| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Thu, 3 Jun 2021 09:29:28 +0800
From: Yafang Shao <laoar.shao@...il.com>
To: Vincent Guittot <vincent.guittot@...aro.org>
Cc: Ingo Molnar <mingo@...hat.com>,
Peter Zijlstra <peterz@...radead.org>,
Juri Lelli <juri.lelli@...hat.com>,
Steven Rostedt <rostedt@...dmis.org>,
Ben Segall <bsegall@...gle.com>, Mel Gorman <mgorman@...e.de>,
Daniel Bristot de Oliveira <bristot@...hat.com>,
linux-kernel <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 1/1] sched: do active load balance on the new idle cpu
On Wed, Jun 2, 2021 at 9:26 PM Vincent Guittot
<vincent.guittot@...aro.org> wrote:
>
> On Wed, 2 Jun 2021 at 14:58, Yafang Shao <laoar.shao@...il.com> wrote:
> >
> > On Wed, Jun 2, 2021 at 8:37 PM Vincent Guittot
> > <vincent.guittot@...aro.org> wrote:
> > >
> > > On Wed, 2 Jun 2021 at 14:26, Yafang Shao <laoar.shao@...il.com> wrote:
> > > >
> > > > We monitored our latency-sensitive RT tasks are randomly preempted by the
> > > > kthreads migration/n, which means to migrate tasks on CPUn to other new
> > > > idle CPU. The logical as follows,
> > > >
> > > > new idle CPU CPU n
> > > > (no task to run) (busy running)
> > > > wakeup migration/n (busy running)
> > > > (idle) migraion/n preempts current task
> > > > run the migrated task (busy running)
> > >
> > > migration thread is only used when we want to migrate the currently
> > > running task of the source cpu.
> >
> > Could you pls explain it in detail ?
>
> CPU A
> become idle
> call newidle_balance()
> ...
> load_balance()
> ... CPU B is the busiest cpu
> env.src_cpu = CPU B;
>
> if (busiest->nr_running > 1) {
> ...
> There is more than 1 runnable threads on CPU B
> Try to migrate cfs runnable but not running tasks from CPU B to CPU A
> in your case, the migration of cfs task should happen here
> because the RT task is running
>
> Handle case of pinned tasks
>
>
> if (!ld_moved)
> no runnable but not running task was moved so we might want to
> migrate the current running task
>
> need_active_balance() should not return true in your case
> because tasks should have been migrated during the step above
>
> wake up stop/migration thread to preempt the current running
> thread so we can migrate it
>
Thanks for the explanation.
>
> so you have has a UC which doesn't migrate task in the 1st step when
> trying to pull runnable and not running tasks but it makes
> need_active_balance() return true. Woudl be good to know which
> condition makes need_active_balance() to return true
>
I will analyze why need_active_balance() returns true.
>
> > But I find the migration/n will pick a task from src_rq->cfs_tasks
> > rather than the current running task, see also detach_one_task():
>
> The current running task is migration/n one at that time
>
> >
> > detach_one_task
> > list_for_each_entry_reverse(p, &env->src_rq->cfs_tasks, se.group_node) {
> > detach_task(p, env);
> > }
> >
> >
> > > This doesn't seem to be your case as it's a RT thread that is
> > > currently running so the migration thread should not be woken up as we
> > > don't need it to migrate a runnable but not running cfs thread from
> > > coin to new idle CPU
> > >
> > > Do you have more details about the UC. Could it be a race between new
> > > idle load balance starting migration thread to pull the cfs running
> > > thread and the RT thread waking up and preempting cfs task before
> > > migration threads which then preempt your RT threads
> > >
> > >
> >
> > No, it is not a race. Below is the detail from sched:sched_swith tracepoint:
> >
> > sensing_node-8880 [007] d... 4300.544185: sched_switch:
> > prev_comm=sensing_node prev_pid=8880 prev_prio=98 prev_state=S ==>
> > next_comm=sensing_node next_pid=8897 next_prio=98
> > sensing_node-8897 [007] d... 4300.544214: sched_switch:
> > prev_comm=sensing_node prev_pid=8897 prev_prio=98 prev_state=S ==>
> > next_comm=sensing_node next_pid=8880 next_prio=98
> > sensing_node-8880 [007] d... 4300.544506: sched_switch:
> > prev_comm=sensing_node prev_pid=8880 prev_prio=98 prev_state=R ==>
> > next_comm=migration/7 next_pid=47 next_prio=0
> > migration/7-47 [007] d... 4300.544509: sched_switch:
> > prev_comm=migration/7 prev_pid=47 prev_prio=0 prev_state=S ==>
> > next_comm=sensing_node next_pid=8880 next_prio=98
> >
> > sensing_node is a RR task and it was preempted by migration/7.
> >
> > >
> > > >
> > > > As the new idle CPU is going to be idle, we'd better move the migration
> > > > work on it instead of burdening the busy CPU. After this change, the
> > > > logic is,
> > > > new idle CPU CPU n
> > > > (no task to run) (busy running)
> > > > migrate task from CPU n (busy running)
> > > > run the migrated task (busy running)
> > > >
> > > > Signed-off-by: Yafang Shao <laoar.shao@...il.com>
> > > > ---
> > > > kernel/sched/fair.c | 17 +++++------------
> > > > 1 file changed, 5 insertions(+), 12 deletions(-)
> > > >
> > > > diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
> > > > index 3248e24a90b0..3e8b98b982ff 100644
> > > > --- a/kernel/sched/fair.c
> > > > +++ b/kernel/sched/fair.c
> > > > @@ -9807,13 +9807,11 @@ static int load_balance(int this_cpu, struct rq *this_rq,
> > > > busiest->push_cpu = this_cpu;
> > > > active_balance = 1;
> > > > }
> > > > - raw_spin_unlock_irqrestore(&busiest->lock, flags);
> > > >
> > > > - if (active_balance) {
> > > > - stop_one_cpu_nowait(cpu_of(busiest),
> > > > - active_load_balance_cpu_stop, busiest,
> > > > - &busiest->active_balance_work);
> > > > - }
> > > > + if (active_balance)
> > > > + active_load_balance_cpu_stop(busiest);
> > >
> > > this doesn't make sense because we reach this point if we want to
> > > migrate the current running task of the busiest cpu and in order to do
> > > this we need the preempt this current running thread
> > >
> > > > +
> > > > + raw_spin_unlock_irqrestore(&busiest->lock, flags);
> > > > }
> > > > } else {
> > > > sd->nr_balance_failed = 0;
> > > > @@ -9923,7 +9921,6 @@ static int active_load_balance_cpu_stop(void *data)
> > > > struct task_struct *p = NULL;
> > > > struct rq_flags rf;
> > > >
> > > > - rq_lock_irq(busiest_rq, &rf);
> > > > /*
> > > > * Between queueing the stop-work and running it is a hole in which
> > > > * CPUs can become inactive. We should not move tasks from or to
> > > > @@ -9933,8 +9930,7 @@ static int active_load_balance_cpu_stop(void *data)
> > > > goto out_unlock;
> > > >
> > > > /* Make sure the requested CPU hasn't gone down in the meantime: */
> > > > - if (unlikely(busiest_cpu != smp_processor_id() ||
> > > > - !busiest_rq->active_balance))
> > > > + if (unlikely(!busiest_rq->active_balance))
> > > > goto out_unlock;
> > > >
> > > > /* Is there any task to move? */
> > > > @@ -9981,13 +9977,10 @@ static int active_load_balance_cpu_stop(void *data)
> > > > rcu_read_unlock();
> > > > out_unlock:
> > > > busiest_rq->active_balance = 0;
> > > > - rq_unlock(busiest_rq, &rf);
> > > >
> > > > if (p)
> > > > attach_one_task(target_rq, p);
> > > >
> > > > - local_irq_enable();
> > > > -
> > > > return 0;
> > > > }
> > > >
> > > > --
> > > > 2.17.1
> > > >
> >
> >
> >
> > --
> > Thanks
> > Yafang
--
Thanks
Yafang
Powered by blists - more mailing lists