On Tue, Jul 11, 2017 at 06:34:22PM +0200, Peter Zijlstra wrote: > On Tue, Jul 11, 2017 at 06:09:27PM +0200, Frederic Weisbecker wrote: > > > > > - tick_nohz_idle_enter costs 7058ns - 10726ns > > > > - tick_nohz_idle_exit costs 8372ns - 20850ns > > > > > > Right, those are horrible expensive, but skipping them isn't 'hard', the > > > only tricky bit is finding a condition that makes sense. > > > > Note you can statically disable it with nohz=0 boot parameter. > > Yeah, but that's bad for power usage, nobody wants that. > > > > See Mike's patch: https://patchwork.kernel.org/patch/2839221/ > > > > > > Combined with the above, and possibly a better condition, that should > > > get rid of most of this. > > > > Such a patch could work well if the decision from the scheduler to not stop > > the tick > > happens on idle entry. > > > > Now if sched_needs_cpu() first allows to stop the tick then refuses it later > > in the end of an idle IRQ, this won't have the desired effect. As long as > > ts->tick_stopped=1, > > it stays so until we really restart the tick. So the whole costly nohz > > machinery stays on. > > > > I guess it doesn't matter though, as we are talking about making fast idle > > entry so the > > decision not to stop the tick is likely to be done once on idle entry, when > > ts->tick_stopped=0. > > > > One exception though: if the tick is already stopped when we enter idle > > (full nohz case). And > > BTW stopping the tick outside idle shouldn't be concerned here. > > > > So I'd rather put that on can_stop_idle_tick(). > > Mike's patch much predates the existence of that function I think ;-) But > sure.. > > > > > > > > - totally from arch_cpu_idle_enter entry to arch_cpu_idle_exit return > > > > costs > > > > 9122ns - 15318ns. > > > > --In this period, rcu_idle_enter costs 1985ns - 2262ns, rcu_idle_exit > > > > costs > > > > 1813ns - 3507ns > > > > > > Is that the POPF being painful? or something else? > > > > Probably that and the atomic_add_return(). > > I got properly lost in the RCU machinery. It wasn't at all clear to me > if rcu_eqs_enter_common() was a slow-path function or not.
It is called on pretty much every transition to idle. > Also, RCU_FAST_NO_HZ will make a fairly large difference here.. Paul > what's the state of that thing, do we actually want that or not? If you are battery powered and don't have tight real-time latency constraints, you want it -- it has represent a 30-40% boost in battery lifetime for some low-utilization battery-powered devices. Otherwise, probably not. > But I think we can at the very least do this; it only gets called from > kernel/sched/idle.c and both callsites have IRQs explicitly disabled by > that point. > > > diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c > index 51d4c3acf32d..dccf2dc8155a 100644 > --- a/kernel/rcu/tree.c > +++ b/kernel/rcu/tree.c > @@ -843,13 +843,8 @@ static void rcu_eqs_enter(bool user) > */ > void rcu_idle_enter(void) > { > - unsigned long flags; > - > - local_irq_save(flags); With this addition, I am all for it: RCU_LOCKDEP_WARN(!irqs_disabled(), "rcu_idle_enter() invoked with irqs enabled!!!"); If you are OK with this addition, may I please have your Signed-off-by? Thanx, Paul > rcu_eqs_enter(false); > - local_irq_restore(flags); > } > -EXPORT_SYMBOL_GPL(rcu_idle_enter); > > #ifdef CONFIG_NO_HZ_FULL > /** >