| 00:22.42 | *** join/#asterisk-dev acidfu_ (~acidfoo@69-165-160-113.dsl.teksavvy.com) |
| 04:35.16 | *** join/#asterisk-dev babak (uid19622@gateway/web/irccloud.com/x-bnvzbliztpplhgdh) |
| 05:07.40 | *** join/#asterisk-dev bof22 (~Thunderbi@185.13.183.107) |
| 06:46.03 | *** join/#asterisk-dev lumasepa (~gestoip@193.145.124.30.local.ull.es) |
| 06:47.59 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 06:55.45 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 06:55.45 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 07:26.16 | *** join/#asterisk-dev pchero_work (~pchero@109.70.54.56) |
| 07:37.06 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 07:37.06 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 08:00.22 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 08:00.22 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 08:27.45 | *** join/#asterisk-dev lumasepa (~gestoip@193.145.124.30.local.ull.es) |
| 10:11.26 | *** join/#asterisk-dev snuff-work (~blatus@203.87.61.22) |
| 10:11.26 | *** mode/#asterisk-dev [+o snuff-work] by ChanServ |
| 11:34.20 | *** join/#asterisk-dev tuxd00d (~tuxd00d@ip72-201-239-241.ph.ph.cox.net) |
| 11:39.51 | *** join/#asterisk-dev acidfoo_ (~acidfoo@modemcable002.114-70-69.static.videotron.ca) |
| 12:26.23 | *** join/#asterisk-dev bof23 (~Thunderbi@185.13.183.107) |
| 12:43.35 | *** join/#asterisk-dev [TK]D-Fender (~Joe@216-191-106-165.dedicated.allstream.net) |
| 13:13.21 | *** join/#asterisk-dev ccc_ (50186c15@gateway/web/freenode/ip.80.24.108.21) |
| 13:35.54 | *** join/#asterisk-dev cresl1n (Adium@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 13:35.54 | *** mode/#asterisk-dev [+o cresl1n] by ChanServ |
| 14:15.52 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-aipwduhukwzbiovq) |
| 14:23.51 | *** join/#asterisk-dev cresl1n (Adium@asterisk/libpri-and-libss7-expert/Cresl1n) |
| 14:23.51 | *** mode/#asterisk-dev [+o cresl1n] by ChanServ |
| 14:31.23 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-yjqqfygrwowdxfxq) |
| 15:08.38 | *** join/#asterisk-dev hehol (~hehol@gatekeeper.loca.net) |
| 15:34.27 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 15:34.27 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 15:39.57 | *** join/#asterisk-dev newtonr (RustyNewto@nat/digium/x-xbyraetauecvtity) |
| 15:39.57 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 16:15.16 | *** join/#asterisk-dev jkroon (~jkroon@uls-154-73-35-201.wall.uls.co.za) |
| 16:16.48 | jkroon | hi all, after spending the day trying to get HANGUPCAUSE and HANGUPCAUSE_KEYS to work I realised that chan_sip sets those variables on the channel created by Dial() and not the requesting channel. |
| 16:17.14 | jkroon | http://pastie.org/10793630 shows my dialplan and the resulting CLI output |
| 16:18.18 | jkroon | http://pastie.org/10793630 - the prov-hangup is based off the example at https://wiki.asterisk.org/wiki/display/AST/Hangup+Cause |
| 16:19.41 | jkroon | I added some debug ast_verbose code to asterisk in chan_sip to log on which channel exactly the cause is being set in order to confirm, and it's definitely on the channel created by Dial() and not the requesting channel. |
| 16:25.55 | jkroon | my questions as follows: |
| 16:26.12 | jkroon | 1. Is Dial() supposed to copy the cause codes from the created channel after it failed? |
| 16:26.25 | jkroon | 2. Is chan_sip supposed to add it to the requesting channel instead? |
| 16:26.38 | jkroon | 3. Is there some other way to get to it? |
| 16:30.13 | file | 1. I believe so, it is involved 2. No - it doesn't have access to that at that point 3. Dunno |
| 16:30.16 | file | rmudgett, thoughts? |
| 16:30.21 | file | cresl1n, or heck why not you |
| 16:31.01 | jkroon | file, 2 - it should have via p->owner which should have a requester. |
| 16:31.16 | jkroon | 1 - would you perhaps mind helping to see if you can locate it? |
| 16:31.45 | file | a requesting channel isn't stored by another channel |
| 16:34.58 | *** part/#asterisk-dev ccc_ (50186c15@gateway/web/freenode/ip.80.24.108.21) |
| 16:35.10 | file | I believe it's conveyed using a control frame with a type of AST_CONTROL_PVT_CAUSE_CODE back to the calling channel, and then when that is indicated using ast_indicate_data it is stored |
| 16:35.14 | jkroon | ast_request takes a requestor channel? |
| 16:35.36 | file | it does however stuff doesn't store that |
| 16:35.48 | file | it's used as part of the channel request process to fill in parts or create a loose association |
| 16:35.53 | jkroon | file, but that's the normal HANGUPCAUSE, not the SIP cause code (ie, I am looking for the 604 or 404 response in this case, not just a value of 1) |
| 16:36.19 | file | the PVT_CAUSE_CODE is for conveying the protocol specific |
| 16:45.07 | file | I have a feeling it's actually a chan_local bug |
| 16:45.26 | jkroon | hmm, why do you say that? |
| 16:46.10 | file | because if I put a verbose on the calling leg before the Local channel I get the info |
| 16:46.21 | file | if I put it on the dialplan being executed by the ;2 Local leg I don't |
| 16:46.35 | file | so it's probably passing the frame completely back |
| 16:50.31 | jkroon | ooh, ok, that makes sense. |
| 16:50.55 | jkroon | let me see if I can confirm. |
| 16:53.24 | file | what version is this? |
| 16:53.27 | jkroon | -- Executing [0845158255@uls-from-trunk:10] NoOp("IAX2/jkroon-15190", "Cause keys: SIP/IS-00000002") in new stack |
| 16:53.34 | jkroon | 11.21.1 |
| 16:53.41 | file | k, so not a regression in 13 |
| 16:53.46 | file | (I did it on 13 using PJSIP) |
| 16:54.13 | jkroon | ok, to be honest, i'm more than happy to file a bug report but I don't even know what to say ... |
| 16:54.52 | file | Hangup causes are passed through Local channels, but are not accessible on the Local channels themselves |
| 16:55.15 | opticron | jkroon, were the channels ever connected? |
| 16:55.36 | jkroon | opticron, the created channel hangs up with a SIP 604 |
| 16:56.02 | opticron | iirc, app_dial doesn't do any copying of that information from channel to channel |
| 16:56.27 | file | the protocol hangup causes stuff uses a control frame to convey it |
| 16:56.32 | file | so it is accessible |
| 16:56.34 | jkroon | file, where do I write the required code? |
| 16:56.43 | file | jkroon, I don't understand the question |
| 16:57.01 | jkroon | if i'd like to write a patch, where do I start? |
| 16:57.09 | opticron | the typical use case is hangup handlers to be executed on dial-created channels where you then read the information after the fact |
| 16:57.11 | file | I don't know where the problem is, so I don't know |
| 16:57.24 | file | it's either in the core or the chan_local stuff... |
| 16:57.45 | jkroon | opticron, the SIP sequence is basically me:INVITE, prov:100 Trying, prov:604 Doesn not exist anywhere ... I'm after the 604. |
| 16:58.59 | jkroon | opticron, ok, so here is the use case: I run Dial(SIP/IS/0845158255), and afterwards if it failed with HANGUPCAUSE=1 (NOTFOUND) I need to distinqguish between a 404 and a 604 response. |
| 17:02.14 | jkroon | file, looks like Dial() gets the frame and uses ast_indicate_data to forward it on, being a chan_local channel it will probably pass it straight on. |
| 17:02.25 | file | potentially! |
| 17:03.30 | jkroon | so now the question becomes, does other channels all have to specifically deal with it? |
| 17:03.52 | jkroon | shouldn't app_dial() just deal with it be adding it to the in channel? |
| 17:05.15 | jkroon | ok, so looking at chan_iax2 (which is just a lot less complicated than chan_sip) it deals with it by storing it. |
| 17:06.16 | jkroon | it's simpler to deal with it in chan_local probably - but is that the right fix? |
| 17:06.20 | file | there is common logic in the core for storing it in the hangupcause stuff |
| 17:07.06 | file | it is only executed if the channel driver did nothing with it |
| 17:07.15 | file | if chan_local returns success, then the logic isn't executed, and it's not stored |
| 17:07.27 | jkroon | ok, so add to chan_local? |
| 17:07.35 | file | add what? |
| 17:08.06 | file | if it returns -1 when told about pvt_cause_code then it'll be accessible on the Local channel |
| 17:08.23 | jkroon | the logic to call ast_channel_hangupcause_hash_set when it tunels an AST_CONTROL_PVT_CAUSE_CODE - sorry - I should be more clear |
| 17:08.51 | file | that would work |
| 17:09.20 | jkroon | so if a AST_CONTROL_PVT_CAUSE_CODE gets sent to local_indiate - first call ast_channel_hangupcause_hash_set before following current logic - although - is passing it through the correct logic? |
| 17:09.46 | file | changing the fact it's passing it through would be a drastic behavior change |
| 17:09.48 | jkroon | eg, in my example, IAX2/incoming gets bridged with Local/...;1 and Local...;2 dials creating SIP/IS/foo |
| 17:10.09 | jkroon | so now IAX2/incoming ends up with the SIP/IS... channel in it's hangupcauses. |
| 17:10.26 | jkroon | I agree it's a drastic change, and most likely it's not something that should happen at this stage. |
| 17:11.55 | jkroon | just found some of my really old chan_local bugs trying to check for dupes |
| 17:29.10 | jkroon | file - does this look sensible: https://issues.asterisk.org/jira/browse/ASTERISK-25912 |
| 17:30.04 | file | yes |
| 17:45.42 | jkroon | file, ok, so if I understand this I just basically need to trigger a return -1 at the end of the function? |
| 17:45.55 | file | if it's a pvt_cause_code I believe so, yes |
| 17:45.57 | file | or you can do the above |
| 17:46.04 | file | with calling ast_channel_hangupcause_hash_set |
| 17:46.06 | file | they do the same thing |
| 17:46.47 | jkroon | chan_sip returns -1, as does chan_iax2 (I was looking at the wrong path earlier) |
| 17:47.09 | jkroon | btw, chan_iax2 does filter it from being transmitted due to being "local information only" |
| 17:47.45 | file | yes |
| 17:48.13 | jkroon | so my question now - is it OK to both forward the frame as well as return -1? I note that AST_CONTROL_RINGING and AST_STATE_RING is blocked already by chan_local. |
| 17:48.36 | file | yes, it's fine |
| 17:48.40 | jkroon | ok, done. |
| 17:49.34 | jkroon | http://pastie.org/10793811 - is it really that simple? |
| 17:49.49 | file | maybe, yes |
| 17:49.57 | jkroon | proceeds to testing. |
| 18:08.50 | rmudgett | jkroon: That patch isn't going to get your -1 returned. You would need to put that check after the local_queue_frame(). |
| 18:09.58 | jkroon | rmudgett, that explains it. well spotted thanks. |
| 18:10.02 | jkroon | is the logic OK? |
| 18:10.12 | jkroon | other than that glaring right at me stupidity. |
| 18:11.05 | rmudgett | Yes. Just move your new lines to after the local_queue_frame() call. |
| 18:12.26 | jkroon | http://pastie.org/10793844 - better? |
| 18:13.58 | rmudgett | yes. You don't need to reference the ASTERISK issue though. |
| 18:14.15 | jkroon | i'll remove that then. |
| 18:15.12 | jkroon | ok, so it's simple enough to upload the patch file to the issue tracker - how do I get it to reviewboard? |
| 18:15.23 | file | we use gerrit these days |
| 18:15.33 | jkroon | ok, onto gerrit then |
| 18:15.34 | file | https://wiki.asterisk.org/wiki/display/AST/Patch+Contribution+Process |
| 18:16.05 | file | https://wiki.asterisk.org/wiki/display/AST/Gerrit+Usage for the full deets |
| 18:16.18 | jkroon | thanks |
| 18:28.47 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-179-40-172.cust.bezeqint.net) |
| 18:28.47 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 18:29.21 | jkroon | could anybody assist with login on gerrit? |
| 18:30.25 | file | what's the problem? |
| 18:30.46 | jkroon | jaco@uls.co.za - credentials are failing. |
| 18:30.55 | file | have you set a username in gerrit? |
| 18:30.57 | jkroon | and the reset email isn't getting to my mail server. |
| 18:31.06 | jkroon | can't log in on web interface yet. |
| 18:31.16 | file | have you submitted a license agreement? |
| 18:31.23 | jkroon | couple of years ago already |
| 18:32.04 | jkroon | https://issues.asterisk.org/jira/secure/DigiumLicense.jspa says https://issues.asterisk.org/jira/secure/DigiumLicense.jspa |
| 18:32.28 | file | are you logging into crowd oauth using your JIRA username and password? |
| 18:32.40 | jkroon | trying to yes |
| 18:32.52 | file | using jkroon as the username? |
| 18:33.10 | jkroon | jaco@uls.co.za ... let me retry |
| 18:33.17 | file | your username is jkroon |
| 18:33.22 | file | not the jaco@uls.co.za |
| 18:33.35 | jkroon | once more i am in your debt |
| 18:54.55 | *** join/#asterisk-dev fornax (~fornax@85.183.43.221) |
| 18:56.23 | jkroon | file, am I correct in saying chan_local no longer exists on trunk and that i'll need to replace that mechanism moving to ast-13? |
| 18:56.32 | file | yes. |
| 18:57.16 | jkroon | origin/11 <-- so that's the branch to fork and patch on? |
| 18:58.15 | file | for 11, yes |
| 18:59.44 | jkroon | well, the code will obviously not apply for trunk or 13, so it has to go to 11 (in this case anyway) |
| 19:01.34 | jkroon | ok, so I did the clone, and I then checked out the 11 branch and created a local branch. |
| 19:01.58 | jkroon | normally I'd commit directly and then push but it seems I should use git review here instead of straight commit? |
| 19:02.22 | file | you commit it to your local branch |
| 19:02.31 | file | and then use git review, which pushes your commit up for review |
| 19:03.26 | rmudgett | Your change would go into either main/core_local.c or main/core_unreal.c |
| 19:03.33 | rmudgett | On v13 and master |
| 19:03.45 | jkroon | anything critical in the changelog? |
| 19:03.53 | jkroon | should we patch 13 too? |
| 19:04.12 | jkroon | Return -1 from local_indicate to force asterisk core to correctly add hangupcauses to the local channel (ASTERISK-25912). <-- ok for changelog? |
| 19:04.22 | file | this is a minor bug fix so it doesn't need to go into the changelog |
| 19:04.27 | file | the commit message needs to follow the formatting though |
| 19:04.40 | jkroon | referring to commit message |
| 19:04.42 | file | https://wiki.asterisk.org/wiki/display/AST/Commit+Messages |
| 19:04.59 | file | and it will need to go into 13 and master |
| 19:06.05 | jkroon | do i need a verbose here? chan_local: Fix hangupcauses not getting set on Local/ channels (ASTERISK-25912). |
| 19:06.38 | file | the ASTERISK-25912 needs to go on a new line and have a #close tag |
| 19:06.55 | file | a more verbose description would be nice, perhaps including a full scenario |
| 19:07.15 | jkroon | well, this won't close it, it only fixes it for 11 currently. |
| 19:07.39 | file | the change won't go in until all branches have a change up |
| 19:07.44 | file | and they are reviewed |
| 19:07.49 | jkroon | ok |
| 19:08.12 | jkroon | I note the Change-Id: tag - how does that flow through the process? |
| 19:08.34 | file | it's automatically added by the gerrit stuff |
| 19:08.41 | file | it links the commit to the review |
| 19:08.50 | file | so that if it is modified, it still is linked to the review |
| 19:09.36 | jkroon | wouldn't a change generate a new commit? or do I update the commit and force push it? |
| 19:09.52 | file | you amend the commit |
| 19:09.55 | file | and do git review again |
| 19:10.08 | file | you can also pull a review down and create a new local branch using git review -d <review number> |
| 19:10.16 | file | make a change, amend, and then git review again |
| 19:10.18 | file | and it'll update |
| 19:10.48 | jkroon | ok, this is going to be a learning curve, but may very well solve a problem we're having here |
| 19:12.31 | jkroon | http://pastie.org/10793917 - sorry for wasting time here |
| 19:12.35 | jkroon | getting that error |
| 19:12.49 | file | you need to specify the branch to git review |
| 19:12.52 | file | git review -D 11 |
| 19:13.27 | file | probably also need to trash that rebase using git rebase --abort |
| 19:13.40 | jkroon | that i've already done :) |
| 19:13.53 | jkroon | not a complete n00b to git however I am new to gerrit |
| 19:14.12 | jkroon | that worked. |
| 19:15.02 | jkroon | i take it merely submitting as a draft is not adequate? |
| 19:15.12 | file | a draft means it's not yet ready for review |
| 19:15.23 | file | if you want it to be up for review then you need to publish it |
| 19:15.34 | jkroon | git review -t ASTERISK-25912 11 <-- to submit it? |
| 19:15.46 | file | you can do it from the web UI, or that may do it |
| 19:15.51 | file | I don't use drafts and haven't done that |
| 19:16.19 | jkroon | i'm trying to follow the guide - the patches I usually write for asterisk are relatively simple of nature. |
| 19:16.50 | file | you can do it, it won't break anything |
| 19:17.13 | jkroon | git review -t ASTERISK-25912 11 -- that failed with ! [remote rejected] HEAD -> refs/publish/11/ASTERISK-25912 (no new changes) |
| 19:17.53 | file | yeah, you'll need to go to https://gerrit.asterisk.org/ and click publish on it |
| 19:19.38 | jkroon | ok, published. |
| 19:22.23 | jkroon | do I need to publish separately for trunk and for 13? |
| 19:22.47 | file | yes, each branch is a completely separate review |
| 19:23.15 | jkroon | should've labeled my branches better ... but next time I'll know. |
| 19:26.08 | jkroon | on master I see core_unreal.c has a slightly different approach, better if I add another case and just ignore the response from unreal_queue_indicate or do I use an if statement again? |
| 19:28.45 | *** join/#asterisk-dev pchero (~pchero@109.70.54.56) |
| 19:33.32 | jkroon | ok, so without having to rebase the branch the whole time between master and 13, is there a way to submit for both master and 13? |
| 19:35.55 | file | you can go into the review on gerrit.asterisk.org and click cherry pick |
| 19:36.00 | file | and then choose the branch to cherry pick it into |
| 19:36.04 | file | ie: type in master for master |
| 19:36.12 | file | it will then create a new review against that branch with the change |
| 19:36.58 | jkroon | ok, i think that should be it. |
| 19:39.34 | jkroon | thanks file. much appreciated. |
| 19:39.44 | file | and now dog walk time! before it rains |
| 19:39.51 | rmudgett | Did the patch fix your problem? |
| 19:40.08 | jkroon | rmudgett, yes it did. on 11 at least. can't test on 13. |
| 19:42.27 | jkroon | if this gets through I think I may have a few more patches that I can then post - most of them I've posted on issues.asterisk.org but this way should get them through faster. |
| 19:43.41 | rmudgett | Much faster |
| 19:44.13 | jkroon | cheers |
| 19:58.09 | gtjoseph | hey file: why did you have to rebase 2525 and 2526? |
| 19:58.18 | gtjoseph | was there an issue with them? |
| 19:58.34 | file | gtjoseph, they conflicted |
| 19:58.40 | file | wouldn't let me submit |
| 19:59.33 | gtjoseph | hmmm |
| 20:08.32 | gtjoseph | rmudgett: can you re +1 2525 and 2526 sicne they were rebased? |
| 20:16.25 | rmudgett | re +1ed |
| 20:16.26 | gtjoseph | thanks |
| 20:18.02 | gtjoseph | file if you +2 2525 and 2526 again, they should go this time |
| 20:23.43 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 20:23.43 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 20:50.00 | *** join/#asterisk-dev [TK]D-Fender (~joe@64.235.216.2) |
| 23:09.32 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-aipwduhukwzbiovq) |
| 23:23.03 | *** join/#asterisk-dev acidfoo_ (~acidfoo@157-52-14-38.cpe.teksavvy.com) |