ipvs: remove check for IP_VS_CONN_F_SYNC from ip_vs_bind_dest
authorJulian Anastasov <ja@ssi.bg>
Tue, 24 Apr 2012 20:46:36 +0000 (23:46 +0300)
committerPablo Neira Ayuso <pablo@netfilter.org>
Tue, 8 May 2012 17:38:26 +0000 (19:38 +0200)
As the IP_VS_CONN_F_INACTIVE bit is properly set
in cp->flags for all kind of connections we do not need to
add special checks for synced connections when updating
the activeconns/inactconns counters for first time. Now
logic will look just like in ip_vs_unbind_dest.

Signed-off-by: Julian Anastasov <ja@ssi.bg>
Signed-off-by: Simon Horman <horms@verge.net.au>
net/netfilter/ipvs/ip_vs_conn.c

index f562e630d35a78854a1973de7c9e3670d70b0bf5..1c1bb309a9558b46f90eca0060218375d8f72466 100644 (file)
@@ -585,11 +585,11 @@ ip_vs_bind_dest(struct ip_vs_conn *cp, struct ip_vs_dest *dest)
 
        /* Update the connection counters */
        if (!(cp->flags & IP_VS_CONN_F_TEMPLATE)) {
-               /* It is a normal connection, so increase the inactive
-                  connection counter because it is in TCP SYNRECV
-                  state (inactive) or other protocol inacive state */
-               if ((cp->flags & IP_VS_CONN_F_SYNC) &&
-                   (!(cp->flags & IP_VS_CONN_F_INACTIVE)))
+               /* It is a normal connection, so modify the counters
+                * according to the flags, later the protocol can
+                * update them on state change
+                */
+               if (!(cp->flags & IP_VS_CONN_F_INACTIVE))
                        atomic_inc(&dest->activeconns);
                else
                        atomic_inc(&dest->inactconns);