Re: [PATCH v3 2/3] arm64: dts: ti: Add k3-am62l3-beaglebadge

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 3 potential issue(s) to consider:
- [High] The `microchip,spi-present-mask` property is specified as an 8-bit value, which will cause the mcp23s18 driver probe to fail.
- [High] Missing `bootph-all` property on `&wkup_i2c0` and its PMIC child node `tps65214`.
- [Low] Unnecessary inclusion of DT binding header files.
--

commit 14331975158856c05618604ffa3aca84008b3237
Author: Judith Mendez <jm@ti.com>

arm64: dts: ti: Add k3-am62l3-beaglebadge

This commit adds the device tree for the BeagleBoard.org BeagleBadge,
a compact single board computer based on the TI AM62L3 SoC. It provides
initial support including GPIOs, LEDs, regulators, and other peripherals.

> diff --git a/arch/arm64/boot/dts/ti/k3-am62l3-beaglebadge.dts b/arch/arm64/boot/dts/ti/k3-am62l3-beaglebadge.dts
> new file mode 100644
> index 0000000000000..b7541e162a90d
> --- /dev/null
> +++ b/arch/arm64/boot/dts/ti/k3-am62l3-beaglebadge.dts
> @@ -0,0 +1,700 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * https://www.beagleboard.org/boards/beaglebadge
> + *
> + * Copyright (C) 2026 Texas Instruments Incorporated - https://www.ti.com/
> + */
> +
> +/dts-v1/;
> +
> +#include <dt-bindings/gpio/gpio.h>
> +#include <dt-bindings/input/input.h>
> +#include <dt-bindings/leds/common.h>
> +#include <dt-bindings/net/ti-dp83867.h>
> +#include <dt-bindings/thermal/thermal.h>

Are these last two includes necessary?

It doesn't look like there are any DP83867 Ethernet PHY nodes in this device
tree to use the networking macros.

Also, the thermal zone trip point uses a raw string literal for the type
rather than the standard macros defined in thermal.h.

[ ... ]

> +&wkup_i2c0 {
> +	clock-frequency = <400000>;
> +	pinctrl-names = "default";
> +	pinctrl-0 = <&wkup_i2c0_pins_default>;
> +	status = "okay";
> +
> +	tps65214: pmic@30 {
> +		compatible = "ti,tps65214";

Will U-Boot SPL drop the I2C bus and PMIC from its device tree, failing to
configure essential SoC voltages like VDD_CORE, VDD_1V8, and VDD_LPDDR4?

While the associated pinmux group wkup_i2c0_pins_default correctly specifies
bootph-all, both the wkup_i2c0 controller node and the PMIC child node do not.

Could this prevent the board from booting?

[ ... ]

> +&spi1 {
> +	pinctrl-names = "default";
> +	pinctrl-0 = <&spi1_pins_default>;
> +	status = "okay";
> +
> +	mcp23s18: gpio@0 {
> +		compatible = "microchip,mcp23s18";
> +		reg = <0>;
> +		spi-max-frequency = <10000000>;
> +		gpio-controller;
> +		#gpio-cells = <2>;
> +		reset-gpios = <&gpio0 25 GPIO_ACTIVE_LOW>;
> +		pinctrl-names = "default";
> +		pinctrl-0 = <&mcp23s18_reset_pins_default>;
> +		microchip,spi-present-mask = /bits/ 8 <0x01>;

Will the mcp23s18 driver probe fail here because of the 8-bit mask value?

The driver in drivers/pinctrl/pinctrl-mcp23s08_spi.c attempts to read this
using device_property_read_u32(), which expects a property of at least 4 bytes.

Reading a 1-byte property will cause the underlying
of_property_read_variable_u32_array() to return -EOVERFLOW, which could
completely disable the GPIO expander and its dependent peripherals.

> +	};
> +};

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260513233447.2713737-1-jm@ti.com?part=2