Re: [RFC 2/3] mutex: Modify the way optimistic spinners are queued

2014-01-15 Thread Jason Low
On Wed, 2014-01-15 at 10:10 -0500, Waiman Long wrote: > On 01/14/2014 07:33 PM, Jason Low wrote: > > * When there's no owner, we might have preempted between the > > @@ -503,8 +504,10 @@ __mutex_lock_common(struct mutex *lock, long state, > > unsigned int subclass, > > *

Re: [RFC 2/3] mutex: Modify the way optimistic spinners are queued

2014-01-15 Thread Waiman Long
On 01/14/2014 07:33 PM, Jason Low wrote: This patch is needed for patch 3, but should also be beneficial in general. The mutex->spin_mlock was introduced in order to ensure that only 1 thread loops on lock->owner field at a time to reduce cache line contention. When lock->owner is NULL and the l