| 02:38.59 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 02:47.32 | *** join/#asterisk-dev leedm777 (~leedm777@50-81-28-152.client.mchsi.com) |
| 02:59.37 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 03:52.11 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 04:10.43 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 04:57.37 | *** join/#asterisk-dev babak (uid19622@gateway/web/irccloud.com/x-tismpckjeyvotixo) |
| 05:09.36 | *** join/#asterisk-dev doome_ (~doome@82.150.48.146) |
| 06:14.33 | *** join/#asterisk-dev infobot_ (ibot@rikers.org) |
| 06:14.33 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Hell hath frozen over: Asterisk is moving to Git -=- Follow on Twitter at @AsteriskDev |
| 06:59.01 | *** join/#asterisk-dev doome_ (~doome@82.150.48.146) |
| 07:08.14 | *** join/#asterisk-dev pchero_work (~pchero@109.70.54.56) |
| 09:29.39 | *** join/#asterisk-dev mjordan (~mjordan@153.97.17.224) |
| 09:29.39 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 10:10.53 | *** part/#asterisk-dev mjordan (~mjordan@153.97.17.224) |
| 10:23.13 | *** join/#asterisk-dev mjordan (~mjordan@153.97.17.224) |
| 10:23.13 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 10:23.18 | mjordan | file: ohai |
| 10:23.23 | file | mjordan, ohai2u |
| 10:23.39 | mjordan | how goes it? |
| 10:23.46 | file | 9 |
| 10:23.48 | file | you? |
| 10:23.51 | mjordan | about the same |
| 10:24.09 | mjordan | alas, I couldn't get my original idea for a dangerous demo to work |
| 10:24.17 | mjordan | jssip was not a fan of me sending it a refer request |
| 10:24.32 | file | not all ideas can become tangible and real |
| 10:24.36 | mjordan | indeed. |
| 10:24.52 | file | but there is always ice cream |
| 10:25.00 | mjordan | mmm |
| 10:25.02 | mjordan | ice cream. |
| 10:25.06 | mjordan | I should have that. |
| 10:26.41 | skrusty | :) |
| 10:47.22 | *** join/#asterisk-dev doome_ (~doome@82.150.48.146) |
| 11:41.13 | *** join/#asterisk-dev mjordan (~mjordan@153.97.17.224) |
| 11:41.13 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 12:01.50 | *** join/#asterisk-dev babak (uid19622@gateway/web/irccloud.com/x-vbzhoewgentpixbt) |
| 12:12.33 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@216-191-106-163.dedicated.allstream.net) |
| 12:23.07 | *** join/#asterisk-dev sgriepentrog (sgriepentr@nat/digium/x-yanrkpdefvfibibm) |
| 12:59.13 | *** join/#asterisk-dev leedm777 (leedm777@nat/digium/x-mtbuberosssarvje) |
| 13:28.40 | *** join/#asterisk-dev sgalarneau (Adium@nat/digium/x-zyrqtqusqlprsxyi) |
| 13:34.49 | *** join/#asterisk-dev newtonr (RustyNewto@nat/digium/x-lzqplhjoqrylqulq) |
| 13:34.49 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 14:00.34 | *** join/#asterisk-dev jsmith (~jsmith@fedora/jsmith) |
| 14:00.34 | *** mode/#asterisk-dev [+o jsmith] by ChanServ |
| 14:07.36 | *** join/#asterisk-dev GMLudo (~GMLudo@srv-05.eyepea.eu) |
| 14:12.14 | GMLudo | newtonr: Hi, I've just discussed with Matt Jordan during KamailioWorld to add a wiki page to list AMI/FastAGI/AGI libraries, like you did for ARI: https://wiki.asterisk.org/wiki/pages/viewpage.action?pageId=29395573#AsteriskRESTInterface(ARI)-ARILibraries |
| 14:12.58 | GMLudo | he was agree with the idea and he suggested me to ping you on IRC, it's because I'm here ;-) |
| 14:13.42 | GMLudo | I told me that an empty page it's ok, users can push comments on this page |
| 14:15.11 | newtonr | GMLudo, yeah I'll throw up a page and you can just comment on it. Then later if you decide you want to contribute regularly we can get you access to edit. |
| 14:15.40 | *** join/#asterisk-dev putnopvut (putnopvut@asterisk/master-of-queues/mmichelson) |
| 14:15.40 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:34.18 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 14:34.28 | GMLudo | newtonr: to be really honest, I'm one of developer a library based on Python-AsyncIO ;-) However, I've used a lot of libraries in the past, I'll add a comment with all librairy I know |
| 14:35.23 | GMLudo | newtonr: If you want to give me a wiki access, I'm interested in, however, with comments, it's enough, I won't have a lot of edit changes |
| 14:41.18 | *** join/#asterisk-dev w9sh (~sph@50-79-224-193-static.hfc.comcastbusiness.net) |
| 14:43.29 | *** join/#asterisk-dev mjordan (~mjordan@153.97.17.224) |
| 14:43.29 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 14:43.38 | *** join/#asterisk-dev sruffell (~sruffell@asterisk/the-kernel-guy/sruffell) |
| 14:43.38 | *** mode/#asterisk-dev [+o sruffell] by ChanServ |
| 14:44.04 | nbjoerg | mjordan? |
| 14:49.07 | *** part/#asterisk-dev w9sh (~sph@50-79-224-193-static.hfc.comcastbusiness.net) |
| 14:50.30 | newtonr | GMLudo, For quality control we try to avoid giving out access until someone is committed to contributing frequently or being a maintainer for certain documentation. Otherwise it's easier to use comments for one-off additions and suggestions. |
| 14:50.47 | newtonr | GMLudo, you can comment here https://wiki.asterisk.org/wiki/display/AST/AMI+Libraries |
| 14:50.57 | newtonr | and here: https://wiki.asterisk.org/wiki/pages/viewpage.action?pageId=32375589 |
| 14:58.05 | gtjoseph | leedm777: mjordan committed 541 and 542 but i'd still like you to take a look and make sure i didn't do something stupid to stasis. |
| 15:12.39 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-kshkeihrrtspsjmi) |
| 16:08.48 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 16:12.22 | coreyfarrell | i'm getting a reference leak on many tests that seems to be from res_sorcery_memory. can anyone who understands sorcery check out the one leak from tests/rest_api/channels/redirect/off-nominal/ - http://pastebin.ubuntu.com/11435035/ |
| 16:15.28 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 16:20.40 | *** join/#asterisk-dev doome_ (~doome@apn-94-44-255-74.vodafone.hu) |
| 16:22.18 | *** join/#asterisk-dev ipengineer (~zconkle@static-71-252-134-63.dllstx.fios.verizon.net) |
| 16:57.15 | *** join/#asterisk-dev superscrat (asanders@nat/digium/x-mepcdnkhcfhfnmtz) |
| 17:31.49 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 18:18.59 | gtjoseph | WTF is 'yappcap' in the testsuite? |
| 19:08.20 | coreyfarrell | gtjoseph: it's a python wrapper for libpcap and used by lib/python/pcap_listener.py |
| 19:08.30 | coreyfarrell | gtjoseph: https://github.com/otherwiseguy/yappcap |
| 19:08.57 | gtjoseph | yeah, the testsuite wiki needs to be updated |
| 19:09.07 | gtjoseph | and apparently the testsuite now needs to be run as root |
| 19:09.14 | coreyfarrell | why? |
| 19:09.19 | gtjoseph | if you want to test all the subscription stuff |
| 19:09.23 | coreyfarrell | I don't think I've ever run it as root |
| 19:09.31 | gtjoseph | i get a persmissions error in pcap |
| 19:09.44 | coreyfarrell | which test? |
| 19:10.07 | gtjoseph | all of the subscriptio tests... tests/channels/pjsip/subscriptions/rls/lists/nominal/presence/batched/basic for instance |
| 19:10.55 | coreyfarrell | trying it out now |
| 19:11.27 | coreyfarrell | oh, I don't have yappcap or pyxb |
| 19:11.34 | gtjoseph | :) |
| 19:12.02 | gtjoseph | i didn't even know about them until bamboo complained of a whole bunch of tests |
| 19:12.23 | gtjoseph | failing |
| 19:12.31 | coreyfarrell | so it might be that those tests require 'rawsocket', or maybe this is a different permission. could be that it requires the same permission as 'tcpdump' requires. |
| 19:12.56 | gtjoseph | maybe. im in the tcpdump group |
| 19:15.03 | coreyfarrell | gtjoseph: what do you get if you run: python -c 'import socket;socket.socket(socket.AF_PACKET, socket.SOCK_RAW).close()' |
| 19:15.22 | coreyfarrell | on my system that is silent as root, but gives a permissions error as non-root. |
| 19:15.34 | gtjoseph | socket.error: [Errno 1] Operation not permitted |
| 19:16.40 | coreyfarrell | hard to say without fully tracing out what yappcap is trying to do, could be that it also requires rawsocket. |
| 19:16.53 | gtjoseph | i think so |
| 19:16.58 | coreyfarrell | which I'm not sure how to grant permission to non-root users for rawsocket. |
| 19:17.02 | gtjoseph | i'll investigate |
| 19:17.05 | gtjoseph | tomorrow |
| 19:34.10 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 19:40.00 | file | the RLS tests concern me less, this unit test failure though |
| 19:41.11 | file | methinks the endpoint stasis change did it |
| 19:43.36 | gtjoseph | file: which unit test? |
| 19:43.44 | file | https://bamboo.asterisk.org/bamboo/browse/AST-ATRUNKUNIT-1062/test |
| 19:44.45 | gtjoseph | hmmm. i ran the unit tests before submitting the patch and they passed. let me check again |
| 19:45.03 | file | I just ran it locally and it did not |
| 19:47.55 | gtjoseph | damn |
| 19:48.35 | gtjoseph | i think when i ran the tests the test_stasis_endpoint module wasn't loaded |
| 19:52.14 | gtjoseph | file: yep, that was me. you might want to revert those 2 commits: reviews 545 and 546 |
| 19:52.54 | file | um |
| 19:53.32 | gtjoseph | actually you want to approve and merge 545 and 546 to do the revert |
| 19:53.55 | file | yeah, uno sec |
| 19:54.27 | file | done |
| 19:54.53 | gtjoseph | cool. this is why i wanted leedm777 to look at these before they were committed. matt jumped the gun. :) |
| 19:55.37 | file | he's in THE FUTURE |
| 20:01.53 | *** join/#asterisk-dev mjordan (~mjordan@81.92.23.58) |
| 20:01.53 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 20:05.53 | file | speak of the devil |
| 20:06.12 | gtjoseph | yeah, it's all his fault. |
| 20:08.10 | coreyfarrell | anyone have advice for how to troubleshoot a res_sorcery_memory reference leak? sorry to pester, I'm trying to test a change that needs to be verified against the whole system with REF_DEBUG enabled and this leak was the only cause of about 100 failures. now posted to https://issues.asterisk.org/jira/browse/ASTERISK-25141 |
| 20:10.08 | file | res_sorcery_memory is a very thin wrapper over ao2, it's more likely that somehow the user of it is not doing something |
| 20:10.18 | gtjoseph | coreyfarrell: the issue is more likely to be in pjsip_options |
| 20:10.27 | gtjoseph | i can look at it |
| 20:11.11 | gtjoseph | coreyfarrell: is there a specific test suite test that demonstrates the leak? |
| 20:11.25 | coreyfarrell | yes.. tests/rest_api/channels/redirect/off-nominal/ |
| 20:11.34 | gtjoseph | let me look |
| 20:11.55 | coreyfarrell | in that case it is an object created by pjsip_options.c (a CONTACT). |
| 20:11.57 | gtjoseph | asterisk 13 or master? |
| 20:12.11 | coreyfarrell | master. possibly 13 but I haven't tested that |
| 20:12.28 | gtjoseph | master it is. looking. |
| 20:13.33 | gtjoseph | while i'm setting up, could you try /tests/channels/pjsip/qualify and see if the leaks are there? |
| 20:13.48 | coreyfarrell | trying it now |
| 20:13.50 | mjordan | file: did I do something? |
| 20:14.19 | coreyfarrell | gtjoseph: any specific one of those tests? |
| 20:14.21 | file | mjordan, gtjoseph didn't want the cache changes to go in yet until dlee looked, and they ended up breaking the unit test at the very least, so he asked for them to be reverted and they were |
| 20:14.33 | mjordan | ah. Sorry |
| 20:14.43 | gtjoseph | coreyfarrell: basic or max_timeout |
| 20:14.48 | file | mjordan, squirrel. |
| 20:14.52 | mjordan | indeed. |
| 20:15.09 | file | mjordan, when are you stateside again? |
| 20:15.13 | mjordan | Sunday. |
| 20:15.27 | file | fancy |
| 20:15.32 | mjordan | This trip has knocked me for a loop. Can't seem to stay well, and the time changes have got my head pretty wonky |
| 20:16.13 | coreyfarrell | gtjoseph: basic leaks, object created by pjsip_options for that one too |
| 20:16.18 | file | drugs mjordan |
| 20:16.27 | gtjoseph | coreyfarrell: ok |
| 20:16.32 | mjordan | yeah, I'm going to get hopped up on nyquil in a bit |
| 20:17.01 | *** join/#asterisk-dev Marquis42 (~mrbmw@d14-69-246-193.try.wideopenwest.com) |
| 20:17.48 | coreyfarrell | hmm, I just tried tests/channels/pjsip/qualify/auth, I got: WARNING: Memory corrupted after free of 0x1dc5b70 allocated at sorcery.c __ast_sorcery_open() line 760 |
| 20:19.21 | gtjoseph | weird. no issues on that one othe rthan the leak |
| 20:19.27 | coreyfarrell | before I forget, can someone with bamboo admin access take a look at the config for the reference check job for master? I tried looking the other day and it is broken (I think since the migration to git). |
| 20:21.15 | coreyfarrell | gtjoseph: so all of tests/channels/pjsip/qualify produce the same leak. that failure I got from auth didn't happen again, must be a race condition :( |
| 20:21.42 | gtjoseph | yep. i'm on it. |
| 20:22.49 | gtjoseph | i'll pull that issue and change it to pjsip_options |
| 20:23.27 | coreyfarrell | pjsip_configuration is the other user of res_sorcery_memory.. so if it's an issue with pjsip_options it's probably an issue with config too |
| 20:23.47 | gtjoseph | i don't think it has anything to do with res_sorcery_memory. |
| 20:24.17 | gtjoseph | i think it has to do with me not cleaning something up correctly in pjsip_options. |
| 20:28.13 | coreyfarrell | gtjoseph: res_pjsip/location.c:permanent_uri_handler looks suspicious: if (!ast_res_pjsip_find_or_create_contact_status(contact)) { |
| 20:28.33 | gtjoseph | looking |
| 20:32.31 | gtjoseph | coreyfarrell: not it |
| 20:34.32 | *** part/#asterisk-dev mjordan (~mjordan@81.92.23.58) |
| 20:37.27 | gtjoseph | coreyfarrell: oops, yes it was. |
| 20:41.25 | coreyfarrell | gtjoseph: so res_pjsip/location.c seems to be the cause of tests/rest_api/channels/redirect/off-nominal/ - but I'm still getting the leak from tests/channels/pjsip/qualify/basic/ |
| 20:41.38 | gtjoseph | not me |
| 20:41.45 | gtjoseph | just passed all 5 tests |
| 20:42.05 | gtjoseph | it's not res_odbc is it? |
| 20:42.34 | gtjoseph | i still get that so i don't compile that module |
| 20:42.38 | coreyfarrell | I don't build res_odbc |
| 20:42.45 | gtjoseph | hmmm |
| 20:42.55 | gtjoseph | how did you fix the leak? |
| 20:43.14 | coreyfarrell | i'll post my patch and the refs.txt from qualify/basic |
| 20:44.04 | gtjoseph | http://pastebin.com/27HVAR5j |
| 20:45.29 | coreyfarrell | I just posted my patch to JIRA, effectively same thing I just did without RAII |
| 20:46.51 | gtjoseph | hmmm. are you SURE you're testing the new code? |
| 20:46.54 | gtjoseph | :) |
| 20:47.51 | coreyfarrell | well I'm sure, I'll delete and reinstall to be SURE :) |
| 20:48.15 | gtjoseph | wait, this is really weird |
| 20:48.36 | gtjoseph | i just ran again and auth failed but the rest passed. then i ran again and all 5 passed |
| 20:50.00 | coreyfarrell | yeah that happened to me before with auth, I got the memory corruption 'warning' on one run, it passed on the next run |
| 20:50.47 | gtjoseph | well, they're all passing consistently now. let me take out the RAII_VAR and add the -1. |
| 20:52.09 | coreyfarrell | I just pull the latest master, only thing I was missing is the revert of your stasis changes.. unfortunately it looks like that's going to force a full rebuild so it'll be a bit before I can retest |
| 20:53.10 | gtjoseph | since the qualify_frequency is set low to test, it's possible that there's another qualify in progress when asterisk shuts down. |
| 20:53.48 | gtjoseph | how long does it take to run the single test? |
| 20:55.32 | coreyfarrell | pjsip/qualify/basic took 28 seconds |
| 20:55.49 | gtjoseph | holy moly!! it takes only 6 seconds for me |
| 20:55.57 | gtjoseph | i think that's the problem |
| 20:56.12 | coreyfarrell | 6 seconds with ref_debug enabled? |
| 20:56.16 | gtjoseph | yep |
| 20:56.29 | coreyfarrell | erm, qualify/basic just passed for me |
| 20:56.44 | gtjoseph | yeah, i think it's a timing thing |
| 20:57.19 | coreyfarrell | only thing I did was pull the stasis reverts.. i'm going to try running it a few more times.. |
| 20:57.33 | gtjoseph | ok, run all the qualify tests |
| 20:59.02 | coreyfarrell | running them.. will take a few minutes |
| 21:00.39 | coreyfarrell | ok so qualify/basic has passed 3 out of 4 times, you're probably right about it being a timing issue |
| 21:01.26 | coreyfarrell | qualify/auth failed (with 3 references to the contact outstanding), the rest of the tests passed |
| 21:02.12 | gtjoseph | i'm certain that's because the test takes long enough to run that more qualifies are scheduled |
| 21:02.25 | gtjoseph | and are pending |
| 21:02.47 | coreyfarrell | I wounder why my desktop runs them so slow |
| 21:03.03 | gtjoseph | hmmm. |
| 21:03.22 | *** join/#asterisk-dev newtonr (RustyNewto@nat/digium/x-xwdlsljrrlskhqop) |
| 21:03.39 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 21:03.52 | gtjoseph | i do have an 6 core/12 thread ssd based machine with 64gb of memory. :) |
| 21:04.03 | coreyfarrell | AH. that explains it |
| 21:04.39 | coreyfarrell | I have a quad core Xeon 3.0ghz with 4GB ram and an old fashioned HD :( |
| 21:05.05 | coreyfarrell | since this is my desktop 4GB of ram means 2GB available |
| 21:05.29 | gtjoseph | try and convince matt to give you a new dev machine at Astricon this year instead of another Cubox. :) |
| 21:05.58 | file | matt did not control that exactly :P |
| 21:15.24 | coreyfarrell | gtjoseph: so how long does it take your system to run the complete testsuite? takes me about 8.5 hours with ref_debug enabled, the dependencies I have allow 666 tests to run |
| 21:16.24 | gtjoseph | about 2 hours |
| 21:16.48 | gtjoseph | same number of tests |
| 21:17.31 | coreyfarrell | how about with ref_debug disabled? is there a big difference on your system? |
| 21:17.46 | gtjoseph | haven't compared |
| 21:18.06 | gtjoseph | i can try the pjsip stack now both ways |
| 21:21.05 | coreyfarrell | so I just tried pjsip/qualify/basic without REF_DEBUG, 12 seconds.. so asterisk is probably getting slowed down by my HD, but also I think refcounter.py is causing my system to swap |
| 21:21.36 | gtjoseph | yeah, 12 seconds is still long enough for another qualify to get scheduled |
| 21:22.02 | gtjoseph | and if sipp is already gone, it'll hang there until shutdown |
| 21:24.24 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 21:24.46 | gtjoseph | coreyfarrell: just uploaded the leak fix reviews |
| 21:25.02 | coreyfarrell | I just tested with REF_DEBUG enabled but the run of refcounter.py from testsuite disabled.. 20 seconds. so 8 seconds overhead to produce the refs log, 10 seconds more to process it. |
| 21:25.06 | coreyfarrell | gtjoseph: checking it out |
| 21:27.07 | coreyfarrell | gtjoseph: so that patch will fix it, but do we really need/want RAII_VAR? it does add overhead.. |
| 21:27.30 | gtjoseph | i can put the -1s in instead. |
| 21:27.54 | gtjoseph | usually if there are more than 1 exit points, i'll use RAII_VAR |
| 21:28.46 | gtjoseph | up to you. let me know. |
| 21:29.42 | coreyfarrell | I think contact was fine (no need to touch it), for status there is only 1 exit point - after "if (!status) {}" |
| 21:29.56 | gtjoseph | i'll change em. give me a sec. |
| 21:30.34 | coreyfarrell | while you are there do you want to put the assignment of status on it's own line? |
| 21:30.58 | gtjoseph | is that a new policy? :) |
| 21:31.45 | coreyfarrell | I'm not aware of any policy.. I just think assignments are harder to read/spot when they're inside if conditions. feel free to ignore that comment I will still +1 |
| 21:33.35 | gtjoseph | done |
| 21:34.13 | coreyfarrell | and +1'ed |
| 21:34.19 | gtjoseph | cool |
| 21:34.32 | gtjoseph | all we need is for file to merge. |
| 21:34.36 | coreyfarrell | thanks for helping me with this gtjoseph! |
| 21:34.43 | gtjoseph | no worries |
| 21:34.47 | gtjoseph | good catch |
| 21:35.20 | coreyfarrell | glad you were here to point me to the right spot.. I was trying to figure out what res_sorcery_memory was doing wrong :) |
| 21:35.24 | gtjoseph | looks like there are other leaks in config_wizard so i'll look at those this weekend |
| 21:35.40 | gtjoseph | sorcery never does anything wrong. :) |
| 21:35.54 | coreyfarrell | funny man |
| 21:37.14 | gtjoseph | now i've got to figure out what i did to stasis. :( |
| 21:37.26 | coreyfarrell | you broke it :P |
| 21:38.01 | gtjoseph | quite true. |
| 22:11.24 | *** join/#asterisk-dev jameswf (uid27319@gateway/web/irccloud.com/x-jhbkfogaubveogeg) |
| 23:35.46 | *** join/#asterisk-dev superscrat (asanders@nat/digium/x-zbanrwtvtnlvvykl) |
| 23:44.15 | *** join/#asterisk-dev superscrat (asanders@nat/digium/x-ttvuhaqqxkfbxdar) |
| 23:52.39 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |