[PATCH nf-next v2] netfilter: nf_flow_table: delay teardown the offload flow until fin packet recv from both direction

[wenxu@xxxxxxxxxxxxxxx]

From: wenxu <wenxu@xxxxxxxxxxxxxxx>

A fin packet receive not always means the tcp connection teardown.
For tcp half close case, only the client shutdown the connection
and the server still can sendmsg to the client. The connection
can still be offloaded until the server shutdown the connection.

Signed-off-by: wenxu <wenxu@xxxxxxxxxxxxxxx>
---
 include/net/netfilter/nf_flow_table.h |  2 ++
 net/netfilter/nf_flow_table_ip.c      | 19 +++++++++++++++----
 2 files changed, 17 insertions(+), 4 deletions(-)

diff --git a/include/net/netfilter/nf_flow_table.h b/include/net/netfilter/nf_flow_table.h
index d5326c4..57481df 100644
--- a/include/net/netfilter/nf_flow_table.h
+++ b/include/net/netfilter/nf_flow_table.h
@@ -164,6 +164,8 @@ enum nf_flow_flags {
 	NF_FLOW_HW_DYING,
 	NF_FLOW_HW_DEAD,
 	NF_FLOW_HW_PENDING,
+	NF_FLOW_FIN_ORIGINAL,
+	NF_FLOW_FIN_REPLY,
 };
 
 enum flow_offload_type {
diff --git a/net/netfilter/nf_flow_table_ip.c b/net/netfilter/nf_flow_table_ip.c
index b350fe9..94f1c00 100644
--- a/net/netfilter/nf_flow_table_ip.c
+++ b/net/netfilter/nf_flow_table_ip.c
@@ -19,7 +19,8 @@
 #include <linux/udp.h>
 
 static int nf_flow_state_check(struct flow_offload *flow, int proto,
-			       struct sk_buff *skb, unsigned int thoff)
+			       struct sk_buff *skb, unsigned int thoff,
+			       enum flow_offload_tuple_dir dir)
 {
 	struct tcphdr *tcph;
 
@@ -27,9 +28,19 @@ static int nf_flow_state_check(struct flow_offload *flow, int proto,
 		return 0;
 
 	tcph = (void *)(skb_network_header(skb) + thoff);
-	if (unlikely(tcph->fin || tcph->rst)) {
+	if (unlikely(tcph->rst)) {
 		flow_offload_teardown(flow);
 		return -1;
+	} else if (unlikely(tcph->fin)) {
+		if (dir == FLOW_OFFLOAD_DIR_ORIGINAL)
+			set_bit(NF_FLOW_FIN_ORIGINAL, &flow->flags);
+		else if (dir == FLOW_OFFLOAD_DIR_REPLY)
+			set_bit(NF_FLOW_FIN_REPLY, &flow->flags);
+
+		if (test_bit(NF_FLOW_FIN_ORIGINAL, &flow->flags) &&
+		    test_bit(NF_FLOW_FIN_REPLY, &flow->flags))
+			flow_offload_teardown(flow);
+		return -1;
 	}
 
 	return 0;
@@ -373,7 +384,7 @@ static unsigned int nf_flow_queue_xmit(struct net *net, struct sk_buff *skb,
 
 	iph = (struct iphdr *)(skb_network_header(skb) + offset);
 	thoff = (iph->ihl * 4) + offset;
-	if (nf_flow_state_check(flow, iph->protocol, skb, thoff))
+	if (nf_flow_state_check(flow, iph->protocol, skb, thoff, dir))
 		return NF_ACCEPT;
 
 	if (!nf_flow_dst_check(&tuplehash->tuple)) {
@@ -635,7 +646,7 @@ static int nf_flow_tuple_ipv6(struct sk_buff *skb, const struct net_device *dev,
 
 	ip6h = (struct ipv6hdr *)(skb_network_header(skb) + offset);
 	thoff = sizeof(*ip6h) + offset;
-	if (nf_flow_state_check(flow, ip6h->nexthdr, skb, thoff))
+	if (nf_flow_state_check(flow, ip6h->nexthdr, skb, thoff, dir))
 		return NF_ACCEPT;
 
 	if (!nf_flow_dst_check(&tuplehash->tuple)) {
-- 
1.8.3.1