Re: [PATCH v2 2/5] mutex: Modify the way optimistic spinners are queued

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Tue, Jan 28, 2014 at 11:13:13AM -0800, Jason Low wrote:
> The mutex->spin_mlock was introduced in order to ensure that only 1 thread
> spins for lock acquisition at a time to reduce cache line contention. When
> lock->owner is NULL and the lock->count is still not 1, the spinner(s) will
> continually release and obtain the lock->spin_mlock. This can generate
> quite a bit of overhead/contention, and also might just delay the spinner
> from getting the lock.
> 
> This patch modifies the way optimistic spinners are queued by queuing before
> entering the optimistic spinning loop as oppose to acquiring before every
> call to mutex_spin_on_owner(). So in situations where the spinner requires
> a few extra spins before obtaining the lock, then there will only be 1 spinner
> trying to get the lock and it will avoid the overhead from unnecessarily
> unlocking and locking the spin_mlock.
> 
> Signed-off-by: Jason Low <jason.low2@hp.com>

One question below.  Also, this might well have a visible effect on
performance, so would be good to see the numbers.

							Thanx, Paul

> ---
>  kernel/locking/mutex.c |   16 +++++++---------
>  1 files changed, 7 insertions(+), 9 deletions(-)
> 
> diff --git a/kernel/locking/mutex.c b/kernel/locking/mutex.c
> index 85c6be1..7519d27 100644
> --- a/kernel/locking/mutex.c
> +++ b/kernel/locking/mutex.c
> @@ -419,6 +419,7 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
>  	struct mutex_waiter waiter;
>  	unsigned long flags;
>  	int ret;
> +	struct mspin_node node;
> 
>  	preempt_disable();
>  	mutex_acquire_nest(&lock->dep_map, subclass, 0, nest_lock, ip);
> @@ -449,9 +450,9 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
>  	if (!mutex_can_spin_on_owner(lock))
>  		goto slowpath;
> 
> +	mspin_lock(MLOCK(lock), &node);
>  	for (;;) {
>  		struct task_struct *owner;
> -		struct mspin_node  node;
> 
>  		if (use_ww_ctx && ww_ctx->acquired > 0) {
>  			struct ww_mutex *ww;
> @@ -466,19 +467,16 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
>  			 * performed the optimistic spinning cannot be done.
>  			 */
>  			if (ACCESS_ONCE(ww->ctx))
> -				goto slowpath;
> +				break;
>  		}
> 
>  		/*
>  		 * If there's an owner, wait for it to either
>  		 * release the lock or go to sleep.
>  		 */
> -		mspin_lock(MLOCK(lock), &node);
>  		owner = ACCESS_ONCE(lock->owner);
> -		if (owner && !mutex_spin_on_owner(lock, owner)) {
> -			mspin_unlock(MLOCK(lock), &node);
> -			goto slowpath;
> -		}
> +		if (owner && !mutex_spin_on_owner(lock, owner))
> +			break;
> 
>  		if ((atomic_read(&lock->count) == 1) &&
>  		    (atomic_cmpxchg(&lock->count, 1, 0) == 1)) {
> @@ -495,7 +493,6 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
>  			preempt_enable();
>  			return 0;
>  		}
> -		mspin_unlock(MLOCK(lock), &node);
> 
>  		/*
>  		 * When there's no owner, we might have preempted between the
> @@ -504,7 +501,7 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
>  		 * the owner complete.
>  		 */
>  		if (!owner && (need_resched() || rt_task(task)))
> -			goto slowpath;
> +			break;
> 
>  		/*
>  		 * The cpu_relax() call is a compiler barrier which forces
> @@ -514,6 +511,7 @@ __mutex_lock_common(struct mutex *lock, long state, unsigned int subclass,
>  		 */
>  		arch_mutex_cpu_relax();
>  	}
> +	mspin_unlock(MLOCK(lock), &node);
>  slowpath:

Are there any remaining goto statements to slowpath?  If so, they need
to release the lock.  If not, this label should be removed.

>  #endif
>  	spin_lock_mutex(&lock->wait_lock, flags);
> -- 
> 1.7.1
>