http://oss.sgi.com/bugzilla/show_bug.cgi?id=860
--- Comment #5 from Carlos E. R. M. <carlos.e.r@xxxxxxxxxxxx> 2010-01-15
14:33:43 CST ---
(In reply to comment #4)
> (In reply to comment #3)
>
> ...
>
> > Thus it will have to be DM (although it only happens with encrypted xfs
> > partitions, not if they are reiserfs or ext3).
>
> ext3, at least upstream, does not issue barriers by default, and as Dave said:
>
> > This looks like a DM lockup. From the output of dmesg you posted,
> > we have a DM barrier flush waiting for completion:
>
> so without barriers on you may not see the problem?
I delayed replying till I tested this. Indeed, mounting "nobarrier" avoids the
issue. I have replayed at least 12 times on 4 filesystems the operation that
caused the problem before (took 5 hours running), and it did not reproduce.
Thanks!
At least on openSUSE 11.2, it doesn't work in openSUSE 11.0: the issue has to
be different there.
In case somebody here has more ideas, I will attach the data I have from the
kernel log, after the lockup, and issuing "echo t > /proc/sysrq-trigger".
--
Configure bugmail: http://oss.sgi.com/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
|