Re: [PATCH v3 2/4] dt-bindings: serial: renesas,scif: Validate 'interrupts' and 'interrupt-names'

[Geert Uytterhoeven]

Hi Krzysztof,

On Wed, Mar 20, 2024 at 9:06 AM Krzysztof Kozlowski
<krzysztof.kozlowski@xxxxxxxxxx> wrote:
> On 19/03/2024 14:25, Geert Uytterhoeven wrote:
> > On Tue, Mar 19, 2024 at 2:04 PM Krzysztof Kozlowski
> > <krzysztof.kozlowski@xxxxxxxxxx> wrote:
> >> On 19/03/2024 13:43, Lad, Prabhakar wrote:
> >>>>>> diff --git a/Documentation/devicetree/bindings/serial/renesas,scifyaml b/Documentation/devicetree/bindings/serial/renesas,scif.yaml
> >>>>>> index af72c3420453..53f18e9810fd 100644
> >>>>>> --- a/Documentation/devicetree/bindings/serial/renesas,scif.yaml
> >>>>>> +++ b/Documentation/devicetree/bindings/serial/renesas,scif.yaml
> >>>>>> @@ -82,38 +82,6 @@ properties:
> >>>>>>    reg:
> >>>>>>      maxItems: 1
> >>>>>>
> >>>>>> -  interrupts:
> >>>>>
> >>>>> I don't understand what is happening with this patchset. Interrupts must
> >>>>> stay here. Where did you receive any different feedback?
> >>>>
> >>>> Look how it is done:
> >>>> https://elixir.bootlin.com/linux/v6.8/source/Documentation/devicetree/bindings/ufs/qcom,ufs.yaml#L44
> >>>>
> >>> Thanks for the pointer, as the above binding doesn't have any
> >>
> >> Yeah, that's just an example to point you the concept: top level
> >> property comes with widest constraints (or widest matching items
> >> description) and each variant narrows the choice.
> >>
> >>> description items as compared to our case, to clarify I have updated
> >>> the binding is below. Is this the correct approach?
> >>>
> >>> option #1
> >>> ---------------
> >>
> >>
> >> Yes, it looks correct.
> >
> > Why duplicate all the descriptions? The only differences are the number
> > of valid interrupts?
> > What was wrong with "[PATCH v2 2/2] dt-bindings: serial: renesas,scif:
> > Validate 'interrupts' and 'interrupt-names'"[1]?
> >
> > https://lore.kernel.org/r/20240307114217.34784-3-prabhakar.mahadev-lad.rj@xxxxxxxxxxxxxx/
>
> I have impression that only two variants out of three have same
> descriptions... but now I see mistake I made in above. I read that first
> interrupt is "Error interrupt" but it is "error or combined". Sorry for
> that, I think most of my comment there is not correct.
>
> It could be made oneOf?
>
>     oneOf:
>      - items:
>           - description: A combined interrupt
>      - items:
>          - ....
>        minItems: 4
> ?

Yes, that would be even better.

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68korg

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds