Re: [PATCH v3] ext4: Fix bug on in ext4_es_cache_extent as ext4_split_extent_at failed

From: Jan Kara
Date: Wed May 05 2021 - 06:41:10 EST


On Wed 05-05-21 11:29:57, yebin wrote:
>
>
> On 2021/4/30 20:58, Jan Kara wrote:
> > On Wed 28-04-21 16:51:58, Ye Bin wrote:
> > > We got follow bug_on when run fsstress with injecting IO fault:
> > > [130747.323114] kernel BUG at fs/ext4/extents_status.c:762!
> > > [130747.323117] Internal error: Oops - BUG: 0 [#1] SMP
> > > ......
> > > [130747.334329] Call trace:
> > > [130747.334553] ext4_es_cache_extent+0x150/0x168 [ext4]
> > > [130747.334975] ext4_cache_extents+0x64/0xe8 [ext4]
> > > [130747.335368] ext4_find_extent+0x300/0x330 [ext4]
> > > [130747.335759] ext4_ext_map_blocks+0x74/0x1178 [ext4]
> > > [130747.336179] ext4_map_blocks+0x2f4/0x5f0 [ext4]
> > > [130747.336567] ext4_mpage_readpages+0x4a8/0x7a8 [ext4]
> > > [130747.336995] ext4_readpage+0x54/0x100 [ext4]
> > > [130747.337359] generic_file_buffered_read+0x410/0xae8
> > > [130747.337767] generic_file_read_iter+0x114/0x190
> > > [130747.338152] ext4_file_read_iter+0x5c/0x140 [ext4]
> > > [130747.338556] __vfs_read+0x11c/0x188
> > > [130747.338851] vfs_read+0x94/0x150
> > > [130747.339110] ksys_read+0x74/0xf0
> > >
> > > If call ext4_ext_insert_extent failed but new extent already inserted, we just
> > > update "ex->ee_len = orig_ex.ee_len", this will lead to extent overlap, then
> > > cause bug on when cache extent.
> > Thanks for the patch but I'm still not quite sure, how overlapping extents
> > in the extent tree can lead to triggering BUG_ON(lblk + len - 1 < lblk) in
> > ext4_es_cache_extent(). Can you ellaborate a bit more how this happens?
> Assume that there is extent [10, 100] (ee_block=10 ee_len=91), call
> ext4_split_extent_at split at 50,
> we get two extent [10, 49] and [50, 100], then call ext4_ext_insert_extent
> to insert new extent [50, 100],
> if insert extent successed, but call ext4_ext_dirty failed(return -EROFS)
> as JBD maybe abort as io error.
> Then fix old extent length with old value, so we get two extent [10,
> 100] (ee_block=10 ee_len=91) and
> [50, 100](ee_block=50 ee_len=51).
> If call ext4_cache_extent to cache above extents as follow:
> prev = 0 lblk = 10 len = 91 --> cache [10, 100] ---> prev = lblk + len =
> 101
> prev = 101 lblk = 50 len = 51 --> prev != 0 && prev != lblk --> cache [prev
> = 101, lblk - prev = 50 - 101 = -51]
> Obvious if call ext4_es_cache_extent cache extent[101, -51] wil trigger
> "BUG_ON(end < lblk)" .

Thanks for great explanation! Now I understand.

> > > If call ext4_ext_insert_extent failed don't update ex->ee_len with old value.
> > > Maybe there will lead to block leak, but it can be fixed by fsck later.
> > >
> > > After we fixed above issue with v2 patch, but we got the same issue.
> > > ext4_split_extent_at:
> > > {
> > > ......
> > > err = ext4_ext_insert_extent(handle, inode, ppath, &newex, flags);
> > > if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
> > > ......
> > > ext4_ext_try_to_merge(handle, inode, path, ex); ->step(1)
> > > err = ext4_ext_dirty(handle, inode, path + path->p_depth); ->step(2)
> > > if (err)
> > > goto fix_extent_len;
> > > ......
> > > }
> > > ......
> > > fix_extent_len:
> > > ex->ee_len = orig_ex.ee_len; ->step(3)
> > > ......
> > > }
> > > If step(1) have been merged, but step(2) dirty extent failed, then go to
> > > fix_extent_len label to fix ex->ee_len with orig_ex.ee_len. But "ex" may not be
> > > old one, will cause overwritten. Then will trigger the same issue as previous.
> > > If step(2) failed, just return error, don't fix ex->ee_len with old value.
> > >
> > > Signed-off-by: Ye Bin <yebin10@xxxxxxxxxx>
> > > ---
> > > fs/ext4/extents.c | 13 +++++--------
> > > 1 file changed, 5 insertions(+), 8 deletions(-)
> > >
> > > diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c
> > > index 77c84d6f1af6..d4aa24a09d8b 100644
> > > --- a/fs/ext4/extents.c
> > > +++ b/fs/ext4/extents.c
> > > @@ -3238,15 +3238,12 @@ static int ext4_split_extent_at(handle_t *handle,
> > > ex->ee_len = cpu_to_le16(ee_len);
> > > ext4_ext_try_to_merge(handle, inode, path, ex);
> > > err = ext4_ext_dirty(handle, inode, path + path->p_depth);
> > > - if (err)
> > > - goto fix_extent_len;
> > > -
> > > - /* update extent status tree */
> > > - err = ext4_zeroout_es(inode, &zero_ex);
> > > -
> > > - goto out;
> > > - } else if (err)
> > > + if (!err)
> > > + /* update extent status tree */
> > > + err = ext4_zeroout_es(inode, &zero_ex);
> > > + } else if (err && err != -EROFS) {
> > I fail to see why EROFS is special here. Can you explain a bit please?
> V1 patch Ted suggest me to fix length only when "err != -EROSFS". As if
> we don't
> fix origin extent with old extent length, it will lead to block leak.
> Ted said as follow:
>
> If you don't want to do that, then a "do no harm" fix would be
> something like this:
>
> ...
> } else if (err == -EROFS) {
> return err;
> } else if (err)
> goto fix_extent_len;
>
> So in the journal abort case, when err is set to EROFS, we don't try
> to reset the length, since in theory the file system is read-only
> already anyway. However, in the ENOSPC case, we won't end up silently
> leaking blocks that will be lost until the user somehow decides to run
> fsck.

I see. Now I understand your patch. Honestly, seeing how fragile is trying
to fix extent tree after split has failed in the middle, I would probably
go even further and make sure we fix the tree properly in case of ENOSPC
and EDQUOT (those are easily user triggerable). Anything else indicates a
HW problem or fs corruption so I'd rather leave the extent tree as is and
don't try to fix it (which also means we will not create overlapping
extents). So something like:


err = ext4_ext_insert_extent(handle, inode, ppath, &newex, flags);
- if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) {
+ if (err == -ENOSPC || err == -EDQUOT) {
+ if (EXT4_EXT_MAY_ZEROOUT & split_flag)
+ err = handle zeroing...
if (err) {
fix extent len
goto out;
}
...
}

and in all other cases just 'goto out' in case of error. What do you think?

Honza
--
Jan Kara <jack@xxxxxxxx>
SUSE Labs, CR