[SCTP]: Do not timestamp every SCTP packet.

We only need the timestamp on COOKIE-ECHO chunks, so instead of always
timestamping every SCTP packet, let common code timestamp if the socket
option is set.  For COOKIE-ECHO, simply get the time of day if we don't
have a timestamp.  This introduces a small possibility that the cookie
may be considered expired, but it will be renegotiated.
Signed-off-by: Vlad Yasevich <vladislav.yasevich@hp.com>
Signed-off-by: Sridhar Samudrala <sri@us.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/sctp/input.c

@@ -218,12 +218,6 @@ int sctp_rcv(struct sk_buff *skb)
 		}
 	}
 
-	/* SCTP seems to always need a timestamp right now (FIXME) */
-	if (skb->tstamp.off_sec == 0) {
-		__net_timestamp(skb);
-		sock_enable_timestamp(sk); 
-	}
-
 	if (!xfrm_policy_check(sk, XFRM_POLICY_IN, skb, family))
 		goto discard_release;
 	nf_reset(skb);

net/sctp/sm_make_chunk.c

@@ -1447,8 +1447,16 @@ no_hmac:
 	/* Check to see if the cookie is stale.  If there is already
 	 * an association, there is no need to check cookie's expiration
 	 * for init collision case of lost COOKIE ACK.
+	 * If skb has been timestamped, then use the stamp, otherwise
+	 * use current time.  This introduces a small possibility that
+	 * that a cookie may be considered expired, but his would only slow
+	 * down the new association establishment instead of every packet.
 	 */
+	if (sock_flag(ep->base.sk, SOCK_TIMESTAMP))
 		skb_get_timestamp(skb, &tv);
+	else
+		do_gettimeofday(&tv);
+
 	if (!asoc && tv_lt(bear_cookie->expiration, tv)) {
 		__u16 len;
 		/*