What handling is expected for GPU ACPI Notify(0xC5) alongside NPCF Notify(0xC0)?

[LIghtJUNction]

I am investigating an ACPI notification on a MECHREVO Kuangshi16Pro Series GM6PX0X, BIOS N.1.09MRO07, Linux 7.2.9-arch1-1, NVIDIA open kernel modules 615.78.08. I have not compared the proprietary module, so this is a request for protocol guidance rather than an open-driver-only bug claim.

The firmware's power-budget update paths send Notify(NPCF, 0xC0) followed by Notify(the GPU ACPI device, 0xC5). In two captured update sequences, tracing showed the NPCF callback, the GPU callback with event 197, and nvidia-powerd's NVPCF DSM subfunctions 1 and 2, both returning NV_OK. The GPU notification reaches rm_acpi_notify's default branch and triggers the assertion at osapi.c:2198. The current upstream main ca2d03e still has the same handlers.

The independent NPCF path therefore appears to run, but that does not establish whether all budget changes were applied, or whether the GPU notification needs separate processing. Before and after these captures, the configured cTGP offset stayed at 25 and nvidia-smi's enforced/default/max power limits were 140/115/140 W. These are limit readings, not measured power or a performance test.

Could NVIDIA clarify:

1. Whether GPU event 0xC5 has a defined power-management meaning.
2. Whether it requires independent processing or acknowledgement, or is expected to coexist with NPCF 0xC0.
3. Which observable state should be checked to validate correct handling.

I have not mapped it to GPS 0xC0, forwarded an extra NVPCF event, or suppressed the assertion.

Relevant code:

• GPU ACPI notification handler.
• Independent NPCF callback.
• ACPI 6.6 device notifications reserve these values for vendor-specific use.

I can provide a narrowly scoped, redacted trace or AML excerpt if it would help; no raw machine logs are attached here.