Re: [PATCH 3/7] arm64: dts: qcom: shikra-iqs-evk: Add AUO G133HAN01 LVDS panel overlay

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] arm64: dts: qcom: missing status okay on i2c3 node in lvds overlay
- [Medium] arm64: dts: qcom: missing backlight property in panel-lvds node
--

commit b4b78d829d507a91f08f278892b57ecc12f421d2
Author: Anand Tiwari <anand.tiwari@oss.qualcomm.com>

arm64: dts: qcom: shikra-iqs-evk: Add AUO G133HAN01 LVDS panel overlay

This commit adds a device tree overlay for shikra-iqs-evk to support the
auo,g133han01 panel connected through a ti,sn65dsi84 bridge on i2c3, while
disabling the default hdmi bridge and connector.

> diff --git a/arch/arm64/boot/dts/qcom/shikra-iqs-evk-lvds-auo,g133han01.dtso b/arch/arm64/boot/dts/qcom/shikra-iqs-evk-lvds-auo,g133han01.dtso
> new file mode 100644
> index 0000000000000..80b9dcc920612
> --- /dev/null
> +++ b/arch/arm64/boot/dts/qcom/shikra-iqs-evk-lvds-auo,g133han01.dtso
> @@ -0,0 +1,127 @@

[ ... ]

> +&{/} {
> +	backlight {
> +		compatible = "gpio-backlight";
> +		gpios = <&pm8150_gpios 1 GPIO_ACTIVE_HIGH>;
> +		default-on;
> +		pinctrl-0 = <&backlight_pwm>;
> +		pinctrl-names = "default";
> +	};
> +
> +	panel-lvds {
> +		compatible = "auo,g133han01";
> +		power-supply = <&vreg_lvds_3p3>;

[Severity: Medium]
In arch/arm64/boot/dts/qcom/shikra-iqs-evk-lvds-auo,g133han01.dtso, the
backlight node is defined without a phandle label, and panel-lvds does not
have a backlight property referencing it.

When panel_simple_probe() in drivers/gpu/drm/panel/panel-simple.c probes
auo,g133han01, it looks up the panel backlight via drm_panel_of_backlight():

	err = drm_panel_of_backlight(&panel->base);
	if (err) {
		dev_err_probe(dev, err, "Could not find backlight\n");
		goto disable_pm_runtime;
	}

Because the backlight property is missing on panel-lvds,
panel->base.backlight remains unset, so drm_panel_enable() and
drm_panel_disable() will not control the gpio-backlight device during
display blanking or system suspend and resume.

Should the backlight node be given a label and referenced from panel-lvds
with backlight = <&backlight>;?

[ ... ]

> +&i2c3 {
> +	#address-cells = <1>;
> +	#size-cells = <0>;
> +
> +	bridge@2c {
> +		compatible = "ti,sn65dsi84";
> +		reg = <0x2c>;

[Severity: High]
In arch/arm64/boot/dts/qcom/shikra.dtsi, i2c3 is disabled by default:

	i2c3: i2c@4a8c000 {
		compatible = "qcom,geni-i2c";
		...
		status = "disabled";
	};

Neither shikra-iqs-som.dtsi, shikra-evk.dtsi, nor shikra-iqs-evk.dts
enables i2c3. In
arch/arm64/boot/dts/qcom/shikra-iqs-evk-lvds-auo,g133han01.dtso, &i2c3
populates the bridge@2c child node and &mdss_dsi0_out is redirected to
&sn65dsi84_in, but &i2c3 does not set status = "okay";.

When this overlay is applied, i2c3 stays disabled, the ti,sn65dsi84 bridge
at bridge@2c is never instantiated, and
msm_dsi_modeset_init() -> drm_of_find_panel_or_bridge() defers probe
indefinitely waiting for sn65dsi84_in.

Does &i2c3 need status = "okay"; here so that the parent controller and
bridge@2c can probe?

[ ... ]

> +&mdss_dsi0_out {
> +	remote-endpoint = <&sn65dsi84_in>;
> +};

[ ... ]

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261021-shikra_display_overlays-v1-0-adaea735999f@oss.qualcomm.com?part=3