test-conntrack: Fix conntrack benchmark by clearing conntrack metadata.

Packets in the benchmark must be treated as new packets, i.e. they
should not have conntrack metadata set.  Current code will set up
'pkt->md.conn' after the first run and all subsequent calls will hit
the 'fast' processing that is intended for recirculated packets making
a false impression that current conntrack implementation is lightning
fast.

Before the change:
  $ ./ovstest test-conntrack benchmark 4 33554432 32 1
  conntrack:   1059 ms

After (correct):
  $ ./ovstest test-conntrack benchmark 4 33554432 32 1
  conntrack:  92785 ms

Fixes: 594570ea1cde ("conntrack: Optimize recirculations.")
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>
Acked-by: Aaron Conole <aconole@redhat.com>

1 files changed, 4 insertions, 0 deletions

diff --git a/tests/test-conntrack.c b/tests/test-conntrack.c
index e7c73220a..24c93e4a4 100644
--- a/tests/test-conntrack.c
+++ b/tests/test-conntrack.c
@@ -82,6 +82,7 @@ ct_thread_main(void *aux_)
 {
     struct thread_aux *aux = aux_;
     struct dp_packet_batch *pkt_batch;
+    struct dp_packet *pkt;
     ovs_be16 dl_type;
     size_t i;
     long long now = time_msec();
@@ -91,6 +92,9 @@ ct_thread_main(void *aux_)
     for (i = 0; i < n_pkts; i += batch_size) {
         conntrack_execute(ct, pkt_batch, dl_type, false, true, 0, NULL, NULL,
                           0, 0, NULL, NULL, now, 0);
+        DP_PACKET_BATCH_FOR_EACH (j, pkt, pkt_batch) {
+            pkt_metadata_init_conn(&pkt->md);
+        }
     }
     ovs_barrier_block(&barrier);
     destroy_packets(pkt_batch);