| 00:06.11 | *** join/#asterisk-dev [TK]D-Fender (~joe@64.235.216.2) |
| 02:23.55 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 02:33.53 | *** join/#asterisk-dev ipengineer (~zconkle@108-206-120-218.lightspeed.mckntx.sbcglobal.net) |
| 03:11.03 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 03:23.45 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 05:52.38 | *** join/#asterisk-dev babak (uid19622@gateway/web/irccloud.com/x-nhidlehimmwodakt) |
| 07:26.33 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 07:27.25 | *** join/#asterisk-dev snuff-work (~blatus@203.87.61.22) |
| 07:27.25 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 08:01.24 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 08:01.24 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 08:07.28 | *** join/#asterisk-dev frenk77 (~frenk77@2a02:8304:2:3:8e89:a5ff:fe57:49be) |
| 08:15.56 | frenk77 | Hello, i want discuss my problem ASTERISK-25321, look last comments. Someone writes me its called by locks held, ok i see in channel.c locked by line ast_channel_lock_both(original, clonechan); then ast_autochan_new_channel(clonechan, original); is called. So this two channels is locked, is old_chan and new_chan in ast_autochan_new_channel function from autochan.c. But there is some list_traverse and another ast_autochan struct ca |
| 08:16.58 | frenk77 | and second question...is somewhere available history for this chat room ? |
| 08:50.03 | *** join/#asterisk-dev fornax (~fornax@85.183.43.221) |
| 09:31.50 | *** join/#asterisk-dev fornax (~fornax@85.183.43.221) |
| 09:40.27 | wdoekes | frenk77: I have history in my local irc client logs |
| 09:41.38 | wdoekes | as for your question. "someone" is @file in this channel. and the rest of your question is truncated at: But there is some list_traverse and another ast_autochan struct ca[truncated] |
| 09:51.14 | wdoekes | you can apparently reproduce the issue locally. if you ASSERT() that chan did not change on 572, you should get an abort() core dump in case it does: then you may be able to see which thread caused it |
| 09:51.27 | *** join/#asterisk-dev fornax (~fornax@85.183.43.221) |
| 09:51.47 | wdoekes | you can also add a bit of sleep between the lock and the unlock, so it reproduces faster |
| 09:55.41 | wdoekes | (possibly it also helps if you add some sleep after inserting the channel into the channels-list (after ao2_link(channels, tmp); in __ast_channel_alloc_ap)) |
| 09:59.25 | frenk77 | reproduce is not a problem, i now looking for changes autochan->chan without locking. So i think it is in ast_autochan_new_channel .... and my question is: it is correctly locked here or not? Sorry for my english i still learning and somtimes i not understand well :) |
| 10:03.51 | file | it seems to be correctly locked, yes |
| 10:04.11 | frenk77 | oh here is truncated part..... But there is some list_traverse and another ast_autochan struct called autochan, and there are some operation witch autochan->chan (lines autochan->chan = ast_channel_unref(old_chan); autochan->chan = ast_channel_ref(new_chan); ) is autochan->chan correctly locked for this? where? :) |
| 10:16.39 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 10:16.39 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 10:32.33 | file | it's protected by the lock of the channel itself, which is new_chan, and since that is locked in the masquerade operation it should be fine |
| 11:05.07 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 11:05.07 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 11:26.55 | frenk77 | ok, i understand so now this bug causes my brain deadlock too :) |
| 11:28.18 | frenk77 | it is most difficult bug ever for me :-( |
| 12:04.35 | wdoekes | frenk77: can you try my suggestions? |
| 12:08.07 | frenk77 | i am going to try them.... |
| 12:09.52 | wdoekes | recompile with ./configure --enable-dev-mode, and set DONT_OPTIMIZE, and add an assertion here: http://junk.devs.nu/frenk77-1.patch |
| 12:10.15 | wdoekes | (something like that) |
| 12:12.03 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@216-191-106-163.dedicated.allstream.net) |
| 12:36.25 | wdoekes | (oh, and DO_CRASH seems not to be enabled by default in menuconfig. you'll want that. or perhaps just replace the ast_assert() by an if(!(...)) { abort(); } ) |
| 12:43.55 | *** join/#asterisk-dev leedm777 (leedm777@nat/digium/x-tzptumeamkflaftl) |
| 12:44.27 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 12:44.27 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:45.15 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 12:45.15 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:45.55 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 12:45.55 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:46.03 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 12:46.03 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:46.25 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 12:46.26 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:46.36 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 12:46.36 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:47.17 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 12:47.17 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:47.29 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 12:47.29 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:48.09 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 12:48.10 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:52.24 | *** part/#asterisk-dev leedm777 (leedm777@nat/digium/x-tzptumeamkflaftl) |
| 13:09.15 | *** join/#asterisk-dev oej (~oej@195.67.25.237) |
| 13:09.15 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 13:42.39 | *** join/#asterisk-dev nix8n82 (~AndChat56@67-130-74-235.dia.static.qwest.net) |
| 13:43.51 | *** join/#asterisk-dev newtonr (RustyNewto@nat/digium/x-nchopvzmdruhndvs) |
| 13:43.51 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 13:45.22 | *** join/#asterisk-dev fornax (~fornax@p548FFD8D.dip0.t-ipconnect.de) |
| 14:07.42 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-rmrnrvzpyarauhha) |
| 14:17.26 | *** join/#asterisk-dev fornax (~fornax@p548FFD8D.dip0.t-ipconnect.de) |
| 14:17.51 | *** join/#asterisk-dev fornax (~fornax@p548FFD8D.dip0.t-ipconnect.de) |
| 14:42.09 | *** join/#asterisk-dev sgriepentrog (sgriepentr@nat/digium/x-qbzpykrlyodutjrc) |
| 14:49.04 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-eauygsifkammocqu) |
| 15:05.21 | stefan27 | it seems after upgrading 13.1 -> 13.5 some SIP channels are visible in "core show channels" long after calls are terminated, so that the output of core show channels can be 100 lines even though it says "active channels 4, active calls 2", is this a known bug? |
| 15:06.03 | file | any known bugs would be in JIRA, I know of none with that |
| 15:06.44 | stefan27 | OK, I'll investigate and make ticket tomorrow |
| 15:18.21 | *** join/#asterisk-dev fornax (~fornax@p548FFD8D.dip0.t-ipconnect.de) |
| 15:18.54 | stefan27 | is there any kind of debug-info I should attach for this type of issue, besides the obvious (sip.conf etc)? is the output of "sip show objects" useful? |
| 15:22.25 | stefan27 | I'll try find the weird use case which causes it |
| 15:28.09 | file | ummm |
| 15:28.33 | stefan27 | I guess if it's a genuine memory leak i'll have to recompile with those mem debug flags |
| 15:32.27 | file | it would be a stasis cache problem |
| 15:32.37 | file | so debug and information about the flow of the calls would be useful |
| 15:39.10 | stefan27 | can it be the case that the sip channels in questions are actually freed only that their names are stored in some cache which isnt cleaned? |
| 15:40.17 | file | it's not names it's snapshots of the channel but yes, it's possible that somehow they did not get removed from the cache |
| 15:40.20 | stefan27 | the AMI client which talks to asterisk seems to get the report that the channels no longer exist and I havent noticed worse performacne of anything so it's not an external visible bad behaviour yet |
| 15:41.14 | *** join/#asterisk-dev babak (uid19622@gateway/web/irccloud.com/x-ncpxvwxduhueuhts) |
| 15:41.25 | stefan27 | but Ill make an issue tomorrow, bbl |
| 15:45.22 | *** join/#asterisk-dev fornax (~fornax@p548FFD8D.dip0.t-ipconnect.de) |
| 15:48.57 | rmudgett | A channel ref leak will cause the channel to still show up in the "core show channels" output even though the channel has hung up. |
| 15:53.53 | file | ah yeah |
| 17:58.39 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 17:58.39 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 18:20.19 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 18:20.19 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 18:22.50 | *** join/#asterisk-dev mjordan_ (mjordan@nat/digium/x-zypprakrubeoihzm) |
| 18:22.50 | *** mode/#asterisk-dev [+o mjordan_] by ChanServ |
| 20:23.36 | *** join/#asterisk-dev gtjoseph (~gtj0@unaffiliated/gtj) |
| 20:36.29 | *** join/#asterisk-dev fornax (~fornax@85.183.43.221) |
| 21:07.59 | *** join/#asterisk-dev [TK]D-Fender (~joe@64.235.216.2) |
| 21:17.10 | *** join/#asterisk-dev Qwell (north@asterisk/developer/Qwell) |
| 21:17.10 | *** mode/#asterisk-dev [+o Qwell] by ChanServ |
| 22:04.47 | seanbright | lots of bug fixes going through gerrit |
| 22:05.28 | seanbright | there's a simple fix for all of that |
| 22:05.32 | seanbright | stop introducing bugs |
| 22:05.39 | seanbright | that'll be $10,000 |
| 22:05.48 | *** kick/#asterisk-dev [seanbright!~file@asterisk/developer-and-muffin-lover/file] by file (seanbright) |
| 22:05.48 | *** join/#asterisk-dev seanbright (~sean@asterisk/contributor-and-bug-marshal/seanbright) |
| 22:05.48 | *** mode/#asterisk-dev [+o seanbright] by ChanServ |
| 22:05.57 | seanbright | YOU'RE WELCOME |
| 22:06.24 | file | hi |
| 22:06.34 | seanbright | o/ |
| 22:07.57 | file | gives seanbright a Dairy Queen Blizzard instead |
| 22:13.36 | seanbright | ugh. i just ate like a pound of peanut m&ms |
| 22:13.41 | seanbright | <-- fatty |
| 22:14.49 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 22:23.23 | file | seanbright, high in protein?!? |
| 22:23.41 | seanbright | and sugarRRRRRrR |
| 22:24.42 | file | the receptionist at my office downtown is on a reduced sugar kick which is mildly amusing because there's been birthdays and she can't bring herself to say no to the offer of cake |
| 22:25.27 | file | sugarrrr |
| 22:40.02 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-rmrnrvzpyarauhha) |
| 23:08.51 | *** join/#asterisk-dev fornax (~fornax@85.183.43.221) |
| 23:47.56 | *** join/#asterisk-dev spditner (~simon@69-196-161-148.dsl.teksavvy.com) |