bridge: move to workqueue gc

Move the fdb garbage collector to a workqueue which fires at least 10
milliseconds apart and cleans chain by chain allowing for other tasks
to run in the meantime. When having thousands of fdbs the system is much
more responsive. Most importantly remove the need to check if the
matched entry has expired in __br_fdb_get that causes false-sharing and
is completely unnecessary if we cleanup entries, at worst we'll get 10ms
of traffic for that entry before it gets deleted.

Signed-off-by: Nikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

1 files changed, 2 insertions, 2 deletions

diff --git a/net/bridge/br_stp_if.c b/net/bridge/br_stp_if.c
index 6c1e21411125..08341d2aa9c9 100644
--- a/net/bridge/br_stp_if.c
+++ b/net/bridge/br_stp_if.c
@@ -57,7 +57,7 @@ void br_stp_enable_bridge(struct net_bridge *br)
 	spin_lock_bh(&br->lock);
 	if (br->stp_enabled == BR_KERNEL_STP)
 		mod_timer(&br->hello_timer, jiffies + br->hello_time);
-	mod_timer(&br->gc_timer, jiffies + HZ/10);
+	mod_delayed_work(system_long_wq, &br->gc_work, HZ / 10);
 
 	br_config_bpdu_generation(br);
 
@@ -88,7 +88,7 @@ void br_stp_disable_bridge(struct net_bridge *br)
 	del_timer_sync(&br->hello_timer);
 	del_timer_sync(&br->topology_change_timer);
 	del_timer_sync(&br->tcn_timer);
-	del_timer_sync(&br->gc_timer);
+	cancel_delayed_work_sync(&br->gc_work);
 }
 
 /* called under bridge lock */