From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([208.118.235.92]:36364)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <stefanha@redhat.com>) id 1TO8Oc-0005aR-8C
	for qemu-devel@nongnu.org; Tue, 16 Oct 2012 10:45:31 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <stefanha@redhat.com>) id 1TO8OW-0001qN-Cf
	for qemu-devel@nongnu.org; Tue, 16 Oct 2012 10:45:26 -0400
Received: from mx1.redhat.com ([209.132.183.28]:34583)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <stefanha@redhat.com>) id 1TO8OW-0001qA-2e
	for qemu-devel@nongnu.org; Tue, 16 Oct 2012 10:45:20 -0400
Date: Tue, 16 Oct 2012 16:42:12 +0200
From: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20121016144212.GA19559@stefanha-thinkpad>
References: <1350305057-6287-1-git-send-email-stefanha@redhat.com>
	<1350305057-6287-3-git-send-email-stefanha@redhat.com>
	<507D32E0.7020006@redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <507D32E0.7020006@redhat.com>
Subject: Re: [Qemu-devel] [PATCH v2 2/3] qemu-img: Detect backing file chain
 infinite loops
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Kevin Wolf <kwolf@redhat.com>
Cc: kashyap.cv@gmail.com, qemu-devel@nongnu.org, =?iso-8859-1?Q?Beno=EEt?= Canet <benoit@irqsave.net>

On Tue, Oct 16, 2012 at 12:11:44PM +0200, Kevin Wolf wrote:
> Am 15.10.2012 14:44, schrieb Stefan Hajnoczi:
> > A malicious or corruption image can contain an infinite loop of backing
> > files.  The qemu-img info --backing-chain command must not hang when
> > such files are encountered.
> > 
> > Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
> 
> This seems to do what is intended, but I think rather than fixing the
> 'qemu-img info' special case I'd have fixed bdrv_open() to detect the
> situation.

When bdrv_open() on the chain fails we'll still need to iterate each
image (with infinite loop detection) in qemu-img info.  This allows
qemu-img info to print out details of the chain and where the chain goes
wrong.

> Do we need to properly communicate what the broken image is with JSON
> output? This patch will only produce broken JSON in the failure case.

Ah, you're right.  We must close the JSON array and add a test case for
this :(.

Will send v3.

Stefan