Hi Miquel, On Fri, Nov 26, 2021 at 12:46 PM Miquel Raynal <miquel.raynal@xxxxxxxxxxx> wrote: > geert@xxxxxxxxxxxxxx wrote on Fri, 19 Nov 2021 10:36:16 +0100: > > On Fri, Nov 19, 2021 at 10:19 AM Miquel Raynal > > <miquel.raynal@xxxxxxxxxxx> wrote: > > > geert@xxxxxxxxxxxxxx wrote on Fri, 19 Nov 2021 09:41:35 +0100: > > > > On Thu, Nov 18, 2021 at 12:19 PM Miquel Raynal > > > > <miquel.raynal@xxxxxxxxxxx> wrote: > > > > > Add a Yaml description for this Renesas NAND controller bindings. > > > > > > > > > > Signed-off-by: Miquel Raynal <miquel.raynal@xxxxxxxxxxx> > > > > > > > --- /dev/null > > > > > +++ b/Documentation/devicetree/bindings/mtd/renesas,r9a06g032-nand-controller.yaml > > > > > @@ -0,0 +1,60 @@ > > > > > +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause) > > > > > +%YAML 1.2 > > > > > +--- > > > > > +$id: http://devicetree.org/schemas/mtd/renesas,r9a06g032-nand-controller.yaml# > > > > > +$schema: http://devicetree.org/meta-schemas/core.yaml# > > > > > + > > > > > +title: Renesas RZ/N1x NAND flash controller device tree bindings > > > > > + > > > > > +maintainers: > > > > > + - Miquel Raynal <miquel.raynal@xxxxxxxxxxx> > > > > > + > > > > > +allOf: > > > > > + - $ref: "nand-controller.yaml" > > > > > + > > > > > +properties: > > > > > + compatible: > > > > > + const: renesas,r9a06g032-nand-controller > > > > > > > > As the NAND Flash Controller is present on all of RZ/N1D, RZ/N1S, > > > > and RZ/N1L, I think you should add a family-specific compatible value > > > > "renesas,rzn1-nand-controller" as a fallback. > > > > > > I see, that's right, I should have added two compatibles. > > > > > > As there is currently only one 'specific' compatible (r9axxx), should I > > > describe the two compatibles as being mandatory? Or should I set the > > > most specific one as optional and the least specific one (rzn1) > > > mandatory? > > > > Yes please. > > I am a little bit confused to which answered you said yes. My apologies: yes to making both mandatory. > > Else you need to match on both in the driver, or we cannot > > differentiate later if the need ever arises. > > I believe you meant "yes the two should be described as mandatory in the > bindings" (at least for now) so that when the need arises, the most > specific one can be replaced with a oneOf choice. Am I right? Exactly. You can already use "enum" for the most-specific one, so it's clear where to add new lines, and to minimize future changes. Gr{oetje,eeting}s, Geert -- Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@xxxxxxxxxxxxxx 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
