tcp: refactor to consolidate TFO passive open code
authorYuchung Cheng <ycheng@google.com>
Mon, 29 Apr 2019 22:46:19 +0000 (15:46 -0700)
committerDavid S. Miller <davem@davemloft.net>
Wed, 1 May 2019 15:47:54 +0000 (11:47 -0400)
Use a helper to consolidate two identical code block for passive TFO.

Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv4/tcp_input.c

index 3a40584..706a99e 100644 (file)
@@ -5989,6 +5989,27 @@ reset_and_undo:
        return 1;
 }
 
+static void tcp_rcv_synrecv_state_fastopen(struct sock *sk)
+{
+       tcp_try_undo_loss(sk, false);
+       inet_csk(sk)->icsk_retransmits = 0;
+
+       /* Once we leave TCP_SYN_RECV or TCP_FIN_WAIT_1,
+        * we no longer need req so release it.
+        */
+       reqsk_fastopen_remove(sk, tcp_sk(sk)->fastopen_rsk, false);
+
+       /* Re-arm the timer because data may have been sent out.
+        * This is similar to the regular data transmission case
+        * when new data has just been ack'ed.
+        *
+        * (TFO) - we could try to be more aggressive and
+        * retransmitting any data sooner based on when they
+        * are sent out.
+        */
+       tcp_rearm_rto(sk);
+}
+
 /*
  *     This function implements the receiving procedure of RFC 793 for
  *     all states except ESTABLISHED and TIME_WAIT.
@@ -6085,22 +6106,8 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb)
                if (!tp->srtt_us)
                        tcp_synack_rtt_meas(sk, req);
 
-               /* Once we leave TCP_SYN_RECV, we no longer need req
-                * so release it.
-                */
                if (req) {
-                       tcp_try_undo_loss(sk, false);
-                       inet_csk(sk)->icsk_retransmits = 0;
-                       reqsk_fastopen_remove(sk, req, false);
-                       /* Re-arm the timer because data may have been sent out.
-                        * This is similar to the regular data transmission case
-                        * when new data has just been ack'ed.
-                        *
-                        * (TFO) - we could try to be more aggressive and
-                        * retransmitting any data sooner based on when they
-                        * are sent out.
-                        */
-                       tcp_rearm_rto(sk);
+                       tcp_rcv_synrecv_state_fastopen(sk);
                } else {
                        tcp_try_undo_spurious_syn(sk);
                        tp->retrans_stamp = 0;
@@ -6138,18 +6145,9 @@ int tcp_rcv_state_process(struct sock *sk, struct sk_buff *skb)
        case TCP_FIN_WAIT1: {
                int tmo;
 
-               /* If we enter the TCP_FIN_WAIT1 state and we are a
-                * Fast Open socket and this is the first acceptable
-                * ACK we have received, this would have acknowledged
-                * our SYNACK so stop the SYNACK timer.
-                */
-               if (req) {
-                       tcp_try_undo_loss(sk, false);
-                       inet_csk(sk)->icsk_retransmits = 0;
-                       /* We no longer need the request sock. */
-                       reqsk_fastopen_remove(sk, req, false);
-                       tcp_rearm_rto(sk);
-               }
+               if (req)
+                       tcp_rcv_synrecv_state_fastopen(sk);
+
                if (tp->snd_una != tp->write_seq)
                        break;