The source region size doesn't have to be aligned to the destination
block size. We should be checking whether the destination region is
aligned to the destination block size.
As with cube textures, some games think it's a good idea to bind a
non-array view to a slot that expects an array view or vice versa.
Fixes clouds not appearing in TrackMania 2 Stadium (#191).
* [d3d11] add some resource validation for CopyResource and CopyResourceSubregion
combine if statement
* [d3d11] added copy extents validation for compressed formats
* correct return values
* fix incorrect logic operators
* set valid copy extents when possible
* [d3d11] Clamp copy region in CopySubresourceRegion
* [dxvk] Add helper methods to deal with block-compressed images
* [d3d11] Clean up validation in CopySubresourceRegion
* [d3d11] Improve error reporting and validation in CopyResource
* [d3d11] Fix inconsistent error messages
* Fix narrowing warnings in spirv_module relating to enum's default width on x64
* Make hashes of states use correct types without casting.
* Fix narrowing conversion in d3d11_sampler.cpp
Fixes invalid shaders being generated in the Blacksmith demo on
some GPUs. Works around a possible issue in the output signature
reader.
Commit #1000, yay.
Fixes a GPU hang when closing Dark Souls 3 as well as similar
undesired behaviour in other games that continue to use the
DXGI swap chain after the window has been destroyed.
We do not have to do this anymore since we'll bind a large
enough dummy buffer. Considerably reduces code size in shaders
which access a large number of shader constants.
We will use this extension in order to implement vertex
binding divisors other than 1 for per-instance attributes.
Will be *required* as soon as support by wine and Vulkan
drivers is widely available.
We no longer use the 'oldSwapchain' member in the swap chain description
to replace an existing swap chain, but rather destroy it altogether and
create a new swap chain. While this is less than optimal, it might help
solve some swap chain-related issues such as #277.
An alternative to manually creating a framebuffer object and binding
it via bindFramebuffer. Future optmizations can use this to bring
down the number of redundant render pass changes.
- Do not report MIP_AUTOGEN if the image format cannot
be used as a color attachment
- Do not report SAMPLE_COMPARISON and GATHER_COMPARISON
if the DXGI format has no corresponding depth format
- Only report image-related features if the image format
can actually be used as a sampled image
