rcu: Don't destroy rcu_torture_boost() callback until it is done
The rcu_torture_boost() cleanup code destroyed debug-objects state before
waiting for the last RCU callback to be invoked, resulting in rare but
very real debug-objects warnings. Move the destruction to after the
waiting to fix this problem.
Signed-off-by: Paul E. McKenney <paul.mckenney@linaro.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
diff --git a/kernel/rcutorture.c b/kernel/rcutorture.c
index 75fca51..6a30ea3f 100644
--- a/kernel/rcutorture.c
+++ b/kernel/rcutorture.c
@@ -796,11 +796,11 @@
/* Clean up and exit. */
VERBOSE_PRINTK_STRING("rcu_torture_boost task stopping");
- destroy_rcu_head_on_stack(&rbi.rcu);
rcutorture_shutdown_absorb("rcu_torture_boost");
while (!kthread_should_stop() || rbi.inflight)
schedule_timeout_uninterruptible(1);
smp_mb(); /* order accesses to ->inflight before stack-frame death. */
+ destroy_rcu_head_on_stack(&rbi.rcu);
return 0;
}
