Multiple calls with PortAudio
This has occurred with a number of recent revisions. If anyone can reproduce it or suggest debugging steps, I'll gladly supply more information. Jira isn't convenient for me due to X issues at the moment, unless there's a way to interact with it other than via a Javascript-capable Web browser. Distribution: Debian Sid, now kernel 2.6.29 (same experience with 2.6.26). FreeSwitch revision 12701. Steps to reproduce Make a call with portAudio to a SIP endpoint (I haven't tried it with other endpoints): pa call <destination> Next, make another call: pa call <other_destination> Result: the party to the first call hears no music on hold, or possibly broken/intermittent music on hold, and the second call fails to go through. Where it hangs seems to vary, and sometimes it does in fact work. Often it hangs before it reaches the calling state, suggesting that there's some kind of locking issue involved. I'll perform further testing. Meanwhile, if anyone else can reproduce it, this may help. Note that the second call often doesn't return an error; it just hangs somewhere in the set-up process, e.g., init.
While I was trying to obtain more detailed logs of my portaudio problems, FreeSWITCH crashed, leaving a core file. The backtraces are here: http://pastebin.freeswitch.org/7998 As far as I can remember, at the time of the segfault, one channel was trying to connect and not succeeding; I had just issued a pa hangup command on it and then a pa call to try connecting again. Since my memory of exactly what was happening isn't as reliable as it should be, the value of the backtraces may be diminished. As to the portaudio problem, with rev. 12701 (Debian Sid, kernel 2.6.29, x86_64 architecture), the situation appears to be that the second and subsequent concurrent portaudio calls sometimes wait for a long time after issuing a log message such as the following: [NOTICE] switch_channel.c:597 switch_channel_set_name() New Channel sofia/internal/1000@192.168.0.2:5070 [d6e56642-1a9b-11de-b23e-c5a9450df57d] These calls do not always complete successfully, but I'm still trying to collect more precise details of when and why they fail. With apologies for being unable to use Jira, if anything valuable appears in the backtraces, you are welcome to let me know via the list or by e-mail. I have previously seen crashes while working with multiple portaudio calls, but I don't yet have a reliable means to reproduce them. If the backtraces are revealing, that's good, but if not, that's fine too and I'll collect better particulars next time.
Please direct the report to http://jira.freeswitch.org /b On Mar 27, 2009, at 2:58 AM, Jason White wrote:
While I was trying to obtain more detailed logs of my portaudio problems, FreeSWITCH crashed, leaving a core file.
The backtraces are here: http://pastebin.freeswitch.org/7998
Brian West brian@freeswitch.org -- Meet us a ClueCon! http://www.cluecon.com
Are you updating with "make current" each time? On Fri, Mar 27, 2009 at 2:58 AM, Jason White <jason@jasonjgw.net> wrote:
While I was trying to obtain more detailed logs of my portaudio problems, FreeSWITCH crashed, leaving a core file.
The backtraces are here: http://pastebin.freeswitch.org/7998
As far as I can remember, at the time of the segfault, one channel was trying to connect and not succeeding; I had just issued a pa hangup command on it and then a pa call to try connecting again. Since my memory of exactly what was happening isn't as reliable as it should be, the value of the backtraces may be diminished.
As to the portaudio problem, with rev. 12701 (Debian Sid, kernel 2.6.29, x86_64 architecture), the situation appears to be that the second and subsequent concurrent portaudio calls sometimes wait for a long time after issuing a log message such as the following:
[NOTICE] switch_channel.c:597 switch_channel_set_name() New Channel sofia/internal/1000@192.168.0.2:5070[d6e56642-1a9b-11de-b23e-c5a9450df57d]
These calls do not always complete successfully, but I'm still trying to collect more precise details of when and why they fail.
With apologies for being unable to use Jira, if anything valuable appears in the backtraces, you are welcome to let me know via the list or by e-mail.
I have previously seen crashes while working with multiple portaudio calls, but I don't yet have a reliable means to reproduce them.
If the backtraces are revealing, that's good, but if not, that's fine too and I'll collect better particulars next time.
_______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Anthony Minessale II FreeSWITCH http://www.freeswitch.org/ ClueCon http://www.cluecon.com/ AIM: anthm MSN:anthony_minessale@hotmail.com <MSN%3Aanthony_minessale@hotmail.com> GTALK/JABBER/PAYPAL:anthony.minessale@gmail.com<PAYPAL%3Aanthony.minessale@gmail.com> IRC: irc.freenode.net #freeswitch FreeSWITCH Developer Conference sip:888@conference.freeswitch.org <sip%3A888@conference.freeswitch.org> iax:guest@conference.freeswitch.org/888 googletalk:conf+888@conference.freeswitch.org<googletalk%3Aconf%2B888@conference.freeswitch.org> pstn:213-799-1400
participants (3)
-
Anthony Minessale
-
Brian West
-
Jason White