**** BEGIN LOGGING AT Wed Jan 08 02:59:59 2014 Jan 08 09:31:04 good morning Jan 08 10:16:09 morning all Jan 08 11:27:41 pb_: ping Jan 08 12:05:32 hi Jan 08 12:12:52 Noor: hello Jan 08 12:49:39 \o everyone Jan 08 12:49:46 I keep having a problem with oe-core and m4-native Jan 08 12:49:59 it gets stuck on do_configure when building core-image-minimal Jan 08 12:50:57 log.do_configure: http://pastebin.com/KU33BAAR Jan 08 13:04:20 pb_: hi ... sorry I was away Jan 08 13:04:49 pb_: we saw your implementation which put dbg packages in separate filesystem Jan 08 13:05:00 which you shared with kergoth sometime back Jan 08 13:05:26 I was curios to ask why that implementation is not part of upstream yet Jan 08 13:07:28 No particular reason as far as I know. Jan 08 13:08:10 pb_: do you expect that it will be in near future? Jan 08 13:08:51 I don't have any particular plans to get it merged though I don't have an objection to doing that either. It's just a question of finding time to sort it out. Jan 08 13:08:59 If you wanted to submit it upstream then you're welcome to do that. Jan 08 13:09:06 pb_: we are planning to do the same .... so just evaluating that if your implementation will be available then we don't try other thing Jan 08 13:09:45 pb_: OK .... we will work on it and send a patch to upstream Jan 08 13:09:57 ok, very good. thanks. Jan 08 14:35:35 jackmitchell, ping Jan 08 14:35:42 Crofton|work: pong Jan 08 14:35:45 hey Jan 08 14:35:49 hi Jan 08 14:35:55 the db using c++ also stuff is my fault Jan 08 14:36:08 but I am not sure why your build tries to link against the host lib Jan 08 14:36:31 Let me update lcoally and see if some fur crept in Jan 08 14:36:32 I wondered if that is what it was trying to do Jan 08 14:37:05 reading the log, it is clear it tries to link agianst something in /usr/lib Jan 08 14:38:21 can you paste the link to the log in here? Jan 08 14:38:36 the same as the one I mentioned in the email? http://ix.io/9Ki Jan 08 14:38:55 /usr/lib/libstdc++.so: could not read symbols: File in wrong format Jan 08 14:38:58 very bad :) Jan 08 14:39:07 indeed :) Jan 08 14:39:48 I do not have that file on my machine Jan 08 14:42:22 bother Jan 08 14:42:25 ERROR: ExpansionError during parsing /home/balister/src/oe-core/meta-oe/meta-oe/recipes-support/vim/vim_7.4.27.bb: Failure expanding variable do_fetch[file-checksums], expression was ${@bb.fetch.get_checksum_file_list(d)} which triggered exception TypeError: supports() takes exactly 4 arguments (3 given) Jan 08 14:42:52 heh, and so the rabbit hole begins Jan 08 14:51:44 not obvious what broke vim Jan 08 14:55:48 I definitely can't find a recipe for Econnamn and I find this quite surprising, what do Enlightenment user use to manage network ? Jan 08 14:58:41 oh my god it's included in the e-wm recipe ?!? Jan 08 15:01:11 talking about connman, it has been configured with --enable-wifi but I don't get the wifi technology listed Jan 08 15:01:18 what can I check ? Jan 08 15:01:25 ok I got it, connman module is included but not econnman, so my question is still valid Jan 08 15:02:21 abelloni, I worked out of box for me, just had to run the daemon. Is your wlan interface up ? Jan 08 15:02:37 abelloni, -I +It Jan 08 15:04:54 yeah Jan 08 15:05:10 wlan0 Link encap:Ethernet HWaddr 48:02:2A:C4:2E:15 UP BROADCAST MULTICAST MTU:1500 Metric:1 Jan 08 15:06:50 oswin: what is your connman version ? Jan 08 15:06:56 I have 1.17 Jan 08 15:07:14 1.20 Jan 08 15:45:29 Crofton: bluelightning: did anyone sort out the location of MeetBot's meeting notes from yesterday? Jan 08 15:46:16 tlwoerner: no, I don't know if Crofton|work got my ping Jan 08 15:46:35 I emailed them to jefro Jan 08 15:46:52 Crofton|work: the partial URLs given by the bot appeared to be invalid... Jan 08 15:46:57 yep Jan 08 15:47:02 bluelightning: is it safe to assume they went *somewhere*? :-) just that the link provided wasn't valid? Jan 08 15:47:02 they are generic urls Jan 08 15:47:19 I copy the minutes to emerald, but not sure if we exported the directory Jan 08 15:47:24 yes Jan 08 15:47:34 they are in the bot dir on my machine Jan 08 15:47:34 will someone update http://www.openembedded.org/wiki/TSC#Meeting_Minutes ? Jan 08 15:47:40 I emailed to jefro Jan 08 15:47:58 I'll see how jefro wants to handle this Jan 08 15:48:02 Crofton|work: it's confusing because the bot points to the main meetbot wiki page which of course tells people files should be found on the debian servers Jan 08 15:48:09 Crofton|work: okay thanks :-) Jan 08 15:48:13 I screwed up and deleted my script that rsyncs, so I need to fix that Jan 08 15:48:19 yeah Jan 08 15:48:26 I should see if that is fixable Jan 08 15:48:42 just automating the log collection etc as a leap forward for us Jan 08 15:49:12 not bad for 2014 (haha) Jan 08 16:50:06 oswin: do you mind giving me the ouput of connmanctl technologies | grep Type ? Jan 08 16:50:54 hi. what's Khem Raj nick on irc? Jan 08 16:52:23 khem Jan 08 16:59:34 anyone seen this? Jan 08 16:59:35 ERROR: ExpansionError during parsing /home/balister/src/oe-core/meta-oe/meta-oe/recipes-support/vim/vim_7.4.27.bb: Failure expanding variable do_fetch[file-checksums], expression was ${@bb.fetch.get_checksum_file_list(d)} which triggered exception TypeError: supports() takes exactly 4 arguments (3 given) Jan 08 17:10:17 Crofton|work: try updating Jan 08 17:10:50 again? Jan 08 17:11:04 when did you last update? Jan 08 17:11:38 this morning Jan 08 17:11:47 I meant update bitbake, btw... http://git.openembedded.org/bitbake/commit/?id=e13acb4113ce75226664c3006a9776cc885e860d Jan 08 17:11:48 I want to looki at jackmitchell db compile fail Jan 08 17:12:08 this update fixced it :) Jan 08 17:12:14 right, blame otavio ;) Jan 08 17:12:22 * Crofton|work blames otavio Jan 08 17:13:04 jackmitchell, so the db vuild try to link to a host libstdc++ today Jan 08 17:15:17 JaMa: thanks, that was easy ;-) that's what I thought, but didn't find the nick, i guess he isn't arount these days.. Jan 08 17:19:30 Crofton|work: so you're getting the same issue? Jan 08 17:19:41 not sure, still rebuilding Jan 08 17:19:48 but since the patch came from me, I suspect not Jan 08 17:20:53 ndec: probably better to send him an e-mail nowadays Jan 08 17:21:05 ok Jan 08 17:28:19 jackmitchell, it builds on my machone Jan 08 17:29:30 jackmitchell, look at a484b35b818768487ff27cf06b8c5d4e128126af Jan 08 17:29:48 try commenting out the compile_prepend and see if that changes anything Jan 08 17:30:06 the patch does two things enable building the C++ lib Jan 08 17:30:19 and the prepend fixes an rpath QA check Jan 08 17:36:49 Crofton|work: commenting out the compile_prepend doesn't work Jan 08 17:37:04 what are you buidling on? Jan 08 17:37:05 I've temporarily reverted that commit to see if it starts working again Jan 08 17:37:25 machine x86_64 multilib, distro archlinux\ Jan 08 17:37:53 can you check what package provided that file? Jan 08 17:37:56 so it compiles fine without the patch Jan 08 17:37:59 yeah Jan 08 17:38:37 http://pastebin.com/zbrR0bLy Jan 08 17:39:04 I don't have the .so in /usr/lib Jan 08 17:39:26 just so+number Jan 08 17:39:53 this is where I hope someone chiomes in about debain etc :) Jan 08 17:40:28 Crofton|work: that's good, isn't it? aren't the *.so's (by themselves) for development only (i.e. -dev)? Jan 08 17:40:51 I think i can compile c++ on my build machine though Jan 08 17:41:05 we test stuff that goes in core Jan 08 17:41:29 so I feel like the builds are OK on Fedora and some form a debian Jan 08 17:41:35 I want to blame arch Jan 08 17:41:39 but that is not fair Jan 08 17:41:56 ubuntu doesn't have it either Jan 08 17:41:57 http://pastebin.com/ufyqJu6N Jan 08 17:42:36 can you send another email highlighting the problem Jan 08 17:42:45 hmm, my arch non-multilib machine also has it in there Jan 08 17:42:52 I do not think anyone clicked through to the build failure but me :) Jan 08 17:43:00 so it's not even that the multilib packages are less cared for Jan 08 17:48:38 jackmitchell, send an email showing the contamination and skip the bit everyone answered :) Jan 08 17:48:52 Jefro, people keep bugging me about the tsc meeting minutes Jan 08 18:40:34 Jefro: and could the webpage be updated? http://www.openembedded.org/wiki/TSC#Meeting_Minutes Jan 08 20:27:04 are core-image-minimal-dev supposed to build fine? master:421e927bd453259f4b3cdbd1676f6e12f97bf34f suggested that "ERROR: User requested feature sdl" "configure was not able to find it" when attempting to configure qemu-native/1.6.1-r0 Jan 08 20:30:31 blindvt_: maybe try installing sdl-devel (or sdl-dev?) on your build host Jan 08 20:31:28 tlwoerner, hmz so there is no checking if a native class is provided by the host or should be built according to recipes? Jan 08 20:33:45 blindvt_: i'm not sure about your "native _class_" phrasing, but there is an ASSUME_PROVIDED (or something like that) to tell the build your host already provides some -native program Jan 08 20:33:51 but i don't think that's what going on here Jan 08 20:34:47 i think your build of qemu-native is looking for native sdl support (qemu runs better with it) and isn't finding it Jan 08 20:35:32 although OE builds qemu-native, i don't think it also builds sdl-native if you are missing it Jan 08 20:38:09 tlwoerner, well, ASSUME suggests that an attempt is made to use the build's stuff or fallback to the oe recipe, but that's not how it works ATM obviously. But either way :) Jan 08 20:39:50 good evening everyone Jan 08 20:40:13 I still have this m4-native issue building core-image-minimal: http://pastebin.com/KU33BAAR Jan 08 20:41:16 that pastebin doesn't actually show a problem Jan 08 20:41:17 cut off Jan 08 20:41:21 well Jan 08 20:41:27 do_configure is stuck Jan 08 20:41:29 at that point Jan 08 20:41:51 unlikely it has anything to do with bitbake, it's just running ./configure in the m4 buildsystem Jan 08 20:41:52 * kergoth shrugs Jan 08 20:42:15 I'll need to do one run now with "just" oe-core, to be sure Jan 08 20:44:49 it might be interesting though, if it happened to anyone else Jan 08 20:45:21 i had wanted to find "official" documentation to verify my knowledge of ASSUME_PROVIDED, but the existing documentation (looking at the mega manual) is rather thin Jan 08 20:45:47 blindvt_: my understanding is that if something is ASSUME_PROVIDED then there is no attempt to build it Jan 08 20:46:00 tlwoerner, that's true, yes Jan 08 20:46:40 tlwoerner, something that would be nice to have extended but i admit that that's a bit tricky Jan 08 20:50:05 tlwoerner, don't worry about this too much. We used to have fewer ASSUME_PROVIDED and that was rather inconvenient (because it was pretty slow to build), now we have more and users have to install more stuff on their build hosts for the time being, everything's fine Jan 08 21:29:37 khem, Re: b4ecf65e20a63d11f14483d213a3931f39225cdd 1) why the SSP_ALL_CFLAGS= 2) why the odd -C... and not just toplevel install_utils ? Jan 08 21:30:56 khem, i think both of these are obsolete since a couple of years by now, fyi Jan 08 21:32:59 well, apparently my problem des not occure with plain oe-core, so nvm Jan 08 21:33:56 so it miught be related to a bad automake version that I have put somewhere Jan 08 21:35:06 hi Jan 08 21:35:23 woglinde, hi Jan 08 21:47:38 kergoth, removing a file:// from SRC_URI should not provoke do_fetch to be run either, ideally. Would be nice to special-case certain selected vars so that their content is re-evaluated in a relevant manner, not if they changed at all Jan 08 22:05:00 okay last openjdk QA issue fixed I think Jan 08 22:05:15 just one rebuild Jan 08 22:37:42 hot off the presses: https://fosdem.org/2014/schedule/track/embedded/ Jan 08 22:39:44 https://fosdem.org/2014/schedule/event/underwater_acoustics_to_opkg/ Jan 08 22:45:10 cool stuff :-) Jan 08 23:03:18 git remote add iio git://git.kernel.org/pub/scm/linux/kernel/git/jic23/iio.git Jan 08 23:03:18 git remote update iio Jan 08 23:03:28 oops **** ENDING LOGGING AT Thu Jan 09 02:59:58 2014