[PATCH -v3 09/10] locking/mutex: Enable optimistic spinning of woken waiter

[Peter Zijlstra <peterz@infradead.org>]

[-- Attachment #1: waiman_long-locking_mutex-enable_optimistic_spinning_of_woken_waiter.patch --]
[-- Type: text/plain, Size: 5089 bytes --]

This patch makes the waiter that sets the HANDOFF flag start spinning
instead of sleeping until the handoff is complete or the owner
sleeps. Otherwise, the handoff will cause the optimistic spinners to
abort spinning as the handed-off owner may not be running.

Cc: Ingo Molnar <mingo@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Imre Deak <imre.deak@intel.com>
Cc: Jason Low <jason.low2@hpe.com>
Cc: "Paul E. McKenney" <paulmck@us.ibm.com>
Cc: Ding Tianhong <dingtianhong@huawei.com>
Cc: Davidlohr Bueso <dave@stgolabs.net>
Cc: Will Deacon <Will.Deacon@arm.com>
Signed-off-by: Waiman Long <Waiman.Long@hpe.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1472254509-27508-2-git-send-email-Waiman.Long@hpe.com
---
 kernel/locking/mutex.c |   76 ++++++++++++++++++++++++++++++++++---------------
 1 file changed, 54 insertions(+), 22 deletions(-)

--- a/kernel/locking/mutex.c
+++ b/kernel/locking/mutex.c
@@ -409,24 +409,39 @@ static inline int mutex_can_spin_on_owne
  *
  * Returns true when the lock was taken, otherwise false, indicating
  * that we need to jump to the slowpath and sleep.
+ *
+ * The waiter flag is set to true if the spinner is a waiter in the wait
+ * queue. The waiter-spinner will spin on the lock directly and concurrently
+ * with the spinner at the head of the OSQ, if present, until the owner is
+ * changed to itself.
  */
 static bool mutex_optimistic_spin(struct mutex *lock,
-				  struct ww_acquire_ctx *ww_ctx, const bool use_ww_ctx)
+				  struct ww_acquire_ctx *ww_ctx,
+				  const bool use_ww_ctx, const bool waiter)
 {
 	struct task_struct *task = current;
 
-	if (!mutex_can_spin_on_owner(lock))
-		goto done;
+	if (!waiter) {
+		/*
+		 * The purpose of the mutex_can_spin_on_owner() function is
+		 * to eliminate the overhead of osq_lock() and osq_unlock()
+		 * in case spinning isn't possible. As a waiter-spinner
+		 * is not going to take OSQ lock anyway, there is no need
+		 * to call mutex_can_spin_on_owner().
+		 */
+		if (!mutex_can_spin_on_owner(lock))
+			goto fail;
 
-	/*
-	 * In order to avoid a stampede of mutex spinners trying to
-	 * acquire the mutex all at once, the spinners need to take a
-	 * MCS (queued) lock first before spinning on the owner field.
-	 */
-	if (!osq_lock(&lock->osq))
-		goto done;
+		/*
+		 * In order to avoid a stampede of mutex spinners trying to
+		 * acquire the mutex all at once, the spinners need to take a
+		 * MCS (queued) lock first before spinning on the owner field.
+		 */
+		if (!osq_lock(&lock->osq))
+			goto fail;
+	}
 
-	while (true) {
+	for (;;) {
 		struct task_struct *owner;
 
 		if (use_ww_ctx && ww_ctx->acquired > 0) {
@@ -442,7 +457,7 @@ static bool mutex_optimistic_spin(struct
 			 * performed the optimistic spinning cannot be done.
 			 */
 			if (READ_ONCE(ww->ctx))
-				break;
+				goto fail_unlock;
 		}
 
 		/*
@@ -450,15 +465,20 @@ static bool mutex_optimistic_spin(struct
 		 * release the lock or go to sleep.
 		 */
 		owner = __mutex_owner(lock);
-		if (owner && !mutex_spin_on_owner(lock, owner))
-			break;
+		if (owner) {
+			if (waiter && owner == task) {
+				smp_mb(); /* ACQUIRE */
+				break;
+			}
 
-		/* Try to acquire the mutex if it is unlocked. */
-		if (__mutex_trylock(lock, false)) {
-			osq_unlock(&lock->osq);
-			return true;
+			if (!mutex_spin_on_owner(lock, owner))
+				goto fail_unlock;
 		}
 
+		/* Try to acquire the mutex if it is unlocked. */
+		if (__mutex_trylock(lock, waiter))
+			break;
+
 		/*
 		 * The cpu_relax() call is a compiler barrier which forces
 		 * everything in this loop to be re-loaded. We don't need
@@ -468,8 +488,17 @@ static bool mutex_optimistic_spin(struct
 		cpu_relax_lowlatency();
 	}
 
-	osq_unlock(&lock->osq);
-done:
+	if (!waiter)
+		osq_unlock(&lock->osq);
+
+	return true;
+
+
+fail_unlock:
+	if (!waiter)
+		osq_unlock(&lock->osq);
+
+fail:
 	/*
 	 * If we fell out of the spin path because of need_resched(),
 	 * reschedule now, before we try-lock the mutex. This avoids getting
@@ -488,7 +517,8 @@ static bool mutex_optimistic_spin(struct
 }
 #else
 static bool mutex_optimistic_spin(struct mutex *lock,
-				  struct ww_acquire_ctx *ww_ctx, const bool use_ww_ctx)
+				  struct ww_acquire_ctx *ww_ctx,
+				  const bool use_ww_ctx, const bool waiter)
 {
 	return false;
 }
@@ -593,7 +623,7 @@ __mutex_lock_common(struct mutex *lock,
 	mutex_acquire_nest(&lock->dep_map, subclass, 0, nest_lock, ip);
 
 	if (__mutex_trylock(lock, false) ||
-	    mutex_optimistic_spin(lock, ww_ctx, use_ww_ctx)) {
+	    mutex_optimistic_spin(lock, ww_ctx, use_ww_ctx, false)) {
 		/* got the lock, yay! */
 		lock_acquired(&lock->dep_map, ip);
 		if (use_ww_ctx)
@@ -648,6 +678,8 @@ __mutex_lock_common(struct mutex *lock,
 		if (__mutex_waiter_is_first(lock, &waiter)) {
 			first = true;
 			__mutex_set_flag(lock, MUTEX_FLAG_HANDOFF);
+			if (mutex_optimistic_spin(lock, ww_ctx, use_ww_ctx, true))
+				break;
 		}
 
 		set_task_state(task, state);