Hi Rob,
On 13/07/20 21:01, Rob Herring wrote:
> On Mon, Jun 22, 2020 at 10:43:27PM +0200, Luca Ceresoli wrote:
>> The definition of "xxx-in-supply" was generic, thus define in detail the
>> possible cases for each chip variant.
>>
>> Also document that the only possible I2C slave address is 0x60 as per the
>> datasheet and fix the second example accordingly.
>>
>> Signed-off-by: Luca Ceresoli <luca@xxxxxxxxxxxxxxxx>
[...]
>> +allOf:
>> + - if:
>> + properties:
>> + compatible:
>> + contains:
>> + enum:
>> + - ti,lp87565
>> + - ti,lp87565-q1
>> + then:
>> + properties:
>> + buck10-in-supply:
>> + description:
>> + Phandle to parent supply node for BUCK0 and BUCK1 converters.
>> +
>> + buck23-in-supply:
>> + description:
>> + Phandle to parent supply node for BUCK2 and BUCK3 converters.
>> +
>> + required:
>> + - buck10-in-supply
>> + - buck23-in-supply
>> +
>> + - if:
>> + properties:
>> + compatible:
>> + contains:
>> + enum:
>> + - ti,lp87561-q1
>> + then:
>> + properties:
>> + buck3210-in-supply:
>> + description:
>> + Phandle to parent supply node for all the four BUCK converters.
>> +
>> + required:
>> + - buck3210-in-supply
>
> I think this should probably be 2 schema documents (and a 3rd for the
> next patch)...
No problem, will do.
Is there a specific rule to know when to split, or is it just when the
amount of changes is relevant?
>> +
>> +examples:
>> + - |
>> + /* TI LP87565-Q1 PMIC (dual 2-phase output configuration) */
>> + i2c@0 {
>> + reg = <0x0 0x100>;
>> + #address-cells = <1>;
>> + #size-cells = <0>;
>> +
>> + pmic@60 {
>> + compatible = "ti,lp87565-q1";
>> + reg = <0x60>;
>> + gpio-controller;
>> + #gpio-cells = <2>;
>> +
>> + buck10-in-supply = <&vsys_3v3>;
>> + buck23-in-supply = <&vsys_3v3>;
>> +
>> + regulators {
>> + buck10_reg: buck10 {
>
> The regulators also need to be documented.
OK.
--
Luca
