[PATCH 2/2] OPP/pmdomain: Fix the assignment of the required-devs

[Ulf Hansson]

On Tue, 3 Sept 2024 at 09:16, Viresh Kumar <viresh.kumar at linaro.org> wrote:
>
> On 03-09-24, 00:48, Ulf Hansson wrote:
> > To fix this problem, let's instead start by letting the OPP core find the
> > device node for the required OPP table and then let genpd search for a
> > corresponding OPP table, allowing us the find the correct required-dev to
> > assign for it.
>
> Why was doing this necessary ?

Let me try to elaborate a bit more.

In the current code, genpd_find_opp_table() tries to find an OPP table
for the genpd that the device is getting attached to. Then genpd
passes that OPP table via devm_pm_opp_set_config(), to let the OPP
core to hook up a required-dev for it. This was a naive approach, as
that OPP table may not be the one that actually corresponds to a
required-opps for the required-dev. Consider the below in DT.

        opp_table_devA: opp-table-devA {
                compatible = "operating-points-v2";

                opp-devA-50 {
                        opp-hz = /bits/ 64 <2500>;
                        required-opps = <&opp_pd_50>; //corresponds to
pd_perf1's OPP table
                };
               ....

        devA {
                compatible = "foo,bar";
                power-domains = <&pd_perf0>, <&pd_perf1>; //both
pd_perf0 and pd_perf1 has OPP tables.
                power-domain-names = "perf0", "perf1";
                operating-points-v2 = <&opp_table_devA>;
        };

To make sure we assign the correct required-dev for cases like the
above, we need to let the OPP core to iterate through the available
required-opps and see if some of them are corresponding to the OPP
table for the genpd the required-dev belongs too.

To manage this in a non-genpd specific way, I added another callback
in struct dev_pm_opp_config. In this way, it should work for any
future possible required-devs types too, I think.

Kind regards
Uffe