linux-kernel - Re: [PATCH] tcpm: allow looking for role

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-Id: <20251215095552.3240340-1-dsimic@manjaro.org>
Date: Mon, 15 Dec 2025 10:55:52 +0100
From: Dragan Simic <dsimic@...jaro.org>
To: arnaud.ferraris@...labora.com
Cc: badhri@...gle.com,
	gregkh@...uxfoundation.org,
	heikki.krogerus@...ux.intel.com,
	linux-kernel@...r.kernel.org,
	linux-usb@...r.kernel.org,
	m.grzeschik@...gutronix.de,
	stable@...r.kernel.org
Subject: Re: [PATCH] tcpm: allow looking for role_sw device in the main node

Hello Arnaud,

Thanks for this patch!  Please, see some comments below.

On Thu, 27 Nov 2025 15:04:15 +0100, Arnaud Ferraris <arnaud.ferraris@...labora.com> wrote:
> When ports are defined in the tcpc main node, fwnode_usb_role_switch_get
> returns an error, meaning usb_role_switch_get (which would succeed)
> never gets a chance to run as port->role_sw isn't NULL, causing a
> regression on devices where this is the case.
> 
> Fix this by turning the NULL check into IS_ERR_OR_NULL, so
> usb_role_switch_get can actually run and the device get properly probed.

It's usual to denote functions by always appending a pair of braces
to their names, so we'd have "fwnode_usb_role_switch_get()" in the
patch description above, for example.

> Fixes: 2d8713f807a4 ("tcpm: switch check for role_sw device with fw_node")
> Cc: stable@...r.kernel.org
> Signed-off-by: Arnaud Ferraris <arnaud.ferraris@...labora.com>
> ---
>  drivers/usb/typec/tcpm/tcpm.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/usb/typec/tcpm/tcpm.c b/drivers/usb/typec/tcpm/tcpm.c
> index cc78770509dbc..37698204d48d2 100644
> --- a/drivers/usb/typec/tcpm/tcpm.c
> +++ b/drivers/usb/typec/tcpm/tcpm.c
> @@ -7877,7 +7877,7 @@ struct tcpm_port *tcpm_register_port(struct device *dev, struct tcpc_dev *tcpc)
>  	port->partner_desc.identity = &port->partner_ident;
>  
>  	port->role_sw = fwnode_usb_role_switch_get(tcpc->fwnode);
> -	if (!port->role_sw)
> +	if (IS_ERR_OR_NULL(port->role_sw))
>  		port->role_sw = usb_role_switch_get(port->dev);
>  	if (IS_ERR(port->role_sw)) {
>  		err = PTR_ERR(port->role_sw);

This is looking good to me.  The usb_role_switch_is_parent() function,
invoked by fwnode_usb_role_switch_get(), can return -EPROBE_DEFER, so
using IS_ERR_OR_NULL() is the way to go.  It's already used in the
fwnode_usb_role_switch_get() function itself for checking against error
conditions, which solidifies this as a proper regression fix.

With the above-mentioned nitpicks addressed, please feel free to include

Reviewed-by: Dragan Simic <dsimic@...jaro.org>