[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#1006149: linux-image-5.16.0-1-686: Fails to boot on T41 Thinkpads



On 4/3/22 22:44, Diederik de Haas wrote:
On Sunday, 3 April 2022 15:16:06 CEST Francesco C wrote:
As you appear to be familiar with kernel building, could you look at the
patches that Damien Le Moal posted in this bug report?

Done ... same behaviour : as before the only way to restart the
machine is by pressing the power button ; Ctl+Alt+Del and Alt+SysReq+b
do not work.

Did you try the patch that Damien posted here:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1006149#131
Asking as you posted a different patch later in your reply.

I _think_ that Damien's patch is supposed to supply more information, not fix
the issue. I'm not sure whether the patch should be applied to the working
5.15 kernel or the non-working 5.16 kernel though.
In the former case a full dmesg output of a boot without the patch and one
with the patch should be attached to this bug report.
In the latter case there likely won't be a log file and in that case a video of
the (5.16) boot without and with the patch should hopefully give us more
information.

Damien: can you clarify whether the patch should be applied to the 5.15 or
5.16 kernel?

My apologies for not following up on this earlier. I have been really busy.

Looking at the code again, this patch will not apply to 5.15 and will apply only to 5.16. And as for 5.16, given that the sd_read_cpr() function will trigger a read log command only for ATA-11 or above drives, which the Thinkpad drive is not, I do not think the patch will make any difference.

My hunch is that this drive simply reacts badly to read log commands and should be marked with a horkage to blacklist that command for it. Cleanups for read log handling in 5.15 actually now use that command to check accesses to any log page. Meaning that prior to 5.15, the command was not issued for that drive.

I will send a patch to blacklist read log for that drive later today, to try. If that patch works, I will queue it and cc stable so that it gets backported.

--
Damien Le Moal
Western Digital Research


Reply to: