On Fri, Jun 28, 2019 at 7:46 AM Maxime Ripard <maxime.ripard@xxxxxxxxxxx> wrote:
>
> On Thu, Jun 27, 2019 at 10:06:57AM -0600, Rob Herring wrote:
> > On Thu, Jun 27, 2019 at 9:57 AM Maxime Ripard <maxime.ripard@xxxxxxxxxxx> wrote:
> > > > > +
> > > > > + reset-gpios = <&gpio2 5 1>;
> > > > > + reset-delay-us = <2>;
> > > > > +
> > > > > + ethphy0: ethernet-phy@1 {
> > > > > + reg = <1>;
> > > >
> > > > Need a child node schema to validate the unit-address and reg property.
> > >
> > > This should be already covered by the ethernet-phy.yaml schemas
> > > earlier in this series.
> >
> > Partially, yes.
> >
> > > Were you expecting something else?
> >
> > That would not prevent having a child node such as 'foo {};' or
> > 'foo@bad {};'. It would also not check valid nodes named something
> > other than 'ethernet-phy'.
>
> Right, but listing the nodes won't either, since we can't enable
> additionalProperties in that schema. So any node that wouldn't match
> ethernet-phy@.* wouldn't be validated, but wouldn't generate a warning
> either.
Perhaps I wasn't clear, but it was missing or incorrect 'reg' property
and unit-address format checks that I was thinking about. Just like we
have for SPI.
Rob
