From mboxrd@z Thu Jan 1 00:00:00 1970 From: Steven Rostedt Subject: Re: [PATCH V2 3/3] sched/deadline: Tracepoints for deadline scheduler Date: Tue, 29 Mar 2016 13:13:43 -0400 Message-ID: <20160329131343.3b2bcf76@gandalf.local.home> References: <14f6caa05f73ceba69eff035ac542cad671552b3.1459182044.git.bristot@redhat.com> <20160329151649.GA12845@twins.programming.kicks-ass.net> <56FAA8FA.5060407@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Cc: Peter Zijlstra , Ingo Molnar , Thomas Gleixner , Juri Lelli , Arnaldo Carvalho de Melo , LKML , linux-rt-users To: Daniel Bristot de Oliveira Return-path: In-Reply-To: <56FAA8FA.5060407@redhat.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-rt-users.vger.kernel.org On Tue, 29 Mar 2016 13:10:34 -0300 Daniel Bristot de Oliveira wrote: > On 03/29/2016 12:16 PM, Peter Zijlstra wrote: > >> + trace_sched_deadline_yield(&rq->curr->dl); > > ouch, it should be trace_sched_deadline_yield(dl_se). It works > as is, but it is really very sad, my bad, sorry. > > >> > dl_se->dl_throttled = 1; > >> > + trace_sched_deadline_throttle(dl_se); > > This is just really very sad. > > Am I missing any other really very sad thing here? Well, we shouldn't have two tracepoints back to back. > > >> > __dequeue_task_dl(rq, curr, 0); > >> > if (unlikely(dl_se->dl_boosted || !start_dl_timer(curr))) > >> > enqueue_task_dl(rq, curr, ENQUEUE_REPLENISH); > >> > @@ -910,6 +917,7 @@ enqueue_dl_entity(struct sched_dl_entity *dl_se, > >> > static void dequeue_dl_entity(struct sched_dl_entity *dl_se) > >> > { > >> > __dequeue_dl_entity(dl_se); > >> > + trace_sched_deadline_block(dl_se); > >> > } > > And that's just not going to happen. > > It will, if a task goes to sleep during the activation, > e.g., when blocking on a system call. For example: > > -0 [007] d..3 78377.688969: sched_switch: prev_comm=swapper/7 prev_pid=0 prev_prio=120 prev_state=R ==> next_comm=b next_pid=18973 next_prio=-1 > b-18973 [007] d..3 78377.688979: sched_deadline_block: now=78377.688976271 deadline=78377.718945137 remaining_runtime=9968866 > b-18973 [007] d..3 78377.688981: sched_switch: prev_comm=b prev_pid=18973 prev_prio=-1 prev_state=S ==> next_comm=swapper/7 next_pid=0 next_prio=120 Why did it go to sleep? The above is still not very useful. What do you mean "blocking on a system call"? -- Steve