summaryrefslogtreecommitdiffstats
path: root/src/OutputUHD.cpp
Commit message (Collapse)AuthorAgeFilesLines
* Put UHD async msg to separate threadMatthias P. Braendli2017-04-211-53/+57
|
* Fix wrong units for UHD TX rate printoutMatthias P. Braendli2017-03-031-2/+2
|
* Fix UHD LO offset loggingMatthias P. Braendli2017-02-251-15/+15
|
* Add ability to set UHD LO offsetMatthias P. Braendli2017-02-251-3/+23
|
* Set uhd timestamps log to std::fixedMatthias P. Braendli2017-02-251-0/+3
|
* Add support for SoapySDRMatthias P. Braendli2017-02-101-0/+3
|
* Fix TIST usage for EDI inputMatthias P. Braendli2017-01-271-5/+7
|
* Remove FAKE_UHD, it's anyway brokenMatthias P. Braendli2017-01-271-19/+0
|
* Replace EDI-to-ETI converter with a dedicated EDI sourceMatthias P. Braendli2017-01-061-5/+5
|
* Refactor flowgraph and puncturing rulesMatthias P. Braendli2016-12-251-1/+1
|
* Remove ModFormatMatthias P. Braendli2016-12-251-1/+1
|
* Fix UHD buffering leading to out of memoryMatthias P. Braendli2016-06-241-6/+3
| | | | | | | | | | | | Since commit d9ef93e, UHD does not backpressure the modulator anymore. If a pipe input is used, the ODR-DabMux before also doesn't get any back-pressure, the modulation thread and the mux run a very high rate. This high rate fills the buffer between OutputUHD and its worker thread, until the out-of-memory killer kills ODR-DabMod. Less impact on ZMQ input, because that is throttled at the mux, but we still have a buffer that can grow in an uncontrolled way
* Set custom names to threadsMatthias P. Braendli2016-06-171-0/+2
|
* Set SCHED_RR prio for all data processing threadsMatthias P. Braendli2016-06-171-0/+6
| | | | This includes Modulator, FIR Filter and UHD worker
* Add TRACE support for timing analysisMatthias P. Braendli2016-06-171-1/+6
|
* Set UHD printout of timestamps to std::fixedMatthias P. Braendli2016-06-171-2/+4
|
* fix segfault on SIGINT before UHD init completesMatthias P. Braendli2016-06-171-2/+1
|
* Use system time instead of usrp time to printMatthias P. Braendli2016-05-261-2/+3
|
* Fix UHD underrun printoutMatthias P. Braendli2016-05-201-2/+2
|
* Embellish warning about timestampMatthias P. Braendli2016-05-201-3/+5
|
* Get rid of FCT discontinuity checkMatthias P. Braendli2016-05-201-18/+0
| | | | The timestamp discontinuity verification takes care of this now
* Improve timestamp handlingMatthias P. Braendli2016-05-201-80/+83
| | | | | Replace PPS representation from double to integer, and analyse in OutputUHD if the timestamp is as expected.
* Refactor OutputUHD, replace double buffering by queueMatthias P. Braendli2016-05-201-106/+66
|
* Fix UHD output timestamp incrementMatthias P. Braendli2015-11-271-1/+1
|
* UHD: do not print length 1 messagesMatthias P. Braendli2015-09-181-1/+10
|
* Keep UHD settings across a modulator restartMatthias P. Braendli2015-09-181-15/+15
|
* Make it possible to use synchronous=1 with internal refMatthias P. Braendli2015-08-261-0/+1
|
* Add support for Ettus GPSDOMatthias P. Braendli2015-08-151-5/+41
|
* Switch to C++11, remove boost::shared_ptrMatthias P. Braendli2015-08-011-2/+1
|
* Replace fixtype by timelock sensor in OutputUHDMatthias P. Braendli2015-07-031-15/+15
|
* Move GPS check out of UHDWorkerMatthias P. Braendli2015-06-121-211/+237
|
* Add some sort of initial gps checkMatthias P. Braendli2015-06-121-74/+147
|
* Cleanup OutputUHD structureMatthias P. Braendli2015-06-121-288/+243
|
* Simplify TIST offset handling, add offset to RCMatthias P. Braendli2015-06-121-1/+1
| | | | This breaks old configuration files using synchronous=1
* Revert priority setting for mod and UHD threadsMatthias P. Braendli2015-06-071-10/+4
| | | | | | | | It is safer to go back to the previous state, given that several users have expressed concerns about reliability. Revert "Set priorities for modulator and UHD threads". This reverts commit c9ea7fb88809d935f3eaeee108415ff5abd17ead.
* Remove debugging printsMatthias P. Braendli2015-06-051-16/+0
|
* Fix intermittent underruns after a restartMatthias P. Braendli2015-06-051-0/+17
| | | | | | | | | | | | | | When using UHD without synchronous=1: After a modulator restart, the input ZMQ buffer is nearly empty, and the modulator actually gets blocked by the ThreadsafeQueue. This is visible by looking at the time deltas printed in the debugging code. This commit adds a minimal form of prebuffering (10 ETI frames) to the ZeroMQ input, and removes the useless minimum required size for the ThreadsafeQueue. In synchronous=1, the issue is not visible because the TIST defines the time to transmit, which will cause the ZMQ buffer to fill.
* Set priorities for modulator and UHD threadsMatthias P. Braendli2015-06-051-4/+10
|
* Logging: use etiLog instead of stderrMatthias Braendli2015-06-021-38/+36
|
* Reduce print of UHD messagesMatthias P. Braendli2015-05-031-11/+40
|
* gps fix detector: add timeoutMatthias P. Braendli2015-05-031-18/+36
|
* Add GPS fix checker, still with debuggingMatthias P. Braendli2015-05-031-3/+75
|
* Use make_shared in DabMod.cppMatthias P. Braendli2015-04-101-14/+14
|
* Add check for FCT validityMatthias P. Braendli2015-04-101-1/+7
|
* Restart whole modulator on FCT discontinuityMatthias P. Braendli2015-02-271-6/+33
|
* Remove incomplete OutputUHD iqbalanceMatthias P. Braendli2015-02-271-6/+0
|
* Restart full modulator on ZeroMQ input overrunMatthias P. Braendli2015-02-221-1/+3
|
* indentationMatthias P. Braendli2015-02-121-42/+43
|
* added reset for static delayJörgen Scott2015-02-091-9/+16
|
* Bug fix for static delayJörgen Scott2015-02-091-4/+4
|
generated by cgit v1.2.3 (git 2.39.1) at 2024-11-11 07:54:18 +0000