IRC log for #asterisk-bugs on 20080312

02:12.05*** join/#asterisk-bugs russellb (n=russell@asterisk/developer-and-stable-maintainer/drumkilla)
02:12.05*** mode/#asterisk-bugs [+o russellb] by ChanServ
02:44.20*** join/#asterisk-bugs putnopvut (n=putnopvu@76.29.245.9)
02:44.20*** mode/#asterisk-bugs [+o putnopvut] by ChanServ
03:00.30*** join/#asterisk-bugs lmadsen (n=Leif@asterisk/documenteur-extraordinaire/blitzrage)
03:00.30*** mode/#asterisk-bugs [+o lmadsen] by ChanServ
04:17.31*** join/#asterisk-bugs putnopvut (n=putnopvu@76.29.245.9)
04:17.31*** mode/#asterisk-bugs [+o putnopvut] by ChanServ
08:03.55*** join/#asterisk-bugs oej (n=olle@62.97.243.70)
09:44.51*** join/#asterisk-bugs jmls (n=jmls@mail.tessera.co.uk)
10:59.01*** join/#asterisk-bugs kinom (n=jonathan@host182-44-static.48-88-b.business.telecomitalia.it)
11:02.30kinomHello, im undergoing a problem that i relate to bug n° 0011181. My asterisk version is Asterisk 1.4.18. i would like to get in touch with who is undergoing testing on that bug, It is assigned to putnopvut any idea how i can ge tin touch with him? cheers
11:48.16*** part/#asterisk-bugs kinom (n=jonathan@host182-44-static.48-88-b.business.telecomitalia.it)
12:09.23*** part/#asterisk-bugs jmls (n=jmls@mail.tessera.co.uk)
12:14.05*** join/#asterisk-bugs caio1982 (i=caio1982@CAcert-br/caio1982)
13:13.12*** join/#asterisk-bugs jsmith (n=jsmith@216.207.245.1)
13:13.12*** mode/#asterisk-bugs [+o jsmith] by ChanServ
13:41.03*** join/#asterisk-bugs putnopvut (n=putnopvu@216.207.245.1)
13:41.03*** mode/#asterisk-bugs [+o putnopvut] by ChanServ
14:12.35*** join/#asterisk-bugs agx (n=AGX@host63-216-static.34-88-b.business.telecomitalia.it)
14:36.10putnopvutM12098
14:36.13MuffinMan[ready for testing] [Asterisk] Channels/chan_sip/General 0012098: Segmentation fault in chan_sip.c reported by wegbert  http://bugs.digium.com/view.php?id=12098
14:36.27Qwellis that the sched race condition?
14:36.31putnopvutfile: would you mind having a look at the patch I submitted there to see if my logic isn't insane?
14:36.34putnopvutQwell: yes it is.
14:37.49fileI can in a bit, I'm working on... something... dry
14:37.57putnopvutfile: thanks :)
14:38.29fileCorydon76 did the last work in that area though, might want to see what he thinks
14:38.31*** join/#asterisk-bugs agx (n=AGX@host63-216-static.34-88-b.business.telecomitalia.it)
14:38.52putnopvutAh...
14:38.55putnopvutCorydon76-dig: ^^^
14:40.30Juggie12187 is interesting
14:40.39Juggieanyone have any idea whats going on there?
14:40.41putnopvutM12187
14:40.42MuffinMan[new] [Asterisk] Applications/app_chanspy 0012187: Crash in ChanSpy /AST_LIST_TRAVERSE_SAFE_BEGIN reported by atis  http://bugs.digium.com/view.php?id=12187
14:40.50putnopvutYikes! Chan_spy!
14:41.37*** join/#asterisk-bugs lmadsen (n=Leif@asterisk/documenteur-extraordinaire/blitzrage)
14:41.37*** mode/#asterisk-bugs [+o lmadsen] by ChanServ
14:44.06jsmiths/chan_spy/chan_sigh/
14:46.44putnopvutlol
14:47.35jsmithI thought you'd enjoy that
14:48.07filethere's still some sort of issue in slinfactory, but nobody has been able to give me access to a setup that exhibits it >_<
14:51.01JunK-Yjsmith: hehehe
15:18.08Corydon76-digputnopvut: looking at your patch now
15:18.30Corydon76-digputnopvut: I originally had that as the source of AST_SCHED_DEL, but I removed it because it could infinitely loop
15:19.05Corydon76-digputnopvut: you might could increase the number of retries, though
15:19.19putnopvutCorydon76-dig: Hmmm... in that specific scenario, I'm not sure how it could infinitely loop.
15:20.30Corydon76-digputnopvut: I don't see how a scheduled item can possibly run through 10 context switches, either, though
15:21.03putnopvutCorydon76-dig: the issue there is that the retrans_pkt is never going to complete because __sip_ack is holding the sip_pvt's lock.
15:21.10Corydon76-digbecause that's what it's doing
15:21.24putnopvutAnd until retrans_pkt completes, the packet will remain not in the scheduler queue.
15:21.49Corydon76-digOh, I see
15:22.36Corydon76-digI'd propose another macro that takes a lock as a third argument
15:22.47Corydon76-digbecause that logic will probably be needed elsewhere
15:22.53putnopvutCorydon76-dig: that sounds like a valid suggestion.
15:23.08Corydon76-digbut keep the short-circuit to prevent infinite loops
15:23.21putnopvutI'll agree to that.
15:24.04Corydon76-digOkey dokey
15:24.43ACiDVI'm testing patch 12098, no crash for now but have some channels that look locked/zombie and "show channels" or "soft hangup <tab>" freeze.
15:24.57ACiDVcore show locks available @ http://pastebin.ca/939746
15:25.47Corydon76-digACiDV: upload it to the bugtracker, please
15:25.59ACiDVok
15:26.15ACiDVI update 12098 or new ticket ?
15:26.23ACiDVnot sure if related
15:27.04Corydon76-digACiDV: update to latest SVN, I fixed that last night
15:27.13putnopvutOoh, cool :)
15:27.29ACiDVRevision: 108030
15:27.31Corydon76-digputnopvut: yes, I'm useful, even at 1 am
15:27.32ACiDValready updated
15:28.42putnopvutCorydon76-dig: I may be misreading the commits list, but it looks like you didn't make the change for 1.4, just trunk and 1.6.0
15:28.54Corydon76-digOh
15:28.59ACiDV:D
15:29.54Corydon76-digWell, in my defense, it was late
15:30.31ACiDVfix related to issues 12143 ?
15:31.01Corydon76-digYes
15:32.50ACiDVyou will commit change to 1.4 ?
15:35.21Corydon76-digCan't.  The code is different
15:36.25ACiDVhave a different approach for 1.4 or must force upgrade to 1.6 ? :P
15:37.52Corydon76-digWait, nevermind.  This is different
15:37.55Corydon76-digOpen a new bug
15:38.09ACiDVok
15:38.25Corydon76-digIt just looked similar
15:48.49ACiDVI presume I must run with valgrind ? or I must use 'ast_grab_core' to get more details ? (asterisk doesnt crash)
15:50.13ACiDVM12196
15:50.15MuffinMan[new] [Asterisk] Channels/General 0012196: Deadlock/Locked channels reported by jvandal  http://bugs.digium.com/view.php?id=12196
15:50.54*** join/#asterisk-bugs Victor_Yure (n=Victor_Y@unaffiliated/victoryure/x-837844)
15:55.36putnopvutI find it a bit discomforting that there are several threads in that core show locks output that are locking a channel and then a pvt. That's a mistake if I understand correctly.
15:56.09putnopvutUnless of course the channel that is locked is not the pvt's owner channel.
15:58.25*** join/#asterisk-bugs russellb (n=russell@asterisk/developer-and-stable-maintainer/drumkilla)
15:58.25*** mode/#asterisk-bugs [+o russellb] by ChanServ
16:11.20ACiDVfile, you found the cause on 11945 ? (audio out of sync) can have details ? :D
16:14.50filetime. reading. writing.
16:15.24ACiDV=)
16:42.43*** join/#asterisk-bugs jsmith (n=jsmith@216.207.245.1)
16:42.43*** mode/#asterisk-bugs [+o jsmith] by ChanServ
17:52.53*** join/#asterisk-bugs oej (n=olle@114.62-97-206.bkkb.no)
18:34.06*** join/#asterisk-bugs jmls_work (n=jmls@mail.tessera.co.uk)
18:49.23*** join/#asterisk-bugs fakhir (n=fakhir@unaffiliated/fakhir)
19:25.51ACiDVCorydon76-dig, I just upload bt for 12196
19:31.29Corydon76-digThank you
19:39.20*** join/#asterisk-bugs oej (n=olle@114.62-97-206.bkkb.no)
19:41.55*** part/#asterisk-bugs jmls_work (n=jmls@mail.tessera.co.uk)
19:42.12*** join/#asterisk-bugs jmls_work (n=jmls@mail.tessera.co.uk)
19:42.28Corydon76-digACiDV: I need the backtrace while the problem is actually happening
19:43.24*** part/#asterisk-bugs jmls_work (n=jmls@mail.tessera.co.uk)
19:43.33*** join/#asterisk-bugs jmls_work (n=jmls@mail.tessera.co.uk)
19:43.37*** join/#asterisk-bugs russellb (n=russell@asterisk/developer-and-stable-maintainer/drumkilla)
19:43.37*** mode/#asterisk-bugs [+o russellb] by ChanServ
19:44.24*** part/#asterisk-bugs jmls_work (n=jmls@mail.tessera.co.uk)
19:57.25ACiDVCorydon76-dig I have attach gdb to my process and execute the 2 commands
19:57.51ACiDVand current process has channel locked
19:57.59Corydon76-digACiDV: okay
19:58.21ACiDVI'm on gdb and process still running, if you need anything else
19:58.35ACiDVmeantime, I search on google how gdb work :P
19:58.50ACiDVand if needed, I can restart * and easily reproduce the problem
19:59.19Corydon76-digBecause this backtrace isn't showing any threads stuck in a hangup routine
19:59.48putnopvutThread 3 is stuck waiting for a lock.
20:00.07putnopvutNo telling where though.
20:00.35Corydon76-digRight
20:02.00Corydon76-digReason I was looking, though, was that his core show locks from earlier showed it stuck with a lock from sip_hangup, yet nothing in this backtrace shows that routine
20:02.53ACiDVI just restart * and check if same bt/lock
20:03.12ACiDVabout lock... its true to say that its not really a lock if it doesnt show ' Locked Here .... ' ?
20:03.36Corydon76-digACiDV: correct
20:04.02Corydon76-digIf you have no locks indicated in 'core show locks' then it's not a deadlock
20:04.26Corydon76-digresource starvation, perhaps, but I don't know how you'd get even that without a lock
20:06.59ACiDVI dial a queue w/ multiple agents, all call passthrough a sip trunk... if I cut the link between 2 servers, I still see lots of call on 'show channels' and on CLI it retry to dial agents at each queue retry
20:07.08ACiDVperhaps that the problem is somewhere else
20:09.38ACiDVuploading a new bt.txt (in fact... error on filename...) that contain core show locks... have  Lock #0 ast_hangup ... Lock #1 sip_hangup ...
20:09.44ACiDVto 12196
20:12.28ACiDVand uploading new backtract.txt
20:19.33ACiDVif I do a show channel <tab> ... I note that all zombies channels are Local/1234@context-hhhh,2  ... always end w/ ,2
20:20.44ACiDVorphan queue members where "caller" has already hangup from queue
20:23.01ACiDVand just see this on cli
20:23.01ACiDV[Mar 12 16:21:57] DEBUG[30902]: chan_iax2.c:1909 iax2_destroy_helper: Unable to cancel schedule ID 1114.  This is probably a bug (chan_iax2.c: iax2_destroy_helper, line 1909).
20:23.05ACiDVwill attach to ticket
20:58.37*** join/#asterisk-bugs jmls (n=jmls@mail.tessera.co.uk)
21:24.55*** join/#asterisk-bugs jsmith (n=jsmith@216.207.245.1)
21:24.55*** mode/#asterisk-bugs [+o jsmith] by ChanServ
23:40.43*** join/#asterisk-bugs ZX81 (n=ZX81@202.49.106.158)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.