8-STABLE/amd64 semi-regular crash with "kernel trap 12 with interrupts disabled" in "process 12 (swi4: clock)"

Jeremy Chadwick freebsd at jdc.parodius.com
Tue Jan 18 21:46:56 UTC 2011


On Wed, Jan 19, 2011 at 12:12:48AM +0300, Lev Serebryakov wrote:
> Hello, Freebsd-stable.
> 
>   One of my servers crashes about once a week, with always same
> diagnostics: "kernel trap 12 with interrupts disabled" and in same
> process: "swi4: clock"
> 
>   It doesn't look as memory failure, as memtest86+ can not find any
> errors in 8 passes.
> 
>   Also, after this crash server refuse to auto-reboot, last message on
> console is "cpu_reset: Stopping other CPUs", and it hangs.
> 
>   Kernel config, booting dmesg & results of "savecore" are attached
> (bzipped).

CC'ing Jack Vogel of Intel, as this looks like it could be something the
em(4) driver might be tickling.  I do see it in the stack trace shortly
before the crash.  In the interim, can you please provide output from the
following command:

# pciconf -lbcv

And include only the entries relevant to your emX devices.

As for the "the server refuses to auto-reboot": that may be a separate
problem.  You might try toggling the hw.acpi.disable_on_reboot and
hw.acpi.handle_reboot sysctls (check what values they have on your
system first) to see if there's any improvement.

For Jack -- the core/stack trace, and dmesg are at the below URL as
attachments (and bzip2 compressed):

http://lists.freebsd.org/pipermail/freebsd-stable/2011-January/061168.html

-- 
| Jeremy Chadwick                                   jdc at parodius.com |
| Parodius Networking                       http://www.parodius.com/ |
| UNIX Systems Administrator                  Mountain View, CA, USA |
| Making life hard for others since 1977.               PGP 4BD6C0CB |



More information about the freebsd-stable mailing list