On Wed, 6 Sep 2017 14:49:52 +0200
Halil Pasic <pa...@linux.vnet.ibm.com> wrote:
> On 09/06/2017 02:42 PM, Cornelia Huck wrote:
> > On Tue, 5 Sep 2017 13:16:43 +0200
> > Halil Pasic <pa...@linux.vnet.ibm.com> wrote:
> >> @@ -488,7 +446,8 @@ static int virtio_ccw_cb(SubchDev *sch, CCW1 ccw)
> >> } else {
> >> virtio_bus_get_vdev_config(&dev->bus, vdev->config);
> >> /* XXX config space endianness */
> >
> > Unrelated: That should be fine, I guess?
> >
> >> - cpu_physical_memory_write(ccw.cda, vdev->config, len);
> >> + /* TODO we may have made -EINVAL out of -EFAULT */
> >
> > Eek.
> >
>
> Actually I wanted to send a patch which gets rid of the -EFAULT case
> in sch_handle_start_func_virtual.
>
> IMHO a program check is more appropriate here. We did the EFAULT if
> cpu_physical_memory_map failed, but we don't have that any more.
OK, such a patch would make this easier to figure out.
