From mboxrd@z Thu Jan 1 00:00:00 1970 From: Peter Zijlstra Subject: Re: [PATCH v2 6/9] perf,tracing: Prepare the perf-trace interface for RCU changes Date: Thu, 13 Feb 2020 09:29:51 +0100 Message-ID: <20200213082951.GK14897@hirez.programming.kicks-ass.net> References: <20200212210139.382424693@infradead.org> <20200212210750.142334759@infradead.org> <20200212232830.GB170680@google.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from bombadil.infradead.org ([198.137.202.133]:47938 "EHLO bombadil.infradead.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726232AbgBMIaM (ORCPT ); Thu, 13 Feb 2020 03:30:12 -0500 Content-Disposition: inline In-Reply-To: <20200212232830.GB170680@google.com> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Joel Fernandes Cc: linux-kernel@vger.kernel.org, linux-arch@vger.kernel.org, rostedt@goodmis.org, mingo@kernel.org, gregkh@linuxfoundation.org, gustavo@embeddedor.com, tglx@linutronix.de, paulmck@kernel.org, josh@joshtriplett.org, mathieu.desnoyers@efficios.com, jiangshanlai@gmail.com, "Steven Rostedt (VMware)" On Wed, Feb 12, 2020 at 06:28:30PM -0500, Joel Fernandes wrote: > On Wed, Feb 12, 2020 at 10:01:45PM +0100, Peter Zijlstra wrote: > > The tracepoint interface will stop providing regular RCU context; make > > sure we do it ourselves, since perf makes use of regular RCU protected > > data. > > > > Suggested-by: Steven Rostedt (VMware) > > Suggested-by: Mathieu Desnoyers > > Signed-off-by: Peter Zijlstra (Intel) > > --- > > kernel/events/core.c | 5 +++++ > > 1 file changed, 5 insertions(+) > > > > --- a/kernel/events/core.c > > +++ b/kernel/events/core.c > > @@ -8950,6 +8950,7 @@ void perf_tp_event(u16 event_type, u64 c > > { > > struct perf_sample_data data; > > struct perf_event *event; > > + unsigned long rcu_flags; > > The flags are not needed I guess, if you agree on not using in_nmi() in > trace_rcu_enter(). Even then we need to store the state: 'didn't do nothing' vs 'did call rcu_needs_to_wake_up_and_pay_attention_noaw'. That is, we only need to do something (expensive!) when !rcu_is_watching().