summaryrefslogtreecommitdiffstats
path: root/src/voicecall.c
AgeCommit message (Collapse)AuthorFilesLines
2009-12-24Use error label instead of err or err_outMarcel Holtmann1-4/+4
2009-12-17Style: Fix a few more occurences of if/elseDenis Kenzior1-1/+2
See Linux coding style document for details
2009-12-16Support 51.011 EFecc format.Andrzej Zaborowski1-4/+43
EFecc has different formats in G2 and G3 UICC specs.
2009-12-04Optimize: call list signal emissionDenis Kenzior1-9/+0
2009-12-04Update voicecall atom to use new callid apiDenis Kenzior1-9/+7
2009-12-04Add utility to get the next free call idDenis Kenzior1-0/+11
2009-12-01Fix: Simplify logic to use g_memdupDenis Kenzior1-5/+3
2009-12-01Fix: Be more signed/unsigned correctDenis Kenzior1-3/+3
2009-11-13Fix: Make sure we don't dial with an incoming callDenis Kenzior1-0/+3
2009-11-13Refactor: Simplify code and add have_incomingDenis Kenzior1-12/+12
2009-11-13Revert: Go back to the old dial behaviorDenis Kenzior1-64/+4
The driver has a better idea of how to handle ATD being invoked when a waiting/held/active call(s) is/are already present. Generally most modems will swap the call automatically preserving the state of the waiting call. We do block dialing another call while a dial is still in progress.
2009-11-12Use release_all_held to release a single held callDenis Kenzior1-0/+8
2009-11-12Remove Voicecall.Busy methodDenis Kenzior1-27/+14
According to 22.030, UDUB or CHLD=0 can only be invoked on waiting calls. Most AT command based modems do not support using CHLD=0 on an incoming call. So we remove the Busy method and invoke set_udub on a call that is in the waiting state.
2009-11-05Use localtime_r instead of localtimeDenis Kenzior1-1/+2
2009-10-30Fix: Refine the logic for CHLD=1 and CHLD=2Denis Kenzior1-4/+8
2009-10-29Fix: Refine logic for hangupDenis Kenzior1-4/+6
2009-10-28In the simple single active call case, use hangupDenis Kenzior1-6/+23
2009-10-28Fix: Take care of the logic for DialingDenis Kenzior1-2/+65
The logic for dialing a call while other calls are in the system is a bit more complicated than what was implemented. There are several cases to consider: - Active Call and Held Call In this situation we cannot dial, the user must drop one of the calls. - Active Call and Waiting Call -> We have to put the active call on hold, unfortunately there is no standard command to do so without affecting the waiting call. If the modem has a vendor specific implementation, then try to use that. - Active Call only -> Put the active call on hold. We use the swap_without_accept implementation (if exists) to be safe, otherwise we fall back to hold active calls. - No Active calls -> Simply dial
2009-10-28Refactor: Avoid unnecessary forward declarationsDenis Kenzior1-45/+44
2009-10-28Refactor: Move more code aroundDenis Kenzior1-69/+67
Avoid unnecessary forward declarations
2009-10-28Refactor: Move code to avoid forward declarationDenis Kenzior1-171/+170
2009-10-28Refactor: Get rid of PENDING_FLAG in voicecallsDenis Kenzior1-128/+44
the pending variable was enough already. Also refactor to use __ofono_dbus_pending_reply to make the code much more readable.
2009-10-28Add utility to parse CLIR optionDenis Kenzior1-7/+17
2009-10-27Fix: Skip empty emergency numbersDenis Kenzior1-1/+3
2009-10-21Migrate from RFC2822 to ISO8601Olivier Le Thanh Duong1-1/+1
2009-10-08Rename function to express its purpose betterDenis Kenzior1-2/+2
2009-10-08Allow sending DTMF tones in alerting statePekka Pessi1-1/+5
2nd stage dialing may be done before call gets connected.
2009-10-08Use swap_without_accept method, if presentPekka Pessi1-0/+20
2009-10-02Refactor: Move elementary file type checkingDenis Kenzior1-5/+4
Every single EF read callback checks the file type reported out of the SIM is what it expects. Instead this should be done in one place and the errors reported accordingly
2009-09-04Some modems don't return from ATD until connectedDenis Kenzior1-1/+2
Yet they send progress indications all the way up to connected, which breaks the current logic.
2009-09-04Don't emit disconnect reason if unknownDenis Kenzior1-1/+3
2009-09-04For incoming call case, send hangupDenis Kenzior1-2/+5
CHLD=1X doesn't work for incoming calls on some devices. This is more portable
2009-09-01change int vendor to unsigned int vendorDenis Kenzior1-1/+1
2009-09-01Add vendor flag capability to voicecallsDenis Kenzior1-4/+4
2009-08-20Remove unused functionDenis Kenzior1-10/+0
2009-08-20Fix implicit declaration of extract_bcd_numberDenis Kenzior1-0/+1
2009-08-20Fix unused variable pathDenis Kenzior1-1/+1
2009-08-20Fix unused variable lDenis Kenzior1-1/+0
2009-08-19Emit the DisconnectReason signalDenis Kenzior1-1/+28
2009-08-19Make sure call history handles waiting callsDenis Kenzior1-1/+2
Missed waiting calls were not being handled properly
2009-08-19Evolve the voicecall driverDenis Kenzior1-437/+458
2009-08-19Update voicecalls to the new SIM APIDenis Kenzior1-12/+29
2009-08-19Refactor ECC: Free the new_en_listDenis Kenzior1-0/+6
Avoid a potential memory leak in case the interface is removed during the SIM read operation.
2009-08-17Refactor ECC Support: Update when EFecc is readDenis Kenzior1-23/+61
We should start with a default set of ECC numbers which are always available according to 22.101, even if no SIM present. Then when EFecc is read, we update the EmergencyNumbers property. According to 22.101 the 112 and 911 numbers must always be available. Since these numbers can also be stored on the SIM itself, we must avoid duplication.
2009-08-17Refactor ECC Support: Don't bother with g_sourceDenis Kenzior1-24/+1
Don't bother playing the g_source tricks for EmergencyNumbers property changed signal emission. The list is not bound to change more than once.
2009-08-17Refactor ECC Support: Free in right spotDenis Kenzior1-7/+6
The freeing of the emergency list should be in voicecalls_destroy function, where the rest of the datastructure is being freed
2009-08-17Support emergency numbersYang Gu1-1/+135
2009-08-14Refactor cssn into ofono_ssnDenis Kenzior1-1/+0
Make this into a fully fledged entity, with a driver instead of the current kludge. This means modem drivers can actually choose whether to instantiate a CSSN atom or not Move the notification functions from voicecall.c into ssn.c. Move the cssn.h header into include/ssn.h and refactor Update call barring to utilize ofono_ssn and use the new atom_watch functionality to detect when ssn has been added or removed
2009-08-14Remove trailing whitespaceYang Gu1-3/+3
2009-08-07Use g_source instead of flag for mpty list changesDenis Kenzior1-6/+11