Donate to e Foundation | Murena handsets with /e/OS | Own a part of Murena! Learn more

Commit 5e5013c6 authored by Linus Torvalds's avatar Linus Torvalds
Browse files

Merge branch 'locking-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip 

Pull locking fix from Ingo Molnar:
 "A single fix for a locking self-test crash"

* 'locking-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  locking/pvqspinlock: Fix kernel panic in locking-selftest
parents c6797654 cba77f03

kernel/locking/qspinlock_paravirt.h

+10 −1
Original line number Diff line number Diff line
@@ -4,6 +4,7 @@ 


#include <linux/hash.h>

#include <linux/bootmem.h>

#include <linux/debug_locks.h>



/*

 * Implement paravirt qspinlocks; the general idea is to halt the vcpus instead
@@ -286,15 +287,23 @@ __visible void __pv_queued_spin_unlock(struct qspinlock *lock) 
{

	struct __qspinlock *l = (void *)lock;

	struct pv_node *node;

	u8 lockval = cmpxchg(&l->locked, _Q_LOCKED_VAL, 0);



	/*

	 * We must not unlock if SLOW, because in that case we must first

	 * unhash. Otherwise it would be possible to have multiple @lock

	 * entries, which would be BAD.

	 */

	if (likely(cmpxchg(&l->locked, _Q_LOCKED_VAL, 0) == _Q_LOCKED_VAL))

	if (likely(lockval == _Q_LOCKED_VAL))

		return;



	if (unlikely(lockval != _Q_SLOW_VAL)) {

		if (debug_locks_silent)

			return;

		WARN(1, "pvqspinlock: lock %p has corrupted value 0x%x!\n", lock, atomic_read(&lock->val));

		return;

	}



	/*

	 * Since the above failed to release, this must be the SLOW path.

	 * Therefore start by looking up the blocked node and unhashing it.