_properties(mem_reqs.memoryTypeBits, required_props, &tex_obj->mem_alloc.memoryTypeIndex); VERIFY(pass == true); result = device.allocateMemory(&tex_obj->mem_alloc, nullptr, &(tex_obj->mem)); VERIFY(result == vk::Result::eSuccess); result = device.bindImageMemory(tex_obj->image, tex_obj->mem, 0); VERIFY(result == vk::Result::eSuccess); if (required_props & vk::MemoryPropertyFlagBits::eHostVisible) { auto const subres = vk::ImageSubresource() .setAspectMask(vk::ImageAspectFlagBits::eColor) .setMipLevel(0) .setArrayLayer(0); vk::SubresourceLayout layout; device.getImageSubresourceLayout(tex_obj->image, &subres, &layout); auto data = device.mapMemory(tex_obj->mem, 0, tex_obj->mem_alloc.allocationSize); VERIFY(data.result == vk::Result::eSuccess); if (!loadTexture(filename, (uint8_t *)data.value, &layout, &tex_width, &tex_height)) { fprintf(stderr, "Error loading texture: %s\n", filename); } device.unmapMemory(tex_obj->mem); } tex_obj->imageLayout = vk::ImageLayout::eShaderReadOnlyOptimal; } void prepare_textures() { vk::Format const tex_format = vk::Format::eR8G8B8A8Unorm; vk::FormatProperties props; gpu.getFormatProperties(tex_format, &props); for (uint32_t i = 0; i < texture_count; i++) { if ((props.linearTilingFeatures & vk::FormatFeatureFlagBits::eSampledImage) && !use_staging_buffer) { /* Device can texture using linear textures */ prepare_texture_image( tex_files[i], &textures[i], vk::ImageTiling::eLinear, vk::ImageUsageFlagBits::eSampled, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent); // Nothing in the pipeline needs to be complete to start, and don't allow fragment // shader to run until layout transition completes set_image_layout(textures[i].image, vk::ImageAspectFlagBits::eColor, vk::ImageLayout::ePreinitialized, textures[i].imageLayout, vk::AccessFlagBits::eHostWrite, vk::PipelineStageFlagBits::eTopOfPipe, vk::PipelineStageFlagBits::eFragmentShader); } else if (props.optimalTilingFeatures & vk::FormatFeatureFlagBits::eSampledImage) { /* Must use staging buffer to copy linear texture to optimized */ texture_object staging_texture; prepare_texture_image( tex_files[i], &staging_texture, vk::ImageTiling::eLinear, vk::ImageUsageFlagBits::eTransferSrc, vk::MemoryPropertyFlagBits::eHostVisible | vk::MemoryPropertyFlagBits::eHostCoherent); prepare_texture_image(tex_files[i], &textures[i], vk::ImageTiling::eOptimal, vk::ImageUsageFlagBits::eTransferDst | vk::ImageUsageFlagBits::eSampled, vk::MemoryPropertyFlagBits::eDeviceLocal); set_image_layout( staging_texture.image, vk::ImageAspectFlagBits::eColor, vk::ImageLayout::ePreinitialized, vk::ImageLayout::eTransferSrcOptimal, vk::AccessFlagBits::eHostWrite, vk::PipelineStageFlagBits::eTopOfPipe, vk::PipelineStageFlagBits::eTransfer); set_image_layout( textures[i].image, vk::ImageAspectFlagBits::eColor, vk::ImageLayout::ePreinitialized, vk::ImageLayout::eTransferDstOptimal, vk::AccessFlagBits::eHostWrite, vk::PipelineStageFlagBits::eTopOfPipe, vk::PipelineStageFlagBits::eTransfer); auto const subresource = vk::ImageSubresourceLayers() .setAspectMask(vk::ImageAspectFlagBits::eColor) .setMipLevel(0) .setBaseArrayLayer(0) .setLayerCount(1); auto const copy_region = vk::ImageCopy() .setSrcSubresource(subresource) .setSrcOffset({0, 0, 0}) .setDstSubresource(subresource) .setDstOffset({0, 0, 0}) .setExtent({(uint32_t)staging_texture.tex_width, (uint32_t)staging_texture.tex_height, 1}); cmd.copyImage( staging_texture.image, vk::ImageLayout::eTransferSrcOptimal, textures[i].image, vk::ImageLayout::eTransferDstOptimal, 1, ©_region); set_image_layout(textures[i].image, vk::ImageAspectFlagBits::eColor, vk::ImageLayout::eTransferDstOptimal, textures[i].imageLayout, vk::AccessFlagBits::eTransferWrite, vk::PipelineStageFlagBits::eTransfer, vk::PipelineStageFlagBits::eFragmentShader); flush_init_cmd(); destroy_texture_image(&staging_texture); } else { assert( !"No support for R8G8B8A8_UNORM as texture image format"); } auto const samplerInfo = vk::SamplerCreateInfo() .setMagFilter(vk::Filter::eNearest) .setMinFilter(vk::Filter::eNearest) .setMipmapMode(vk::SamplerMipmapMode::eNearest) .setAddressModeU(vk::SamplerAddressMode::eClampToEdge) .setAddressModeV(vk::SamplerAddressMode::eClampToEdge) .setAddressModeW(vk::SamplerAddressMode::eClampToEdge) .setMipLodBias(0.0f) .setAnisotropyEnable(VK_FALSE) .setMaxAnisotropy(1) .setCompareEnable(VK_FALSE) .setCompareOp(vk::CompareOp::eNever) .setMinLod(0.0f) .setMaxLod(0.0f) .setBorderColor(vk::BorderColor::eFloatOpaqueWhite) .setUnnormalizedCoordinates(VK_FALSE); auto result = device.createSampler(&samplerInfo, nullptr, &textures[i].sampler); VERIFY(result == vk::Result::eSuccess); auto const viewInfo = vk::ImageViewCreateInfo() .setImage(textures[i].image) .setViewType(vk::ImageViewType::e2D) .setFormat(tex_format) .setSubresourceRange(vk::ImageSubresourceRange( vk::ImageAspectFlagBits::eColor, 0, 1, 0, 1)); result = device.createImageView(&viewInfo, nullptr, &textures[i].view); VERIFY(result == vk::Result::eSuccess); } } vk::ShaderModule prepare_vs() { size_t size = 0; void *vertShaderCode = read_spv("cube-vert.spv", &size); vert_shader_module = prepare_shader_module(vertShaderCode, size); free(vertShaderCode); return vert_shader_module; } char *read_spv(const char *filename, size_t *psize) { FILE *fp = fopen(filename, "rb"); if (!fp) { return nullptr; } fseek(fp, 0L, SEEK_END); long int size = ftell(fp); fseek(fp, 0L, SEEK_SET); void *shader_code = malloc(size); size_t retval = fread(shader_code, size, 1, fp); VERIFY(retval == 1); *psize = size; fclose(fp); return (char *)shader_code; } void resize() { uint32_t i; // Don't react to resize until after first initialization. if (!prepared) { return; } // In order to properly resize the window, we must re-create the // swapchain // AND redo the command buffers, etc. // // First, perform part of the cleanup() function: prepared = false; auto result = device.waitIdle(); VERIFY(result == vk::Result::eSuccess); for (i = 0; i < swapchainImageCount; i++) { device.destroyFramebuffer(framebuffers[i], nullptr); } device.destroyDescriptorPool(desc_pool, nullptr); device.destroyPipeline(pipeline, nullptr); device.destroyPipelineCache(pipelineCache, nullptr); device.destroyRenderPass(render_pass, nullptr); device.destroyPipelineLayout(pipeline_layout, nullptr); device.destroyDescriptorSetLayout(desc_layout, nullptr); for (i = 0; i < texture_count; i++) { device.destroyImageView(textures[i].view, nullptr); device.destroyImage(textures[i].image, nullptr); device.freeMemory(textures[i].mem, nullptr); device.destroySampler(textures[i].sampler, nullptr); } device.destroyImageView(depth.view, nullptr); device.destroyImage(depth.image, nullptr); device.freeMemory(depth.mem, nullptr); device.destroyBuffer(uniform_data.buf, nullptr); device.freeMemory(uniform_data.mem, nullptr); for (i = 0; i < swapchainImageCount; i++) { device.destroyImageView(buffers[i].view, nullptr); device.freeCommandBuffers(cmd_pool, 1, &buffers[i].cmd); } device.destroyCommandPool(cmd_pool, nullptr); if (separate_present_queue) { device.destroyCommandPool(present_cmd_pool, nullptr); } // Second, re-perform the prepare() function, which will re-create the // swapchain. prepare(); } void set_image_layout(vk::Image image, vk::ImageAspectFlags aspectMask, vk::ImageLayout oldLayout, vk::ImageLayout newLayout, vk::AccessFlags srcAccessMask, vk::PipelineStageFlags src_stages, vk::PipelineStageFlags dest_stages) { if (!cmd) { auto const cmd = vk::CommandBufferAllocateInfo() .setCommandPool(cmd_pool) .setLevel(vk::CommandBufferLevel::ePrimary) .setCommandBufferCount(1); auto result = device.allocateCommandBuffers(&cmd, &this->cmd); VERIFY(result == vk::Result::eSuccess); auto const cmd_buf_info = vk::CommandBufferBeginInfo().setPInheritanceInfo(nullptr); result = this->cmd.begin(&cmd_buf_info); VERIFY(result == vk::Result::eSuccess); } auto DstAccessMask = [](vk::ImageLayout const &layout) { vk::AccessFlags flags; switch (layout) { case vk::ImageLayout::eTransferDstOptimal: // Make sure anything that was copying from this image has // completed flags = vk::AccessFlagBits::eTransferWrite; break; case vk::ImageLayout::eColorAttachmentOptimal: flags = vk::AccessFlagBits::eColorAttachmentWrite; break; case vk::ImageLayout::eDepthStencilAttachmentOptimal: flags = vk::AccessFlagBits::eDepthStencilAttachmentWrite; break; case vk::ImageLayout::eShaderReadOnlyOptimal: // Make sure any Copy or CPU writes to image are flushed flags = vk::AccessFlagBits::eShaderRead | vk::AccessFlagBits::eInputAttachmentRead; break; case vk::ImageLayout::eTransferSrcOptimal: flags = vk::AccessFlagBits::eTransferRead; break; case vk::ImageLayout::ePresentSrcKHR: flags = vk::AccessFlagBits::eMemoryRead; break; default: break; } return flags; }; auto const barrier = vk::ImageMemoryBarrier() .setSrcAccessMask(srcAccessMask) .setDstAccessMask(DstAccessMask(newLayout)) .setOldLayout(oldLayout) .setNewLayout(newLayout) .setSrcQueueFamilyIndex(0) .setDstQueueFamilyIndex(0) .setImage(image) .setSubresourceRange(vk::ImageSubresourceRange( aspectMask, 0, 1, 0, 1)); cmd.pipelineBarrier(src_stages, dest_stages, vk::DependencyFlagBits(), 0, nullptr, 0, nullptr, 1, &barrier); } void update_data_buffer() { mat4x4 VP; mat4x4_mul(VP, projection_matrix, view_matrix); // Rotate 22.5 degrees around the Y axis mat4x4 Model; mat4x4_dup(Model, model_matrix); mat4x4_rotate(model_matrix, Model, 0.0f, 1.0f, 0.0f, (float)degreesToRadians(spin_angle)); mat4x4 MVP; mat4x4_mul(MVP, VP, model_matrix); auto data = device.mapMemory(uniform_data.mem, 0, uniform_data.mem_alloc.allocationSize, vk::MemoryMapFlags()); VERIFY(data.result == vk::Result::eSuccess); memcpy(data.value, (const void *)&MVP[0][0], sizeof(MVP)); device.unmapMemory(uniform_data.mem); } bool loadTexture(const char *filename, uint8_t *rgba_data, vk::SubresourceLayout *layout, int32_t *width, int32_t *height) { FILE *fPtr = fopen(filename, "rb"); if (!fPtr) { return false; } char header[256]; char *cPtr = fgets(header, 256, fPtr); // P6 if (cPtr == nullptr || strncmp(header, "P6\n", 3)) { fclose(fPtr); return false; } do { cPtr = fgets(header, 256, fPtr); if (cPtr == nullptr) { fclose(fPtr); return false; } } while (!strncmp(header, "#", 1)); sscanf(header, "%u %u", width, height); if (rgba_data == nullptr) { fclose(fPtr); return true; } char *result = fgets(header, 256, fPtr); // Format VERIFY(result != nullptr); if (cPtr == nullptr || strncmp(header, "255\n", 3)) { fclose(fPtr); return false; } for (int y = 0; y < *height; y++) { uint8_t *rowPtr = rgba_data; for (int x = 0; x < *width; x++) { size_t s = fread(rowPtr, 3, 1, fPtr); (void)s; rowPtr[3] = 255; /* Alpha of 1 */ rowPtr += 4; } rgba_data += layout->rowPitch; } fclose(fPtr); return true; } bool memory_type_from_properties(uint32_t typeBits, vk::MemoryPropertyFlags requirements_mask, uint32_t *typeIndex) { // Search memtypes to find first index with those properties for (uint32_t i = 0; i < VK_MAX_MEMORY_TYPES; i++) { if ((typeBits & 1) == 1) { // Type is available, does it match user properties? if ((memory_properties.memoryTypes[i].propertyFlags & requirements_mask) == requirements_mask) { *typeIndex = i; return true; } } typeBits >>= 1; } // No memory types matched, return failure return false; } #if defined(VK_USE_PLATFORM_WIN32_KHR) void run() { if (!prepared) { return; } update_data_buffer(); draw(); curFrame++; if (frameCount != INT_MAX && curFrame == frameCount) { PostQuitMessage(validation_error); } } void create_window() { WNDCLASSEX win_class; // Initialize the window class structure: win_class.cbSize = sizeof(WNDCLASSEX); win_class.style = CS_HREDRAW | CS_VREDRAW; win_class.lpfnWndProc = WndProc; win_class.cbClsExtra = 0; win_class.cbWndExtra = 0; win_class.hInstance = connection; // hInstance win_class.hIcon = LoadIcon(nullptr, IDI_APPLICATION); win_class.hCursor = LoadCursor(nullptr, IDC_ARROW); win_class.hbrBackground = (HBRUSH)GetStockObject(WHITE_BRUSH); win_class.lpszMenuName = nullptr; win_class.lpszClassName = name; win_class.hIconSm = LoadIcon(nullptr, IDI_WINLOGO); // Register window class: if (!RegisterClassEx(&win_class)) { // It didn't work, so try to give a useful error: printf("Unexpected error trying to start the application!\n"); fflush(stdout); exit(1); } // Create window with the registered class: RECT wr = {0, 0, static_cast(width), static_cast(height)}; AdjustWindowRect(&wr, WS_OVERLAPPEDWINDOW, FALSE); window = CreateWindowEx(0, name, // class name name, // app name WS_OVERLAPPEDWINDOW | // window style WS_VISIBLE | WS_SYSMENU, 100, 100, // x/y coords wr.right - wr.left, // width wr.bottom - wr.top, // height nullptr, // handle to parent nullptr, // handle to menu connection, // hInstance nullptr); // no extra parameters if (!window) { // It didn't work, so try to give a useful error: printf("Cannot create a window in which to draw!\n"); fflush(stdout); exit(1); } // Window client area size must be at least 1 pixel high, to prevent // crash. minsize.x = GetSystemMetrics(SM_CXMINTRACK); minsize.y = GetSystemMetrics(SM_CYMINTRACK) + 1; } #elif defined(VK_USE_PLATFORM_XLIB_KHR) || defined(VK_USE_PLATFORM_XCB_KHR) #if defined(VK_USE_PLATFORM_XLIB_KHR) void create_xlib_window() { display = XOpenDisplay(nullptr); long visualMask = VisualScreenMask; int numberOfVisuals; XVisualInfo vInfoTemplate = {}; vInfoTemplate.screen = DefaultScreen(display); XVisualInfo *visualInfo = XGetVisualInfo( display, visualMask, &vInfoTemplate, &numberOfVisuals); Colormap colormap = XCreateColormap(display, RootWindow(display, vInfoTemplate.screen), visualInfo->visual, AllocNone); XSetWindowAttributes windowAttributes = {}; windowAttributes.colormap = colormap; windowAttributes.background_pixel = 0xFFFFFFFF; windowAttributes.border_pixel = 0; windowAttributes.event_mask = KeyPressMask | KeyReleaseMask | StructureNotifyMask | ExposureMask; xlib_window = XCreateWindow( display, RootWindow(display, vInfoTemplate.screen), 0, 0, width, height, 0, visualInfo->depth, InputOutput, visualInfo->visual, CWBackPixel | CWBorderPixel | CWEventMask | CWColormap, &windowAttributes); XSelectInput(display, xlib_window, ExposureMask | KeyPressMask); XMapWindow(display, xlib_window); XFlush(display); xlib_wm_delete_window = XInternAtom(display, "WM_DELETE_WINDOW", False); } void handle_xlib_event(const XEvent *event) { switch (event->type) { case ClientMessage: if ((Atom)event->xclient.data.l[0] == xlib_wm_delete_window) { quit = true; } break; case KeyPress: switch (event->xkey.keycode) { case 0x9: // Escape quit = true; break; case 0x71: // left arrow key spin_angle += spin_increment; break; case 0x72: // right arrow key spin_angle -= spin_increment; break; case 0x41: pause = !pause; break; } break; case ConfigureNotify: if (((int32_t)width != event->xconfigure.width) || ((int32_t)height != event->xconfigure.height)) { width = event->xconfigure.width; height = event->xconfigure.height; resize(); } break; default: break; } } void run_xlib() { while (!quit) { XEvent event; if (pause) { XNextEvent(display, &event); handle_xlib_event(&event); } else { while (XPending(display) > 0) { XNextEvent(display, &event); handle_xlib_event(&event); } } update_data_buffer(); draw(); curFrame++; if (frameCount != UINT32_MAX && curFrame == frameCount) { quit = true; } } } #endif #if defined(VK_USE_PLATFORM_XCB_KHR) void handle_xcb_event(const xcb_generic_event_t *event) { uint8_t event_code = event->response_type & 0x7f; switch (event_code) { case XCB_EXPOSE: // TODO: Resize window break; case XCB_CLIENT_MESSAGE: if ((*(xcb_client_message_event_t *)event).data.data32[0] == (*atom_wm_delete_window).atom) { quit = true; } break; case XCB_KEY_RELEASE: { const xcb_key_release_event_t *key = (const xcb_key_release_event_t *)event; switch (key->detail) { case 0x9: // Escape quit = true; break; case 0x71: // left arrow key spin_angle += spin_increment; break; case 0x72: // right arrow key spin_angle -= spin_increment; break; case 0x41: pause = !pause; break; } } break; case XCB_CONFIGURE_NOTIFY: { const xcb_configure_notify_event_t *cfg = (const xcb_configure_notify_event_t *)event; if ((width != cfg->width) || (height != cfg->height)) { width = cfg->width; height = cfg->height; resize(); } } break; default: break; } } void run_xcb() { xcb_flush(connection); while (!quit) { xcb_generic_event_t *event; if (pause) { event = xcb_wait_for_event(connection); } else { event = xcb_poll_for_event(connection); while (event) { handle_xcb_event(event); free(event); event = xcb_poll_for_event(connection); } } update_data_buffer(); draw(); curFrame++; if (frameCount != UINT32_MAX && curFrame == frameCount) { quit = true; } } } void create_xcb_window() { uint32_t value_mask, value_list[32]; xcb_window = xcb_generate_id(connection); value_mask = XCB_CW_BACK_PIXEL | XCB_CW_EVENT_MASK; value_list[0] = screen->black_pixel; value_list[1] = XCB_EVENT_MASK_KEY_RELEASE | XCB_EVENT_MASK_EXPOSURE | XCB_EVENT_MASK_STRUCTURE_NOTIFY; xcb_create_window(connection, XCB_COPY_FROM_PARENT, xcb_window, screen->root, 0, 0, width, height, 0, XCB_WINDOW_CLASS_INPUT_OUTPUT, screen->root_visual, value_mask, value_list); /* Magic code that will send notification when window is destroyed */ xcb_intern_atom_cookie_t cookie = xcb_intern_atom(connection, 1, 12, "WM_PROTOCOLS"); xcb_intern_atom_reply_t *reply = xcb_intern_atom_reply(connection, cookie, 0); xcb_intern_atom_cookie_t cookie2 = xcb_intern_atom(connection, 0, 16, "WM_DELETE_WINDOW"); atom_wm_delete_window = xcb_intern_atom_reply(connection, cookie2, 0); xcb_change_property(connection, XCB_PROP_MODE_REPLACE, xcb_window, (*reply).atom, 4, 32, 1, &(*atom_wm_delete_window).atom); free(reply); xcb_map_window(connection, xcb_window); // Force the x/y coordinates to 100,100 results are identical in // consecutive // runs const uint32_t coords[] = {100, 100}; xcb_configure_window(connection, xcb_window, XCB_CONFIG_WINDOW_X | XCB_CONFIG_WINDOW_Y, coords); } #endif #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) void run() { while (!quit) { update_data_buffer(); draw(); curFrame++; if (frameCount != UINT32_MAX && curFrame == frameCount) { quit = true; } } } void create_window() { window = wl_compositor_create_surface(compositor); if (!window) { printf("Can not create wayland_surface from compositor!\n"); fflush(stdout); exit(1); } shell_surface = wl_shell_get_shell_surface(shell, window); if (!shell_surface) { printf("Can not get shell_surface from wayland_surface!\n"); fflush(stdout); exit(1); } wl_shell_surface_add_listener(shell_surface, &shell_surface_listener, this); wl_shell_surface_set_toplevel(shell_surface); wl_shell_surface_set_title(shell_surface, APP_SHORT_NAME); } #endif #if defined(VK_USE_PLATFORM_WIN32_KHR) HINSTANCE connection; // hInstance - Windows Instance HWND window; // hWnd - window handle POINT minsize; // minimum window size char name[APP_NAME_STR_LEN]; // Name to put on the window/icon #elif defined(VK_USE_PLATFORM_XLIB_KHR) || defined(VK_USE_PLATFORM_XCB_KHR) Display *display; Window xlib_window; Atom xlib_wm_delete_window; xcb_connection_t *connection; xcb_screen_t *screen; xcb_window_t xcb_window; xcb_intern_atom_reply_t *atom_wm_delete_window; #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) wl_display *display; wl_registry *registry; wl_compositor *compositor; wl_surface *window; wl_shell *shell; wl_shell_surface *shell_surface; #endif vk::SurfaceKHR surface; bool prepared; bool use_staging_buffer; bool use_xlib; bool separate_present_queue; vk::Instance inst; vk::PhysicalDevice gpu; vk::Device device; vk::Queue graphics_queue; vk::Queue present_queue; uint32_t graphics_queue_family_index; uint32_t present_queue_family_index; vk::Semaphore image_acquired_semaphores[FRAME_LAG]; vk::Semaphore draw_complete_semaphores[FRAME_LAG]; vk::Semaphore image_ownership_semaphores[FRAME_LAG]; vk::PhysicalDeviceProperties gpu_props; std::unique_ptr queue_props; vk::PhysicalDeviceMemoryProperties memory_properties; uint32_t enabled_extension_count; uint32_t enabled_layer_count; char const *extension_names[64]; char const *enabled_layers[64]; uint32_t width; uint32_t height; vk::Format format; vk::ColorSpaceKHR color_space; uint32_t swapchainImageCount; vk::SwapchainKHR swapchain; std::unique_ptr buffers; vk::Fence fences[FRAME_LAG]; uint32_t frame_index; vk::CommandPool cmd_pool; vk::CommandPool present_cmd_pool; struct { vk::Format format; vk::Image image; vk::MemoryAllocateInfo mem_alloc; vk::DeviceMemory mem; vk::ImageView view; } depth; static int32_t const texture_count = 1; texture_object textures[texture_count]; struct { vk::Buffer buf; vk::MemoryAllocateInfo mem_alloc; vk::DeviceMemory mem; vk::DescriptorBufferInfo buffer_info; } uniform_data; vk::CommandBuffer cmd; // Buffer for initialization commands vk::PipelineLayout pipeline_layout; vk::DescriptorSetLayout desc_layout; vk::PipelineCache pipelineCache; vk::RenderPass render_pass; vk::Pipeline pipeline; mat4x4 projection_matrix; mat4x4 view_matrix; mat4x4 model_matrix; float spin_angle; float spin_increment; bool pause; vk::ShaderModule vert_shader_module; vk::ShaderModule frag_shader_module; vk::DescriptorPool desc_pool; vk::DescriptorSet desc_set; std::unique_ptr framebuffers; bool quit; uint32_t curFrame; uint32_t frameCount; bool validate; bool use_break; bool suppress_popups; uint32_t current_buffer; uint32_t queue_family_count; }; #if _WIN32 // Include header required for parsing the command line options. #include Demo demo; // MS-Windows event handling function: LRESULT CALLBACK WndProc(HWND hWnd, UINT uMsg, WPARAM wParam, LPARAM lParam) { switch (uMsg) { case WM_CLOSE: PostQuitMessage(validation_error); break; case WM_PAINT: demo.run(); break; case WM_GETMINMAXINFO: // set window's minimum size ((MINMAXINFO *)lParam)->ptMinTrackSize = demo.minsize; return 0; case WM_SIZE: // Resize the application to the new window size, except when // it was minimized. Vulkan doesn't support images or swapchains // with width=0 and height=0. if (wParam != SIZE_MINIMIZED) { demo.width = lParam & 0xffff; demo.height = (lParam & 0xffff0000) >> 16; demo.resize(); } break; default: break; } return (DefWindowProc(hWnd, uMsg, wParam, lParam)); } int WINAPI WinMain(HINSTANCE hInstance, HINSTANCE hPrevInstance, LPSTR pCmdLine, int nCmdShow) { // TODO: Gah.. refactor. This isn't 1989. MSG msg; // message bool done; // flag saying when app is complete int argc; char **argv; // Use the CommandLine functions to get the command line arguments. // Unfortunately, Microsoft outputs // this information as wide characters for Unicode, and we simply want the // Ascii version to be compatible // with the non-Windows side. So, we have to convert the information to // Ascii character strings. LPWSTR *commandLineArgs = CommandLineToArgvW(GetCommandLineW(), &argc); if (nullptr == commandLineArgs) { argc = 0; } if (argc > 0) { argv = (char **)malloc(sizeof(char *) * argc); if (argv == nullptr) { argc = 0; } else { for (int iii = 0; iii < argc; iii++) { size_t wideCharLen = wcslen(commandLineArgs[iii]); size_t numConverted = 0; argv[iii] = (char *)malloc(sizeof(char) * (wideCharLen + 1)); if (argv[iii] != nullptr) { wcstombs_s(&numConverted, argv[iii], wideCharLen + 1, commandLineArgs[iii], wideCharLen + 1); } } } } else { argv = nullptr; } demo.init(argc, argv); // Free up the items we had to allocate for the command line arguments. if (argc > 0 && argv != nullptr) { for (int iii = 0; iii < argc; iii++) { if (argv[iii] != nullptr) { free(argv[iii]); } } free(argv); } demo.connection = hInstance; strncpy(demo.name, "cube", APP_NAME_STR_LEN); demo.create_window(); demo.init_vk_swapchain(); demo.prepare(); done = false; // initialize loop condition variable // main message loop while (!done) { PeekMessage(&msg, nullptr, 0, 0, PM_REMOVE); if (msg.message == WM_QUIT) // check for a quit message { done = true; // if found, quit app } else { /* Translate and dispatch to event queue*/ TranslateMessage(&msg); DispatchMessage(&msg); } RedrawWindow(demo.window, nullptr, nullptr, RDW_INTERNALPAINT); } demo.cleanup(); return (int)msg.wParam; } #elif __linux__ #if defined(VK_USE_PLATFORM_WAYLAND_KHR) static void handle_ping(void *data, wl_shell_surface *shell_surface, uint32_t serial) { wl_shell_surface_pong(shell_surface, serial); } static void handle_configure(void *data, wl_shell_surface *shell_surface, uint32_t edges, int32_t width, int32_t height) {} static void handle_popup_done(void *data, wl_shell_surface *shell_surface) {} static const wl_shell_surface_listener shell_surface_listener = { handle_ping, handle_configure, handle_popup_done}; #endif int main(int argc, char **argv) { Demo demo; demo.init(argc, argv); #if defined(VK_USE_PLATFORM_XLIB_KHR) && defined(VK_USE_PLATFORM_XCB_KHR) if (demo.use_xlib) { demo.create_xlib_window(); } else { demo.create_xcb_window(); #elif defined(VK_USE_PLATFORM_XCB_KHR) demo.create_xcb_window(); #elif defined(VK_USE_PLATFORM_XLIB_KHR) demo.create_xlib_window(); #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) demo.create_window(); #endif } demo.init_vk_swapchain(); demo.prepare(); #if defined(VK_USE_PLATFORM_XLIB_KHR) && defined(VK_USE_PLATFORM_XCB_KHR) if (demo.use_xlib) { demo.run_xlib(); } else { demo.run_xcb(); #elif defined(VK_USE_PLATFORM_XCB_KHR) demo.run_xcb(); #elif defined(VK_USE_PLATFORM_XLIB_KHR) demo.run_xlib(); #elif defined(VK_USE_PLATFORM_WAYLAND_KHR) demo.run(); #endif } demo.cleanup(); return validation_error; } #else #error "Platform not supported" #endif