diff options
| author | Yuchung Cheng <ycheng@google.com> | 2013-03-24 10:42:25 +0000 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2013-03-24 17:27:28 -0400 |
| commit | 7ebe183c6d444ef5587d803b64a1f4734b18c564 (patch) | |
| tree | 8e49b71e1d0ffc90ce725830b7bb6d2363495b70 | |
| parent | 8fe7f99a9e11a43183bc27420309ae105e1fec1a (diff) | |
tcp: undo spurious timeout after SACK reneging
On SACK reneging the sender immediately retransmits and forces a
timeout but disables Eifel (undo). If the (buggy) receiver does not
drop any packet this can trigger a false slow-start retransmit storm
driven by the ACKs of the original packets. This can be detected with
undo and TCP timestamps.
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Acked-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
| -rw-r--r-- | net/ipv4/tcp_input.c | 7 |
1 files changed, 2 insertions, 5 deletions
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c index 0d9bdacce99..3bd55bad230 100644 --- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -2059,11 +2059,8 @@ void tcp_enter_loss(struct sock *sk, int how) if (tcp_is_reno(tp)) tcp_reset_reno_sack(tp); - if (!how) { - /* Push undo marker, if it was plain RTO and nothing - * was retransmitted. */ - tp->undo_marker = tp->snd_una; - } else { + tp->undo_marker = tp->snd_una; + if (how) { tp->sacked_out = 0; tp->fackets_out = 0; } |