| 00:05.21 | *** join/#asterisk-dev anthm (n=anthm@freeswitch/developer/anthm) |
| 00:20.00 | *** join/#asterisk-dev moy (n=moy@189.169.91.147) |
| 00:27.08 | wedhorn | mvanbaak: for info |
| 00:27.10 | wedhorn | M13524 |
| 00:27.12 | MuffinMan | [assigned] [Asterisk] Channels/chan_skinny 0013524: [patch] Fixes skinny unload reported by wedhorn (Karma: +14.25) http://bugs.digium.com/view.php?id=13524 |
| 00:28.18 | *** join/#asterisk-dev Iamnacho (i=Iamnacho@ip68-103-153-140.ks.ok.cox.net) |
| 00:33.24 | *** join/#asterisk-dev qdk (n=qdk@0x573fe57e.bynqu2.static.dsl.tele.dk) |
| 00:59.58 | *** join/#asterisk-dev mog (n=mog@74.95.48.254) |
| 00:59.58 | *** mode/#asterisk-dev [+o mog] by ChanServ |
| 01:00.37 | *** join/#asterisk-dev anthm (n=anthm@freeswitch/developer/anthm) |
| 01:47.16 | *** join/#asterisk-dev Cresl1n (n=matt@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 01:47.16 | *** mode/#asterisk-dev [+o Cresl1n] by ChanServ |
| 02:42.33 | *** join/#asterisk-dev mateo_au (i=chatzill@12.144.159.231) |
| 03:07.58 | *** join/#asterisk-dev obnauticus (n=obnautic@about/windows/regular/obnauticus) |
| 03:16.32 | mateo_au | Is anyone working on support for SIP 100rel/PRACK at the moment? |
| 03:16.47 | snuff-home | not that i know of |
| 03:17.29 | mateo_au | I checked out the latest trunk, looks like support is not there, just wondering if something might have been going on in another branch somewhere |
| 03:19.27 | snuff-home | nope if u want to check just browse the team directory on svn |
| 03:19.52 | mateo_au | ok, thanks for the tip |
| 03:23.23 | *** join/#asterisk-dev jameswf-home (n=james@ip68-2-99-240.ph.ph.cox.net) |
| 03:43.26 | *** join/#asterisk-dev Cresl1n (n=matt@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 03:43.26 | *** mode/#asterisk-dev [+o Cresl1n] by ChanServ |
| 03:45.11 | *** join/#asterisk-dev korihor (n=korihor@190.78.32.60) |
| 04:33.37 | *** join/#asterisk-dev philippel (n=p_lindhe@pool-98-111-70-106.sttlwa.fios.verizon.net) |
| 05:13.35 | *** join/#asterisk-dev denon (n=denon@tooth.decay.org) |
| 05:22.17 | *** join/#asterisk-dev Cresl1n (n=matt@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 05:22.17 | *** mode/#asterisk-dev [+o Cresl1n] by ChanServ |
| 05:52.35 | *** join/#asterisk-dev obnauticus (n=obnautic@about/windows/regular/obnauticus) |
| 05:59.25 | *** join/#asterisk-dev mog (n=mog@c-68-62-219-86.hsd1.al.comcast.net) |
| 05:59.25 | *** mode/#asterisk-dev [+o mog] by ChanServ |
| 06:01.12 | *** join/#asterisk-dev snuff-home (n=snuffy@60-240-117-7-nsw-pppoe.tpgi.com.au) |
| 06:24.11 | *** join/#asterisk-dev l0verb0y (n=l0verb0y@119.92.213.100) |
| 06:26.27 | *** join/#asterisk-dev korihor (n=korihor@190.39.165.176) |
| 07:02.27 | *** join/#asterisk-dev sucituanbo (n=free@c-24-21-121-148.hsd1.wa.comcast.net) |
| 07:48.50 | *** join/#asterisk-dev jeffg (i=h83Cb6E5@pdpc/supporter/active/jeffg) |
| 08:47.30 | *** join/#asterisk-dev friendly12345 (n=friendly@ppp59-167-79-93.lns1.mel6.internode.on.net) |
| 09:03.22 | *** join/#asterisk-dev steliosk (n=Stelios@91.140.114.193) |
| 09:21.14 | mvanbaak | wedhorn: I just uploaded a new version of your patch for skinny_unload |
| 09:21.20 | mvanbaak | it coredumped my asterisk :) |
| 09:32.16 | *** join/#asterisk-dev JT (n=j@unaffiliated/jt) |
| 10:32.50 | wedhorn | mvanbaak: ping |
| 10:33.00 | mvanbaak | pong |
| 10:33.36 | wedhorn | uploaded my skinny.conf. Forgot all about it but I'll rewrite the sample as part of the next patch |
| 10:34.14 | mvanbaak | I just wanted to know if I had the correct schema in my mind before actually trying it |
| 10:34.32 | wedhorn | ok, it should be pretty obvious from my conf |
| 10:34.39 | mvanbaak | yup |
| 10:34.51 | mvanbaak | btw, did you see my patch for the skinny_unload ? |
| 10:35.28 | mvanbaak | without that unregister_exten(l) it coredumps |
| 10:35.52 | wedhorn | yep, looks fine. I'll test tomorrow if you want, or just commit it. The changes are fairly benign (except for unloading) |
| 10:40.12 | mvanbaak | committed |
| 10:41.38 | wedhorn | thought I'd have to twist a bit harder than that ;) |
| 10:41.54 | mvanbaak | ;) |
| 10:42.51 | wedhorn | I've got some ideas to get reload working nicely, but I think I'll wait until we've finished with the config stuff |
| 10:43.17 | mvanbaak | yeah |
| 10:43.18 | mvanbaak | :) |
| 10:44.41 | mvanbaak | btw, you might want to try compiling under devmode ;) |
| 10:44.53 | wedhorn | what's devmode? |
| 10:45.08 | mvanbaak | ./configure --enable-dev-mode |
| 10:45.22 | mvanbaak | it makes compile warnings fatal errors |
| 10:45.35 | mvanbaak | chan_skinny.c:6044: warning: `oldregcontext' might be used uninitialized in this function |
| 10:46.02 | wedhorn | ok, hadn't even noticed that warning |
| 10:50.38 | wedhorn | yep, that'll need fixing. Funny cause I thought I always checked the warnings. Mind you it would have been a pain using dev-mode for the original patch given the number of globals I removed |
| 10:51.31 | mvanbaak | lol, yup |
| 10:51.56 | mvanbaak | I know. I have that from time to time as well |
| 10:52.07 | mvanbaak | but I always use it before uploading patches |
| 10:52.13 | mvanbaak | just a tip :) |
| 10:52.56 | wedhorn | yeah yeah, testing before commit is handy as well ;) |
| 10:53.33 | mvanbaak | eh ? |
| 10:53.42 | mvanbaak | what did I bork ? |
| 10:54.09 | wedhorn | nothing that I know of (yet) |
| 10:54.18 | mvanbaak | gheh |
| 10:56.41 | wedhorn | I'll try, but to be truthful, I didn't even actually compile the patch for the unload before I posted it (although that only entailed cleaning up the code) |
| 10:57.23 | mvanbaak | I tested it |
| 10:57.33 | mvanbaak | works ok |
| 10:57.50 | wedhorn | ok, ok. I tested a previous (dirty) version of it as well |
| 11:07.59 | mvanbaak | wedhorn: nat |
| 11:08.08 | mvanbaak | shouldn't that be device specific as well ? |
| 11:12.47 | mvanbaak | hhmm, it actually works ;) |
| 11:13.53 | mvanbaak | well, it compiles runs and loads/parses the config file |
| 11:22.00 | wedhorn | nat should probably be device only (unless it has rtp implications) |
| 11:22.36 | wedhorn | yeah I did actually compile and test that patch. I couldn't find any issues, but I didn't test every config option |
| 11:23.47 | mvanbaak | I have a load of options set |
| 11:23.51 | wedhorn | actually nat is on the line and sub structures, so probably better off leaving it at line |
| 11:24.02 | mvanbaak | yeah, noticed that while reading the code |
| 11:24.03 | mvanbaak | :) |
| 11:24.04 | *** join/#asterisk-dev _charly_ (i=kroseneg@sunrise.schmidham.net) |
| 11:24.26 | mvanbaak | I can test on my home box |
| 11:24.28 | wedhorn | strange. Probably another bit that need fixing |
| 11:26.07 | wedhorn | actually, it seems that nat stuff is purely rtp |
| 11:28.25 | mvanbaak | hhmm, it coredumps |
| 11:29.38 | wedhorn | ouch, any hints where |
| 11:30.11 | mvanbaak | cleanup_stale_contexts |
| 11:31.12 | wedhorn | yeah, didn't you mention that one earlier? |
| 11:32.48 | mvanbaak | ah |
| 11:32.55 | mvanbaak | I did a wget again on my home box ;) |
| 11:33.26 | mvanbaak | ah |
| 11:33.33 | mvanbaak | adding = NULL fixed it indeed |
| 11:33.51 | mvanbaak | hhmm, if I add two speeddials to a device, only 1 shows up |
| 11:34.23 | mvanbaak | but they do both show up in 'skinny show device SEP.....' |
| 11:34.33 | wedhorn | probably should set oldregcontext rather than null. |
| 11:34.40 | wedhorn | ast_copy_string(oldcontexts, regcontext, sizeof(oldcontexts)); |
| 11:34.45 | wedhorn | oldregcontext = oldcontexts; |
| 11:34.53 | wedhorn | (from the old reload_config) |
| 11:35.18 | wedhorn | maybe a counting bug |
| 11:43.58 | mvanbaak | ah well |
| 11:44.04 | mvanbaak | have to look at it later |
| 11:44.07 | mvanbaak | I'm off |
| 11:44.22 | wedhorn | yeah, I'll have a look tomorrow. cya |
| 12:08.23 | *** join/#asterisk-dev mateo_au (i=chatzill@12.144.159.231) |
| 12:47.11 | *** join/#asterisk-dev Daviey (n=dave@ubuntu/member/daviey) |
| 13:28.57 | *** join/#asterisk-dev mvanbaak_ (n=michiel@asterisk/contributor-and-bug-marshal/mvanbaak) |
| 13:28.57 | *** mode/#asterisk-dev [+o mvanbaak_] by ChanServ |
| 14:23.41 | mvanbaak | wedhorn: I fixed the issue with the speeddials |
| 14:42.12 | mvanbaak | shoot |
| 14:58.58 | *** join/#asterisk-dev jeffg (i=3GHiOgCn@pdpc/supporter/active/jeffg) |
| 15:45.46 | *** join/#asterisk-dev Cresl1n (n=matt@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 15:45.46 | *** mode/#asterisk-dev [+o Cresl1n] by ChanServ |
| 16:10.33 | *** join/#asterisk-dev anthm (n=anthm@freeswitch/developer/anthm) |
| 16:41.33 | *** join/#asterisk-dev caio1982 (i=caio1982@CAcert-br/caio1982) |
| 16:50.57 | *** join/#asterisk-dev jeffg (i=Hw8y5xKC@pdpc/supporter/active/jeffg) |
| 17:25.48 | *** join/#asterisk-dev jameswf-home (n=james@ip68-2-99-240.ph.ph.cox.net) |
| 17:29.30 | *** join/#asterisk-dev mvanbaak_ (n=michiel@asterisk/contributor-and-bug-marshal/mvanbaak) |
| 17:29.30 | *** mode/#asterisk-dev [+o mvanbaak_] by ChanServ |
| 17:32.36 | *** join/#asterisk-dev jeffg (i=spp3aKHt@pdpc/supporter/active/jeffg) |
| 18:33.28 | *** join/#asterisk-dev mvanbaak (n=michiel@asterisk/contributor-and-bug-marshal/mvanbaak) |
| 18:33.28 | *** mode/#asterisk-dev [+o mvanbaak] by ChanServ |
| 18:50.47 | *** join/#asterisk-dev jeffg (i=rs68Nkcv@pdpc/supporter/active/jeffg) |
| 19:03.42 | *** join/#asterisk-dev CunningPike (n=arodgers@S010600095b33697f.vc.shawcable.net) |
| 19:13.52 | mvanbaak | hey all |
| 19:14.26 | *** join/#asterisk-dev bbryant (n=brett@c-68-59-20-153.hsd1.sc.comcast.net) |
| 19:14.26 | *** mode/#asterisk-dev [+o bbryant] by ChanServ |
| 19:14.30 | mvanbaak | hey bbryant |
| 19:14.51 | bbryant | hey mvanbaak |
| 19:15.16 | mvanbaak | bbryant: did you have a look at the latest appdocsxml stuff ? |
| 19:15.33 | mvanbaak | it's crazy |
| 19:15.52 | bbryant | I looked at it a little bit |
| 19:15.59 | bbryant | I haven't had much time for asterisk stuff recently |
| 19:16.20 | mvanbaak | ugh, that sux |
| 19:16.26 | mvanbaak | I know how that feels |
| 19:16.49 | mvanbaak | have had a couple of weeks recently,.... took over 12 hours of work a day |
| 19:16.54 | mvanbaak | and work != asterisk |
| 19:17.35 | mvanbaak | but, monday I'll have 16 hours in planes with a loaded ipod :) |
| 19:17.38 | mvanbaak | zzzzzzzzzzzzzzzzz time |
| 19:17.48 | mvanbaak | :) |
| 19:19.40 | mvanbaak | anywayz .... |
| 19:19.50 | mvanbaak | have a look at it if you have a spare hour |
| 19:19.53 | mvanbaak | it's nice |
| 19:21.50 | tzanger | mvanbaak: I just got back from romania, that's exactly what I did |
| 19:21.56 | tzanger | mr bean, james bond, and tunes |
| 19:22.05 | tzanger | got a little APC USB battery |
| 19:22.14 | tzanger | when the phone started getting low on juice, plugged the sync cable in |
| 19:22.21 | tzanger | best $50 investment I ever made |
| 19:22.29 | tzanger | charges any USB charge-capable device |
| 19:22.38 | mvanbaak | cool |
| 19:22.44 | tzanger | claims like 50 hours additional life for an ipod |
| 19:39.56 | *** join/#asterisk-dev davevg-btwtech (n=davevg-b@nj-67-76-177-147.sta.embarqhsd.net) |
| 20:17.54 | *** join/#asterisk-dev wedhorn (n=nnnnnnnw@123-243-209-34.static.tpgi.com.au) |
| 20:24.51 | *** join/#asterisk-dev mvanbaak_ (n=michiel@asterisk/contributor-and-bug-marshal/mvanbaak) |
| 20:24.52 | *** mode/#asterisk-dev [+o mvanbaak_] by ChanServ |
| 20:24.59 | mvanbaak_ | wedhorn: !!!! |
| 20:25.51 | mvanbaak_ | wedhorn: I altered your patch a little, and now it's working in production |
| 20:26.23 | wedhorn | ok, I'll have a look |
| 20:26.52 | wedhorn | any structural changes, or just the various parameters |
| 20:27.04 | mvanbaak_ | wedhorn: I fixed that oldregcontext stuff |
| 20:27.14 | mvanbaak_ | wedhorn: and one more serious change |
| 20:27.43 | mvanbaak_ | wedhorn: while adding lines to a device, I increment the lineInstance now |
| 20:27.55 | mvanbaak_ | so the speeddials get added at instance after the lines |
| 20:28.07 | mvanbaak_ | instead of having overlapping instance numbers |
| 20:28.13 | wedhorn | yeah, I just read your note. Was that an issue prior to the patch? |
| 20:28.21 | mvanbaak_ | no |
| 20:29.06 | mvanbaak_ | in the old version, build_device would first add all the lines, incrementing lineInstance for every line, and after that add the speeddials using the last lineInstance value |
| 20:29.16 | mvanbaak_ | my patch does this in the new setup as well |
| 20:29.41 | mvanbaak_ | I just moved the line stuff above the speeddial stuff and added 'lineInstance++' to it |
| 20:30.07 | *** join/#asterisk-dev mateo_au (n=chatzill@12.144.159.231) |
| 20:30.16 | wedhorn | ah, that'd explain it, but what if you mix up lines and speeddials |
| 20:30.32 | mvanbaak_ | no problem |
| 20:30.34 | wedhorn | that wouldn't have made sense before, but it's reasonable with this patch |
| 20:31.00 | mvanbaak_ | it just increments this stuff |
| 20:31.33 | mvanbaak_ | and you use AST_LIST_INSERT_HEAD |
| 20:31.41 | mvanbaak_ | so the lines are always added on top |
| 20:31.58 | mvanbaak_ | it's just that the lineInstance counter is incremented for both lines and speeddials |
| 20:33.04 | mvanbaak_ | it doesn't matter. we do lines first, then we add speeddials |
| 20:33.55 | wedhorn | ok, if you say so. It's 0630 here, only just had my first coffee and the brain hasn't kicked into gear yet |
| 20:34.22 | wedhorn | will have a look and test it later on. |
| 20:34.23 | mvanbaak_ | it's 22:34 here |
| 20:34.27 | mvanbaak_ | cool |
| 20:34.28 | mvanbaak_ | thank you |
| 20:34.45 | mvanbaak_ | it's working with my 7960 and 7905 |
| 20:34.58 | mvanbaak_ | if you want I can upload my skinny.conf to the bug as well |
| 20:34.59 | wedhorn | there are a couple of other issue that I noted in the patch with FIXME, any thoughts on them |
| 20:35.07 | mvanbaak_ | so you can see how that is |
| 20:35.35 | wedhorn | it's ok, I've a bunch of them floating around (and I need to fix the samples) |
| 20:37.02 | mvanbaak_ | wedhorn: hhmm, making sure a line is attached to one device only eh ? |
| 20:38.12 | wedhorn | yeah, we can probably just skip it, but it'll through our instances out of whack |
| 20:38.12 | mvanbaak_ | cant we do: if (l->device) { ast_log() && bail_out() } |
| 20:38.33 | mvanbaak_ | else { l->device = d } |
| 20:39.06 | wedhorn | I don't think we want to bail out. Just log a message, don't attach the line, and continue |
| 20:39.19 | mvanbaak_ | indeed |
| 20:39.26 | wedhorn | but not attaching could have issues (that I didn't want to deal with at the time) |
| 20:39.50 | mvanbaak_ | just log a warning |
| 20:40.14 | mvanbaak_ | 'line was on another device before. now it's here. you bad admin fix your damn config' |
| 20:40.40 | mvanbaak_ | or, even better |
| 20:40.56 | mvanbaak_ | make l->device a list so a line CAN be attached to multiple devices |
| 20:40.57 | wedhorn | yeah, but now instance 1 will actually be line 2 |
| 20:42.00 | mvanbaak_ | this whole instance thing is evil if you ask me |
| 20:42.40 | wedhorn | It's the line link between the phones and asterisk. Sort of required |
| 20:42.45 | mvanbaak_ | because we never record what the last taken instance of a device lineslot is |
| 20:42.50 | mvanbaak_ | yup |
| 20:42.58 | mvanbaak_ | thats ok |
| 20:43.10 | mvanbaak_ | but the stuff we use to find the instancenumber is not ok |
| 20:43.29 | mvanbaak_ | I mean, we are adding speeddials that have the isHint flag set as next lineinstance |
| 20:43.59 | wedhorn | yeah, it's on the todo list (like everything else). The sub grabbing is related to lines and instances and needs work |
| 20:44.07 | mvanbaak_ | [2008-09-20 22:44:00] WARNING[3735]: chan_skinny.c:1497 find_line_by_instance: Could not find line with instance '3' on device 'office' |
| 20:44.28 | mvanbaak_ | that's the warning I get for the first speeddial on my 7960 |
| 20:45.04 | mvanbaak_ | I have two lines on it, and the 3rd speeddial is 'speeddial => 6002@hints,Livingroom |
| 20:45.16 | mvanbaak_ | that speeddial is giving that warning |
| 20:45.30 | wedhorn | yep, you want to see the number of warning I get because the old device basically don't send instance information (apart from line buttons) |
| 20:45.43 | mvanbaak_ | yup |
| 20:46.13 | mvanbaak_ | that's why I said the instance thing is evil as it is now |
| 20:51.55 | wedhorn | the other issue is actually trying to retain correct instancing, so if you want to connect lines to button 1,3 and 4 (eg) you can. At the moment it'd force those into 1,2 and 3. |
| 20:52.53 | mvanbaak_ | yeah |
| 20:54.10 | wedhorn | the easiest way I could see would be to have an intervening structure between the lines and devices which would hold info of instance, line and device. |
| 20:54.27 | wedhorn | Add a couple of lists in those structure and you've got shared lines ;) |
| 20:55.30 | mvanbaak_ | gheh, you might be right |
| 20:55.40 | mvanbaak_ | damn, I see no reason why that shouldn't work |
| 20:55.59 | wedhorn | it does, wrote a patch for it about 18 months ago |
| 20:56.49 | wedhorn | actually 2 years |
| 20:58.05 | mvanbaak_ | gheh |
| 21:11.17 | wedhorn | mvanbaak_: I guess it depends on whether we want to fix up the instance stuff now or later |
| 21:11.32 | wedhorn | changes would be quite extensive, albeit simple |
| 21:21.17 | *** join/#asterisk-dev jeffg (i=yED1UL6P@pdpc/supporter/active/jeffg) |
| 21:21.19 | *** join/#asterisk-dev lmadsen (n=Leif@asterisk/documenteur-extraordinaire/blitzrage) |
| 21:21.19 | *** mode/#asterisk-dev [+o lmadsen] by ChanServ |
| 21:56.51 | *** join/#asterisk-dev kamanashisroy (n=kamanash@119.30.34.3) |