Changelog:
20041209
- (dtucker) [version.h] Release 3.6.1p1.
20041208
- (dtucker) [Makefile.in configure.ac] Add --with-privsep-path configure
option, based on patch from Andrew Stribblehill (ads at debian org).
- (dtucker) [Makefile.in configure.ac] Strip installed binaries by default,
add --disable-strip configure option, taken from OpenSSH. Noticed by otto@
20041207
- (dtucker) OpenBSD CVS Sync
- mickey@cvs.openbsd.org 2004/12/06 17:52:33
[ntpd.h]
ensure the most excellent alignment in the structs; henning@ ok
- mickey@cvs.openbsd.org 2004/12/06 21:57:17
[ntpd.8 ntpd.c ntpd.h]
do not log tiny local clock drifts; w/ help from Joerg Sonnenberger
<joerg@britannica.bec.de>; henning@ ok
20041204
- (dtucker) OpenBSD CVS Sync
- jmc@cvs.openbsd.org 2004/11/07 22:42:33
[ntpd.conf.5]
document that keywords can be specified multiple times;
from otto and myself; prodded by henning;
- otto@cvs.openbsd.org 2004/11/08 20:09:19
[ntpd.conf.5]
Advice user to use multiple servers. Prodded by Daniel Polak, help and ok
jmc@ ok henning@
- henning@cvs.openbsd.org 2004/11/10 12:27:54
[ntpd.c ntpd.h parse.y]
const'ify conffile
From: Joerg Sonnenberger <joerg@britannica.bec.de>
- henning@cvs.openbsd.org 2004/11/10 12:47:28
[client.c ntp.c ntpd.h]
ntp_adjtime() -> priv_adjtime()
ntp_settime() -> priv_settime()
ntp_host_dns() -> priv_host_dns()
- henning@cvs.openbsd.org 2004/11/12 18:24:52
[ntp.c ntpd.h util.c]
some missing includes, from Joerg Sonnenberger <joerg@britannica.bec.de>
- henning@cvs.openbsd.org 2004/11/25 07:27:41
[parse.y]
fix "listen on hostname"
fallout from the deferred dns lookups
noticed by dhartmei@
- (dtucker) [y.tab.c] Regen.
20041203
- (dtucker) [ntpd.c openbsd-compat/bsd-misc.c openbsd-compat/openbsd-compat.h]
Correctly initialise __progname on platforms that don't have it natively,
based on OpenSSH's ssh_get_progname(). Reported by ihsan at dogan.ch.
20041106
- (dtucker) [client.c] Back out portable-specific SA_LEN bits.
- (dtucker) OpenBSD CVS Sync
- henning@cvs.openbsd.org 2004/11/05 15:28:29
[parse.y]
memleaks in error pathes, patrick latifi, Thanks!
- dtucker@cvs.openbsd.org 2004/11/06 00:39:46
[client.c]
Use SA_LEN() instead of ss.ss_len. Evaluates to the same result but it's
easier on portable. ok henning@
20041105
- (dtucker) [configure.ac ntpd.c] Remove workaround for signal(SIGCHLD,
SIG_IGN) on Linux.
- (dtucker) OpenBSD CVS Sync
- henning@cvs.openbsd.org 2004/11/02 19:00:38
[ntpd.8]
superfluous comma, From: James Herbert <lists@artyzan.net>
- henning@cvs.openbsd.org 2004/11/05 00:04:22
[ntpd.c]
use SIG_DFL instead of SIG_IGN when we are not interested in SIG_CHILD
anymore, same thing for us and it makes darren's life easier for the
portable
- (dtucker) [configure.ac ntpd.h] Add a --with-privsep-user option to
configure.
20041103
- (dtucker) [configure.ac] Check for snprintf too.
- (dtucker) [ntpd.8] Remove references to OpenBSD-specific startup files,
from Christian Gut (cycloon at is-root org).
20041028
- (dtucker) [openbsd-compat/bsd-misc.c] Use precision from adjtimex for
clock_getres.
20041027
- (dtucker) OpenBSD CVS Sync
- henning@cvs.openbsd.org 2004/10/27 10:55:27
[ntp.c]
use clock_getres(3) and calculate precision from that, and fill the
precision field when we reply in server mode accordingly. from phessler
- dtucker@cvs.openbsd.org 2004/10/27 14:19:12
[ntp.c]
Calculate Hz and round up; ok henning@
- (dtucker) [configure.ac openbsd-compat/bsd-misc.c
openbsd-compat/openbsd-compat.h] Add clock_getres compat function.
20041026
- (dtucker) [configure.ac includes.h openbsd-compat/bsd-misc.c
openbsd-compat/openbsd-compat.h] Add vsyslog() to compat library.
- (dtucker) [configure.ac openbsd-compat/Makefile.in
openbsd-compat/bsd-snprintf.c] Import snprintf replacement from OpenSSH
Portable.
- (dtucker) [configure.ac defines.h] Add a few definitions needed to compile
on older Solaris version.
- (dtucker) [README] Update for Solaris.
- (dtucker) [configure.ac] Look for socklen_t in <sys/socket.h>
- (dtucker) [configure.ac includes.h] Include <ctype.h>, do better checking
for socklen_t.
- (dtucker) [openbsd-compat/openbsd-compat.h] Add prototypes for snprintf
and vsnprintf.
- (dtucker) [configure.ac defines.h] Use sockaddr_storage.__ss_family where
available.
20041023
- (dtucker) OpenBSD CVS Sync
- henning@cvs.openbsd.org 2004/09/14 22:01:28
[client.c]
paranoia: reset query->fd to -1 after close, from canacar some time ago
- henning@cvs.openbsd.org 2004/09/15 00:05:29
[buffer.c]
remove buf_write(), not used in ntpd. found by theo
- henning@cvs.openbsd.org 2004/09/15 00:07:20
[ntp.c]
missing include, from theo
- henning@cvs.openbsd.org 2004/09/15 00:08:06
[ntp.c ntpd.c]
unused variables, theo
- henning@cvs.openbsd.org 2004/09/15 00:18:12
[ntpd.h parse.y]
remove the unused variable/macro code, ok theo
- henning@cvs.openbsd.org 2004/09/15 00:23:08
[parse.y]
kill another unused function and two debugging printfs
- henning@cvs.openbsd.org 2004/09/15 19:14:11
[ntp.c ntpd.c]
malloc the imsg buffers instead of having them statically, suggested by
micsky some time ago, ok otto
- henning@cvs.openbsd.org 2004/09/15 19:21:25
[imsg.c ntp.c ntpd.c ntpd.h]
imsg framework cleanup:
-kill the _pid flavors of imsg_create and imsg_compose, and just add pid as
argument to those
-use imsg_create in imsg_compose instead of duplicating code
-check for datalen overflow
- henning@cvs.openbsd.org 2004/09/15 19:22:55
[imsg.c]
need buf_free() to free buf, free() is not good enough
- henning@cvs.openbsd.org 2004/09/16 01:02:37
[imsg.c]
ewps...
- henning@cvs.openbsd.org 2004/09/16 01:06:51
[imsg.c]
use imsg_add instead of the lower level buf_add in imsg_create; it already
does the error checking for us.
- henning@cvs.openbsd.org 2004/09/16 01:10:05
[imsg.c]
in imsg_compose:
-don't buf_free() on imsg_add() errors, it already does that for us
-use imsg_close() instead of buf_close(), does error handling already
- henning@cvs.openbsd.org 2004/09/16 01:13:42
[imsg.c ntpd.h]
the "type" param to imsg_compose and imsg_create is really an
enum imsg_type and not an int
- henning@cvs.openbsd.org 2004/09/18 07:33:14
[ntp.c ntpd.h]
do not bother overallocating and shrinking the pfd and idx2peer arrays,
doesn't by us anything. discussed with ryan during dinner at original joe's
- henning@cvs.openbsd.org 2004/09/18 20:01:38
[client.c ntp.c ntpd.8 ntpd.c ntpd.h]
add a new -s option, that tells ntpd to set the time using settimeofday()
once at startup. ntpd delays daemonizing until it has done the intial
time setting (or ran into the timeout) in this mode to make sure stuff
started later in rc is not subject to time jumps.
this eleminates the need to run rdate -n beforehands.
with some input from & ok ryan and bob, march music from mickey
- henning@cvs.openbsd.org 2004/09/18 20:27:57
[ntpd.c ntpd.h]
don't call settimeofday() when the offset is smaller than 180 seconds,
adjtime() will fix that fast enough, from discussion in theo's living room
ok mcbride beck
- henning@cvs.openbsd.org 2004/09/18 20:31:46
[ntpd.8]
say when we run settimeofday() with -s and when not, help from bob
- henning@cvs.openbsd.org 2004/09/18 20:37:12
[ntpd.8 ntpd.c]
implement -S to override earlier -s, requested by theo
- henning@cvs.openbsd.org 2004/09/18 23:21:35
[ntpd.c]
jmc says S before s and not s before S, sssssssso we do.
- henning@cvs.openbsd.org 2004/09/18 23:22:13
[ntpd.8]
greatly improved by jmc with some tweaks by yours truly
- henning@cvs.openbsd.org 2004/09/23 01:53:07
[ntpd.c]
reset chld_pid to 0 when acting upon a SIGCHLD so we don't try to send it
a kill then - tiny possible race there
pointed out by Brian Poole <raj@cerias.purdue.edu>
- henning@cvs.openbsd.org 2004/09/24 14:51:16
[client.c ntp_msg.c]
connect() the client-side sockets. idea & test & ok camield@
- henning@cvs.openbsd.org 2004/09/30 10:19:43
[client.c]
now that we connect() the client sockets we need to handle ECONNREFUSED as
non-fatal too, from camield@
- henning@cvs.openbsd.org 2004/10/04 11:12:58
[ntp.c]
do not take the average offset from all peers when calculating the total
offset to correct the local clock, but use the median.
given a reasonable sized set of servers this makes us nearly immune against
outliers or flasetickers, without the need for a horribly complicated
outliers detection which does not yield to better results anyway.
- henning@cvs.openbsd.org 2004/10/05 11:23:28
[client.c]
in client_addr_init() and client_nextaddr(), do not set up the socket and
connect it, instead leave it at -1.
in client_query, set up and connect the socket if it is -1.
and, the real reason for this change: handle connect failures gracefully
ok otto
- henning@cvs.openbsd.org 2004/10/08 12:42:25
[client.c]
whitespace both in comment; from bernd
- henning@cvs.openbsd.org 2004/10/13 09:20:41
[ntp.c]
when we get back a IMSG_HOST_DNS message from the parent the peer ID within
might have become invalid (because the peer showed up, dns request sent to
parent, peer vanishes, and then the reply comes back), so do not fatal() in
that case but just log_warnx(). provoked by brad
- henning@cvs.openbsd.org 2004/10/13 12:22:39
[ntp.c server.c]
correctly set refid in replies with NTP protocol versions < 4.
code path for NTP4 remains unchanged, we already set refid correctly there.
NTP3 and older uses an IPv4 address as refid.
use the IP of the server we last synced to if it was a IPv4 one.
sometimes we use the average offset between two, in that case just pick
one for the IP.
this scheme naturally fails when we query IPv6 servers and have to reply
to IPv4 NTP3 (or even older NTP versions) clients - refid stays at 0 then.
this is a protocol limitation, nothing we can do about it.
- henning@cvs.openbsd.org 2004/10/13 12:37:47
[ntp_msg.c]
fall cleaning
- henning@cvs.openbsd.org 2004/10/13 13:19:44
[client.c]
thinko, htonl() -> ntohl(). as we don't use the value in question effect
zero
- henning@cvs.openbsd.org 2004/10/13 13:35:19
[client.c ntp.h ntp_msg.c]
in struct ntp_msg, rename "distance" to "rootdelay" to closer match RFCs
and such
- henning@cvs.openbsd.org 2004/10/13 14:02:50
[ntp.c server.c]
set rootdelay in replies.
inherit rootdelay from the delay from the last client update from the peer
that we picked last time to adjust the local clock.
in some cases we use the average offset between two peers' client updates,
then use the average delay between the two as well.
- dtucker@cvs.openbsd.org 2004/10/14 09:35:48
[client.c ntpd.h server.c]
Have ntpd use IPTOS_LOWDELAY; ok henning@
- dtucker@cvs.openbsd.org 2004/10/15 01:58:04
[client.c server.c]
Only set IPTOS_LOWDELAY on IPv4 interfaces; pointed out by phessler,
ok henning
- henning@cvs.openbsd.org 2004/10/22 21:17:37
[client.c ntp.c ntp_msg.c ntpd.h server.c]
in server mode reply with stratum from the peer that we currently prefer
plus one
- henning@cvs.openbsd.org 2004/10/22 21:24:20
[ntp_msg.c]
oups
- (dtucker) [client.c] Use SA_LEN instead of ss_len.
- (dtucker) [ntpd.c] Move seed_rng earlier.
- (dtucker) [includes.h ntpd.c version.h] Add a version identifier.
- (dtucker) [configure.ac ntp.c] Test for the presence of <paths.h>
- (dtucker) [defines.h] Define MAX() macro if not already defined.
- (dtucker) [ntpd.c] Add SCCS tag marker so 'what' works too.
35a952c759