Give each in-flight framebuffer a complete post-processing chain and
its own COMPUTE recording slot while retaining one physical queue.
Claim framebuffer ownership before submitting compute work, and use
the frame index for both COMPUTE and COPY recording state. Drain
in-flight work before reconfiguring either chain so COPY never
references replaced effect resources.
Require both chains to contain the same effects, preserve pending
damage, and release ownership on every pre-copy failure path.
Replace the rotating D3D12 copy queues with two framebuffer-bound
recording slots on one physical COPY queue. Keep allocator, command
list, query range, callback state, and fence target independent per
slot.
Submit source waits, execution, and signaling on the shared timeline.
Track framebuffer ownership through completion and serialize LGMP
publication around the last successfully published frame.
Combine the previous and current frame damage before recording copy
commands. Remove contained regions, merge beneficial overlaps, and use
one full-frame copy when partial work reaches the full surface size.
This prevents a previous full-frame update from being copied again
alongside the current damage.
Use calibrated copy-queue timestamps to separate source and effect
waits from the actual framebuffer copy.
Exclude producer readiness waits from client import timing so the
same interval is not counted in both Copy and Import.
Treat IddCx's single empty dirty rectangle as no image update instead
of promoting it to full damage.
Preserve pending damage and format changes, and fall back to full damage
when legacy move regions are present.
Monitor-default HDR metadata describes the virtual display rather than
the captured content. Do not expose it through KVMFR or use it to select
downstream processing limits.
Only publish explicit NEW frame metadata and otherwise use the complete
PQ range internally.
Publish BT.2020 chromaticities in the virtual monitor EDID so
Windows receives a valid HDR colour volume.
Extend cursor messages with their IddCx SDR white level and an
explicit visibility-valid flag. Consumers can then calibrate cursors
without treating colour-transform-only messages as cursor hides.
Honor the IddCx 3x4 XYZ matrix and post-transfer LUT in the IDD
conversion pipeline. Carry transform changes to the client so EGL applies
the same calibration to hardware cursors, and invalidate the full frame
whenever calibration changes.
Handle DEFAULT, NEW, and UNCHANGED IDDCX HDR10 metadata states and
publish valid mastering metadata through KVMFR. Preserve the metadata
across post-processing without reallocating resources for
metadata-only changes.
Invalidate in-progress swap chain setup when IddCx unassigns the
monitor, and serialize replacement assignments so workers are
never started on abandoned handles.
Also initialize all worker dependencies before starting the processing
thread and treat teardown during SetDevice as normal cancellation.
Keep the replug lifecycle gated until the old swap chain has drained and
released its pending frame, and the replacement swap chain is fully
initialized. Coalesce additional requests into a follow-up replug to
prevent overlapping topology changes and IddCx release-frame timeout
bugchecks.
Drop frames when the LGMP queue is full instead of blocking
while holding an IddCx surface. Always finish every
acquired frame, including duplicate frame numbers, so
replug teardown cannot trigger the release-frame watchdog.
Submit the D3D12 copy before exposing the frame through LGMP, preventing
clients from waiting on work that has not yet been queued. Track the
last published buffer separately for safe subscriber resends and handle
submission failures.
While it would make much more sense to use these then a full replug to
change modes, Microsoft have not properly implemented the API to clear
the cached monitor mode states internally, making these calls useless.
Revert to just replugging the device on mode change
If the screen is blanked or removed through a user action
`IddCxMonitorQueryHardwareCursor` will return with a failure. This
is normal and we should not log this as an error.
It is invalid to call `IddCxMonitorSetupHardwareCursor` before
`IddCxSwapChainSetDevice`. This fixes this by moving the thread into
CSwapChainProcessor and starting it after `IddCxSwapChainSetDevice`
has succeeded.
The callback runs in a random thread, we can't call directx methods
safely from it, so move reset so it's called automatically when a free
copy list is obtained.
`IddCxSwapChainFinishedProcessingFrame` must be called after every
frame, but we should do it as early as possible once all commands
are queued that use the frame.
As the resource size can be larger then the actual frame data, we
need to track this seperately so that we don't waste cycles copying
data the client will never use.
