[Bug 220069] [6.13.9] regression USB controller dies

[bugzilla-daemon@kernel.org]

https://bugzilla.kernel.org/show_bug.cgi?id=220069

--- Comment #38 from Claudio Wunder (cwunder@gnome.org) ---
That's very useful (all the commentary)!

By the way,

> Looks like a 3rd crash happened, interestingly enough, the script did not
> detect a 3rd crash and did not send an unbind/bind command; But the Bus (5)
> got completely disconnected, Bus (7) which previously had the webcam, now is
> still working. I did not have the autopsy command running for Bus (5) so no
> logs; But Ill start it now and keep an eye on it.

Ignore this, I facepalmed myself afterwards, of course it would only
unbind/bind and capture 6a since that's what I provided to the script. I
modified the script adding support to "multiple" pci ids and generating the
folders and dumps prefixed with the pciid.

> I would suggest to reenable power management, and try something else. If you
> moved the problematic hubs to bus 5, try to move the problematic webcam to
> bus 7.

Noted, Ill first keep it off for one more day to see a crash happens; Then Ill
re-enable it afterwards.

> If bus 5 crashes, the problem is only hubs and the webcam is harmless.
If bus 7 crashes, the problem is the webcam.
If neither bus crashes for a few days, it looks like it takes two to the tango.

Ill attempt these distribution steps.

> If your kernel has xhci_hcd and xhci_pci built as modules, run:

`lsmod` doesn't seem to list either xhci_pci and xhci_hcd as modules;

> If they are built into the kernel image, you need to figure out how to add

Seems the case. Can I add this flag via grub?

-- 
You may reply to this email to add a comment.

You are receiving this mail because:
You are watching the assignee of the bug.