Bug 1217282 - kernel-default-6.6.1 fails to boot
Summary: kernel-default-6.6.1 fails to boot
Status: RESOLVED FIXED
Alias: None
Product: openSUSE Tumbleweed
Classification: openSUSE
Component: Kernel (show other bugs)
Version: Current
Hardware: Other Other
: P5 - None : Critical (vote)
Target Milestone: ---
Assignee: openSUSE Kernel Bugs
QA Contact: E-mail List
URL: https://gitlab.freedesktop.org/drm/in...
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2023-11-17 12:35 UTC by Tristan Miller
Modified: 2024-06-25 18:01 UTC (History)
2 users (show)

See Also:
Found By: ---
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---


Attachments
Output of hwinfo (2.13 MB, text/plain)
2023-11-17 12:35 UTC, Tristan Miller
Details
System log for kernel 6.6.1 (213.30 KB, text/x-log)
2023-11-17 14:10 UTC, Tristan Miller
Details

Note You need to log in before you can comment on or make changes to this bug.
Description Tristan Miller 2023-11-17 12:35:07 UTC
Created attachment 870822 [details]
Output of hwinfo

Today I upgraded from openSUSE-release 20231108-2584.1 with kernel-default-6.5.9-1.3 to openSUSE-release 20231115-2592.1 with kernel-default-6.6.1-1.1.

I'm unable to boot with 6.6.1.  Almost as soon as it's selected from the GRUB menu, my monitors lose the video signal and the computer becomes unresponsive, though the hard drive light continues to flash for about half a minute.  The only thing I can do is to power-cycle the machine and reboot with an older kernel.  As far as I can tell, no logs are generated by the failed boot.

Attached is the output of hwinfo.
Comment 1 Takashi Iwai 2023-11-17 13:01:05 UTC
What if you boot with nomodeset boot option?  Does it continue to boot, at least?
Comment 2 Tristan Miller 2023-11-17 13:38:21 UTC
Yes, nomodeset allows me to boot into a text console.
Comment 3 Takashi Iwai 2023-11-17 13:44:57 UTC
So it's likely a regression of i915 driver.

Could you reload i915 driver forcibly with modeset=1 option on the system with nomodeset boot?
  modprobe -r i915
  modprobe i915 modeset=1

Does this also lead to a graphics problem, too?  You should be able to catch more kernel logs with this, so please attach the kernel log after reloading the i915 driver.
Comment 4 Tristan Miller 2023-11-17 14:10:47 UTC
Created attachment 870823 [details]
System log for kernel 6.6.1
Comment 5 Tristan Miller 2023-11-17 14:11:24 UTC
Yes, the screen goes blank and the system becomes unresponsive to keyboard input upon manually loading the i915 driver.  I've attached the full system log.
Comment 6 Takashi Iwai 2023-11-17 14:20:01 UTC
OK, it shows lots of kernel WARNINGs from i915.

At best, please report it to the upstream, gitlab.freedesktop.org Issues DRM/Intel, together with the obtained log.
Comment 7 Tristan Miller 2023-11-17 16:35:08 UTC
I'm able to reproduce the issue also with the 6.7-rc1 kernel.

As requested, I've filed an upstream bug report, including a system log with debug information: https://gitlab.freedesktop.org/drm/intel/-/issues/9686
Comment 8 Tristan Miller 2023-11-20 12:32:17 UTC
Upstream https://gitlab.freedesktop.org/drm/intel/-/issues/9686 was marked as a duplicate of https://gitlab.freedesktop.org/drm/intel/-/issues/9636 for which a patch https://gitlab.freedesktop.org/drm/intel/uploads/fcccbfe2833d6f4679dff3608c735ee2/0001-drm-i915-Also-check-for-VGA-converter-in-eDP-probe.patch was posted a few days ago.  Applying the patch to 6.6.1 works around the problem for me.  Can we get this patch into the Tumbleweed kernel-default package?
Comment 9 Takashi Iwai 2023-11-20 16:05:48 UTC
Thanks.

I backported the fix patch and pushed to stable/for-next.  It'll be included likely in a release after 6.6.2 (which was already submitted today).

A test kernel with the fix patch is also being built in OBS home:tiwai:bsc1217282 repo.
Comment 10 Takashi Iwai 2023-11-27 14:19:51 UTC
The patch was merged to stable tree for TW.  I close the bug for now.
Feel free to reopen if the kernel still shows the very same problem.
Comment 11 Tristan Miller 2023-12-07 14:12:45 UTC
Yes, everything's working fine now; thanks.