On Thu, Oct 31, 2019 at 03:52:24PM +0100, Torsten Duwe wrote:
> On Thu, Oct 31, 2019 at 01:51:00PM +0100, Maxime Ripard wrote:
> > On Tue, Oct 29, 2019 at 01:16:57PM +0100, Torsten Duwe wrote:
> > > +
> > > +  ports:
> > > +    anyOf:
> > > +      - port@0:
> > > +        description: Video port for LVTTL input
> > > +      - port@1:
> > > +        description: Video port for eDP output (panel or connector).
> > > +                     May be omitted if EDID works reliably.
> > > +    required:
> > > +      - port@0
> >
> > Have you tried to validate those two ports in a DT?
>
> Yes, it validates as expected, like I wrote. Various sources told me that
> json-schema is not always straightforward so I assumed anyOf was OK.
>
> > I'm not quite sure what you wanted to express with that anyOf, but if
> > it was something like port@0 is mandatory, and port@1 is optional, it
> > should be something like this:
> >
> > properties:
> >
> >   ...
> >
> >   ports:
> >     type: object
> >
> >     properties:
> >       port@0:
> >         type: object
> >         description: |
> > 	  Video port for LVTTL input
> >
> >       port@1:
> >         type: object
> >         description: |
> > 	  Video port for eDP output (..)
> >
> >     required:
> >       - port@0
> >
> > This way, you express that both port@0 and port@1 must by nodes, under
> > a node called ports, and port@0 is mandatory.
>
> That validates, too. Looks better, admittedly. I don't have a strong
> opinion here. It's just that Rob wrote in
> <CAL_JsqKAU3WG3L=KP8A8u4vW=q_BQWPN-m_c+ADOwTioJ2-cmg@mail.gmail.com>:
>
> | For this case specifically, we do need to define a common graph
> | schema, but haven't yet. You can assume we do and only really need to
> | capture what Maxime said above.
> (your points back then were port@N descriptions and neccessity for port@0)
>
> Are you sure that "object" is specific enough?

Possibly not, but at least it checks that there's indeed something
called port@0 (and port@1), and that they are both nodes (and not
properties).

We can probably refine this further, but this is good enough at the
moment.

> > You should even push this a bit further by adding
> > additionalProperties: false to prevent a DT from having undocumented
> > properties and children for the main node and ports node.
>
> You mean like
>
> | jsonschema.exceptions.SchemaError: Additional properties are not allowed ('unevaluatedProperties' was unexpected)
> [...]
> | On schema:
> |    {'$id': 'http://devicetree.org/schemas/watchdog/allwinner,sun4i-a10-wdt.yaml#',
> [...]
> |      'unevaluatedProperties': False}
>
> ? ;-)

That would be on the meta-schema, but yes, we want to trigger warnings
on something that isn't described.

>
> But yes, this patch series passes even with additionalProperties: false.
>
> In which form would you like to receive the update?

Please send a new version.

Thanks!
Maxime