[PATCH 1/2] dt-bindings: phy: qcom, qmp: Mark '#clock-cells' as a 'optional' property

Bhupesh Sharma bhupesh.sharma at linaro.org
Sat Feb 26 20:19:31 PST 2022


On Sun, 27 Feb 2022 at 05:08, Dmitry Baryshkov
<dmitry.baryshkov at linaro.org> wrote:
>
> On 25/02/2022 12:16, Bhupesh Sharma wrote:
> > Hi Dmitry,
> >
> > Thanks for your review comments.
> >
> > On Thu, 24 Feb 2022 at 20:22, Dmitry Baryshkov
> > <dmitry.baryshkov at linaro.org> wrote:
> >>
> >> On Thu, 24 Feb 2022 at 15:33, Bhupesh Sharma <bhupesh.sharma at linaro.org> wrote:
> >>>
> >>> Since '#clock-cells' is not a required property for several
> >>> QCoM boards supporting qmp-phy(s) (for e.g. sm8150, sm8250 or sm8350 SoC
> >>> based boards), mark it as an optional property instead.
> >>
> >> I believe the description is not correct. the clock-cells should not
> >> be used at all, so it should be removed from the root node.
> >
> > Hmm.. 'clock-cells' is still used for describing qmp phy pcie nodes for certain
> > devices like 'msm8996.dtsi'. I am not sure if removing it would impact existing
> > dts files. But let me try cleaning them up in v2.
>
> Checked. Unless I'm mistaken, msm8996.dtsi doesn't use #clock-cells in
> root QMP PHY nodes.

Indeed. I meant that we made some cleanups already via
82d61e19fccb ("arm64: dts: qcom: msm8996: Move '#clock-cells' to QMP
PHY child node"),
but it seems a few dts still need to be cleaned up to move the
'#clock-cells' to QMP PHY child node instead of the root node.

I will send a fixed version in v2 :)

Regards,
Bhupesh



More information about the linux-phy mailing list