From mboxrd@z Thu Jan  1 00:00:00 1970
From: viro@parcelfarce.linux.theplanet.co.uk
Subject: Re: [PATCH] ext3 [linux-2.6.2.]: accessing already freed inodes when under memory pressure
Date: Fri, 2 Apr 2004 20:10:00 +0100
Sender: linux-fsdevel-owner@vger.kernel.org
Message-ID: <20040402191000.GD31500@parcelfarce.linux.theplanet.co.uk>
References: <OFB7BF1E31.8824C0B7-ONC1256E67.004015BB-C1256E67.0041B813@de.ibm.com> <1080653969.24117.192.camel@hades.cambridge.redhat.com> <Pine.LNX.4.58.0403300745040.1096@ppc970.osdl.org> <20040402161223.GZ31500@parcelfarce.linux.theplanet.co.uk> <20040402180111.GB31500@parcelfarce.linux.theplanet.co.uk> <Pine.LNX.4.58.0404021050220.1122@ppc970.osdl.org> <Pine.LNX.4.58.0404021100020.1122@ppc970.osdl.org>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Cc: David Woodhouse <dwmw2@infradead.org>,
	Martin Schwidefsky <schwidefsky@de.ibm.com>,
	Andrew Morton <akpm@osdl.org>, Carsten Otte <cotte@de.ibm.com>,
	Carsten Otte <cotte@freenet.de>, linux-fsdevel@vger.kernel.org,
	sct@redhat.com, Dave Kleikamp <shaggy@austin.ibm.com>
Return-path: <linux-fsdevel-owner@vger.kernel.org>
Received: from parcelfarce.linux.theplanet.co.uk ([195.92.249.252]:52636 "EHLO
	www.linux.org.uk") by vger.kernel.org with ESMTP id S264157AbUDBTKD
	(ORCPT <rfc822;linux-fsdevel@vger.kernel.org>);
	Fri, 2 Apr 2004 14:10:03 -0500
To: Linus Torvalds <torvalds@osdl.org>
Content-Disposition: inline
In-Reply-To: <Pine.LNX.4.58.0404021100020.1122@ppc970.osdl.org>
List-Id: linux-fsdevel.vger.kernel.org

On Fri, Apr 02, 2004 at 11:02:24AM -0800, Linus Torvalds wrote:
> 
> 
> On Fri, 2 Apr 2004, Linus Torvalds wrote:
> > 
> > I think that -EXDEV is not only OK, it's _required_. It may be the same 
> > physical filesystem, but it's a different mount and thus a different 
> > filesystem.
> 
> Actually, we already do this in "sys_link()" and "do_rename()":
> 
> sys_link:
> 
>         error = -EXDEV;
>         if (old_nd.mnt != nd.mnt)
>                 goto out_release;
>         new_dentry = lookup_create(&nd, 0);
 
> so this case should already be ok, no?

This case is OK now, but we'll need to do explicit lookup_mnt() since we are
switching to LOOKUP_PARENT for old_mnt.