xfs
[Top] [All Lists]

Re: [PATCH 3/5] xfs: fold xfs_attr_remove_int into xfs_attr_remove

To: Brian Foster <bfoster@xxxxxxxxxx>
Subject: Re: [PATCH 3/5] xfs: fold xfs_attr_remove_int into xfs_attr_remove
From: Dave Chinner <david@xxxxxxxxxxxxx>
Date: Tue, 6 May 2014 06:56:53 +1000
Cc: Christoph Hellwig <hch@xxxxxx>, xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <20140505202127.GD12448@xxxxxxxxxxxxxx>
References: <1399130415-5382-1-git-send-email-hch@xxxxxx> <1399130415-5382-4-git-send-email-hch@xxxxxx> <20140505202127.GD12448@xxxxxxxxxxxxxx>
User-agent: Mutt/1.5.21 (2010-09-15)
On Mon, May 05, 2014 at 04:21:28PM -0400, Brian Foster wrote:
> On Sat, May 03, 2014 at 05:20:13PM +0200, Christoph Hellwig wrote:
> > Also remove a useless ilock roundtrip for the first attr fork check, it's
> > racy anyway and we redo it later under the ilock before we start the 
> > removal.
> > 
> > Plus various minor style fixes to the new xfs_attr_remove.
> > 
> > Signed-off-by: Christoph Hellwig <hch@xxxxxx>
.....
> > @@ -487,35 +497,26 @@ xfs_attr_remove_int(xfs_inode_t *dp, struct xfs_name 
> > *name, int flags)
> >      */
> >     xfs_trans_ijoin(args.trans, dp, 0);
> >  
> > -   /*
> > -    * Decide on what work routines to call based on the inode size.
> > -    */
> >     if (!xfs_inode_hasattr(dp)) {
> >             error = XFS_ERROR(ENOATTR);
> 
> I suppose we probably want to nuke the XFS_ERROR() while we're here..?
> Otherwise, it looks good.

No need, I'll do that at the end of the dev cycle for everything.

Cheers,

Dave.

-- 
Dave Chinner
david@xxxxxxxxxxxxx

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