From: Fernando Luis Vazquez Cao <fernando@kic.ac.jp>
To: Dave Chinner <david@fromorbit.com>
Cc: "Eric Sandeen" <sandeen@redhat.com>,
"Fernando Luis Vázquez Cao" <fernando@oss.ntt.co.jp>,
"Jan Kara" <jack@suse.cz>, "Theodore Tso" <tytso@MIT.EDU>,
"Alan Cox" <alan@lxorguk.ukuu.org.uk>,
"Pavel Machek" <pavel@suse.cz>,
"kernel list" <linux-kernel@vger.kernel.org>,
"Jens Axboe" <jens.axboe@oracle.com>,
"Ric Wheeler" <rwheeler@redhat.com>
Subject: Re: vfs: Add MS_FLUSHONFSYNC mount flag
Date: Sat, 14 Feb 2009 01:29:28 +0900 [thread overview]
Message-ID: <1234542568.9916.183.camel@bladerunner> (raw)
In-Reply-To: <20090213122051.GX8830@disturbed>
On Fri, 2009-02-13 at 23:20 +1100, Dave Chinner wrote:
> On Fri, Feb 13, 2009 at 12:20:17AM -0600, Eric Sandeen wrote:
> > I'm just a little leery of the "dangerous" mount option proliferation, I
> > guess.
>
> You're not the only one, Eric. It's bad enough having to explain to
> users what barriers do once they have lost data after a power loss,
> let alone confusing them further by adding more mount options they
> will get wrong by accident....
That is precisely the reason why we should use sensible defaults, which
in this case means enabling barriers and flushing disk caches on
fsync()/fdatasync() by default.
Adding either a new mount option (as you yourself suggest below) or a
sysfs tunable is desirable for those cases when we really do not need to
flush the disk write cache to guarantee integrity (battery-backed block
devices come to mind), or we want to be fast at the cost of potentially
losing some data.
> Quite frankly, the VFS should do stuff that is slow and safe
> and filesystems can choose to ignore the VFS (via filesystem
> specific mount options) if they want to be fast and potentially
> unsafe.
To avoid unnecessary flushes and allow for filesystem-specific
optimizations I was considering the following approach:
1- Add flushonfsync mount option (as an aside, I am of the opinion that
it should be set by default).
2- Modify file_fsync() so that it checks whether FLUSHONFSYNC is set and
flushes the underlying device accordingly. With this we would cover all
filesystems that use the vfs-provided file_fsync() as their fsync method
(commonly used filesystems such as fat fall in this group).
3- Advanced filesystems (ext3/4, XFS, btrfs, etc) which provide their
own fsync implementations are allowed to perform filesystem-specific
optimizations there to minimize the number of flushes and maximize
throughput.
In this patch set I implemented (1) and (3) for ext3/4 to have some code
to comment on.
Does this approach make sense? Thoughts?
- Fernando
next prev parent reply other threads:[~2009-02-13 16:36 UTC|newest]
Thread overview: 73+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-01-13 13:14 ext2 + -osync: not as easy as it seems Pavel Machek
2009-01-13 13:45 ` Alan Cox
2009-01-13 14:03 ` Theodore Tso
2009-01-13 14:07 ` Jens Axboe
2009-01-13 14:26 ` [PATCH] block: Fix documentation for blkdev_issue_flush() Theodore Ts'o
2009-01-13 14:28 ` Jens Axboe
2009-01-13 14:30 ` ext2 + -osync: not as easy as it seems Jan Kara
2009-01-13 14:46 ` Theodore Tso
2009-01-14 3:37 ` Fernando Luis Vázquez Cao
2009-01-14 10:35 ` Jan Kara
2009-01-14 13:21 ` Theodore Tso
2009-01-14 14:05 ` Jan Kara
2009-01-14 14:08 ` Jens Axboe
2009-01-14 14:34 ` Theodore Tso
2009-01-14 14:43 ` Jens Axboe
2009-02-12 16:43 ` Eric Sandeen
2009-02-16 12:09 ` Jens Axboe
2009-01-14 14:12 ` Theodore Tso
2009-01-14 14:37 ` Jan Kara
2009-01-14 16:59 ` Theodore Tso
2009-01-15 12:06 ` Fernando Luis Vázquez Cao
2009-01-15 23:45 ` Jan Kara
2009-01-16 12:31 ` Fernando Luis Vázquez Cao
2009-01-16 13:55 ` ext3: call blkdev_issue_flush on fsync Fernando Luis Vázquez Cao
2009-01-16 16:30 ` Jan Kara
2009-01-17 9:47 ` Fernando Luis Vázquez Cao
2009-01-17 10:00 ` Fernando Luis Vázquez Cao
2009-01-19 12:03 ` Jan Kara
2009-01-28 9:45 ` Fernando Luis Vázquez Cao
2009-01-28 9:55 ` Jan Kara
2009-02-12 10:33 ` Fernando Luis Vázquez Cao
2009-02-12 10:35 ` vfs: Improve readability off mount flag definitins by using offsets Fernando Luis Vázquez Cao
2009-02-12 10:36 ` vfs: Add MS_FLUSHONFSYNC mount flag Fernando Luis Vázquez Cao
2009-02-12 17:13 ` Eric Sandeen
2009-02-12 17:29 ` Jeff Garzik
2009-02-14 15:36 ` Christoph Hellwig
2009-02-15 7:23 ` Fernando Luis Vázquez Cao
2009-02-15 22:54 ` Theodore Tso
2009-02-16 4:29 ` Eric Sandeen
2009-02-16 7:47 ` Fernando Luis Vázquez Cao
2009-02-12 21:23 ` Jan Kara
2009-02-12 21:30 ` Eric Sandeen
2009-02-13 1:47 ` Fernando Luis Vázquez Cao
2009-02-13 6:07 ` Eric Sandeen
2009-02-13 2:23 ` Theodore Tso
2009-02-22 14:15 ` Pavel Machek
2009-02-22 20:51 ` Eric Sandeen
2009-02-22 23:19 ` Theodore Tso
2009-02-22 23:42 ` Jeff Garzik
2009-02-22 23:46 ` Jeff Garzik
2009-02-23 1:23 ` Theodore Tso
2009-02-13 1:14 ` Fernando Luis Vázquez Cao
2009-02-13 6:20 ` Eric Sandeen
2009-02-13 10:36 ` Fernando Luis Vázquez Cao
2009-02-13 12:20 ` Dave Chinner
2009-02-13 16:29 ` Fernando Luis Vazquez Cao [this message]
2009-02-14 11:24 ` Dave Chinner
2009-02-14 13:03 ` Fernando Luis Vázquez Cao
2009-02-14 13:19 ` Fernando Luis Vázquez Cao
2009-02-15 2:48 ` Dave Chinner
2009-02-15 7:11 ` Fernando Luis Vázquez Cao
2009-02-12 10:37 ` util-linux: Add new mount options flushonfsync and noflushonfsync to mount Fernando Luis Vázquez Cao
2009-02-12 10:38 ` util-linux: Add explanation for new mount options flushonfsync and noflushonfsync to mount(8) man page Fernando Luis Vázquez Cao
2009-02-12 10:38 ` block: Add block_flush_device() Fernando Luis Vázquez Cao
2009-02-12 10:39 ` ext3: call blkdev_issue_flush on fsync Fernando Luis Vázquez Cao
2009-02-12 10:40 ` ext4: " Fernando Luis Vázquez Cao
2009-02-15 22:46 ` Theodore Tso
2009-02-16 7:09 ` Fernando Luis Vázquez Cao
2009-02-16 7:25 ` [PATCH 1/3] block: Add block_flush_device() Fernando Luis Vázquez Cao
2009-02-16 7:29 ` [2/3] ext3: call block_flush_device() on fsync Fernando Luis Vázquez Cao
2009-02-16 7:31 ` [PATCH 3/3] ext4: " Fernando Luis Vázquez Cao
2009-01-16 13:59 ` ext4: call blkdev_issue_flush " Fernando Luis Vázquez Cao
2009-01-13 14:42 ` ext2 + -osync: not as easy as it seems Pavel Machek
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=1234542568.9916.183.camel@bladerunner \
--to=fernando@kic.ac.jp \
--cc=alan@lxorguk.ukuu.org.uk \
--cc=david@fromorbit.com \
--cc=fernando@oss.ntt.co.jp \
--cc=jack@suse.cz \
--cc=jens.axboe@oracle.com \
--cc=linux-kernel@vger.kernel.org \
--cc=pavel@suse.cz \
--cc=rwheeler@redhat.com \
--cc=sandeen@redhat.com \
--cc=tytso@MIT.EDU \
/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