lkml.org 
[lkml]   [2023]   [Dec]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
SubjectRe: [PATCH v3 5/5] dt-bindings: net: ipq4019-mdio: Document ipq5332 platform
From


On 12/15/2023 6:21 PM, Krzysztof Kozlowski wrote:
> On 15/12/2023 11:03, Jie Luo wrote:
>>>>>> + cmn-reference-clock:
>>>>>> + $ref: /schemas/types.yaml#/definitions/uint32
>>>>>
>>>>> Nothing improved here
>>>>
>>>> With this change, the warning is not reported when i run
>>>> dt_binding_check, looks the new added property needs
>>>> the type ref to avoid the warning reported.
>>>
>>> Nothing improved in the property name, nor its style, nor in the actual
>>> contents/values.
>>
>> This property is for CMN PLL block reference clock configuration,
>> so i use this property name.
>>
>> it will be appreciated if you can suggest a suitable name, thanks.
>
> See example-schema about naming. Read writing-bindings. You need vendor
> prefix for custom properties.

Ok, thanks.

>
>>
>>>
>>> ...
>>>
>>>>>> + reset-gpios:
>>>>>> + maxItems: 1
>>>>>> +
>>>>>> + reset-assert-us:
>>>>>> + maxItems: 1
>>>>>
>>>>> This does not look related to ipq5332.
>>>>
>>>> The reset gpio properties are needed on ipq5332, since qca8084 phy is
>>>> connected, which uses the MDIO bus level gpio reset.
>>>
>>> I am talking about this property, not these properties.
>>
>> ok.
>>
>>>
>>>>
>>>> Without declaring these gpio properties, the warning will be reported
>>>> by dt_binding_check.
>>>
>>> How is it even possible to have warnings if there is no such node in
>>> DTS? We do not care about warnings in your downstream code.
>>>
>>>
>>> Best regards,
>>> Krzysztof
>>>
>>
>> If i do not declare the property "reset-assert-us" and
>> "reset-deassert-us", the warning will be reported by "make
>> dt_binding_check" since i
>> add a example in this file.
>
> This argument does not make sense, sorry. Obviously if property is not
> allowed, it should be removed.
>
> Provide rationale, in terms of hardware, why this property must be added
> and why it cannot be deduced from the compatible.
>
> Best regards,
> Krzysztof
>

So i can remove "reset-assert-us" and "reset-deassert-us" from the added
example to avoid the dt check warning? even these two properties are
needed to be defined in the device tree to make this driver working
correctly.

\
 
 \ /
  Last update: 2023-12-15 13:04    [W:0.441 / U:0.412 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site