From: Eric Biggers <ebiggers@kernel.org>
To: Jan Kara <jack@suse.cz>
Cc: linux-fsdevel@vger.kernel.org, linux-ext4@vger.kernel.org,
linux-f2fs-devel@lists.sourceforge.net,
linux-xfs@vger.kernel.org, Theodore Ts'o <tytso@mit.edu>,
Christoph Hellwig <hch@lst.de>,
stable@vger.kernel.org
Subject: Re: [PATCH 01/13] fs: avoid double-writing inodes on lazytime expiration
Date: Thu, 7 Jan 2021 13:46:37 -0800 [thread overview]
Message-ID: <X/eBPZ+kLGuz2NDC@gmail.com> (raw)
In-Reply-To: <20210107144709.GG12990@quack2.suse.cz>
On Thu, Jan 07, 2021 at 03:47:09PM +0100, Jan Kara wrote:
> > diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
> > index acfb55834af23..081e335cdee47 100644
> > --- a/fs/fs-writeback.c
> > +++ b/fs/fs-writeback.c
> > @@ -1509,11 +1509,22 @@ __writeback_single_inode(struct inode *inode, struct writeback_control *wbc)
> >
> > spin_unlock(&inode->i_lock);
> >
> > - if (dirty & I_DIRTY_TIME)
> > - mark_inode_dirty_sync(inode);
> > /* Don't write the inode if only I_DIRTY_PAGES was set */
> > if (dirty & ~I_DIRTY_PAGES) {
> > - int err = write_inode(inode, wbc);
> > + int err;
> > +
> > + /*
> > + * If the inode is being written due to a lazytime timestamp
> > + * expiration, then the filesystem needs to be notified about it
> > + * so that e.g. the filesystem can update on-disk fields and
> > + * journal the timestamp update. Just calling write_inode()
> > + * isn't enough. Don't call mark_inode_dirty_sync(), as that
> > + * would put the inode back on the dirty list.
> > + */
> > + if ((dirty & I_DIRTY_TIME) && inode->i_sb->s_op->dirty_inode)
> > + inode->i_sb->s_op->dirty_inode(inode, I_DIRTY_SYNC);
> > +
> > + err = write_inode(inode, wbc);
> > if (ret == 0)
> > ret = err;
> > }
>
> I have to say I dislike this special call of ->dirty_inode(). It works but
> it makes me wonder, didn't we forget about something or won't we forget in
> the future? Because it's very easy to miss this special case...
>
> I think attached patch (compile-tested only) should actually fix the
> problem as well without this special ->dirty_inode() call. It basically
> only moves the mark_inode_dirty_sync() before inode->i_state clearing.
> Because conceptually mark_inode_dirty_sync() is IMO the right function to
> call. It will take care of clearing I_DIRTY_TIME flag (because we are
> setting I_DIRTY_SYNC), it will also not touch inode->i_io_list if the inode
> is queued for sync (I_SYNC_QUEUED is set in that case). The only problem
> with calling it was that it was called *after* clearing dirty bits from
> i_state... What do you think?
>
> Honza
> --
> Jan Kara <jack@suse.com>
> SUSE Labs, CR
> From 80ccc6a78d1c0532f600b98884f7a64e58333485 Mon Sep 17 00:00:00 2001
> From: Jan Kara <jack@suse.cz>
> Date: Thu, 7 Jan 2021 15:36:05 +0100
> Subject: [PATCH] fs: Make sure inode is clean after __writeback_single_inode()
>
> Signed-off-by: Jan Kara <jack@suse.cz>
> ---
> fs/fs-writeback.c | 23 ++++++++++++-----------
> 1 file changed, 12 insertions(+), 11 deletions(-)
>
> diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c
> index acfb55834af2..b9356f470fae 100644
> --- a/fs/fs-writeback.c
> +++ b/fs/fs-writeback.c
> @@ -1473,22 +1473,25 @@ __writeback_single_inode(struct inode *inode, struct writeback_control *wbc)
> ret = err;
> }
>
> + /*
> + * If inode has dirty timestamps and we need to write them, call
> + * mark_inode_dirty_sync() to notify filesystem about it.
> + */
> + if (inode->i_state & I_DIRTY_TIME &&
> + (wbc->for_sync || wbc->sync_mode == WB_SYNC_ALL ||
> + time_after(jiffies, inode->dirtied_time_when +
> + dirtytime_expire_interval * HZ))) {
> + trace_writeback_lazytime(inode);
> + mark_inode_dirty_sync(inode);
> + }
> +
> /*
> * Some filesystems may redirty the inode during the writeback
> * due to delalloc, clear dirty metadata flags right before
> * write_inode()
> */
> spin_lock(&inode->i_lock);
> -
> dirty = inode->i_state & I_DIRTY;
> - if ((inode->i_state & I_DIRTY_TIME) &&
> - ((dirty & I_DIRTY_INODE) ||
> - wbc->sync_mode == WB_SYNC_ALL || wbc->for_sync ||
> - time_after(jiffies, inode->dirtied_time_when +
> - dirtytime_expire_interval * HZ))) {
> - dirty |= I_DIRTY_TIME;
> - trace_writeback_lazytime(inode);
> - }
> inode->i_state &= ~dirty;
>
> /*
> @@ -1509,8 +1512,6 @@ __writeback_single_inode(struct inode *inode, struct writeback_control *wbc)
>
> spin_unlock(&inode->i_lock);
>
> - if (dirty & I_DIRTY_TIME)
> - mark_inode_dirty_sync(inode);
> /* Don't write the inode if only I_DIRTY_PAGES was set */
> if (dirty & ~I_DIRTY_PAGES) {
> int err = write_inode(inode, wbc);
It looks like that's going to work, and it fixes the XFS bug too.
Note that if __writeback_single_inode() is called from writeback_single_inode()
(rather than writeback_sb_inodes()), then the inode might not be queued for
sync, in which case mark_inode_dirty_sync() will move it to a writeback list.
That's okay because afterwards, writeback_single_inode() will delete the inode
from any writeback list if it's been fully cleaned, right? So clean inodes
won't get left on a writeback list.
It's confusing because there are comments in writeback_single_inode() and above
__writeback_single_inode() that say that the inode must not be moved between
writeback lists. I take it that those comments are outdated, as they predate
I_SYNC_QUEUED being introduced by commit 5afced3bf281 ("writeback: Avoid
skipping inode writeback")?
- Eric
next prev parent reply other threads:[~2021-01-07 21:47 UTC|newest]
Thread overview: 37+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-05 0:54 [PATCH 00/13] lazytime fixes and cleanups Eric Biggers
2021-01-05 0:54 ` [PATCH 01/13] fs: avoid double-writing inodes on lazytime expiration Eric Biggers
2021-01-07 14:47 ` Jan Kara
2021-01-07 14:58 ` Matthew Wilcox
2021-01-07 21:46 ` Eric Biggers [this message]
2021-01-08 8:54 ` Christoph Hellwig
2021-01-08 9:01 ` Christoph Hellwig
2021-01-09 17:11 ` Eric Biggers
2021-01-05 0:54 ` [PATCH 02/13] gfs2: don't worry about I_DIRTY_TIME in gfs2_fsync() Eric Biggers
2021-01-08 8:54 ` Christoph Hellwig
2021-01-05 0:54 ` [PATCH 03/13] fs: only specify I_DIRTY_TIME when needed in generic_update_time() Eric Biggers
2021-01-08 8:57 ` Christoph Hellwig
2021-01-05 0:54 ` [PATCH 04/13] fat: only specify I_DIRTY_TIME when needed in fat_update_time() Eric Biggers
2021-01-07 13:13 ` Jan Kara
2021-01-07 19:10 ` Eric Biggers
2021-01-05 0:54 ` [PATCH 05/13] fs: don't call ->dirty_inode for lazytime timestamp updates Eric Biggers
2021-01-07 13:17 ` Jan Kara
2021-01-07 13:18 ` Jan Kara
2021-01-08 9:02 ` Christoph Hellwig
2021-01-05 0:54 ` [PATCH 06/13] fs: pass only I_DIRTY_INODE flags to ->dirty_inode Eric Biggers
2021-01-08 9:02 ` Christoph Hellwig
2021-01-05 0:54 ` [PATCH 07/13] fs: correctly document the inode dirty flags Eric Biggers
2021-01-08 9:03 ` Christoph Hellwig
2021-01-05 0:54 ` [PATCH 08/13] ext4: simplify i_state checks in __ext4_update_other_inode_time() Eric Biggers
2021-01-07 13:24 ` Jan Kara
2021-01-07 19:06 ` Eric Biggers
2021-01-05 0:54 ` [PATCH 09/13] fs: drop redundant checks from __writeback_single_inode() Eric Biggers
2021-01-08 9:12 ` Christoph Hellwig
2021-01-05 0:54 ` [PATCH 10/13] fs: clean up __mark_inode_dirty() a bit Eric Biggers
2021-01-08 9:10 ` Christoph Hellwig
2021-01-05 0:54 ` [PATCH 11/13] fs: add a lazytime_expired method Eric Biggers
2021-01-07 14:02 ` Jan Kara
2021-01-07 22:05 ` Eric Biggers
2021-01-08 9:14 ` Christoph Hellwig
2021-01-05 0:54 ` [PATCH 12/13] xfs: remove a stale comment from xfs_file_aio_write_checks() Eric Biggers
2021-01-08 9:15 ` Christoph Hellwig
2021-01-05 0:54 ` [PATCH 13/13] xfs: implement lazytime_expired Eric Biggers
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=X/eBPZ+kLGuz2NDC@gmail.com \
--to=ebiggers@kernel.org \
--cc=hch@lst.de \
--cc=jack@suse.cz \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-f2fs-devel@lists.sourceforge.net \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-xfs@vger.kernel.org \
--cc=stable@vger.kernel.org \
--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;
as well as URLs for NNTP newsgroup(s).