Re: [PATCH net-next v3 09/13] net: dsa: rzn1-a5psw: Get switch info via OF data

[sashiko-bot@kernel.org]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] net: dsa: rzn1-a5psw: missing NULL check for of_device_get_match_data
--

commit d0e877d4584f59434a728598a2547f8f5e97b982
Author: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com>

net: dsa: rzn1-a5psw: Get switch info via OF data

This commit introduces an a5psw_of_data structure to hold per-SoC switch 
information, replacing previously embedded constants to support newer hardware 
variants with different port counts.

> diff --git a/drivers/net/dsa/rzn1_a5psw.c b/drivers/net/dsa/rzn1_a5psw.c
> index bdfa84745eda4..c8197ef62fd2d 100644
> --- a/drivers/net/dsa/rzn1_a5psw.c
> +++ b/drivers/net/dsa/rzn1_a5psw.c
[ ... ]
> @@ -1226,7 +1228,8 @@ static int a5psw_probe(struct platform_device *pdev)
>  	if (IS_ERR(a5psw->base))
>  		return PTR_ERR(a5psw->base);
>  
> -	a5psw->bridged_ports = BIT(A5PSW_CPU_PORT);
> +	a5psw->of_data = of_device_get_match_data(dev);
> +	a5psw->bridged_ports = BIT(a5psw->of_data->cpu_port);

[Severity: High]
Does this code dereference a NULL pointer if the driver is bound via sysfs?

If a privileged user manually binds the driver to a platform device via sysfs 
(e.g., using /sys/bus/platform/drivers/rzn1_a5psw/bind) that lacks a device 
tree node (dev->of_node is NULL), of_device_get_match_data() will return NULL.

Since a5psw->of_data is dereferenced immediately on the next line in 
a5psw_probe() without any validation, this could lead to a kernel panic during 
probe. Could a check be added here to safely handle a missing OF match?

>  
>  	ret = a5psw_pcs_get(a5psw);
>  	if (ret)

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261014131906.990077-1-prabhakar.mahadev-lad.rj@bp.renesas.com?part=9