summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Adding interface support for remote app (chan_lcr).Andreas Eversberg2010-12-1328-481/+924
| | | | | | | | | | | | | | chan_lcr can be handled as an interface. This way it is possible to (e.g.): - make a SIP phone become an LCR extension with all LCR features. - make conference calls. (untested) - perform parallel ringing. (ISDN phone and SIP phones can ring in parallel.) - do voice recoding. It is still also possible to link chan_lcr directly without interface (as before). Documentation/howto for that will follow.
* Adding various arguments to 'execute' condition and 'execute' action.Andreas Eversberg2010-12-132-2/+18
|
* Workaround for "noise" problems with app_rxfax in current LCR.Peter Schlaile2010-12-131-3/+17
| | | | | | | | | | | The problem: for some reason, asterisk silently switches write_format from SLINEAR to ALAW when using app_rxfax, thereby effectively disabling automatic conversion SLINEAR->ALAW and thereby breaking app_rxfax after the first call to ast_write(). Read: instantly, after the first data frame. Older versions of LCR work flawlessly, until we find out, what really triggers the problem, we at least switch the channel back, since it never makes sense to send something different than ALAW to chan_lcr.
* See commit 728d... It was not applied correctly.Andreas Eversberg2010-12-131-240/+350
|
* Fixed channel selection on outgoing calls on NT-mode multipointAndreas Eversberg2010-12-134-2/+44
| | | | | | | | | | | | | | | On a multipoint bus, it is required to assign a channel with no other alternative allowed. This is required, because an individual phone on the bus may not choose a different channel, while other phones accept the indicated channel. Also an individual phone does not have the information about other available channels. On a point-to-point configuration it is possible anyway. In this multipoint case, the channel assignment is now forced automatically, even if the 'force' keyword is not specified in the "out-channel" of interface.conf. Additionally "lcradmin portinfo" shows the channel selection settings now.
* Remove the last commit, because it was in the wrong branch.Andreas Eversberg2010-12-131-350/+240Star
|
* Completed GSM mobile station support with OsmocomBBAndreas Eversberg2010-12-1311-4859/+2154Star
| | | | OsmocomBB stack can now be attached to LCRl
* Adding forwarding of DTMF and keypad between peersAndreas Eversberg2010-12-133-1/+59
| | | | | | | | | This is needed for peers that require DTMF messages, rather than tones: - GSM mobile stations - Asterisk channel API modified: apppbx.cpp modified: apppbx.h
* Make LCR compile with the latest OpenBSC master branchAndreas Eversberg2010-10-306-1940/+4845
| | | | | | | | | modified: Makefile.in modified: aclocal.m4 modified: config.h.in modified: configure modified: configure.ac modified: gsm_bs.cpp
* [chan_lcr] Fixed caller ID by clearing ast->cid firstAndreas Eversberg2010-10-301-0/+1
| | | | | | | Tests with asterisk 1.6.2.13 showed that callerid did not work unless the ast->cid structure is set to zero. Thanx for Marcello for prividing this fix.
* [chan_lcr] Fixed uninitialized variable in ast_read()Andreas Eversberg2010-09-281-1/+1
| | | | Thanx to Martin for this bug report.
* [chan_lcr] Added second caller ID (ANI) in case the caller ID is user provided.Andreas Eversberg2010-09-281-2/+2
|
* Fix of vbox announcing bugAndreas Eversberg2010-09-201-5/+4Star
| | | | | | | | | Hi, here's a simple patch that makes the announcement work again when using the vbox in lcr. Bodo
* Fixed numbering plan in GSM MS supportAndreas Eversberg2010-09-201-7/+4Star
|
* Removed debugging output from chan_lcrAndreas Eversberg2010-08-041-0/+2
|
* Removed ast_log again for testing. Seems that it also causes dead-locks when ↵Andreas Eversberg2010-07-261-1/+2
| | | | it is used this way.
* Hi all,Andreas Eversberg2010-07-252-64/+34Star
| | | | | | | | | as described in http://www.gnu.org/software/hello/manual/autoconf/Files.html, AC_CHECK_FILE breaks when cross configuring/compiling LCR. Attached patch replaces AC_CHECK_FILE by 'test -d' and some extra code in configure.ac. Cheers Arny
* Fixed API and incomming calls from Osmocom-BB.Andreas Eversberg2010-07-219-595/+878
| | | | | | | Conflicts: Makefile.in configure
* Enabled ast_log again and changed usleep during trylock to 1.Andreas Eversberg2010-07-211-5/+4Star
|
* Removed the lock-debugging for chan_lcr.Andreas Eversberg2010-07-212-117/+0Star
| | | | | | | Apply reverse for future debugging. modified: chan_lcr.c modified: select.c
* Another test patch.Andreas Eversberg2010-07-091-4/+5
|
* Since this banch only is for debugging the chan_lcr deadlock problem, i will ↵Andreas Eversberg2010-07-081-2/+2
| | | | commit from now on without comments.
* Another "trylock" test.Andreas Eversberg2010-07-071-1/+6
|
* The "trylock" test patch.Andreas Eversberg2010-07-061-2/+6
|
* Disabled log_ast and replaced it by simple printf.Andreas Eversberg2010-07-051-4/+5
|
* Hopefully fixed the double log bug that causes so much trouble.Andreas Eversberg2010-06-301-2/+5
|
* More debug.Andreas Eversberg2010-06-241-0/+8
|
* More debugAndreas Eversberg2010-06-222-0/+7
|
* Debugging select loop. Here the process freezes.Andreas Eversberg2010-06-171-0/+11
|
* Fixed call forwarding after no response. Instread of dialing the old extensionAndreas Eversberg2010-06-153-8/+8
| | | | | | | | again, the forwarding number is dialed after timeout. modified: apppbx.cpp modified: apppbx.h modified: socket_server.c
* Added more detailed debugging for the locking problem.Andreas Eversberg2010-06-141-20/+39
|
* Added additional debug logging to lcr_read.Andreas Eversberg2010-06-091-1/+4
|
* Removed LF from debug code.Andreas Eversberg2010-06-091-1/+1
|
* Updated debugging patch.Andreas Eversberg2010-06-081-1/+3
|
* Test patch #1 for chan_lcr, to debug the Asterisk freeze problem.Andreas Eversberg2010-06-044-4801/+1984Star
|
* Added detection of tone_zone_sound structure name. Thanx to Janis for thisAndreas Eversberg2010-06-034-2/+176
| | | | | | | | | patch. modified: Makefile.am modified: Makefile.in modified: configure modified: configure.ac
* Tested and fixed the new OpenBSC API.Andreas Eversberg2010-05-304-1900/+4750
| | | | | | | modified: Makefile.in modified: config.h.in modified: configure modified: gsm_bs.cpp
* Splitted GSM support into BS (network) and MS (mobile) part.Andreas Eversberg2010-05-3126-5685/+4126Star
|
* Fixed LCR to work with the current API of OpenBSC.Andreas Eversberg2010-05-314-36/+70
| | | | | | | modified: Makefile.am modified: Makefile.in modified: README modified: gsm.cpp
* Added patch to play ringing tone when connected but the call is forwardedAndreas Eversberg2010-05-282-0/+18
| | | | and ringing again. Thanx to Jacek for this patch.
* Fix for redirection number. Thanx to Dennis for this bugfix.Andreas Eversberg2010-05-282-2/+2
| | | | | | | I changed it a bit, so the redirection number is converted from national/international format into numerical format. (adding prefixes) If there is still a problem with it, let me know.
* Added Progress messages. (receive only)Andreas Eversberg2010-05-238-6/+139
|
* Fixed forwarding of sending-complete information.Andreas Eversberg2010-05-204-7/+11
| | | | | | | modified: Makefile.in modified: README modified: action.cpp modified: apppbx.cpp
* Fixes a locking bug in chan_lcr. Thanx to WIMPy for that report.Andreas Eversberg2010-04-301-1/+1
| | | | modified: chan_lcr.c
* fix last remnant of "extern" vs. incorrect "external" confusion, correct ↵Andreas Eversberg2010-04-062-2/+2
| | | | | | | | | spelling Patch against current git. Signed-off-by: Andreas Mohr <andi@lisas.de> Signed-off-by: Andreas Eversberg <jolly@eversberg.eu>
* Add .gitignoreKarsten Keil2010-03-251-0/+18
| | | | Signed-off-by: Karsten Keil <kkeil@linux-pingi.de>
* Revert "Add .gitignore"Karsten Keil2010-03-255-868/+564Star
| | | | | | Only .gitignore should be updated This reverts commit 51655a18970b6117e6102f6ca7541b53fb6f568e.
* Add .gitignoreKarsten Keil2010-03-255-564/+868
| | | | Signed-off-by: Karsten Keil <kkeil@linux-pingi.de>
* Just a small bug fix: Extension matching will be performed by extension's ↵Andreas Eversberg2010-03-241-1/+1
| | | | | | number, not by their name. I really don't know why there is suddenly this bug. Earlier versions did this correctly.
* Bugreport by Christian:Andreas Eversberg2010-03-231-1/+1
| | | | | | | | | | gsm.cpp: In member function .void Pgsm::bchannel_send(unsigned int, unsigned int, unsigned char*, int).: gsm.cpp:249: error: overloaded function with no contextual type information make[1]: *** [gsm.o] Error 1 I hope that this will fix the compiler problem.