Hardware:
- Laptop: Dell Alienware x14 (Service Tag: 665YHR3)
- GPU: NVIDIA GA106M [GeForce RTX 3060 Mobile / Max-Q] (PCI 0000:01:00.0, rev a1)
- CPU: Intel Core i7-12700H
- BIOS: 1.30.0 (latest available from Dell)
Software:
- OS: Debian GNU/Linux 12 (bookworm)
- Kernel: 6.1.0-50-amd64
- NVIDIA Driver: 535.261.03 (installed via Debian package nvidia-driver)
- DKMS: kernel module built and signed with enrolled MOK key
Problem:
Loading the proprietary NVIDIA driver causes a repeating kernel error loop:
NVRM: GPU 0000:01:00.0: RmInitAdapter failed! (0x24:0x72:1447)
This error repeats every ~13 seconds indefinitely, causing severe system instability. The GPU never initializes successfully. The system becomes unresponsive and requires a hard reset.
Steps to reproduce:
- Install nvidia-driver 535.261.03 on Debian bookworm
- Ensure nouveau is blacklisted
- Boot — RmInitAdapter failed (0x24:0x72:1447) begins looping immediately
What was tried:
- Multiple driver versions via DKMS
- Blacklisting nouveau before loading nvidia
- Disabling nvidia-persistenced
- Secure Boot enrolled MOK signing
- All combinations of kernel parameters
Current workaround:
Blacklisted the proprietary driver entirely. Running nouveau as the display driver. System is stable but CUDA, NVENC, and hardware acceleration are unavailable.
Blacklist config (/etc/modprobe.d/blacklist-nvidia.conf):
blacklist nvidia
blacklist nvidia_drm
blacklist nvidia_modeset
blacklist nvidia_uvm
blacklist nvidia_peermem
Additional context:
The nvidia-bug-report.sh tool is not runnable in the current state since the driver is blacklisted to maintain system stability. The error was captured from persistent journal logs from the period when the driver was active.
The nouveau driver loads cleanly on the same hardware with no errors, suggesting the hardware itself is functional.
Request:
Any guidance on what 0x24:0x72:1447 indicates internally, and whether there is a known fix or workaround for GA106M on this specific Alienware platform would be greatly appreciated.