Re: [PATCH RFC] tracepoint: Introduce tracepoint callbacks executing with preempt on

[Joel Fernandes]

On Fri, Apr 27, 2018 at 9:13 AM, Steven Rostedt <rostedt@xxxxxxxxxxx> wrote:
> On Fri, 27 Apr 2018 08:57:01 -0700
> "Paul E. McKenney" <paulmck@xxxxxxxxxxxxxxxxxx> wrote:
>
>> > +           if (preempt_on) {                                       \
>> > +                   WARN_ON_ONCE(in_nmi()); /* no srcu from nmi */  \
>>
>> Very good on this check, thank you!
>
> I think you need to return and not call the read lock.
>
>                         if (WARN_ON_ONCE(in_nmi()))
>                                 return;

Cool, I'll do that.

>>
>> > +                   idx = srcu_read_lock(&tracepoint_srcu);         \
>>
>> Hmmm...  Do I need to create a _notrace variant of srcu_read_lock()
>> and srcu_read_unlock()?
>
> I think so.

Oh yes, since otherwise we call into lockdep.

Paul, then I think that's true we'd need srcu _notrace variants that
don't do the rcu_lock_acquire. Sorry for my earlier email saying its
not needed. Thanks,

- Joel