**** BEGIN LOGGING AT Tue Jun 03 02:59:58 2014 Jun 03 07:31:08 do_fetch says "looking for in paths:". wouldn't it be useful if it also told us where it found it? Jun 03 07:32:20 good morning Jun 03 07:33:46 morning :) Jun 03 07:52:23 morning Jun 03 07:52:46 oe.utils.contains disapeared ? AttributeError: 'module' object has no attribute 'contains' Jun 03 07:54:15 ok i will use bb.utils.contains instead Jun 03 08:01:32 isn't the kernel normally added to /boot ? mine is empty. Jun 03 08:03:05 mine is added to deploy/images/ARCH/bzImage... Jun 03 08:06:11 deploy/images/MACHINE/bzImage sorry Jun 03 08:07:04 gm Jun 03 08:07:20 afournier: yeah. I get it there as well, but I would like to have it in the /boot folder of my image. I have 2 sw-images and need to load kernel and dtb from there. Jun 03 08:08:40 hi guys .. i trying to recompile a old oe-classic toolchain on a recent host , but i fails on coreutils-native ( and probably more) , is there any other way to bring this to live or is the only way to carefully move the version up until it works Jun 03 08:39:37 what should i do to get my patch reviewed/accepted ? Jun 03 08:55:08 I can add kernel-image and kernel-devicetree to get stuff into /boot, but thought that was automatic. Jun 03 09:40:13 tasslehoff: no, afaik it's not automatic Jun 03 09:47:15 ack Jun 03 11:31:27 since last update openembedded stoped working inside source directory and creates a build directory next to it, is it possible to disable this behavior ? Jun 03 11:31:54 B = "${S}" or 'inherit autotools-brokensep' Jun 03 11:32:01 ok Jun 03 11:32:02 thanks Jun 03 11:32:04 but the right way is to fix the issues you're seeing Jun 03 11:33:14 i guessed this but i don't have time for that right now :( Jun 03 11:38:12 anyway this new behavior looks way cleaner than before Jun 03 11:56:13 http://pastebin.com/abCRJFas <- trying to rebuild my meta data on x86_64 host , new linux mint 17 ? any ideas why this fails like that Jun 03 12:00:17 rob_w: on what branch/commit are you? Jun 03 12:00:53 there was a problem with newer tar versions on some hosts, but that has been fixed a while ago, it should be in dora too Jun 03 12:01:00 specifically http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=bf19a8617d05f17506dcb7154e19e2aa2b61838b Jun 03 12:02:45 this is not dora data .. thanks for the hint Jun 03 12:47:28 that's a bit kamikaze how the autotools / autotools-broken thing has been managed, now i find that tftp-hpa does not build, would not have been less dangerous to replace all inherit autotools by inherit autotools-brokensep first, so you keep the old behavior, and then change every recipe to inherit autotools gradually, so mainteners can swich to the new behavior after testing it ? Jun 03 12:48:09 or makes this new behavior in another class like autotools-sep Jun 03 12:48:21 afournier that would have been to easy Jun 03 12:49:04 :) Jun 03 12:49:33 easy is good :) Jun 03 12:50:26 strip: Unable to recognise the format of the input file Jun 03 12:50:41 Tag_CPU_name: "7-A" Jun 03 12:50:41 Tag_CPU_arch: v7 Jun 03 12:51:18 i cannot image how many layers are broken right now because of this autotools class :) Jun 03 12:51:57 afournier they should be fixed until the next release comes out Jun 03 13:00:38 hi; me again :) Jun 03 13:01:07 anyone knowledgeable about imx.6 yocto with vivante (again...) - in particular, when/why the vivante samples just quit without any message at all? Jun 03 13:02:37 [hmm maybe this belongs in #yocto instead] Jun 03 13:11:07 anyone here have a clue what the environment variable would be for me to set an vanila make install to use the host-arch's strip tool? Jun 03 13:11:16 I tried STRIP=${STRIP} Jun 03 13:11:31 any other lucky guess before I troll 10000000 lines of bullshit makefiles? Jun 03 13:11:57 what is the convention for strip? Jun 03 13:12:30 what is the command called even? Jun 03 13:12:32 pompomJuice: STRIP=echo Jun 03 13:12:39 pompomJuice: OE will strip it better for you Jun 03 13:12:42 oh my what does that do? Jun 03 13:12:45 (and preserve the symbols) Jun 03 13:12:50 perfect Jun 03 13:12:51 thanks Jun 03 13:13:02 having trouble creating a mono 3.6 with llvm back end Jun 03 13:13:04 it's best to disable any stripping inthe makefiles Jun 03 13:13:13 if I get this going I would like to contribute it to angstrom Jun 03 13:13:33 how do you disable makefile stripping? Jun 03 13:14:10 basically I took the normal lvvm from meta-oe and adapted them for mono llvm Jun 03 13:14:22 almost there Jun 03 13:14:48 look fro 'strip', 'STRIP' or 'install -s' in the makefiles Jun 03 13:14:55 ok thanks Jun 03 13:15:09 Just needed confermation that that is what I must look for Jun 03 13:15:21 these makefiles recurse to hell & gone Jun 03 13:16:44 ok you disable with the echo move Jun 03 13:16:45 take it slow Jun 03 13:17:16 why are the mono recipies on OE so behind? Jun 03 13:17:33 still stuck on mono 2.x Jun 03 13:18:00 building mono 3.4 from github is almost trivial now Jun 03 13:18:43 when im done im gonna pastbin the recipies then you can put them in Jun 03 13:19:01 since mono on embedded is the next big thing IMO Jun 03 13:19:09 we are doing it, it is working brilliantly Jun 03 13:19:20 .net on linux Jun 03 13:19:21 woop Jun 03 13:21:43 I do a make install Jun 03 13:21:48 it fails with a strip error Jun 03 13:21:56 I grep the word "strip" Jun 03 13:21:58 0 hits Jun 03 13:22:01 boggles the mind Jun 03 13:23:32 grep -i ? Jun 03 13:23:44 well I use midnight commander Jun 03 13:24:17 part of the problem is makefile output errors are so hard to locate what makefile is causing the problem Jun 03 13:24:29 if htey just added the file location Jun 03 13:25:33 ok but I found the solution Jun 03 13:25:40 configure has a switch Jun 03 13:25:41 --enable-keep-symbols Jun 03 13:25:50 take it slow Jun 03 13:27:06 is there a way to make sstate go from state Jun 03 13:27:17 do_install -> do_configure Jun 03 13:27:29 im at do_install Jun 03 13:27:44 nm it just seems to work Jun 03 14:02:43 in a bitbake recipy, how do you mark a file so that it lands in sysroots/bin but not on the device's /usr/bin? Jun 03 14:03:18 nm Jun 03 14:10:16 what is the bitbake variale for the "package" folder? Jun 03 15:25:12 so Jun 03 15:25:18 an extra long TSC meeting today? Jun 03 15:27:50 in run.do_compile i have BUILD_CXX="g++ ", CXX="i586-oe-linux-g++ <...>", CXX_FOR_BUILD="g++ ", is it normal for BUILD_CXX and CXX_FOR_BUILD to not have i586-oe-linux-g++ ? Jun 03 15:29:40 yes, if you're building for another arch Jun 03 15:29:57 i am building for i586 Jun 03 16:00:43 * RP notes its TSC time (or one of them anyway) Jun 03 16:00:55 * RP can't be around in a couple of hours but is around now for a bit Jun 03 16:04:20 https://gitorious.org/dbus-cplusplus/mainline/source/1f6f3e6e966e0b453edc4a82338dc27966c37505:tools/Makefile.am the comment says it well :) removing CXX = $(CXX_FOR_BUILD) make the compilation go super smooth :) Jun 03 16:11:41 * koen is here as well, off and on Jun 03 16:34:10 * koen drops a pin Jun 03 16:35:03 * Jefro heard something Jun 03 16:43:38 * RP did too, belated Jun 03 16:43:52 speed of sound isn't great :) Jun 03 16:59:57 Hello, question for a few people: can I set a conditional INHERTIS in distro/.conf? Jun 03 17:02:31 WarheadsSE: using an indirect variable, yes Jun 03 17:03:03 FOO = "" Jun 03 17:03:03 FOO_machine = "" Jun 03 17:03:03 INHERITS += "${FOO}" Jun 03 17:03:21 Okay, so, say, we only want this on a release image .. Jun 03 17:03:27 or use a conditional python expression in there Jun 03 17:03:57 * RP -> afk Jun 03 17:03:59 I was told by one of my team members, they attempted that and got an error about undefined Jun 03 19:59:35 anyone know anything about stub.h file? Jun 03 19:59:43 stubs.h Jun 03 19:59:47 I have a weird problem Jun 03 20:00:31 frak Jun 03 20:00:41 why am I defing hard and soft abi .... Jun 03 20:00:45 never mind Jun 03 20:19:11 :P **** ENDING LOGGING AT Wed Jun 04 02:59:58 2014