On Fri, 2012-10-05 at 07:59 -0400, Brian Foster wrote:
> On 10/04/2012 05:55 PM, Ben Hutchings wrote:
> > On Tue, Oct 02, 2012 at 12:53:47AM +0200, Willy Tarreau wrote:
> >> 2.6.32-longterm review patch.  If anyone has any objections, please let me know.
> >>
> >> ------------------
> >>
> >> From: Brian Foster <bfoster@redhat.com>
> >>
> >> commit 97795d2a5b8d3c8dc4365d4bd3404191840453ba upstream.
> >>
> >> If we hit a condition where we have allocated metadata blocks that
> >> were not appropriately reserved, we risk underflow of
> >> ei->i_reserved_meta_blocks.  In turn, this can throw
> >> sbi->s_dirtyclusters_counter significantly out of whack and undermine
> >> the nondelalloc fallback logic in ext4_nonda_switch().  Warn if this
> >> occurs and set i_allocated_meta_blocks to avoid this problem.
> >>
> >> This condition is reproduced by xfstests 270 against ext2 with
> >> delalloc enabled:
> >>
> >> Mar 28 08:58:02 localhost kernel: [  171.526344] EXT4-fs (loop1): delayed block allocation failed for inode 14 at logical offset 64486 with max blocks 64 with error -28
> >> Mar 28 08:58:02 localhost kernel: [  171.526346] EXT4-fs (loop1): This should not happen!! Data will be lost
> >>
> >> 270 ultimately fails with an inconsistent filesystem and requires an
> >> fsck to repair.  The cause of the error is an underflow in
> >> ext4_da_update_reserve_space() due to an unreserved meta block
> >> allocation.
> > [...]
> >> +	if (unlikely(ei->i_allocated_meta_blocks > ei->i_reserved_meta_blocks)) {
> >> +		ext4_msg(inode->i_sb, KERN_NOTICE, "%s: ino %lu, allocated %d "
> >> +			 "with only %d reserved metadata blocks\n", __func__,
> >> +			 inode->i_ino, ei->i_allocated_meta_blocks,
> >> +			 ei->i_reserved_meta_blocks);
> >> +		WARN_ON(1);
> >> +		ei->i_allocated_meta_blocks = ei->i_reserved_meta_blocks;
> >> +	}
> > [...]
> > 
> > This seems to be working around a bug elsewhere.  Has the underlying
> > bug been fixed in mainline yet?
> > 
> 
> Yes, the bug was fixed in:
> 
> 03179fe92318e7934c180d96f12eff2cb36ef7b6
> ext4: undo ext4_calc_metadata_amount if we fail to claim space

OK, and that's been applied to stable as:

3.2: d9af293 ext4: undo ext4_calc_metadata_amount if we fail to claim space
3.4: c0ce1fd ext4: undo ext4_calc_metadata_amount if we fail to claim space
3.5: 564dfa3 ext4: undo ext4_calc_metadata_amount if we fail to claim space

Presumably it will need some backporting for older versions.

Ben.

-- 
Ben Hutchings
You can't have everything.  Where would you put it?
[unhandled content-type:application/pgp-signature]