The sorted subtable ranking patch introduced a classifier instance per
ingress port with its subtables ranked on the frequency of hits. The PMD
thread can have more classifier instances now and solely depends on the
number of ingress ports currently handled by the pmd thread.
Fixes: 3453b4d62a98 ("dpif-netdev: dpcls per in_port with sorted
subtables")
Signed-off-by: Bhanuprakash Bodireddy <bhanuprakash.bodire...@intel.com>
Signed-off-by: Darrell Ball <dlu...@gmail.com>
---
lib/dpif-netdev.c | 8 +++++---
1 file changed, 5 insertions(+), 3 deletions(-)
diff --git a/lib/dpif-netdev.c b/lib/dpif-netdev.c
index d215156..e031c33 100644
--- a/lib/dpif-netdev.c
+++ b/lib/dpif-netdev.c
@@ -507,9 +507,11 @@ struct tx_port {
* I/O of all non-pmd threads. There will be no actual thread created
* for the instance.
*
- * Each struct has its own flow table and classifier. Packets received
- * from managed ports are looked up in the corresponding pmd thread's
- * flow table, and are executed with the found actions.
+ * Each struct has its own flow cache and classifier per managed ingress port.
+ * For packets received on ingress port, a look up is done on corresponding PMD
+ * thread's flow cache and in case of a miss, lookup is performed in the
+ * corresponding classifier of port. Packets are executed with the found
+ * actions in either case.
* */
struct dp_netdev_pmd_thread {
struct dp_netdev *dp;
--
2.4.11
_______________________________________________
dev mailing list
d...@openvswitch.org
https://mail.openvswitch.org/mailman/listinfo/ovs-dev
