srcu: Use rcu_seq_done_exact() for polling API
authorJoel Fernandes <joelagnelf@nvidia.com>
Thu, 13 Feb 2025 16:47:27 +0000 (11:47 -0500)
committerJoel Fernandes <joelagnelf@nvidia.com>
Tue, 8 Apr 2025 18:55:55 +0000 (14:55 -0400)
commitf50ad4b73e1bdf7fb40566650fa11bcb438ad5ce
tree14d890279de338b5f1bfdaf6da7c29da8a1036f9
parent4d949edbc4026eeb81a8f931586cfd4f1de5957e
srcu: Use rcu_seq_done_exact() for polling API

poll_state_synchronize_srcu() uses rcu_seq_done() unlike
poll_state_synchronize_rcu() which uses rcu_seq_done_exact().

The  rcu_seq_done_exact() makes more sense for polling API, as with
this API, there is a higher chance that there is a significant delay
between the get_state..() and poll_state..() calls since a cookie
can be stored and reused at a later time. During such a delay, if
the gp_seq counter progresses more than ULONG_MAX/2 distance, then
poll_state..() may return false for a long time unwantedly.

Fix by using the more accurate rcu_seq_done_exact() API which is
exactly what straight RCU's polling does.

It may make sense, as future work, to add debug code here as well, where
we compare a physical timestamp between get_state..() and poll_state()
calls and yell if significant time has past but the grace period has
still not progressed.

Reviewed-by: Neeraj Upadhyay <Neeraj.Upadhyay@amd.com>
Reviewed-by: Paul E. McKenney <paulmck@kernel.org>
Reviewed-by: Kent Overstreet <kent.overstreet@linux.dev>
Cc: Kent Overstreet <kent.overstreet@linux.dev>
Signed-off-by: Joel Fernandes <joelagnelf@nvidia.com>
kernel/rcu/srcutree.c