* Unlock images correctly on Err
If the CommandBuffer encountered an error in the lock_submit function,
it would unlock all the previously locked resources and return the
error. It would unlock all the images with a None value which would
panic in an assertion in the unlock method of the images. Now it checks
for the final_layout and unlocks the image this way.
This change should help with the panics mentioned in #1441 and #1447.
With this fix it should return the error that is occuring instead of
panicking.
* Add fix to the change log to inform users.
* Implement caching for GraphicsPipeline
Added the method build_with_cache to set an Option that holds an Arc
containing a PipelineCache. This could probably be optimized to just
hold a vk::PipelineCache(u64) and start with a vk::NULL_HANDLE and
set the value to the internal_object of the PipelineCache. Using the Arc
is probably safer for now, but it is an idea.
* Add changes to the CHANGELOG
* Added a cache option to the ComputePipeline
This is a breaking change as there is no ComputePipelineBuilder and the
cache has to be added in the new function and the underlying functions
that build the ComputePipeline.
This needs to be discussed.
* Document the cache feature for the ComputePipeline
* Formatting with cargo fmt
* Remove outdated documentation
* Added testing for the PipelineCache
Checking that the cache returns the same data if the get_data method is
invoked twice without building a new pipeline.
Checking that the cache returns the same data if the same pipeline is
built two times with the cache.
Checking that the cache returns different data if there is a different
pipeline built in between the get_data calls.
* Create an example for the PipelineCache
* Formatting
* Move SubpassContents to command_buffer module, use it as parameter type instead of bool, add RenderPassState internally
* Add some safety checks to execute_commands
* Move Kind and related types to top-level module, add CommandBuffer::kind trait method, keep Kind value in AutoCommandBuffer(Builder)
* Some minor code cleanup
* Documentation update
* Use winit's hinstance() function
* Require HasRawWindowHandle for the window and remove handle parameter
Co-authored-by: Antonino Siena <a.siena@gmx.de>
* Feature: Create a Surface from a RawWindowHandle
* Comment on the new features
* Updated changelog
* Corrected feature name
* Compiler error corrections
* Empty commit for CI
Co-authored-by: Antonino Siena <a.siena@gmx.de>
* add additional ImageUsage constructors
* update examples to use ImageUsage::color_attachment instead of caps.supported_usage_flags
* fix additional usage in multi-window example
* update ignore to not include mandelbrot.png and triangle.png
* update ignore to not include mandelbrot.png and triangle.png
This feature was recently promoted from an extension to the core spec in Vulkan 1.2. It is already supported on many recent drivers: NVidia, Mesa RADV, and AMDVLK.
Other than supporting enabling the feature and the basic buffer address query, I made no attempt to create safe interfaces for its use. Such will come later, after I've experimented with its use in Geobacter.
* Change documentation.
Changed the documentation to explain how the data layout of an image is changed as it is copied from image to buffer.
* Fixed the trailing whitespace
Not sure how widespread this issue is, I'm on Mesa 20.0 currently. I've re-enabled a heap size check only when the reported size is != 0, because MoltenVK apparently has an issue with the reported heap sizes.
Without this, an application has to create their own loader for the
Vulkan function pointers if they need to drop down to raw Vulkan
calls. It's not a lot of effort to load them, but it's nicer to just
have them available on the Instance or Device rather than having to
pass an extra parameter around for function pointer access.
Co-authored-by: Austin Johnson <me@austinj.work>
* Use traingle example as template
* Remove triangle example comments
* Switch to CpuBufferPool chunks instead of CpuAccessibleBuffer for vertex data
* rustfmt
* Keep unwrap()s on same line
Almost none of the options are valid anymore. The few that are, seem to be default. I move to delete the .rustfmt.toml completely, to allow rustfmt to be run without config errors.
* implement exclusive fullscreen
* fullscreen ext should be device ext
* add acquire/release
* require VK_KHR_get_surface_capabilities2
* fullscreen_exclusive is now an enum; explicit acquire/release methods added; FullscreenExclusiveError added to handle errors; FullscreenExclusiveLost varients added to acquire and present related errors
* add checks for double acquire/release; require AppControlled for acquire/release functions
* update examples and tests; resolve issue with try_gpu_lock() on SwapchainImage; update changelogs
* Add Eq and Hash implementations to various buffer and image types
* Changelog entry
* Add Eq and Hash to graphics pipelines too
* Changelog entry
* Remove Eq and Hash from dyn GraphicsPipelineAbstract, as the device cannot be checked
* Add Eq and Hash to GraphicsPipelineAbstract, include size in buffer equality
* Add Eq and Hash to descriptor sets
* Changelog entry
* Add Eq and Hash to Instance, check for it in Device
