- Reinstantiate the after_idle hook call in tcp_output(), which got lost

  somewhere along the way due to mismerging r211464 in our development tree.

- Capture the essence of r211464 in NewReno's after_idle() hook. We don't
  use V_ss_fltsz/V_ss_fltsz_local yet which needs to be revisited.

Sponsored by:	FreeBSD Foundation
Submitted by:	David Hayes <dahayes at swin edu au>
MFC after:	9 weeks
X-MFC with:	r215166

1 files changed, 17 insertions, 4 deletions

diff --git a/sys/netinet/cc/cc_newreno.c b/sys/netinet/cc/cc_newreno.c
index e383510..511186a 100644
--- a/sys/netinet/cc/cc_newreno.c
+++ b/sys/netinet/cc/cc_newreno.c
@@ -216,15 +216,28 @@ newreno_post_recovery(struct cc_var *ccv)
 void
 newreno_after_idle(struct cc_var *ccv)
 {
+	int rw;
+
 	/*
-	 * We have been idle for "a while" and no acks are expected to clock out
-	 * any data we send -- slow start to get ack "clock" running again.
+	 * If we've been idle for more than one retransmit timeout the old
+	 * congestion window is no longer current and we have to reduce it to
+	 * the restart window before we can transmit again.
+	 *
+	 * The restart window is the initial window or the last CWND, whichever
+	 * is smaller.
+	 *
+	 * This is done to prevent us from flooding the path with a full CWND at
+	 * wirespeed, overloading router and switch buffers along the way.
+	 *
+	 * See RFC5681 Section 4.1. "Restarting Idle Connections".
 	 */
 	if (V_tcp_do_rfc3390)
-		CCV(ccv, snd_cwnd) = min(4 * CCV(ccv, t_maxseg),
+		rw = min(4 * CCV(ccv, t_maxseg),
 		    max(2 * CCV(ccv, t_maxseg), 4380));
 	else
-		CCV(ccv, snd_cwnd) = CCV(ccv, t_maxseg) * 2;
+		rw = CCV(ccv, t_maxseg) * 2;
+
+	CCV(ccv, snd_cwnd) = min(rw, CCV(ccv, snd_cwnd));
 }