Re: [PATCH 1/1] dt-bindings: clock: imx5: add assigned-clock* properties

[Frank Li]

On Wed, Sep 02, 2026 at 08:14:46AM +0200, Krzysztof Kozlowski wrote:
> On Tue, Sep 01, 2026 at 03:37:19PM -0400, Frank.Li@xxxxxxxxxxx wrote:
> > From: Frank Li <Frank.Li@xxxxxxx>
> >
> > The i.MX5 CCM is a pure clock provider with no 'clocks' consumer
> > property in its binding. The dtschema framework only merges the
> > clocks.yaml meta-schema properties (which include assigned-clocks,
> > assigned-clock-parents, and assigned-clock-rates) into a binding
> > when the binding itself declares 'clocks:'.
> >
> > For example, imx53-m53menlo.dts configures the CKO1 output clock
> > by setting assigned-clocks on the &clks node:
> >
> >   &clks {
> >     assigned-clocks = <&clks IMX5_CLK_CKO1_SEL>, ...;
> >     assigned-clock-parents = <&clks IMX5_CLK_AHB>;
> >     assigned-clock-rates = <133333334>, <33333334>, <33333334>;
> >   };
> >
> > Without explicit declarations this triggers a dtbs_check error:
> >   arch/arm/boot/dts/nxp/imx/imx53-m53menlo.dtb: ccm@53fd4000 (fsl,imx53-ccm): 'assigned-clock-parents', 'assigned-clock-rates', 'assigned-clocks' do not match any of the regexes: '^pinctrl-[0-9]+$'
> >
> > Assisted-by: Claude:claude-sonnet-4.6
> > Signed-off-by: Frank Li <Frank.Li@xxxxxxx>
>
> Then the DTS is wrong. If this device does not have a parent clock
> input, then assigning clock parents is wrong because it is either
> contradictory (you said no parents) or misplaced (clock consumer should
> set it).

Make sense, CLKO may just use for debug to measure one of internal clock.

The boards is too old, not sure if it is really used for some periphal.

Frank

>
> Best regards,
> Krzysztof
>