[NETFILTER]: Fix oops in conntrack event cache

ip_ct_refresh_acct() can be called without a valid "skb" pointer.
This used to work, since ct_add_counters() deals with that fact.
However, the recently-added event cache doesn't handle this at all.

This patch is a quick fix that is supposed to be replaced soon by a cleaner
solution during the pending redesign of the event cache.

Signed-off-by: Harald Welte <laforge@netfilter.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/ipv4/netfilter/ip_conntrack_core.c

@@ -1143,7 +1143,10 @@ void ip_ct_refresh_acct(struct ip_conntrack *ct,
 		if (del_timer(&ct->timeout)) {
 			ct->timeout.expires = jiffies + extra_jiffies;
 			add_timer(&ct->timeout);
-			ip_conntrack_event_cache(IPCT_REFRESH, skb);
+			/* FIXME: We loose some REFRESH events if this function
+			 * is called without an skb.  I'll fix this later -HW */
+			if (skb)
+				ip_conntrack_event_cache(IPCT_REFRESH, skb);
 		}
 		ct_add_counters(ct, ctinfo, skb);
 		write_unlock_bh(&ip_conntrack_lock);