linux-kernel - Re: [PATCH 2/7] sched/fair: prepare throttle path for task based throttle

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

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20250521115115.GB24746@bytedance>
Date: Wed, 21 May 2025 19:51:15 +0800
From: Aaron Lu <ziqianlu@...edance.com>
To: Florian Bezdeka <florian.bezdeka@...mens.com>
Cc: Valentin Schneider <vschneid@...hat.com>,
	Ben Segall <bsegall@...gle.com>,
	K Prateek Nayak <kprateek.nayak@....com>,
	Peter Zijlstra <peterz@...radead.org>,
	Josh Don <joshdon@...gle.com>, Ingo Molnar <mingo@...hat.com>,
	Vincent Guittot <vincent.guittot@...aro.org>,
	Xi Wang <xii@...gle.com>, linux-kernel@...r.kernel.org,
	Juri Lelli <juri.lelli@...hat.com>,
	Dietmar Eggemann <dietmar.eggemann@....com>,
	Steven Rostedt <rostedt@...dmis.org>, Mel Gorman <mgorman@...e.de>,
	Chengming Zhou <chengming.zhou@...ux.dev>,
	Chuyi Zhou <zhouchuyi@...edance.com>,
	Jan Kiszka <jan.kiszka@...mens.com>
Subject: Re: [PATCH 2/7] sched/fair: prepare throttle path for task based
 throttle

On Wed, May 21, 2025 at 02:37:05PM +0800, Aaron Lu wrote:
> On Tue, May 20, 2025 at 02:02:54PM +0200, Florian Bezdeka wrote:
> > On Tue, 2025-05-20 at 18:41 +0800, Aaron Lu wrote:
> > > @@ -6744,6 +6805,7 @@ static bool check_cfs_rq_runtime(struct cfs_rq *cfs_rq) { return false; }
> > >  static void check_enqueue_throttle(struct cfs_rq *cfs_rq) {}
> > >  static inline void sync_throttle(struct task_group *tg, int cpu) {}
> > >  static __always_inline void return_cfs_rq_runtime(struct cfs_rq *cfs_rq) {}
> > > +static void task_throttle_setup_work(struct task_struct *p) {}
> > >  
> > >  static inline int cfs_rq_throttled(struct cfs_rq *cfs_rq)
> > >  {
> > > @@ -8851,6 +8913,7 @@ static struct task_struct *pick_task_fair(struct rq *rq)
> > >  {
> > >  	struct sched_entity *se;
> > >  	struct cfs_rq *cfs_rq;
> > > +	struct task_struct *p;
> > >  
> > >  again:
> > >  	cfs_rq = &rq->cfs;
> > > @@ -8871,7 +8934,14 @@ static struct task_struct *pick_task_fair(struct rq *rq)
> > >  		cfs_rq = group_cfs_rq(se);
> > >  	} while (cfs_rq);
> > >  
> > > -	return task_of(se);
> > > +	p = task_of(se);
> > > +	if (throttled_hierarchy(cfs_rq_of(se))) {
> > > +		/* Shuold not happen for now */
> > 
> > Typo: s/Shuold/Should
> >
> 
> Ah right, thanks.
> 
> > Btw: Is there a trace point in place when throttling/unthrottling
> > happens? Would be nice to see that in a trace, but might be that I
> > missed those events in my configuration up to now.
> 
> There is no trace point for throttle related events yet.
> 
> I tried to see if bpftrace can do the job but unfortunately, bpftrace
> does not allow variable as array indice so I failed to get cfs_rq when
> probing tg_throttle_down(tg, data)...
> 

@ajor from IRC helped me solve this problem so I think the following
bpftrace script can somehow serve the purpose before trace events for
throttle/unthrottle are supported:

kfunc:tg_throttle_down
{
	$rq = (struct rq *)args->data;
	$cpu = $rq->cpu;
	$cfs_rq = *(args->tg->cfs_rq + $cpu);

	if ($cfs_rq->throttle_count == 0) {
		printf("%llu cfs_rq %s/%d throttled\n", nsecs, str(args->tg->css.cgroup->kn->name), $cpu);
	}
}

kfunc:tg_unthrottle_up
{
	$rq = (struct rq *)args->data;
	$cpu = $rq->cpu;
	$cfs_rq = *(args->tg->cfs_rq + $cpu);

	if ($cfs_rq->throttle_count == 1) {
		printf("%llu cfs_rq %s/%d unthrottled\n", nsecs, str(args->tg->css.cgroup->kn->name), $cpu);
	}
}