summaryrefslogtreecommitdiff
path: root/ntpshm.c
Commit message (Collapse)AuthorAgeFilesLines
* splint and cppcheck cleanup. All regression tests pass.Eric S. Raymond2013-09-291-0/+2
|
* Change gpsd_report to no longer use a global. All regression tests pass.Eric S. Raymond2013-09-291-60/+89
| | | | | | | | This is a large, ugly change. But without it we can't troubleshoot the ICP/IP-source initialization bug properly - colliding definitions of gpsd_report() were interfering with error reporting early in gpsd runs. More cleanup work remains to be done, but at least this is working.
* Remove some thread debugging that stepped on other messages.Eric S. Raymond2013-09-281-2/+1
|
* More cppcheck cleanup.Eric S. Raymond2013-09-171-1/+2
|
* Remove dead code found by clang.Eric S. Raymond2013-06-071-1/+0
|
* Clean up more static-analysis warnings.Eric S. Raymond2013-04-301-1/+1
|
* Refactoring ntpshm.c.Michael Tatarinov2013-03-251-6/+3
| | | | Signed-off-by: Eric S. Raymond <esr@thyrsus.com>
* Lose the assumption that socket_t is an integer.Eric S. Raymond2012-09-221-1/+1
|
* Nuke trailing whitespace in C source.Eric S. Raymond2012-09-071-31/+31
|
* Add #ifdef line which was deleted accidentally.Bernd Zeimetz2012-07-011-0/+1
|
* Include unistd.h as workaround for timepps.h missing it.Bernd Zeimetz2012-07-011-12/+0
| | | | | Also remove a duplicate include of timepps.h from ntpshm.h, moving the comment to gpsd.h-tail
* Fix three PPS factoring issues.Michael Tatarinov2012-06-011-79/+77
| | | | | | | | | 1. Activates PPS support for RS-232 or USB devices only. 2. Сorrectly completes the gpsd_ppsmonitor thread. 3. and I hope this fixes the PPS support when the TIOCMIWAIT isn't defined but kPPS is present. Signed-off-by: Eric S. Raymond <esr@thyrsus.com>
* Do not activate pps subsystem for nmea2000.Reinhard Arlt2012-05-301-0/+8
|
* More Coverity suppressions.Eric S. Raymond2012-05-121-1/+1
| | | | | Also, remove our custom scan script, I'm going to publish a more polished version.
* Still working on pacifying Coverity. All regression tests pass.Eric S. Raymond2012-05-111-0/+3
|
* More comment fixes.Eric S. Raymond2012-05-111-3/+3
|
* Shameless microtweaking.Eric S. Raymond2012-05-101-1/+1
| | | | | | | for (;;) generates tighter code than while (1). No practical difference but it bugs me anyway. All regression tests pass.
* Fix type error in ntpshm.cGary E. Miller2012-05-091-1/+1
| | | | Found by: Beat Bolli <bbolli@ewanet.ch>
* Fix a typo in a previous patch.Gary E. Miller2012-05-031-2/+6
| | | | | | | Fix a serious bug in ntpshm_put() caused by using ship_to_ntp, which is packet synchronous, in the PPS thread context which is asynchronous to packets. I'm not totally content yet that things are aas they should be.
* Add logging for unlogged NTP error path.Gary E. Miller2012-05-031-8/+16
|
* Allow use of the CTS line for NNTP.Eric S. Raymond2012-05-011-1/+1
| | | | | | | | | | | | Nathan Knotts wrote: "I'm currently using the Xtreme104 12 port serial board which only has the CTS handshaking lines exposed. Previous gpsd releases had support for the cts handshaking pin via compile time options, then later, command line options. As of release 3.5, these options have been removed and automatic detection occurs. However, the bitmask defining which control lines to listen to (PPS_LINE_TIOC) does not include the CTS line. I've added TIOCM_CTS to that bitmask to include CTS capability. This has been tested to work with a ublox chip outputting NMEA."
* Leap-second warning status from GPS subframe info it's passed to ntpd.Eric S. Raymond2012-04-141-1/+4
| | | | | Patch concept from Michael Tatarinov. Modified to use a session context member rather than a global.
* Slightly better error message for PPS thread.Gary E. Miller2012-04-041-1/+2
|
* Log the last way PPS thread can exit silently.Gary E. Miller2012-04-041-1/+6
|
* Add an error out message for PPS thread.Gary E. Miller2012-04-041-1/+3
|
* Remove pps-pin option; the PPS code now listens on all handshake pins.Eric S. Raymond2012-04-041-6/+11
| | | | All regression tests pass.
* Fix the byuild when pps=false and sys/timepps.h is present.Eric S. Raymond2012-03-011-2/+2
|
* Only ship to chrony when we ship to NTP direct.Eric S. Raymond2012-02-291-1/+1
|
* Revert the timing of the PPS threasd launch to what it was in 3.4.Eric S. Raymond2012-02-281-7/+5
| | | | | mlichvar says kernel PPS worked in 3.4, though userspace PPS didn't due to problems with the privilege-dropping.
* Add some PPS progress logging.Eric S. Raymond2012-02-271-2/+6
|
* PPS fixes from Michael Tatarinov.Eric S. Raymond2012-02-271-12/+19
| | | | | | | With these, he erports PPS working under BSD and Savannah bug #34959: "gpsd can't activate the device after deactivation" fixed. All regression tests pass.
* Typo fix.Eric S. Raymond2012-02-271-1/+1
|
* Attempted fix for Savannah bug #34959.Eric S. Raymond2012-02-271-12/+19
|
* Attempted fix workaround for the TIOCMIWAIT-hang bug.Eric S. Raymond2012-02-261-10/+25
| | | | | | | | | | The PPS thread hangs whenever you changew baud (or possibly other serial params) on the devices, so spawning it needs to be deferred until after autobaud and sync lock have been achieved. The call to terminatec the thread on device deactivation han't been put in yet.Pushing this version as a proof of concept so Gary can test it.
* Trial version of PPS drift message. Protocol minor version bumped.Eric S. Raymond2012-02-251-2/+5
| | | | All regression tests pass, code splints clean.
* Typo fix.Eric S. Raymond2012-02-251-1/+1
|
* Shut splint up.Eric S. Raymond2011-10-261-0/+2
|
* Used clock_gettime() instead of gettimeofday() if possible.Michael Tatarinov2011-10-201-1/+1
| | | | Signed-off-by: Eric S. Raymond <esr@thyrsus.com>
* Comment tweaks.Michael Tatarinov2011-10-201-2/+2
| | | | Signed-off-by: Eric S. Raymond <esr@thyrsus.com>
* Document kernel bug that interferes with non-kernel PPS, and the workaround.Eric S. Raymond2011-10-121-1/+1
|
* Improve PPS error handling.Gary E. Miller2011-10-111-2/+4
|
* Another PPS log tweak.Gary E. Miller2011-10-111-1/+1
|
* Improve PPS log messages.Gary E. Miller2011-10-111-15/+14
|
* const and splint cleanup.Eric S. Raymond2011-10-091-17/+18
|
* We can't say no request for PPS on RTS or RI has been seen, any more.Eric S. Raymond2011-10-051-5/+1
|
* Replace pps_on_cts with pps_pin for more flexibility.Eric S. Raymond2011-10-051-7/+1
| | | | | | | Yes, we had a user request for this - a guy using a Jackson Labs rubidium-oscillator clock whose easy alternatives are RTS and RI. All regression tests pass. Code splints clean.
* Documentation tweaks.Eric S. Raymond2011-10-051-0/+14
|
* Small refactoring step and optimization.Eric S. Raymond2011-09-221-11/+17
|
* Fix typo in log message.Gary E. Miller2011-08-221-1/+1
|
* Try to open the right chrony socket...Eric S. Raymond2011-08-221-13/+7
| | | | ...now that command-line devices are opened vefore privilege-dropping.
View Lorry Controller Status