Re: [RFC][PATCH v2 5/5] mutex: Give spinners a chance tospin_on_owner if need_resched() triggered while queued

From: Peter Zijlstra
Date: Sun Feb 02 2014 - 16:13:22 EST


On Sun, Feb 02, 2014 at 01:01:23PM -0800, Jason Low wrote:
> On Fri, 2014-01-31 at 21:08 +0100, Peter Zijlstra wrote:
> > On Fri, Jan 31, 2014 at 12:01:37PM -0800, Jason Low wrote:
> > > Currently still getting soft lockups with the updated version.
> >
> > Bugger.. ok clearly I need to think harder still. I'm fairly sure this
> > cancelation can work though, just seems tricky to get right :-)
>
> Ok, I believe I have found a race condition between m_spin_lock() and
> m_spin_unlock().
>
> In m_spin_unlock(), we do "next = ACCESS_ONCE(node->next)". Then, if
> next is not NULL, we proceed to set next->locked to 1.
>
> A thread in m_spin_lock() in the unqueue path could execute
> "next = cmpxchg(&prev->next, node, NULL)" after the thread in
> m_spin_unlock() accesses its node->next and finds that it is not NULL.
> Then, the thread in m_spin_lock() could check !node->locked before
> the thread in m_spin_unlock() sets next->locked to 1.

Yes indeed. How silly of me to not spot that!

> The following addition change was able to solve the initial lockups that were
> occurring when running fserver on a 2 socket box.
>
> ---
> diff --git a/kernel/locking/mutex.c b/kernel/locking/mutex.c
> index 9eb4dbe..e71a84a 100644
> --- a/kernel/locking/mutex.c
> +++ b/kernel/locking/mutex.c
> @@ -513,8 +513,13 @@ static void m_spin_unlock(struct m_spinlock **lock)
> return;
>
> next = ACCESS_ONCE(node->next);
> - if (unlikely(next))
> - break;
> +
> + if (unlikely(next)) {
> + next = cmpxchg(&node->next, next, NULL);
> +
> + if (next)

The cmpxchg could fail and next still be !NULL I suppose.

> + break;
> + }


The way I wrote that same loop in step-B, is:


for (;;) {
if (*lock == node && cmpxchg(lock, node, prev) == node)
return

next = xchg(&node->next, NULL); /* B -> A */
if (next)
break;

arch_mutex_cpu_relax();
}

I suppose we can make that something like:


if (node->next) {
next = xchg(&node->next, NULL);
if (next)
break
}

To avoid the xchg on every loop.

I had wanted to avoid the additional locked op in the unlock path, but
yes that does make things easier.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/