On Mon, Sep 08, 2025 at 03:43:22PM +0000, Aleksandrs Vinarskis wrote:
>
>
>
>
>
> On Monday, September 8th, 2025 at 16:15, Rob Herring (Arm) <robh@xxxxxxxxxx> wrote:
>
> >
> >
> >
> > On Mon, 08 Sep 2025 01:18:03 +0200, Aleksandrs Vinarskis wrote:
> >
> > > Introduce common generic led consumer binding, where consumer defines
> > > led(s) by phandle, as opposed to trigger-source binding where the
> > > trigger source is defined in led itself.
> > >
> > > Add already used in some schemas 'leds' parameter which expects
> > > phandle-array. Additionally, introduce 'led-names' which could be used
> > > by consumers to map LED devices to their respective functions.
> > >
> > > Signed-off-by: Aleksandrs Vinarskis alex@xxxxxxxxxxxxx
> > > ---
> > > .../devicetree/bindings/leds/leds-consumer.yaml | 89 ++++++++++++++++++++++
> > > 1 file changed, 89 insertions(+)
> >
> >
> > My bot found errors running 'make dt_binding_check' on your patch:
> >
> > yamllint warnings/errors:
> >
> > dtschema/dtc warnings/errors:
> > /builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/leds/leds-consumer.example.dtb: camera@36 (ovti,ov02c10): Unevaluated properties are not allowed ('led-names', 'leds' were unexpected)
> > from schema $id: http://devicetree.org/schemas/media/i2c/ovti,ov02e10.yaml#
>
> Rob: this is because the 1st patch that adds `led-consumer.yaml` uses
> ov02e10 in its example, while the property is added to
> `video-interface-devices.yaml` (which is used by ov02e10 and other
> cameras) only in the 2nd patch. As I see it, reversing the order of 1st
> and 2nd patch would likewise cause 'error' when 1st patch is checked
> without the 2nd one. I don't think that 1st and 2nd patches should be
> combined, but if you prefer to do that so the warning goes away - please
> let me know.
>
Just drop the compatible (and most of the properties) from the example.
That's a common binding, so adding there some other device DTS only
complicates things. See access-controllers, for example.
Best regards,
Krzysztof
