Re: [PATCH v2 2/2] arm64: dts: rockchip: Add EmbedFire LubanCat 4 board

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] The PWM fan period is configured to 5000 ns (200 kHz), which is likely a typo for 50000 ns (20 kHz).
- [Medium] The PWM fan is not bound to any SoC thermal zones, rendering automatic thermal management non-functional.
--

--- Patch [2]: [PATCH v2 2/2] arm64: dts: rockchip: Add EmbedFire LubanCat 4 board ---
commit 1f3302a5304b9bff69dfd891bf8f0a9d2c32995c
Author: Pufan Jin <2254650260@qq.com>

arm64: dts: rockchip: Add EmbedFire LubanCat 4 board

Add device tree for the EmbedFire LubanCat 4 single board computer,
featuring the Rockchip RK3588S SoC.

> diff --git a/arch/arm64/boot/dts/rockchip/rk3588s-lubancat-4.dts b/arch/arm64/boot/dts/rockchip/rk3588s-lubancat-4.dts
> new file mode 100644
> index 0000000000000..0eae4b7d85e54
> --- /dev/null
> +++ b/arch/arm64/boot/dts/rockchip/rk3588s-lubancat-4.dts
[ ... ]
> +	fan: pwm-fan {
> +		compatible = "pwm-fan";
> +		cooling-levels = <0 100 150 200 255>;
> +		pwms = <&pwm0 0 5000 0>;

[Severity: Medium]
Is the PWM period here intended to be 5000 ns (200 kHz)?

This is far above the standard 21-28 kHz range specified for 4-wire cooling
fans, which can cause the fan's internal controller to misinterpret the signal.
This can result in the fan either running at 100% speed continuously or
failing to spin at all.

Could this be a typo for 50000 ns (20 kHz)?

[Severity: Medium]
Does this node also need to be bound to the SoC's thermal zones via a
&package_thermal override?

The fan defines cooling-levels and #cooling-cells, but without being bound to
any SoC thermal zones, the kernel's thermal subsystem will not automatically
adjust the fan speed when the SoC temperature rises.

> +		#cooling-cells = <2>;
> +	};

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260705135014.1004166-1-2254650260@qq.com?part=2