backends/vhost-user.c: prevent using uninitialized vqs

Similar rational to: e6cc11d64f

For vhost scsi and vhost-user-scsi an issue was observed
where, of the 3 virtqueues, seabios would only set cmd,
leaving ctrl and event without a physical address.
This can caused vhost_verify_ring_part_mapping to return
ENOMEM, causing the following logs:

qemu-system-x86_64: Unable to map available ring for ring 0
qemu-system-x86_64: Verify ring failure on region 0

The issue has already been fixed elsewhere, but it was noted
that in backends/vhost-user.c, the vhost_user_backend_dev_init()
function, which other vdevs use in their realize() to initialize
their vqs, was not being properly zeroing out the queues. This
commit ensures hardware modules using the
vhost_user_backend_dev_init() API properly zero out their vqs on
initialization.

Suggested-by: Philippe Mathieu-Daude <philmd@redhat.com>
Signed-off-by: Raphael Norwitz <raphael.norwitz@nutanix.com>
Message-Id: <1566498865-55506-2-git-send-email-raphael.norwitz@nutanix.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
This commit is contained in:
Raphael Norwitz 2019-08-22 11:34:25 -07:00 committed by Michael S. Tsirkin
parent d8438bd24a
commit b8487a3578

View File

@ -46,7 +46,7 @@ vhost_user_backend_dev_init(VhostUserBackend *b, VirtIODevice *vdev,
b->vdev = vdev;
b->dev.nvqs = nvqs;
b->dev.vqs = g_new(struct vhost_virtqueue, nvqs);
b->dev.vqs = g_new0(struct vhost_virtqueue, nvqs);
ret = vhost_dev_init(&b->dev, &b->vhost_user, VHOST_BACKEND_TYPE_USER, 0);
if (ret < 0) {