Bug 199585 - AMDGPU related kernel dump in /home/kernel/COD/linux/mm/slub.c:296 after monitors power off
Summary: AMDGPU related kernel dump in /home/kernel/COD/linux/mm/slub.c:296 after moni...
Status: RESOLVED UNREPRODUCIBLE
Alias: None
Product: Drivers
Classification: Unclassified
Component: Video(DRI - non Intel) (show other bugs)
Hardware: x86-64 Linux
: P1 high
Assignee: drivers_video-dri
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2018-05-02 05:19 UTC by Jason Hicks
Modified: 2018-07-15 10:32 UTC (History)
3 users (show)

See Also:
Kernel Version: 4.16.6
Subsystem:
Regression: Yes
Bisected commit-id:


Attachments
dmesg (156.65 KB, text/plain)
2018-05-02 05:19 UTC, Jason Hicks
Details
kernel bug (6.43 KB, text/plain)
2018-05-02 05:19 UTC, Jason Hicks
Details
lspci -vk (27.61 KB, text/plain)
2018-05-02 05:22 UTC, Jason Hicks
Details
inxi -F (5.83 KB, application/octet-stream)
2018-05-02 05:23 UTC, Jason Hicks
Details
xrandr (1.62 KB, text/plain)
2018-05-02 05:30 UTC, Jason Hicks
Details
Xorg.0.log (81.02 KB, text/plain)
2018-05-02 05:32 UTC, Jason Hicks
Details
dmesg - second lock up (2.50 MB, text/plain)
2018-05-02 06:04 UTC, Jason Hicks
Details

Description Jason Hicks 2018-05-02 05:19:00 UTC
Created attachment 275707 [details]
dmesg

Using three monitors, one connected via display port, one DVI, and one HDMI. After they reach the display timeout and power off, the machine hard locks leaving a kernel bug stacktrace.
Comment 1 Jason Hicks 2018-05-02 05:19:26 UTC
Created attachment 275709 [details]
kernel bug
Comment 2 Jason Hicks 2018-05-02 05:22:01 UTC
Created attachment 275711 [details]
lspci -vk
Comment 3 Jason Hicks 2018-05-02 05:23:03 UTC
Created attachment 275713 [details]
inxi -F
Comment 4 Jason Hicks 2018-05-02 05:30:30 UTC
Created attachment 275715 [details]
xrandr
Comment 5 Jason Hicks 2018-05-02 05:32:42 UTC
Created attachment 275717 [details]
Xorg.0.log
Comment 6 Jason Hicks 2018-05-02 06:04:33 UTC
Created attachment 275719 [details]
dmesg - second lock up

Disabled DC mode, updated grub, rebooted, and then set the monitor power off timeout to 5 minutes. Log didn't capture a full stack trace this time, but the error looks different. Will go back to kernel 4.15 and see if the issue still occurs.
Comment 7 Jason Hicks 2018-05-02 06:25:13 UTC
I can confirm this is not an issue on kernel 4.15, but now I'm running into an unrelated issue with the EDID information from the DVI monitor being incorrect. Switch the type to regression.
Comment 8 Christian König 2018-05-04 15:08:36 UTC
Looks like a double free in the DC code to me. Question is does it works with the latest code on amd-staging-drm-next ?
Comment 9 Jason Hicks 2018-07-15 10:32:47 UTC
Not able to reproduce this freezing issue anymore on kernel 4.17.5 with the same hardware.

Note You need to log in before you can comment on or make changes to this bug.