Re: [PATCH V2 2/2] pinctrl: bcm: add driver for BCM4908 pinmux

["Rafał Miłecki" <zajec5@gmail.com>]

On 22.12.2021 13:13, Andy Shevchenko wrote:
> On Wed, Dec 22, 2021 at 1:11 PM Rafał Miłecki <zajec5@gmail.com> wrote:
>> From: Rafał Miłecki <rafal@milecki.pl>
>>
>> BCM4908 has its own pins layout so it needs a custom binding and a Linux
>> driver.
> 
> ...
> 
>> V2: Formatting fixes
>>      Kconfig fix
>>      Cleanup of #include-s
>>      Use devm_kasprintf_strarray()
> 
> Thanks, but it seems there are unsettled down points as per v1.
> Can you comment on them there?

Those remaining comments are a matter of personal taste & details of
personal coding style. We don't have defined rules for such details.

If developer submitted code that matches *defined* rules and is fine to
read I don't see why we should enforce someone's coding style. We may
easily get into pointless and time wasting argues between multiple
developers.

Empty line after one-line comment isn't against rules and checkpatch.pl
doesn't complain about it.

I've never heard of rule of sorting #include-s from the most generic to
the most particular one. I don't even know how to meter that. Actually
coding-style.rst suggests #include-s should be sorted but without
specifying how. My first guess is alphabetical order.

If you think some extra coding style should be enforced for Linux code
please kindly update coding-style.rst and checkpatch.pl so that:

1. We have clear rules
2. We keep code consistent across subsystems
3. It can be automatically verified
4. There are not more argues about what's the preferred format

As I pointed out we have over 1000 examples of empty line above
module_platform_driver() so clearly what you describe as common sense
isn't clear for all developers.