Re: [PATCH v3 06/10] iio: afe: rescale: add offset support

["Liam Beguin" <liambeguin@gmail.com>]

On Sun Jul 4, 2021 at 12:53 PM EDT, Jonathan Cameron wrote:
> On Wed, 30 Jun 2021 21:00:30 -0400
> Liam Beguin <liambeguin@gmail.com> wrote:
>
> > From: Liam Beguin <lvb@xiphos.com>
> > 
> > This is a preparatory change required for the addition of temperature
> > sensing front ends.
> > 
> > Signed-off-by: Liam Beguin <lvb@xiphos.com>
> Hi Liam.
>
> A few remaining things inline.
>
> Looking good in general.
>
> Jonathan
>
> > ---
> >  drivers/iio/afe/iio-rescale.c | 63 +++++++++++++++++++++++++++++++++++
> >  1 file changed, 63 insertions(+)
> > 
> > diff --git a/drivers/iio/afe/iio-rescale.c b/drivers/iio/afe/iio-rescale.c
> > index 8f79c582519c..c8750286c308 100644
> > --- a/drivers/iio/afe/iio-rescale.c
> > +++ b/drivers/iio/afe/iio-rescale.c
> > @@ -32,6 +32,7 @@ struct rescale {
> >  	bool chan_processed;
> >  	s32 numerator;
> >  	s32 denominator;
> > +	s32 offset;
> >  };
> >  
> >  static int rescale_read_raw(struct iio_dev *indio_dev,
> > @@ -39,6 +40,8 @@ static int rescale_read_raw(struct iio_dev *indio_dev,
> >  			    int *val, int *val2, long mask)
> >  {
> >  	struct rescale *rescale = iio_priv(indio_dev);
> > +	int scale, scale2;
> > +	int schan_off = 0;
> >  	s64 tmp, tmp2;
> >  	u32 factor;
> >  	int ret;
> > @@ -99,6 +102,62 @@ static int rescale_read_raw(struct iio_dev *indio_dev,
> >  			dev_err(&indio_dev->dev, "unsupported type %d\n", ret);
> >  			return -EOPNOTSUPP;
> >  		}
> > +	case IIO_CHAN_INFO_OFFSET:
> > +		/*
> > +		 * Processed channels are scaled 1-to-1 and source offset is
> > +		 * already taken into account.
> > +		 *
> > +		 * In other cases, the final offset value is defined by:
> > +		 *	offset = schan_offset + rescaler_offset / schan_scale
>
> Maths is right, but perhaps useful to express how this is derived as I
> had
> to scribble it down to check.
>
> Want to express real world measurement as:
> scale * (raw + offset)
> Given we applying scale and offset recursively we have.
> rescaler_scale * (schan_scale * (raw + schan_offset) + rescaler_offset)
> which can be rearrange into the correct form at.
> rescaler_scale *schan_scale* (raw + (schan_offset +
> rescaler_offset/schan_scale))
> Thus the scale and offset we expose to userspace should be.
> scale = rescaler_scale * schan_scale
> offset = schan_offset + rescaler_offset/schan_scale;
>

Understood, I'll update the comment with more details.

>
> > +		 */
> > +		if (rescale->chan_processed) {
> > +			*val = rescale->offset;
> > +			return IIO_VAL_INT;
> > +		}
> > +
> > +		if (iio_channel_has_info(rescale->source->channel,
> > +					 IIO_CHAN_INFO_OFFSET)) {
> > +			ret = iio_read_channel_offset(rescale->source,
> > +						      &schan_off, NULL);
> > +			if (ret < 0)
> > +				return ret;
> if you've returned in the first branch of the if, no need to worry about
> the
> else for checking the second condition.
>
> if (ret != IIO_VAL_INT)
> return...

Right, I'll combine both statements.

>
>
> > +			else if (ret != IIO_VAL_INT)
> > +				return -EOPNOTSUPP;
> > +		}
> > +
> > +		ret = iio_read_channel_scale(rescale->source, &scale, &scale2);
> > +		switch (ret) {
> > +		case IIO_VAL_FRACTIONAL:
> > +			tmp = (s64)rescale->offset * scale2;
> > +			*val = div_s64(tmp, scale) + schan_off;
> > +			return IIO_VAL_INT;
> > +		case IIO_VAL_INT:
> > +			*val = div_s64(rescale->offset, scale) + schan_off;
> > +			return IIO_VAL_INT;
> > +		case IIO_VAL_FRACTIONAL_LOG2:
> > +			tmp = (s64)rescale->offset * (1 << scale2);
> > +			*val = div_s64(tmp, scale) + schan_off;
> > +			return IIO_VAL_INT;
> > +		case IIO_VAL_INT_PLUS_NANO:
> > +			tmp = (s64)rescale->offset * 1000000000UL;
> > +			tmp2 = ((s64)scale * 1000000000UL) + scale2;
> > +			factor = gcd(tmp, tmp2);
> > +			tmp /= factor;
> > +			tmp2 /= factor;
>
> What is the benefit to doing gcd division before div_s64?
>

You're right the GCD division is unnecessary here, will drop.

> > +			*val = div_s64(tmp, tmp2) + schan_off;
> > +			return IIO_VAL_INT;
> > +		case IIO_VAL_INT_PLUS_MICRO:
> > +			tmp = (s64)rescale->offset * 1000000UL;
> > +			tmp2 = ((s64)scale * 1000000UL) + scale2;
> > +			factor = gcd(tmp, tmp2);
> > +			tmp /= factor;
> > +			tmp2 /= factor;
> > +			*val = div_s64(tmp, tmp2) + schan_off;
> > +			return IIO_VAL_INT;
> > +		default:
> > +			dev_err(&indio_dev->dev, "unsupported type %d\n", ret);
> > +			return -EOPNOTSUPP;
> > +		}
> >  	default:
> >  		return -EINVAL;
> >  	}
> > @@ -175,6 +234,9 @@ static int rescale_configure_channel(struct device *dev,
> >  	chan->info_mask_separate = BIT(IIO_CHAN_INFO_RAW) |
> >  		BIT(IIO_CHAN_INFO_SCALE);
> >  
> > +	if (rescale->offset)
> > +	    chan->info_mask_separate |= BIT(IIO_CHAN_INFO_OFFSET);
> > +
> >  	/*
> >  	 * Using .read_avail() is fringe to begin with and makes no sense
> >  	 * whatsoever for processed channels, so we make sure that this cannot
> > @@ -339,6 +401,7 @@ static int rescale_probe(struct platform_device *pdev)
> >  	rescale->cfg = of_device_get_match_data(dev);
> >  	rescale->numerator = 1;
> >  	rescale->denominator = 1;
> > +	rescale->offset = 0;
> >  
> >  	ret = rescale->cfg->props(dev, rescale);
> >  	if (ret)