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:   Mon, 26 Aug 2019 09:29:45 -0700
From:   "Paul E. McKenney" <paulmck@linux.ibm.com>
To:     Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc:     Joel Fernandes <joel@joelfernandes.org>,
        Scott Wood <swood@redhat.com>, linux-rt-users@vger.kernel.org,
        linux-kernel@vger.kernel.org, Thomas Gleixner <tglx@linutronix.de>,
        Peter Zijlstra <peterz@infradead.org>,
        Juri Lelli <juri.lelli@redhat.com>,
        Clark Williams <williams@redhat.com>
Subject: Re: [PATCH RT v2 2/3] sched: migrate_enable: Use sleeping_lock to
 indicate involuntary sleep
Message-ID: <20190826162945.GE28441@linux.ibm.com>
Reply-To: paulmck@linux.ibm.com
References: <20190821231906.4224-1-swood@redhat.com>
 <20190821231906.4224-3-swood@redhat.com>
 <20190823162024.47t7br6ecfclzgkw@linutronix.de>
 <433936e4c720e6b81f9b297fefaa592fd8a961ad.camel@redhat.com>
 <20190824031014.GB2731@google.com>
 <20190826152523.dcjbsgyyir4zjdol@linutronix.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20190826152523.dcjbsgyyir4zjdol@linutronix.de>
User-Agent: Mutt/1.5.21 (2010-09-15)
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

On Mon, Aug 26, 2019 at 05:25:23PM +0200, Sebastian Andrzej Siewior wrote:
> On 2019-08-23 23:10:14 [-0400], Joel Fernandes wrote:
> > On Fri, Aug 23, 2019 at 02:28:46PM -0500, Scott Wood wrote:
> > > On Fri, 2019-08-23 at 18:20 +0200, Sebastian Andrzej Siewior wrote:
> > > > 
> > > > this looks like an ugly hack. This sleeping_lock_inc() is used where we
> > > > actually hold a sleeping lock and schedule() which is okay. But this
> > > > would mean we hold a RCU lock and schedule() anyway. Is that okay?
> > > 
> > > Perhaps the name should be changed, but the concept is the same -- RT-
> > > specific sleeping which should be considered involuntary for the purpose of
> > > debug checks.  Voluntary sleeping is not allowed in an RCU critical section
> > > because it will break the critical section on certain flavors of RCU, but
> > > that doesn't apply to the flavor used on RT.  Sleeping for a long time in an
> > > RCU critical section would also be a bad thing, but that also doesn't apply
> > > here.
> > 
> > I think the name should definitely be changed. At best, it is super confusing to
> > call it "sleeping_lock" for this scenario. In fact here, you are not even
> > blocking on a lock.
> > 
> > Maybe "sleeping_allowed" or some such.
> 
> The mechanism that is used here may change in future. I just wanted to
> make sure that from RCU's side it is okay to schedule here.

Good point.

The effect from RCU's viewpoint will be to split any non-rcu_read_lock()
RCU read-side critical section at this point.  This alrady happens in a
few places, for example, rcu_note_context_switch() constitutes an RCU
quiescent state despite being invoked with interrupts disabled (as is
required!).  The __schedule() function just needs to understand (and does
understand) that the RCU read-side critical section that would otherwise
span that call to rcu_node_context_switch() is split in two by that call.

However, if this was instead an rcu_read_lock() critical section within
a PREEMPT=y kernel, then if a schedule() occured within stop_one_task(),
RCU would consider that critical section to be preempted.  This means
that any RCU grace period that is blocked by this RCU read-side critical
section would remain blocked until stop_one_cpu() resumed, returned,
and so on until the matching rcu_read_unlock() was reached.  In other
words, RCU would consider that RCU read-side critical section to span
the call to stop_one_cpu() even if stop_one_cpu() invoked schedule().

On the other hand, within a PREEMPT=n kernel, the call to schedule()
would split even an rcu_read_lock() critical section.  Which is why I
asked earlier if sleeping_lock_inc() and sleeping_lock_dec() are no-ops
in !PREEMPT_RT_BASE kernels.  We would after all want the usual lockdep
complaints in that case.

Does that help, or am I missing the point?

							Thanx, Paul