Quoting Manivannan Sadhasivam (2020-11-24 19:49:24)
>
>
> On 25 November 2020 7:44:10 AM IST, Stephen Boyd <[email protected]> wrote:
> >Quoting Manivannan Sadhasivam (2020-11-18 23:27:11)
> >> diff --git
> >a/Documentation/devicetree/bindings/clock/qcom,gcc-sdx55.yaml
> >b/Documentation/devicetree/bindings/clock/qcom,gcc-sdx55.yaml
> >> new file mode 100644
> >> index 000000000000..9d8981817ae3
> >> --- /dev/null
> >> +++ b/Documentation/devicetree/bindings/clock/qcom,gcc-sdx55.yaml
> >> @@ -0,0 +1,73 @@
> >[...]
> >> +
> >> +properties:
> >> + compatible:
> >> + const: qcom,gcc-sdx55
> >> +
> >> + clocks:
> >> + items:
> >[...]
> >> + - description: PLL test clock source
> >> +
> >> + clock-names:
> >> + items:
> >[...]
> >> + - const: core_bi_pll_test_se
> >
> >Is it optional? As far as I know this clk has never been implemented
> >because it's a hardware validation thing and not used otherwise.
>
> It is implemented in drivers but not used as you said. But since it is the
> parent clk of PLLs I'm not sure we can make it optional.
We can leave it out completely if the bootloader code never uses it as a
parent of the PLL. That scenario would be pretty weird and is why we
removed it from the video clk controller in commit abc8f93f33e7 ("clk:
qcom: Get rid of the test clock for videocc-sc7180"). I'm fine if you
want to keep it, but I'm confused why you care so much :)