* Net Llama! <netllama@xxxxxxxxxxxxxxxxxx> [2004-07-12 12:33:27 -0500]:
> I'm running a vanilla 2.6.7 on a workstation. It appears to have Oops'ed
> yesterday morning while the slocate/updatedb cronjob was running:
[...]
> Jul 11 04:07:39 netllama kernel: <6>note: updatedb[25042] exited with
> preempt_count 1
> Jul 11 04:07:39 netllama kernel: bad: scheduling while atomic!
> Jul 11 04:07:39 netllama kernel: [schedule+1210/1264]
> schedule+0x4ba/0x4f0
[...]
> The box seems to be stable since then (i've not rebooted it yet), although
> the updatedb cronjob is hung. I'm not 100% certain that this is a XFS
> issue, but since i see a ton of xfs references in the oops, i'm asking
> here first. thanks.
Well, I have seen such problems on other kernels as well (2.6.7 ->
2.6.7-bk20), running on machines without any XFS partition. Thus I tend
to say it is an other kernel bug (scheduling problems are not related to
XFS anyway IMHO). It seems running on 2.8.1-rc1 is a better idea, as I
heard 2.6.7 had different small problems as well.
Regards,
Laszlo
|