| 00:36.02 | *** join/#asterisk-dev moy (n=moy@CPE001cdfec4cee-CM00080dab8485.cpe.net.cable.rogers.com) |
| 00:41.35 | *** join/#asterisk-dev jameswf (n=james@unaffiliated/jameswf-home) |
| 01:01.38 | *** join/#asterisk-dev Juggie (n=Juggie@CPE001601df17fb-CM001ceac25ada.cpe.net.cable.rogers.com) |
| 01:20.35 | *** join/#asterisk-dev kpfleming (n=kpflemin@asterisk/digium-director-of-software-tech/kpfleming) |
| 01:20.36 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 01:48.52 | *** join/#asterisk-dev Deeewayne (n=dwayne@76.29.245.9) |
| 01:48.52 | *** mode/#asterisk-dev [+o Deeewayne] by ChanServ |
| 02:03.39 | leifmadsen | snuff-work: pong |
| 02:12.40 | snuff-work | M15392 |
| 02:12.42 | MuffinMan | [new] [Asterisk] PBX/General 0015392: [patch] 1.6.1.1: Memory handling error in main/pbx.c (pbx_extension_helper) reported by yrashk (Karma: neutral) https://issues.asterisk.org/view.php?id=15392 |
| 02:13.19 | snuff-work | i'm unsure what to do about it in general..basically in this guys case the context is null.. |
| 02:13.28 | snuff-work | solaris + printf with null = segfault |
| 02:17.12 | leifmadsen | hmmm |
| 02:17.15 | leifmadsen | let me look quickly |
| 02:18.07 | leifmadsen | ah yes, that one |
| 02:18.50 | Juggie | nice nice |
| 02:19.04 | Juggie | canada revenue agency (my employer) has decided to do a mini audit on me |
| 02:19.09 | snuff-work | i'm not sure if calling pbx_helper is ok with null or not |
| 02:19.14 | Juggie | the funny part is they want proof of shit that happened like 3 years ago |
| 02:19.46 | Juggie | 'please show us you documents proving you bought a house and used your rrsp money you withdrew towards it' |
| 02:19.48 | Juggie | is the gist of it |
| 02:21.24 | Juggie | crazy bastards |
| 02:21.46 | Juggie | guy includes his name too so i think when i send him the info i'm going to send him a little internal mail telling him the information has been faxed :) |
| 02:22.01 | snuff-work | heh |
| 02:22.51 | Juggie | hehe |
| 02:22.58 | Juggie | i could also actually send it internal mail (post) |
| 02:24.27 | leifmadsen | Juggie: ya, that's what I'd do |
| 02:24.35 | leifmadsen | "hey... do you *really* need this?" |
| 02:24.43 | leifmadsen | -- Donny, CRA. |
| 02:24.48 | *** join/#asterisk-dev jameswf (n=james@unaffiliated/jameswf-home) |
| 02:25.08 | russellb | waves his hand like a jedi and says "There is no audit." |
| 02:25.44 | leifmadsen | nice |
| 02:26.58 | snuff-work | leifmadsen: i'll relook @ that bug tonight.. provide an update to u in ur am |
| 02:27.24 | leifmadsen | sweet, thank you sir! |
| 02:28.03 | snuff-work | is eating a hot n spicy noodle box that he told them to make extra spicy.. |
| 02:28.15 | snuff-work | taking a bit longer to eat now |
| 02:28.40 | leifmadsen | lol |
| 02:28.41 | leifmadsen | nice |
| 02:29.10 | snuff-work | think there is bout 4 chillies + hot sauce |
| 02:30.04 | leifmadsen | damn yo |
| 02:31.27 | snuff-work | brings tears to my eyes |
| 02:31.47 | leifmadsen | I'm jealous |
| 02:32.02 | leifmadsen | it is 10:30pm here, so it is too late for spicy food |
| 02:32.42 | snuff-work | shame |
| 02:40.03 | Juggie | leifmadsen, my sister got hit too |
| 02:40.10 | leifmadsen | ouch |
| 02:40.11 | Juggie | she has to provide like a zillion school receipts though |
| 02:40.20 | leifmadsen | hopes he doesn't get hit |
| 02:40.36 | Juggie | for me it is just the purchase of my house (which i had already dug out because of selling it) and then the transfer of my RRSP from mutual funds to my account |
| 02:40.41 | Juggie | which i found in my house file |
| 02:40.45 | Juggie | pays to keep files i guess :) |
| 02:49.41 | leifmadsen | yep |
| 02:49.45 | leifmadsen | I have all that stuff on file |
| 02:49.52 | leifmadsen | I keep the important stuff |
| 02:49.57 | leifmadsen | anyays, bedtime |
| 04:12.48 | *** join/#asterisk-dev lmsteffan (n=laurent@reef.ac-noumea.nc) |
| 05:09.38 | *** join/#asterisk-dev lmsteffan (n=laurent@114.69.178.97) |
| 05:47.47 | *** join/#asterisk-dev |Cybex| (n=John@80.100.126.176) |
| 06:41.06 | *** join/#asterisk-dev tzafrir_laptop (n=tzafrir@local.xorcom.com) |
| 06:41.06 | *** mode/#asterisk-dev [+o tzafrir_laptop] by ChanServ |
| 07:06.15 | *** join/#asterisk-dev obnauticus (n=obnautic@about/windows/regular/obnauticus) |
| 07:32.53 | *** join/#asterisk-dev tamiel (n=tamiel@213.30.183.226) |
| 07:59.41 | *** join/#asterisk-dev Dovid (n=annon@213.8.121.90) |
| 08:00.53 | *** join/#asterisk-dev codefreeze-lap (n=murf@72.21.67.40) |
| 08:00.53 | *** mode/#asterisk-dev [+o codefreeze-lap] by ChanServ |
| 09:57.51 | *** join/#asterisk-dev qdk (n=qdk@0x573c23bb.bynqu1.dynamic.dsl.tele.dk) |
| 10:59.53 | *** join/#asterisk-dev kpfleming (n=kpflemin@asterisk/digium-director-of-software-tech/kpfleming) |
| 10:59.53 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 11:33.10 | *** join/#asterisk-dev J4zen (n=j4zen@a82-95-153-17.adsl.xs4all.nl) |
| 11:33.23 | J4zen | I was wondering, in the ASterisk Manager API theres a function called ExtensionState. It returns a status-code when given an extension id |
| 11:33.27 | J4zen | now, on http://www.voip-info.org/wiki/index.php?page=Asterisk+Manager+API+Action+ExtensionState some of those status codes are listed |
| 11:33.44 | J4zen | Does anyone know of a complete documentation of this function? |
| 11:34.01 | J4zen | Because the documentation on voip-info is not close to complete im afraid |
| 11:37.00 | *** join/#asterisk-dev J4zen (n=j4zen@95-36-106-201.dsl.alice.nl) |
| 11:41.50 | snuff-home | well J4zen look at the source.. main/manager.c |
| 11:42.14 | snuff-home | then u might have to look @ the code itself for extensionstgat |
| 11:42.56 | J4zen | will do, no full documentation on methods available then? |
| 11:43.05 | J4zen | or extensive |
| 12:03.37 | snuff-home | not that i can think of |
| 12:03.47 | snuff-home | or read doc/ directory |
| 12:04.09 | snuff-home | the asterisk.pdf might also help |
| 12:56.24 | *** join/#asterisk-dev eliel (n=eliels@200.61.172.61) |
| 12:56.24 | *** mode/#asterisk-dev [+o eliel] by ChanServ |
| 13:10.41 | *** join/#asterisk-dev jeffg (n=jeffg@pdpc/supporter/active/jeffg) |
| 13:25.45 | *** join/#asterisk-dev Greek-Boy (n=greek@41.188.154.137) |
| 13:26.18 | *** join/#asterisk-dev Greek-Boy (n=greek@41.188.154.137) |
| 13:33.25 | *** join/#asterisk-dev Greek-B0y (n=greek@41.188.154.137) |
| 13:59.48 | *** join/#asterisk-dev xpot-mobile (n=james@71-213-54-186.slkc.qwest.net) |
| 14:04.33 | *** join/#asterisk-dev putnopvut (n=putnopvu@asterisk/master-of-queues/mmichelson) |
| 14:04.33 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:18.50 | *** join/#asterisk-dev leifmadsen (n=Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 14:18.50 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 14:24.59 | *** join/#asterisk-dev dioedu (n=dioedu@201.7.117.114) |
| 14:27.03 | *** join/#asterisk-dev riksta (n=rick@92.63.131.41) |
| 14:28.20 | *** join/#asterisk-dev Deeewayne (n=dwayne@75.76.254.162) |
| 14:28.20 | *** mode/#asterisk-dev [+o Deeewayne] by ChanServ |
| 14:29.30 | riksta | Hi, I have been using asterisk 1.6.1 and 1.6.2 and I am encountering problems with CDRs not being set to BUSY correctly. Has anyone else seen the issue? (SIP->SIP calls) From the sip debug information i see asterisk gets the hangup cause like so: CSeq: 102 BYE X-Asterisk-HangupCause: User busy X-Asterisk-HangupCauseCode: 17 |
| 14:31.30 | leifmadsen | riksta: have you checked the issue tracker to determine if there is a bug opened? Your question is also probably better suited for #asterisk, where more people are going to be using asterisk rather than developing for it |
| 14:32.11 | riksta | leifmadsen: I have searched quite heavily but not found anything i can see relevant. I will try #asterisk |
| 14:33.06 | leifmadsen | okie |
| 14:33.23 | leifmadsen | if you can't find anything, then you can open an issue with the appropriate information for how to reproduce it |
| 14:33.59 | riksta | yeah, i certainly will |
| 14:34.02 | riksta | cheers |
| 14:35.16 | leifmadsen | peas |
| 14:50.00 | *** join/#asterisk-dev sruffell (n=sruffell@asterisk/the-kernel-guy/sruffell) |
| 14:50.00 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 14:56.35 | *** join/#asterisk-dev ctooley (n=ctooley@c-68-52-165-168.hsd1.tn.comcast.net) |
| 14:56.35 | *** mode/#asterisk-dev [+o ctooley] by ChanServ |
| 15:01.27 | *** join/#asterisk-dev jpeeler (n=jpeeler@asterisk/digium-software-dev/jpeeler) |
| 15:01.27 | *** mode/#asterisk-dev [+o jpeeler] by ChanServ |
| 15:05.29 | *** join/#asterisk-dev kpfleming (n=kpflemin@asterisk/digium-director-of-software-tech/kpfleming) |
| 15:05.29 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 15:06.14 | *** join/#asterisk-dev thansen (n=thansen@76.27.110.194) |
| 15:15.32 | *** join/#asterisk-dev jameswf (n=james@unaffiliated/jameswf-home) |
| 15:16.13 | *** join/#asterisk-dev atis_work (n=atis_wor@193.238.212.171) |
| 15:25.23 | *** join/#asterisk-dev xpot-mobile (n=james@71-213-54-186.slkc.qwest.net) |
| 15:27.15 | *** join/#asterisk-dev marv[work] (n=timr@router.asteriasgi.com) |
| 15:37.01 | *** join/#asterisk-dev snuff-home (n=snuffy@203-219-29-182-bri-ts6-2600.tpgi.com.au) |
| 15:37.01 | *** mode/#asterisk-dev [+o snuff-home] by ChanServ |
| 15:43.35 | *** join/#asterisk-dev anthm (n=anthm@freeswitch/developer/anthm) |
| 15:44.30 | *** join/#asterisk-dev jeffg (n=jeffg@pdpc/supporter/active/jeffg) |
| 16:23.26 | kaii | i encountered a strange problem and wonder if this is a bug. i've looked at the logs and corresponding parts of the asterisk source and of course i searched for similar problems in the issue tracker -- but didnt came up with a solution (or a reported issue) |
| 16:24.27 | kaii | after some strange log messages calling any SIP extension is no more possible, although passing from external PRI to internal BRI is still working. |
| 16:24.55 | kaii | i wrote down all information i have on http://pastebin.ca/1512395 |
| 16:25.28 | kaii | would be soo happy, if you could look at this and tell your thoughts ... |
| 16:25.38 | kaii | any comments are appreciated. |
| 16:26.29 | kaii | i'm willing to fix by myself, but at this point i'm stuck with debugging, having no idea on where the error occurs |
| 16:27.27 | kaii | i think it has to do with this message: file.c: Unexpected control subclass '16' |
| 16:27.43 | kaii | after it, all sip channels are auto-congested and after that, no sip channels are established anymore. |
| 16:33.27 | *** join/#asterisk-dev Laureano (n=Laureano@190.245.101.140) |
| 16:39.05 | kaii | asterisk does not crash/segfault and chan_sip.so still responds .. ("sip show channels") |
| 16:40.45 | Corydon76-dig | kaii: this is not a support channel |
| 16:41.33 | Corydon76-dig | Try in #asterisk |
| 16:49.21 | kaii | Corydon76-dig: im afraid nobody except the people who are in #asterisk-dev too will grasp the problem |
| 16:49.45 | kaii | this is not a configuration issue, its definitely a bug. |
| 16:49.50 | anthm | sounds like a deadlock in chan_sip get a gcore and run thread apply all bt and look for any threads stuck on a mutex lock |
| 16:50.35 | kaii | anthm: problem is, this is a production machine and was already restarted by a collegue in the support |
| 16:50.52 | anthm | next time it happes tell him to gcore it first |
| 16:50.54 | kaii | i'll advise all supporters that in case this happens, a gcore is necessary |
| 16:53.06 | kaii | i wonder where the unexpected control packet with class 16 comes from and what it has to do with that issue |
| 16:54.01 | Qwell | kaii: 16 is hold, if I understand that code correctly |
| 16:54.42 | kaii | Qwell: so the PRI where the call has come from is putting me on hold? lol |
| 16:54.49 | kaii | *trying to put me on hold |
| 16:58.40 | Qwell | dunno, just saying what 16 stands for |
| 17:01.59 | *** join/#asterisk-dev mvanbaak (n=michiel@asterisk/contributor-and-bug-marshal/mvanbaak) |
| 17:01.59 | *** mode/#asterisk-dev [+o mvanbaak] by ChanServ |
| 17:04.24 | kaii | funny |
| 17:12.05 | *** join/#asterisk-dev ctooley (n=ctooley@c-68-52-165-168.hsd1.tn.comcast.net) |
| 17:12.06 | *** mode/#asterisk-dev [+o ctooley] by ChanServ |
| 17:16.37 | *** join/#asterisk-dev atis_work (n=atis_wor@193.238.212.171) |
| 17:57.36 | *** join/#asterisk-dev leifmadsen (n=Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 17:57.36 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 18:16.33 | *** join/#asterisk-dev atis_work (n=atis_wor@193.238.212.171) |
| 18:23.02 | mvanbaak | hhmm, SFA only available for linux ? |
| 18:23.48 | mvanbaak | either that, or the README needs an update |
| 18:26.38 | Corydon76-dig | mvanbaak: ping |
| 18:27.40 | *** join/#asterisk-dev outtolunc (n=me@c-98-248-96-110.hsd1.ca.comcast.net) |
| 18:29.06 | *** join/#asterisk-dev kpfleming (n=kpflemin@asterisk/digium-director-of-software-tech/kpfleming) |
| 18:29.06 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 18:29.33 | mvanbaak | Corydon76-dig: pong |
| 18:29.56 | Corydon76-dig | mvanbaak: did you see the commit to malloc_hold last night? |
| 18:30.07 | mvanbaak | saw it 5 minutes ago |
| 18:30.20 | Corydon76-dig | Testing appreciated |
| 18:31.28 | mvanbaak | :) working on that |
| 18:40.36 | *** join/#asterisk-dev Greek-Boy (n=greek@41.188.154.137) |
| 18:43.59 | *** join/#asterisk-dev tzafrir_laptop (n=tzafrir@bzq-179-75-202.static.bezeqint.net) |
| 18:43.59 | *** mode/#asterisk-dev [+o tzafrir_laptop] by ChanServ |
| 18:45.06 | *** join/#asterisk-dev malcolmd (n=malcolmd@pdpc/sponsor/digium/malcolmd) |
| 18:45.06 | *** mode/#asterisk-dev [+o malcolmd] by ChanServ |
| 18:49.49 | mvanbaak | Corydon76-dig: still not compiling |
| 18:49.52 | mvanbaak | Corydon76-dig: working on it |
| 18:50.13 | mvanbaak | astmm.c:469: error: `MAP_ANONYMOUS' undeclared (first use in this function) |
| 18:50.57 | mvanbaak | Corydon76-dig: http://pastebin.ca/1512584 |
| 18:56.02 | mvanbaak | it's MAP_ANON |
| 18:56.55 | Qwell | mvanbaak: MAP_ANON is deprecated |
| 19:00.53 | seanbright | not on openbsd it isn't |
| 19:01.30 | mvanbaak | indeed |
| 19:01.45 | mvanbaak | on OpenBSD there is no MAP_ANONYMOUS |
| 19:01.46 | seanbright | wonders if asterisk will run on irix |
| 19:01.57 | mvanbaak | ultrix |
| 19:02.01 | seanbright | or minix |
| 19:02.02 | seanbright | heh |
| 19:02.16 | mvanbaak | as/400 |
| 19:02.41 | seanbright | is that the only change required to get her compiling? MAP_ANONYMOUS to MAP_ANON? |
| 19:03.17 | mvanbaak | seanbright: nope. look at the pastebin |
| 19:04.37 | *** join/#asterisk-dev dbrooks (n=dbrooks@nat/digium/x-vyaaoadhuoogxnpu) |
| 19:05.41 | seanbright | good times |
| 19:07.31 | mvanbaak | yup |
| 19:11.18 | Corydon76-dig | mvanbaak: What's the context of line 342 there? |
| 19:11.26 | Corydon76-dig | I seem to have different line numbers |
| 19:12.13 | mvanbaak | <PROTECTED> |
| 19:12.28 | eliel | hello |
| 19:12.32 | Corydon76-dig | That's nuts. I'm already casting there |
| 19:13.04 | mvanbaak | dont shoot the messenger |
| 19:13.05 | mvanbaak | ;) |
| 19:13.35 | Corydon76-dig | svn update |
| 19:14.55 | mvanbaak | btw |
| 19:14.57 | Corydon76-dig | btw, just so you realize, this IS 1.4 |
| 19:15.00 | mvanbaak | at the top of the file |
| 19:15.11 | mvanbaak | the #include <malloc.h> |
| 19:15.21 | mvanbaak | malloc.h is deprecated |
| 19:15.30 | mvanbaak | so to fix that i did: |
| 19:15.37 | mvanbaak | #if defined(__OpenBSD__) |
| 19:15.43 | mvanbaak | #include <stdlib.h> |
| 19:15.46 | mvanbaak | #else |
| 19:15.53 | mvanbaak | #include <malloc.h> |
| 19:15.55 | mvanbaak | #endif |
| 19:17.50 | mvanbaak | ok, getting there |
| 19:17.54 | Corydon76-dig | Does it really need stdlib? |
| 19:18.27 | mvanbaak | http://pastebin.ca/1512660 |
| 19:18.28 | Corydon76-dig | mmap should be coming from sys/mman.h, not stdlib |
| 19:19.08 | mvanbaak | I dont know. I just followed the error message on my screen |
| 19:19.21 | mvanbaak | 'malloc.h has been deprecated in favor of stdlib.h' |
| 19:19.28 | Corydon76-dig | What was the error when you removed both? |
| 19:19.52 | Corydon76-dig | I only included malloc.h for memalign |
| 19:20.17 | mvanbaak | hhmm |
| 19:20.28 | mvanbaak | if I remove both I dont see any extra error |
| 19:20.39 | mvanbaak | only the two ones from the last pastebin |
| 19:20.49 | mvanbaak | I know have: |
| 19:20.58 | mvanbaak | #if !defined(__OpenBSD__) |
| 19:21.03 | mvanbaak | #include <malloc.h> |
| 19:21.06 | mvanbaak | #endif |
| 19:21.20 | Corydon76-dig | arg6 on line 99 is a string, not an int |
| 19:21.35 | *** join/#asterisk-dev errotan (n=errotan@62.201.123.35) |
| 19:21.39 | Corydon76-dig | err, line 199 |
| 19:21.57 | Corydon76-dig | unless I'm looking at the wrong line again |
| 19:22.08 | mvanbaak | hhmm |
| 19:22.20 | mvanbaak | format string #6 is %s |
| 19:22.32 | Corydon76-dig | That's what I said! |
| 19:22.52 | Corydon76-dig | So why's it saying it's an int format |
| 19:22.53 | mvanbaak | but arg6 isn't (at least that's what the error tells me) |
| 19:22.58 | seanbright | does arg 6 mean format sepcific 6 or astmm_log arg 6? |
| 19:23.06 | seanbright | format specifier* |
| 19:23.41 | mvanbaak | cur->alloc_file is not a string I think |
| 19:24.12 | Corydon76-dig | seanbright: either way, it's a string |
| 19:24.21 | Corydon76-dig | mvanbaak: it's a char[40] |
| 19:24.57 | Corydon76-dig | arg 6 is alloc_func, format 6 is alloc_file. Both are strings |
| 19:26.00 | Corydon76-dig | Ah, but astmm_log is a macro. |
| 19:26.03 | Corydon76-dig | Found it. |
| 19:26.08 | mvanbaak | sweet ! |
| 19:26.41 | Corydon76-dig | svn update |
| 19:27.39 | Corydon76-dig | Hopefully shouldn't be any trouble porting this to trunk and 1.6.x |
| 19:28.09 | mvanbaak | same orror on line 237 |
| 19:28.21 | mvanbaak | that's the last one (on my screen right now) |
| 19:29.03 | Corydon76-dig | revision 209600 |
| 19:29.32 | mvanbaak | uhhuh |
| 19:29.39 | mvanbaak | saw it on #asterisk-commits |
| 19:29.42 | mvanbaak | compiles now ! |
| 19:29.53 | Corydon76-dig | waits for the first crash |
| 19:30.13 | mvanbaak | lol |
| 19:30.22 | seanbright | annnnnnnd go |
| 19:30.27 | Corydon76-dig | btw, on the morning conference call for the last 2 days, I've been running this branch continuously |
| 19:30.51 | Corydon76-dig | Memory holding steady at 26M/call |
| 19:31.06 | Corydon76-dig | (+58M for the core) |
| 19:31.36 | seanbright | 26MB per call? |
| 19:31.42 | Corydon76-dig | Yep |
| 19:31.44 | seanbright | that seems... high |
| 19:31.52 | mvanbaak | anima*CLI> core show version |
| 19:31.52 | mvanbaak | Asterisk SVN-tilghman-malloc_hold-r209600-/branches/1.4 built by mvanbaak @ anima.vanbaak.info on a amd64 running OpenBSD on 2009-07-30 19:30:54 UTC |
| 19:31.55 | Corydon76-dig | Remember that it holds each frame in memory for 90 seconds |
| 19:31.59 | seanbright | ahh |
| 19:32.02 | seanbright | well there's that. |
| 19:33.08 | Corydon76-dig | I currently have 9189 1 page segments in cache |
| 19:33.13 | Corydon76-dig | ready for allocation |
| 19:33.24 | Corydon76-dig | mvanbaak: "memory show available" |
| 19:34.12 | mvanbaak | 1278461 bytes allocated in 2711 allocations |
| 19:34.29 | mvanbaak | anima*CLI> memory show available |
| 19:34.29 | mvanbaak | Size 4096 * 1: 2434 |
| 19:34.29 | mvanbaak | Size 4096 * 2: 2 |
| 19:34.29 | mvanbaak | Size 4096 * 3: 6 |
| 19:34.29 | mvanbaak | Size 4096 * 4: 2 |
| 19:34.31 | mvanbaak | Size 4096 * 5: 7 |
| 19:34.34 | mvanbaak | Size 4096 * 6: 1 |
| 19:34.36 | mvanbaak | Size 4096 * 9: 2 |
| 19:34.39 | mvanbaak | anima*CLI> |
| 19:34.44 | Corydon76-dig | Pretty cool, eh? |
| 19:34.50 | mvanbaak | freaking awesome ! |
| 19:35.23 | Corydon76-dig | Oh, and "memory show polluted" should come up with no lines |
| 19:35.32 | mvanbaak | it does |
| 19:35.46 | mvanbaak | anima*CLI> memory show polluted |
| 19:35.46 | mvanbaak | anima*CLI> |
| 19:36.12 | Corydon76-dig | So you now have something that does illegal memory access similar to how valgrind works, without having to run valgrind |
| 19:36.24 | Corydon76-dig | s/access/access detection/ |
| 19:36.34 | mvanbaak | gheh |
| 19:36.46 | mvanbaak | classic quote there :) |
| 19:36.54 | seanbright | how much performance impact? |
| 19:37.19 | Corydon76-dig | seanbright: Increased memory usage, only |
| 19:37.19 | Corydon76-dig | The performance impact is negligible |
| 19:37.22 | seanbright | excellent |
| 19:37.42 | mvanbaak | memory is cheap. (if you run amd64, otherwise you're limited to 2GB on OpenBSD) |
| 19:37.43 | Corydon76-dig | You know what was the trickiest part of writing this? |
| 19:37.57 | mvanbaak | getting it running on !Linux ? |
| 19:38.17 | Corydon76-dig | My own illegal memory accesses by my own routines weren't killing Asterisk and coredumping |
| 19:38.36 | mvanbaak | whehehehe |
| 19:39.02 | Corydon76-dig | Instead, they were just flagged by the reporting thread and sent to the log |
| 19:39.05 | mvanbaak | btw, this only works for stuff using the ast_* functions for memory allocation right ? |
| 19:39.07 | seanbright | segv handlers have always been voodoo to me |
| 19:39.52 | seanbright | Corydon76-dig: one thing i noticed which i am sure is a non-issue |
| 19:39.52 | mvanbaak | so if there's an malloc in one of the .c files, it wont be handled right ? |
| 19:39.52 | seanbright | in handle_sigsegv you have a comment that says: |
| 19:39.52 | seanbright | The general rule of thumb is don't try to allocate memory or output to a |
| 19:39.53 | Corydon76-dig | mvanbaak: I think that's correct, although we might intercept all of them with the header file |
| 19:39.53 | seanbright | <PROTECTED> |
| 19:40.10 | seanbright | but on the 4th line of that you are fprintf'ing to stderr |
| 19:40.14 | seanbright | is that kosher? |
| 19:40.17 | mvanbaak | Corydon76-dig: you might be right there |
| 19:41.04 | Corydon76-dig | seanbright: The issue really is that we don't want to output to a real file, because we might be in the middle of another access of the same file. But stderr is a special file handle where that isn't an issue |
| 19:41.17 | seanbright | excellent |
| 19:41.29 | Corydon76-dig | You cannot rewind the pointer on stderr, for example |
| 19:42.37 | Corydon76-dig | mvanbaak: so yes, this might could be extracted and put to use in another project |
| 19:42.53 | mvanbaak | Corydon76-dig: already working on that ;) |
| 19:43.16 | Corydon76-dig | Just remember that it MUST be compiled in, as written |
| 19:43.24 | mvanbaak | yup |
| 19:43.33 | Corydon76-dig | I suppose you could do a LD_PRELOAD trick, if that works on obsd |
| 19:43.49 | mvanbaak | maybe, but I wont count on it |
| 19:44.06 | mvanbaak | that works on obsd, but still ... |
| 19:44.23 | Corydon76-dig | Ask your obsd buddies whether this is generally useful |
| 19:44.42 | mvanbaak | that's why I was extracting it |
| 19:44.44 | Corydon76-dig | I mean, it's all GPL, for what it's worth |
| 19:44.54 | mvanbaak | we cant use it in OpenBSD itself |
| 19:44.58 | Corydon76-dig | unless you obtain another license for it from Digium |
| 19:45.11 | Corydon76-dig | I can't rule that out; it's not up to me |
| 19:45.18 | mvanbaak | I know |
| 19:45.31 | mvanbaak | I'm only extracting it to have it previewed and tuff |
| 19:45.37 | mvanbaak | s/tuff/stuff |
| 19:56.10 | mvanbaak | if digium was to provide it under MIT or BSD we could use it |
| 19:56.20 | mvanbaak | but I'll ask legal |
| 19:56.30 | mvanbaak | dont annoy devs with it |
| 20:25.39 | *** join/#asterisk-dev rawko (n=rawko@adsl-070-155-127-242.sip.asm.bellsouth.net) |
| 20:25.43 | rawko | i got $50 via paypal for someone who can help me setup my trunks and dialing rules with a tdm800 with elastix |
| 20:26.22 | jsmith | rawko: You'd be better off asking in #asterisk |
| 20:26.23 | Qwell | rawko: #freepbx... |
| 20:26.25 | mvanbaak | try #asterisk-biz and/or the asterisk-biz mailinglist |
| 20:26.41 | Qwell | well, at least we all agree on one thing - this sure isn't the place for it. |
| 20:26.42 | jsmith | Anywhere but here? :-) |
| 20:27.02 | rawko | gota try |
| 20:27.04 | mvanbaak | indeed |
| 20:27.08 | Qwell | No, you don't. |
| 20:27.09 | rawko | figured people in here know what the scoop is |
| 20:29.42 | *** part/#asterisk-dev rawko (n=rawko@adsl-070-155-127-242.sip.asm.bellsouth.net) |
| 20:52.35 | *** join/#asterisk-dev dbrooks (n=dbrooks@nat/digium/x-jbhpcdckxvjmthlo) |
| 21:04.58 | *** join/#asterisk-dev anthm (n=anthm@freeswitch/developer/anthm) |
| 21:20.36 | *** join/#asterisk-dev lmsteffan (n=laurent@reef.ac-noumea.nc) |
| 21:54.49 | *** join/#asterisk-dev kpfleming (n=kpflemin@asterisk/digium-director-of-software-tech/kpfleming) |
| 21:54.49 | *** mode/#asterisk-dev [+o kpfleming] by ChanServ |
| 22:02.08 | lmsteffan | putnovput: would you have a few minutes to spend for a discussion about the recent (early July) changes you made to chan_sip.c ? I am having problems with that section of the code |
| 22:07.02 | Corydon76-dig | ~ask |
| 22:07.03 | infobot | Questions in the channel should be specific, informative, complete, concise, and on-topic. Don't ask if you can ask a question first. Don't ask if a person is there; just ask what you intended to ask them. Better questions more frequently yield better answers. We are all here voluntarily or against our will. |
| 22:13.06 | putnopvut | lmsteffan: what changes are you referring to? |
| 22:13.45 | lmsteffan | OK. At one point in the code a comment claims that it's not necessary to hold a lock (target.chan1) beacuse it was already set in another function. |
| 22:14.43 | lmsteffan | It looks to me like this lock is not in fact held. The function in which this happens is local_attended_transfer. |
| 22:15.23 | putnopvut | right in get_sip_pvt_byid_locked, the channel *should* be locked. |
| 22:17.17 | putnopvut | Also what version of Asterisk are you referring to? |
| 22:17.56 | lmsteffan | I'm testing on SVN trunk, usually the latest version. |
| 22:18.00 | putnopvut | All right. |
| 22:18.30 | putnopvut | In get_sip_pvt_byid_locked, we get and lock both targetcall_pvt and targetcall_pvt->owner. |
| 22:18.52 | putnopvut | You can see that target.chan1 is set to targetcall_pvt->owner later. |
| 22:19.32 | lmsteffan | yes but a few lines before the comment it has been unlocked |
| 22:19.48 | putnopvut | Ah, in the else case...I see what you're saying. |
| 22:20.12 | putnopvut | Oh yuck, it's getting unlocked twice |
| 22:20.19 | lmsteffan | and in the failed case also, if indeed targetcall_pvt->owner and targe.chan1 are synonymous |
| 22:22.00 | putnopvut | Yeah, that needs to be taken care of. Thanks for pointing it out! |
| 22:23.48 | lmsteffan | so I tried to lock it again, got a deadlock with the ¤t->chan1 lock (I got a trace if you need it) but this gave rise to an infinite loop of errors (Bad Magic Number for an astobj2 object) |
| 22:27.00 | lmsteffan | and also, I noticed that this behaviour is not visible when compiling the code with "too much" debugging - like AST_DEV_MODE with deadlock detection, and other options |
| 22:27.47 | lmsteffan | I suspect this is due to race conditions between threads, so this bug is timing-dependent. |
| 22:47.16 | putnopvut | oh this is going to be *FUN* |
| 22:49.15 | putnopvut | That thread has to lock two sip_pvts and two channels. Making that work without deadlocking is just ... yuck. |
| 22:51.56 | russellb | and by *FUN* you really mean "FUN" |
| 22:52.10 | lmsteffan | Well I am relieved, I thought it was only me :-D |
| 22:55.53 | lmsteffan | Now that you've confirmed that it deos look like a bug, would you like me to enter a formal bug report ? |
| 22:57.25 | putnopvut | lmsteffan: yes please. I won't be able to get that fixed today, and if there's no report, I'll likely forget ;) |
| 23:05.50 | snuff-work | i have a question about pbx_extension_helper.. |
| 23:06.29 | snuff-work | basically is it valid to call with null a context/exten/prio ? |
| 23:43.16 | *** join/#asterisk-dev moy (n=moy@CPE001cdfec4cee-CM00080dab8485.cpe.net.cable.rogers.com) |