GB205M [RTX 5070 Ti Mobile] - GSP RM heartbeat timeout on cold boot after long power-off

NVIDIA Open GPU Kernel Modules Version

595.58.03-1 (CachyOS package: nvidia-utils 595.58.03-1)

Please confirm this issue does not happen with the proprietary driver (of the same version). This issue tracker is only for bugs specific to the open kernel driver.

• I confirm that this does not happen with the proprietary driver package.

Operating System and Version

CachyOS

Kernel Release

6.19.10-1-cachyos

Please confirm you are running a stable release kernel (e.g. not a -rc). We do not accept bug reports for unreleased kernels.

• I am running on a stable kernel release.

Hardware: GPU

GPU 0: NVIDIA GeForce RTX 5070 Ti Laptop GPU (UUID: GPU-68965bb5-bf7e-9743-705b-8bbfff0d93a8)

Describe the bug

After a long power-off period (overnight or several hours), the NVIDIA GB205M fails to initialize on cold boot due to a GSP RM heartbeat timeout. The GPU sends no heartbeats at all (heartbeat value remains 0), causing the driver to time out and the display compositor (KWin) to lose its DRM output configuration, resulting in a complete system freeze requiring a hard reset.

After the hard reset and reboot, the system boots and runs normally for the rest of the day with no further issues. The problem only occurs on the first boot after a long power-off.

The issue also occurs after long suspend/sleep sessions where the GPU enters a deep power state.

To Reproduce

1. Fully shut down the system
2. Leave powered off for 4+ hours (overnight is most reliable)
3. Power on — system boots, NVIDIA GSP fails to initialize
4. System freezes within ~60-90 seconds of reaching the desktop
5. Hard reset required
6. On second boot the system works normally

Bug Incidence

Always

nvidia-bug-report.log.gz

nvidia-bug-report.log.gz

More Info

What does NOT fix the issue:

• nvidia.NVreg_EnableGpuFirmware=0 kernel parameter — has no effect on GB205M/Blackwell, GSP appears mandatory on this architecture
• nvidia.NVreg_PreserveVideoMemoryAllocations=1 — does not prevent the cold boot failure
• pcie_port_pm=off and pcie_aspm=off — no effect
• processor.max_cstate=5 — no effect
• Blacklisting i915/xe Intel GPU drivers — no effect
• Early KMS NVIDIA module loading via mkinitcpio — no effect
• Enabling nvidia-persistenced, nvidia-suspend, nvidia-resume, nvidia-hibernate services — no effect
• Switching from dGPU-only BIOS mode to MSHybrid — reduces frequency but does not eliminate the issue
• Driver versions tested: 595.45.04, 595.58.03, 590.48.01 — All affected

Start with the attached nvidia-bug-report.log.gz and reproduce the cold-boot or long-suspend failure using the listed hardware, kernel, and driver versions. Trace the GSP RM heartbeat timeout and KWin freeze through the open kernel module sources; done would mean the GPU initializes reliably without the heartbeat timeout or required hard reset. No source file or test is named in the report.