Hi, On Tue, May 14, 2002 at 10:58:42AM +0100, Richard Kimber wrote: > On Mon, 13 May 2002 15:06:48 +1000 (EST) > Neil Brown <neilb@cse.unsw.edu.au> wrote: > > > I just got bitten by this Assertion. The one that starts as in the > > subject, and ends with: > > "!(((jh2bh(jh))->b_state & (1UL << BH_Lock)) != 0)" > > After getting one of these, when using rsync to a PD drive, the device > could not be unmounted. After re-booting I did e2fsck /dev/sdb5 and got: > > e2fsck 1.24a (02-Sep-2001) > backup: recovering journal > backup: Attempt to read block from filesystem resulted in short read while > reading block 1331 > > JFS: Failed to read block at offset 813 > JFS: IO error -5 recovering block 813 in log > e2fsck: Input/output error while recovering ext3 journal of backup > > What should my next step be? Looks like you've got a bad block in the journal itself --- that's bad news. We should really try to recover from this in e2fsck, but there's really nothing that the kernel can do --- an unrecoverable journal means that a full fsck is absolutely necessary. For now, try removing the journal and fsck'ing: tune2fs -O ^has_journal should do the trick. Cheers, Stephen
