Re: [PATCH v3 1/3] dt-bindings: leds: Add LP5812 LED driver

[Krzysztof Kozlowski <krzk@kernel.org>]

On 31/03/2025 17:31, Nam Tran wrote:
>>>
>>>> +
>>>> +  "#address-cells":
>>>> +    const: 1
>>>> +
>>>> +  "#size-cells":
>>>> +    const: 0
>>>
>>> What are these?
>>>
>>> Nam: I included the #address-cells and #size-cells properties to resolve a warning encountered when running:
>>> make dt_binding_check DT_SCHEMA_FILES=Documentation/devicetree/bindings/leds/ti,lp5812.yaml
>>> The specific warning was:
>>> Documentation/devicetree/bindings/leds/ti,lp5812.example.dts:23.17-30: Warning (reg_format): /example-0/i2c/led-controller@1b:reg: property has invalid length (4 bytes) (#address-cells == 2, #size-cells == 1)
>>
>> This makes no sense.
>>
>>> This warning suggests that the default values for #address-cells and #size-cells in the schema context are not aligned with the LP5812's expected usage. To explicitly define the correct values, I set these properties as mentioned.
>>> This ensures that the binding schema validation passes without warnings. If you believe a different approach is more appropriate, I’m happy to adjust the binding accordingly.
>>
>> I can barely parse your messages. They are neither properly quoting my
>> replies, nor wrapped according to email style. Use standard format,
>> expressed in countless guides bout netiquette and mailing lists.
>>
>> You added properties to hide warning, instead of fixing the warning, but
>> these properties make nos sense here.
> 
> #address-cells is 1 because I only one cell to define address of a child node of LP5812
> #size-cells is 0 because I don't need to define size of reg properly of child node.
> Reg properly of child node is describe for address only.


Please do not explain me how DT works, we all know. You do not have
child node. If you disagree - point me to the line in schema having it.

Best regards,
Krzysztof