From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:49955) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1clYHm-0006V4-VE for qemu-devel@nongnu.org; Wed, 08 Mar 2017 04:53:35 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1clYHi-0003hE-4q for qemu-devel@nongnu.org; Wed, 08 Mar 2017 04:53:35 -0500 Received: from mx1.redhat.com ([209.132.183.28]:38274) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1clYHh-0003h4-Uj for qemu-devel@nongnu.org; Wed, 08 Mar 2017 04:53:30 -0500 References: <1488876478-6889-1-git-send-email-jasowang@redhat.com> <313e32c5-140d-ec13-6b10-98043756f2e6@redhat.com> <61938aa1-2698-f55e-ea1c-e7d96da0b5d7@redhat.com> <20170308103023.5e0d9d81.cornelia.huck@de.ibm.com> From: Jason Wang Message-ID: <31903f0c-a3e7-6acc-5e31-84d08507c513@redhat.com> Date: Wed, 8 Mar 2017 17:53:23 +0800 MIME-Version: 1.0 In-Reply-To: <20170308103023.5e0d9d81.cornelia.huck@de.ibm.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH] virtio: destroy region cache during reset List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Cornelia Huck Cc: Paolo Bonzini , mst@redhat.com, qemu-devel@nongnu.org, peterx@redhat.com On 2017=E5=B9=B403=E6=9C=8808=E6=97=A5 17:30, Cornelia Huck wrote: > On Wed, 8 Mar 2017 14:22:06 +0800 > Jason Wang wrote: > >> On 2017=E5=B9=B403=E6=9C=8808=E6=97=A5 11:21, Jason Wang wrote: >>> On 2017=E5=B9=B403=E6=9C=8807=E6=97=A5 18:55, Paolo Bonzini wrote: >>>> On 07/03/2017 09:47, Jason Wang wrote: >>>>> We don't destroy region cache during reset which can make the maps >>>>> of previous driver leaked to a buggy or malicious driver that don't >>>>> set vring address before starting to use the device. >>>> I'm still not sure as to how this can happen. Reset does clear >>>> desc/used/avail, which should then be checked before accessing the >>>> caches. >>> But the code does not check them in fact? (E.g the attached qtest >>> patch can still pass check-qtest). >>> >>> Thanks >> Ok, the reproducer seems wrong. And I think what you mean is something >> like the check done in virtio_queue_ready(). But looks like not all >> virtqueue check for this. One example is virtio_net_handle_ctrl(), and > Shouldn't the check for desc in virtio_queue_notify_vq() already take > care of that? Yes, I miss this. > >> there may be even more. So you want to fix them all? > Obviously not speaking for Paolo, but I think the virtio core should > have be audited for missing guards. > Probably, otherwise we need a careful auditing of all callers of caches. Thanks