Thanks for the follow-up! The next time I reboot I will provide the full log up to the crash.  **It's almost fully reproducible.**   For now I'm reproducing it with the behaviour of the gnome automounting, as soon as i plug the 5-disk dock (with btrfs raid-1) the crash happens at the 3rd or 4th mount. But some intermediate kernels fail to start the graphical session, so I think I need a short script to reproduce it in multiuser mode without X/wayland.

It was OK with 6.14 rc1. The crash seems to start in the middle of the 6.14 series.  It's still crashing with 7.0.9. I've not yet rebooted to 7.0.10.

Last week I attempted a bisect using stock kernel, but it takes time...

I'm at this point:

$ git bisect log
git bisect start
# status: waiting for both good and bad commits
# good: [bf89928ffeb731c623a15ee7327261131a80ddcb] Linux 6.12.86
git bisect good bf89928ffeb731c623a15ee7327261131a80ddcb
# status: waiting for bad commit, 1 good commit known
# bad: [fd590381da18a801b7fea1a258e1760335607b10] Linux 6.16.3
git bisect bad fd590381da18a801b7fea1a258e1760335607b10
# good: [adc218676eef25575469234709c2d87185ca223a] Linux 6.12
git bisect good adc218676eef25575469234709c2d87185ca223a
# skip: [e50da555ca4d42b1b98d0f26789db64f26a0919a] Merge tag 'sound-6.15-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/tiwai/sound
git bisect skip e50da555ca4d42b1b98d0f26789db64f26a0919a
# good: [9b2955bae7025190c8d15983b37554da96e43e21] arm64: dts: qcom: pm660l: add flash leds
git bisect good 9b2955bae7025190c8d15983b37554da96e43e21
# skip: [7d06015d936c861160803e020f68f413b5c3cd9d] Merge tag 'pci-v6.15-changes' of git://git.kernel.org/pub/scm/linux/kernel/git/pci/pci
git bisect skip 7d06015d936c861160803e020f68f413b5c3cd9d
# good: [94788792f37902f1f4d417f6f9663831cf7e91fc] ipv4: initialize inet socket cookies with sockcm_init
git bisect good 94788792f37902f1f4d417f6f9663831cf7e91fc
# bad: [325f510fcd9cda5a44bcb662b74ba4e3dabaca10] watchdog: da9052_wdt: respect TWDMIN
git bisect bad 325f510fcd9cda5a44bcb662b74ba4e3dabaca10
# skip: [a10c7949adf94356e56d5c8878f6fc3f25bd0c15] Merge tag 'linux_kselftest-kunit-6.15-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/shuah/linux-kselftest
git bisect skip a10c7949adf94356e56d5c8878f6fc3f25bd0c15
# skip: [c925e9936def79c6a1af7c61b02438670d0ab65c] Merge tag 'amd-drm-fixes-6.15-2025-04-16' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes
git bisect skip c925e9936def79c6a1af7c61b02438670d0ab65c
# good: [7a2dd31359b0761c48ae44d376d2f083fc41f0ef] iio: accel: msa311: convert to use maple tree register cache
git bisect good 7a2dd31359b0761c48ae44d376d2f083fc41f0ef

I'm open to any advice to help finding it !

regards,
Christophe


On 03/06/2026 16:34, Salvatore Bonaccorso wrote:
Control: found -1 7.0.9-1

On Sat, May 30, 2026 at 10:51:21AM +0200, Christophe Combelles wrote:
After upgrading to :

7.0.9+deb13-amd64 #1 SMP PREEMPT_DYNAMIC Debian 7.0.9-1~bpo13+1 (2026-05-26)
x86_64 GNU/Linux

I plugged again the same device after a fresh reboot, I heard several audio
ticks (I remember more than one per plugged disk) like nautilus or something
from the gnome session, then the whole system froze with the caps lock key
blinking, and no QR code.
Thanks for reporting back. I'm unsure if you have seen the earlier
questions:

- Can you reproduce the issue, and can you get a complete log from the
   beginning of the log?
- Is that a regression for an earlier kernel, which is the last one
   which works?
- 6.18.12-1~bpo13+1 is not a current kernel, can you reproduce the
   issue with the most current kernel in backports? Ideally please as
   well test the current 6.19.8-1 in unstable and report back if the
   problem is triggerable there as well (if yes, the full log from this
   is valuable to potentially make an upstream report).


It might be usefull to attach a netconsole to get some boot messages
right from the beginning,
https://docs.kernel.org/networking/netconsole.html explains how to do
it. Can you pleas give that a try?

Please test in any case as well the most current kernel, 7.0.10-1 at
the moment in unstable.

Regards,
Salvatore

Reply via email to