Paweł Staszewski wrote:
I have Linux that works as Router + traffic management with connection
tracking enabled but without making NAT
About 600Mbit/s TX and 600Mbit/s RX forwarded traffic
iptables -L -n -v -t raw
Chain PREROUTING (policy ACCEPT 3340M packets, 2623G bytes)
pkts bytes target prot opt in out source
destination
3339M 2621G NOTRACK all -- * * 0.0.0.0/0
0.0.0.0/0
kernel 2.6.28.2
CPU: Core 2, speed 2999.98 MHz (estimated)
Counted CPU_CLK_UNHALTED events (Clock cycles when not halted) with a
unit mask of 0x00 (Unhalted core cycles) count 100000
samples % image name app name
symbol name
14544205 15.8604 vmlinux vmlinux
ctnetlink_del_expect
This doesn't seem to be accurate, I'm guessing its really
ctnetlink_conntrack_event().
Does this patch make any difference?
diff --git a/include/net/netfilter/nf_conntrack_core.h b/include/net/netfilter/nf_conntrack_core.h
index e78afe7..c25068e 100644
--- a/include/net/netfilter/nf_conntrack_core.h
+++ b/include/net/netfilter/nf_conntrack_core.h
@@ -59,7 +59,7 @@ static inline int nf_conntrack_confirm(struct sk_buff *skb)
struct nf_conn *ct = (struct nf_conn *)skb->nfct;
int ret = NF_ACCEPT;
- if (ct) {
+ if (ct && ct != &nf_conntrack_untracked) {
if (!nf_ct_is_confirmed(ct) && !nf_ct_is_dying(ct))
ret = __nf_conntrack_confirm(skb);
nf_ct_deliver_cached_events(ct);
