linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jan Kara <jack@suse.cz>
To: Jens Axboe <jens.axboe@oracle.com>
Cc: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org,
	chris.mason@oracle.com, hch@infradead.org, tytso@mit.edu,
	akpm@linux-foundation.org, jack@suse.cz,
	trond.myklebust@fys.uio.no
Subject: Re: [PATCH 05/11] Assign bdi in super_block
Date: Wed, 16 Sep 2009 14:16:20 +0200	[thread overview]
Message-ID: <20090916121620.GD26030@duck.suse.cz> (raw)
In-Reply-To: <1253038617-30204-6-git-send-email-jens.axboe@oracle.com>

On Tue 15-09-09 20:16:51, Jens Axboe wrote:
> We do this automatically in get_sb_bdev() from the set_bdev_super()
> callback. Filesystems that have their own private backing_dev_info
> must assign that in ->fill_super().
> 
> Note that ->s_bdi assignment is required for proper writeback!
> 
> Acked-by: Christoph Hellwig <hch@infradead.org>
> Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
> ---
>  fs/btrfs/disk-io.c |    1 +
>  fs/fuse/inode.c    |    2 ++
>  fs/super.c         |    6 ++++++
>  fs/sync.c          |    9 ++++++++-
>  fs/ubifs/super.c   |    1 +
>  include/linux/fs.h |    1 +
>  6 files changed, 19 insertions(+), 1 deletions(-)
  I think we agreed that NFS needs to assign it's private BDI to sb->s_bdi,
didn't we? From Tronds comment, that should be enough.

									Honza
> diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c
> index 15831d5..8b81927 100644
> --- a/fs/btrfs/disk-io.c
> +++ b/fs/btrfs/disk-io.c
> @@ -1600,6 +1600,7 @@ struct btrfs_root *open_ctree(struct super_block *sb,
>  
>  	sb->s_blocksize = 4096;
>  	sb->s_blocksize_bits = blksize_bits(4096);
> +	sb->s_bdi = &fs_info->bdi;
>  
>  	/*
>  	 * we set the i_size on the btree inode to the max possible int.
> diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c
> index 4567db6..e5dbecd 100644
> --- a/fs/fuse/inode.c
> +++ b/fs/fuse/inode.c
> @@ -894,6 +894,8 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent)
>  	if (err)
>  		goto err_put_conn;
>  
> +	sb->s_bdi = &fc->bdi;
> +
>  	/* Handle umasking inside the fuse code */
>  	if (sb->s_flags & MS_POSIXACL)
>  		fc->dont_mask = 1;
> diff --git a/fs/super.c b/fs/super.c
> index 9cda337..b03fea8 100644
> --- a/fs/super.c
> +++ b/fs/super.c
> @@ -707,6 +707,12 @@ static int set_bdev_super(struct super_block *s, void *data)
>  {
>  	s->s_bdev = data;
>  	s->s_dev = s->s_bdev->bd_dev;
> +
> +	/*
> +	 * We set the bdi here to the queue backing, file systems can
> +	 * overwrite this in ->fill_super()
> +	 */
> +	s->s_bdi = &bdev_get_queue(s->s_bdev)->backing_dev_info;
>  	return 0;
>  }
>  
> diff --git a/fs/sync.c b/fs/sync.c
> index 1923409..c08467a 100644
> --- a/fs/sync.c
> +++ b/fs/sync.c
> @@ -27,6 +27,13 @@
>   */
>  static int __sync_filesystem(struct super_block *sb, int wait)
>  {
> +	/*
> +	 * This should be safe, as we require bdi backing to actually
> +	 * write out data in the first place
> +	 */
> +	if (!sb->s_bdi)
> +		return 0;
> +
>  	/* Avoid doing twice syncing and cache pruning for quota sync */
>  	if (!wait) {
>  		writeout_quota_sb(sb, -1);
> @@ -101,7 +108,7 @@ restart:
>  		spin_unlock(&sb_lock);
>  
>  		down_read(&sb->s_umount);
> -		if (!(sb->s_flags & MS_RDONLY) && sb->s_root)
> +		if (!(sb->s_flags & MS_RDONLY) && sb->s_root && sb->s_bdi)
>  			__sync_filesystem(sb, wait);
>  		up_read(&sb->s_umount);
>  
> diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c
> index 51763aa..c4af069 100644
> --- a/fs/ubifs/super.c
> +++ b/fs/ubifs/super.c
> @@ -1980,6 +1980,7 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent)
>  	if (err)
>  		goto out_bdi;
>  
> +	sb->s_bdi = &c->bdi;
>  	sb->s_fs_info = c;
>  	sb->s_magic = UBIFS_SUPER_MAGIC;
>  	sb->s_blocksize = UBIFS_BLOCK_SIZE;
> diff --git a/include/linux/fs.h b/include/linux/fs.h
> index db29588..90162fb 100644
> --- a/include/linux/fs.h
> +++ b/include/linux/fs.h
> @@ -1342,6 +1342,7 @@ struct super_block {
>  	int			s_nr_dentry_unused;	/* # of dentry on lru */
>  
>  	struct block_device	*s_bdev;
> +	struct backing_dev_info *s_bdi;
>  	struct mtd_info		*s_mtd;
>  	struct list_head	s_instances;
>  	struct quota_info	s_dquot;	/* Diskquota specific options */
> -- 
> 1.6.4.1.207.g68ea
> 
-- 
Jan Kara <jack@suse.cz>
SUSE Labs, CR

  reply	other threads:[~2009-09-16 12:16 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-09-15 18:16 [PATCH 0/11] Post merge per-bdi writeback patches v3 Jens Axboe
2009-09-15 18:16 ` [PATCH 01/11] fs: remove bdev->bd_inode_backing_dev_info Jens Axboe
2009-09-16 12:06   ` Jan Kara
2009-09-15 18:16 ` [PATCH 02/11] writeback: get rid of wbc->for_writepages Jens Axboe
2009-09-16 12:07   ` Jan Kara
2009-09-15 18:16 ` [PATCH 03/11] writeback: merely wakeup flusher thread if work allocation fails for WB_SYNC_NONE Jens Axboe
2009-09-16 12:08   ` Jan Kara
2009-09-15 18:16 ` [PATCH 04/11] writeback: make wb_writeback() take an argument structure Jens Axboe
2009-09-16 12:53   ` Jan Kara
2009-09-16 13:06     ` Jens Axboe
2009-09-15 18:16 ` [PATCH 05/11] Assign bdi in super_block Jens Axboe
2009-09-16 12:16   ` Jan Kara [this message]
2009-09-16 13:00     ` Jens Axboe
2009-09-15 18:16 ` [PATCH 06/11] writeback: only use bdi_writeback_all() for WB_SYNC_NONE writeout Jens Axboe
2009-09-15 18:16 ` [PATCH 07/11] writeback: use RCU to protect bdi_list Jens Axboe
2009-09-15 18:16 ` [PATCH 08/11] writeback: inline allocation failure handling in bdi_alloc_queue_work() Jens Axboe
2009-09-15 18:16 ` [PATCH 09/11] writeback: separate starting of sync vs opportunistic writeback Jens Axboe
2009-09-16 13:05   ` Jan Kara
2009-09-16 13:07     ` Jens Axboe
2009-09-15 18:16 ` [PATCH 10/11] writeback: splice dirty inode entries to default bdi on bdi_destroy() Jens Axboe
2009-09-16 13:12   ` Jan Kara
2009-09-16 13:21     ` Jens Axboe
2009-09-16 13:29       ` Jan Kara
2009-09-16 18:31         ` Jens Axboe
2009-09-17  9:33           ` Jan Kara
2009-09-15 18:16 ` [PATCH 11/11] writeback: add comments to bdi_work structure Jens Axboe
2009-09-16 13:15   ` Jan Kara

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=20090916121620.GD26030@duck.suse.cz \
    --to=jack@suse.cz \
    --cc=akpm@linux-foundation.org \
    --cc=chris.mason@oracle.com \
    --cc=hch@infradead.org \
    --cc=jens.axboe@oracle.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=trond.myklebust@fys.uio.no \
    --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).