From: Theodore Ts'o <tytso@mit.edu>
To: Jan Kara <jack@suse.cz>, Dmitry Monakhov <dmonakhov@openvz.org>
Cc: linux-ext4@vger.kernel.org
Subject: REGRESSION: [PATCH 04/12] ext4: Disable merging of uninitialized extents
Date: Sat, 9 Feb 2013 12:10:15 -0500 [thread overview]
Message-ID: <20130209171015.GC8091@thunk.org> (raw)
In-Reply-To: <1358510446-19174-5-git-send-email-jack@suse.cz>
On Fri, Jan 18, 2013 at 01:00:38PM +0100, Jan Kara wrote:
> Merging of uninitialized extents creates all sorts of interesting race
> possibilities when writeback / DIO races with fallocate. Thus
> ext4_convert_unwritten_extents_endio() has to deal with a case where
> extent to be converted needs to be split out first. That isn't nice
> for two reasons:
>
> 1) It may need allocation of extent tree block so ENOSPC is possible.
> 2) It complicates end_io handling code
>
> So we disable merging of uninitialized extents which allows us to simplify
> the code. Extents will get merged after they are converted to initialized
> ones.
>
> Reviewed-by: Zheng Liu <wenqing.lz@taobao.com>
> Signed-off-by: Jan Kara <jack@suse.cz>
Sorry for not noticing this earlier, but this patch is causing a
regression. It is loading to test 113 failing when dioread_nolock is
used:
113 [ 47.619363] EXT4-fs error (device vdb): ext4_convert_unwritten_extents_endio:3411: i
node #10951: comm kworker/u:0: Written extent modified before IO finished: extent logical block
1024, len 1024; IO logical block 1024, len 127
[ 47.619363]
[ 47.623239] EXT4-fs warning (device vdb): ext4_convert_unwritten_extents:4522: inode #10951:
block 1024: len 127: ext4_ext_map_blocks returned -5
[ 47.628975] EXT4-fs (vdb): failed to convert unwritten extents to written extents -- potenti
al data loss! (inode 10951, offset 4194304, size 520192, error -5)
As a result, I am considering whether or not I should to drop the
following patches from the ext4 tree:
e63dd9c ext4: disable merging of uninitialized extents
de39534 ext4: remove unnecessary wait for extent conversion in ext4_fallocate()
37bf0a8 ext4: ext4_split_extent should take care of extent zeroout
I know that these patches fix other potential races which causes data
loss, but they've been around for a while, and in practice seem to be
relatively rarely hit.
Jan, Dmitry, any chance you can take a look at this, and give me
suggestions about how to proceed at this point?
Thanks,
- Ted
next prev parent reply other threads:[~2013-02-09 17:10 UTC|newest]
Thread overview: 53+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-01-18 12:00 [PATCH 0/12 v2] ext4: Several simplifications and fixes Jan Kara
2013-01-18 12:00 ` [PATCH 01/12] ext4: Always use ext4_bio_write_page() for writeout Jan Kara
2013-01-28 14:31 ` Theodore Ts'o
2013-01-18 12:00 ` [PATCH 02/12] ext4: Use redirty_page_for_writepage() in ext4_bio_write_page() Jan Kara
2013-01-28 14:34 ` Theodore Ts'o
2013-01-18 12:00 ` [PATCH 03/12] ext4: Remove bogus wait for unwritten extents in ext4_ind_direct_IO Jan Kara
2013-01-22 11:11 ` Dmitry Monakhov
2013-01-22 13:44 ` Jan Kara
2013-01-22 14:12 ` Dmitry Monakhov
2013-01-22 15:21 ` Jan Kara
2013-01-22 14:22 ` Zheng Liu
2013-01-22 15:22 ` Jan Kara
2013-01-22 16:00 ` Zheng Liu
2013-01-22 23:14 ` Jan Kara
2013-01-23 6:11 ` Zheng Liu
2013-01-23 9:42 ` Jan Kara
2013-01-18 12:00 ` [PATCH 04/12] ext4: Disable merging of uninitialized extents Jan Kara
2013-01-24 9:49 ` Dmitry Monakhov
2013-01-24 15:12 ` Jan Kara
2013-01-24 15:32 ` Dmitry Monakhov
2013-01-28 14:36 ` Theodore Ts'o
2013-01-28 15:02 ` Dmitry Monakhov
2013-01-28 15:38 ` Theodore Ts'o
2013-01-29 7:41 ` Dmitry Monakhov
2013-01-29 8:37 ` Zheng Liu
2013-01-31 7:47 ` Dmitry Monakhov
2013-01-31 12:39 ` Jan Kara
2013-01-31 14:09 ` Dmitry Monakhov
2013-01-31 16:54 ` Theodore Ts'o
2013-02-09 17:10 ` Theodore Ts'o [this message]
2013-02-12 21:58 ` REGRESSION: " Jan Kara
2013-02-13 4:57 ` Theodore Ts'o
2013-02-13 7:26 ` Dmitry Monakhov
2013-02-13 15:08 ` Merge window planning for ext4 and Ted's vacation Theodore Ts'o
2013-02-14 10:47 ` REGRESSION: [PATCH 04/12] ext4: Disable merging of uninitialized extents Jan Kara
2013-02-14 16:11 ` Jan Kara
2013-02-14 19:05 ` Theodore Ts'o
2013-02-14 21:32 ` Jan Kara
2013-01-18 12:00 ` [PATCH 05/12] ext4: Remove unnecessary wait for extent conversion in ext4_fallocate() Jan Kara
2013-01-18 12:00 ` [PATCH 06/12] ext4: Move work from io_end to inode Jan Kara
2013-01-28 14:45 ` Theodore Ts'o
2013-01-18 12:00 ` [PATCH 07/12] ext4: Simplify list handling in ext4_do_flush_completed_IO() Jan Kara
2013-01-28 14:51 ` Theodore Ts'o
2013-01-18 12:00 ` [PATCH 08/12] ext4: Remove __ext4_journalled_writepage() from mpage_da_submit_io() Jan Kara
2013-01-28 14:40 ` Theodore Ts'o
2013-01-18 12:00 ` [PATCH 09/12] ext4: Dirty page has always buffers attached Jan Kara
2013-01-28 17:55 ` Theodore Ts'o
2013-01-18 12:00 ` [PATCH 10/12] ext4: Simplify mpage_add_bh_to_extent() Jan Kara
2013-01-28 18:06 ` Theodore Ts'o
2013-01-18 12:00 ` [PATCH 11/12] ext4: Make ext4_bio_writepage() handle unprepared buffers Jan Kara
2013-01-29 1:59 ` Theodore Ts'o
2013-01-18 12:00 ` [PATCH 12/12] ext4: Fix ext4_writepage() to achieve data=ordered guarantees Jan Kara
2013-01-29 2:08 ` Theodore Ts'o
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20130209171015.GC8091@thunk.org \
--to=tytso@mit.edu \
--cc=dmonakhov@openvz.org \
--cc=jack@suse.cz \
--cc=linux-ext4@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).