Received-SPF: pass (google.com: domain of linux-kernel+bounces-79363-linux.lists.archive=gmail.com@vger.kernel.org designates 2604:1380:45d1:ec00::1 as permitted sender) client-ip=2604:1380:45d1:ec00::1;
Date: Fri, 23 Feb 2024 16:28:19 -0800
From: "Paul E. McKenney" <paulmck@kernel.org>
To: Frederic Weisbecker <frederic@kernel.org>
Cc: Boqun Feng <boqun.feng@gmail.com>, linux-kernel@vger.kernel.org,
	rcu@vger.kernel.org, Neeraj Upadhyay <Neeraj.Upadhyay@amd.com>,
	Chen Zhongjin <chenzhongjin@huawei.com>,
	Yang Jihong <yangjihong1@huawei.com>,
	Neeraj Upadhyay <quic_neeraju@quicinc.com>,
	Joel Fernandes <joel@joelfernandes.org>,
	Josh Triplett <josh@joshtriplett.org>,
	Steven Rostedt <rostedt@goodmis.org>,
	Mathieu Desnoyers <mathieu.desnoyers@efficios.com>,
	Lai Jiangshan <jiangshanlai@gmail.com>,
	Zqiang <qiang.zhang1211@gmail.com>
Subject: Re: [PATCH v2 4/6] rcu-tasks: Maintain lists to eliminate
 RCU-tasks/do_exit() deadlocks
Message-ID: <dcea0803-7f9d-4d27-af68-1efb9015dbff@paulmck-laptop>
Reply-To: paulmck@kernel.org
References: <20240217012745.3446231-1-boqun.feng@gmail.com>
 <20240217012745.3446231-5-boqun.feng@gmail.com>
 <ZdiNXmO3wRvmzPsr@lothringen>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <ZdiNXmO3wRvmzPsr@lothringen>

On Fri, Feb 23, 2024 at 01:19:42PM +0100, Frederic Weisbecker wrote:
> On Fri, Feb 16, 2024 at 05:27:39PM -0800, Boqun Feng wrote:
> > -void exit_tasks_rcu_start(void) __acquires(&tasks_rcu_exit_srcu)
> > +void exit_tasks_rcu_start(void)
> >  {
> > -	current->rcu_tasks_idx = __srcu_read_lock(&tasks_rcu_exit_srcu);
> > +	unsigned long flags;
> > +	struct rcu_tasks_percpu *rtpcp;
> > +	struct task_struct *t = current;
> > +
> > +	WARN_ON_ONCE(!list_empty(&t->rcu_tasks_exit_list));
> > +	preempt_disable();
> > +	rtpcp = this_cpu_ptr(rcu_tasks.rtpcpu);
> > +	t->rcu_tasks_exit_cpu = smp_processor_id();
> > +	raw_spin_lock_irqsave_rcu_node(rtpcp, flags);
> > +	if (!rtpcp->rtp_exit_list.next)
> 
> And then you might want to turn that into a WARN_ONCE.

Excellent point, thank you!

I am queueing a separate patch for this with your Reported-by.  (This
change can lag behind, just in case this series needs to go in quickly.)

						Thanx, Paul

> Thanks.
> 
> > +		INIT_LIST_HEAD(&rtpcp->rtp_exit_list);
> > +	list_add(&t->rcu_tasks_exit_list, &rtpcp->rtp_exit_list);
> > +	raw_spin_unlock_irqrestore_rcu_node(rtpcp, flags);
> > +	preempt_enable();
> >  }