Re: [PATCH v1 2/2] pinctrl: nuvoton: add NPCM8XX pinctrl and GPIO driver

[Krzysztof Kozlowski <krzysztof.kozlowski@xxxxxxxxxx>]

On 13/07/2022 16:51, Tomer Maimon wrote:
> On Wed, 13 Jul 2022 at 17:29, Krzysztof Kozlowski
> <krzysztof.kozlowski@xxxxxxxxxx> wrote:
>>
>> On 13/07/2022 15:35, Tomer Maimon wrote:
>>
>>>>> +static int npcm8xx_pinctrl_probe(struct platform_device *pdev)
>>>>> +{
>>>>> +     struct npcm8xx_pinctrl *pctrl;
>>>>> +     int ret;
>>>>> +
>>>>> +     pctrl = devm_kzalloc(&pdev->dev, sizeof(*pctrl), GFP_KERNEL);
>>>>> +     if (!pctrl)
>>>>> +             return -ENOMEM;
>>>>> +
>>>>> +     pctrl->dev = &pdev->dev;
>>>>> +     dev_set_drvdata(&pdev->dev, pctrl);
>>>>> +
>>>>> +     pctrl->gcr_regmap =
>>>>> +             syscon_regmap_lookup_by_compatible("nuvoton,npcm845-gcr");
>>>>
>>>> No. Use property. By this patchset, I would expect that you learnt from
>>>> previous mistakes around this. Why repeating the same trouble second time?
>>> You suggest to use phandle property like nuvoton,sysgcr even that the
>>> NPCM8XX pin controller driver is used only NPCM8XX SoC, so the only
>>> GCR node in the NPCM8XX SoC is nuvoton,npcm845-gcr?
>>
>> Yes. The previous case (reset driver, AFAIR) was also about driver used
>> only in one SoC, wasn't it?
> Actually not, the NPCM reset driver serves NPCM7XX and NPCM8XX and
> probably other future BMC SoC's

No, when someone developed reset driver, it served only NPCM7XX. So
using this argument - only one SoC is supported - that person use
exactly the same API as here.

And it was wrong...

Now you use the same argument - only one SoC is supported.

I clearly see a pattern here...

> Still, you suggest using the phandle property in the driver even if
> the driver serves one SoC?
>>
>> Best regards,
>> Krzysztof
> 
> Best regards,
> 
> Tomer


Best regards,
Krzysztof