**** BEGIN LOGGING AT Mon Apr 02 02:59:58 2012 Apr 02 08:52:57 Hello, I'm using oe-core and having issue with custom kernel bb recipe, getting: "ERROR: Multiple .bb files are due to be built which each provide virtual/kernel...", when building a kernel (bitbake linux-mytest) which is not specified in machine's PREFERRED_PROVIDER_virtual/kernel. The build builds two kernels, one specified in PREFERRED_PROVIDER_virtual/kernel and the one specified on command line. I could resolve it with hard setting of PROVIDES to "${P} ${PF} ${PN Apr 02 12:07:47 ericben: i applied the meta-fsl-arm patches Apr 02 13:28:30 Does someone has any PoE IC suggestion to us to use for an i.MX based project? Apr 02 14:32:53 hmmmm Apr 02 14:32:57 hi Apr 02 14:33:28 openembedded-core/meta/recipes-extended/shadow/shadow-native_4.1.4.3.bb has inherit gettext but still fails with | configure.in:496: error: possibly undefined macro: AM_GNU_GETTEXT_VERSION Apr 02 14:37:08 anyone got a ballpark estimate for the TMPDIR size for an angstrom console-image from oe-core (with inherit rm_work)? Apr 02 14:48:10 tasslehoff, it depend on the number of machines and on what you build Apr 02 14:49:14 denisATeukrea: building just for beagleboard. I'm on a mac and need to create a virtual disk for my TMPDIR Apr 02 14:49:32 personally I build for imx53qsb and I'm doing du right now Apr 02 14:49:48 tried giving it 60GB, and hope that will be enough. Apr 02 14:49:50 15G for sstate-cache Apr 02 14:50:06 the tmpdir is beeing calculated Apr 02 14:55:24 hmmm Apr 02 14:55:33 maybe I'll workarround it Apr 02 14:55:43 but I really wonder why it's broken Apr 02 14:56:01 what is broken? Apr 02 14:57:21 for me openembedded-core/meta/recipes-extended/shadow/shadow-native_4.1.4.3.bb, doesn't find the gettext m4 Apr 02 14:57:30 there is inherit gettext tough Apr 02 14:57:40 but still: Apr 02 14:57:41 | configure.in:496: error: possibly undefined macro: AM_GNU_GETTEXT_VERSION Apr 02 14:58:25 hmmm bitbake -e says: Apr 02 14:58:26 DEPENDS="autoconf-native automake-native libtool-native gnu-config-native gettext-minimal-native" Apr 02 14:58:41 I wonder what's the minimal version Apr 02 14:58:46 I'll try to see if it's there Apr 02 14:58:46 denisATeukrea: is sounds like the gettext-native / gettext-minimal-native issue that RP just fixed Apr 02 14:58:53 ah ok Apr 02 14:58:55 thanks Apr 02 14:59:02 if you update you should have the fix :) Apr 02 14:59:04 I'll update Apr 02 14:59:07 thanks a lot Apr 02 14:59:10 np Apr 02 15:33:21 http://pastebin.com/SKP5LdHa <- seen this before? what is bitbake trying to tell me? Apr 02 15:35:50 hmmm Apr 02 15:35:56 bitbake -e is your friend Apr 02 15:36:14 bitbake -e the package Apr 02 15:36:21 and redirect the output in a file Apr 02 15:36:24 then vim the file Apr 02 15:36:37 and look at the line that generate METADATA_BRANCH Apr 02 15:36:44 ah sorry Apr 02 15:36:58 maybe it won't produce -e output Apr 02 15:37:02 because of that error Apr 02 15:37:03 hmmmm Apr 02 15:37:50 maybe it expects git in every layer? Apr 02 15:42:14 denisATeukrea: metadata_scm.bbclass can really only handle git and monotone; but it works fine without git Apr 02 15:42:33 ok Apr 02 15:42:49 that is a puzzling error, I've never seen it before Apr 02 15:42:58 neither do I Apr 02 15:43:18 maybe not enough memory in the virtual machine? Apr 02 15:45:14 denisATeukrea: that could be. Apr 02 15:47:09 it had only 768MB :) Apr 02 15:51:00 on work computer I've 6G and on personal computer 4G Apr 02 15:52:38 denisATeukrea: gave it 4GB, and it seems much happier Apr 02 15:53:09 ok Apr 02 15:53:19 can you give it 6G? Apr 02 15:53:24 because 4G is a bit low Apr 02 15:57:27 . Apr 02 15:59:29 well it depend on what you cross compile but for stuff involving webkit linking and sometimes parsing is better with 6G than 4G Apr 02 16:35:13 4G and compiling with one core in a VM seems a bit low, but I've got the time :) **** BEGIN LOGGING AT Mon Apr 02 20:24:03 2012 Apr 02 20:33:05 Hello Apr 02 20:35:00 I recently set up an OE environment (trying to learn OE actually), and the first package I'm trying to build is failing miserably Apr 02 20:35:25 oe-core or oe classic? Apr 02 20:35:37 classic, I suppose Apr 02 20:35:51 if you are starting new, I suggest you start w/ oe-core.. Apr 02 20:36:01 I cloned the OE git repository Apr 02 20:36:05 then check out the documentation on the yocto project website.. Apr 02 20:36:25 http://git.openembedded.org/openembedded-core/ Apr 02 20:36:29 thats the repository you want.. Apr 02 20:36:36 the other is still maintained, but deprecated Apr 02 20:37:12 I'm actually using the jlime distro, is available in oe-core ? Apr 02 20:37:20 s/is/is it/ Apr 02 20:37:34 I don't know about that one, sorry Apr 02 20:39:49 apelete`: someone was working on that but I don't know how far they got - you should probably ping the jlime folks Apr 02 20:44:49 while trying to build tha package 'nano' I got an error about the svn.openmoko.org server being unreachable Apr 02 20:44:49 how do you resolve such issues ? Apr 02 20:44:49 is there a mirror somewhere ? Apr 02 20:47:53 here is the relevant part of error message: Apr 02 20:47:53 0: opkg-utils-native-0.1.8+svnr4747-r7 do_fetch (pid 9573) Apr 02 20:47:53 svn: OPTIONS of 'http://svn.openmoko.org/trunk/src/host/opkg-utils': could not connect to server (http://svn.openmoko.org) Apr 02 20:47:53 NOTE: package opkg-utils-native-0.1.8+svnr4747-r7: task Fetch failed: Unable to fetch URL svn://svn.openmoko.org/trunk/src/host/;module=opkg-utils;proto=http from any source.: Failed Apr 02 20:47:56 Apr 02 20:48:24 how can I resolve this ? Apr 02 20:48:25 ya, the upstream location for that item has gone away Apr 02 20:48:37 post to the oe mailing list and ask if anyone has suggestions Apr 02 20:48:49 ok Apr 02 20:53:00 what's the difference between oe and oe-core btw ? Apr 02 20:53:49 I do not know much about oe, and even less about oe-core :) Apr 02 20:56:15 opkg-utils is hosted on yp.org now Apr 02 20:56:28 so just use that location for SRC_URI Apr 02 20:56:32 http://git.yoctoproject.org/cgit/cgit.cgi/opkg-utils/ Apr 02 20:57:04 apelete`: http://www.openembedded.org/wiki/OpenEmbedded-Core Apr 02 20:57:06 and I guess, send a patch to the list with the SRC set to the new location, once you made sure it works for you. Apr 02 20:57:59 I guess at some point, someone will have to port jlime to oe-core. Boy I wish I'd have more time ... Apr 02 20:59:10 there is a meta-jlime, I don't think it's official though hence why it's not listed on the layer index page Apr 02 21:01:36 I can't see it by browsing the oe-core tree Apr 02 21:01:41 care to share a link? Apr 02 21:03:30 Dodji: https://github.com/tworaz/oe-tworaz Apr 02 21:04:13 it has a *lot* of recipes in it that should not be in a proper distro layer Apr 02 21:04:24 but I guess it's not meant to be yet Apr 02 21:04:31 thanks. **** ENDING LOGGING AT Tue Apr 03 02:59:58 2012