Hi,
On Mon, Feb 22, 2021 at 12:45 PM Stephen Boyd <swboyd@xxxxxxxxxxxx> wrote:
>
> Quoting Matthias Kaehlcke (2021-02-22 12:38:46)
> > On Mon, Feb 22, 2021 at 12:20:04PM -0800, Stephen Boyd wrote:
> > > Quoting Matthias Kaehlcke (2021-02-19 18:10:59)
> > > > Lazor rev3 and older are stuffed with a 47k NTC as thermistor for
> > > > the charger temperature which currently isn't supported by the
> > > > PM6150 ADC driver. Delete the charger thermal zone and ADC channel
> > > > to avoid the use of bogus temperature values.
> > > >
> > > > Signed-off-by: Matthias Kaehlcke <mka@xxxxxxxxxxxx>
> > > > ---
> > > >
> > > > arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r0.dts | 9 +++++++++
> > > > arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r1.dts | 9 +++++++++
> > > > arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r3.dts | 9 +++++++++
> > > > 3 files changed, 27 insertions(+)
> > > >
> > > > diff --git a/arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r0.dts b/arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r0.dts
> > > > index 30e3e769d2b4..0974dbd424e1 100644
> > > > --- a/arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r0.dts
> > > > +++ b/arch/arm64/boot/dts/qcom/sc7180-trogdor-lazor-r0.dts
> > > > @@ -14,6 +14,15 @@ / {
> > > > compatible = "google,lazor-rev0", "qcom,sc7180";
> > > > };
> > > >
> > > > +/*
> > > > + * rev <= 3 are stuffed with a 47k NTC as charger thermistor which is currently
> > > > + * not supported by the PM6150 ADC driver. Delete the thermal zone and ADC
> > > > + * channel to avoid the use of bogus temperature values.
> > > > + */
> > > > +/delete-node/ &charger_thermal;
> > > > +/delete-node/ &pm6150_adc_charger_thm;
> > > > +/delete-node/ &pm6150_adc_tm_charger_thm;
> > >
> > > Can we disable pm6150_adc_tm instead on <= rev3 boards? It would be the
> > > same number of lines, but is simpler to reason about disabled nodes vs.
> > > deleted nodes usually.
> >
> > For Lazor theoretically this could be done since it doesn't use other ADC
> > channels, however it won't work for other trogdor devices that will be
> > upstreamed eventually. Some of these boards have the same problem, however
> > they have other thermistors connected to the ADC. One could argue that it's
> > preferable to do things in a uniform way, but I'm open to do it either way
> > for Lazor.
> >
>
> I see. Can the thermal-zone be disabled then vs. deleting three nodes? I
> think the thermal driver uses for_each_available_child_of_node() so that
> would work?
FWIW: +1 to what Stephen suggests assuming it works.
-Doug
