rcu: handle unstable rdp in rcu_read_unlock_strict()

rcu_read_unlock_strict() can be called with preemption enabled
which can make for an unstable rdp and a racy norm value.

Fix this by dropping the preempt-count in __rcu_read_unlock()
after the call to rcu_read_unlock_strict(), adjusting the
preempt-count check appropriately.

Suggested-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Ankur Arora <ankur.a.arora@oracle.com>
Reviewed-by: Frederic Weisbecker <frederic@kernel.org>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Boqun Feng <boqun.feng@gmail.com>

Author: terminus

include/linux/rcupdate.h

@@ -95,9 +95,9 @@ static inline void __rcu_read_lock(void)
 
 static inline void __rcu_read_unlock(void)
 {
-	preempt_enable();
 	if (IS_ENABLED(CONFIG_RCU_STRICT_GRACE_PERIOD))
 		rcu_read_unlock_strict();
+	preempt_enable();
 }
 
 static inline int rcu_preempt_depth(void)

kernel/rcu/tree_plugin.h

@@ -833,8 +833,17 @@ void rcu_read_unlock_strict(void)
 {
 	struct rcu_data *rdp;
 
-	if (irqs_disabled() || preempt_count() || !rcu_state.gp_kthread)
+	if (irqs_disabled() || in_atomic_preempt_off() || !rcu_state.gp_kthread)
 		return;
+
+	/*
+	 * rcu_report_qs_rdp() can only be invoked with a stable rdp and
+	 * from the local CPU.
+	 *
+	 * The in_atomic_preempt_off() check ensures that we come here holding
+	 * the last preempt_count (which will get dropped once we return to
+	 * __rcu_read_unlock().
+	 */
 	rdp = this_cpu_ptr(&rcu_data);
 	rdp->cpu_no_qs.b.norm = false;
 	rcu_report_qs_rdp(rdp);