Implement the clock filter as descirbed by David Mills:

form the last 8 replied received from a peer, find the one with the lowest
delay. Use that as the peer's update taken into account for calculating
the local clock's offset.
Invalidate that reply and all ones received earlier than it so that they do
not get used again.

1 files changed, 8 insertions, 5 deletions

diff --git a/usr.sbin/ntpd/client.c b/usr.sbin/ntpd/client.c
index 70560973270..f877e1f944e 100644
--- a/usr.sbin/ntpd/client.c
+++ b/usr.sbin/ntpd/client.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: client.c,v 1.10 2004/07/06 19:06:43 henning Exp $ */
+/*	$OpenBSD: client.c,v 1.11 2004/07/06 23:26:38 henning Exp $ */
 
 /*
  * Copyright (c) 2003, 2004 Henning Brauer <henning@openbsd.org>
@@ -138,16 +138,19 @@ client_dispatch(struct ntp_peer *p)
 	T2 = lfp_to_d(msg.rectime);
 	T3 = lfp_to_d(msg.xmttime);
 
-	p->offset[p->shift] = ((T2 - T1) + (T3 - T4)) / 2;
-	p->delay[p->shift] = (T2 - T1) - (T3 - T4);
+	p->reply[p->shift].offset = ((T2 - T1) + (T3 - T4)) / 2;
+	p->reply[p->shift].delay = (T4 - T1) - (T2 - T3);
+	p->reply[p->shift].error = (T2 - T1) - (T3 - T4);
+	p->reply[p->shift].rcvd = time(NULL);
+	p->reply[p->shift].good = 1;
 
 	p->state = STATE_REPLY_RECEIVED;
 	p->next = time(NULL) + INTERVAL_QUERY;
 	p->deadline = 0;
 
 	log_debug("reply received from %s: offset %f delay %f",
-	    log_sockaddr((struct sockaddr *)&fsa), p->offset[p->shift],
-	    p->delay[p->shift]);
+	    log_sockaddr((struct sockaddr *)&fsa), p->reply[p->shift].offset,
+	    p->reply[p->shift].delay);
 
 	if (++p->shift >= OFFSET_ARRAY_SIZE) {
 		p->shift = 0;