From: Ethan Jackson <ethan@nicira.com>
Date: Tue, 2 Apr 2013 19:32:22 +0000 (-0700)
Subject: ofproto-dpif: Don't rate limit facet_learn() with fin_timeouts.
X-Git-Tag: sliver-openvswitch-1.10.90-1~10^2~1
X-Git-Url: http://git.onelab.eu/?p=sliver-openvswitch.git;a=commitdiff_plain;h=62bd7349adaf5ee740cd9da58305f7523d7c9e6e

ofproto-dpif: Don't rate limit facet_learn() with fin_timeouts.

In the standard case, rate limiting facet_learn() to once ever
500ms, makes sense.  The worst that can happen is a learning entry
is expired half a second to early.  However, when using
fin_timeouts, we really need react quickly to delete the newly
stale flow.

Bug #15915.
Signed-off-by: Ethan Jackson <ethan@nicira.com>
---

diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
index 32ed0919a..47830c19e 100644
--- a/ofproto/ofproto-dpif.c
+++ b/ofproto/ofproto-dpif.c
@@ -4585,13 +4585,14 @@ facet_learn(struct facet *facet)
     struct ofproto_dpif *ofproto = ofproto_dpif_cast(facet->rule->up.ofproto);
     struct subfacet *subfacet= CONTAINER_OF(list_front(&facet->subfacets),
                                             struct subfacet, list_node);
+    long long int now = time_msec();
     struct action_xlate_ctx ctx;
 
-    if (time_msec() < facet->learn_rl) {
+    if (!facet->has_fin_timeout && now < facet->learn_rl) {
         return;
     }
 
-    facet->learn_rl = time_msec() + 500;
+    facet->learn_rl = now + 500;
 
     if (!facet->has_learn
         && !facet->has_normal