**** BEGIN LOGGING AT Thu Jun 12 02:59:59 2014 Jun 12 04:28:30 hi all Jun 12 04:30:47 im having some troubles building gstreamer1.0_git under openembedded Jun 12 04:31:24 i know i need to patch it somehow, but im not sure exactly how. wondering if anyone could spare a minute to help me Jun 12 07:19:16 good morning Jun 12 07:37:07 gm Jun 12 07:54:18 hi woglinde, all Jun 12 08:37:39 morning Jun 12 08:38:12 morning afournier, all Jun 12 13:29:30 if i install the python recipe that ships with meta-openembedded on my device, it seems to be a slimmed down version of python standard library. is there another recipe for the "full" python version? Jun 12 13:30:03 for example, it does not seem to have the module contextlib Jun 12 13:30:50 mago_: most python modules are split out into their own packages Jun 12 13:31:10 mago_: you can install python-modules to get absolutely everything if you want; but that's often too much Jun 12 13:33:15 can i list all available python packages somehow? Jun 12 13:34:25 (how do I know which package i should include to get contextlib module?) Jun 12 13:34:58 what I usually do is to do a find in the packages-split directory of the workdir for python Jun 12 13:36:31 great, python-misc it is! thank you Jun 12 13:44:29 suihkulokki: hi there - do you know where I should xf86-video-omapfb now? Jun 12 14:24:46 when packaging the kernel it looks like some files are missing in -dev Jun 12 14:25:14 like arch/arm/kernel/traps.c Jun 12 14:28:24 bluelightning: nobody should need it anymore I think Jun 12 14:28:39 there is like 5 different other X servers for omap already Jun 12 14:33:56 trying to create a standard "setuptools" recipe for a python module, and get this QA error: QA Issue: non debug package contains .debug directory: python-zmq path /work/armv7a-vfp-neon-oe-linux-gnueabi/python-zmq/14.1.1-r0/packages-split/python-zmq/usr/lib/python2.7/site-packages/zmq/backend/cython/.debug/context.so Jun 12 14:34:20 seems like I should add something to FILES_${PN}-dbg .. but i don't fully understand what it is, and why i have to do it? Jun 12 15:03:45 suihkulokki: the problem is we have old stable releases which may fail to build with that repo gone Jun 12 15:04:01 maybe we just live with that, I don't know Jun 12 15:22:01 lets say I have the default behavior of enabling a TTY on ttyS0 in my distro. Now, a product needs to use tty0 for something else, so they don't need a tty. I went into local.conf and cleared SERIAL_CONSOLES, and my inittab was updated accordingly. Problem fix. However, if the product wants to manage these sorts of things in a layer, where would it go? does this qualify as a new "distro"? So they would create a new conf/distro/something.conf that i Jun 12 15:22:01 ncludes only this SERIAL_CONSOLES = "" and then copy everything from the default distro? Jun 12 15:35:12 hi all Jun 12 15:38:44 hi pb_ Jun 12 15:39:13 mago_: typically yes Jun 12 15:39:36 mago_: I'd be surprised if that was the sole bit of policy configuration that you have... Jun 12 16:16:34 ok Jun 12 16:16:49 can a distro inherit from another distro? Jun 12 16:17:38 i guess i could just include conf/distro/otherdistro.conf Jun 12 16:18:28 pb_: Do you have much knowledge about elf sections? Jun 12 16:19:16 pb_: Specifically,   [Nr] Name              Type            Addr     Off    Size   ES Flg Lk Inf Al Jun 12 16:19:16   [12] .note.GNU-stack   PROGBITS        00000000 0010ce 000000 00      0   0  1 Jun 12 16:19:37 For the GNU-stack section whether AL 1 and/or 0 is valid Jun 12 16:29:04 RP: I believe 0 and 1 are equivalent, both mean no constraints on alignment Jun 12 16:29:32 If the section is zero sized then it seems reasonable that it wouldn't require any alignment. Jun 12 16:30:28 pb_: elfutils complains that 0 is invalid Jun 12 16:30:53 pb_: gold emits a 0 there in some cases, ld.bfd emits 1 Jun 12 16:31:06 RP: I think elfutils is probably broken then. I seem to remember that prelink had a similar bug at one point. Jun 12 16:31:54 pb_: So that would make http://git.yoctoproject.org/cgit/cgit.cgi/poky-contrib/commit/?h=rpurdie/t222&id=d89a7283918962522046fa80d1a417288e0eba38 a correct fix... Jun 12 16:33:08 RP: Pretty much. I don't think it needs to be conditional on size=0. Jun 12 16:34:52 pb_: I don't quite follow your comment about size=0 Jun 12 17:17:21 Anyone else hit 'The remote end hung up unexpectedly' from git.ti.com from time to time? Jun 12 17:17:30 denix: ever run into this? Jun 12 17:22:53 kergoth: is it just today or always? Jun 12 17:35:55 denix: i dont think its just today, but randomly in general Jun 12 17:44:21 Jun 12 17:47:59 denix, ping Jun 12 17:49:34 JaMa, could you do a blog posting somewhere on how to interpret and fix the erros seen in your world reports? Jun 12 18:06:42 Crofton: you mean world test-dependencies builds? Jun 12 18:07:00 Crofton: because normal world builds should be clear Jun 12 18:20:48 Crofton: I've sent explanation to ML Jun 12 18:47:18 Crofton: pong Jun 12 19:00:08 denix, I am doing a build for a keystone board for a k2hk for a gsoc student Jun 12 19:00:10 any tricks? Jun 12 19:00:38 clever one to, he jsut pointed me at http://processors.wiki.ti.com/index.php/MCSDK_UG_Chapter_Exploring#Yocto Jun 12 19:03:41 Crofton: BSP is being merged to meta-ti, so you can use that. otherwise MCSDK is quite messy and a fork of old stuff. trying to clean that up now... Jun 12 22:27:35 what is the difference between a ubi and ubifs file? Jun 12 22:44:54 Crofton: ubi is more, ubifs is like partition Jun 12 22:46:25 so you need both to flash something Jun 12 23:52:36 hmmm, http://layers.openembedded.org/layerindex/branch/master/recipes/?q=devmem **** ENDING LOGGING AT Fri Jun 13 02:59:59 2014