ipf: Cancel fragment pkt copy.

Canceling packet copying can better improve the performance of handling
fragmented packets. In 640d4db, pkt copying was added to fix the crash,
but the crash has been fixed in 7e6b41a, so there is no need to copy the
pkt any longer.

Acked-by: Mike Pattrick <mkp@redhat.com>
Acked-by: Aaron Conole <aconole@redhat.com>
Signed-off-by: fang <fangjiannan@cmss.chinamobile.com>
Signed-off-by: Aaron Conole <aconole@redhat.com>

AUTHORS.rst

@@ -171,6 +171,7 @@ Ethan Rahn                         erahn@arista.com
 Eziz Durdyyev                      ezizdurdy@gmail.com
 Fabrizio D'Angelo                  fdangelo@redhat.com
 Faicker Mo                         faicker.mo@ucloud.cn
+fang                               fangjiannan@cmss.chinamobile.com
 Fangrui Song                       maskray@google.com
 Felix Huettner                     felix.huettner@mail.schwarz
 Fengqi Li                          lifengqi@inspur.com

lib/ipf.c

@@ -820,7 +820,7 @@ ipf_process_frag(struct ipf *ipf, struct ipf_list *ipf_list,
              * recommend not setting the mempool number of buffers too low
              * and also clamp the number of fragments. */
             struct ipf_frag *frag = &ipf_list->frag_list[last_inuse_idx + 1];
-            frag->pkt = dp_packet_clone(pkt);
+            frag->pkt = pkt;
             frag->start_data_byte = start_data_byte;
             frag->end_data_byte = end_data_byte;
             ipf_list->last_inuse_idx++;
@@ -959,7 +959,6 @@ ipf_extract_frags_from_batch(struct ipf *ipf, struct dp_packet_batch *pb,
                     dp_packet_batch_refill(pb, rp->pkt, pb_idx);
                     rp->list->reass_execute_ctx = rp->pkt;
                 }
-                dp_packet_delete(pkt);
             }
             ovs_mutex_unlock(&ipf->ipf_lock);
         } else {