Re: Query regarding synchronize_sched_expedited and resched_cpu

From: Paul E. McKenney
Date: Thu Sep 21 2017 - 12:00:58 EST

On Thu, Sep 21, 2017 at 03:59:46PM +0200, Peter Zijlstra wrote:
> On Tue, Sep 19, 2017 at 08:31:26AM -0700, Paul E. McKenney wrote:
> > So I have this one queued. Objections?
> Changelog reads like its whitespace damaged.

It does, now that you mention it. How about the updated version below?

Thanx, Paul


commit c21c9b78182e35eb0e72ef4e3bba3054f26eaaea
Author: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
Date: Mon Sep 18 08:54:40 2017 -0700

sched: Make resched_cpu() unconditional

The current implementation of synchronize_sched_expedited() incorrectly
assumes that resched_cpu() is unconditional, which it is not. This means
that synchronize_sched_expedited() can hang when resched_cpu()'s trylock
fails as follows (analysis by Neeraj Upadhyay):

o CPU1 is waiting for expedited wait to complete:

rdp->exp_dynticks_snap & 0x1 // returns 1 for CPU5
IPI sent to CPU5

ret = swait_event_timeout(rsp->expedited_wq,

expmask = 0x20, CPU 5 in idle path (in cpuidle_enter())

o CPU5 handles IPI and fails to acquire rq lock.

Handles IPI
returns while failing to try lock acquire rq->lock
need_resched is not set

o CPU5 calls rcu_idle_enter() and as need_resched is not set, goes to
idle (schedule() is not called).

o CPU 1 reports RCU stall.

Given that resched_cpu() is now used only by RCU, this commit fixes the
assumption by making resched_cpu() unconditional.

Reported-by: Neeraj Upadhyay <neeraju@xxxxxxxxxxxxxx>
Suggested-by: Neeraj Upadhyay <neeraju@xxxxxxxxxxxxxx>
Signed-off-by: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Acked-by: Steven Rostedt (VMware) <rostedt@xxxxxxxxxxx>

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index cab8c5ec128e..b2281971894c 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -505,8 +505,7 @@ void resched_cpu(int cpu)
struct rq *rq = cpu_rq(cpu);
unsigned long flags;

- if (!raw_spin_trylock_irqsave(&rq->lock, flags))
- return;
+ raw_spin_lock_irqsave(&rq->lock, flags);
raw_spin_unlock_irqrestore(&rq->lock, flags);