[PATCH 06/13] xfs: xfs_sync_data is redundant.

[Ben Myers]

Hey Dave,

On Tue, Oct 02, 2012 at 10:14:03AM +1000, Dave Chinner wrote:
> On Mon, Oct 01, 2012 at 12:44:10PM -0500, Ben Myers wrote:
> > Hi Dave,
> > 
> > On Fri, Sep 28, 2012 at 08:34:43AM -0400, Christoph Hellwig wrote:
> > > >  /*
> > > > + * Flush all dirty data to disk. Must not be called while holding an XFS_ILOCK
> > > 
> > > slightly too long line.
> > 
> > It comes out at 78 columns when the patch is applied.  Not a problem.
> > 
> > > > + * or a page lock.
> > > > + */
> > > > +static inline void
> > > > +xfs_flush_inodes(struct xfs_inode *ip)
> > > > +{
> > > > +	writeback_inodes_sb_if_idle(VFS_I(ip)->i_sb, WB_REASON_FS_FREE_SPACE);
> > > > +}
> > > 
> > > Passing an inode here seems very odd, an xfs_mount would be the proper
> > > interface. 
> > > 
> > > Looks fine modulo these cosmetic changes, so:
> > >
> > > Reviewed-by: Christoph Hellwig <hch at lst.de>
> > 
> > I cleaned up these cosmetic changes on your behalf.  Here is the result:
> 
> Looks good. Thanks for doing this, Ben.

No problem.  I wanted to pull this in before the merge window opened.
Otherwise I'd have let you go after it.

> FWIW, The patch I just posted to change
> writeback_inodes_sb_if_idle() to sync_inodes_sb() will need
> massaging to apply to your tree now. I'll update it if you push your
> tree out to oss.sgi.com before taking it...

Sure.  You can find it here:

git://oss.sgi.com/xfs/xfs.git, tmp-for-dave-worker-changes branch

Can you repost the full patch rather than a delta?  That would be cleaner.

Regards,
	Ben