Bug 251579 - kernel crash after an upgrade from 12.0-RELEASE-p13 to 12.1-RELEASE-p10
Summary: kernel crash after an upgrade from 12.0-RELEASE-p13 to 12.1-RELEASE-p10
Status: Open
Alias: None
Product: Base System
Classification: Unclassified
Component: kern (show other bugs)
Version: 12.1-RELEASE
Hardware: amd64 Any
: --- Affects Only Me
Assignee: freebsd-bugs (Nobody)
URL:
Keywords: panic, regression
Depends on:
Blocks:
 
Reported: 2020-12-04 14:37 UTC by Michaël
Modified: 2020-12-08 14:23 UTC (History)
2 users (show)

See Also:


Attachments
core file (71.72 KB, text/plain)
2020-12-04 14:37 UTC, Michaël
no flags Details
another core file (167.52 KB, text/plain)
2020-12-04 14:40 UTC, Michaël
no flags Details
and another (90.49 KB, text/plain)
2020-12-04 14:43 UTC, Michaël
no flags Details
this one is cool too (114.84 KB, text/plain)
2020-12-04 14:43 UTC, Michaël
no flags Details

Note You need to log in before you can comment on or make changes to this bug.
Description Michaël 2020-12-04 14:37:10 UTC
Created attachment 220249 [details]
core file

Hi! 

After a Freebsd upgrade the system reboot after a random time with this error:
"Fatal trap 12: page fault while in kernel mode"

When I return to 12.0 the system looks stable.

As attachment the last core.txt.
Comment 1 Michaël 2020-12-04 14:40:53 UTC
Created attachment 220250 [details]
another core file
Comment 2 Michaël 2020-12-04 14:43:00 UTC
Created attachment 220251 [details]
and another
Comment 3 Michaël 2020-12-04 14:43:22 UTC
Created attachment 220252 [details]
this one is cool too
Comment 4 Michaël 2020-12-05 14:17:42 UTC
Hi,

I have just realized that I have this problem with old kernel too. Sorry for that. It's a random error, I didn't have this problem in the past and after I switched on old kernel the system had been working for a while before it crashed in the same condition.

Perhaps it's a hardware problem. Is it possible that a problem with memory or SSD cause this kind of error ?
Comment 5 Mark Johnston freebsd_committer 2020-12-07 15:28:14 UTC
(In reply to Michaël from comment #4)
Yes, this looks like some sort of hardware problem.  Faulting on an address like 0xffffffb481ea9b18 is pretty strange.  I would suggest trying to run a memory checker.
Comment 6 Michaël 2020-12-07 20:53:15 UTC
(In reply to Mark Johnston from comment #5)
Thx. I tried memtest but I didn't succeed to identify a fault with this.
I have noticed however that the memory size checked was under the real
memory installed.  I think that's becaus this tool run in userland, and
I need to do that at boot time. Do you know a good memory checker I
could use ?
Comment 7 Mark Johnston freebsd_committer 2020-12-08 14:09:57 UTC
(In reply to Michaël from comment #6)
memtest86+ is probably the best bet.
Comment 8 Michaël 2020-12-08 14:23:16 UTC
(In reply to Mark Johnston from comment #7)

Ok Thanks for your help. I will try this.