FS#80205 - amdgpu hard lockup when applying any clock, voltage, or memory adjustment after kernel 6.6.1 update

Attached to Project: Arch Linux
Opened by Sean Kloos (guava) - Thursday, 09 November 2023, 06:23 GMT
Last edited by Toolybird (Toolybird) - Thursday, 09 November 2023, 19:56 GMT
Task Type Bug Report
Category Kernel
Status Closed
Assigned To No-one
Architecture x86_64
Severity Low
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

OS: Arch Linux Kernel 6.6.1
CPU: Intel 12700k
GPU: AMD RX 6900 XT
Motherboard: ASUS Z690P BIOS 2802

Steps to reproduce:
Use Kernel 6.6.1 and Linux_Firmware 20231030.2b304bfe-1
Add "amdgpu.ppfeaturemask=0xffffffff" to kernel command line
apply any setting to pp_od_clk_voltage, for example "echo 's 1 2000' > p_od_clk_voltage; echo 'c' > pp_od_clk_voltage"

other kernel command line options set:
intel_iommu=on iommu=pt lsm=landlock,lockdown,integrity,apparmor,bpf lockdown=confidentiality audit=1 audit_backlog_limit=8192

Workaround:
use lts kernel 6.1.61
This task depends upon

Closed by  Toolybird (Toolybird)
Thursday, 09 November 2023, 19:56 GMT
Reason for closing:  Upstream
Additional comments about closing:  Please refer to the comments section
Comment by loqs (loqs) - Thursday, 09 November 2023, 08:56 GMT
See Kernel Debugging regressions [1] and report amdgpu issues on [2].

[1]: https://wiki.archlinux.org/title/Kernel#Debugging_regressions
[2]: https://gitlab.freedesktop.org/drm/amd/-/issues
Comment by Toolybird (Toolybird) - Thursday, 09 November 2023, 19:55 GMT
Clearly an upstream kernel regression. @loqs has already guided you in the right direction. "git bisect" will aid in identifying the causal commit for reporting to upstream. Please visit the Arch support channels (Forum, etc) if you need help.

Loading...