]>
git.ipfire.org Git - thirdparty/chrony.git/log
Miroslav Lichvar [Fri, 24 Jun 2011 11:27:30 +0000 (13:27 +0200)]
Add configure option for sendmail path
Miroslav Lichvar [Fri, 24 Jun 2011 10:30:48 +0000 (12:30 +0200)]
Generate version and date in man pages
Miroslav Lichvar [Fri, 24 Jun 2011 10:27:54 +0000 (12:27 +0200)]
Convert make_release to shell script
Miroslav Lichvar [Thu, 23 Jun 2011 15:49:18 +0000 (17:49 +0200)]
Remove chrony.lsm
Miroslav Lichvar [Thu, 23 Jun 2011 13:23:16 +0000 (15:23 +0200)]
Step also cooked select timestamp in scheduler slew handler
Miroslav Lichvar [Thu, 23 Jun 2011 11:47:55 +0000 (13:47 +0200)]
Detect infinite loop in scheduler
If more timeouts were handled than there were in the timer queue on
start, assume some code is scheduling timeouts with negative delays and
abort. Make the actual limit higher in case the machine is temporarily
overloaded and dispatching the handlers takes more time than was delay
of a scheduled timeout.
Miroslav Lichvar [Thu, 23 Jun 2011 11:42:04 +0000 (13:42 +0200)]
Add asserts for timeout delays
Miroslav Lichvar [Thu, 23 Jun 2011 10:13:51 +0000 (12:13 +0200)]
Add waitsync command
Miroslav Lichvar [Wed, 15 Jun 2011 10:11:32 +0000 (12:11 +0200)]
Fix current_total_tick calculation
Miroslav Lichvar [Tue, 14 Jun 2011 18:45:48 +0000 (20:45 +0200)]
Don't call driver read_freq in LCL_ReadAbsoluteFrequency
Miroslav Lichvar [Tue, 14 Jun 2011 17:30:42 +0000 (19:30 +0200)]
Create logdir before making first tracking write
Miroslav Lichvar [Tue, 14 Jun 2011 17:18:41 +0000 (19:18 +0200)]
Don't update drift file on first reference update
Miroslav Lichvar [Tue, 14 Jun 2011 19:03:16 +0000 (21:03 +0200)]
Don't reset kernel frequency on start without drift file
Miroslav Lichvar [Mon, 13 Jun 2011 16:17:33 +0000 (18:17 +0200)]
Don't use uninitialized values
Miroslav Lichvar [Mon, 13 Jun 2011 15:03:30 +0000 (17:03 +0200)]
Validate leap status in refclock samples
Miroslav Lichvar [Mon, 13 Jun 2011 15:02:42 +0000 (17:02 +0200)]
Set leap status by enum
Miroslav Lichvar [Mon, 13 Jun 2011 13:34:16 +0000 (15:34 +0200)]
Store reference IDs in uint32_t
Miroslav Lichvar [Mon, 13 Jun 2011 11:49:46 +0000 (13:49 +0200)]
Don't limit refclock driver name to 4 chars
Miroslav Lichvar [Fri, 10 Jun 2011 16:55:20 +0000 (18:55 +0200)]
Don't try to recover from our own time steps
Miroslav Lichvar [Fri, 10 Jun 2011 16:29:41 +0000 (18:29 +0200)]
Try to handle unexpected backward time jumps
Miroslav Lichvar [Thu, 9 Jun 2011 12:32:22 +0000 (14:32 +0200)]
Set version string in config.h
Miroslav Lichvar [Thu, 9 Jun 2011 12:31:04 +0000 (14:31 +0200)]
Add .deps to .gitignore
Miroslav Lichvar [Thu, 9 Jun 2011 11:54:46 +0000 (13:54 +0200)]
Make .deps order-only prerequisite
Miroslav Lichvar [Mon, 6 Jun 2011 19:32:38 +0000 (21:32 +0200)]
Remove kernel version check from rtc code
It should work with all currently supported kernels (>= 2.2.0).
Miroslav Lichvar [Mon, 6 Jun 2011 15:41:14 +0000 (17:41 +0200)]
Determine hz and shift from sysconf(_SC_CLK_TCK) when available
Miroslav Lichvar [Mon, 6 Jun 2011 15:00:27 +0000 (17:00 +0200)]
Log final version specific details
Miroslav Lichvar [Mon, 6 Jun 2011 14:13:12 +0000 (16:13 +0200)]
Drop support for old readonly adjtime
Miroslav Lichvar [Mon, 6 Jun 2011 14:04:42 +0000 (16:04 +0200)]
Drop support for pre 2.2 Linux kernels
Miroslav Lichvar [Mon, 6 Jun 2011 11:56:27 +0000 (13:56 +0200)]
Support Linux 3.0 and later
Miroslav Lichvar [Wed, 25 May 2011 14:59:40 +0000 (16:59 +0200)]
Accept packets with compatible NTP versions
All incoming NTP packets are now required to have version 2, 3 or 4.
Miroslav Lichvar [Tue, 24 May 2011 16:06:01 +0000 (18:06 +0200)]
Set source IPv6 address on NTP reply
This is needed on systems with multiple IPv6 addresses to reply with
the same source address as the destination address of the NTP request.
Miroslav Lichvar [Mon, 23 May 2011 15:57:28 +0000 (17:57 +0200)]
Merge CCWARNFLAGS with CFLAGS
Miroslav Lichvar [Mon, 23 May 2011 15:51:21 +0000 (17:51 +0200)]
Use config.h
Miroslav Lichvar [Mon, 23 May 2011 13:38:14 +0000 (15:38 +0200)]
Use object dependencies in Makefile
Miroslav Lichvar [Wed, 4 May 2011 10:29:40 +0000 (12:29 +0200)]
Update NEWS
Miroslav Lichvar [Mon, 2 May 2011 11:21:50 +0000 (13:21 +0200)]
Update versions in man pages
Miroslav Lichvar [Thu, 28 Apr 2011 15:32:13 +0000 (17:32 +0200)]
Modify weight calculation again
Dividing the weights by variance or unweighted variance seems to have a
significant negative impact on response with normally distributed
network delays.
Divide by the difference between the mean and minimum distance instead.
It should be stable as there is no loop and the response seems to be a
good compromise between the original minimum distance weighting which
works well with normally distributed delays and the variance weighting
which works well with exponentially distributed delays.
Miroslav Lichvar [Thu, 28 Apr 2011 15:12:39 +0000 (17:12 +0200)]
Revert using unweighted variance in weight calculation
This reverts commit
165e6805ab2e46b6fe6b627fcc5fd72d95802c39 .
Miroslav Lichvar [Wed, 20 Apr 2011 10:38:34 +0000 (12:38 +0200)]
Increase smoothing factor in refclock variance
Miroslav Lichvar [Mon, 18 Apr 2011 10:46:26 +0000 (12:46 +0200)]
Update NEWS
Miroslav Lichvar [Mon, 18 Apr 2011 10:52:52 +0000 (12:52 +0200)]
Add another chrony.conf example
Miroslav Lichvar [Mon, 18 Apr 2011 10:36:02 +0000 (12:36 +0200)]
Ignore extra samples in reported nruns
Miroslav Lichvar [Wed, 13 Apr 2011 16:43:25 +0000 (18:43 +0200)]
In weight calculation use unweighted variance from last regression
This fixes a positive feedback where weights could reach inf.
Also change the SD_TO_DIST_RATIO constant to get close to the original
response.
Miroslav Lichvar [Tue, 12 Apr 2011 14:40:22 +0000 (16:40 +0200)]
Allow changing tick up to max_tick_bias
Miroslav Lichvar [Mon, 11 Apr 2011 15:54:01 +0000 (17:54 +0200)]
Change default maxclockerror to 1 ppm
Miroslav Lichvar [Mon, 11 Apr 2011 15:52:04 +0000 (17:52 +0200)]
Set reference time to last sample instead of time on update
This is done mainly to fix reported root dispersion to include max clock
error after selecting another source without new sample.
Miroslav Lichvar [Fri, 8 Apr 2011 14:53:11 +0000 (16:53 +0200)]
Update refclock documentation
Miroslav Lichvar [Thu, 7 Apr 2011 16:32:50 +0000 (18:32 +0200)]
Add new commands to protocol comment in candm.h
Miroslav Lichvar [Thu, 7 Apr 2011 16:20:35 +0000 (18:20 +0200)]
Update client copyright message
Miroslav Lichvar [Thu, 7 Apr 2011 16:17:37 +0000 (18:17 +0200)]
Assert there are no unhandled commands in cmdmon
Miroslav Lichvar [Thu, 7 Apr 2011 16:16:39 +0000 (18:16 +0200)]
Add reselectdist command
Miroslav Lichvar [Thu, 7 Apr 2011 14:17:58 +0000 (16:17 +0200)]
Don't add \n to chronyc command line arguments
This fixes parsing of some commands.
Miroslav Lichvar [Thu, 7 Apr 2011 12:44:56 +0000 (14:44 +0200)]
Revert marking offline sources as unreachable
Miroslav Lichvar [Wed, 6 Apr 2011 14:59:40 +0000 (16:59 +0200)]
Don't crash when sources report is requested soon after start
Miroslav Lichvar [Wed, 6 Apr 2011 14:58:12 +0000 (16:58 +0200)]
Add include directive
Miroslav Lichvar [Tue, 5 Apr 2011 16:14:05 +0000 (18:14 +0200)]
Add configure option to force retry on DNS failure
This is apparently needed on system which keep nameservers specified
in /etc/resolv.conf even when there is no network connection. Should be
used with care as invalid names will be retried forever.
Miroslav Lichvar [Tue, 5 Apr 2011 14:32:50 +0000 (16:32 +0200)]
Don't update empty sourcestats on clock update
Miroslav Lichvar [Tue, 15 Feb 2011 17:55:34 +0000 (18:55 +0100)]
Fix warnings produced by latest gcc
Miroslav Lichvar [Tue, 15 Feb 2011 16:22:40 +0000 (17:22 +0100)]
Reduce Linux driver verbosity
Miroslav Lichvar [Tue, 15 Feb 2011 15:56:30 +0000 (16:56 +0100)]
Don't log error on opening driftfile
Log frequency and skew values read from the driftfile instead.
Miroslav Lichvar [Mon, 14 Feb 2011 17:21:38 +0000 (18:21 +0100)]
Make starting log even in debug mode
Miroslav Lichvar [Mon, 14 Feb 2011 17:19:58 +0000 (18:19 +0100)]
Don't log in signal handler
Miroslav Lichvar [Fri, 11 Feb 2011 16:54:45 +0000 (17:54 +0100)]
Don't use uninitialized memory when setting RTC time
Miroslav Lichvar [Fri, 11 Feb 2011 16:43:56 +0000 (17:43 +0100)]
Use system headers for Linux RTC support
Miroslav Lichvar [Fri, 11 Feb 2011 16:31:38 +0000 (17:31 +0100)]
Fix compiler warnings in PPS configure test
Miroslav Lichvar [Fri, 28 Jan 2011 12:21:56 +0000 (13:21 +0100)]
Update NEWS
Miroslav Lichvar [Fri, 28 Jan 2011 11:57:03 +0000 (12:57 +0100)]
Update credits
Miroslav Lichvar [Fri, 28 Jan 2011 11:56:09 +0000 (12:56 +0100)]
Remove CVS headers
Miroslav Lichvar [Thu, 27 Jan 2011 12:05:26 +0000 (13:05 +0100)]
Update copyright
Miroslav Lichvar [Tue, 25 Jan 2011 16:51:19 +0000 (17:51 +0100)]
Ignore reselectdist for refclocks
Miroslav Lichvar [Tue, 25 Jan 2011 16:22:00 +0000 (17:22 +0100)]
Make importance of stratum in source selection configurable
Instead of always selecting the source with minimum stratum, add weighted
stratum to the distance when comparing selectable sources. The weight
can be configured with new stratumweight directive and can be set to
zero to ignore stratum completely, by default 1.0.
Miroslav Lichvar [Mon, 24 Jan 2011 17:04:33 +0000 (18:04 +0100)]
Disable variance source test for now
It seems it triggers even with one source alone if it's close and
accurate or polling interval is long enough. It would need to include
max_clock_error.
Miroslav Lichvar [Mon, 24 Jan 2011 16:23:34 +0000 (17:23 +0100)]
Add outlyer source status to cmdmon protocol
This is not used yet.
Miroslav Lichvar [Mon, 24 Jan 2011 16:13:28 +0000 (17:13 +0100)]
Print sources with bad stats in client as unreachable
Miroslav Lichvar [Mon, 24 Jan 2011 16:09:00 +0000 (17:09 +0100)]
Select source with minimum distance using a scoring system
Each source has a score against currently selected source which is
updated (multiplied by ratio of their distances) when one of the two
sources has a new sample. When the score reaches a limit, the source
will be selected. This should allow to slowly select the source with
minimum distance without frequent reselecting.
To avoid switching between sources with very variable distances (e.g. on
LAN or when upstream server uses a longer polling interval), sources
that are currently not selected are penalized by a fixed distance. This
can be configured with new reselectdist directive (100 microseconds by
default).
Miroslav Lichvar [Wed, 19 Jan 2011 14:29:49 +0000 (15:29 +0100)]
Set status on doffset and dfreq commands
Miroslav Lichvar [Wed, 19 Jan 2011 13:44:10 +0000 (14:44 +0100)]
Reply with status invalid instead of bad length on invalid command
Miroslav Lichvar [Tue, 18 Jan 2011 17:07:46 +0000 (18:07 +0100)]
Don't leak descriptors to sendmail
Miroslav Lichvar [Mon, 20 Dec 2010 13:12:47 +0000 (14:12 +0100)]
Don't forget to shift last class dispatch timevals on step
Miroslav Lichvar [Fri, 17 Dec 2010 17:05:09 +0000 (18:05 +0100)]
Don't send packet in last auto_offline transmit timeout
Miroslav Lichvar [Fri, 17 Dec 2010 16:51:00 +0000 (17:51 +0100)]
Fix switching offline auto_offline source to online
Miroslav Lichvar [Fri, 17 Dec 2010 15:09:12 +0000 (16:09 +0100)]
Enforce timeout class separation from last dispatched timeout
Miroslav Lichvar [Fri, 17 Dec 2010 14:38:23 +0000 (15:38 +0100)]
Use enum for scheduler timeout classes
Miroslav Lichvar [Fri, 17 Dec 2010 13:52:39 +0000 (14:52 +0100)]
Fix crash when timeout is removed from its handler
Remove the timeout before dispatching the handler, and allow
calling SCH_RemoveTimeout() with nonexistent id.
Miroslav Lichvar [Thu, 16 Dec 2010 17:51:40 +0000 (18:51 +0100)]
Make unsynchronised entry in tracking log on start
Miroslav Lichvar [Thu, 16 Dec 2010 17:27:38 +0000 (18:27 +0100)]
Delay selecting source on start
On start, when servers are reachable and use the same polling interval,
wait for them to have the same reachability register (which corresponds
to the number of samples in sourcestats) before selecting one.
Miroslav Lichvar [Thu, 16 Dec 2010 15:55:59 +0000 (16:55 +0100)]
Don't update local reference unnecessarily
Local reference is now updated only when a new source is selected or
match_addr is equal to the selected source, match_addr 0 is no longer
treated specially.
Miroslav Lichvar [Thu, 16 Dec 2010 15:01:29 +0000 (16:01 +0100)]
Track reachability in sources module
Add new flag to source struct to indicate when source is selectable
(packets with good headers are received) and use a reachability
register for last 8 samples instead of the reachable flag. Source
drivers now provide only reachability updates.
Miroslav Lichvar [Thu, 16 Dec 2010 12:28:44 +0000 (13:28 +0100)]
Apply jitter test only on NTP sources for now
With refclocks it seems to fail easily, users would have to set the delay
option appropriately.
Miroslav Lichvar [Wed, 15 Dec 2010 16:48:30 +0000 (17:48 +0100)]
Allow selecting source only when last regression was successful
Miroslav Lichvar [Wed, 15 Dec 2010 13:25:40 +0000 (14:25 +0100)]
Fix jitter test in source selection
Distance should be compared to square root of variance.
Miroslav Lichvar [Tue, 14 Dec 2010 16:00:53 +0000 (17:00 +0100)]
Divide regression weights by stddev instead of minimum distance
This improves accuracy with large but stable network delays.
Miroslav Lichvar [Thu, 9 Dec 2010 15:06:42 +0000 (16:06 +0100)]
Add small randomness to spacing between samples for one server
Miroslav Lichvar [Thu, 9 Dec 2010 14:14:05 +0000 (15:14 +0100)]
Shorten initial delay and sampling separation to 0.2 seconds
Miroslav Lichvar [Thu, 9 Dec 2010 13:34:29 +0000 (14:34 +0100)]
Fix separation of timeouts scheduled for exactly the same time
Miroslav Lichvar [Tue, 7 Dec 2010 15:21:24 +0000 (16:21 +0100)]
Add reselect command
Miroslav Lichvar [Mon, 6 Dec 2010 14:32:06 +0000 (15:32 +0100)]
Require password for clients command
Miroslav Lichvar [Mon, 6 Dec 2010 14:01:01 +0000 (15:01 +0100)]
Update chronyc help text
Miroslav Lichvar [Mon, 6 Dec 2010 13:53:37 +0000 (14:53 +0100)]
Slowly back off sampling rate when selected source loses connectivity
Miroslav Lichvar [Mon, 6 Dec 2010 13:29:40 +0000 (14:29 +0100)]
Require valid_header before accumulating sample
Split test7 into two and to accumulate the sample require that all tests
pass, except the one which checks packet stratum is not higher than ours.
Also, don't mark the source as unreachable when only test4c fails.