Re: [PATCH v6 06/10] regulator: use linear_ranges helper

["Vaittinen, Matti" <Matti.Vaittinen@fi.rohmeurope.com>]

Hello Mark,

On Wed, 2020-03-25 at 11:34 +0000, Mark Brown wrote:
> On Wed, Mar 25, 2020 at 06:58:56AM +0000, Vaittinen, Matti wrote:
> > On Tue, 2020-03-24 at 14:24 +0000, Mark Brown wrote:
> > > On Tue, Mar 24, 2020 at 10:30:54AM +0200, Matti Vaittinen wrote:
> > > >  
> > > > -		if (!(min_uV <= linear_max_uV && max_uV >=
> > > > range-
> > > > > min_uV))
> > > > +		ret = linear_range_get_selector_high(range,
> > > > min_uV,
> > > > &sel,
> > > > +						     &found);
> > > > +		if (ret)
> > > >  			continue;
> > > Why are these going for _get_selector_high()?  Usually we try to
> > > choose
> > > the lowest voltage we can.
> > I tried to maintain the existing logic - which is (if I am not
> > mistaken) to search for first value which is equal or higher than
> > the
> > minimum. Hence we use the linear_range_get_selector_high() - which
> > does
> > not return selector for voltage smaller than given value (minimum).
> 
> OK, that's a bit confusingly named then - I'd read that as finding
> the
> highest value within the range without registering that the function
> only takes a single value so can't verify if the selector is in range
> or
> not.  I'll also need to check that the modified code verifies that
> the
> selector returned is within range.

In this API the boolan "found" reflects whether the value referred by
the selector is in range. I didn't check the "found" here because the
returned sel is in any case passed to rdev->desc->ops-
>list_voltage(rdev, sel) - and returned voltage is checked to be in
valid range.

And Rob once wrote to me that "naming is hard" :) And I do agree. I
selected this name because API returns value which is equal or higher
than given. _low variant returns value which is lower or equal.

Just a note - Dropped some email from CC as I keep getting delivery
failures.

Br,
	Matti
_______________________________________________
Linux-mediatek mailing list
Linux-mediatek@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-mediatek