On Tue, Oct 01, 2013 at 12:26:07PM -0600, Stephen Warren wrote:
> On 09/23/2013 03:40 PM, Thierry Reding wrote:
> > In preparation for adding an optional regulator and enable GPIO to the
> > driver, split the power on and power off sequences into separate
> > functions to reduce code duplication at the multiple call sites.
>
> > diff --git a/drivers/video/backlight/pwm_bl.c b/drivers/video/backlight/pwm_bl.c
>
> > +static void pwm_backlight_power_off(struct pwm_bl_data *pb)
> > +{
> > + pwm_disable(pb->pwm);
>
> Both the call-sites you're replacing do the following before pwm_disable():
>
> pwm_config(pb->pwm, 0, pb->period);
>
> While I agree that probably shouldn't be necessary, I think it's at
> least worth mentioning that in the commit description just to make it
> obvious that it was a deliberate change. Splitting that change into a
> separate patch might be reasonable in order to keep refactoring and
> functional changes separate, although perhaps it's not worth it.
Actually I'll add that back. I'm not sure exactly why I dropped it (it
may just have been oversight) and there have been reports that some HW
actually requires pwm_config(..., 0, ...) before pwm_disable() to turn
off properly.
> There are also a couple unrelated whitespace changes thrown in here.
I think they increase readability, but I can certainly move them into a
separate patch.
Thierry
Attachment:
pgpXcAKTjkKHc.pgp
Description: PGP signature
