Re: [PATCH v2 10/10] arm64: dts: qcom: Add the support of cpufreq on SDX75

[Pavan Kondeti]

On Mon, Jun 05, 2023 at 08:21:39PM +0200, Konrad Dybcio wrote:
> 
> 
> On 5.06.2023 18:29, Rohit Agarwal wrote:
> > Add the support of cpufreq to enable the cpufreq scaling
> > on SDX75 SoC. Also add CPU specific information to build
> > energy model for EAS.
> > 
> > Signed-off-by: Rohit Agarwal <quic_rohiagar@xxxxxxxxxxx>
> > ---
> >  arch/arm64/boot/dts/qcom/sdx75.dtsi | 30 ++++++++++++++++++++++++++++++
> >  1 file changed, 30 insertions(+)
> > 
> > diff --git a/arch/arm64/boot/dts/qcom/sdx75.dtsi b/arch/arm64/boot/dts/qcom/sdx75.dtsi
> > index 47170ae..e1887a4 100644
> > --- a/arch/arm64/boot/dts/qcom/sdx75.dtsi
> > +++ b/arch/arm64/boot/dts/qcom/sdx75.dtsi
> > @@ -47,10 +47,14 @@
> >  			device_type = "cpu";
> >  			compatible = "arm,cortex-a55";
> >  			reg = <0x0 0x0>;
> > +			clocks = <&cpufreq_hw 0>;
> >  			enable-method = "psci";
> >  			power-domains = <&CPU_PD0>;
> >  			power-domain-names = "psci";
> >  			next-level-cache = <&L2_0>;
> > +			qcom,freq-domain = <&cpufreq_hw 0>;
> > +			capacity-dmips-mhz = <1024>;
> > +			dynamic-power-coefficient = <100>;
> >  			L2_0: l2-cache {
> >  				compatible = "cache";
> >  				next-level-cache = <&L3_0>;
> > @@ -64,10 +68,14 @@
> >  			device_type = "cpu";
> >  			compatible = "arm,cortex-a55";
> >  			reg = <0x0 0x100>;
> > +			clocks = <&cpufreq_hw 0>;
> >  			enable-method = "psci";
> >  			power-domains = <&CPU_PD1>;
> >  			power-domain-names = "psci";
> >  			next-level-cache = <&L2_100>;
> > +			qcom,freq-domain = <&cpufreq_hw 0>;
> > +			capacity-dmips-mhz = <1024>;
> > +			dynamic-power-coefficient = <100>;
> >  			L2_100: l2-cache {
> >  				compatible = "cache";
> >  				next-level-cache = <&L3_0>;
> > @@ -78,10 +86,14 @@
> >  			device_type = "cpu";
> >  			compatible = "arm,cortex-a55";
> >  			reg = <0x0 0x200>;
> > +			clocks = <&cpufreq_hw 0>;
> >  			enable-method = "psci";
> >  			power-domains = <&CPU_PD2>;
> >  			power-domain-names = "psci";
> >  			next-level-cache = <&L2_200>;
> > +			qcom,freq-domain = <&cpufreq_hw 0>;
> > +			capacity-dmips-mhz = <1024>;
> > +			dynamic-power-coefficient = <100>;
> >  			L2_200: l2-cache {
> >  				compatible = "cache";
> >  				next-level-cache = <&L3_0>;
> > @@ -92,10 +104,14 @@
> >  			device_type = "cpu";
> >  			compatible = "arm,cortex-a55";
> >  			reg = <0x0 0x300>;
> > +			clocks = <&cpufreq_hw 0>;
> >  			enable-method = "psci";
> >  			power-domains = <&CPU_PD3>;
> >  			power-domain-names = "psci";
> >  			next-level-cache = <&L2_300>;
> > +			qcom,freq-domain = <&cpufreq_hw 0>;
> > +			capacity-dmips-mhz = <1024>;
> That sounds a bit bogus.. Thinking about it, it sounds bogus on most
> platforms we have support for! I guess SM8250 big cores aren't *really*
> equally as powerful..
> 

These are relative values on a given CPU. If a b.L SoC has 2 LITTLE CPUs
and 2 big CPUs. The LITTLE CPU capacity is relative to big CPU capacity
i.e 1024. Thats the reason for SM8250 having capacity-dmips-mhz as 448
for LITTLE CPUs (0-3) and 1024 as Big CPUs.

Since SDx75 has all similar capacity CPUs, setting 1024 or for that
matter any value is good enough.

Thanks,
Pavan