Re: [PATCH] ARM: dts: aspeed: sanmiguel: Remove HSC nodes

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] Removing actual hardware from the Device Tree to bypass kernel probing for userspace preference violates DT principles.
--

--- Patch [1]: [PATCH] ARM: dts: aspeed: sanmiguel: Remove HSC nodes ---
commit 4baedd7709172e87dfefc81a46cefd1b968b6e6a
Author: Potin Lai <potin.lai.pt@gmail.com>

ARM: dts: aspeed: sanmiguel: Remove HSC nodes

This commit removes Hot-Swap Controller power monitor nodes from the device
tree. The commit message states this is done to prevent automatic kernel
probing and allow userspace to handle the device instantiation.

> diff --git a/arch/arm/boot/dts/aspeed/aspeed-bmc-facebook-sanmiguel.dts b/arch/arm/boot/dts/aspeed/aspeed-bmc-facebook-sanmiguel.dts
> index 3cc6e772cc8c4..98f224d2c091f 100644
> --- a/arch/arm/boot/dts/aspeed/aspeed-bmc-facebook-sanmiguel.dts
> +++ b/arch/arm/boot/dts/aspeed/aspeed-bmc-facebook-sanmiguel.dts
> @@ -442,50 +442,6 @@ &i2c8 {
>  &i2c9 {
>  	status = "okay";
>  
> -	pdb_mps_hsc1: power-monitor@10 {
> -		compatible = "mps,mp5926";
> -		reg = <0x10>;
> -	};

[Severity: Medium]
Does removing these nodes to accommodate userspace policy conflict with the
principle that the device tree should be an OS-agnostic description of the
physical hardware?

Since the device tree is meant to describe the physical hardware rather than
dictate software configuration, removing physically present devices to prevent
kernel probing is typically avoided. If these components are on hot-pluggable
boards or are conditionally populated, would it be more appropriate to use
device tree overlays, or to mark the nodes with status = "disabled" instead
of deleting the hardware description outright?

[ ... ]

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260730-upstream-sanmiguel-remove-hsc-nodes-v1-1-f0308eb8a00c@gmail.com?part=1