[PATCH 3/3] rcu: Keep qsmaskinitnext fresh for rcu_boost_kthread_setaffinity()

[Pingfan Liu <kernelfans@gmail.com>]

rcutree_online_cpu() is concurrent, so there is the following race
scene:

        CPU 1                               CPU2
mask_old = rcu_rnp_online_cpus(rnp);
...

                                   mask_new = rcu_rnp_online_cpus(rnp);
                                   ...
                                   set_cpus_allowed_ptr(t, cm);

set_cpus_allowed_ptr(t, cm);

Consequently, the old mask will overwrite the new mask in the task's cpus_ptr.

Since there is a mutex ->boost_kthread_mutex, using it to build an
order, then the latest ->qsmaskinitnext will be fetched for updating cpus_ptr.

Notes about the cpu teardown: The cpu hot-removing initiator executes
rcutree_dead_cpu() for the teardown cpu. If in future, an initiator can
hot-remove several cpus at a time, it executes rcutree_dead_cpu() in
series for each cpu. So it is still race-free without the mutex.  But
considering this is a cold path, applying that redundant mutex is
harmless.

Signed-off-by: Pingfan Liu <kernelfans@gmail.com>
Cc: "Paul E. McKenney" <paulmck@kernel.org>
Cc: David Woodhouse <dwmw@amazon.co.uk>
Cc: Frederic Weisbecker <frederic@kernel.org>
Cc: Neeraj Upadhyay <quic_neeraju@quicinc.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Lai Jiangshan <jiangshanlai@gmail.com>
Cc: Joel Fernandes <joel@joelfernandes.org>
Cc: "Jason A. Donenfeld" <Jason@zx2c4.com>
To: rcu@vger.kernel.org
---
 kernel/rcu/tree_plugin.h | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

diff --git a/kernel/rcu/tree_plugin.h b/kernel/rcu/tree_plugin.h
index 0bf6de185af5..b868ac6c6ac8 100644
--- a/kernel/rcu/tree_plugin.h
+++ b/kernel/rcu/tree_plugin.h
@@ -1223,7 +1223,7 @@ static void rcu_spawn_one_boost_kthread(struct rcu_node *rnp)
 static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp)
 {
 	struct task_struct *t = rnp->boost_kthread_task;
-	unsigned long mask = rcu_rnp_online_cpus(rnp);
+	unsigned long mask;
 	cpumask_var_t cm;
 	int cpu;
 
@@ -1232,6 +1232,11 @@ static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp)
 	if (!zalloc_cpumask_var(&cm, GFP_KERNEL))
 		return;
 	mutex_lock(&rnp->boost_kthread_mutex);
+	/*
+	 * Relying on the lock to serialize, so the latest qsmaskinitnext is for
+	 * cpus_ptr.
+	 */
+	mask = rcu_rnp_online_cpus(rnp);
 	for_each_leaf_node_possible_cpu(rnp, cpu)
 		if ((mask & leaf_node_cpu_bit(rnp, cpu)))
 			cpumask_set_cpu(cpu, cm);
-- 
2.31.1