Re: [PATCH] usb: typec: fix unreleased fwnode_handle in typec_port_register_altmodes()

[Heikki Krogerus]

On Mon, Oct 21, 2024 at 04:06:30PM +0200, Javier Carrasco wrote:
> On 21/10/2024 15:42, Heikki Krogerus wrote:
> > Hi,
> > 
> > On Mon, Oct 21, 2024 at 03:55:43PM +0300, Heikki Krogerus wrote:
> >> On Sat, Oct 19, 2024 at 10:40:19PM +0200, Javier Carrasco wrote:
> >>> The 'altmodes_node' fwnode_handle is never released after it is no
> >>> longer required, which leaks the resource.
> >>>
> >>> Add the required call to fwnode_handle_put() when 'altmodes_node' is no
> >>> longer required.
> >>>
> >>> Cc: stable@xxxxxxxxxxxxxxx
> >>> Fixes: 7b458a4c5d73 ("usb: typec: Add typec_port_register_altmodes()")
> >>> Signed-off-by: Javier Carrasco <javier.carrasco.cruz@xxxxxxxxx>
> >>
> >> Reviewed-by: Heikki Krogerus <heikki.krogerus@xxxxxxxxxxxxxxx>
> >>
> >>> ---
> >>>  drivers/usb/typec/class.c | 1 +
> >>>  1 file changed, 1 insertion(+)
> >>>
> >>> diff --git a/drivers/usb/typec/class.c b/drivers/usb/typec/class.c
> >>> index d61b4c74648d..1eb240604cf6 100644
> >>> --- a/drivers/usb/typec/class.c
> >>> +++ b/drivers/usb/typec/class.c
> >>> @@ -2341,6 +2341,7 @@ void typec_port_register_altmodes(struct typec_port *port,
> >>>  		altmodes[index] = alt;
> >>>  		index++;
> >>>  	}
> >>> +	fwnode_handle_put(altmodes_node);
> >>>  }
> >>>  EXPORT_SYMBOL_GPL(typec_port_register_altmodes);
> > 
> > Sorry to go back to this, but I guess we should actually use those
> > scope based helpers with fwnodes in this case. So instead of a
> > dedicated fwnode_handle_put() call like that, just introduce
> > altmodes_node like this:
> > 
> >         ...
> >         struct fwnode_handle *altmodes_node __free(fwnode_handle) =
> >                 device_get_named_child_node(&port->dev, "altmodes");
> > 
> >         if (IS_ERR(altmodes_node))
> >                 return;
> > 
> >         fwnode_for_each_child_node(altmodes_node, child) {
> >         ...
> > 
> > thanks,
> > 
> 
> That would have to be a second patch, because it does not apply to all
> affected stable kernels. I can send it separately, though.

Great, thanks!

-- 
heikki