From: Jan Kara <jack@suse.cz>
To: Niu Yawei <yawei.niu@gmail.com>
Cc: Christoph Hellwig <hch@infradead.org>,
jack@suse.cz, linux-fsdevel@vger.kernel.org,
linux-ext4@vger.kernel.org, yawei.niu@intel.com,
andreas.dilger@intel.com, lai.siyao@intel.com
Subject: Re: [PATCH 2/3 v2] quota: avoid unnecessary dqget()/dqput() calls
Date: Mon, 2 Jun 2014 09:42:45 +0200 [thread overview]
Message-ID: <20140602074245.GB3224@quack.suse.cz> (raw)
In-Reply-To: <538541A3.2060102@gmail.com>
On Wed 28-05-14 09:53:39, Niu Yawei wrote:
> Avoid unnecessary dqget()/dqput() calls in __dquot_initialize(),
> that will introduce global lock contention otherwise.
The patch looks good. Just two minor nits below.
Honza
> Signed-off-by: Lai Siyao <lai.siyao@intel.com>
> Signed-off-by: Niu Yawei <yawei.niu@intel.com>
> ---
> fs/quota/dquot.c | 13 ++++++++++++-
> 1 files changed, 12 insertions(+), 1 deletions(-)
>
> diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c
> index 9cd5f63..dc6f711 100644
> --- a/fs/quota/dquot.c
> +++ b/fs/quota/dquot.c
> @@ -1400,7 +1400,7 @@ static int dquot_active(const struct inode *inode)
> */
> static void __dquot_initialize(struct inode *inode, int type)
> {
> - int cnt;
> + int cnt, dq_get = 0;
> struct dquot *got[MAXQUOTAS];
> struct super_block *sb = inode->i_sb;
> qsize_t rsv;
> @@ -1416,6 +1416,13 @@ static void __dquot_initialize(struct inode *inode, int type)
> got[cnt] = NULL;
> if (type != -1 && cnt != type)
> continue;
> + /* the i_dquot should have been initialized in most case,
cases ^^^
> + * we check it without locking here to avoid unnecessary
> + * dqget()/dqput() calls. */
Please fix the comment style - scripts/checkpatch.pl is your friend.
> + if (inode->i_dquot[cnt])
> + continue;
> + dq_get = 1;
init_needed or something like that would be a better name I think.
> +
> switch (cnt) {
> case USRQUOTA:
> qid = make_kqid_uid(inode->i_uid);
> @@ -1427,6 +1434,10 @@ static void __dquot_initialize(struct inode *inode, int type)
> got[cnt] = dqget(sb, qid);
> }
>
> + /* all required i_dquot has been initialized */
> + if (!dq_get)
> + return;
> +
> down_write(&sb_dqopt(sb)->dqptr_sem);
> if (IS_NOQUOTA(inode))
> goto out_err;
> --
> 1.7.1
>
>
--
Jan Kara <jack@suse.cz>
SUSE Labs, CR
next prev parent reply other threads:[~2014-06-02 7:42 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-05-22 10:47 [PATCH] quota: remove dqptr_sem for scalability Niu Yawei
2014-05-22 13:25 ` Jan Kara
2014-05-23 3:37 ` Niu Yawei
2014-05-27 10:13 ` [PATCH 2/3] quota: avoid unnecessary dqget()/dqput() calls Niu Yawei
2014-05-27 10:15 ` [PATCH 3/3] quota: remove dqptr_sem Niu Yawei
2014-05-23 4:02 ` [PATCH] quota: remove dqptr_sem for scalability Eric Sandeen
2014-05-23 5:22 ` Niu Yawei
2014-05-23 13:02 ` Eric Sandeen
2014-05-27 10:10 ` [PATCH 1/3] quota: protect Q_GETFMT by dqonoff_mutex Niu Yawei
2014-05-27 10:12 ` Christoph Hellwig
2014-05-27 10:28 ` Niu Yawei
2014-05-28 1:52 ` [PATCH 1/3 v2] " Niu Yawei
2014-06-02 7:32 ` Jan Kara
2014-05-28 1:53 ` [PATCH 2/3 v2] quota: avoid unnecessary dqget()/dqput() calls Niu Yawei
2014-06-02 7:42 ` Jan Kara [this message]
2014-05-28 1:55 ` [PATCH 3/3 v2] quota: remove dqptr_sem Niu Yawei
2014-05-28 2:01 ` Niu Yawei
2014-06-02 8:34 ` Jan Kara
2014-06-03 9:51 ` Niu Yawei
2014-06-03 15:43 ` Jan Kara
2014-06-04 4:19 ` [PATCH 1/5] quota: protect Q_GETFMT by dqonoff_mutex Niu Yawei
2014-06-04 15:36 ` Jan Kara
2014-06-04 4:20 ` [PATCH 2/5] quota: avoid unnecessary dqget()/dqput() calls Niu Yawei
2014-06-04 4:21 ` [PATCH 3/5] quota: simplify remove_inode_dquot_ref() Niu Yawei
2014-06-04 4:22 ` [PATCH 4/5] quota: missing lock in dqcache_shrink_scan() Niu Yawei
2014-06-04 4:23 ` [PATCH 5/5] quota: remove dqptr_sem Niu Yawei
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=20140602074245.GB3224@quack.suse.cz \
--to=jack@suse.cz \
--cc=andreas.dilger@intel.com \
--cc=hch@infradead.org \
--cc=lai.siyao@intel.com \
--cc=linux-ext4@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=yawei.niu@gmail.com \
--cc=yawei.niu@intel.com \
/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).