| 00:02.05 | *** join/#asterisk-dev w9sh (~sph@50-79-224-193-static.hfc.comcastbusiness.net) |
| 00:04.34 | mjordan | file: ASTERISK-22936 |
| 00:04.52 | mjordan | if we don't synchronously push the fix up, then can we end up in the following: |
| 00:05.26 | mjordan | nevermind. |
| 00:05.29 | mjordan | Thought about it some more. |
| 00:05.30 | mjordan | HM. |
| 00:07.25 | file | mjordan, HM = hiptastic medicine! |
| 00:07.38 | mjordan | trying to think about the implications of that patch. |
| 00:07.57 | mjordan | basically, the fixup operation won't complete synchronously. That means we'll return to the masquerade routine before the fixup has finished. |
| 00:08.11 | mjordan | We unlock the channel. Something else locks it, and has a channel that is not yet completely "fixed". |
| 00:08.44 | mjordan | Basically, anything that pushes a SIP task - if it uses the Asterisk channel object - has to double check it |
| 00:11.08 | mjordan | the question is: can you requeue the task if the channel is a zombie? |
| 00:11.13 | mjordan | or do you have to abandon it? |
| 00:11.45 | mjordan | cus the session isn't necessarily dead, just the original channel. Although in the outline of the issue, the PJSIP channel is effectively toast |
| 00:12.13 | mjordan | and actually, with how masquerades survive in 12, they have to be effectively toast. We don't have any situations where the session survives on a new channel |
| 00:12.52 | file | I think you are wrong in that. |
| 00:13.15 | file | an async_goto can do it |
| 00:13.25 | mjordan | async_goto doesn't happen anymore, IIRC |
| 00:14.06 | mjordan | hm. |
| 00:14.18 | mjordan | yeah, we can grab a channel out of an app, which does that. |
| 00:14.20 | mjordan | Oh well. |
| 00:14.26 | mjordan | blech. |
| 00:15.25 | mjordan | I'm curious, why does the chan_pjsip fixup callback have to happen on a PJSIP thread? |
| 00:15.46 | file | because session is not protected at all |
| 00:15.49 | mjordan | ah. |
| 00:18.26 | mjordan | so we either protect session, or we update all of the taskprocessor pushes to check whether or not the channel is a masqueraded channel and then.... what? |
| 00:18.39 | mjordan | I'm not sure the pushed tasks can survive until the masqueraded task is handled. |
| 00:18.48 | mjordan | Is there a way to push a task to the front of the queue? |
| 00:19.06 | file | not presently, and that does not guarantee that something is not already executing |
| 00:19.35 | mjordan | hm. |
| 00:19.49 | mjordan | what a mess. |
| 00:19.56 | mjordan | I don't think his patch will work however. |
| 00:20.04 | file | it's the boundary between Asterisk old and PJSIP. |
| 00:22.54 | mjordan | yup. But you can't delay a masquerade. |
| 04:49.53 | *** join/#asterisk-dev mjordan (~matt@nat/digium/x-trjwkfikxkweqnuz) |
| 04:49.53 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 05:36.02 | *** join/#asterisk-dev mjordan (~matt@75.76.55.191) |
| 05:36.02 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 10:16.09 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 10:16.09 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 13:15.03 | *** join/#asterisk-dev hehol (~Adium@2a01:198:71d:0:91:51ed:4796:65c) |
| 13:19.12 | *** join/#asterisk-dev hehol (~Adium@2a01:198:71d:0:91:51ed:4796:65c) |
| 13:54.20 | *** join/#asterisk-dev hehol (~Adium@ip-78-94-76-253.unitymediagroup.de) |
| 14:34.41 | *** join/#asterisk-dev dar123 (~IceChat77@77.30.127.192) |
| 14:57.45 | *** join/#asterisk-dev mjordan (~matt@75.76.55.191) |
| 14:57.45 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 15:25.00 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@64.235.216.2) |
| 15:55.31 | *** join/#asterisk-dev dar123 (~IceChat77@77.30.127.192) |
| 15:59.48 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@64.235.216.2) |
| 16:09.01 | *** part/#asterisk-dev mjordan (~matt@75.76.55.191) |
| 16:23.51 | *** join/#asterisk-dev leifmadsen (~Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 16:23.51 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 17:05.18 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@64.235.216.2) |
| 17:16.58 | *** join/#asterisk-dev serafie (~erin@24.96.64.240) |
| 17:26.08 | *** join/#asterisk-dev mjordan (~matt@nat/digium/x-lazsiigoduqzrfmt) |
| 17:26.09 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 17:36.28 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@64.235.216.2) |
| 18:52.37 | *** join/#asterisk-dev dar123 (~IceChat77@77.30.127.192) |
| 18:58.05 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 18:58.05 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 19:33.34 | *** join/#asterisk-dev danjenkins (~danjenkin@cpc3-folk2-2-0-cust34.1-2.cable.virginm.net) |
| 20:30.21 | *** join/#asterisk-dev serafie (~erin@24.96.64.240) |
| 20:46.32 | *** join/#asterisk-dev vlad_starkov (~vlad_star@109.188.124.76) |
| 21:32.26 | *** join/#asterisk-dev otherwiseguy (~otherwise@2605:a601:437:ce01:3e97:eff:fe14:ef66) |
| 22:33.15 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-28-58.cablep.bezeqint.net) |
| 22:33.16 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 22:35.59 | *** part/#asterisk-dev mjordan (~matt@nat/digium/x-lazsiigoduqzrfmt) |
| 23:11.02 | *** join/#asterisk-dev snuff-work (~snuffy@210.8.167.13) |
| 23:11.03 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |