Re: [PATCH] arm64: dts: imx8mn: Add display DT overlays for the ifm VHIP4 EvalBoard

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] The 1.8V regulator `reg_adv_1v8` is incorrectly configured to output 3.3V, violating hardware requirements for the ADV7535.
- [Medium] The `hdmi-transmitter@3d` node (ADV7535) is missing `clocks` and `clock-names` properties to link the defined `cec_clock`, breaking CEC support.
--

commit fdacb0c2eecf5c57dea7a716eb4e6d3417011685
Author: Fedor Ross <fedor.ross@ifm.com>

arm64: dts: imx8mn: Add display DT overlays for the ifm VHIP4 EvalBoard

This commit adds display DT overlays for the ifm VHIP4 EvalBoard v1, adding
support for the Displaytech DT050BTFT-PTS panel and the NXP ADV7533 DSI-to-HDMI
adapter. It supports different DSI lane counts to allow testing multiple
configurations.

> diff --git a/arch/arm64/boot/dts/freescale/imx8mn-vhip4-evalboard-v1-overlay-nxp-adv7533-hdmi.dtso b/arch/arm64/boot/dts/freescale/imx8mn-vhip4-evalboard-v1-overlay-nxp-adv7533-hdmi.dtso
> new file mode 100644
> index 0000000000000..e1a7aa6806c48
> --- /dev/null
> +++ b/arch/arm64/boot/dts/freescale/imx8mn-vhip4-evalboard-v1-overlay-nxp-adv7533-hdmi.dtso
[ ... ]
> +	reg_adv_1v8: regulator-adv-1v8 {
> +		compatible = "regulator-fixed";
> +		regulator-always-on;
> +		regulator-min-microvolt = <3300000>;
> +		regulator-max-microvolt = <3300000>;
                                           ^^^^^^^

[Severity: Medium]
Is the voltage configuration correct here? The reg_adv_1v8 node appears to
represent a 1.8V power supply used by the ADV7535, but its voltage limits
are set to 3.3V. This misinforms the regulator core about the hardware
voltage.

> +		regulator-name = "ADV-1V8";
> +	};
[ ... ]
> +&i2c3 {
> +	#address-cells = <1>;
> +	#size-cells = <0>;
> +	status = "okay";
> +
> +	hdmi-transmitter@3d {
> +		compatible = "adi,adv7535";
> +		reg = <0x3d>, <0x3e>, <0x3c>, <0x3f>;
> +		reg-names = "main", "edid", "cec", "packet";
> +		a2vdd-supply = <&reg_adv_1v8>;
> +		adi,dsi-lanes = <4>;
                 ^^^^^^^^^^^^^^^^^^^

[Severity: Medium]
Is there a missing clock reference for the CEC block? A cec_clock is defined
earlier in the file but is not linked here.

Without clocks and clock-names properties linking to cec_clock, the DRM bridge
core will fail to find it when calling devm_clk_get(dev, "cec"), which returns
-ENOENT. This causes adv7511_cec_init() to abort, leaving CEC uninitialized.

> +		avdd-supply = <&reg_adv_1v8>;
> +		dvdd-supply = <&reg_adv_1v8>;
> +		pvdd-supply = <&reg_adv_1v8>;
> +		v3p3-supply = <&reg_adv_3v3>;
> +		status = "okay";

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260908215409.32993-1-marex@nabladev.com?part=1