On Tue, Feb 23, 2021 at 10:50:18AM -0800, Curtis Malainey wrote:

> > Are you sure that *zero* debounce is needed for the button presses?
> > 250ms does look like a lot of time but zero might be going from one
> > extreme to the other.

> Fair point, I was looking at some other codecs and why they respond so
> quickly, it appears they have no fixed delay and just call schedule
> work. That being said, I can easily double tap <100ms. So Ideally i
> would like to keep this on the order of ~50ms at most. I am guessing
> Realtek will want to keep the 250ms for jack detect still.

Those feel like plausible numbers to me assuming there's no hardware
debounce.

> Would queueing two separate jobs with two different delays be the
> simple way to go? Realtek does this sound fine to you?

Possibly just queuing the same job with different timeouts?  I don't
have particularly strong feelings assuming the resulting code make
sense.