tcp: do not enqueue skb with SYN flag

If we remove the SYN flag from the skbs that tcp_fastopen_add_skb()
places in socket receive queue, then we can remove the test that
tcp_recvmsg() has to perform in fast path.

All we have to do is to adjust SEQ in the slow path.

For the moment, we place an unlikely() and output a message
if we find an skb having SYN flag set.
Goal would be to get rid of the test completely.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/ipv4/tcp.c

@@ -1466,8 +1466,10 @@ static struct sk_buff *tcp_recv_skb(struct sock *sk, u32 seq, u32 *off)
 
 	while ((skb = skb_peek(&sk->sk_receive_queue)) != NULL) {
 		offset = seq - TCP_SKB_CB(skb)->seq;
-		if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)
+		if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) {
+			pr_err_once("%s: found a SYN, please report !\n", __func__);
 			offset--;
+		}
 		if (offset < skb->len || (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)) {
 			*off = offset;
 			return skb;
@@ -1657,8 +1659,10 @@ int tcp_recvmsg(struct sock *sk, struct msghdr *msg, size_t len, int nonblock,
 				break;
 
 			offset = *seq - TCP_SKB_CB(skb)->seq;
-			if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)
+			if (unlikely(TCP_SKB_CB(skb)->tcp_flags & TCPHDR_SYN)) {
+				pr_err_once("%s: found a SYN, please report !\n", __func__);
 				offset--;
+			}
 			if (offset < skb->len)
 				goto found_ok_skb;
 			if (TCP_SKB_CB(skb)->tcp_flags & TCPHDR_FIN)

net/ipv4/tcp_fastopen.c

@@ -143,6 +143,9 @@ void tcp_fastopen_add_skb(struct sock *sk, struct sk_buff *skb)
 	__skb_pull(skb, tcp_hdrlen(skb));
 	skb_set_owner_r(skb, sk);
 
+	TCP_SKB_CB(skb)->seq++;
+	TCP_SKB_CB(skb)->tcp_flags &= ~TCPHDR_SYN;
+
 	tp->rcv_nxt = TCP_SKB_CB(skb)->end_seq;
 	__skb_queue_tail(&sk->sk_receive_queue, skb);
 	tp->syn_data_acked = 1;