Hello all, After a long time, I upgraded the pjsip stack from 0.9x to 1.0x (latest svn as of yesterday) and I am getting some strange crashes occasionally. I'll have to use the debug build for a while, to get the details of the crash but for now it seems as though 'pjmedia_snd_stream *strm' is not proper (0x0fa). My app is pjsua_lib based, and the crash happens inside pjsua_set_ec (SIPLocalEcho,0) which is called within the on_call_media_state callback I'll add the important bits of my code to make things more understandable ... void SIPCaller::on_call_media_state(pjsua_call_id call_id) { CCErr=pjsua_call_get_info(call_id, &CMSCallInfo); if (CMSCallInfo.media_status == PJSUA_CALL_MEDIA_ACTIVE) { CCErr=pjsua_conf_connect(CMSCallInfo.conf_slot, 0); CCErr=pjsua_conf_connect(0, CMSCallInfo.conf_slot); pjsua_set_ec (SIPLocalEcho,0); } } the crash happens at (on the if statement): if (strm->play_strm) { paPlaySI = Pa_GetStreamInfo(strm->play_strm); } inside the pjmedia_snd_stream_get_info (pasound.c). ... any ideas? Greets, Toni

Don't know if it is related but I am getting intermediate crashes on pjaus_conf_connect()...with 1.0 rc1... I am currently trying to dig in to see what might be going on. What is strange is the call otherwise seems to work. I would expect something to be definitely broken after a access violation. David Clark At 09:33 AM 2/9/2009, buldozer@aufbix.org wrote: > Hello all, > > > After a long time, I upgraded the pjsip stack from 0.9x to 1.0x > (latest svn as >of yesterday) and I am getting some strange crashes occasionally. I'll have to >use the debug build for a while, to get the details of the crash but >for now it >seems as though 'pjmedia_snd_stream *strm' is not proper (0x0fa). > > My app is pjsua_lib based, and the crash happens inside pjsua_set_ec >(SIPLocalEcho,0) which is called within the on_call_media_state callback > >I'll add the important bits of my code to make things more understandable ... > >void SIPCaller::on_call_media_state(pjsua_call_id call_id) { > > CCErr=pjsua_call_get_info(call_id, &CMSCallInfo); > if (CMSCallInfo.media_status == PJSUA_CALL_MEDIA_ACTIVE) { > CCErr=pjsua_conf_connect(CMSCallInfo.conf_slot, 0); > CCErr=pjsua_conf_connect(0, CMSCallInfo.conf_slot); > pjsua_set_ec (SIPLocalEcho,0); > } > >} > >the crash happens at (on the if statement): > if (strm->play_strm) { > paPlaySI = Pa_GetStreamInfo(strm->play_strm); > } > >inside the pjmedia_snd_stream_get_info (pasound.c). > > > ... any ideas? > > > Greets, Toni > > >_______________________________________________ >Visit our blog: http://blog.pjsip.org > >pjsip mailing list >pjsip@lists.pjsip.org >http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org

If supported any special configuration required to enable it? Thanks, David Clark

You can create your own SDP offer/answer for T.38 if you use pjsip-ua API. regards, Gang On Tue, Feb 10, 2009 at 7:28 AM, David Clark <vdc1048@tx.rr.com> wrote: > If supported any special configuration required to enable it? > > Thanks, > > David Clark > > > _______________________________________________ > Visit our blog: http://blog.pjsip.org > > pjsip mailing list > pjsip@lists.pjsip.org > http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org >

Heelo all, Finally found the reason for crash - there was no sound card in the system (actually there was, but it slipped out of the slot a bit). Maybe it would be good, if pjmedia checked the sound devices, before trying to set the EC on nonexistent streams. Greets, Toni

Hi, Sorry, I think I couldn't get it. When there is no sound card, pjsua_var.snd_port should be NULL (it fails on opening sound device), and I think NULL conditions are always checked at the beginning of functions. Regards, nanang On Tue, Feb 10, 2009 at 9:22 PM, <buldozer@aufbix.org> wrote: > Heelo all, > > > Finally found the reason for crash - there was no sound card in the system > (actually there was, but it slipped out of the slot a bit). Maybe it would > be > good, if pjmedia checked the sound devices, before trying to set the EC on > nonexistent streams. > > > Greets, Toni > > > > _______________________________________________ > Visit our blog: http://blog.pjsip.org > > pjsip mailing list > pjsip@lists.pjsip.org > http://lists.pjsip.org/mailman/listinfo/pjsip_lists.pjsip.org >

Hi Nanang, Yes - it gets checked in the Debug builds (ASSERT), but not in the Release builds, where the asserts are skipped (checked the disassembly). Greets, Toni

On Wed, Feb 11, 2009 at 3:18 PM, <buldozer@aufbix.org> wrote: > Hi Nanang, > > > Yes - it gets checked in the Debug builds (ASSERT), but not in the Release > builds, where the asserts are skipped (checked the disassembly). > > It should still check it in Release build, unless you disable it by declaring PJ_ENABLE_EXTRA_CHECK to zero ( http://www.pjsip.org/pjlib/docs/html/group__pj__assert.htm) Cheers Benny > > > Greets, Toni > >