Re: [PATCH v3 01/14] dt-bindings: display: renesas,cmm: Add R-Car CMM documentation

[Laurent Pinchart <laurent.pinchart@ideasonboard.com>]

Hi Jacopo,

On Fri, Aug 30, 2019 at 08:01:09PM +0200, Jacopo Mondi wrote:
> On Mon, Aug 26, 2019 at 01:15:50PM +0300, Laurent Pinchart wrote:
> > On Mon, Aug 26, 2019 at 09:59:43AM +0200, Jacopo Mondi wrote:
> >> On Mon, Aug 26, 2019 at 09:34:41AM +0200, Geert Uytterhoeven wrote:
> >>> On Sun, Aug 25, 2019 at 3:50 PM Jacopo Mondi <jacopo+renesas@jmondi.org> wrote:
> >>>> Add device tree bindings documentation for the Renesas R-Car Display
> >>>> Unit Color Management Module.
> >>>>
> >>>> CMM is the image enhancement module available on each R-Car DU video
> >>>> channel on R-Car Gen2 and Gen3 SoCs (V3H and V3M excluded).
> >>>>
> >>>> Signed-off-by: Jacopo Mondi <jacopo+renesas@jmondi.org>
> >>>
> >>> Thanks for your patch!
> >>>
> >>>> --- /dev/null
> >>>> +++ b/Documentation/devicetree/bindings/display/renesas,cmm.txt
> >>>> @@ -0,0 +1,33 @@
> >>>> +* Renesas R-Car Color Management Module (CMM)
> >>>> +
> >>>> +Renesas R-Car image enhancement module connected to R-Car DU video channels.
> >>>> +
> >>>> +Required properties:
> >>>> + - compatible: shall be one or more of the following:
> >>>> +   - "renesas,cmm-r8a7795": for R8A7795 (R-Car H3) compatible CMM.
> >>>> +   - "renesas,cmm-r8a7796": for R8A7796 (R-Car M3-W) compatible CMM.
> >>>> +   - "renesas,cmm-r8a77965": for R8A77965 (R-Car M3-N) compatible CMM.
> >>>> +   - "renesas,cmm-r8a77990": for R8A77990 (R-Car E3) compatible CMM.
> >>>> +   - "renesas,cmm-r8a77995": for R8A77995 (R-Car D3) compatible CMM.
> >>>
> >>> Please use "renesas,<socype->-cmm" instead of "renesas,cmm-<soctype>".
> >>
> >> I actually copied it from the r-car gpio bindings, and I liked
> >> cmm-<soctype> better. If you prefer I can change it though.
> >>
> >>>> +   - "renesas,rcar-gen3-cmm": for a generic R-Car Gen3 compatible CMM.
> >>>> +   - "renesas,rcar-gen2-cmm": for a generic R-Car Gen2 compatible CMM.
> >>>> +
> >>>> +   When the generic compatible string is specified, the SoC-specific
> >>>> +   version corresponding to the platform should be listed first.
> >>>> +
> >>>> + - reg: the address base and length of the memory area where CMM control
> >>>> +   registers are mapped to.
> >>>> +
> >>>> + - clocks: phandle and clock-specifier pair to the CMM functional clock
> >>>> +   supplier.
> >>>
> >>> Thinking about yaml validation:
> >>>
> >>> power-domains?
> >>> resets?
> >>
> >> They should indeed be documented.
> >
> > How about converting this binding to yaml alreay ? It should be fairly
> > simple.
> 
> I'm trying to, and I'm having my portion of fun time at it.
> 
> The definition of the schema itself seems good, but I wonder, is this
> the first renesas schema we have? Because it seems to me the schema
> validator is having an hard time to digest the examplea 'clocks' and
> 'power-domains' properties, which have 1 phandle and 2 specifiers and 1
> phandle and 1 specifier respectively for Rensas SoCs.
> 
> In other words, if in the example I have:
> 
>  examples:
>    - |
>      cmm0: cmm@fea40000 {
>           compatible = "renesas,r8a7796-cmm";
>           reg = <0 0xfea40000 0 0x1000>;
>           clocks = <&cpg 711>              <---- 1 phandle + 1 specifier
>           resets = <&cpg 711>;
>           power-domains = <&sysc>;         <---- 1 phandle
>      };
> 
> The schema validation is good.
> 
> While if I use an actual example
>    - |
>      cmm0: cmm@fea40000 {
>           compatible = "renesas,r8a7796-cmm";
>           reg = <0 0xfea40000 0 0x1000>;
>           clocks = <&cpg CPG_MOD 711>         <---- 1 phandle + 2 specifier
>           resets = <&cpg 711>;
>           power-domains = <&sysc R8A7796_PD_ALWAYS_ON>; <---- 1 phandle
>      };                                                       + 1 specfier
> 
> The schema validation fails...
> Error: Documentation/devicetree/bindings/display/renesas,cmm.example.dts:20.29-30 syntax error
> FATAL ERROR: Unable to parse input tree
> 
> Are clocks properties with > 2 entries and power-domains properties with
> > 1 entries supported?
> 
> Because from what I read here:
> https://github.com/robherring/yaml-bindings/blob/master/schemas/clock/clock.yaml
> "The length of a clock specifier is defined by the value of a #clock-cells
> property in the clock provider node."
> 
> And that's expected, but is the examples actually validated against the
> clock provider pointed by the phandle? Because in that case, if we had a
> yaml schema for the cpg-mssr provider, it would indeed specify clock-cells=2.
> 
> Do we need a schema for cpg-mssr first, or am I doing something else
> wrong?

I think you just need to #include the headers that define CPG_MOD and
R8A7796_PD_ALWAYS_ON.

> >>>> +Example:
> >>>> +--------
> >>>> +
> >>>> +       cmm0: cmm@fea40000 {
> >>>> +               compatible = "renesas,cmm-r8a7796";
> >>>> +               reg = <0 0xfea40000 0 0x1000>;
> >>>> +               power-domains = <&sysc R8A7796_PD_ALWAYS_ON>;
> >>>> +               clocks = <&cpg CPG_MOD 711>;
> >>>> +               resets = <&cpg 711>;
> >>>> +       };

-- 
Regards,

Laurent Pinchart