Re: [PATCH] tracing/sched: Check preempt_count() for current when reading task->state

[Ingo Molnar <mingo@kernel.org>]

* Steven Rostedt <rostedt@goodmis.org> wrote:

> On Thu, 11 Dec 2014 07:38:11 +0100
> Ingo Molnar <mingo@kernel.org> wrote:
> 
> > > Cc: stable@vger.kernel.org # 3.13+
> > > Fixes: 01028747559a "sched: Create more preempt_count accessors"
> > > Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
> > > ---
> > >  include/trace/events/sched.h | 6 +++++-
> > >  1 file changed, 5 insertions(+), 1 deletion(-)
> > > 
> > > diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h
> > > index 0a68d5ae584e..13fbadcc172b 100644
> > > --- a/include/trace/events/sched.h
> > > +++ b/include/trace/events/sched.h
> > > @@ -97,10 +97,14 @@ static inline long __trace_sched_switch_state(struct task_struct *p)
> > >  	long state = p->state;
> > >  
> > >  #ifdef CONFIG_PREEMPT
> > > +	unsigned long pc;
> > > +
> > > +	pc = (p == current) ? preempt_count() : task_preempt_count(p);
> > > +
> > >  	/*
> > >  	 * For all intents and purposes a preempted task is a running task.
> > >  	 */
> > > -	if (task_preempt_count(p) & PREEMPT_ACTIVE)
> > > +	if (pc & PREEMPT_ACTIVE)
> > >  		state = TASK_RUNNING | TASK_STATE_MAX;
> > 
> > I really don't like the overhead around here.
> 
> Hi Ingo!
> 
> What overhead are you worried about? Note, this is in the 
> schedule tracepoint and does not affect the scheduler itself 
> (as long as the tracepoint is not enabled).

Scheduler tracepoints are pretty popular, so I'm worried about 
their complexity when they are activated.

> I'm also thinking that as long as "prev" is always guaranteed 
> to be "current" we can remove the check and just use 
> preempt_count() always. But I'm worried that we can't 
> guaranteed that.

You could add a WARN_ON_ONCE() or so to double check that 
assumption?

> What other ideas do you have? Because wrong data is worse than 
> the overhead of the above code. If Thomas taught me anything, 
> it's that!

My idea is to have simpler, yet correct code. And ponies!

Thanks,

	Ingo