Re: [PATCH 2/2] arm64: dts: qcom: sdm845: cheza doesn't support LMh node

[Dmitry Baryshkov]

On Thu, 24 Aug 2023 at 21:25, Doug Anderson <dianders@xxxxxxxxxxxx> wrote:
>
> Hi,
>
> On Wed, Aug 23, 2023 at 3:28 PM David Heidelberg <david@xxxxxxx> wrote:
> >
> > Cheza firmware doesn't allow controlling LMh from the operating system.
> >
> > Fixes: 36c6581214c4 ("arm64: dts: qcom: sdm845: Add support for LMh node")
> > Suggested-by: Dmitry Baryshkov <dmitry.baryshkov@xxxxxxxxxx>
> > Signed-off-by: David Heidelberg <david@xxxxxxx>
> > ---
> >  arch/arm64/boot/dts/qcom/sdm845-cheza.dtsi | 14 ++++++++++++++
> >  1 file changed, 14 insertions(+)
> >
> > diff --git a/arch/arm64/boot/dts/qcom/sdm845-cheza.dtsi b/arch/arm64/boot/dts/qcom/sdm845-cheza.dtsi
> > index d86b0d112110..8cc8fc290fd3 100644
> > --- a/arch/arm64/boot/dts/qcom/sdm845-cheza.dtsi
> > +++ b/arch/arm64/boot/dts/qcom/sdm845-cheza.dtsi
> > @@ -143,6 +143,10 @@ panel_in_edp: endpoint {
> >         };
> >  };
> >
> > +&cpufreq_hw {
> > +       /delete-property/ interrupts-extended; /* reference to lmh_cluster[01] */
> > +};
> > +
> >  &psci {
> >         /delete-node/ power-domain-cpu0;
> >         /delete-node/ power-domain-cpu1;
> > @@ -275,6 +279,14 @@ &BIG_CPU_SLEEP_1
> >                            &CLUSTER_SLEEP_0>;
> >  };
> >
> > +&lmh_cluster0 {
> > +       status = "disabled";
> > +};
> > +
> > +&lmh_cluster1 {
> > +       status = "disabled";
> > +};
>
> It's not a huge deal to me, but as I understand it usually you'd put
> the "disabled" in sdm845.dtsi and then it would be up to all the other
> sdm845 boards to mark this as "okay".

That depends. If a device is usually available (like LMH) and only
disabled in some obscure cases, it is better to disable it only in
this obscure case (in my opinion).

> >  /*
> >   * Reserved memory changes
> >   *
> > @@ -338,6 +350,8 @@ flash@0 {
> >
> >
> >  &apps_rsc {
> > +       /delete-property/ power-domains;
> > +
>
> Is the deletion of the "power-domains" here related to LMh? That seems
> like it was added to sdm845.dtsi in a separate commit that doesn't
> talk about LMh at all...

CLUSTER_PD becomes unavailable after the previous patch. So maybe you
are right and this chunk should be moved to the previous patch.

-- 
With best wishes
Dmitry