On Fri, 2017-01-20 at 07:39 +0100, Thierry Reding wrote: > On Thu, Jan 19, 2017 at 05:52:10PM +0100, Clemens Gruber wrote: > > On Thu, Jan 19, 2017 at 06:10:08PM +0200, Andy Shevchenko wrote: > > > Combining with your proposal I would see the best approach is to > > > set > > > pca->period_ns accordingly to current prescaler value if you want > > > to.
> > I'll send v2 of patch 2/2 with aforementioned changes in the next > > days. > > Like I said above, I think atomic API conversion wouldn't be very > difficult for this driver and it has the added advantage of giving you > the proper infrastructure to do this rather than having to duplicate > it in the driver. > > That would be my preference, but I'm willing to take v2 of 2/2 as well > if it ends up being really nice and compact. =) I think we may split to this fix and separate change to move to atomic API. Does it sound reasonable? P.S. <offtopic>Can you comment further pwm-lpss changes Mika and I did?</offtopic> -- Andy Shevchenko <andriy.shevche...@linux.intel.com> Intel Finland Oy
