Re: [RFC] Make need_resched() return true when rcu_urgent_qs requested

["Paul E. McKenney" <paulmck@linux.vnet.ibm.com>]

On Wed, Jul 11, 2018 at 07:43:03AM -0700, Paul E. McKenney wrote:
> On Wed, Jul 11, 2018 at 03:23:45PM +0100, David Woodhouse wrote:
> > 
> > 
> > On Mon, 2018-07-09 at 15:08 -0700, Paul E. McKenney wrote:
> > > index f9c0ca2ccf0c..3350ece366ab 100644
> > > --- a/kernel/rcu/tree.c
> > > +++ b/kernel/rcu/tree.c
> > > @@ -2839,6 +2839,15 @@ void rcu_check_callbacks(int user)
> > >                 rcu_bh_qs();
> > >         }
> > >         rcu_preempt_check_callbacks();
> > > +       /* The load-acquire pairs with the store-release setting to true. */
> > > +       if (smp_load_acquire(this_cpu_ptr(&rcu_dynticks.rcu_urgent_qs))) {
> > > +               /* Idle and userspace execution already are quiescent states. */
> > > +               if (rcu_is_cpu_rrupt_from_idle() && !user) {
> > 
> > if (idle && !user) seems tautological and... illogical.
> > 
> > If I make it 'if (!rcu_is_cpu_rrput_from_idle() && !user)' it seems to
> > work better. Ripping out my debugging printks now to check that's still
> > true...
> 
> Right you are!  I will step away for a bit to put a paper bag over
> my head...
> 
> > (Also, isn't userspace execution only a quiescent state if NO_HZ_FULL?)
> 
> Userspace execution is a quiescent state in all cases.  However, you
> are quite right that NO_HZ_FULL makes a difference, namely, it allows
> one CPU to reliably determine whether or not some other CPU is
> currently executing either in userspace or in idle.
> 
> Without NO_HZ_FULL, CPUs can only detect their own userspace execution.
> Which is what is happening here because rcu_check_callbacks() is being
> invoked from the scheduling-clock interrupt, which is where the "user"
> parameter comes from.
> 
> So the above code can reliably detect the usermode-execution quiescent
> state because it is always running on the CPU in question.

And here is an updated v4.15 patch with Marius's Reported-by and David's
fix to my lost exclamation point.

							Thanx, Paul

------------------------------------------------------------------------

commit 83c4beae36f2a2b38c1a0fa80538af7ce2477823
Author: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Date:   Mon Jul 9 13:47:30 2018 -0700

    rcu: Make need_resched() respond to urgent RCU-QS needs
    
    The per-CPU rcu_dynticks.rcu_urgent_qs variable communicates an urgent
    need for an RCU quiescent state from the force-quiescent-state processing
    within the grace-period kthread to context switches and to cond_resched().
    Unfortunately, such urgent needs are not communicated to need_resched(),
    which is sometimes used to decide when to invoke cond_resched(), for
    but one example, within the KVM vcpu_run() function.  As of v4.15, this
    can result in synchronize_sched() being delayed by up to ten seconds,
    which can be problematic, to say nothing of annoying.
    
    This commit therefore checks rcu_dynticks.rcu_urgent_qs from within
    rcu_check_callbacks(), which is invoked from the scheduling-clock
    interrupt handler.  If the current task is not an idle task and is
    not executing in usermode, a context switch is forced, and either way,
    the rcu_dynticks.rcu_urgent_qs variable is set to false.  If the current
    task is an idle task, then RCU's dyntick-idle code will detect the
    quiescent state, so no further action is required.  Similarly, if the
    task is executing in usermode, other code in rcu_check_callbacks() and
    its called functions will report the corresponding quiescent state.
    
    Reported-by: Marius Hillenbrand <mhillenb@amazon.de>
    Reported-by: David Woodhouse <dwmw2@infradead.org>
    Suggested-by: Peter Zijlstra <peterz@infradead.org>
    Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
    [ paulmck: Backported to v4.15.  Probably applies elsewhere. ]

diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
index f9c0ca2ccf0c..de2f91cb2a0c 100644
--- a/kernel/rcu/tree.c
+++ b/kernel/rcu/tree.c
@@ -2839,6 +2839,15 @@ void rcu_check_callbacks(int user)
 		rcu_bh_qs();
 	}
 	rcu_preempt_check_callbacks();
+	/* The load-acquire pairs with the store-release setting to true. */
+	if (smp_load_acquire(this_cpu_ptr(&rcu_dynticks.rcu_urgent_qs))) {
+		/* Idle and userspace execution already are quiescent states. */
+		if (!rcu_is_cpu_rrupt_from_idle() && !user) {
+			set_tsk_need_resched(current);
+			set_preempt_need_resched();
+		}
+		__this_cpu_write(rcu_dynticks.rcu_urgent_qs, false);
+	}
 	if (rcu_pending())
 		invoke_rcu_core();
 	if (user)