summaryrefslogtreecommitdiff
path: root/usr.sbin/ntpd
AgeCommit message (Expand)Author
2010-08-03fix linecount bug with comments spanning multiple linesHenning Brauer
2010-07-01don't need to setup the resolver early now that dns lookups are doneEric Faurot
2010-05-26Move imsg into libutil and add a man page.Nicholas Marriott
2010-05-26Rename some imsg bits to make namespace collisions less likely buf toNicholas Marriott
2010-04-27imsg.h does not need sys/tree.h.Nicholas Marriott
2010-04-20Get rid of MAXSENSORDEVICES. Gaps in sensordev lists are now handledTheo de Raadt
2010-04-07Remove XXX comment and just close received fd if calloc() fails.Nicholas Marriott
2009-09-15Enclose repeated buffer draining code in a new msgbuf_drain()Jacek Masiulaniec
2009-08-08Make imsg_clear close and free any fds in the received fd queue as well asNicholas Marriott
2009-07-23make buf_write() behave like msgbuf_write(): send out only theEric Faurot
2009-06-25incorrect .PATH setting, discovered with nicmTheo de Raadt
2009-06-24recvmsg ENOENT noncritical tooHenning Brauer
2009-06-17ENOPROTOOPT is non-fatal on recvfrom, can apparently happen with ipvshitHenning Brauer
2009-06-11effectively un-do r1.111, it fucks machines with bad clocks royally.Henning Brauer
2009-06-08revert this change by eric@:David Gwynne
2009-06-07Change the way fds passed over a socket are retreived on the receiving side.Eric Faurot
2009-06-07Make the imsg protocol network-safe.Eric Faurot
2009-06-06change the imsg header fields a bit to prepare for upcoming changes.Eric Faurot
2009-06-06If ntpd decides the clock has become unsynced, reset the number of savedChris Kuethe
2009-06-06Turns up the polling rate of sensors and makes ntpd calculate frequencyChris Kuethe
2009-06-06make ntpd imsg-in-a-lib ready as well. extensive testing done, noPierre-Yves Ritschard
2009-06-04After calling adjfreq to correct the clock's rate, measure and fix the clockChris Kuethe
2009-06-01no we don't need endpwentHenning Brauer
2009-05-31use the SCM_TIMESTAMP socket option to get the time we received theHenning Brauer
2009-05-20fix looking at the mode bits so that we do not reply to broadcast and otherHenning Brauer
2009-05-18move text on weight to where it is first referred to;Kevin Steves
2009-05-13when using a timedelta sensor for -s, if the sensor is invalid duringKevin Steves
2009-04-22ignore replies with timestamps after 2030 to prevent time_t / tv_sec wrapsHenning Brauer
2009-03-31Fixed memory leaks which would occur if the second of two memoryTobias Stoeckmann
2009-03-04I should know outside data must be vis(3)'d.Kevin Steves
2009-03-04received refid should not be converted to host byte order; ok henning@Kevin Steves
2009-03-04log reason when not synced; ok henning@Kevin Steves
2009-02-12-v logs debug adjtime calls also; ok henning@Kevin Steves
2009-02-11move prototypes that are really in ntp.c; ok henning@Kevin Steves
2009-02-10log tiny frequency adjustments at debug only.Kevin Steves
2009-02-10endservent() not needed here; ok henning@Kevin Steves
2009-02-08spelling in commentKevin Steves
2009-02-06rename a #define to be clearer; no binary changeKevin Steves
2009-01-31adjtime has been logged at 32ms and greater for 2.5 years now.Kevin Steves
2009-01-29log warning when can't read drift file; allows detection ofKevin Steves
2009-01-27fix printf format specification to print max len 4 for refid,Kevin Steves
2009-01-26return the correct refid for a sensor; ok henning@Christian Weisgerber
2009-01-26use monotime instead of wallclock for the report thing as wellHenning Brauer
2008-12-10Correctly skip the internal fd when checking the server sockets. Use PFD_MAXClaudio Jeker
2008-11-10delete operation not needed; from naddyTheo de Raadt
2008-11-10If a bind() spuriously fails (as appears to happen once in a blue moon dueTheo de Raadt
2008-10-17bring in findeol() fix from pfctlHenning Brauer
2008-10-10- use nmea(4) for general sensor example rather than udcf(4)Stuart Henderson
2008-10-02in priv_settime(), we must not adjust the 'next' timestamps on their forHenning Brauer
2008-09-28ntpd does not require you to run rdate -n before it -- it notices such anTheo de Raadt
generated by cgit v1.2.3 (git 2.25.1) at 2024-11-24 11:51:28 +0000