| 00:43.37 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 01:00.32 | *** join/#asterisk-dev sgriepentrog (~sgriepent@162-239-32-132.lightspeed.brhmal.sbcglobal.net) |
| 01:19.12 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 01:24.45 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 01:47.16 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 01:50.14 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 03:22.57 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 03:36.50 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 03:55.48 | *** join/#asterisk-dev jagruti_ (~root@182.70.116.47) |
| 03:58.01 | *** part/#asterisk-dev jagruti_ (~root@182.70.116.47) |
| 04:05.23 | *** join/#asterisk-dev superscrat (~asanders@173-17-133-2.client.mchsi.com) |
| 05:36.13 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 05:36.13 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 06:17.14 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 06:17.14 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 07:12.12 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 07:12.12 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 07:21.08 | *** join/#asterisk-dev tzafrir (~tzafrir@local.xorcom.com) |
| 07:21.08 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 07:26.28 | oej | Morning folks |
| 07:40.15 | stefan27 | hey oej |
| 07:40.21 | stefan27 | are you really Swedish? |
| 07:54.26 | oej | Yes, really. And you? |
| 08:11.37 | *** join/#asterisk-dev hehol (~hehol@2001:1438:1009:200:e524:b778:4ab4:a9cd) |
| 09:52.27 | *** join/#asterisk-dev infobot (ibot@rikers.org) |
| 09:52.27 | *** topic/#asterisk-dev is Asterisk Development Discussion -=- http://www.asterisk.org/developers -=- Tier 2 and 3.14159265 support is in #asterisk -=- Asterisk 13.0.0 now released. Do the release dance! -=- Follow on Twitter at @AsteriskDev |
| 11:42.42 | *** join/#asterisk-dev oej (~oej@178-78-201-42.customers.ownit.se) |
| 11:42.42 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:01.59 | *** join/#asterisk-dev oej (~oej@178-78-201-42.customers.ownit.se) |
| 12:01.59 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 12:12.02 | *** join/#asterisk-dev [TK]D-Fender (~chatzilla@216-191-106-163.dedicated.allstream.net) |
| 12:37.09 | *** join/#asterisk-dev jagruti_ (~root@182.70.116.47) |
| 12:37.11 | *** part/#asterisk-dev jagruti_ (~root@182.70.116.47) |
| 13:37.36 | *** join/#asterisk-dev danjenkins (~anonymous@cpc11-folk2-2-0-cust171.1-2.cable.virginm.net) |
| 13:49.22 | *** join/#asterisk-dev sgriepentrog (sgriepentr@nat/digium/x-lwbozsvjfksketxt) |
| 13:50.59 | *** join/#asterisk-dev leedm777 (leedm777@nat/digium/x-vjosdfvmhqznycvj) |
| 14:08.49 | *** join/#asterisk-dev newtonr (RustyNewto@nat/digium/x-abudwrsbrlmswnru) |
| 14:08.49 | *** mode/#asterisk-dev [+o newtonr] by ChanServ |
| 14:14.22 | *** join/#asterisk-dev kharwell (kharwell@nat/digium/x-dfmefyswyhitvtlo) |
| 14:19.01 | *** join/#asterisk-dev putnopvut (putnopvut@asterisk/master-of-queues/mmichelson) |
| 14:19.01 | *** mode/#asterisk-dev [+o putnopvut] by ChanServ |
| 14:23.33 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-eekzpoevzfngudsy) |
| 14:25.06 | *** join/#asterisk-dev danjenkins_ (anonymous@nat/digium/x-yrbzaampdsxtjcqs) |
| 14:37.11 | *** part/#asterisk-dev rmudgett (rmudgett@nat/digium/x-eekzpoevzfngudsy) |
| 14:37.34 | *** join/#asterisk-dev rmudgett (rmudgett@nat/digium/x-eekzpoevzfngudsy) |
| 15:19.14 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 15:19.14 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 15:40.31 | *** join/#asterisk-dev superscrat (asanders@nat/digium/x-ncwpiypizshocskd) |
| 16:20.05 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 16:20.05 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 17:10.07 | *** join/#asterisk-dev anthm (~anthm@freeswitch/developer/anthm) |
| 17:31.48 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 17:31.48 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 17:54.32 | *** join/#asterisk-dev mjordan (mjordan@nat/digium/x-feabmrqinzfixntl) |
| 17:54.32 | *** mode/#asterisk-dev [+o mjordan] by ChanServ |
| 17:58.40 | *** join/#asterisk-dev russellb (~russellb@redhat/russellb) |
| 17:58.40 | *** mode/#asterisk-dev [+o russellb] by ChanServ |
| 18:02.25 | *** join/#asterisk-dev jsmith (~jsmith@fedora/jsmith) |
| 18:02.25 | *** mode/#asterisk-dev [+o jsmith] by ChanServ |
| 18:19.36 | *** join/#asterisk-dev russellb (~russellb@redhat/russellb) |
| 18:19.36 | *** mode/#asterisk-dev [+o russellb] by ChanServ |
| 18:35.30 | mjordan | superscrat: if you get a chance, can you verify coreyfarrell's latest patch set fixes your findings? |
| 18:35.31 | mjordan | https://gerrit.asterisk.org/#/c/15/ |
| 18:36.31 | coreyfarrell | mjordan: did you get a chance to add res_pjsip to the ref checks bamboo job? |
| 18:36.50 | mjordan | ooo, thanks for reminding me. I'll do it now |
| 18:37.06 | mjordan | really hopes that a move to jenkins makes that just part of a git repo |
| 18:37.51 | coreyfarrell | mjordan: you don't enjoy making tedious config changes for everyone? |
| 18:38.06 | mjordan | coreyfarrell: it's my passion |
| 18:45.44 | mjordan | coreyfarrell: Okay, I removed the '-d chan_pjsip -d res_pjsip' from the menuselect options |
| 18:45.58 | mjordan | er, actually, build-asterisk-only.sh script |
| 18:46.00 | mjordan | it should now build both |
| 18:46.10 | mjordan | I'll kick off a run now though so we can see later today if I was right :-) |
| 18:47.06 | *** join/#asterisk-dev russellb (~russellb@redhat/russellb) |
| 18:47.06 | *** mode/#asterisk-dev [+o russellb] by ChanServ |
| 18:48.37 | coreyfarrell | sounds great.. will get us memory testing coverage for pjsip itself and any features that are only tested with pjsip |
| 18:51.52 | *** join/#asterisk-dev leifmadsen (~leifmadse@asterisk/documenteur-extraordinaire/blitzrage) |
| 18:51.52 | *** mode/#asterisk-dev [+o leifmadsen] by ChanServ |
| 18:58.14 | *** join/#asterisk-dev danjenkins (anonymous@nat/digium/x-ynhbmvejhzhdtazf) |
| 19:03.57 | *** join/#asterisk-dev eremitah (~int@unaffiliated/eremitah) |
| 19:04.21 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 19:04.21 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 19:06.09 | *** join/#asterisk-dev eremitah (~int@unaffiliated/eremitah) |
| 20:16.26 | *** join/#asterisk-dev oej (~oej@h87-96-134-129.dynamic.se.alltele.net) |
| 20:16.26 | *** mode/#asterisk-dev [+o oej] by ChanServ |
| 20:26.32 | gtjoseph | sgriepentrog: are you sure you applied the patch? I checked out a clean copy of 13, applied the patch and I get a clean compile. |
| 20:27.05 | sgriepentrog | I'll retry. |
| 20:27.17 | gtjoseph | thanks |
| 20:30.15 | *** join/#asterisk-dev jrose_atDigium (jrose_atDi@nat/digium/x-wzyoydzshozyjzyw) |
| 20:32.42 | *** join/#asterisk-dev Demon_VoIP (~demon@ip253.net222.n37.ru) |
| 20:34.03 | sgriepentrog | gtjoseph: apprently I goofed something, seems to be working better now. |
| 20:34.18 | gtjoseph | whew. |
| 20:34.39 | gtjoseph | Does better mean it compiles successfully? |
| 20:34.51 | sgriepentrog | waiting |
| 20:38.38 | gtjoseph | sgriepentrog: if you've got a few minutes, revert the change in dsp.c and look at the array-bounds error. I can't figure out why there's an error there. |
| 20:40.20 | sgriepentrog | It compiles and runs. Checking your dsp error. |
| 20:42.38 | sgriepentrog | That's interesting. They're deprecating yum. |
| 20:43.12 | gtjoseph | yep |
| 20:44.11 | sgriepentrog | looks like you could pretty much just alias yum=dnf though. |
| 20:44.19 | gtjoseph | exactly |
| 20:44.26 | Qwell | sgriepentrog: The packaging voice of reason died a few years ago. |
| 20:44.49 | Qwell | There's nobody left to tell them their doing dumb things anymore. ;/ |
| 20:44.54 | Qwell | ...they're* |
| 20:45.16 | mjordan | uhm |
| 20:45.20 | mjordan | deprecating yum? |
| 20:45.49 | sgriepentrog | fedora 22 alpha |
| 20:46.36 | sgriepentrog | gtjoseph: I don't see why either. I mean, how is it supposed to know what value dsp->freqcount could get to? Yeah, it could be too large, but it has no way of knowing that at compile time. |
| 20:47.37 | gtjoseph | well, it can because the loop that sets dsp-freqcount is deterministic. I think it's just broke. |
| 20:47.43 | gtjoseph | gcc5 that is. |
| 20:47.46 | sgriepentrog | I suppose you could assert on the dsp->freqs being less than sizeof first, maybe it would pick up on that. |
| 20:49.09 | gtjoseph | possibly |
| 20:49.14 | gtjoseph | i'll mess with it more. |
| 20:50.39 | sgriepentrog | Yeah, it does. This compiles: if (y<sizeof(dsp->freqs)) |
| 20:50.39 | sgriepentrog | <PROTECTED> |
| 20:51.27 | gtjoseph | yeah, but that will pass NULL pointers to goertzel_sample |
| 20:52.34 | gtjoseph | anyway, I can rewrite it so it still works and passes the test but i'd still like to know if it's a gcc bug or are we missing something. |
| 20:52.39 | sgriepentrog | Apparently the compiler is convinced that freqcount can be 7 or more. |
| 20:52.44 | gtjoseph | exactly |
| 20:52.50 | gtjoseph | which it can't |
| 20:52.52 | sgriepentrog | I think it's our bug honestly. |
| 20:52.56 | gtjoseph | unless i'm going blind. |
| 20:53.17 | gtjoseph | look down at ast_dsp_prog_reset |
| 20:53.29 | sgriepentrog | Yup. |
| 20:54.36 | Qwell | jsmith: https://lists.fedoraproject.org/pipermail/devel/2012-June/168941.html :( |
| 20:55.07 | sgriepentrog | What if the compiler thought the value could be uniitialized? |
| 20:55.30 | gtjoseph | I tried initializing the variable, and hard coding it to 7, still no go. |
| 20:56.05 | mjordan | Qwell: interesting |
| 20:56.06 | jsmith | Qwell: I see your mailing list post (by some crazy guy) and raise you this: https://lists.fedoraproject.org/pipermail/devel/2015-April/209580.html |
| 20:56.12 | rmudgett | ast_dsp_prog_reset() can set dsp->freqcount to 8 |
| 20:56.12 | mjordan | I can't claim to know enough to have a good opinion |
| 20:56.34 | gtjoseph | rmudgett: how? |
| 20:56.50 | jsmith | Qwell: You want the long complicated explanation, or the short explanation? |
| 20:56.50 | rmudgett | max = x + 1 |
| 20:56.54 | mjordan | jsmith: Qwell: for my own curiosity - why the move away from yum? |
| 20:56.58 | mjordan | jsmith: I'll take both! |
| 20:56.59 | rmudgett | x ranges 0 to 7 |
| 20:57.12 | gtjoseph | rmudgett: but max can only be <= 6 |
| 20:57.24 | jsmith | mjordan: Yum was a giant mess of python code -- no clean APIs, slow in spots, and less and less maintainable over time |
| 20:57.29 | rmudgett | oh righ |
| 20:57.31 | rmudgett | t |
| 20:57.40 | mjordan | jsmith: sounds like another open source project I know :-P |
| 20:57.44 | jsmith | mjordan: So the idea was to start from scratch with something that *wasn't* yum, but very much like unto it... |
| 20:57.56 | Qwell | jsmith: but it was just a fork of yum, wasn't it? |
| 20:58.09 | jsmith | mjordan: With a not-invented-here depsolver, real APIs, and the core parts written in C |
| 20:58.12 | mjordan | jsmith: relevant - https://www.youtube.com/watch?v=aNaXdLWt17A |
| 20:58.21 | gtjoseph | rmudgett: unless ARRAY_LEN is calculating the length incorrectly |
| 20:58.23 | jsmith | mjordan: So now that dnf has matured, it's "taking over" for yum |
| 20:58.37 | jsmith | mjordan: In the next version of Fedora (or two), when you run "yum", you'll really be running the dnf code |
| 20:58.50 | sgriepentrog | I'm starting to believe that gtjoseph is right and it's a compiler glitch. |
| 20:58.55 | jsmith | mjordan: You can always go back to the "old" yum code, at least for now |
| 20:59.11 | mjordan | are the spec file definitions changing? |
| 20:59.16 | jsmith | Nope |
| 20:59.19 | Qwell | mjordan: it's still just RPM |
| 20:59.21 | mjordan | \o/ |
| 20:59.21 | gtjoseph | There was an bug like this back in 4.6 |
| 20:59.27 | rmudgett | gtjoseph: If ARRAY_LEN is wrong then there should be other problems. It is used a lot |
| 20:59.36 | jsmith | mjordan: Still RPM, still the same packaging, same build system, etc. |
| 20:59.48 | mjordan | I can't really blame someone for wanting to maintain a nicer codebase |
| 20:59.53 | gtjoseph | yeah and I just hard coded it to 7 and it still failed |
| 20:59.56 | jsmith | mjordan: The depsolver is *slightly* different, but actually quite a bit better (and is the same one used in OpenSuSE) |
| 21:01.00 | jsmith | Qwell: As to the finer points of whether or not it was a "fork"... well, I think it was more of a titanium spork :-p |
| 21:01.17 | Qwell | jsmith: I just remember the yum guys (proper) being pissed about lack of cooperation. |
| 21:01.18 | jsmith | Qwell: More or less the same shape, but different materials, and slightly more useful :-p |
| 21:02.06 | jsmith | Qwell: Well, yeah... |
| 21:02.47 | sgriepentrog | If I check dsp->freqcount > 7 and return 0, then it compiles fine as long as that's inside the while len loop. Put it above that and it still fails. Gcc is trying to be smart, but it's not consistent. |
| 21:03.11 | gtjoseph | rmudgett: sgriepentrog I'm just going to rewrite the test so it doesn't need freqcount and if I can reproduce the bug in a test snippet, i'll file it with gcc. |
| 21:03.29 | jsmith | sgriepentrog: I'd suggest trying with GCC 5, but that's apt to bring even more pain :-( |
| 21:03.41 | gtjoseph | jsmith: this IS with gcc5 |
| 21:03.51 | *** join/#asterisk-dev [TK]D-Fender (~joe@64.235.216.2) |
| 21:03.57 | jsmith | gtjoseph: Ah, right... I should have looked at the scrollback |
| 21:04.16 | gtjoseph | and it's the only remaining issue |
| 21:04.24 | gtjoseph | :) |
| 21:04.34 | jsmith | Oh, even better. |
| 21:06.13 | sgriepentrog | The while (len) loop is really confusing gcc. Inside that, it's got everything figured out. Outside of that, forcing freqcount to a valid number doesn't matter it still complaints. |
| 21:06.21 | sgriepentrog | s/complaints/complains/ |
| 21:06.54 | gtjoseph | yeah, it's really weird. |
| 21:07.30 | gtjoseph | i even tries with no optimization and it still fails. |
| 21:07.35 | gtjoseph | s/tries/tried/ |
| 21:09.59 | sgriepentrog | How about : if (dsp->freqcount < ARRAY_LEN(dsp->freqs)) |
| 21:09.59 | sgriepentrog | <PROTECTED> |
| 21:09.59 | sgriepentrog | <PROTECTED> |
| 21:09.59 | sgriepentrog | <PROTECTED> |
| 21:10.30 | sgriepentrog | It's low overhead, compiles, and shouldn't ever be false. Better yet would be an assert just in case.. |
| 21:10.32 | gtjoseph | does it pass? |
| 21:10.39 | sgriepentrog | yup. Compiles. |
| 21:10.52 | sgriepentrog | Of course it's missing a level of curlies. |
| 21:12.09 | gtjoseph | I think I'll just check dsp->freqs[y]->fac |
| 21:12.24 | gtjoseph | it can only be 0 if that slot wasn't initialized |
| 21:12.51 | gtjoseph | i can get rid of freqcount altogether then |
| 21:13.20 | sgriepentrog | Problem is if you make a significant change you need a test to prove it didn't break. |
| 21:13.35 | gtjoseph | yeah true |
| 21:13.51 | rmudgett | It didn't break. The compiler said so :) |
| 21:14.10 | gtjoseph | :) |
| 21:15.54 | gtjoseph | Do we have a testsuite test that exercises this? |
| 21:16.22 | sgriepentrog | shrugs |
| 21:17.00 | rmudgett | DTMF detection is the main thing I can think of that the testsuite would exercise of the dsp code. |
| 21:17.45 | gtjoseph | but only for analog or in-band right? |
| 21:17.54 | rmudgett | Touching channels.h causes slow builds... |
| 21:18.05 | rmudgett | gtjoseph: yes |
| 21:20.44 | *** join/#asterisk-dev tzafrir (~tzafrir@bzq-218-28-58.cablep.bezeqint.net) |
| 21:20.44 | *** mode/#asterisk-dev [+o tzafrir] by ChanServ |
| 21:38.30 | *** join/#asterisk-dev Synthase__ (uid63346@gateway/web/irccloud.com/x-jgaldacfyobstavf) |
| 21:39.04 | *** join/#asterisk-dev otherwiseguy (~otherwise@23-228-175-118.mci.googlefiber.net) |
| 21:41.08 | *** join/#asterisk-dev sgriepentrog (sgriepentr@nat/digium/x-eipirpnckmstigls) |
| 21:46.30 | *** join/#asterisk-dev tm1000 (sid6728@gateway/web/irccloud.com/x-idsmymjgkxkdwgwg) |
| 22:24.35 | *** join/#asterisk-dev jameswf (uid27319@gateway/web/irccloud.com/x-ydmnlmuhqgfpleaz) |
| 23:20.05 | *** part/#asterisk-dev kharwell (kharwell@nat/digium/x-dfmefyswyhitvtlo) |
| 23:39.38 | *** join/#asterisk-dev leedm777 (~leedm777@216.186.181.138) |