| To: | Christoph Hellwig <hch@xxxxxx>, xfs@xxxxxxxxxxx |
|---|---|
| Subject: | Re: [PATCH] sanitize xfs_initialize_vnode |
| From: | Christoph Hellwig <hch@xxxxxx> |
| Date: | Thu, 24 Jul 2008 08:20:02 +0200 |
| In-reply-to: | <20080724061615.GR6761@disturbed> |
| References: | <20080502105215.GA17870@lst.de> <20080723195110.GA6645@lst.de> <20080724061615.GR6761@disturbed> |
| Sender: | xfs-bounce@xxxxxxxxxxx |
| User-agent: | Mutt/1.3.28i |
On Thu, Jul 24, 2008 at 04:16:15PM +1000, Dave Chinner wrote: > > + } > > + > > + xfs_iflags_clear(ip, XFS_INEW); > > + barrier(); > > + > > + unlock_new_inode(inode); > > +} > > Do we still need that barrier()? Or has the reason for it > existing been lost in the mists of time? Regardless, it was > there before so this is not a reason to stop the patch from > going in... Good question. I wonder why it's there in the first place. |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [PATCH] sanitize xfs_initialize_vnode, Dave Chinner |
|---|---|
| Next by Date: | Re: [PATCH 5/5] implement IHOLD/IRELE directly, Christoph Hellwig |
| Previous by Thread: | Re: [PATCH] sanitize xfs_initialize_vnode, Dave Chinner |
| Next by Thread: | Re: [PATCH] sanitize xfs_initialize_vnode, Christoph Hellwig |
| Indexes: | [Date] [Thread] [Top] [All Lists] |