**** BEGIN LOGGING AT Mon Oct 31 02:59:57 2011 Oct 31 08:53:03 morning all Oct 31 09:23:18 morning Oct 31 09:26:48 hi JaMa Oct 31 09:54:05 hi JaMa, bluelightning Oct 31 09:54:13 hi RP__ Oct 31 09:58:47 hi :) Oct 31 09:59:15 RP__: do you want me to send PR bumps for python deps now when it's usable? (ie for testing from master-next)? Oct 31 10:01:58 JaMa: sure, if it works now I'll put it in master-next Oct 31 10:02:14 JaMa: did the arm issues get fixed too? Oct 31 10:02:26 yes Oct 31 10:03:03 and runtime issues seems also resolved (it was because another subpackages missing few .so due to linking errors during build) Oct 31 10:04:35 JaMa: great! :) Oct 31 11:52:12 someone seen "Cannot register service: RPC: Unable to receive; errno = Connection refused" while starting qemu with runqemu? full log http://paste.pocoo.org/show/500772/ Oct 31 11:57:31 hi ant_work Oct 31 12:04:11 helo bluelightning Oct 31 13:59:12 hi, I'm wondering how to deal with LIC_FILES_CHECKSUM for an image recipe (whcih does not have a license file or so), already tried to give an empty var but still get: ERROR: Licensing Error: LIC_FILES_CHKSUM does not match, please fix Oct 31 13:59:54 this is for edison, Oct 31 14:00:35 1) its LIC_FILES_CHKSUM not LIC_FILES_CHECKSUM Oct 31 14:01:00 2) see task or image recipes for example, usually they poinst to COREBASE mit license Oct 31 14:04:04 JaMa: thanks, actually I indeed used CHKSUM in my recipe, I peeked into recipes-core/images but no one in there sets LIC_FILES_CHKSUM (but they do set LICENSE to MIT, I set LICENSE to unknown for now, can try MIT if needed Oct 31 14:05:48 eFfeM_work: LIC_FILES_CHKSUM = "file://${COREBASE}/LICENSE;md5=3f40d7994397109285ec7b81fdeb3b58 Oct 31 14:06:35 ah thanks *but where dit you get that from ? Oct 31 14:06:48 eFfeM_work: or inherit core-image as recipes-core/images does Oct 31 14:07:35 will try that, thanks Oct 31 14:07:52 LIC_FILES_CHKSUM = "file://${COREBASE}/meta/COPYING.MIT;md5=3da9cfbcb788c80a0384361b4de20420" Oct 31 14:08:05 ^ is better or include both to LIC_FILES_CHKSUM as core-image does Oct 31 14:10:07 decided as a backburner project to see if I can get my current project build with 1.1 Oct 31 14:10:59 with inherit it is much better :-) thanks JaMa Oct 31 14:15:43 we have a session now Oct 31 14:15:51 'toolchain and yocto' at Linaro Connect Oct 31 14:15:51 hm: WARNING: QA Issue: No GNU_HASH in the elf binary: '/home/frans/edison-6.0-build/tmp/work/ppc603e-poky-linux/libgcc-4.6.1+svnr175454-r10/packages-split/libgcc/lib/libgcc_s.so.1' Oct 31 14:27:16 hrw: I'll be interested to hear how it goes Oct 31 14:29:38 almost there, but get three lines like this: | /bin/sh is needed by dropbear-0.52-r3.ppc603e Oct 31 14:30:13 my image does include busybox which I would expect to provide sh (but I used the default recipe with the default config) Oct 31 14:30:30 eFfeM_work: that means noting in your image provides /bin/sh. busybox should be providing that Oct 31 14:34:19 RP__: my idea, but busybox is in my image recipe Oct 31 14:35:10 eFfeM_work: are you 100% sure busybox is getting in there? Oct 31 14:35:54 #ubuntu-uds-grand-sierra-g if someone want to attend Oct 31 14:38:01 RP__: fragmetn from image file: http://pastebin.com/gHLCAwAE Oct 31 14:38:06 checing the work dir Oct 31 14:39:33 hm, busybox is not build thought things in IMAGE_INSTALL automatically made it into DEPENDS Oct 31 14:42:35 btw handbuildign busybox gives two warnings: Oct 31 14:42:36 WARNING: busybox: No generic license file exists for: bzip2 at /home/frans/poky-edison-6.0/meta/files/common-licenses Oct 31 14:42:36 WARNING: busybox: There is also no SPDXLICENSEMAP for this license type: bzip2 at /home/frans/poky-edison-6.0/meta/files/common-licenses Oct 31 14:43:10 eFfeM_work: those aren't anything to worry about in connection with this problem Oct 31 14:43:38 RP__: I know, just wanted to report these (like the GNU_HASH I reported earlier) Oct 31 14:44:21 RP__: anyway, made progress. adding this: Oct 31 14:44:22 DEPENDS += ${IMAGE_INSTALL} Oct 31 14:44:22 RDEPENDS += ${IMAGE_INSTALL} Oct 31 14:44:41 to the recipe did NOT help, but bitbake busybox then build my image DID work Oct 31 14:44:58 not really sure how I should have done this though Oct 31 14:46:29 eFfeM_work: something odd is going on with your image recipe then Oct 31 14:46:57 RP__: yeah, guess so, will peek into it Oct 31 14:49:05 I will try a clean build first; had a disk full halfway so probalby a good idea anyway Oct 31 16:03:13 hi , is there any low cost and good quality SMT Machine for prototype development board ? Oct 31 16:50:19 fray: ping Oct 31 19:40:01 I'm working on some IA32 x32 ABI stuff.. Anyone here familiar with that -- and/or ELF? Oct 31 19:40:10 trying to determine why something isn't matching what I'm expecting Oct 31 20:44:21 fray: ping Oct 31 20:45:47 hey Oct 31 20:48:29 Anyone happen to know what version of oe-core was included in 1.1/edison? think that branch was before the start of using the combo-layer tool, so no .conf to inspect. thinking about generating a git object for the meta tree and then searching oe-core commits for that tree Oct 31 20:49:53 kergoth: I believe it was 2011-1 Oct 31 20:50:25 Kergoth: Tag and Branch Oct 31 20:50:34 ah, right Oct 31 20:50:36 okay, thanks Oct 31 21:49:43 Has anyone messed with combining meta-toolchain and external-toolchain thus far? E.g. creating an external-toolchain-crosssdk which extracts the actual cross toolchain bits from external and packages them for re-deployment into meta-toolchain along with other bits? Oct 31 21:49:50 I'm going to guess no :) Oct 31 21:50:54 kergoth: Not that way around I don't think Oct 31 21:51:50 would think it'd be doable, figured I'd see if anyone tried it before I do :) Oct 31 22:59:54 NOTE: package gcc-runtime-nativesdk-4.6.1+svnr180099-r18: task do_compile: Failed Oct 31 22:59:57 * kergoth digs **** ENDING LOGGING AT Tue Nov 01 02:59:57 2011