Add a protocol-independent recovery channel outside LGMP so clients can
request a usable guest display even when transport versions differ.
Synchronize recovery state with the helper across driver restarts and
restore the configured topology without persisting temporary changes.
Define a fixed, protocol-independent recovery ABI in the final 64 KiB
of IVSHMEM.
Provide crash-safe multi-client requests and coherent producer status.
Exclude the region from LGMP allocation and frame capacity so future
protocol mismatches retain a stable control path.
Make CSRWLock own the native lock and provide scoped guards for shared,
exclusive, early-unlock, and non-blocking use.
Replace direct SRW lock management throughout the IDD, input driver,
and helper while preserving the existing lock scopes.
Expose a Consumer Control HID report for the system volume usages that
Windows does not handle through the Keyboard/Keypad usage page.
Translate the existing mute and volume keyboard usages at the LGInput
boundary, preserve normal keyboard state, and force a neutral consumer
report across resets and HID device reactivation.
Move named pipe writes off the LGMP input worker so a stalled LGInput
endpoint cannot block queue draining or lease maintenance.
Coalesce motion only under queue pressure while preserving mode, button,
wheel, and keyboard transitions. Reset HID state after discontinuities
and carry all 32 mouse button bits through the pipe and HID reports.
Add an optional input transport and a dedicated LGMP receiver that
validates source ownership, generations, ordered sequences, and leases.
Poll input outside the 10 ms control timer. Forward mouse and keyboard
state through the LGInput pipe, and neutralize the endpoint before
ownership changes or transport failures can leave input held.
Replace raw HID report forwarding with typed absolute mouse, relative
mouse, and keyboard messages.
Expose absolute and relative modes as separate Mouse collections because
Windows MouHID requires one X/Y motion mode per mouse device.
Move the reusable named-pipe endpoint and shared support code into
the LGCommon static library.
Run a dedicated server in LGIdd and a reconnecting client in
LGInput, with device-lifecycle handling and report framing.
Refactor the helper pipe to use the same endpoint implementation.
Build LGInput as an independent UMDF driver with its own entry point,
service, tracing, and binary.
Expose absolute pointer, relative mouse, and keyboard collections with a
guarded report queue.
Keep LGIdd as the startup and deployment project. Give it a non-linking
build/package dependency on LGInput so F5 stages both driver stacks and
installs them together through LGIddInstall, while the two UMDF binaries
remain independent for future IPC.
Stage both DLLs for the NSIS installer, retain the WDK UMDF remote-debug
startup attachment, and move CSRWLock into LGCommon for the input
driver's report queue.
Use an explicit placed buffer for software frame copies so the D3D copy
uses the same row pitch published through KVMFR.
GetCopyableFootprints describes buffer copies and cannot be used to infer
the physical layout of a placed row-major texture. This caused padded
resolutions to use the wrong row starts in every consumer.
Keep the copy direct to IVSHMEM without a staging or CPU copy.
Move cadence and pipeline declarations under capture, and move local
named-pipe handling under ipc.
Flatten display contexts and rename files after their contained classes.
Keep transport headers limited to transport contracts and capabilities.
Consolidate CSwapChainProcessor definitions so each source file matches
the class it implements.
Introduce transport, frame, and control interfaces with an LGMP factory
backend.
Move LGMP and IVSHMEM implementation details under transport/lgmp and
expose direct frame-buffer memory through a neutral capability.
Group the IDD sources and Visual Studio filters by subsystem.
Split the device and swap-chain implementations into focused units,
rename the context classes, and reduce header coupling.
Move cadence-aware hardware capture and immediate software capture
behind a common frame processor interface.
Keep shared damage tracking and frame-buffer ownership in the base
processor, and move stateless format, resource, and rectangle helpers
into CFrameProcessorUtil.
Decouple frame-buffer resources from CSwapChainProcessor by passing the
device dependencies they use directly.
Bypass cadence retention when the software render adapter is active.
Publish each available source frame immediately with one damage-aware
D3D copy into its final IVSHMEM resource.
Keep only one software copy in flight so newer frames are dropped
instead of queued behind stale work. Retain accumulated damage for the
next frame and skip static re-encodes when no image update is pending.
Use a row-major IVSHMEM texture when the shared heap supports it and
fall back to a direct IVSHMEM buffer copy otherwise. In indirect mode,
copy only damaged rows from readback memory into IVSHMEM.
Store refresh rates in millihertz and preserve three decimal places.
Advertise exact rational rates while accepting legacy integer values.
Accept rates from 23.900 through 1000.000 Hz.
Rebuild the driver's cached mode list before replugging the monitor so
Windows enumerates the settings saved by the helper.
Keep saved mode refresh rates intact and update only ExtraMode to the
configured default refresh.
Permit retained-buffer reuse when both private queues are occupied by a
previous timing owner, even if the new owner has no pending deliveries.
Use the shared fallback when possible. Otherwise retain the fresh frame
locally and republish it instead of waiting for stale private deliveries.
