This should help understand some virgl/ChromeOS-specific WineD3D
behavior better.

Based on a patch by Michael Müller.

This should more closely match the algorithm used by native, discovered through
extensive manual testing.

This does not include any automated tests, both because I could not get winmm to
recognize a second plugged joystick, and (as I eventually discovered while
testing) the number of tests that would be necessary to conclusively prove this
algorithm turned out to be impractical.

Ensures the Direct 2D lock is held before attempting to access Direct 2D
exclusive resources.

This fixes periodic crashes in PowerPoint 365

Tests that Direct2D can access D2D exclusive resources whilst the
Direct 2D lock is held if the factory type is single threaded

Tests that Direct2D can not access D2D exclusive resources whilst the
Direct 2D lock is held

This is a simple passthrough implementation to the native Vulkan driver.

This is a simple passthrough implementation to the native Vulkan driver,
with the addition of a VkSurfaceKHR invalidation check.

We cannot depend on the main surface being committed to after
the client subsurface is created, so perform a commit to ensure
the subsurface position takes effect on creation.

The client area subsurface (the target of Vulkan rendering) is not going
to be presented properly (or at all) by the compositor if its parent
surface is not mapped or doesn't have the right size, even though the
window may be visible from Wine's perspective. To avoid this issue,
ensure that the parent surface has up-to-date contents in terms of size,
in case these haven't been provided yet, or will never be provided
(e.g., when the Vulkan rendering is fullscreen).

If the VkSwapchainKHR extent does not match the target window client
area, report VK_ERROR_OUT_OF_DATE_KHR, to notify the app that it should
recreate the swapchain.

We also report VK_ERROR_OUT_OF_DATE_KHR if the HWND is no longer valid,
to match win32 behavior (e.g., nvidia). The application will eventually
get VK_ERROR_SURFACE_LOST_KHR during the swapchain recreation process.

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=55115

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=55536

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=55959

Floats with fractional part 0.5000 should be rounded to next even number.

Wine-Bug: https://bugs.winehq.org/show_bug.cgi?id=55185

This corresponds to vkd3d as of a52604da8cae8ce5639eeb20ef263c12bf70e1df commit.