Re: [PATCH v8 2/3] drm/panel: Add support for S6E3HA2 panel driver on TM2 board

[Inki Dae]

2017-01-31 20:30 GMT+09:00 Krzysztof Kozlowski <krzk@xxxxxxxxxx>:
> On Tue, Jan 31, 2017 at 12:37 PM, Inki Dae <daeinki@xxxxxxxxx> wrote:
>> 2017-01-31 19:01 GMT+09:00 Krzysztof Kozlowski <krzk@xxxxxxxxxx>:
>>> On Tue, Jan 31, 2017 at 11:34 AM, Inki Dae <inki.dae@xxxxxxxxxxx> wrote:
>>>>
>>>>
>>>> 2017ë 01ì 31ì 18:22ì Krzysztof Kozlowski ì(ê) ì ê:
>>>>> On Tue, Jan 31, 2017 at 2:01 AM, Inki Dae <inki.dae@xxxxxxxxxxx> wrote:
>>>>>>
>>>>>>
>>>>>> 2017ë 01ì 24ì 10:50ì Hoegeun Kwon ì(ê) ì ê:
>>>>>>> Dear Thierry,
>>>>>>>
>>>>>>> Could you please review this patch?
>>>>>>
>>>>>> Thierry, I think this patch has been reviewed enough but no comment from you. Seems you are busy. I will pick up this.
>>>>>>
>>>>>
>>>>> Comments from v8 were not resolved and I think we are waiting for v9:
>>>>> https://lkml.org/lkml/2017/1/11/178
>>>>>
>>>>> If that is not correct then please clarify.
>>>>
>>>> Seems you pointed to change te-gpios bindings to optional. right?
>>>>
>>>> I thought Rob left ack so it's no problem.
>>>> https://lkml.org/lkml/2017/1/13/626
>>>
>>> Yes, change them to optional. I think it is a problem (regardless of
>>> Rob's ack) because you are merging a driver requiring a property which
>>> soon you want to remove. If you merge it (like this) removal of
>>> te-gpios property will be breakage of ABI. This is not a serious
>>> problem but knowing such plan of te-gpios removal upfront, it would be
>>> wrong to commit such driver.
>>
>> This is a trivial thing so it doesn't make breakage of ABI because the
>> property, te-gpios, is *optional*.
>
> So why bindings document is not changed?

Seems Hoegeun missed this.
I also think it should specify *te-gpios is optional property* like
mipi-dsi binding document did but this doesn't breakage of ABI you
mentioned. And I'd like to say we should feel ABI breakage is a
serious problem.

Sad to say, Exynos is already broken like Marek declared. This is
really a serious problem Exynos maintainers have to fix including you
and me.

Thanks,
Inki Dae

>
> +Required properties:
> (...)
> +  - te-gpios: a GPIO spec for the tearing effect synchronization signal
> +    gpio pin (active high)
>
> Andrzej pointed this out and it is not fixed since then. What is the
> problem with fixing the bindings documentation?
>
>> Anyway, this wouldn't be *the things* Thierry mentioned.
>
> Probably not, I did not respond to Thierry's feedback.
>
> Best regards,
> Krzysztof