On 23/01/2023 14:00, Stephan Gerhold wrote:
Unless this conclusion changes with your CPR patch set this means that
both the DTS and the DT schema will need changes anyway, because you
wouldn't need power-domain-names = "cpr", but rather
power-domains = <&rpmpd MSM8939_VDDMX_AO>, <&vreg_dummy>;
power-domain-names = "mx", "cpr";
I have not been owning the CPR for 8939 so far but, this what we have in
our 4.19 tree.
CPU0: cpu@100 {
device_type = "cpu";
compatible = "arm,cortex-a53", "arm,armv8";
reg = <0x100>;
next-level-cache = <&L2_1>;
enable-method = "qcom,kpss-acc-v2";
qcom,acc = <&acc0>;
qcom,saw = <&saw0>;
clocks = <&apcs1>;
operating-points-v2 = <&cluster1_opp_table>;
power-domains = <&cpr>;
power-domain-names = "cpr";
#cooling-cells = <2>;
capacity-dmips-mhz = <1024>;
};
cpr: power-controller@b018000 {
compatible = "qcom,msm8939-cpr", "qcom,cpr";
reg = <0x0b018000 0x1000>;
interrupts = <0 15 IRQ_TYPE_EDGE_RISING>;
clocks = <&rpmcc CXO_SMD_CXO_A_CLK>;
clock-names = "ref";
power-domains = <&rpmpd MSM8939_VDDMX_AO>;
#power-domain-cells = <0>;
operating-points-v2 = <&cpr_opp_table>;
};
So the CPR code not the CPU code owns VDDMX_AO. I'm not sure if there's
a good reason why it has been done that way.
Anyway, this feels like a bit of a departure from our core discussion. I
will see if it is possible to drop the CPU power-domain entirely
contingent on the patch you flagged.
---
bod
