On Thu, Nov 28, 2019 at 03:19:38PM +0300, Nikita Yushchenko wrote:

> > I'm not seeing any integration with DAPM here, I'd expect to see that so
> > we don't cut off the start of audio especially with the longer times
> > available (which I'm frankly not sure are seriously usable).

> I believe driver already has that integration, there is
> aic31xx_dapm_power_event() that is called on DAPM events, and polls state in
> register bits waiting for operation to complete.

> Btw, the default setting for register fields in question is "304ms" /
> "3.9ms" thus some delay is already there. This patch just makes it
> explicitly controllable by those who wait it.

Can you confirm that this does take effect (should be easy with the
longer delays) and put a comment in indicating that please in case
someone is cut'n'pasting?