From mboxrd@z Thu Jan  1 00:00:00 1970
From: Wengang Wang <wen.gang.wang@oracle.com>
Date: Tue, 18 Aug 2009 14:32:22 +0800
Subject: [Cluster-devel] [PATCH] gfs2: free disk inode which is deleted
	by remote node
In-Reply-To: <1250501379.3377.10.camel@localhost.localdomain>
References: <200908161044.n7GAijjO019926@rgminet15.oracle.com>
	<1250501379.3377.10.camel@localhost.localdomain>
Message-ID: <4A8A4AF6.9060000@oracle.com>
List-Id: <cluster-devel.redhat.com>
To: cluster-devel.redhat.com
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit

Hi Steve,

Steven Whitehouse wrote:
> Hi,
> 
> Don't worry about the wrong list, I did spot the patch anyway. I agree
> with the principle of what you are trying to do, but something isn't
> quite right. When I tried this on Friday I got a NULL pointer deref the
> first time I tried to touch a file on the fs.
> 
> I think it needs a bit more work, but its certainly going in the right
> direction,

encouraged :)
and yes the dentry->d_inode is set to NULL in path 
vfs_unlink()->d_delete()->dentry_iput(). I didn't test it out since I 
wasn't running on the patched code on the other node, sorry for that.
I'm going to send the v2 version of the patch soon.
if you are going to commit the patch, please help(and feel free) to 
correct mistakes, it there are, in my patch log --my English is not good.

regards,
wengang.