Re: [PATCH v3 4/4] arm64: dts: qcom: Fix pm4125 vbus regulator compatible and constraints

[Konrad Dybcio]

On 7/7/26 10:33 AM, Rakesh Kota wrote:
> On Mon, Jul 06, 2026 at 02:49:33PM +0100, Bryan O'Donoghue wrote:
>> On 06/07/2026 13:31, Rakesh Kota wrote:
>>> Remove pm8150b fallback compatible from pm4125_vbus and fix regulator
>>> constraints in qrb2210 DTS files to use microvolt instead of
>>> microamp.
>>>
>>> Signed-off-by: Rakesh Kota <rakesh.kota@xxxxxxxxxxxxxxxx>
>> You should break this patch up at the "and"
>>
>> One to remove the fallabck, one to fixup your regulator stuff.
>>
> I understand the preference for splitting at the "and", however
> splitting this patch is not straightforward in this case. The PM4125
> compatible and its regulator constraints are tightly coupled in the same
> .yaml schema file — if we split them, dtbs_check will fail on the
> intermediate state.
> 
> I've confirmed this locally. Removing the pm8150b fallback compatible in
> isolation causes the following schema validation failures:
> 
> /local/mnt/workspace/kernel_versions/kernel-checkers/linux-next/temp-out/arch/arm64/boot/dts/qcom/qrb2210-rb1-vision-mezzanine.dtb:
> usb-vbus-regulator@1100 (qcom,pm4125-vbus-reg): 'regulator-min-microamp' is a required property
> 	 from schema $id: http://devicetree.org/schemas/regulator/qcom,usb-vbus-regulator.yaml#
> /local/mnt/workspace/kernel_versions/kernel-checkers/linux-next/temp-out/arch/arm64/boot/dts/qcom/qrb2210-rb1-vision-mezzanine.dtb:
> usb-vbus-regulator@1100 (qcom,pm4125-vbus-reg): 'regulator-max-microamp' is a required property
> 	from schema $id: http://devicetree.org/schemas/regulator/qcom,usb-vbus-regulator.yaml#
> 
> This is because the qcom,pm4125-vbus-reg compatible requires
> regulator-min-microamp and regulator-max-microamp to be present per the
> schema, and both changes are needed together for a bisectable,
> build-clean series.
> 
> Could you please advise if you'd like me to proceed with the combined
> patch, or if there's a preferred way to structure this that avoids the
> intermediate dtbs_check breakage?

This change is fine as-is.

Konrad