There was something wrong with the old FS and the kernel 184.108.40.206 (unable to
replay log) has not been of much help.
All works fine with kernel 220.127.116.11 and new clear FS for a few days .
On Tue, Dec 29, 2009 at 10:22:04AM +1100, Dave Chinner wrote:
> On Mon, Dec 28, 2009 at 04:10:48PM +0100, DS wrote:
> > Hello,
> > after upgrade from Debian etch -> Debian lenny with custom kernel
> > 18.104.22.168 (before 22.214.171.124) I see this error and server is
> > resetting sometimes..
> > Filesystem was creating far far away (about 4years ago).
> > XFS internal error XFS_WANT_CORRUPTED_RETURN at line 366 of file
> > fs/xfs/xfs_alloc.c. Caller 0xffffffff81169ae3
> Any errors in the log before this (e.g. IO errors)?
> > Pid: 4022, comm: procmail Not tainted 126.96.36.199 #3
> > Call Trace:
> > [<ffffffff8118edfe>] xfs_error_report+0x3e/0x40
> > [<ffffffff811680fa>] xfs_alloc_fixup_trees+0x2ba/0x340
> > [<ffffffff81167de9>] ? xfs_alloc_get_rec+0x29/0x60
> > [<ffffffff81169ae3>] xfs_alloc_ag_vextent_near+0x3f3/0x940
> > [<ffffffff8116a105>] xfs_alloc_ag_vextent+0xd5/0x130
> > [<ffffffff8116aa68>] xfs_alloc_vextent+0x3a8/0x4f0
> Looks like a corrupted free space tree - that could have happened a
> long time ago and only now are you tripping over it. I'd run
> xfs_repair on the filesystem to fix up the problem. If it happens
> again then we'll have to look deeper.
> Dave Chinner