Re: [Qemu-devel] [PATCH] virtio-pci: warn when both legacy and modern modes are disabled

[Cornelia Huck <cornelia.huck@de.ibm.com>]

On Thu, 21 Jul 2016 11:24:45 +0200
Greg Kurz <groug@kaod.org> wrote:

> On Thu, 12 May 2016 11:45:58 +0200
> Greg Kurz <gkurz@linux.vnet.ibm.com> wrote:
> > Without presuming if we got there because of a user mistake or some
> > more subtile bug in the tooling, it doesn't hurt to log somewhere that

s/subtile/subtle/

> > the device won't be functional.
> > 
> > Signed-off-by: Greg Kurz <gkurz@linux.vnet.ibm.com>
> > ---
> 
> Michael,
> 
> This patch still applies to your tree. Please nack if this is not worth
> being upstreamed.
> 
> Cc'ing Connie for broader audience :)
> 
> Cheers.
> 
> --
> Greg
> 
> >  hw/virtio/virtio-pci.c |    4 ++++
> >  1 file changed, 4 insertions(+)
> > 
> > diff --git a/hw/virtio/virtio-pci.c b/hw/virtio/virtio-pci.c
> > index 0f634d2d776e..a74978cb5e83 100644
> > --- a/hw/virtio/virtio-pci.c
> > +++ b/hw/virtio/virtio-pci.c
> > @@ -1639,6 +1639,10 @@ static void virtio_pci_device_plugged(DeviceState *d, Error **errp)
> >      uint32_t size;
> >      VirtIODevice *vdev = virtio_bus_get_device(&proxy->bus);
> > 
> > +    if (!legacy && !modern) {
> > +        error_report("Warning: device is unserviceable when both legacy and modern modes are disabled. At least one of the disable-modern or disable-legacy properties should be set to false.");
> > +    }

Do you maybe want to fail this at the realize stage already? A device
that is neither legacy nor modern should not exist at all.

> > +
> >      config = proxy->pci_dev.config;
> >      if (proxy->class_code) {
> >          pci_config_set_class(config, proxy->class_code);
> > 
> > 
>