On Sat, 21 Nov 2020 14:58:49 +0000
Jonathan Cameron <[email protected]> wrote:

> On Fri, 20 Nov 2020 10:29:44 -0800
> Stephen Boyd <[email protected]> wrote:
> 
> > This DT property can be 0, 16, and then 64, but not 32. The math here
> > doesn't recognize this slight bump in the power of 2 numbers and
> > translates a DT property of 64 into the register value '3' when it
> > really should be '2'. Fix it by subtracting one more if the number being
> > translated is larger than 16.
> > 
> > Cc: Daniel Campello <[email protected]>
> > Cc: Lars-Peter Clausen <[email protected]>
> > Cc: Peter Meerwald-Stadler <[email protected]>
> > Cc: Douglas Anderson <[email protected]>
> > Cc: Gwendal Grignou <[email protected]>
> > Cc: Evan Green <[email protected]>
> > Signed-off-by: Stephen Boyd <[email protected]>
> > ---
> > 
> > Changes from v1 
> > (https://lore.kernel.org/r/[email protected]):
> >  * Changed ternary to consider 17 to 31 as the same as 16
> > 
> >  drivers/iio/proximity/sx9310.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/drivers/iio/proximity/sx9310.c b/drivers/iio/proximity/sx9310.c
> > index a2f820997afc..ee1b4ff05a37 100644
> > --- a/drivers/iio/proximity/sx9310.c
> > +++ b/drivers/iio/proximity/sx9310.c
> > @@ -1305,7 +1305,7 @@ sx9310_get_default_reg(struct sx9310_data *data, int 
> > i,
> >             if (ret)
> >                     break;
> >  
> > -           pos = min(max(ilog2(pos), 3), 10) - 3;
> > +           pos = min(max(ilog2(pos), 3), 11) - (pos >= 32 ? 4 : 3);  
> 
> Given this is now rather tricky to read, I'd just do an explicit lookup of
> the allowed values.  Probably just a switch statement.
As an alternative, a comment would help saying that it's powers of 2 except
for a gap at 64.

Thanks,

J

> 
> Thanks,
> 
> Jonathan
> 
> >             reg_def->def &= ~SX9310_REG_PROX_CTRL7_AVGPOSFILT_MASK;
> >             reg_def->def |= 
> > FIELD_PREP(SX9310_REG_PROX_CTRL7_AVGPOSFILT_MASK,
> >                                        pos);
> > 
> > base-commit: 5b19ca2c78a0838976064c0347e46a2c859b541d  
> 

Reply via email to