| 01:24.54 | *** join/#asterisk-bugs seanbright-home (n=seanbrig@c-69-251-175-43.hsd1.md.comcast.net) |
| 01:38.22 | *** join/#asterisk-bugs elguero (n=elguero@ns1.nashuacs.com) |
| 01:38.40 | *** part/#asterisk-bugs jmls_work (n=jmls@mail.tessera.co.uk) |
| 03:19.34 | *** join/#asterisk-bugs russellb (n=russell@asterisk/developer-and-stable-maintainer/drumkilla) |
| 03:19.34 | *** mode/#asterisk-bugs [+o russellb] by ChanServ |
| 03:24.27 | *** join/#asterisk-bugs d3wayne (n=deeewayn@76.29.245.9) |
| 03:32.49 | *** join/#asterisk-bugs angler (n=angler@pdpc/sponsor/digium/angler) |
| 05:22.25 | *** join/#asterisk-bugs d3wayne (n=deeewayn@76.29.245.9) |
| 05:29.17 | *** join/#asterisk-bugs codefreeze (n=steve_mu@216.166.159.235) |
| 05:29.17 | *** mode/#asterisk-bugs [+o codefreeze] by irc.freenode.net |
| 05:30.36 | *** join/#asterisk-bugs Deeewayne (n=dwayne@216.207.245.1) [NETSPLIT VICTIM] |
| 05:30.36 | *** join/#asterisk-bugs Juggie (i=Juggie@CPE001601df17fb-CM001a6682d7b6.cpe.net.cable.rogers.com) [NETSPLIT VICTIM] |
| 05:30.36 | *** join/#asterisk-bugs Qwell (i=north@pdpc/sponsor/digium/Qwell) [NETSPLIT VICTIM] |
| 05:30.36 | *** mode/#asterisk-bugs [+ooo Deeewayne Juggie Qwell] by irc.freenode.net |
| 07:03.42 | *** join/#asterisk-bugs angler (n=angler@pdpc/sponsor/digium/angler) |
| 07:50.29 | *** join/#asterisk-bugs agx (n=AGX@host63-216-static.34-88-b.business.telecomitalia.it) |
| 08:21.07 | *** join/#asterisk-bugs angler (n=angler@pdpc/sponsor/digium/angler) |
| 08:21.07 | *** join/#asterisk-bugs Deeewayne (n=dwayne@216.207.245.1) |
| 08:21.07 | *** mode/#asterisk-bugs [+o Deeewayne] by ChanServ |
| 10:41.28 | *** join/#asterisk-bugs angler (n=angler@pdpc/sponsor/digium/angler) |
| 11:46.37 | *** join/#asterisk-bugs jmls_home (n=jmls@mail.tessera.co.uk) |
| 12:06.15 | *** join/#asterisk-bugs caio1982 (i=caio1982@CAcert-br/caio1982) |
| 13:16.11 | *** join/#asterisk-bugs Victor_Yure (n=Victor_Y@unaffiliated/victoryure/x-837844) |
| 14:20.00 | *** join/#asterisk-bugs angler (n=angler@pdpc/sponsor/digium/angler) |
| 14:22.54 | *** join/#asterisk-bugs Juggie (i=Juggie@CPE001601df17fb-CM001a6682d7b6.cpe.net.cable.rogers.com) |
| 14:22.54 | *** mode/#asterisk-bugs [+o Juggie] by ChanServ |
| 15:07.18 | *** join/#asterisk-bugs putnopvut (n=putnopvu@216.207.245.1) |
| 15:07.18 | *** mode/#asterisk-bugs [+o putnopvut] by ChanServ |
| 15:27.52 | *** join/#asterisk-bugs seanbright (i=seanbrig@65.207.74.18) |
| 15:28.53 | *** join/#asterisk-bugs russellb (n=russell@asterisk/developer-and-stable-maintainer/drumkilla) |
| 15:28.53 | *** mode/#asterisk-bugs [+o russellb] by ChanServ |
| 15:37.19 | *** part/#asterisk-bugs Yourname`` (n=chatzill@unaffiliated/yourname/x-837320) |
| 15:47.36 | *** join/#asterisk-bugs fakhir (n=fakhir@unaffiliated/fakhir) |
| 15:49.04 | *** join/#asterisk-bugs d1mas (n=chatzill@host-99.dataart.net) |
| 15:50.10 | d1mas | russellb: hello. Are you there? |
| 16:04.47 | d1mas | looks like there is a problem with autoservice. Again :) |
| 16:06.08 | putnopvut | d1mas: yikes! What's wrong? |
| 16:07.09 | *** join/#asterisk-bugs fakhir (n=fakhir@unaffiliated/fakhir) |
| 16:07.45 | d1mas | The ast_autoservice_stop methods waits while channel list version (as_chan_list_state) changes before exiting. This is to make sure channel we are removing is not in the list which autoservice_run processes anymore |
| 16:08.14 | d1mas | The problem is that if autoservice_run is already in the ast_waitfor_n, it can take up to 500 ms before list version gets incremented |
| 16:08.50 | d1mas | so ast_autoservice_stop is delayed up to 500 ms. An NO ONE services the channel at that time - neither autoservice nor application |
| 16:09.12 | d1mas | as result, voice frames are lost on Zap channels. In my case this results of losing DTMF |
| 16:10.26 | d1mas | (because any delay in reading Zap devices longer than 4*20ms = 80ms definitely leads to loss of frames) |
| 16:16.29 | russellb | that should not happen |
| 16:16.38 | russellb | <PROTECTED> |
| 16:16.38 | russellb | <PROTECTED> |
| 16:16.43 | russellb | that will make it break out of poll |
| 16:16.48 | d1mas | I saw this |
| 16:17.11 | d1mas | However the fact is, I spend sometimes ~300ms in the last loop of autoservice_stop |
| 16:17.48 | d1mas | (can show you modifed code used to measure time and stats collected) |
| 16:19.16 | d1mas | as a workaround I just changed 500 ms to 50 ms on production machine because people started complaining loudly after today's update. Seems to help |
| 16:19.46 | *** topic/#asterisk-bugs by Qwell -> Asterisk Bugs Discussion -=- http://bugs.digium.com/ -=- http://asterisk.org/developers/bug-guidelines - If you are not discussing a report already posted to the bugtracker, the discussion probably belongs in #asterisk. -=- Overall issue count: 279 -=- Monitor new reports in #asterisk-commits -=- Test bugs marked Ready For Testing -=- Mantis happy now. |
| 16:19.53 | Qwell | 279... |
| 16:19.56 | d1mas | Aha! |
| 16:20.01 | d1mas | it is race condition |
| 16:20.20 | d1mas | the AS tread can be not in the poll yet when kill is called |
| 16:20.47 | russellb | true |
| 16:23.39 | russellb | http://pastebin.ca/950450 |
| 16:26.37 | d1mas | russellb: when BLOCKING file is set? |
| 16:27.46 | russellb | when the channel is being poll'd on |
| 16:28.24 | d1mas | is the patch against 1.4 ? it does not apply... |
| 16:28.40 | russellb | no, it was trunk i think |
| 16:28.45 | russellb | code is pretty much the same though |
| 16:28.51 | d1mas | ok |
| 16:28.53 | russellb | i'm surprised it didn't apply |
| 16:30.04 | d1mas | it maybe because of whitespaces |
| 16:30.11 | d1mas | instead of tabs |
| 16:31.30 | d1mas | well.... |
| 16:31.53 | d1mas | I believe there still can be a race condition |
| 16:32.27 | d1mas | what if autoservice_run thread gets preempted right after AST_LIST_UNLOCK(&aslist) but before ast_waitfor_n ? |
| 16:32.40 | russellb | ok, well i guess put it on the bug tracker |
| 16:32.43 | russellb | i can't really work on it this minute |
| 16:32.50 | d1mas | ok |
| 16:36.43 | d1mas | russellb: why not? the autoservice_run just finished building its list and released the lock. Then scheduler preempts the thread and autoservice_stop is run. It removes the item from the list and check blocking flag. The flag is off because autoservice_run has not entered into wait_nfd yet |
| 16:36.57 | d1mas | sorry, ignore the last one |
| 16:37.21 | d1mas | while it is true I was responding on your ealier message because I scrolled window back :)))) |
| 17:25.57 | Qwell | hmm, hey russellb ? think this is the same as the double free Dwayne found the other day? |
| 17:26.00 | Qwell | M12225 |
| 17:26.01 | MuffinMan | [feedback] [Asterisk] Channels/chan_sip/General 0012225: Crash shortly after launching reported by rickross http://bugs.digium.com/view.php?id=12225 |
| 17:26.14 | *** join/#asterisk-bugs d1mas (n=chatzill@host-99.dataart.net) |
| 17:26.49 | Qwell | in sipsock_read, it looks like |
| 17:26.58 | russellb | maybe ... |
| 17:27.05 | russellb | check to see if their revision is after my change |
| 17:27.14 | Qwell | trying to find it |
| 17:27.30 | Qwell | 109833 vs their 108926 |
| 17:27.32 | russellb | it's one of the last changes to chan_sip |
| 17:27.33 | Qwell | SO, nope |
| 17:27.34 | russellb | ah |
| 17:28.06 | russellb | this polycom speaker has some bass, haha ... |
| 17:28.25 | Qwell | so, I'm gonna close, with a note to reopen if it still happens after updating to 109833 or later |
| 17:28.31 | Qwell | sound good? |
| 17:28.36 | russellb | sounds good. |
| 17:28.51 | mvanbaak | good, GOOOOOOOOOD |
| 17:29.03 | russellb | is happy that he has working g722 MOH |
| 17:29.12 | russellb | core dump. |
| 17:29.14 | russellb | curses |
| 17:29.29 | mvanbaak | covers ears |
| 17:29.36 | russellb | oh, that was from my debug code .. |
| 17:29.40 | russellb | false alarm ! |
| 17:30.03 | mvanbaak | did you have DO_CRASH enabled ? ;) |
| 17:30.14 | russellb | haha ... nope |
| 17:33.36 | *** join/#asterisk-bugs fakhir (n=fakhir@unaffiliated/fakhir) |
| 17:43.03 | Juggie | that was a funny bug |
| 17:43.12 | Juggie | i enabled DO_CRASH and my asterisk is crashing |
| 17:43.13 | Juggie | what do i do |
| 17:44.40 | Qwell | Juggie: you saw what I did because of that, right? |
| 17:44.59 | Qwell | r109974 in trunk |
| 18:01.33 | Qwell | M7274 |
| 18:01.35 | MuffinMan | [assigned] [Asterisk] Addons/cdr_addon_mysql 0007274: [patch] record spooling support in cdr_addon_mysql.c reported by jpabuyer http://bugs.digium.com/view.php?id=7274 |
| 18:01.45 | Qwell | Is there any reason to keep that open? We all know that the branch is coming |
| 18:02.12 | mvanbaak | you main the nolosscdr stuff right ? |
| 18:02.17 | Qwell | yeah |
| 18:02.30 | mvanbaak | then you can close it (I think) |
| 18:02.38 | Qwell | russellb: thoughts? |
| 18:02.49 | russellb | i have no thoughts about anything |
| 18:03.04 | russellb | sure, that's fine |
| 18:03.10 | russellb | explain that we have a generic solution coming |
| 18:03.13 | russellb | and point to the branch |
| 18:03.15 | russellb | thank them for the patch |
| 18:03.17 | Qwell | it's been well documented |
| 18:03.19 | russellb | give him karma ... |
| 18:03.23 | Qwell | all of that has been done :D |
| 18:03.25 | Qwell | except the karma |
| 18:03.26 | russellb | k |
| 18:39.04 | *** join/#asterisk-bugs citats (n=james@mrplow.gnuinternet.com) |
| 18:39.04 | *** mode/#asterisk-bugs [+o citats] by ChanServ |
| 18:52.34 | Qwell | file: ping! |
| 18:52.37 | Qwell | M11901 |
| 18:52.39 | MuffinMan | [assigned] [Asterisk] Core/Channels 0011901: bridging chan_h323 and chan_sip reported by pj http://bugs.digium.com/view.php?id=11901 |
| 18:52.57 | Qwell | is there any way to tell if p2p bridging is enabled from the CLI or anything? |
| 18:53.10 | file | ummm, rtp debug |
| 18:53.34 | Qwell | no 'core show plz2bep2pbridging'? |
| 18:53.41 | file | nosir |
| 18:53.47 | mvanbaak | BUG |
| 18:53.55 | Qwell | <PROTECTED> |
| 18:54.13 | mvanbaak | core make pizza |
| 18:54.29 | file | RTP sessions aren't kept in a list so you can't iterate through, core show channel has no concept of RTP sessions on channels, and sip show channel for example has no idea of bridging... |
| 18:54.55 | Qwell | k |
| 19:02.23 | Juggie | Qwell, now you've made it so i have to deliver! :) |
| 19:06.52 | *** join/#asterisk-bugs d3wayne (n=deeewayn@76.29.245.9) |
| 19:11.07 | *** join/#asterisk-bugs implicit (n=implicit@200.12.227.181) |
| 19:11.07 | *** mode/#asterisk-bugs [+o implicit] by ChanServ |
| 19:40.53 | *** join/#asterisk-bugs infobot (i=ibot@pdpc/supporter/active/TimRiker/bot/apt) |
| 19:40.53 | *** topic/#asterisk-bugs is Asterisk Bugs Discussion -=- http://bugs.digium.com/ -=- http://asterisk.org/developers/bug-guidelines - If you are not discussing a report already posted to the bugtracker, the discussion probably belongs in #asterisk. -=- Overall issue count: 279 -=- Monitor new reports in #asterisk-commits -=- Test bugs marked Ready For Testing -=- Mantis happy now. |
| 20:11.43 | *** join/#asterisk-bugs angler (n=angler@pdpc/sponsor/digium/angler) |
| 21:18.19 | *** join/#asterisk-bugs implicit_ (n=implicit@200.12.227.181) |
| 23:07.47 | *** join/#asterisk-bugs implicit (n=implicit@200.12.227.181) |
| 23:07.47 | *** mode/#asterisk-bugs [+o implicit] by ChanServ |
| 23:26.31 | *** topic/#asterisk-bugs by russellb -> Asterisk Bugs Discussion -=- http://bugs.digium.com/ -=- http://asterisk.org/developers/bug-guidelines - If you are not discussing a report already posted to the bugtracker, the discussion probably belongs in #asterisk. -=- Overall issue count: 277 -=- Monitor new reports in #asterisk-commits -=- Test bugs marked Ready For Testing -=- Mantis happy now. |