The stat counter struct no longer has to be passed to
the pipeline compiler function.
The new implementation uses atomic counters of the pipeline manager
rather than per-command list counters, which removes the need to
pass the counter structure to the compiler function.
We don't support rasterization with a sample count different from
the framebuffer sample count, but if there are no attachments, any
sample count is allowed.
Adding an extension will now require an additional step, but
this evolved version of the extension list code is more sane
and the structures can be more easily copied around.
Works around an issue with some games not setting the D3D11 depth
bias state correctly, which can result in an excessive number of
pipelines being compiled.
This should come in handy when compiling an optimized version of
a pipeline asynchronously. This can be extended to hold multiple
pipeline handles, i.e. one optimized one and one without opts.
Collateral damage: We're not using derivative pipelines anymore,
needs to be re-added at a later point.
Drivers from both major vendors implement their own shader cache
already, and storing a cache per game causes more issues than it
solves. Should fix#261.
* [util] Adds getTempDirectory() function
Will be used by on-disk pipeline caching
* [dxvk] Implement on-disk shader caching
Saving the pipeline cache to disk when the application exits
should be sufficient but the DxvkPipelineCache destructor isn't
reliably called on exit (ref-counting issue?).
As a workaround every frame we check and save the cache if the
size increased by some amount or after one minute elapsed.
* [dxvk] Periodically update shader cache file in separate thread
Reduces the CPU overhead of descriptor set updates, which usually
happen once per draw call. Gains seem to be minor in most games,
some outliers show significantly better performance (i.e. Tomb Raider).
When using DXVK_LOG_LEVEL=debug, the graphics pipeline manager now
prints the names of the shaders used by the pipeline. This shall
help debug driver crashes in vkCreate*Pipelines.
Checks whether all input slots consumed by the vertex shader
are provided by the input layout, and disables rendering in
case the state validation fails. This should hopefully fix
GPU lockups in Nier:Automata.
