From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([208.118.235.92]:37035) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UMonI-0003ep-Eu for qemu-devel@nongnu.org; Mon, 01 Apr 2013 20:09:45 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1UMonH-00050p-G9 for qemu-devel@nongnu.org; Mon, 01 Apr 2013 20:09:44 -0400 Received: from mail.linux-iscsi.org ([67.23.28.174]:39660 helo=linux-iscsi.org) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1UMonH-00050l-Bv for qemu-devel@nongnu.org; Mon, 01 Apr 2013 20:09:43 -0400 From: "Nicholas A. Bellinger" Date: Mon, 1 Apr 2013 23:58:24 +0000 Message-Id: <1364860704-11896-4-git-send-email-nab@linux-iscsi.org> In-Reply-To: <1364860704-11896-1-git-send-email-nab@linux-iscsi.org> References: <1364860704-11896-1-git-send-email-nab@linux-iscsi.org> Subject: [Qemu-devel] [PATCH-v2 3/3] vhost: Skip uninitialized VQs in vhost_virtqueue_[start, stop] List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: target-devel Cc: kvm-devel , "Michael S. Tsirkin" , qemu-devel , Nicholas Bellinger , lf-virt , Anthony Liguori , Stefan Hajnoczi , Paolo Bonzini , Asias He From: Nicholas Bellinger This patch adds virtio_queue_valid() checks in vhost_virtqueue_start() and vhost_virtqueue_stop() to avoid uninitialized VQs during vhost-scsi-pci seabios operation, where we currently expect only the request VQ to have been initialized before virtio-scsi LLD guest hand-off. Also, go ahead and skip the same uninitialized VQs during sanity checks within vhost_verify_ring_mappings() by checking vq->ring_[phys,size] directly. Cc: Michael S. Tsirkin Cc: Asias He Cc: Paolo Bonzini Signed-off-by: Nicholas Bellinger --- hw/vhost.c | 12 ++++++++++++ 1 files changed, 12 insertions(+), 0 deletions(-) diff --git a/hw/vhost.c b/hw/vhost.c index 4d6aee3..832cc89 100644 --- a/hw/vhost.c +++ b/hw/vhost.c @@ -314,6 +314,9 @@ static int vhost_verify_ring_mappings(struct vhost_dev *dev, hwaddr l; void *p; + if (!vq->ring_phys || !vq->ring_size) { + continue; + } if (!ranges_overlap(start_addr, size, vq->ring_phys, vq->ring_size)) { continue; } @@ -645,6 +648,10 @@ static int vhost_virtqueue_start(struct vhost_dev *dev, assert(idx >= dev->vq_index && idx < dev->vq_index + dev->nvqs); + if (!virtio_queue_valid(vdev, idx)) { + return 0; + } + vq->num = state.num = virtio_queue_get_num(vdev, idx); r = ioctl(dev->control, VHOST_SET_VRING_NUM, &state); if (r) { @@ -732,6 +739,11 @@ static void vhost_virtqueue_stop(struct vhost_dev *dev, }; int r; assert(idx >= dev->vq_index && idx < dev->vq_index + dev->nvqs); + + if (!virtio_queue_valid(vdev, idx)) { + return; + } + r = ioctl(dev->control, VHOST_GET_VRING_BASE, &state); if (r < 0) { fprintf(stderr, "vhost VQ %d ring restore failed: %d\n", idx, r); -- 1.7.2.5