releases/4.4.53/w1-don-t-leak-refcount-on-slave-attach-failure-in-w1_attach_slave_device.patch

   1 From d2ce4ea1a0b0162e5d2d7e7942ab6f5cc2063d5a Mon Sep 17 00:00:00 2001
   2 From: "Maciej S. Szmigiero" <mail@maciej.szmigiero.name>
   3 Date: Sat, 21 Jan 2017 23:50:18 +0100
   4 Subject: w1: don't leak refcount on slave attach failure in w1_attach_slave_device()
   5
   6 From: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
   7
   8 commit d2ce4ea1a0b0162e5d2d7e7942ab6f5cc2063d5a upstream.
   9
  10 Near the beginning of w1_attach_slave_device() we increment a w1 master
  11 reference count.
  12 Later, when we are going to exit this function without actually attaching
  13 a slave device (due to failure of __w1_attach_slave_device()) we need to
  14 decrement this reference count back.
  15
  16 Signed-off-by: Maciej S. Szmigiero <mail@maciej.szmigiero.name>
  17 Fixes: 9fcbbac5ded489 ("w1: process w1 netlink commands in w1_process thread")
  18 Cc: Evgeniy Polyakov <zbr@ioremap.net>
  19 Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
  20
  21 ---
  22  drivers/w1/w1.c |    1 +
  23  1 file changed, 1 insertion(+)
  24
  25 --- a/drivers/w1/w1.c
  26 +++ b/drivers/w1/w1.c
  27 @@ -763,6 +763,7 @@ int w1_attach_slave_device(struct w1_mas
  28                 dev_err(&dev->dev, "%s: Attaching %s failed.\n", __func__,
  29                          sl->name);
  30                 w1_family_put(sl->family);
  31 +               atomic_dec(&sl->master->refcnt);
  32                 kfree(sl);
  33                 return err;
  34         }