On Thu, Mar 30, 2017 at 11:06:27AM +0200, Cornelia Huck wrote:
> On Wed, 29 Mar 2017 20:14:44 +0300
> "Michael S. Tsirkin" <m...@redhat.com> wrote:
>
> > Some drivers can't support all features in all configurations. At the
> > moment we blindly set FEATURES_OK and later FAILED. Support this better
> > by adding a callback drivers can use to do some early checks.
>
> Looks reasonable. Do we need to document that the driver must not do
> anything beyond dealing with features and reading the config space that
> early?
It's up to the driver - we probably should document that on failure
neither probe nor remove will be called. On success we proceed
to probe.
> >
> > Signed-off-by: Michael S. Tsirkin <m...@redhat.com>
> > ---
> > drivers/virtio/virtio.c | 6 ++++++
> > include/linux/virtio.h | 1 +
> > 2 files changed, 7 insertions(+)
> >
> > diff --git a/drivers/virtio/virtio.c b/drivers/virtio/virtio.c
> > index 400d70b..48230a5 100644
> > --- a/drivers/virtio/virtio.c
> > +++ b/drivers/virtio/virtio.c
> > @@ -232,6 +232,12 @@ static int virtio_dev_probe(struct device *_d)
> > if (device_features & (1ULL << i))
> > __virtio_set_bit(dev, i);
> >
> > + if (drv->validate) {
> > + err = drv->validate(dev);
> > + if (err)
> > + goto err;
> > + }
> > +
> > err = virtio_finalize_features(dev);
> > if (err)
> > goto err;
> > diff --git a/include/linux/virtio.h b/include/linux/virtio.h
> > index 193fea9..ed04753 100644
> > --- a/include/linux/virtio.h
> > +++ b/include/linux/virtio.h
> > @@ -176,6 +176,7 @@ struct virtio_driver {
> > unsigned int feature_table_size;
> > const unsigned int *feature_table_legacy;
> > unsigned int feature_table_size_legacy;
> > + int (*validate)(struct virtio_device *dev);
> > int (*probe)(struct virtio_device *dev);
> > void (*scan)(struct virtio_device *dev);
> > void (*remove)(struct virtio_device *dev);
>
> Would be good to add some doc; but other members are undocumented here
> already...
True. Patches welcome.
