Re: [PATCH v3 3/7] locking/pvqspinlock: Collect slowpath lock statistics

[Waiman Long]

On 07/26/2015 09:14 PM, Davidlohr Bueso wrote:
> On Wed, 2015-07-22 at 16:12 -0400, Waiman Long wrote:
> > This patch enables the accumulation of kicking and waiting related
> > PV qspinlock statistics when the new QUEUED_LOCK_STAT configuration
> > option is selected. It also enables the collection of kicking and
> > wakeup latencies which have a heavy dependency on the CPUs being used.
> >
> > The measured latencies for different CPUs are:
> >
> > 	CPU		Wakeup		Kicking
> > 	---		------		-------
> > 	Haswell-EX	89.8us		 7.4us
> > 	Westmere-EX	67.6us		 9.3us
> But you don't mention anything about the overhead of enabling
> QUEUED_LOCK_STAT. This does several atomic ops, thus potentially
> thrashing workloads.

Yes, QUEUED_LOCK_STAT will slow performance a bit. It is like enabling 
LOCK_STAT and you will expect some slow down too. It is essentially a 
debugging option to see what had actually happened in the system. It 
should be turned off in a production system.

Cheers,
Longman
--
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/