Re: [PATCH v2] sched/core: Reduce cost of sched_move_task when config autogroup

From: chi wu
Date: Tue Mar 21 2023 - 07:51:35 EST

Peter Zijlstra <peterz@xxxxxxxxxxxxx> 于2023年3月21日周二 17:36写道：
>
> On Tue, Mar 21, 2023 at 02:44:59PM +0800, wuchi wrote:
>
> > kernel/sched/core.c | 30 ++++++++++++++++++++++++++++--
> > 1 file changed, 28 insertions(+), 2 deletions(-)
> >
> > diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> > index a380f34789a2..1e7d6a8c3455 100644
> > --- a/kernel/sched/core.c
> > +++ b/kernel/sched/core.c
> > @@ -10330,7 +10330,7 @@ void sched_release_group(struct task_group *tg)
> > spin_unlock_irqrestore(&task_group_lock, flags);
> > }
> >
> > -static void sched_change_group(struct task_struct *tsk)
> > +static struct task_group *sched_get_task_group(struct task_struct *tsk)
> > {
> > struct task_group *tg;
> >
> > @@ -10342,7 +10342,28 @@ static void sched_change_group(struct task_struct *tsk)
> > tg = container_of(task_css_check(tsk, cpu_cgrp_id, true),
> > struct task_group, css);
> > tg = autogroup_task_group(tsk, tg);
> > - tsk->sched_task_group = tg;
> > +
> > + return tg;
> > +}
> > +
> > +static bool sched_task_group_changed(struct task_struct *tsk)
> > +{
> > + /*
> > + * Some sched_move_task calls are useless because that
> > + * task_struct->sched_task_group maybe not changed (equals
> > + * task_group of cpu_cgroup) when system enable autogroup.
> > + * So do some checks in sched_move_task.
> > + */
> > +#ifdef CONFIG_SCHED_AUTOGROUP
> > + return sched_get_task_group(tsk) != tsk->sched_task_group;
> > +#else
> > + return true;
> > +#endif /* CONFIG_SCHED_AUTOGROUP */
> > +}
> > +
> > +static void sched_change_group(struct task_struct *tsk)
> > +{
> > + tsk->sched_task_group = sched_get_task_group(tsk);
> >
> > #ifdef CONFIG_FAIR_GROUP_SCHED
> > if (tsk->sched_class->task_change_group)
> > @@ -10367,6 +10388,10 @@ void sched_move_task(struct task_struct *tsk)
> > struct rq *rq;
> >
> > rq = task_rq_lock(tsk, &rf);
> > +
> > + if (!sched_task_group_changed(tsk))
> > + goto unlock;
> > +
> > update_rq_clock(rq);
> >
> > running = task_current(rq, tsk);
> > @@ -10391,6 +10416,7 @@ void sched_move_task(struct task_struct *tsk)
> > resched_curr(rq);
> > }
> >
> > +unlock:
> > task_rq_unlock(rq, tsk, &rf);
> > }
>
> Would you mind terribly if I change it like so?
>
> ---
> Subject: sched/core: Reduce cost of sched_move_task when config autogroup
> From: wuchi <wuchi.zero@xxxxxxxxx>
> Date: Tue, 21 Mar 2023 14:44:59 +0800
>
> From: wuchi <wuchi.zero@xxxxxxxxx>
>
> Some sched_move_task calls are useless because that
> task_struct->sched_task_group maybe not changed (equals task_group
> of cpu_cgroup) when system enable autogroup. So do some checks in
> sched_move_task.
>
> sched_move_task eg:
> task A belongs to cpu_cgroup0 and autogroup0, it will always belong
> to cpu_cgroup0 when do_exit. So there is no need to do {de|en}queue.
> The call graph is as follow.
>
> do_exit
> sched_autogroup_exit_task
> sched_move_task
> dequeue_task
> sched_change_group
> A.sched_task_group = sched_get_task_group (=cpu_cgroup0)
> enqueue_task
>
> Performance results:
> ===========================
> 1. env
> cpu: bogomips=4600.00
> kernel: 6.3.0-rc3
> cpu_cgroup: 6:cpu,cpuacct:/user.slice
>
> 2. cmds
> do_exit script:
>
> for i in {0..10000}; do
> sleep 0 &
> done
> wait
>
> Run the above script, then use the following bpftrace cmd to get
> the cost of sched_move_task:
>
> bpftrace -e 'k:sched_move_task { @ts[tid] = nsecs; }
> kr:sched_move_task /@ts[tid]/
> { @ns += nsecs - @ts[tid]; delete(@ts[tid]); }'
>
> 3. cost time(ns):
> without patch: 43528033
> with patch: 18541416
> diff:-24986617 -57.4%
>
> As the result show, the patch will save 57.4% in the scenario.
>
> Signed-off-by: wuchi <wuchi.zero@xxxxxxxxx>
> Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx>
> Link: https://lkml.kernel.org/r/20230321064459.39421-1-wuchi.zero@xxxxxxxxx
> ---
> kernel/sched/core.c | 22 +++++++++++++++++++---
> 1 file changed, 19 insertions(+), 3 deletions(-)
>
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -10351,7 +10351,7 @@ void sched_release_group(struct task_gro
> spin_unlock_irqrestore(&task_group_lock, flags);
> }
>
> -static void sched_change_group(struct task_struct *tsk)
> +static struct task_group *sched_get_task_group(struct task_struct *tsk)
> {
> struct task_group *tg;
>
> @@ -10363,7 +10363,13 @@ static void sched_change_group(struct ta
> tg = container_of(task_css_check(tsk, cpu_cgrp_id, true),
> struct task_group, css);
> tg = autogroup_task_group(tsk, tg);
> - tsk->sched_task_group = tg;
> +
> + return tg;
> +}
> +
> +static void sched_change_group(struct task_struct *tsk, struct task_group *group)
> +{
> + tsk->sched_task_group = group;
>
> #ifdef CONFIG_FAIR_GROUP_SCHED
> if (tsk->sched_class->task_change_group)
> @@ -10384,10 +10390,19 @@ void sched_move_task(struct task_struct
> {
> int queued, running, queue_flags =
> DEQUEUE_SAVE | DEQUEUE_MOVE | DEQUEUE_NOCLOCK;
> + struct task_group *group;
> struct rq_flags rf;
> struct rq *rq;
>
> rq = task_rq_lock(tsk, &rf);
> + /*
> + * Esp. with SCHED_AUTOGROUP enabled it is possible to get superfluous
> + * group changes.
> + */
> + group = sched_get_task_group(tsk);
> + if (group == tsk->sched_task_group)
> + goto unlock;
> +
> update_rq_clock(rq);
>
> running = task_current(rq, tsk);
> @@ -10398,7 +10413,7 @@ void sched_move_task(struct task_struct
> if (running)
> put_prev_task(rq, tsk);
>
> - sched_change_group(tsk);
> + sched_change_group(tsk, group);
>
> if (queued)
> enqueue_task(rq, tsk, queue_flags);
> @@ -10412,6 +10427,7 @@ void sched_move_task(struct task_struct
> resched_curr(rq);
> }
>
> +unlock:
> task_rq_unlock(rq, tsk, &rf);
> }
>

It‘s more efficient and concise, I learn more from the code. Thanks!