Re: [PATCH v4 1/1] pwm: pca9685: fix gpio-only operation.

From: Andy Shevchenko
Date: Thu Apr 20 2017 - 03:29:45 EST


On Tue, Apr 18, 2017 at 6:52 PM, Sven Van Asbroeck <thesven73@xxxxxxxxx> wrote:
> Thanks for the feedback Andy !!

You're welcome.

>
>> I would go with
>>
>> /* Wait for @sleep microseconds for the oscillator to be back up */
>> if (sleep)
>> udelay(sleep);
>>
>> Otherwise int sleep is oddly here.
>>
>> Or
>>
>> bool sleep
>>
>> /* Wait 500us ... */
>> if (sleep)
>> udelay(500);
>>
>>> +}
>
> I think you may be getting confused between:
> - the chip's SLEEP bit (int sleep)
> - the amount of time to delay after chip comes _out of_ sleep.
> (always 500 us)
>
> If it's confusing for you, it might be confusing for others?
> Perhaps change the parameter to 'bool sleep_bit' or 'bool do_sleep'
> to make the distinction clearer?

Taking above into consideration perhaps sleep is not quite good word
at all. By functional description it sounds like latency tolerance to
me.

>> __maybe_unused and remove ugly #ifdef:ery.
>
> If this works on non- CONFIG_PM systems, I'm all for it !
> Grepping the drivers/ directory, I see that some drivers use
> #ifdef CONFIG_PM, some use __maybe_unused for runtime_pm.

This approach kinda new that's why you see variety of approaches.

> Mika and Thierry, thoughts ?

At the end it's Thierry's call, so, I'm not insisting.

--
With Best Regards,
Andy Shevchenko