tcp: improve undo on timeout

Upon timeout, undo (via both timestamps/Eifel and DSACKs) was
disabled if any retransmits were still in flight.  The concern was
perhaps that spurious retransmission sent in a previous recovery
episode may trigger DSACKs to falsely undo the current recovery.

However, this inadvertently misses undo opportunities (using either
TCP timestamps or DSACKs) when timeout occurs during a loss episode,
i.e.  recurring timeouts or timeout during fast recovery. In these
cases some retransmissions will be in flight but we should allow
undo. Furthermore, we should only reset undo_marker and undo_retrans
upon timeout if we are starting a new recovery episode. Finally,
when we do reset our undo state, we now do so in a manner similar
to tcp_enter_recovery(), so that we require a DSACK for each of
the outstsanding retransmissions. This will achieve the original
goal by requiring that we receive the same number of DSACKs as
retransmissions.

This patch increases the undo events by 50% on Google servers.

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>

1 files changed, 1 insertions, 1 deletions

diff --git a/include/linux/tcp.h b/include/linux/tcp.h
index fa5258f322e7..e567f0dbf282 100644
--- a/include/linux/tcp.h
+++ b/include/linux/tcp.h
@@ -276,7 +276,7 @@ struct tcp_sock {
 	u32	retrans_stamp;	/* Timestamp of the last retransmit,
 				 * also used in SYN-SENT to remember stamp of
 				 * the first SYN. */
-	u32	undo_marker;	/* tracking retrans started here. */
+	u32	undo_marker;	/* snd_una upon a new recovery episode. */
 	int	undo_retrans;	/* number of undoable retransmissions. */
 	u32	total_retrans;	/* Total retransmits for entire connection */