Linear tiling is faster for small, frequently updated textures
When lots of textures are updated each frame, a skipped frame will
destroy/recreate all textures, causing another skipped frame, etc. So
in-flight texture tracking is disabled, except for RTT textures
Limit swap chain to 2 images
Refactor base and oit renderers to avoid code dup
setImageLayout was missing TransferSrcOptimal as source layout
Vulkan needs the first vertex to be the provoking vertex so copy 3rd on
1st when flat shading.
VRAM stride was incorrect (RTT to VRAM)
Keep in-use imave and image views when the texture is updated.
Fix inside clipping (18 wheeler rear mirror)
Wait until the graphics queue is idle before changing shared buffers and
attachments.
Refactoring
stencil must be reset to zero if poly isn't shadowed -> fixes jet set
radio "see-through" shadows
take culling into account for modvols
oit: reduce per-pixel and per-poly mem usage
oit: framebuffers weren't resized when needed
oit: poly with depth write disabled weren't drawn (Soul calibur missing
water surface)
Fix glsl error when two-volume mode needs fog attachment
BufferData was always allocated in host visible mem, regardless of the
mem property argument.
NVidia gpus don't support blitting to 565 and 5551, which is needed for
mipmaps, so convert to 8888 instead.
Use optimal tiling for oit pointers image (linear not supported by
nvidia and probably slower anyway)
oit screen drawer uses multiple final attachments
Set non-oit depth attachment as transient and lazily allocate mem
Flyinghead