tcp: fix false undo corner cases
The undo code assumes that, upon entering loss recovery, TCP
1) always retransmit something
2) the retransmission never fails locally (e.g., qdisc drop)
so undo_marker is set in tcp_enter_recovery() and undo_retrans is
incremented only when tcp_retransmit_skb() is successful.
When the assumption is broken because TCP's cwnd is too small to
retransmit or the retransmit fails locally. The next (DUP)ACK
would incorrectly revert the cwnd and the congestion state in
tcp_try_undo_dsack() or tcp_may_undo(). Subsequent (DUP)ACKs
may enter the recovery state. The sender repeatedly enter and
(incorrectly) exit recovery states if the retransmits continue to
fail locally while receiving (DUP)ACKs.
The fix is to initialize undo_retrans to -1 and start counting on
the first retransmission. Always increment undo_retrans even if the
retransmissions fail locally because they couldn't cause DSACKs to
undo the cwnd reduction.
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>
diff --git a/net/ipv4/tcp_input.c b/net/ipv4/tcp_input.c
index b5c2375..40639c2 100644
--- a/net/ipv4/tcp_input.c
+++ b/net/ipv4/tcp_input.c
@@ -1106,7 +1106,7 @@
}
/* D-SACK for already forgotten data... Do dumb counting. */
- if (dup_sack && tp->undo_marker && tp->undo_retrans &&
+ if (dup_sack && tp->undo_marker && tp->undo_retrans > 0 &&
!after(end_seq_0, prior_snd_una) &&
after(end_seq_0, tp->undo_marker))
tp->undo_retrans--;
@@ -1187,7 +1187,7 @@
/* Account D-SACK for retransmitted packet. */
if (dup_sack && (sacked & TCPCB_RETRANS)) {
- if (tp->undo_marker && tp->undo_retrans &&
+ if (tp->undo_marker && tp->undo_retrans > 0 &&
after(end_seq, tp->undo_marker))
tp->undo_retrans--;
if (sacked & TCPCB_SACKED_ACKED)
@@ -1893,7 +1893,7 @@
tp->lost_out = 0;
tp->undo_marker = 0;
- tp->undo_retrans = 0;
+ tp->undo_retrans = -1;
}
void tcp_clear_retrans(struct tcp_sock *tp)
@@ -2665,7 +2665,7 @@
tp->prior_ssthresh = 0;
tp->undo_marker = tp->snd_una;
- tp->undo_retrans = tp->retrans_out;
+ tp->undo_retrans = tp->retrans_out ? : -1;
if (inet_csk(sk)->icsk_ca_state < TCP_CA_CWR) {
if (!ece_ack)