Re: [4.4-rt PATCH] trace: use rcuidle version for preemptoff_hist trace point

[Yang Shi]

On 3/7/2016 10:00 AM, Sebastian Andrzej Siewior wrote:
> * Yang Shi | 2016-02-23 13:23:23 [-0800]:
>
> > I recall the rcuidle version is used by 4.1-rt, but not sure why it is dropped
> > in 4.4-rt. It looks such fix is still needed.
> I don't recall while I removed it. It was durring v4.1 -> v4.4 port. In
> v4.1 we had the idle version only in time_hardirqs_on() + â_off(). You
> introduced it also to start_critical_timings() + stop_â(). Is this
> required?

Yes, the test showed the same problem in start{stop}_critical_timings too.

> In the meantime I bring back the empty macro of
> trace_preemptirqsoff_hist_rcuidle() back in -RT sine it breaks compile
> latest v4.4.3-RT9

Yes, sorry to forget to bring this in too. The 4.1 patch was submitted 
by me, but forgot to have it in 4.4.

Yang

> > kernel/trace/trace_irqsoff.c | 8 ++++----
> > 1 file changed, 4 insertions(+), 4 deletions(-)
> >
> > diff --git a/kernel/trace/trace_irqsoff.c b/kernel/trace/trace_irqsoff.c
> > index 36e584f..069942c 100644
> > --- a/kernel/trace/trace_irqsoff.c
> > +++ b/kernel/trace/trace_irqsoff.c
> > @@ -421,13 +421,13 @@ void start_critical_timings(void)
> > {
> > 	if (preempt_trace() || irq_trace())
> > 		start_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
> > -	trace_preemptirqsoff_hist(TRACE_START, 1);
> > +	trace_preemptirqsoff_hist_rcuidle(TRACE_START, 1);
> > }
> > EXPORT_SYMBOL_GPL(start_critical_timings);
> >
> > void stop_critical_timings(void)
> > {
> > -	trace_preemptirqsoff_hist(TRACE_STOP, 0);
> > +	trace_preemptirqsoff_hist_rcuidle(TRACE_STOP, 0);
> > 	if (preempt_trace() || irq_trace())
> > 		stop_critical_timing(CALLER_ADDR0, CALLER_ADDR1);
> > }
> Sebastian