Hi Thierry,

On Thu, 10 Mar 2016 18:54:38 +0100
Thierry Reding [off-list ref] wrote:

On Mon, Mar 07, 2016 at 08:34:19AM -0800, Doug Anderson wrote:

quoted

Thierry,

On Thu, Feb 25, 2016 at 3:14 PM, Doug Anderson [off-list ref] wrote:

quoted

So just to summarize:

* Add pwm_get_state(), pwm_apply_state(), pwm_get_args().
pwm_get_state() initially returns 0 for duty cycle if driver doesn't
support readout.

* Re-implement pwm_get_period() (and maybe other similar functions)
atop pwm_get_state() as you describe earlier in the thread.

* Document pwm_get_period() (and maybe other similar functions) as deprecated.

* Fix drivers for all current 2 users of PWM regulator to support
hardware readout.

* Update PWM regulator as you described earlier in the thread (Feb 23).

* If PWM regulator is ever used on a new board whose PWM doesn't
support hardware readout, the voltage will change at probe time.


Did I get all that right?  Thanks!

Can you provide a "yes, you got that right" or a "no, you didn't
understand"?  That will unblock Boris, I think.

Sounds about right. Hopefully this will eliminate any objections that
others had about the series.

Okay, I'll rework the series accordingly and send a new version soon.

Thanks,

Boris

-- 
Boris Brezillon, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com