Re: [RFC PATCH 2/5] sched/events: Introduce cfs_rq load tracking trace event

[Dietmar Eggemann]

On 03/28/2017 07:37 PM, Steven Rostedt wrote:
> On Tue, 28 Mar 2017 13:36:26 -0400
> Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:
>
> > But why play games, and rely on the design of the code? A
> > TRACE_EVENT_CONDTION() is more robust and documents that this
> > tracepoint should not be called when cfs_rq is NULL.
>
> In other words, what are you trying to save for not using the
> TRACE_EVENT_CONDITION()?

IMHO, if we could avoid this

  if(cfs_rq)
    trace_sched_load_cfs_rq(cfs_rq);
  else
    trace_sched_load_se(container_of(sa, struct sched_entity, avg));

in __update_load_avg(), then we can use 'unconditional' TRACE_EVENTs in 
all call-sites:

__update_load_avg{_cfs_rq}(), propagate_entity_load_avg(), 
attach_entity_load_avg(), detach_entity_load_avg() for cfs_rq and

__update_load_avg_blocked_se(), __update_load_avg_se(), 
propagate_entity_load_avg() for se.

[...]