On Fri, Jul 10, 2015 at 09:27:45PM -0400, Waiman Long wrote:
> Anyway, I have an alternative fix that should better capture the problem:
>
> -------------------------------
> diff --git a/kernel/locking/qspinlock_paravirt.h
> b/kernel/locking/qspinlock_paravirt.h
> index 04ab181..92fc54f 100644
> --- a/kernel/locking/qspinlock_paravirt.h
> +++ b/kernel/locking/qspinlock_paravirt.h
> @@ -286,15 +286,24 @@ __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)) {
> + printk(KERN_WARNING
> + "pvqspinlock: lock 0x%lx has corrupted value 0x%x!\n",
> + (unsigned long)lock, atomic_read(&lock->val));
> + WARN_ON_ONCE(1);
WARN_ONCE(1, "foo");
> + return;
> + }
Right, so since this should not ever happen in 'sane' code, its a shame
to have to put in this condition. But yes, this works too.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
