Re: [PATCH 3/4] dt-bindings: clock: Add AST2700 clock bindings

[Krzysztof Kozlowski <krzk@kernel.org>]

On 19/08/2024 08:42, Ryan Chen wrote:
>> Subject: Re: [PATCH 3/4] dt-bindings: clock: Add AST2700 clock bindings
>>
>> On 19/08/2024 07:55, Ryan Chen wrote:
>>>> Subject: Re: [PATCH 3/4] dt-bindings: clock: Add AST2700 clock
>>>> bindings
>>>>
>>>> On 13/08/2024 03:53, Ryan Chen wrote:
>>>>>> Drop the define for number of clocks from the header, because it is
>>>>>> not a
>>>>
>>>> *NUMBER OF CLOCKS*
>>>>
>>>>>> binding. You can put it in the driver or not, I don't care and do
>>>>>> not provide guidance on this because I don't know if it makes sense at all.
>>>>>> What I know is that number of clocks is not related to binding. It
>>>>>> is not needed
>>>>
>>>> *NUMBER OF CLOCKS*
>>>>
>>>>>> in the binding, either.
>>>>>
>>>>> Sorry, I am confused.
>>>>> if you think that number of clocks is not related to binding.
>>>>
>>>> *NUMBER OF CLOCKS*
>>>>
>>>>> How dtsi claim for clk?
>>>>> For example in dtsi.
>>>>> include <dt-bindings/clock/aspeed,ast2700-clk.h>
>>>>> usb3bhp: usb3bhp {
>>>>> ....
>>>>> clocks = <&syscon0 SCU0_CLK_GATE_PORTAUSB>;
>>>>
>>>> And where is *NUMBER OF CLOCKS* here? I don't see any problem. No
>>>> useless SCU0_CLK_GATE_NUM define here.
>>>>
>>> Understood now, I will remove those *NUMBER OF CLOCKS*.
>>> And will replace to
>>> #define SCU0_CLK_END  34
>>
>> NAK, it's like you keep ignoring my comments entirely. Even if you call it
>> "SCU0_CLK_NOT_END" it does not change. Do you understand that it is not
>> about name? Read my first comment.
>>
>>>
>>> Refer:
>>> https://github.com/torvalds/linux/blob/master/include/dt-bindings/cloc
>>> k/imx8-clock.h#L87
>>
>> So you found a bug and this allows you to create the same bug?
>>
> Sorry, I don't see this is a bug.

No, it's not a bug, but I do not agree for using arguments like "someone
did it, so I can do the same". Why did you pick up exactly this example
instead of others who removed the clock number?

> But I try to understand your point, you prefer following for clock nums, am I correct?
> https://github.com/torvalds/linux/blob/master/drivers/clk/meson/g12a.c#L5558-L5559

I said that this is not a binding. Don't add to the binding things which
are not a binding.

I don't care how do you implement in drivers - there are several ways
how to achieve it.


Best regards,
Krzysztof