Re: [PATCH v5 2/4] pinctrl: qcom: Add SM4450 pinctrl driver

[Tengfei Fan <quic_tengfan@quicinc.com>]

在 12/4/2023 4:14 PM, Krzysztof Kozlowski 写道:
> On 04/12/2023 09:06, Tengfei Fan wrote:
>>
>>
>> 在 12/4/2023 3:56 PM, Krzysztof Kozlowski 写道:
>>> On 04/12/2023 02:57, Tengfei Fan wrote:
>>>>
>>>>
>>>> 在 11/30/2023 7:57 PM, Krzysztof Kozlowski 写道:
>>>>> On 30/11/2023 03:40, Tengfei Fan wrote:
>>>>>> Add pinctrl driver for TLMM block found in SM4450 SoC.
>>>>>>
>>>>>> Reviewed-by: Bjorn Andersson <andersson@kernel.org>
>>>>>> Signed-off-by: Tengfei Fan <quic_tengfan@quicinc.com>
>>>>>> ---
>>>>>>     drivers/pinctrl/qcom/Kconfig.msm      |    8 +
>>>>>>     drivers/pinctrl/qcom/Makefile         |    1 +
>>>>>>     drivers/pinctrl/qcom/pinctrl-sm4450.c | 1013 +++++++++++++++++++++++++
>>>>>>     3 files changed, 1022 insertions(+)
>>>>>>     create mode 100644 drivers/pinctrl/qcom/pinctrl-sm4450.c
>>>>>>
>>>>>
>>>>> Hm, was this patch ever built?
>>>>>
>>>>> Best regards,
>>>>> Krzysztof
>>>>>
>>>> This patch has been built before, I will check and compare if there are
>>>> any errors and changes when I submitted this patch series.
>>>>
>>>
>>> No, it wasn't built. I just tried - applied it and:
>>>
>>> pinctrl-sm4450.c:996:19: error: initialization of ‘int (*)(struct
>>> platform_device *)’ from incompatible pointer type ‘void (*)(struct
>>> platform_device *)’ [-Werror=incompatible-pointer-types]
>>>     996 |         .remove = msm_pinctrl_remove,
>>>         |                   ^~~~~~~~~~~~~~~~~~
>>> ../drivers/pinctrl/qcom/pinctrl-sm4450.c:996:19: note: (near
>>> initialization for ‘sm4450_tlmm_driver.remove’)
>>>
>>> So you just sent a patch which was not even compiled.
>>>
>>> NAK.
>>>
>>> Best regards,
>>> Krzysztof
>>>
>> I compiled all the related patches together, but I did not compile this
>> patch separately.
> 
> We talk about this patch here. Please do not send knowingly wrong code,
> because it does not make sense and hurts bisectability.
Sure, I will avoid similar problems in the future.
> 
>> The fact that there is a compilation problem is known, but because the
>> patch is already reviewed-by, so a separate patch(patch 3) is submitted
>> to fix the compilation error.
> 
> That's not the process. Each patch must be correct. Each.
Yes, the correctness of each patch will be ensured in the future.
> 
> Best regards,
> Krzysztof
> 

-- 
Thx and BRs,
Tengfei Fan