[ovs-dev] [PATCH 1.7] linux: Increase accuracy of ingress_policing_rate at low rates

Thomas Graf tgraf at redhat.com
Fri Jan 25 19:13:55 UTC 2013


The current method of calculating the ingress policer rate
can lead to inaccuracy if ingress_policing_rate is set to
a smallish values because the rate is divided by 8 first
which causes rounding errors.

Signed-off-by: Thomas Graf <tgraf at redhat.com>
---
 lib/netdev-linux.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/netdev-linux.c b/lib/netdev-linux.c
index 4d2f3ac..5fe3e2a 100644
--- a/lib/netdev-linux.c
+++ b/lib/netdev-linux.c
@@ -3686,7 +3686,7 @@ tc_add_policer(struct netdev *netdev, int kbits_rate, int kbits_burst)
     memset(&tc_police, 0, sizeof tc_police);
     tc_police.action = TC_POLICE_SHOT;
     tc_police.mtu = mtu;
-    tc_fill_rate(&tc_police.rate, kbits_rate/8 * 1000, mtu);
+    tc_fill_rate(&tc_police.rate, (kbits_rate * 1000)/8, mtu);
     tc_police.burst = tc_bytes_to_ticks(tc_police.rate.rate,
                                         kbits_burst * 1024);
 
-- 
1.7.11.7




More information about the dev mailing list