| 00:19.41 | Entomologist | *** CLOSED (16643) [Channels/chan_sip/General] [patch] Setting "timerb" on chan_sip.conf doesn't work at all, in [general] or peer |
| 00:19.42 | Entomologist | Reported by: nahuelgreco Karma: 13 |
| 00:19.42 | Entomologist | https://issues.asterisk.org/view.php?id=16643 |
| 00:19.43 | Entomologist | ********************************************************* |
| 04:59.39 | *** join/#asterisk-bugs Deeewayne (~dwayne@c-71-207-214-190.hsd1.al.comcast.net) |
| 04:59.40 | *** mode/#asterisk-bugs [+o Deeewayne] by ChanServ |
| 06:23.29 | *** join/#asterisk-bugs oej (~olle@ns.webway.se) |
| 07:41.07 | *** join/#asterisk-bugs mpe (~mpe@217.74.220.92) |
| 08:28.09 | *** join/#asterisk-bugs oej (~olle@static-213-115-251-100.sme.bredbandsbolaget.se) |
| 08:36.53 | *** join/#asterisk-bugs Antonjs (~anton@mm-headoffice.clients.fishnet.net.nz) |
| 10:52.15 | *** join/#asterisk-bugs |Cybex| (~John@atwork-21.r-212.178.82.atwork.nl) |
| 10:53.22 | *** join/#asterisk-bugs |Cybex| (~John@atwork-21.r-212.178.82.atwork.nl) |
| 14:54.54 | *** join/#asterisk-bugs putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 14:54.55 | *** mode/#asterisk-bugs [+o putnopvut] by ChanServ |
| 15:04.35 | *** join/#asterisk-bugs leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 15:04.35 | *** mode/#asterisk-bugs [+o leifmadsen] by ChanServ |
| 15:52.20 | *** join/#asterisk-bugs Deeewayne (~dwayne@75.76.254.162) |
| 15:52.20 | *** mode/#asterisk-bugs [+o Deeewayne] by ChanServ |
| 16:00.22 | *** join/#asterisk-bugs jpeeler (~jpeeler@asterisk/digium-software-dev/jpeeler) |
| 17:12.34 | *** join/#asterisk-bugs oej (~olle@ns.webway.se) |
| 17:36.18 | Entomologist | *** CLOSED (16697) [Channels/chan_iax2] IAX always attempts authentication against first (alphabetically) user |
| 17:36.19 | Entomologist | Assigned to: dvossel |
| 17:36.19 | Entomologist | Reported by: clint |
| 17:36.20 | Entomologist | https://issues.asterisk.org/view.php?id=16697 |
| 17:36.20 | Entomologist | ********************************************************* |
| 18:00.31 | bzing2 | anyone have a good handle of how asterisk channel structures work, specifically the locking? |
| 18:00.38 | russellb | yes. |
| 18:01.01 | bzing2 | I have found (and can repeat) a deadlock in chan_local |
| 18:01.09 | russellb | version? |
| 18:01.16 | bzing2 | 1.6.2.2 |
| 18:02.36 | russellb | Okay, compile with DEBUG_THREADS, and grab "core show locks" output when the deadlock occurs |
| 18:02.59 | bzing2 | yup, got that, and a backtrace from a core dump, submitted with the bug report |
| 18:03.05 | russellb | ah, perfect |
| 18:03.17 | russellb | now for someone to get to it ... it's in a pile with 600 other open tickets ... |
| 18:03.34 | bzing2 | yeh, thats always the problem |
| 18:03.56 | bzing2 | I "have" to have a stab at it myself, its effecting somthing thats opperational |
| 18:04.15 | bzing2 | what I am concerned about is what the right way to lock things is |
| 18:04.20 | russellb | carefully :-) |
| 18:04.33 | bzing2 | lol .... and in the same order |
| 18:04.41 | bzing2 | the latter being the problem |
| 18:05.07 | russellb | there is a section on locking in doc/CODING-GUIDELINES |
| 18:05.13 | russellb | and discusses channel locking and such |
| 18:05.16 | russellb | take a look at that first |
| 18:05.34 | bzing2 | ah, yes, that would seem a good place to start |
| 18:05.42 | russellb | chan_local is on of, if not the most complicated piece of code that deals with this problem, unfortunately |
| 18:05.46 | russellb | s/on of/one of/ |
| 18:06.20 | bzing2 | yes, guessed that due to it needing to optinize itself out the loop |
| 18:06.25 | bzing2 | ... that being the problem |
| 18:15.39 | bzing2 | ok all good general advice on locking |
| 18:16.19 | bzing2 | but should one hold a lock before calling ast_bridged_channel? |
| 18:16.29 | russellb | should, yes |
| 18:16.34 | russellb | welllllllllllll |
| 18:16.36 | russellb | it's complicated. |
| 18:16.53 | bzing2 | which was the answer I was expecting |
| 18:17.06 | russellb | technically you should, but if you're in the channel thread that owns all channels involved, it's not going to change on you |
| 18:17.25 | bzing2 | ah ha |
| 18:17.28 | russellb | so in that case, you don't really _have_ to |
| 18:18.03 | bzing2 | well in this case there isn't a lock |
| 18:18.12 | russellb | ah |
| 18:18.27 | bzing2 | but locking it would only move the deadlock |
| 18:18.51 | bzing2 | which may not be a bad thing, if one can then do somthing to avoid it |
| 18:20.35 | bzing2 | problem is around p->chan->_bridge, in theory one should lock chan first |
| 18:22.42 | russellb | indeed |
| 18:22.52 | russellb | and you must do so if you're not in the channel thread for p->chan |
| 18:23.03 | russellb | for example, if you're in code because of a callback on p->other or whatever it's called |
| 18:24.45 | *** join/#asterisk-bugs putnopvut (~putnopvut@asterisk/master-of-queues/mmichelson) |
| 18:24.45 | *** mode/#asterisk-bugs [+o putnopvut] by ChanServ |
| 18:24.56 | bzing2 | ok, one assumes your the thread is you see somthing like pbx_thread (data=0x1b5ff740) |
| 18:27.03 | russellb | right |
| 18:27.54 | bzing2 | now perhaps a stupid question ... two ast_channel strutures point to the save local_pvt structure |
| 18:28.45 | bzing2 | at some level this hurts my head, are they the same channel? Or it this the wierdness that is the local channel? |
| 18:29.45 | Entomologist | *** CLOSED (16809) [Tests/NewFeature] [patch] Design functionality test for dialplan pattern matching |
| 18:29.46 | Entomologist | Assigned to: mmichelson |
| 18:29.46 | Entomologist | Reported by: lmadsen Karma: 39.25 |
| 18:29.47 | Entomologist | https://issues.asterisk.org/view.php?id=16809 |
| 18:29.47 | Entomologist | ********************************************************* |
| 18:30.05 | bzing2 | appologieses if he is thinking out loud |
| 18:33.14 | russellb | they are not the same channel |
| 18:33.23 | russellb | they are different, and they owned by 2 different channel threads |
| 18:33.28 | russellb | _that_ is why it gets complicated |
| 18:33.50 | russellb | This is where I usually break out the whiteboard and draw some diagrams ... |
| 18:34.02 | bzing2 | I have already filled up two |
| 18:35.07 | seanbright | so i have a patch to change all of that to multiplexed i/o |
| 18:35.19 | seanbright | so asterisk will only ever have 2 threads |
| 18:35.20 | seanbright | it's awesome |
| 18:35.24 | bzing2 | ok, one assumes one is the channel created by Dial(Local/thing) and the other is magied into being as a result |
| 18:35.29 | seanbright | (i'm kidding) |
| 18:35.34 | bzing2 | lol |
| 18:35.47 | bzing2 | it would be one mother of a patch! |
| 18:35.55 | seanbright | meh |
| 18:36.11 | seanbright | #define pthread_thread_create (void) 0; |
| 18:36.18 | seanbright | piece of cake |
| 18:36.31 | russellb | Let me try to explain the channels in chan_local, and if I fail, we can just get on the phone *gasp* |
| 18:37.09 | russellb | Scenario: incoming SIP call (SIP/123), executes Dial(Local/456@default) ... 456@default does a Dial(SIP/456) |
| 18:37.15 | bzing2 | tell you what, unfortunatly I have to run to a meeting, but could I take some of your time (planned) to talk about this? |
| 18:38.18 | russellb | sure |
| 18:38.41 | russellb | maybe I should just document what I was going to say ... |
| 18:38.52 | bzing2 | that may be a plan, its hurting my head |
| 18:38.56 | russellb | heh |
| 18:39.04 | bzing2 | I will IM you my email |
| 18:39.06 | Qwell | russellb: /* \brief MAGIC */ |
| 18:39.12 | Corydon76-dig | hires a camera crew to follow Russell around and record everything he says |
| 18:39.21 | russellb | eep |
| 18:42.31 | Deeewayne | sits on the edge of his seat waiting for the release of Corydon76-dig's movie about what Russell was going to say |
| 18:44.46 | putnopvut | All russell talks about any more are indigenous popcorn species of the Andes. |
| 18:44.58 | putnopvut | It gets *really* old |
| 18:45.04 | seanbright | popcorn species? |
| 18:45.15 | seanbright | shakes his head |
| 18:45.17 | putnopvut | Varieties? |
| 18:45.21 | putnopvut | I don't know. |
| 18:45.21 | seanbright | popcorn is corn |
| 18:45.34 | putnopvut | Yeah, but there are like 3 million different types of it. |
| 18:45.44 | Qwell | seanbright: sodacorn |
| 18:45.45 | Qwell | err, wait no |
| 18:45.57 | seanbright | whatever |
| 18:46.00 | seanbright | you had me at hello |
| 18:46.05 | Deeewayne | popped is the corn the corn that makes my kids happiest |
| 18:46.05 | Qwell | I never said hello |
| 18:46.17 | Deeewayne | hrm... the corn the corn |
| 18:46.47 | seanbright | Qwell: hi |
| 18:48.05 | leifmadsen | Deeewayne: ping! |
| 18:48.14 | Deeewayne | leifmadsen, yo! |
| 18:48.23 | leifmadsen | Deeewayne: did I miss a ping from you the other day? |
| 18:48.28 | Deeewayne | yes :-( |
| 18:48.41 | leifmadsen | it wasn't my fault! It was Qwell! |
| 18:49.02 | Qwell | saw that coming |
| 18:49.28 | Deeewayne | leifmadsen, I was just going to say thanks for being you |
| 18:49.34 | leifmadsen | awww :) |
| 18:49.47 | leifmadsen | Qwell: you saw it coming because you were guilty! |
| 18:50.02 | Deeewayne | but that wasn't intended for Valentine's Day weekend |
| 18:50.57 | leifmadsen | ;) |
| 19:20.47 | seanbright | i say we tell bklang that asterisk is not supported on opensolaris |
| 19:45.35 | Entomologist | *** CLOSED (16316) [Channels/chan_sip/DatabaseSupport] [regression] After upgrading i see a lot more Notices peers status |
| 19:45.35 | Entomologist | Reported by: barthpbx |
| 19:45.36 | Entomologist | https://issues.asterisk.org/view.php?id=16316 |
| 19:45.36 | Entomologist | ********************************************************* |
| 19:57.55 | seanbright | Corydon76-dig: ping |
| 19:58.48 | seanbright | i think you can just use AC_TYPE_UINT64_T for 16830 instead of a hand-rolled autoconf test |
| 20:16.51 | Corydon76-dig | seanbright: I don't think even that is necessary |
| 20:34.43 | *** join/#asterisk-bugs davidbrillert (~aragon125@h216-235-8-130.host.egate.net) |
| 20:54.50 | Entomologist | *** CLOSED (16673) [Core/Portability] [patch] TLS socket file descriptor fails to open (with no error message in log) |
| 20:54.51 | Entomologist | Assigned to: dvossel |
| 20:54.51 | Entomologist | Reported by: michaesc Karma: 0 |
| 20:54.52 | Entomologist | https://issues.asterisk.org/view.php?id=16673 |
| 20:54.52 | Entomologist | ********************************************************* |
| 23:13.32 | seanbright | Corydon76-dig: ok. just pointing out there was a built-in test for types like that. |
| 23:14.32 | Corydon76-dig | Ah, couldn't find it in the autoconf manual |
| 23:15.23 | seanbright | for ref: http://www.gnu.org/software/hello/manual/autoconf/Particular-Types.html |
| 23:55.25 | *** join/#asterisk-bugs corretico (~laguilar@201.201.46.106) |