From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:53928) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1gU0S9-0001mx-M6 for qemu-devel@nongnu.org; Mon, 03 Dec 2018 21:28:50 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1gU0S4-0006JI-KX for qemu-devel@nongnu.org; Mon, 03 Dec 2018 21:28:49 -0500 Received: from mx1.redhat.com ([209.132.183.28]:41718) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1gU0S4-0006IE-84 for qemu-devel@nongnu.org; Mon, 03 Dec 2018 21:28:44 -0500 References: <20181203100608.28538-1-jasowang@redhat.com> From: Jason Wang Message-ID: Date: Tue, 4 Dec 2018 10:28:32 +0800 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH V4 for 3.1 0/4] Fix possible OOB during queuing packets List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Peter Maydell Cc: QEMU Developers , "Michael S. Tsirkin" , P J P , Li Qiang , Li Qiang , Paolo Bonzini , Thomas Huth , Eric Blake On 2018/12/4 =E4=B8=8A=E5=8D=8812:18, Peter Maydell wrote: > On Mon, 3 Dec 2018 at 10:06, Jason Wang wrote: >> Hi: >> >> This series tries to fix a possible OOB during queueing packets >> through qemu_net_queue_append_iov(). This could happen when it tries >> to queue a packet whose size is larger than INT_MAX which may lead >> integer overflow. We've fixed similar issue in the past during >> qemu_net_queue_deliver_iov() by ignoring large packets there. Let's >> just move the check earlier to qemu_sendv_packet_async() and reduce >> the limitation to NET_BUFSIZE. A simple qtest were also added this. >> >> Please review. > I did a test build and run, and the new test generates warning > messages during "make check": > /ppc64/virtio/net/pci/large_tx: > qemu-system-ppc64: warning: hub 0 is not connected to host network > (similarly for /i386/ and /x86_64/). > > thank > -- PMM This is intended, we don't need any host networking device. This makes=20 it very simpler and easier to trigger the queuing at hub. I can add a=20 patch to suppress this warning if qtest is enabled. Thanks