rcu-tasks: Enclose task-list scan in rcu_read_lock()

commit f747c7e upstream.

The rcu_tasks_trace_postgp() function uses for_each_process_thread()
to scan the task list without the benefit of RCU read-side protection,
which can result in use-after-free errors on task_struct structures.
This error was missed because the TRACE01 rcutorture scenario enables
lockdep, but also builds with CONFIG_PREEMPT_NONE=y.  In this situation,
preemption is disabled everywhere, so lockdep thinks everywhere can
be a legitimate RCU reader.  This commit therefore adds the needed
rcu_read_lock() and rcu_read_unlock().

Note that this bug can occur only after an RCU Tasks Trace CPU stall
warning, which by default only happens after a grace period has extended
for ten minutes (yes, not a typo, minutes).

Fixes: 4593e77 ("rcu-tasks: Add stall warnings for RCU Tasks Trace")
Cc: Alexei Starovoitov <alexei.starovoitov@gmail.com>
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: <bpf@vger.kernel.org>
Cc: <stable@vger.kernel.org> # 5.7.x
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

kernel/rcu/tasks.h

@@ -1078,9 +1078,11 @@ static void rcu_tasks_trace_postgp(struct rcu_tasks *rtp)
 		if (ret)
 			break;  // Count reached zero.
 		// Stall warning time, so make a list of the offenders.
+		rcu_read_lock();
 		for_each_process_thread(g, t)
 			if (READ_ONCE(t->trc_reader_special.b.need_qs))
 				trc_add_holdout(t, &holdouts);
+		rcu_read_unlock();
 		firstreport = true;
 		list_for_each_entry_safe(t, g, &holdouts, trc_holdout_list) {
 			if (READ_ONCE(t->trc_reader_special.b.need_qs))