xfs
[Top] [All Lists]

Re: [PATCH 1/3] xfs: rcu free inodes

To: Dave Chinner <david@xxxxxxxxxxxxx>
Subject: Re: [PATCH 1/3] xfs: rcu free inodes
From: "Paul E. McKenney" <paulmck@xxxxxxxxxxxxxxxxxx>
Date: Tue, 14 Dec 2010 12:23:01 -0800
Cc: xfs@xxxxxxxxxxx, eric.dumazet@xxxxxxxxx
In-reply-to: <1292203957-15819-2-git-send-email-david@xxxxxxxxxxxxx>
References: <1292203957-15819-1-git-send-email-david@xxxxxxxxxxxxx> <1292203957-15819-2-git-send-email-david@xxxxxxxxxxxxx>
Reply-to: paulmck@xxxxxxxxxxxxxxxxxx
User-agent: Mutt/1.5.20 (2009-06-14)
On Mon, Dec 13, 2010 at 12:32:35PM +1100, Dave Chinner wrote:
> From: Dave Chinner <dchinner@xxxxxxxxxx>
> 
> Introduce RCU freeing of XFS inodes so that we can convert lookup
> traversals to use rcu_read_lock() protection. This patch only
> introduces the RCU freeing to minimise the potential conflicts with
> mainline if this is merged into mainline via a VFS patchset. It
> abuses the i_dentry list for the RCU callback structure because the
> VFS patches make this a union so it is safe to use like this and
> simplifies and merge issues.
> 
> This patch uses basic RCU freeing rather than SLAB_DESTROY_BY_RCU.
> The later lookup patches need the same "found free inode" protection
> regardless of the RCU freeing method used, so once again the RCU
> freeing method can be dealt with apprpriately at merge time without
> affecting any other code.

There are only two call sites that free into xfs_inode_zone.  One of them
is initialization, before readers have access to the data strcuture,
and the other is covered by this patch.  So looks good to me!

Reviewed-by: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>

> Signed-off-by: Dave Chinner <dchinner@xxxxxxxxxx>
> ---
>  fs/xfs/xfs_iget.c |   14 +++++++++++++-
>  1 files changed, 13 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c
> index cdb1c25..9fae475 100644
> --- a/fs/xfs/xfs_iget.c
> +++ b/fs/xfs/xfs_iget.c
> @@ -105,6 +105,18 @@ xfs_inode_alloc(
>  }
> 
>  void
> +__xfs_inode_free(
> +     struct rcu_head         *head)
> +{
> +     struct inode            *inode = container_of((void *)head,
> +                                                     struct inode, i_dentry);
> +     struct xfs_inode        *ip = XFS_I(inode);
> +
> +     INIT_LIST_HEAD(&inode->i_dentry);
> +     kmem_zone_free(xfs_inode_zone, ip);
> +}
> +
> +void
>  xfs_inode_free(
>       struct xfs_inode        *ip)
>  {
> @@ -147,7 +159,7 @@ xfs_inode_free(
>       ASSERT(!spin_is_locked(&ip->i_flags_lock));
>       ASSERT(completion_done(&ip->i_flush));
> 
> -     kmem_zone_free(xfs_inode_zone, ip);
> +     call_rcu((struct rcu_head *)&VFS_I(ip)->i_dentry, __xfs_inode_free);
>  }
> 
>  /*
> -- 
> 1.7.2.3
> 

<Prev in Thread] Current Thread [Next in Thread>