linux-ext4.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Zheng Liu <gnehzuil.liu@gmail.com>
To: Jan Kara <jack@suse.cz>
Cc: Ted Tso <tytso@mit.edu>, Akira Fujita <a-fujita@rs.jp.nec.com>,
	linux-ext4@vger.kernel.org
Subject: Re: [PATCH 1/3] ext4: Retry allocation when inline->extent conversion failed
Date: Thu, 12 Dec 2013 18:30:20 +0800	[thread overview]
Message-ID: <20131212103020.GA3239@gmail.com> (raw)
In-Reply-To: <20131212093936.GA1886@quack.suse.cz>

On Thu, Dec 12, 2013 at 10:39:36AM +0100, Jan Kara wrote:
> On Thu 12-12-13 14:44:22, Zheng Liu wrote:
> > Hi Jan,
> > 
> > On Tue, Dec 10, 2013 at 11:13:58AM +0100, Jan Kara wrote:
> > > Similarly as other ->write_begin functions in ext4, also
> > > ext4_da_write_inline_data_begin() should retry allocation if the
> > > conversion failed because of ENOSPC. This avoids returning ENOSPC
> > > prematurely because of uncommitted block deletions.
> > 
> > I don't see why we need to try again here.  If there is no enough space
> > in inline data, ext4_da_convert_inline_data_to_extent() just read the
> > inline data into a page and clear EXT4_STATE_MAY_INLINE_DATA flag.  If
> > we try again, we will encounter an ENOSPC error again.  So why do we
> > need to retry allocation?
>   Umm, I don't think so. ext4_da_convert_inline_data_to_extent() calls
> ext4_da_get_block_prep() and that can falsely fail with ENOSPC if there are
> blocks freed in currently running transaction. So we call
> ext4_should_retry_alloc() which forces commit of the current transaction
> and then we retry the allocation.

Ah, I see.  Thanks for your explanation.

                                                - Zheng

> 
> 								Honza
> 
> > > 
> > > Signed-off-by: Jan Kara <jack@suse.cz>
> > > ---
> > >  fs/ext4/inline.c | 7 +++++++
> > >  1 file changed, 7 insertions(+)
> > > 
> > > diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c
> > > index bae987549dc3..ed6e71fe5e9d 100644
> > > --- a/fs/ext4/inline.c
> > > +++ b/fs/ext4/inline.c
> > > @@ -849,11 +849,13 @@ int ext4_da_write_inline_data_begin(struct address_space *mapping,
> > >  	handle_t *handle;
> > >  	struct page *page;
> > >  	struct ext4_iloc iloc;
> > > +	int retries;
> > >  
> > >  	ret = ext4_get_inode_loc(inode, &iloc);
> > >  	if (ret)
> > >  		return ret;
> > >  
> > > +retry_journal:
> > >  	handle = ext4_journal_start(inode, EXT4_HT_INODE, 1);
> > >  	if (IS_ERR(handle)) {
> > >  		ret = PTR_ERR(handle);
> > > @@ -875,6 +877,11 @@ int ext4_da_write_inline_data_begin(struct address_space *mapping,
> > >  							    inode,
> > >  							    flags,
> > >  							    fsdata);
> > > +		ext4_journal_stop(handle);
> > > +		handle = NULL;
> > > +		if (ret == -ENOSPC &&
> > > +		    ext4_should_retry_alloc(inode->i_sb, &retries))
> > > +			goto retry_journal;
> > >  		goto out;
> > >  	}
> > >  
> > > -- 
> > > 1.8.1.4
> > > 
> > > --
> > > To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> > > the body of a message to majordomo@vger.kernel.org
> > > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> -- 
> Jan Kara <jack@suse.cz>
> SUSE Labs, CR

  reply	other threads:[~2013-12-12 10:27 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-12-10 10:13 [PATCH 0/3] Fix ext4 deadlock when running xfstests/269 Jan Kara
2013-12-10 10:13 ` [PATCH 1/3] ext4: Retry allocation when inline->extent conversion failed Jan Kara
2013-12-12  6:44   ` Zheng Liu
2013-12-12  9:39     ` Jan Kara
2013-12-12 10:30       ` Zheng Liu [this message]
2013-12-18  5:45   ` Theodore Ts'o
2013-12-10 10:13 ` [PATCH 2/3] ext4: Standardize error handling in ext4_da_write_inline_data_begin() Jan Kara
2013-12-18  5:45   ` Theodore Ts'o
2013-12-10 10:14 ` [PATCH 3/3] ext4: Fix deadlock when writing in ENOSPC conditions Jan Kara
2013-12-12  6:51   ` Zheng Liu
2013-12-12  9:41     ` Jan Kara
2013-12-18  5:45   ` 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=20131212103020.GA3239@gmail.com \
    --to=gnehzuil.liu@gmail.com \
    --cc=a-fujita@rs.jp.nec.com \
    --cc=jack@suse.cz \
    --cc=linux-ext4@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).