Re: [PATCH RFC tip/core/rcu 5/5] rcu: Limit expedited helping to every 10 ms or every 4th GP

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

On Wed, Jul 01, 2015 at 12:07:30PM +0200, Peter Zijlstra wrote:
> On Tue, Jun 30, 2015 at 02:48:30PM -0700, Paul E. McKenney wrote:
> > From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
> 
> This seems like a good place to explain why this is a desirable thing,
> no?

Good point.

> Why would you want to limit this?

Because the unconditional wakeup is a two-edges sword.  It reduces
the latency of normal RCU grace periods on the one hand, but it makes
rcu_sched consume even more CPU on the other.

							Thanx, Paul

> > Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
> > ---
> >  kernel/rcu/tree.c | 15 ++++++++++++---
> >  1 file changed, 12 insertions(+), 3 deletions(-)
> > 
> > diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
> > index 308b6acb4260..247aa1120c4c 100644
> > --- a/kernel/rcu/tree.c
> > +++ b/kernel/rcu/tree.c
> > @@ -3505,10 +3505,19 @@ void synchronize_sched_expedited(void)
> >  			   !atomic_read(&rsp->expedited_need_qs));
> >  
> >  	rcu_exp_gp_seq_end(rsp);
> > -	mutex_unlock(&rnp->exp_funnel_mutex);
> >  	smp_mb(); /* ensure subsequent action seen after grace period. */
> > -	if (rsp->gp_kthread && rcu_gp_in_progress(rsp))
> > -		wake_up(&rsp->gp_wq);
> > +	if (rsp->gp_kthread && rcu_gp_in_progress(rsp)) {
> > +		static unsigned long nextgp;
> > +		static unsigned long nextjiffy;
> > +
> > +		if (time_after_eq(jiffies, nextgp) ||
> > +		    ULONG_CMP_GE(rsp->gpnum, nextgp)) {
> > +			nextgp = rsp->gpnum + 4;
> > +			nextjiffy = jiffies + 10;
> > +			wake_up(&rsp->gp_wq);
> > +		}
> > +	}
> > +	mutex_unlock(&rnp->exp_funnel_mutex);
> >  
> >  	put_online_cpus();
> >  }
> > -- 
> > 1.8.1.5
> > 
>