Re: [PATCH v1] dts: arm64: freescale: move imx9*-clock.h imx9*-power.h into dt-bindings

[E Shattow <e@freeshell.de>]

On 9/10/25 00:07, Peng Fan wrote:
> Hi Marek,
> 
>> Subject: Re: [PATCH v1] dts: arm64: freescale: move imx9*-clock.h
>> imx9*-power.h into dt-bindings
>>
>> On 9/4/25 11:34 AM, Peng Fan wrote:
>>
>> Hi,
>>
>> sorry for my late reply.
>>
>>>> Instead of playing this "I found this code somewhere, so I can do
>>>>> whatever the same" answer the first implied question - why these
>> are
>>>>> bindings? Provide arguments what do they bind.
>>>>
>>>> I am not sure how to answer this, but what I can write is, that if I
>>>> scramble these IDs in either the DT or the firmware (which provides
>>>> the SCMI clock service), then the system cannot work. I am not sure
>>>> if this is the answer you are looking for.
>>>
>>> Marek,
>>>    Some U-Boot code indeed directly use the IDs to configure the clock
>> without
>>>    relying on any drivers. Since the SCMI IDs could not be moved to
>>> dt-bindings,
>>
>> Why can they not be moved to DT bindings ?
> 
> DT maintainers suggested to keep the files under
> arch/arm64/boot/dts/freescale/
> 
> https://lore.kernel.org/all/75ab9162-ed02-479d-92a1-7cfabff6b32e@linaro.org/
> 
> Regards
> Peng

Follow up FYI to all this is resolved from U-Boot [1] as not using the
header and duplicating the information.

1:
https://lore.kernel.org/u-boot/20250923-imx94-v1-5-cb3b65169048@oss.nxp.com/

There was yet some IRC discussion about what justification can be for
avoiding this maintenance and duplication.

I am dropping this suggestion to move imx9*-clock.h imx9*-power.h into
dt-bindings. There is some better way of doing this, maybe a future
dt-firmware includes? That is not at all my area of interest or
expertise, however. I just noticed something out-of-place and bring it
to our attention.

My great appreciation is due for the reviews and discussion about this.
Thanks, everyone!

-E