Skip to contentSkip to conversation

scheduler/locking updates for 4.2

GIT PULL Thomas Gleixner 1 participant 1 message Jun 23, 2015 Locking primitives

lore
Share & tools
lore
#0 lore raw

Linus,

please pull the latest sched-locking-for-linus git tree from:

git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip.git sched-locking-for-linus

These locking updates depend on the alreay merged sched/core branch:

- Lockless top waiter wakeup for rtmutex (Davidlohr)

- Reduce hash bucket lock contention for PI futexes (Sebastian)

- Documentation update (Davidlohr)

Thanks,

tglx

------------------>
Davidlohr Bueso (2):
locking/rtmutex: Implement lockless top-waiter wakeup locking/rtmutex: Update stale plist comments

Sebastian Andrzej Siewior (1):
futex: Lower the lock contention on the HB lock during wake up

 kernel/futex.c                  | 32 +++++++++++++--
 kernel/locking/rtmutex.c        | 87 ++++++++++++++++++++++++++---------------
 kernel/locking/rtmutex_common.h |  3 ++
 3 files changed, 88 insertions(+), 34 deletions(-)
kernel/futex.c +29 -3
index f9984c363e9a..a0cf6fa953de 100644
@@ -1117,11 +1117,14 @@ static void mark_wake_futex(struct wake_q_head *wake_q, struct futex_q *q)
 	q->lock_ptr = NULL;
 }
 
-static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this)
+static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this,
+			 struct futex_hash_bucket *hb)
 {
 	struct task_struct *new_owner;
 	struct futex_pi_state *pi_state = this->pi_state;
 	u32 uninitialized_var(curval), newval;
+	WAKE_Q(wake_q);
+	bool deboost;
 	int ret = 0;
 
 	if (!pi_state)
@@ -1173,7 +1176,19 @@ static int wake_futex_pi(u32 __user *uaddr, u32 uval, struct futex_q *this)
 	raw_spin_unlock_irq(&new_owner->pi_lock);
 
 	raw_spin_unlock(&pi_state->pi_mutex.wait_lock);
-	rt_mutex_unlock(&pi_state->pi_mutex);
+
+	deboost = rt_mutex_futex_unlock(&pi_state->pi_mutex, &wake_q);
+
+	/*
+	 * First unlock HB so the waiter does not spin on it once he got woken
+	 * up. Second wake up the waiter before the priority is adjusted. If we
+	 * deboost first (and lose our higher priority), then the task might get
+	 * scheduled away before the wake up can take place.
+	 */
+	spin_unlock(&hb->lock);
+	wake_up_q(&wake_q);
+	if (deboost)
+		rt_mutex_adjust_prio(current);
 
 	return 0;
 }
@@ -2413,13 +2428,23 @@ retry:
 	 */
 	match = futex_top_waiter(hb, &key);
 	if (match) {
-		ret = wake_futex_pi(uaddr, uval, match);
+		ret = wake_futex_pi(uaddr, uval, match, hb);
+		/*
+		 * In case of success wake_futex_pi dropped the hash
+		 * bucket lock.
+		 */
+		if (!ret)
+			goto out_putkey;
 		/*
 		 * The atomic access to the futex value generated a
 		 * pagefault, so retry the user-access and the wakeup:
 		 */
 		if (ret == -EFAULT)
 			goto pi_faulted;
+		/*
+		 * wake_futex_pi has detected invalid state. Tell user
+		 * space.
+		 */
 		goto out_unlock;
 	}
 
@@ -2440,6 +2465,7 @@ retry:
 
 out_unlock:
 	spin_unlock(&hb->lock);
+out_putkey:
 	put_futex_key(&key);
 	return ret;
 
kernel/locking/rtmutex.c +56 -31

193 lines not on this page · raw message

kernel/locking/rtmutex_common.h +3 -0
index 855212501407..7844f8f0e639 100644
@@ -131,6 +131,9 @@ extern int rt_mutex_finish_proxy_lock(struct rt_mutex *lock,
 				      struct hrtimer_sleeper *to,
 				      struct rt_mutex_waiter *waiter);
 extern int rt_mutex_timed_futex_lock(struct rt_mutex *l, struct hrtimer_sleeper *to);
+extern bool rt_mutex_futex_unlock(struct rt_mutex *lock,
+				  struct wake_q_head *wqh);
+extern void rt_mutex_adjust_prio(struct task_struct *task);
 
 #ifdef CONFIG_DEBUG_RT_MUTEXES
 # include "rtmutex-debug.h"