tcp: improve fastopen icmp handling

If a fast open socket is already accepted by the user, it should
be treated like a connected socket to record the ICMP error in
sk_softerr, so the user can fetch it. Do that in both tcp_v4_err
and tcp_v6_err.

Also refactor the sequence window check to improve readability
(e.g., there were two local variables named 'req').

Signed-off-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: Daniel Lee <longinus00@gmail.com>
Signed-off-by: Jerry Chu <hkchu@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index 1665f0f..a2780e5 100644 (file)
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -336,8 +336,8 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info)
        const int code = icmp_hdr(icmp_skb)->code;
        struct sock *sk;
        struct sk_buff *skb;
-       struct request_sock *req;
-       __u32 seq;
+       struct request_sock *fastopen;
+       __u32 seq, snd_una;
        __u32 remaining;
        int err;
        struct net *net = dev_net(icmp_skb->dev);
@@ -378,12 +378,12 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info)
 
        icsk = inet_csk(sk);
        tp = tcp_sk(sk);
-       req = tp->fastopen_rsk;
        seq = ntohl(th->seq);
+       /* XXX (TFO) - tp->snd_una should be ISN (tcp_create_openreq_child() */
+       fastopen = tp->fastopen_rsk;
+       snd_una = fastopen ? tcp_rsk(fastopen)->snt_isn : tp->snd_una;
        if (sk->sk_state != TCP_LISTEN &&
-           !between(seq, tp->snd_una, tp->snd_nxt) &&
-           (req == NULL || seq != tcp_rsk(req)->snt_isn)) {
-               /* For a Fast Open socket, allow seq to be snt_isn. */
+           !between(seq, snd_una, tp->snd_nxt)) {
                NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
                goto out;
        }
@@ -426,11 +426,9 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info)
                if (code != ICMP_NET_UNREACH && code != ICMP_HOST_UNREACH)
                        break;
                if (seq != tp->snd_una  || !icsk->icsk_retransmits ||
-                   !icsk->icsk_backoff)
+                   !icsk->icsk_backoff || fastopen)
                        break;
 
-               /* XXX (TFO) - revisit the following logic for TFO */
-
                if (sock_owned_by_user(sk))
                        break;
 
@@ -462,14 +460,6 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info)
                goto out;
        }
 
-       /* XXX (TFO) - if it's a TFO socket and has been accepted, rather
-        * than following the TCP_SYN_RECV case and closing the socket,
-        * we ignore the ICMP error and keep trying like a fully established
-        * socket. Is this the right thing to do?
-        */
-       if (req && req->sk == NULL)
-               goto out;
-
        switch (sk->sk_state) {
                struct request_sock *req, **prev;
        case TCP_LISTEN:
@@ -502,10 +492,13 @@ void tcp_v4_err(struct sk_buff *icmp_skb, u32 info)
                goto out;
 
        case TCP_SYN_SENT:
-       case TCP_SYN_RECV:  /* Cannot happen.
-                              It can f.e. if SYNs crossed,
-                              or Fast Open.
-                            */
+       case TCP_SYN_RECV:
+               /* Only in fast or simultaneous open. If a fast open socket is
+                * is already accepted it is treated as a connected one below.
+                */
+               if (fastopen && fastopen->sk == NULL)
+                       break;
+
                if (!sock_owned_by_user(sk)) {
                        sk->sk_err = err;
 

diff --git a/net/ipv6/tcp_ipv6.c b/net/ipv6/tcp_ipv6.c
index 7fa6743..a7a62ce 100644 (file)
--- a/net/ipv6/tcp_ipv6.c
+++ b/net/ipv6/tcp_ipv6.c
@@ -340,7 +340,8 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
        struct sock *sk;
        int err;
        struct tcp_sock *tp;
-       __u32 seq;
+       struct request_sock *fastopen;
+       __u32 seq, snd_una;
        struct net *net = dev_net(skb->dev);
 
        sk = inet6_lookup(net, &tcp_hashinfo, &hdr->daddr,
@@ -371,8 +372,11 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
 
        tp = tcp_sk(sk);
        seq = ntohl(th->seq);
+       /* XXX (TFO) - tp->snd_una should be ISN (tcp_create_openreq_child() */
+       fastopen = tp->fastopen_rsk;
+       snd_una = fastopen ? tcp_rsk(fastopen)->snt_isn : tp->snd_una;
        if (sk->sk_state != TCP_LISTEN &&
-           !between(seq, tp->snd_una, tp->snd_nxt)) {
+           !between(seq, snd_una, tp->snd_nxt)) {
                NET_INC_STATS_BH(net, LINUX_MIB_OUTOFWINDOWICMPS);
                goto out;
        }
@@ -436,8 +440,13 @@ static void tcp_v6_err(struct sk_buff *skb, struct inet6_skb_parm *opt,
                goto out;
 
        case TCP_SYN_SENT:
-       case TCP_SYN_RECV:  /* Cannot happen.
-                              It can, it SYNs are crossed. --ANK */
+       case TCP_SYN_RECV:
+               /* Only in fast or simultaneous open. If a fast open socket is
+                * is already accepted it is treated as a connected one below.
+                */
+               if (fastopen && fastopen->sk == NULL)
+                       break;
+
                if (!sock_owned_by_user(sk)) {
                        sk->sk_err = err;
                        sk->sk_error_report(sk);                /* Wake people up to see the error (see connect in sock.c) */
@@ -1760,6 +1769,7 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i)
        const struct inet_sock *inet = inet_sk(sp);
        const struct tcp_sock *tp = tcp_sk(sp);
        const struct inet_connection_sock *icsk = inet_csk(sp);
+       struct fastopen_queue *fastopenq = icsk->icsk_accept_queue.fastopenq;
 
        dest  = &sp->sk_v6_daddr;
        src   = &sp->sk_v6_rcv_saddr;
@@ -1802,7 +1812,9 @@ static void get_tcp6_sock(struct seq_file *seq, struct sock *sp, int i)
                   jiffies_to_clock_t(icsk->icsk_ack.ato),
                   (icsk->icsk_ack.quick << 1) | icsk->icsk_ack.pingpong,
                   tp->snd_cwnd,
-                  tcp_in_initial_slowstart(tp) ? -1 : tp->snd_ssthresh
+                  sp->sk_state == TCP_LISTEN ?
+                       (fastopenq ? fastopenq->max_qlen : 0) :
+                       (tcp_in_initial_slowstart(tp) ? -1 : tp->snd_ssthresh)
                   );
 }