Re: [EXT] Re: [PATCH v3 1/4] dt-bindings: soc: imx8m: add DT Binding doc for soc unique ID

[Krzysztof Kozlowski]

On Mon, Nov 16, 2020 at 07:04:13AM +0000, Alice Guo wrote:
> 
> > -----Original Message-----
> > From: Krzysztof Kozlowski <krzk@xxxxxxxxxx>
> > Sent: 2020年11月15日 0:50
> > To: Alice Guo <alice.guo@xxxxxxx>
> > Cc: robh+dt@xxxxxxxxxx; shawnguo@xxxxxxxxxx; s.hauer@xxxxxxxxxxxxxx;
> > dl-linux-imx <linux-imx@xxxxxxx>; Peng Fan <peng.fan@xxxxxxx>;
> > devicetree@xxxxxxxxxxxxxxx; linux-kernel@xxxxxxxxxxxxxxx;
> > linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> > Subject: [EXT] Re: [PATCH v3 1/4] dt-bindings: soc: imx8m: add DT Binding doc
> > for soc unique ID
> > 
> > Caution: EXT Email
> > 
> > On Fri, Nov 13, 2020 at 07:04:06PM +0800, Alice Guo wrote:
> > > Add DT Binding doc for the Unique ID of i.MX 8M series.
> > >
> > > Signed-off-by: Alice Guo <alice.guo@xxxxxxx>
> > > ---
> > >  .../devicetree/bindings/arm/fsl.yaml          | 25
> > +++++++++++++++++++
> > >  1 file changed, 25 insertions(+)
> > >
> > > diff --git a/Documentation/devicetree/bindings/arm/fsl.yaml
> > > b/Documentation/devicetree/bindings/arm/fsl.yaml
> > > index e4db0f9ed664..f4faebbb57da 100644
> > > --- a/Documentation/devicetree/bindings/arm/fsl.yaml
> > > +++ b/Documentation/devicetree/bindings/arm/fsl.yaml
> > > @@ -901,6 +901,31 @@ properties:
> > >                - fsl,s32v234-evb           # S32V234-EVB2 Customer
> > Evaluation Board
> > >            - const: fsl,s32v234
> > >
> > > +  soc:
> > > +    type: object
> > > +    properties:
> > > +      compatible:
> > > +        oneOf:
> > > +          - description: i.MX8M SoCs
> > > +            items:
> > > +              - enum:
> > > +                  - fsl,imx8mm-soc
> > > +                  - fsl,imx8mn-soc
> > > +                  - fsl,imx8mp-soc
> > > +                  - fsl,imx8mq-soc
> > > +              - const: simple-bus
> > > +
> > > +          - description: Other SoCs
> > > +            items:
> > > +              - const: simple-bus
> > > +
> > > +      nvmem-cells:
> > > +        maxItems: 1
> > > +        description: Phandle to the SOC Unique ID provided by a nvmem
> > > + node
> > > +
> > > +      nvmem-cells-names:
> > > +        const: soc_unique_id
> > > +
> > >  additionalProperties: true
> > 
> > The "soc" node should be required for these compatibles. Otherwise you will
> > have to stay with this backwards-compatible-DTB-device-initcall-glue for many
> > years... I think you can achieve it with allOf.
> > 
> [Alice Guo]Hi,
> I find that my description for compatible is ambiguous. There are two kinds of compatible of the "soc" node:
> 1. For dtsi files used for SoCs other than imx8m series SoCs and old version dtsi files used for imx8m series SoCs,
> compatible of the "soc" node is "simple-bus".
> 2. For new version dtsi files used for imx8m series SoCs, compatible of the "soc" node is {"fsl,imx8mX-soc","simple-bus"}.
> 
> "nvmem-cell" is an optional property.
> 
> I do not understand what you mean. You mean that limit the compatible of "soc" node must include " fsl,imx8mX-soc" in new version dts files. Is my understanding correct?

All new DTS files should have soc node with "fsl,imx8mX-soc" and this
should be required by dtschema. The nvmem-cells and related properties
do not look like optional. From the hardware perspective - they are
always present. From the driver point of view: they are required and
driver will fail to work.

Best regards,
Krzysztof