On Tue, Apr 18, 2017 at 6:52 PM, Sven Van Asbroeck <thesve...@gmail.com> 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
