Merge branch 'tcp_fast_open_synack_fin'
Eric Dumazet says:
====================
tcp: fastopen: accept data/FIN present in SYNACK
Implements RFC 7413 (TCP Fast Open) 4.2.2, accepting payload and/or FIN
in SYNACK messages, and prepare removal of SYN flag in tcp_recvmsg()
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
diff --git a/include/net/tcp.h b/include/net/tcp.h
index f6f8f03..27f4c73 100644
--- a/include/net/tcp.h
+++ b/include/net/tcp.h
@@ -1437,6 +1437,7 @@
extern struct tcp_fastopen_context __rcu *tcp_fastopen_ctx;
int tcp_fastopen_reset_cipher(void *key, unsigned int len);
+void tcp_fastopen_add_skb(struct sock *sk, struct sk_buff *skb);
struct sock *tcp_try_fastopen(struct sock *sk, struct sk_buff *skb,
struct request_sock *req,
struct tcp_fastopen_cookie *foc,
diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c
index 19746b3..c507577 100644
--- a/net/ipv4/tcp.c
+++ b/net/ipv4/tcp.c
@@ -1466,8 +1466,10 @@
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 @@
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)
diff --git a/net/ipv4/tcp_fastopen.c b/net/ipv4/tcp_fastopen.c
index 55be6ac..6a6e11e 100644
--- a/net/ipv4/tcp_fastopen.c
+++ b/net/ipv4/tcp_fastopen.c
@@ -124,6 +124,38 @@
return false;
}
+
+/* If an incoming SYN or SYNACK frame contains a payload and/or FIN,
+ * queue this additional data / FIN.
+ */
+void tcp_fastopen_add_skb(struct sock *sk, struct sk_buff *skb)
+{
+ struct tcp_sock *tp = tcp_sk(sk);
+
+ if (TCP_SKB_CB(skb)->end_seq == tp->rcv_nxt)
+ return;
+
+ skb = skb_clone(skb, GFP_ATOMIC);
+ if (!skb)
+ return;
+
+ skb_dst_drop(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;
+
+ /* u64_stats_update_begin(&tp->syncp) not needed here,
+ * as we certainly are not changing upper 32bit value (0)
+ */
+ tp->bytes_received = skb->len;
+}
+
static struct sock *tcp_fastopen_create_child(struct sock *sk,
struct sk_buff *skb,
struct dst_entry *dst,
@@ -132,7 +164,6 @@
struct tcp_sock *tp;
struct request_sock_queue *queue = &inet_csk(sk)->icsk_accept_queue;
struct sock *child;
- u32 end_seq;
bool own_req;
req->num_retrans = 0;
@@ -178,35 +209,11 @@
tcp_init_metrics(child);
tcp_init_buffer_space(child);
- /* Queue the data carried in the SYN packet.
- * We used to play tricky games with skb_get().
- * With lockless listener, it is a dead end.
- * Do not think about it.
- *
- * XXX (TFO) - we honor a zero-payload TFO request for now,
- * (any reason not to?) but no need to queue the skb since
- * there is no data. How about SYN+FIN?
- */
- end_seq = TCP_SKB_CB(skb)->end_seq;
- if (end_seq != TCP_SKB_CB(skb)->seq + 1) {
- struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
+ tp->rcv_nxt = TCP_SKB_CB(skb)->seq + 1;
- if (likely(skb2)) {
- skb_dst_drop(skb2);
- __skb_pull(skb2, tcp_hdrlen(skb));
- skb_set_owner_r(skb2, child);
- __skb_queue_tail(&child->sk_receive_queue, skb2);
- tp->syn_data_acked = 1;
+ tcp_fastopen_add_skb(child, skb);
- /* u64_stats_update_begin(&tp->syncp) not needed here,
- * as we certainly are not changing upper 32bit value (0)
- */
- tp->bytes_received = end_seq - TCP_SKB_CB(skb)->seq - 1;
- } else {
- end_seq = TCP_SKB_CB(skb)->seq + 1;
- }
- }
- tcp_rsk(req)->rcv_nxt = tp->rcv_nxt = end_seq;
+ tcp_rsk(req)->rcv_nxt = tp->rcv_nxt;
/* tcp_conn_request() is sending the SYNACK,
* and queues the child into listener accept queue.
*/
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index 1c2a734..4add3eb 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -5509,6 +5509,9 @@
tp->syn_data_acked = tp->syn_data;
if (tp->syn_data_acked)
NET_INC_STATS_BH(sock_net(sk), LINUX_MIB_TCPFASTOPENACTIVE);
+
+ tcp_fastopen_add_skb(sk, synack);
+
return false;
}