Re: [PATCH RFC] sched/fair: Avoid unnecessary IPIs for ILB

[Vincent Guittot]

On Sun, 8 Oct 2023 at 19:35, Joel Fernandes <joel@xxxxxxxxxxxxxxxxx> wrote:
>
> On Fri, Oct 06, 2023 at 12:51:41PM +0200, Ingo Molnar wrote:
> >
> > * Joel Fernandes (Google) <joel@xxxxxxxxxxxxxxxxx> wrote:
> >
> > > From: Vineeth Pillai <vineethrp@xxxxxxxxxx>
> > >
> > > Whenever a CPU stops its tick, it now requires another idle CPU to handle the
> > > balancing for it because it can't perform its own periodic load balancing.
> > > This means it might need to update 'nohz.next_balance' to 'rq->next_balance' if
> > > the upcoming nohz-idle load balancing is too distant in the future. This update
> > > process is done by triggering an ILB, as the general ILB handler
> > > (_nohz_idle_balance) that manages regular nohz balancing also refreshes
> > > 'nohz.next_balance' by looking at the 'rq->next_balance' of all other idle CPUs
> > > and selecting the smallest value.
> > >
> > > Triggering this ILB can be achieved by setting the NOHZ_NEXT_KICK flag. This
> > > primarily results in the ILB handler updating 'nohz.next_balance' while
> > > possibly not doing any load balancing at all. However, sending an IPI merely to
> > > refresh 'nohz.next_balance' seems excessive, and there ought to be a more
> > > efficient method to update 'nohz.next_balance' from the local CPU.
> > >
> > > Fortunately, there already exists a mechanism to directly invoke the ILB
> > > handler (_nohz_idle_balance) without initiating an IPI. It's accomplished by
> > > setting the NOHZ_NEWILB_KICK flag. This flag is set during regular "newly idle"
> > > balancing and solely exists to update a CPU's blocked load if it couldn't pull
> > > more tasks during regular "newly idle balancing" - and it does so without
> > > having to send any IPIs. Once the flag is set, the ILB handler is called
> > > directly from do_idle()-> nohz_run_idle_balance(). While its goal is to update
> > > the blocked load without an IPI, in our situation, we aim to refresh
> > > 'nohz.next_balance' without an IPI but we can piggy back on this.
> > >
> > > So in this patch, we reuse this mechanism by also setting the NOHZ_NEXT_KICK to
> > > indicate nohz.next_balance needs an update via this direct call shortcut. Note
> > > that we set this flag without knowledge that the tick is about to be stopped,
> > > because at the point we do it, we have no way of knowing that. However we do
> > > know that the CPU is about to enter idle. In our testing, the reduction in IPIs
> > > is well worth updating nohz.next_balance a few more times.
> > >
> > > Also just to note, without this patch we observe the following pattern:
> > >
> > > 1. A CPU is about to stop its tick.
> > > 2. It sets nohz.needs_update to 1.
> > > 3. It then stops its tick and goes idle.
> > > 4. The scheduler tick on another CPU checks this flag and decides an ILB kick is needed.
> > > 5. The ILB CPU ends up being the one that just stopped its tick!
> > > 6. This results in an IPI to the tick-stopped CPU which ends up waking it up
> > >    and disturbing it!
> > >
> > > Testing shows a considerable reduction in IPIs when doing this:
> > >
> > > Running "cyclictest -i 100 -d 100 --latency=1000 -t -m" on a 4vcpu VM
> > > the IPI call count profiled over 10s period is as follows:
> > > without fix: ~10500
> > > with fix: ~1000
> > >
> > > Fixes: 7fd7a9e0caba ("sched/fair: Trigger nohz.next_balance updates when a CPU goes NOHZ-idle")
> > >
> > > [ Joel: wrote commit messages, collaborated on fix, helped reproduce issue etc. ]
> > >
> > > Cc: Suleiman Souhlal <suleiman@xxxxxxxxxx>
> > > Cc: Steven Rostedt <rostedt@xxxxxxxxxxx>
> > > Cc: Hsin Yi <hsinyi@xxxxxxxxxx>
> > > Cc: Frederic Weisbecker <frederic@xxxxxxxxxx>
> > > Cc: Paul E. McKenney <paulmck@xxxxxxxxxx>
> > > Signed-off-by: Vineeth Pillai <vineethrp@xxxxxxxxxx>
> > > Co-developed-by: Joel Fernandes (Google) <joel@xxxxxxxxxxxxxxxxx>
> > > Signed-off-by: Joel Fernandes (Google) <joel@xxxxxxxxxxxxxxxxx>
> > > ---
> > >  kernel/sched/fair.c | 21 ++++++++++++++-------
> > >  1 file changed, 14 insertions(+), 7 deletions(-)
> > >
> > > diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> > > index cb225921bbca..2ece55f32782 100644
> > > --- a/kernel/sched/fair.c
> > > +++ b/kernel/sched/fair.c
> > > @@ -11786,13 +11786,12 @@ void nohz_balance_enter_idle(int cpu)
> > >     /*
> > >      * Ensures that if nohz_idle_balance() fails to observe our
> > >      * @idle_cpus_mask store, it must observe the @has_blocked
> > > -    * and @needs_update stores.
> > > +    * stores.
> > >      */
> > >     smp_mb__after_atomic();
> > >
> > >     set_cpu_sd_state_idle(cpu);
> > >
> > > -   WRITE_ONCE(nohz.needs_update, 1);
> > >  out:
> > >     /*
> > >      * Each time a cpu enter idle, we assume that it has blocked load and
> > > @@ -11945,21 +11944,25 @@ static bool nohz_idle_balance(struct rq *this_rq, enum cpu_idle_type idle)
> > >  }
> > >
> > >  /*
> > > - * Check if we need to run the ILB for updating blocked load before entering
> > > - * idle state.
> > > + * Check if we need to run the ILB for updating blocked load and/or updating
> > > + * nohz.next_balance before entering idle state.
> > >   */
> > >  void nohz_run_idle_balance(int cpu)
> > >  {
> > >     unsigned int flags;
> > >
> > > -   flags = atomic_fetch_andnot(NOHZ_NEWILB_KICK, nohz_flags(cpu));
> > > +   flags = atomic_fetch_andnot(NOHZ_NEWILB_KICK | NOHZ_NEXT_KICK, nohz_flags(cpu));
> > > +
> > > +   if (!flags)
> > > +           return;
> > >
> > >     /*
> > >      * Update the blocked load only if no SCHED_SOFTIRQ is about to happen
> > >      * (ie NOHZ_STATS_KICK set) and will do the same.
> > >      */
> > > -   if ((flags == NOHZ_NEWILB_KICK) && !need_resched())
> > > -           _nohz_idle_balance(cpu_rq(cpu), NOHZ_STATS_KICK);
> > > +   if ((flags == (flags & (NOHZ_NEXT_KICK | NOHZ_NEWILB_KICK))) &&
> > > +       !need_resched())
> > > +           _nohz_idle_balance(cpu_rq(cpu), flags);
> > >  }
> > >
> > >  static void nohz_newidle_balance(struct rq *this_rq)
> > > @@ -11977,6 +11980,10 @@ static void nohz_newidle_balance(struct rq *this_rq)
> > >     if (this_rq->avg_idle < sysctl_sched_migration_cost)
> > >             return;
> > >
> > > +   /* If rq->next_balance before nohz.next_balance, trigger ILB */
> > > +   if (time_before(this_rq->next_balance, READ_ONCE(nohz.next_balance)))
> > > +           atomic_or(NOHZ_NEXT_KICK, nohz_flags(this_cpu));
> > > +
> > >     /* Don't need to update blocked load of idle CPUs*/
> > >     if (!READ_ONCE(nohz.has_blocked) ||
> > >         time_before(jiffies, READ_ONCE(nohz.next_blocked)))
> >
> > Ok, judging by your IPI reduction numbers this is definitely an
> > optimization we want to do.
> >
> > The patch does make _nohz_idle_balance() run more parallel, as previously
> > it would be generally run by the first-idle CPU in nohz.idle_cpus_mask (at
> > least for next_balance updates), but I think it's still SMP-safe, as all
> > key data structure updates are already rq-locked AFAICS.
>
> One thing I am confused about in the original code is:
>
> tick_nohz_idle_stop_tick() is what sets the nohz.idle_cpus_mask.
> However, nohz_run_idle_balance() is called before that can happen, in
> do_idle(). So it is possible that NOHZ_NEWILB_KICK is set for a CPU but it is
> not yet in the mask.

2 things:
- update of nohz.idle_cpus_mask is not strictly aligned with cpu
entering/exiting idle state. It is set when entering but only cleared
during the next tick on the cpu because nohz.idle_cpus_mask is a
bottleneck when all CPUs enter/exit idle at high rate (usec). This
implies that a cpu entering idle can already be set in
nohz.idle_cpus_mask
- NOHZ_NEWILB_KICK is more about updating blocked load of others
already idle CPUs than the one entering idle which has already updated
its blocked load in newidle_balance()

The goal of nohz_run_idle_balance() is to run ILB only for updating
the blocked load of already idle CPUs without waking up one of those
idle CPUs and outside the preempt disable / irq off phase of the local
cpu about to enter idle because it can take a long time.

>
> So will this code in _nohz_idle_balance() really run in such a scenario?
>
>                 if (flags & NOHZ_STATS_KICK)
>                         has_blocked_load |= update_nohz_stats(rq);
>
> AFAICS, this loop may not select the CPU due to its absence from the mask:
>         for_each_cpu_wrap(balance_cpu,  nohz.idle_cpus_mask, this_cpu+1)
>
> I must be missing something. I'll go trace this path later as well.
>
> thanks,
>
>  - Joel
>