Skip to content

A Rust project that implements multi-threaded computation output from a Vulkan asynchronous queue with per-frame synchronization

License

Notifications You must be signed in to change notification settings

jknoptrix/VulkanAsyncQueue

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

19 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Vulkan Queue

This shit provides a simple and easy-to-use interface for working with Vulkan queues and swapchains in Rust.

Features

  • Asynchronous queue submission using Tokio
  • Easy swapchain creation and management
  • Support for multiple threads

Usage

You can create a new VulkanQueue instance and use it to submit command buffers to the graphics queue:

let queue = VulkanQueue::new(
    &instance,
    &device,
    physical_device,
    surface,
    &surface_loader,
    queue_index,
    buffer,
    frame_buffer,
    window_width,
    window_height
);

let command_generator = |i: usize| {
    // Generate command buffers here...
};

queue.submit_commands(1, command_generator).await.unwrap();

In this example, we create a new VulkanQueue instance and pass in the necessary parameters, such as the Vulkan instance and device handles, the physical device, the surface, and the surface loader. We also specify the queue family index, buffer, frame buffer, window width, and window height.

Then we create a closure that generates command buffers for submission to the graphics queue. This closure takes an index as an argument and returns a vector of command buffers. Then we call the submit_commands method on our VulkanQueue instance and pass in our command generator closure. This will submit our command buffers to the graphics queue for execution.

Swapchain

This bullshit also provides support for creating and managing swapchains. When you create a new VulkanQueue instance, a default swapchain is automatically created for you using the specified parameters.

You can access the swapchain images for rendering by using the swapchain.images field of your VulkanQueue instance:

let framebuffers: Vec<vk::Framebuffer> = queue.swapchain
    .images
    .iter()
    .map(|&image| {
        // create framebuffers here...
    })
    .collect();

Here we iterating over the swapchain images and use them to create framebuffers for rendering.

Once you’ve rendered your scene to a swapchain image, you can present it to the surface using the queue_present_khr method of the Swapchain object:

let present_info = vk::PresentInfoKHR::builder()
    .wait_semaphores(&[semaphore])
    .swapchains(&[queue.swapchain.handle])
    .image_indices(&[image_index])
    .build();

unsafe {
    queue.swapchain.loader.queue_present(queue.graphics_queue, &present_info).unwrap();
}

PresentInfoKHR struct specifies which swapchain and image to present, as well as which semaphore to wait on before presenting. Then we call the queue_present_khr method on our Swapchain object to present the specified image to the surface.

Examples

Commands (rendering triangles):

let pipeline = create_pipeline(device, swapchain.image_format);
let command_generator = |i: usize| {
    let command_buffer = create_command_buffer(device, command_pool);

    let begin_info = vk::CommandBufferBeginInfo::builder()
        .flags(vk::CommandBufferUsageFlags::ONE_TIME_SUBMIT)
        .build();
    unsafe {
        device
            .begin_command_buffer(command_buffer, &begin_info)
            .expect("Failed to begin recording Command Buffer at beginning!");
    }

    let clear_values = [vk::ClearValue {
        color: vk::ClearColorValue {
            float32: [0.0, 0.0, 0.0, 1.0],
        },
    }];

    let render_pass_begin_info = vk::RenderPassBeginInfo::builder()
        .render_pass(render_pass)
        .framebuffer(framebuffers[i])
        .render_area(vk::Rect2D {
            offset: vk::Offset2D { x: 0, y: 0 },
            extent: swapchain.extent,
        })
        .clear_values(&clear_values)
        .build();

    unsafe {
        device.cmd_begin_render_pass(
            command_buffer,
            &render_pass_begin_info,
            vk::SubpassContents::INLINE,
        );
        device.cmd_bind_pipeline(command_buffer, vk::PipelineBindPoint::GRAPHICS, pipeline);
        device.cmd_draw(command_buffer, 3, 1, 0, 0);
        device.cmd_end_render_pass(command_buffer);
        device
            .end_command_buffer(command_buffer)
            .expect("Failed to record Command Buffer at Ending!");
    }

    vec![command_buffer]
};
queue.submit_commands(1, command_generator).await.unwrap();

Framebuffers using swapchain:

You can extend it by yourself, but you can already call it and 'setting this shit up' the queue settings:

let framebuffers: Vec<vk::Framebuffer> = swapchain
    .images
    .iter()
    .map(|&image| {
        let image_view_create_info = vk::ImageViewCreateInfo::builder()
            .view_type(vk::ImageViewType::TYPE_2D)
            .format(swapchain.image_format)
            .components(vk::ComponentMapping {
                r: vk::ComponentSwizzle::IDENTITY,
                g: vk::ComponentSwizzle::IDENTITY,
                b: vk::ComponentSwizzle::IDENTITY,
                a: vk::ComponentSwizzle::IDENTITY,
            })
            .subresource_range(vk::ImageSubresourceRange {
                aspect_mask: vk::ImageAspectFlags::COLOR,
                base_mip_level: 0,
                level_count: 1,
                base_array_layer: 0,
                layer_count: 1,
            })
            .image(image)
            .build();

        let image_view = unsafe {
            device
                .create_image_view(&image_view_create_info, None)
                .unwrap()
        };

        let framebuffer_create_info = vk::FramebufferCreateInfo::builder()
            .render_pass(render_pass)
            .attachments(&[image_view])
            .width(swapchain.extent.width)
            .height(swapchain.extent.height)
            .layers(1)
            .build();

        unsafe { device.create_framebuffer(&framebuffer_create_info, None).unwrap() }
    })
    .collect();

// ...
let present_info = vk::PresentInfoKHR::builder()
    .wait_semaphores(&[semaphore])
    .swapchains(&[swapchain.handle])
    .image_indices(&[image_index])
    .build();

unsafe {
    swapchain.loader.queue_present(queue.graphics_queue, &present_info).unwrap();
}

Resource managment:

let vertex_data = [/* ... */];
let buffer_size = (vertex_data.len() * std::mem::size_of::<Vertex>()) as u64;

let (vertex_buffer, vertex_buffer_memory) = queue.resource_manager.create_buffer(
    buffer_size,
    vk::BufferUsageFlags::VERTEX_BUFFER,
    vk::MemoryPropertyFlags::HOST_VISIBLE | vk::MemoryPropertyFlags::HOST_COHERENT,
);

let data_ptr = unsafe {
    device
        .map_memory(vertex_buffer_memory, 0, buffer_size, vk::MemoryMapFlags::empty())
        .unwrap()
};
unsafe {
    std::ptr::copy_nonoverlapping(vertex_data.as_ptr() as *const u8, data_ptr, buffer_size as usize);
    device.unmap_memory(vertex_buffer_memory);
}

Releases

No releases published

Packages

No packages published

Languages