On Wed, Feb 05, 2020 at 02:56:57PM +, Stefan Hajnoczi wrote:
> On Wed, Feb 05, 2020 at 01:13:37AM -0500, Michael S. Tsirkin wrote:
> > On Tue, Feb 04, 2020 at 03:16:18PM +, Stefan Hajnoczi wrote:
> > > hw/virtio.c checks vq->vring.desc != NULL to see if the vring has been
> > > set up succe
On Wed, 5 Feb 2020 14:49:46 +
Stefan Hajnoczi wrote:
> On Tue, Feb 04, 2020 at 05:02:39PM +0100, Cornelia Huck wrote:
> > On Tue, 4 Feb 2020 15:16:18 +
> > Stefan Hajnoczi wrote:
> > > diff --git a/hw/virtio/virtio.c b/hw/virtio/virtio.c
> > > index 2c5410e981..5d7f619a1e 100644
> > >
On Wed, Feb 05, 2020 at 01:13:37AM -0500, Michael S. Tsirkin wrote:
> On Tue, Feb 04, 2020 at 03:16:18PM +, Stefan Hajnoczi wrote:
> > hw/virtio.c checks vq->vring.desc != NULL to see if the vring has been
> > set up successfully.
> >
> > When virtio_queue_set_rings() fails due to an invalid v
On Tue, Feb 04, 2020 at 05:02:39PM +0100, Cornelia Huck wrote:
> On Tue, 4 Feb 2020 15:16:18 +
> Stefan Hajnoczi wrote:
>
> > hw/virtio.c checks vq->vring.desc != NULL to see if the vring has been
> > set up successfully.
> >
> > When virtio_queue_set_rings() fails due to an invalid vring m
On Tue, Feb 04, 2020 at 03:16:18PM +, Stefan Hajnoczi wrote:
> hw/virtio.c checks vq->vring.desc != NULL to see if the vring has been
> set up successfully.
>
> When virtio_queue_set_rings() fails due to an invalid vring memory
> address it must clear vq->vring.desc (and related fields) so we
On Tue, Feb 04, 2020 at 03:16:18PM +, Stefan Hajnoczi wrote:
> hw/virtio.c checks vq->vring.desc != NULL to see if the vring has been
> set up successfully.
>
> When virtio_queue_set_rings() fails due to an invalid vring memory
> address it must clear vq->vring.desc (and related fields) so we
On Tue, 4 Feb 2020 15:16:18 +
Stefan Hajnoczi wrote:
> hw/virtio.c checks vq->vring.desc != NULL to see if the vring has been
> set up successfully.
>
> When virtio_queue_set_rings() fails due to an invalid vring memory
> address it must clear vq->vring.desc (and related fields) so we don't
hw/virtio.c checks vq->vring.desc != NULL to see if the vring has been
set up successfully.
When virtio_queue_set_rings() fails due to an invalid vring memory
address it must clear vq->vring.desc (and related fields) so we don't
treat this virtqueue as successfully initialized later on.
This bug