Re: [PATCH net-next v2 3/3] net: dsa: realtek: add LED drivers for rtl8366rb

[Linus Walleij]

On Thu, May 16, 2024 at 11:05 PM Linus Walleij <linus.walleij@xxxxxxxxxx> wrote:
> On Thu, May 16, 2024 at 7:30 PM Luiz Angelo Daros de Luca
> <luizluca@xxxxxxxxx> wrote:
> > > On Sat, 27 Apr 2024 02:11:30 -0300 Luiz Angelo Daros de Luca wrote:
> > > > +static int rtl8366rb_setup_leds(struct realtek_priv *priv)
> > > > +{
> > > > +     struct device_node *leds_np, *led_np;
> > > > +     struct dsa_switch *ds = &priv->ds;
> > > > +     struct dsa_port *dp;
> > > > +     int ret = 0;
> > > > +
> > > > +     dsa_switch_for_each_port(dp, ds) {
> > > > +             if (!dp->dn)
> > > > +                     continue;
> > > > +
> > > > +             leds_np = of_get_child_by_name(dp->dn, "leds");
> > > > +             if (!leds_np) {
> > > > +                     dev_dbg(priv->dev, "No leds defined for port %d",
> > > > +                             dp->index);
> > > > +                     continue;
> > > > +             }
> > > > +
> > > > +             for_each_child_of_node(leds_np, led_np) {
> > > > +                     ret = rtl8366rb_setup_led(priv, dp,
> > > > +                                               of_fwnode_handle(led_np));
> > > > +                     if (ret) {
> > > > +                             of_node_put(led_np);
> > > > +                             break;
> > > > +                     }
> > > > +             }
> > > > +
> > > > +             of_node_put(leds_np);
> > > > +             if (ret)
> > > > +                     return ret;
> > > > +     }
> > > > +     return 0;
> > > > +}
> > >
> > > coccicheck generates this warning:
> > >
> > > drivers/net/dsa/realtek/rtl8366rb.c:1032:4-15: ERROR: probable double put.
> > >
> > > I think it's a false positive.
> >
> > Me too. I don't think it is a double put. The put for led_np is called
> > in the increment code inside the for_each_child_of_node macro. With a
> > break, we skip that part and we need to put it before leaving. I don't
> > know coccicheck but maybe it got confused by the double for.
>
> Maybe I can use for_each_child_of_node_scoped() and
> get the handling for free? The checkers should learn about
> *_scoped now.
>
> (I'm still working on the patch, I'm just slow.)

Referring to my Marvell LED patch, confusing it for this one,
damn I need to get to sleep :D

Luiz: try to use for_each_child_of_node_scoped()!

Yours,
Linus Walleij