Re: [PATCH 6/6] dt-bindings: display: samsung,exynos7-decon: add exynos7870 compatible

[Krzysztof Kozlowski <krzk@kernel.org>]

On 25/09/2024 22:05, Kaustabh Chakraborty wrote:
> On 2024-09-25 19:56, Krzysztof Kozlowski wrote:
>> On 25/09/2024 21:36, Kaustabh Chakraborty wrote:
>>> On 2024-09-25 19:25, Krzysztof Kozlowski wrote:
>>>> On 25/09/2024 20:42, Kaustabh Chakraborty wrote:
>>>>> On 2024-09-20 12:39, Krzysztof Kozlowski wrote:
>>>>>> On 19/09/2024 17:20, Kaustabh Chakraborty wrote:
>>>>>>> Add the compatible string of Exynos7870 to the existing list.
>>>>>>>
>>>>>>> Signed-off-by: Kaustabh Chakraborty <kauschluss@disroot.org>
>>>>>>
>>>>>> ... and the DTS is <please provide lore ink in changelog>?
>>>>>
>>>>> Didn't quite understand. The patch adds the compatible string
>>>>> for Exynos7870 DECON in documentation. There's no DTS involved
>>>>> in here, right?
>>>>
>>>> Provide lore link to the DTS submission.
>>>
>>> There aren't any DTS submissions *yet* which use the compatible.
>>> Is that an issue?
>>>
>>
>> Yeah, users are supposed to be upstream. Not downstream.
> 
> I understand that. I had plans to submit it in the future.
> If that's how it's meant to be done, I'll have to revisit this
> submission at a later date then.
> 

Partial, asynchronous bringup of a device is fine, so if the basic
support is there, I understand that drivers come in different pace.
Although I don't understand why DTS for this piece of hardware would
come in different pace, considering you cannot test it without DTS. You
have there DTS, so it should be sent.

But even without the DTS for DECON, the problem is earlier - lack of
basic support for this device. There is nothing for this chip.

This means it cannot be tested and is trickier to verify. That's not the
usual upstreaming way we expect, especially that you did not provide
rationale for such way.

Best regards,
Krzysztof