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:   Fri, 19 Oct 2018 05:07:58 -0700
From:   "Paul E. McKenney" <paulmck@linux.ibm.com>
To:     Joel Fernandes <joel@joelfernandes.org>
Cc:     Steven Rostedt <rostedt@goodmis.org>,
        Nikolay Borisov <nborisov@suse.com>,
        linux-kernel@vger.kernel.org, Jonathan Corbet <corbet@lwn.net>,
        Josh Triplett <josh@joshtriplett.org>,
        Lai Jiangshan <jiangshanlai@gmail.com>,
        linux-doc@vger.kernel.org,
        Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Subject: Re: [PATCH RFC] doc: rcu: remove obsolete (non-)requirement about
 disabling preemption
Reply-To: paulmck@linux.ibm.com
References: <20181017203324.GS2674@linux.ibm.com>
 <20181018020751.GB99677@joelaf.mtv.corp.google.com>
 <20181018144637.GD2674@linux.ibm.com>
 <20181019000350.GB89903@joelaf.mtv.corp.google.com>
 <20181019001932.GR2674@linux.ibm.com>
 <20181019012645.GC89903@joelaf.mtv.corp.google.com>
 <20181018215035.5e8ff553@vmware.local.home>
 <20181019022529.GA155753@joelaf.mtv.corp.google.com>
 <20181018225223.42641c73@vmware.local.home>
 <20181019035844.GA141835@joelaf.mtv.corp.google.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20181019035844.GA141835@joelaf.mtv.corp.google.com>
User-Agent: Mutt/1.5.21 (2010-09-15)
Message-Id: <20181019120758.GV2674@linux.ibm.com>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Thu, Oct 18, 2018 at 08:58:44PM -0700, Joel Fernandes wrote:
> On Thu, Oct 18, 2018 at 10:52:23PM -0400, Steven Rostedt wrote:
> > On Thu, 18 Oct 2018 19:25:29 -0700
> > Joel Fernandes <joel@joelfernandes.org> wrote:
> > 
> > > On Thu, Oct 18, 2018 at 09:50:35PM -0400, Steven Rostedt wrote:
> > > > On Thu, 18 Oct 2018 18:26:45 -0700
> > > > Joel Fernandes <joel@joelfernandes.org> wrote:
> > > >   
> > > > > Yes, local_irq_restore is light weight, and does not check for reschedules.
> > > > > 
> > > > > I was thinking of case where ksoftirqd is woken up, but does not run unless
> > > > > we set the NEED_RESCHED flag. But that should get set anyway since probably
> > > > > ksoftirqd is of high enough priority than the currently running task..
> > > > > 
> > > > > Roughly speaking the scenario could be something like:
> > > > > 
> > > > > rcu_read_lock();
> > > > >                  <-- IPI comes in for the expedited GP, sets exp_hint
> > > > > local_irq_disable();
> > > > > // do a bunch of stuff
> > > > > rcu_read_unlock();   <-- This calls the rcu_read_unlock_special which raises
> > > > >                          the soft irq, and wakesup softirqd.  
> > > > 
> > > > If softirqd is of higher priority than the current running task, then
> > > > the try_to_wake_up() will set NEED_RESCHED of the current task here.
> > > >   
> > > 
> > > Yes, only *if*. On my system, ksoftirqd is CFS nice 0. I thought expedited
> > > grace periods are quite important and they should complete quickly which is
> > > the whole reason for interrupting rcu read sections with an IPI and stuff.
> > > IMO there should be no harm in setting NEED_RESCHED unconditionally anyway
> > > for possible benefit of systems where the ksoftirqd is not of higher priority
> > > than the currently running task, and we need to run it soon on the CPU. But
> > > I'm Ok with whatever Paul and you want to do here.
> > 
> > 
> > Setting NEED_RESCHED unconditionally wont help. Because even if we call
> > schedule() ksoftirqd will not be scheduled! If it's CFS nice 0, and the
> > current task still has quota to run, if you call schedule, you'll just
> > waste time calculating that the current task should still be running.
> > It's equivalent to calling yield() (which is why we removed all yield()
> > users in the kernel, because *all* of them were buggy!). This is *why*
> > it only calls schedule *if* softirqd is of higher priority.
> 
> Yes, ok. you are right the TTWU path should handle setting the NEED_RESCHED
> flag or not and unconditionally setting it does not get us anything. I had to
> go through the code a bit since it has been a while since I explored it.
> 
> So Paul, I'm Ok with your latest patch for the issue we discussed and don't
> think much more can be done barring raising of ksofitrqd priorities :-) So I
> guess the synchronize_rcu_expedited will just cope with the deal between
> local_irq_enable and the next scheduling point.. :-)

Thank you both!

Indeed, real-time systems need to be configured carefully, especially if
you are crazy enough to run them under high load.  I interpreted "Ok with
your latest patch" as an Acked-by, but please let me know if that is a
misinterpretation.

							Thanx, Paul