Re: [PATCH v3] dt-bindings: pinctrl: convert semtech,sx150xq bindings to dt-schema

[Neil Armstrong]

On 18/11/2022 16:19, Sander Vanheule wrote:
> Hi,
>
> On Tue, 2022-11-15 at 11:06 +0100, Neil Armstrong wrote:
> > This converts the Semtech SX150Xq bindings to dt-schemas, add necessary
> > bindings documentation to cover all differences between HW variants
> > and current bindings usage.
> >
> > Signed-off-by: Neil Armstrong <neil.armstrong@xxxxxxxxxx>
> > Reviewed-by: Rob Herring <robh@xxxxxxxxxx>
> > ---
> > To: Linus Walleij <linus.walleij@xxxxxxxxxx>
> > To: Rob Herring <robh+dt@xxxxxxxxxx>
> > To: Krzysztof Kozlowski <krzysztof.kozlowski+dt@xxxxxxxxxx>
> > Cc: linux-gpio@xxxxxxxxxxxxxxx
> > Cc: devicetree@xxxxxxxxxxxxxxx
> > Cc: linux-kernel@xxxxxxxxxxxxxxx
> > ---
> > Changes in v3:
> > - Resent with missing To: Linus Walleij
> > - Link to v2: https://lore.kernel.org/r/20221005-mdm9615-sx1509q-yaml-v2-0-a4a5b8eecc7b@xxxxxxxxxx
> >
> > Changes in v2:
> > - fixed rob comments
> > - added rob's Reviewed-by
> > - Link to v1: https://lore.kernel.org/r/20221005-mdm9615-sx1509q-yaml-v1-0-0c26649b637c@xxxxxxxxxx
> > ---
>
> [snip]
>
> > +  - if:
> > +      properties:
> > +        compatible:
> > +          contains:
> > +            enum:
> > +              - semtech,sx1503q
> > +              - semtech,sx1506q
> > +    then:
> > +      patternProperties:
> > +        '-cfg$':
> > +          properties:
> > +            pins:
> > +              items:
> > +                pattern: '^gpio[0-15]$'
>
> [snip]
>
> > +  - if:
> > +      properties:
> > +        compatible:
> > +          contains:
> > +            const: semtech,sx1509q
> > +    then:
> > +      patternProperties:
> > +        '-cfg$':
> > +          properties:
> > +            pins:
> > +              items:
> > +                pattern: '^(oscio|gpio[0-15])$'
>
> Sorry to be so late to reply, but don't these patterns only match "gpio0", "gpio1", and "gpio5"?
>
> A quick search for some datasheets turned up the SX1503 and SX1509Q with 16 GPIOs, so I assume the
> intention was to match "gpio0" to "gpio15". I think this should be "^(gpio[0-9]|gpio1[0-5])$" (or
> something equivalent).

Damn you're right, Linus should I resend or send a fixup ?

Thanks,
Neil

> Best,
> Sander