The existing implementation only supports 64 invoc-per-subgroup GPUs, and misbehaves on adreno when invocations need to be split into 4 emulated subgroups.

Reuses most of the existing code for generating the gl_Layer passthrough. Fixes geometry in Nier: Automata on GPUs without HW passthrough support.

We can take advantage of SSBO addresses being passed in a constant bufer to account for the extra alignment requirements in the shader itself.

Adreno drivers will crash compiling geometry shaders if the input position is not wrapped in a gl_in struct.

video_core/vulkan: Implemented `VkPipelineCache` to store Vulkan pipelines

So that old cache can get deleted when the cache version changes and does not grow infinitely

The addition of the use_vulkan_driver_pipeline_cache option into the default ini string literal caused the 16,384-byte limit of the MSVC compiler to be exceeded.

As an optional feature which can be enabled in the advanced graphics configuration, all pipelines that get built at the initial shader loading are stored in a VkPipelineCache object and are dumped to the disk.

These vendor specific pipeline cache files are located at `/shader/GAME_ID/vulkan_pipelines.bin`. This feature was mainly added because of an issue with the AMD driver (see yuzu-emu#8507) causing invalidation of the cache files the driver builds automatically.

Vulkan: Fix drivers that don't support dynamic_state_2 up

Yuzu Fried Chicken Part 1.5: MacroHLE Rework and Dynamic State

Revert "k_page_group: synchronize"

TAS: Immediately switch stick to TAS on input

Co-Authored-By: Narr the Reg <5944268+german77@users.noreply.github.com>

cmake: move find-modules to root cmake dir

cmake: allow options shadowing with normal variables

TAS: Record sanitized instead of raw stick inputs

cmake: improve find_package failure messages

service: nifm: Initialize request state