Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
Date:   Tue, 11 Feb 2020 05:23:13 -0800
From:   "Paul E. McKenney" <paulmck@kernel.org>
To:     Peter Zijlstra <peterz@infradead.org>
Cc:     Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
        rostedt <rostedt@goodmis.org>,
        linux-kernel <linux-kernel@vger.kernel.org>,
        Ingo Molnar <mingo@kernel.org>,
        "Joel Fernandes, Google" <joel@joelfernandes.org>,
        Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
        "Gustavo A. R. Silva" <gustavo@embeddedor.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Josh Triplett <josh@joshtriplett.org>,
        Lai Jiangshan <jiangshanlai@gmail.com>
Subject: Re: [PATCH] tracing/perf: Move rcu_irq_enter/exit_irqson() to perf
 trace point hook
Message-ID: <20200211132313.GJ2935@paulmck-ThinkPad-P72>
Reply-To: paulmck@kernel.org
References: <20200210170643.3544795d@gandalf.local.home>
 <576504045.617212.1581381032132.JavaMail.zimbra@efficios.com>
 <20200211120015.GL14914@hirez.programming.kicks-ass.net>
 <20200211130301.GH2935@paulmck-ThinkPad-P72>
 <20200211131637.GS14914@hirez.programming.kicks-ass.net>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20200211131637.GS14914@hirez.programming.kicks-ass.net>
User-Agent: Mutt/1.9.4 (2018-02-28)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Tue, Feb 11, 2020 at 02:16:37PM +0100, Peter Zijlstra wrote:
> On Tue, Feb 11, 2020 at 05:03:01AM -0800, Paul E. McKenney wrote:
> 
> > > It is not the fact that perf issues rcu_read_lock() that is the problem.
> > > As we established yesterday, I can probably remove most rcu_read_lock()
> > > calls from perf today (yay RCU flavour unification).
> > 
> > Glad some aspect of this unification is actually helping you.  ;-)
> 
> rcu_read_lock() is exceedingly cheap though, so I never really worried
> about it. But now that RCU includes RCU-sched (again) we can go and
> remove a bunch of them.
> 
> > > As per nmi_enter() calling rcu_nmi_enter() I've always assumed that NMIs
> > > are fully covered by RCU.
> > > 
> > > If this isn't so, RCU it terminally broken :-)
> > 
> > All RCU can do is respond to calls to rcu_nmi_enter() and rcu_nmi_exit().
> > It has not yet figured out how to force people to add these calls where
> > they are needed.  ;-)
> > 
> > But yes, it would be very nice if architectures arranged things so
> > that all NMI handlers were visible to RCU.  And we no longer have
> > half-interrupts, so maybe there is hope...
> 
> Well,.. you could go back to simply _always_ watching :-)

The idle loop always was unwatched, even back in DYNIX/ptx.  And watching
the idle loop requires waking up idle CPUs, which makes lots of people
quite unhappy.  But it could be done with something sort of like
synchronize_rcu_tasks(), as long as this didn't need to be used in
production on battery-powered systems.

							Thanx, Paul