XFS hang during xfs_fsr run

[Dave Chinner]

On Wed, Mar 10, 2010 at 09:46:20AM +0100, Michael Weissenbacher wrote:
> >Odds are you'll still oops, just earlier.  Send anything interesting
> >you see in the dmesg if/when it happens...
> >
> Those are the assertion failures i saw during yesterday's run. The
> process xfs_fsr hung on inode 536898141 and i couldn't kill it, even
> with "kill -9". I had to reboot to get everything back to normal.

> Mar  9 22:32:43 email kernel: [57269.991906] Assertion failed: ip->i_delayed_blks == 0, file: fs/xfs/xfs_bmap.c, line: 5657

Hi Michael - have you got any idea what the files are that are
hitting this? This failure is implying that the inode is still dirty
after syncing all the data. Is something trying to modify it while
XFS is trying to map it?

We've seen this occasionally in the past, but we've never been able
to reproduce it with any reliability.  Hence any information you can
extract would certainly help us here.

Cheers,

Dave.
-- 
Dave Chinner
david at fromorbit.com