Re: [PATCH v2 4/4] arm64: dts: qcom: hamoa-pmics: Add ADC support

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [Medium] Missing Thermal Sensor Provider for VADC causing thermal zones to indefinitely defer
- [Low] Devicetree Naming Violation: Underscore in node name
--

commit 7b5021a0220835eeee60f0b85faf679ed9b09977
Author: Ayyagari Ushasreevalli <aushasre@qti.qualcomm.com>

arm64: dts: qcom: hamoa-pmics: Add ADC support

This commit adds the ADC node and defines channels for die temperature,
VPH power, and system thermistors for PMK8550, PM8550VE*, and PMC8380*
PMICs. It also defines several thermal zones corresponding to the off-PMIC
system thermistors.

> diff --git a/arch/arm64/boot/dts/qcom/hamoa-pmics.dtsi b/arch/arm64/boot/dts/qcom/hamoa-pmics.dtsi
> index 6a31a0adf8be4..2e746ede850f8 100644
> --- a/arch/arm64/boot/dts/qcom/hamoa-pmics.dtsi
> +++ b/arch/arm64/boot/dts/qcom/hamoa-pmics.dtsi
[ ... ]
> @@ -189,6 +191,90 @@ trip1 {
>  				};
>  			};
>  		};
> +
> +		sys-0-thermal {
> +			polling-delay-passive = <0>;
> +			thermal-sensors = <&pmk8550_vadc ADC5_GEN3_AMUX1_GPIO_100K_PU(1)>;
[ ... ]
> @@ -277,6 +363,142 @@ pmk8550_pwm: pwm {
>  
>  			status = "disabled";
>  		};
> +
> +		pmk8550_vadc: adc@9000 {
> +			compatible = "qcom,spmi-adc5-gen3";
> +			reg = <0x9000>, <0x9100>;
> +			#address-cells = <1>;
> +			#size-cells = <0>;
> +			interrupts = <0x0 0x90 0x1 IRQ_TYPE_EDGE_RISING>,
> +				     <0x0 0x91 0x1 IRQ_TYPE_EDGE_RISING>;
> +			#thermal-sensor-cells = <1>;

[Severity: Medium]
Does the driver for pmk8550_vadc (qcom-spmi-adc5-gen3.c) actually register a
thermal zone provider?

It looks like the driver currently registers IIO channels but delegates
thermal monitor capabilities to an auxiliary device named adc5_tm_gen3,
which currently has no driver in the kernel to bind to it.

Without a driver registering the thermal zones via
devm_thermal_of_zone_register(), will the generic thermal framework
indefinitely defer probing for these thermal zones, leaving the system
thermistor monitoring uninitialized?

[ ... ]
> @@ -301,6 +525,18 @@ pm8550_gpios: gpio@8800 {
>  			#gpio-cells = <2>;
>  			interrupt-controller;
>  			#interrupt-cells = <2>;
> +
> +			sys_therm_0_gpio3: sys-therm_0-gpio3-state {

[Severity: Low]
This isn't a bug, but does this node name violate devicetree conventions?

Devicetree specifications strongly discourage underscores in node names,
recommending hyphens instead. While the node label (sys_therm_0_gpio3)
can use underscores, should the node name be sys-therm-0-gpio3-state?

The sibling node sys-therm-1-gpio4-state correctly uses hyphens. Leaving
the underscore might trigger a node_name_chars warning in dtc when compiled
with W=1.

> +				pins = "gpio3";
> +				function = PMIC_GPIO_FUNC_NORMAL;
> +				bias-high-impedance;
> +			};

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260614-adc5_gen3_dt-v2-0-32ec576c5865@oss.qualcomm.com?part=4