Presents to the swapchain waiting on the device semaphore.

Same as above but recreate default pipeline and its bundled storages and handles.

Same as above, recreates vulkan framebuffers

Recreates all render passes associated with the pipeline for swapchain recreation.

Makes new semaphores for draw synchronization. Useful for swapchain recreation. Unsafe because they must not be in use.

Mark the image as in use by the given frame.

TODO(issue#2) PIPELINES handle when there is more than one pipeline. Save the handles to the base shaders so they don't have to be recreated for no re

When a memory barrier is inserted that transfers queue ownership, the accept end of the memory barrier must also be run in a command buffer of the que

Waits for all the in flight frames, ie device idle.

Checks if all the validation layers specified are supported supported in this machine.

Selects Mailbox if available, but if not tries to fallback to FIFO. See the [spec](https://renderdoc.org/vkspec_chunked/chap32.html#VkPresentModeKHR)

Selects the resolution of the swap chain images. This is almost always equal to the resolution of the Surface we're drawing too, but we need to double

If drawing to a surface, chooses the best format from the ones available for the surface. Tries to use B8g8r8a8Srgb format with SRGB_NONLINEAR colors

Must be called by the backend when cleaning up all resources, if they are managed by the application (as in Vulkan/D3D12). # Safety Unsafe because af

Creates the base pipeline for Sarekt. A user can load custom shaders, etc, to create custom pipelines (passed back as opaque handles) based off this

================================================================================ Pipeline Helper Methods =============================================

================================================================================ Instance Creation ===================================================

================================================================================ Logical Device Helper Methods =======================================

Creates command buffer for main thread to make draw calls on. TODO(issue#1) MULTITHREADING one secondary per thread.

================================================================================ Storage Creation Methods ============================================

================================================================================ Command & Descriptor Pool/Buffer Methods ============================

Given the render target images and format, create an image view suitable for rendering on. (one level, no mipmapping, color bit access).

Creates a shader store in the vulkan backend configuration to load and delete shaders from.

================================================================================ Presentation and Swapchain Helper Methods ===========================

Same as above but allows for MSAA to be used, useful when creating internal render targets.

It is invariant in the vulkan renderer setup that p_user_data is of type DebugUserData, it is set up in new.

Goes through and checks if the device supports all needed extensions for current configuration, such as swapchains when drawing to a window.

Finds the queue family indices to use for the rendering command submissions. Right now only picks the first suitable queue family for each type of co

================================================================================ Instance Helper Methods =============================================

Tells us if this device is compatible with Sarekt. This means it has what is needed by this configuration in terms of: Supported Queue Families (Graph

Load a buffer and allocate memory into the backend/GPU and return a handle.

I could create a buffer myself and allocate memory with VMA, but their recomended approach is to allow the library to create a buffer and bind the mem

Same as above but keeps the buffer in cpu accessible memory. Useful for things updated frequently like MVP uniforms.

The procedure for loading an image in vulkan could use a staging image, but its just as well we use a staging buffer, which is easier and [could even

Load a shader into the driver and return a handle.

Logs extensions that are available and what was requested.

Create with a group of methods to load/destroy shaders.

Creates an image and imageview pairing, with a Drop implementation. Unsafe because you must clean up the vk::Image and vk::ImageView still.