[PATCH 2/4] rcu: Add warning to ensure rcu_seq_done_exact() is working

[Joel Fernandes <joelagnelf@nvidia.com>]

The previous patch improved the rcu_seq_done_exact() function by adding
a meaningful constant for the guardband.

Ensure that this is working for the future by a quick check during
rcu_gp_init().

Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
---
 kernel/rcu/tree.c | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/kernel/rcu/tree.c b/kernel/rcu/tree.c
index 659f83e71048..6b1bb85b2a56 100644
--- a/kernel/rcu/tree.c
+++ b/kernel/rcu/tree.c
@@ -1798,6 +1798,7 @@ static noinline_for_stack bool rcu_gp_init(void)
 	struct rcu_data *rdp;
 	struct rcu_node *rnp = rcu_get_root();
 	bool start_new_poll;
+	unsigned long old_gp_seq;
 
 	WRITE_ONCE(rcu_state.gp_activity, jiffies);
 	raw_spin_lock_irq_rcu_node(rnp);
@@ -1825,7 +1826,12 @@ static noinline_for_stack bool rcu_gp_init(void)
 	 */
 	start_new_poll = rcu_sr_normal_gp_init();
 	/* Record GP times before starting GP, hence rcu_seq_start(). */
+	old_gp_seq = rcu_state.gp_seq;
 	rcu_seq_start(&rcu_state.gp_seq);
+	/* Ensure that rcu_seq_done_exact() guardband doesn't give false positives. */
+	WARN_ON_ONCE(IS_ENABLED(CONFIG_PROVE_RCU) &&
+		     rcu_seq_done_exact(&old_gp_seq, rcu_seq_snap(&rcu_state.gp_seq)));
+
 	ASSERT_EXCLUSIVE_WRITER(rcu_state.gp_seq);
 	trace_rcu_grace_period(rcu_state.name, rcu_state.gp_seq, TPS("start"));
 	rcu_poll_gp_seq_start(&rcu_state.gp_seq_polled_snap);
-- 
2.43.0