I agree this is a bit subtle. Here's an updated version of the patch.
---
A "packet out" ordinarily isn't accounted to any OpenFlow rule,
because its actions come directly from the controller instead of
from an OpenFlow rule. However, any "resubmit" actions or output
to OFPP_TABLE within the set of actions do bring in actions from a
rule, and the "packet out" should be accounted to the rule in that
case. This commit implements that behavior.
Reported-by: Jesse Gross <je...@nicira.com>
Signed-off-by: Ethan Jackson <et...@nicira.com>
---
ofproto/ofproto-dpif.c | 15 ++++++++++++---
1 files changed, 12 insertions(+), 3 deletions(-)
diff --git a/ofproto/ofproto-dpif.c b/ofproto/ofproto-dpif.c
index f5280ff..e206f43 100644
--- a/ofproto/ofproto-dpif.c
+++ b/ofproto/ofproto-dpif.c
@@ -5534,15 +5534,24 @@ packet_out(struct ofproto *ofproto_, struct ofpbuf
*packet,
ofproto->max_ports);
if (!error) {
struct odputil_keybuf keybuf;
- struct action_xlate_ctx ctx;
struct ofpbuf *odp_actions;
+ struct ofproto_push push;
struct ofpbuf key;
ofpbuf_use_stack(&key, &keybuf, sizeof keybuf);
odp_flow_key_from_flow(&key, flow);
- action_xlate_ctx_init(&ctx, ofproto, flow, flow->vlan_tci, 0, packet);
- odp_actions = xlate_actions(&ctx, ofp_actions, n_ofp_actions);
+ action_xlate_ctx_init(&push.ctx, ofproto, flow, flow->vlan_tci, 0,
+ packet);
+
+ /* Ensure that resubmits in 'ofp_actions' get accounted to their
+ * matching rules. */
+ push.packets = 1;
+ push.bytes = packet->size;
+ push.used = time_msec();
+ push.ctx.resubmit_hook = push_resubmit;
+
+ odp_actions = xlate_actions(&push.ctx, ofp_actions, n_ofp_actions);
dpif_execute(ofproto->dpif, key.data, key.size,
odp_actions->data, odp_actions->size, packet);
ofpbuf_delete(odp_actions);
--
1.7.7.1
_______________________________________________
dev mailing list
dev@openvswitch.org
http://openvswitch.org/mailman/listinfo/dev
