From: Jan Kara <jack@suse.cz>
To: Dave Chinner <david@fromorbit.com>
Cc: linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org,
akpm@linux-foundation.org, davej@redhat.com,
viro@zeniv.linux.org.uk, jack@suse.cz, glommer@parallels.com
Subject: Re: [PATCH 02/11] inode: add IOP_NOTHASHED to avoid inode hash lock in evict
Date: Wed, 31 Jul 2013 16:44:51 +0200 [thread overview]
Message-ID: <20130731144451.GD22930@quack.suse.cz> (raw)
In-Reply-To: <1375244150-27296-3-git-send-email-david@fromorbit.com>
On Wed 31-07-13 14:15:41, Dave Chinner wrote:
> From: Dave Chinner <dchinner@redhat.com>
>
> Some filesystems don't use the VFS inode hash and fake the fact they
> are hashed so that all the writeback code works correctly. However,
> this means the evict() path still tries to remove the inode from the
> hash, meaning that the inode_hash_lock() needs to be taken
> unnecessarily. Hence under certain workloads the inode_hash_lock can
> be contended even if the inode is never actually hashed.
>
> To avoid this, add an inode opflag to allow inode_hash_remove() to
> avoid taking the hash lock on inodes have never actually been
> hashed.
>
> Signed-off-by: Dave Chinner <dchinner@redhat.com>
> ---
> fs/xfs/xfs_iops.c | 2 ++
> include/linux/fs.h | 3 ++-
> 2 files changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c
> index 96dda62..68a8264 100644
> --- a/fs/xfs/xfs_iops.c
> +++ b/fs/xfs/xfs_iops.c
> @@ -1168,8 +1168,10 @@ xfs_setup_inode(
> inode->i_state = I_NEW;
>
> inode_sb_list_add(inode);
> +
> /* make the inode look hashed for the writeback code */
> hlist_add_fake(&inode->i_hash);
> + inode->i_opflags |= IOP_NOTHASHED;
>
> inode->i_mode = ip->i_d.di_mode;
> set_nlink(inode, ip->i_d.di_nlink);
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index b09ddc0..51cf6ed 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -515,6 +515,7 @@ struct posix_acl;
> #define IOP_FASTPERM 0x0001
> #define IOP_LOOKUP 0x0002
> #define IOP_NOFOLLOW 0x0004
> +#define IOP_NOTHASHED 0x0008 /* inode never hashed, avoid unhashing */
>
> /*
> * Keep mostly read-only and often accessed (especially for
> @@ -2371,7 +2372,7 @@ static inline void insert_inode_hash(struct inode *inode)
> extern void __remove_inode_hash(struct inode *);
> static inline void remove_inode_hash(struct inode *inode)
> {
> - if (!inode_unhashed(inode))
> + if (!((inode->i_opflags & IOP_NOTHASHED) || inode_unhashed(inode)))
It's somewhat easier for me to read if the condition if written as:
if (!(inode->i_opflags & IOP_NOTHASHED) && !inode_unhashed(inode))
But whatever... The patch looks good. You can add:
Reviewed-by: Jan Kara <jack@suse.cz>
> __remove_inode_hash(inode);
> }
Honza
--
Jan Kara <jack@suse.cz>
SUSE Labs, CR
next prev parent reply other threads:[~2013-07-31 14:44 UTC|newest]
Thread overview: 32+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-07-31 4:15 [PATCH 00/11] Sync and VFS scalability improvements Dave Chinner
2013-07-31 4:15 ` [PATCH 01/11] writeback: plug writeback at a high level Dave Chinner
2013-07-31 14:40 ` Jan Kara
2013-08-01 5:48 ` Dave Chinner
2013-08-01 8:34 ` Jan Kara
2013-07-31 4:15 ` [PATCH 02/11] inode: add IOP_NOTHASHED to avoid inode hash lock in evict Dave Chinner
2013-07-31 14:44 ` Jan Kara [this message]
2013-08-01 8:12 ` Christoph Hellwig
2013-08-02 1:11 ` Dave Chinner
2013-08-02 14:32 ` Christoph Hellwig
2013-07-31 4:15 ` [PATCH 03/11] inode: convert inode_sb_list_lock to per-sb Dave Chinner
2013-07-31 14:48 ` Jan Kara
2013-07-31 4:15 ` [PATCH 04/11] sync: serialise per-superblock sync operations Dave Chinner
2013-07-31 15:12 ` Jan Kara
2013-07-31 4:15 ` [PATCH 05/11] inode: rename i_wb_list to i_io_list Dave Chinner
2013-07-31 14:51 ` Jan Kara
2013-07-31 4:15 ` [PATCH 06/11] bdi: add a new writeback list for sync Dave Chinner
2013-07-31 15:11 ` Jan Kara
2013-08-01 5:59 ` Dave Chinner
2013-07-31 4:15 ` [PATCH 07/11] writeback: periodically trim the writeback list Dave Chinner
2013-07-31 15:15 ` Jan Kara
2013-08-01 6:16 ` Dave Chinner
2013-08-01 9:03 ` Jan Kara
2013-07-31 4:15 ` [PATCH 08/11] inode: convert per-sb inode list to a list_lru Dave Chinner
2013-08-01 8:19 ` Christoph Hellwig
2013-08-02 1:06 ` Dave Chinner
2013-07-31 4:15 ` [PATCH 09/11] fs: Use RCU lookups for inode cache Dave Chinner
2013-07-31 4:15 ` [PATCH 10/11] list_lru: don't need node lock in list_lru_count_node Dave Chinner
2013-07-31 4:15 ` [PATCH 11/11] list_lru: don't lock during add/del if unnecessary Dave Chinner
2013-07-31 6:48 ` [PATCH 00/11] Sync and VFS scalability improvements Sedat Dilek
2013-08-01 6:19 ` Dave Chinner
2013-08-01 6:31 ` Sedat Dilek
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=20130731144451.GD22930@quack.suse.cz \
--to=jack@suse.cz \
--cc=akpm@linux-foundation.org \
--cc=davej@redhat.com \
--cc=david@fromorbit.com \
--cc=glommer@parallels.com \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=viro@zeniv.linux.org.uk \
/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).