Re: [RFC PATCH] tracing: add sched_prio_update

2016-08-11 Thread Peter Zijlstra
On Mon, Jul 04, 2016 at 03:46:04PM -0400, Julien Desfossez wrote: > The effective priority of running threads can also be temporarily > changed in the PI code, but a dedicated tracepoint is already in place > to cover this case. So while we have that tracepoint its not really all that useful. I

Re: [RFC PATCH] tracing: add sched_prio_update

2016-08-11 Thread Peter Zijlstra
On Mon, Jul 04, 2016 at 03:46:04PM -0400, Julien Desfossez wrote: > The effective priority of running threads can also be temporarily > changed in the PI code, but a dedicated tracepoint is already in place > to cover this case. So while we have that tracepoint its not really all that useful. I

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-15 Thread Daniel Bristot de Oliveira
On 07/06/2016 10:53 AM, Julien Desfossez wrote: >> But still, it's a >> > rather hefty tracepoint (lots of fields), probably want to keep from >> > adding comm too. > Yes, I agree we can remove the comm field, it is easy to get from the > previous sched_switch. > Sorry for the delay. I do liked

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-15 Thread Daniel Bristot de Oliveira
On 07/06/2016 10:53 AM, Julien Desfossez wrote: >> But still, it's a >> > rather hefty tracepoint (lots of fields), probably want to keep from >> > adding comm too. > Yes, I agree we can remove the comm field, it is easy to get from the > previous sched_switch. > Sorry for the delay. I do liked

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-06 Thread Steven Rostedt
On Wed, 6 Jul 2016 09:53:24 -0400 Julien Desfossez wrote: > > I'm not convinced this should be needed. I hate adding back to back > > tracepoints. > > Indeed, having two tracepoints back to back is not pretty. We placed it > here to get the priority of the newly

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-06 Thread Steven Rostedt
On Wed, 6 Jul 2016 09:53:24 -0400 Julien Desfossez wrote: > > I'm not convinced this should be needed. I hate adding back to back > > tracepoints. > > Indeed, having two tracepoints back to back is not pretty. We placed it > here to get the priority of the newly created threads. Maybe a

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-06 Thread Julien Desfossez
On 06-Jul-2016 09:13:25 AM, Steven Rostedt wrote: > On Tue, 5 Jul 2016 21:50:34 + (UTC) > Mathieu Desnoyers wrote: > > > > > > >> + > > >> +TP_PROTO(struct task_struct *tsk), > > >> + > > >> +TP_ARGS(tsk), > > >> + > > >> +

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-06 Thread Julien Desfossez
On 06-Jul-2016 09:13:25 AM, Steven Rostedt wrote: > On Tue, 5 Jul 2016 21:50:34 + (UTC) > Mathieu Desnoyers wrote: > > > > > > >> + > > >> +TP_PROTO(struct task_struct *tsk), > > >> + > > >> +TP_ARGS(tsk), > > >> + > > >> +TP_STRUCT__entry( > > >> +

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-06 Thread Steven Rostedt
On Tue, 5 Jul 2016 21:50:34 + (UTC) Mathieu Desnoyers wrote: > > > >> + > >> + TP_PROTO(struct task_struct *tsk), > >> + > >> + TP_ARGS(tsk), > >> + > >> + TP_STRUCT__entry( > >> + __array( char, comm, TASK_COMM_LEN ) > > > > I could

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-06 Thread Steven Rostedt
On Tue, 5 Jul 2016 21:50:34 + (UTC) Mathieu Desnoyers wrote: > > > >> + > >> + TP_PROTO(struct task_struct *tsk), > >> + > >> + TP_ARGS(tsk), > >> + > >> + TP_STRUCT__entry( > >> + __array( char, comm, TASK_COMM_LEN ) > > > > I could imagine this being a high frequency

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-05 Thread Mathieu Desnoyers
- On Jul 5, 2016, at 11:19 AM, rostedt rost...@goodmis.org wrote: > On Mon, 4 Jul 2016 15:46:04 -0400 > Julien Desfossez wrote: > > >> diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h >> index 9b90c57..fcb0f29 100644 >> ---

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-05 Thread Mathieu Desnoyers
- On Jul 5, 2016, at 11:19 AM, rostedt rost...@goodmis.org wrote: > On Mon, 4 Jul 2016 15:46:04 -0400 > Julien Desfossez wrote: > > >> diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h >> index 9b90c57..fcb0f29 100644 >> --- a/include/trace/events/sched.h >> +++

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-05 Thread Steven Rostedt
On Mon, 4 Jul 2016 15:46:04 -0400 Julien Desfossez wrote: > diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h > index 9b90c57..fcb0f29 100644 > --- a/include/trace/events/sched.h > +++ b/include/trace/events/sched.h > @@ -8,6 +8,34 @@ > #include

Re: [RFC PATCH] tracing: add sched_prio_update

2016-07-05 Thread Steven Rostedt
On Mon, 4 Jul 2016 15:46:04 -0400 Julien Desfossez wrote: > diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h > index 9b90c57..fcb0f29 100644 > --- a/include/trace/events/sched.h > +++ b/include/trace/events/sched.h > @@ -8,6 +8,34 @@ > #include > #include > >

[RFC PATCH] tracing: add sched_prio_update

2016-07-04 Thread Julien Desfossez
This tracepoint allows to keep track of all priority changes of a task. It outputs the scheduling policy, the nice value, the rt_priority and the deadline-related attributes (dl_runtime, dl_deadline and dl_period). It is emitted in the code path of the sched_setscheduler, sched_setattr,

[RFC PATCH] tracing: add sched_prio_update

2016-07-04 Thread Julien Desfossez
This tracepoint allows to keep track of all priority changes of a task. It outputs the scheduling policy, the nice value, the rt_priority and the deadline-related attributes (dl_runtime, dl_deadline and dl_period). It is emitted in the code path of the sched_setscheduler, sched_setattr,