[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAMpxmJX5673AmGDwrb=DMUu7=8Xi2VTtWE72F2hgitK9QUt-RA@mail.gmail.com>
Date: Thu, 20 Feb 2020 13:51:18 +0100
From: Bartosz Golaszewski <bgolaszewski@...libre.com>
To: Srinivas Kandagatla <srinivas.kandagatla@...aro.org>
Cc: Bartosz Golaszewski <brgl@...ev.pl>,
Linus Walleij <linus.walleij@...aro.org>,
Khouloud Touil <ktouil@...libre.com>,
Geert Uytterhoeven <geert@...ux-m68k.org>,
linux-gpio <linux-gpio@...r.kernel.org>,
LKML <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v4 2/4] gpiolib: use kref in gpio_desc
czw., 20 lut 2020 o 13:05 Srinivas Kandagatla
<srinivas.kandagatla@...aro.org> napisaĆ(a):
>
>
>
> On 20/02/2020 10:01, Bartosz Golaszewski wrote:
> > --- a/drivers/gpio/gpiolib.c
> > +++ b/drivers/gpio/gpiolib.c
> > @@ -2798,6 +2798,8 @@ static int gpiod_request_commit(struct gpio_desc *desc, const char *label)
> > goto done;
> > }
> >
> > + kref_init(&desc->ref);
> > +
>
> Should we not decrement refcount on the error path of this function?
>
On error the descriptor will still be unrequested so there's no point
in potentially calling gpiod_free(). Also: the next time someone
requests it and succeeds, we'll set it back to 1.
Bartosz
Powered by blists - more mailing lists