From: Thomas Gleixner <[email protected]> 3.12-stable review patch. If anyone has any objections, please let me know.
=============== commit fb75a4282d0d9a3c7c44d940582c2d226cf3acfb upstream. If the proxy lock in the requeue loop acquires the rtmutex for a waiter then it acquired also refcount on the pi_state related to the futex, but the waiter side does not drop the reference count. Add the missing free_pi_state() call. Signed-off-by: Thomas Gleixner <[email protected]> Cc: Peter Zijlstra <[email protected]> Cc: Darren Hart <[email protected]> Cc: Davidlohr Bueso <[email protected]> Cc: [email protected] Cc: Andy Lowe <[email protected]> Link: http://lkml.kernel.org/r/[email protected] Signed-off-by: Thomas Gleixner <[email protected]> Signed-off-by: Jiri Slaby <[email protected]> --- kernel/futex.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/kernel/futex.c b/kernel/futex.c index 3ee1b3ce78df..509bdd404414 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -2497,6 +2497,11 @@ static int futex_wait_requeue_pi(u32 __user *uaddr, unsigned int flags, if (q.pi_state && (q.pi_state->owner != current)) { spin_lock(q.lock_ptr); ret = fixup_pi_state_owner(uaddr2, &q, current); + /* + * Drop the reference to the pi state which + * the requeue_pi() code acquired for us. + */ + free_pi_state(q.pi_state); spin_unlock(q.lock_ptr); } } else { -- 2.7.1
