Re: [PATCH v3 1/2] dt-bindings: pinctrl: qcom: update compatible name for match with driver

[Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx>]

On 13/03/2024 08:55, Tengfei Fan wrote:
>>>>>> Wasn't this applied?
>>>>>
>>>>> My test code base on tag: next-20240308, this patch is still not applied.
>>>>>
>>>>> In fact, the following dt binding check warning only can be got before
>>>>> this patch is applied.
>>>>>
>>>>
>>>> Please read all emails in the previous thread. You ignored two emails in
>>>> the past and apparently one more recent.
>>>
>>> I don't know if you mean I ignored the email which related with "Patch
>>> applied" tag from Linus Walleij. If so, the following is the reasion why
>>> I still include this patch:
>>
>> Yep, that's the one. Please do not send patches which were already
>> applied. It causes unnecessary effort on reviewer and maintainer side.
>>
>>>
>>> I synced the latest upstream code on 03/12/2024, the latest tag is
>>> next-20240308, this tag still doesn't include this patch[PATCH v3 1/2].
>>
>> Happens, considering Linus applied it after 8th of March, I think.
>>
>>>
>>> Dt binding check still get warning if I only send [PATCH v3 2/2] patch
>>> to upstream base on next-20240308. so I include this patch[PATCH v3 1/2]
>>
>> If you send patch 1+2, dt_binding_check will have exactly the same
>> result. I don't know about what sort of dt binding check you talk, but
>> for all cases: you changed nothing by sending these two patches in that
>> regard. Only noise on the lists.
> 
> The dt binding check failed which Rob Herring remind me in previous 
> patch series as the following:

This does not make any sense. Whether Rob runs his test on previous or
future next, changes nothing in regard of this patchset being sent with
duplicated patch or not. The result will be exactly the same for Rob.

> 
> Documentation/devicetree/bindings/pinctrl/qcom,sm4450-tlmm.example.dtb:
> /example-0/pinctrl@f100000: failed to match any schema with
> compatible: ['qcom,sm4450-tlmm']
> 
> This failed is introduced by 
> https://lore.kernel.org/linux-arm-msm/20231206020840.33228-2-quic_tengfan@xxxxxxxxxxx/. 
> Something got broken aroud -m flags for dtschema, so indeed no reports 
> this unmatched compatibles warning when this patch was revriwed. We also 
> have some discusstion in patch email.

Again, not related at all whether you send patch *which was applied* or not.

> 
> The patch[PATCH v3 1/2] is made for fix this previous patch dt binding 
> check failed. So dt binding check failed will disappear after this 
> patch[PATCH v3 1/2] is applied.

And who is supposed to run that dt binding check and on what base? Your
patch changes absolutely nothing in that regard, just creates confusion.

And the fact that you keep arguing over this simple case, reminds me
other clueless discussions I had with some Qualcomm folks. None of the
arguments you brought here justify sending patch which was applied.

> 
>>
>>> in patch series even if this patch have "Patch applied" tag.
>>>
>>> Looking forward to getting your advice if submitting patch series this
>>> way is problematic.
>>
>> Do not send patches which are known to be applied.
> 
> Yes, I will be careful not to resend the patch which have already been 
> applied in the future work.

Then why do you keep arguing that sending this duplicated patch was
correct approach?

> 
> Do you think it is necessary to send another version patch series for 
> remove this applied patch[PATCH v3 1/2] from patch series?

No. It is merge window, please read process documents in Documentation
directory. Then please read Qualcomm upstreaming guide.

Best regards,
Krzysztof