Messages in this thread | | | Date | Fri, 6 Sep 2013 08:41:04 -0400 | From | Steven Rostedt <> | Subject | Re: [PATCH v8 01/10] tracing: Add support for SOFT_DISABLE to syscall events |
| |
On Fri, 06 Sep 2013 21:22:15 +0900 Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com> wrote:
> > And AFAICS, the tracepoint (on which the syscall tracer based) > call-site uses rcu_read_lock_sched_notrace() instead of rcu_read_lock(), > in that case, I think we should use synchronize_sched(). is that wrong? >
Ah, no, you are correct. I was thinking that the tracepoints used rcu_read_lock(), but looking at the code, it's rcu_read_lock_sched(). Thus a synchronize_sched() *is* required.
Thanks, and please ignore my other comments about using "synchronize_rcu()".
-- Steve
| |