**** BEGIN LOGGING AT Mon Feb 04 02:59:59 2013 Feb 04 08:24:00 good morning Feb 04 10:40:52 morning all Feb 04 10:48:17 morning Feb 04 12:10:21 morning all Feb 04 12:23:58 # insmod myDriver.ko Feb 04 12:23:58 Error: could not insert module myDriver.ko: Invalid module format Feb 04 12:24:10 myDriver: : disagrees about version of symbol module_layout Feb 04 12:24:49 erros tried to load .ko driver Feb 04 12:26:12 but # uname -r Feb 04 12:26:13 3.4.26-yocto-standard Feb 04 12:26:13 root@jasperforest:~# modinfo myDriver Feb 04 12:26:40 .... depends: Feb 04 12:26:41 vermagic: 3.4.26-yocto-standard SMP preempt mod_unload modversions Feb 04 12:28:31 have anyone tried kernel loadable module .ko on yocto x86_64 target? Feb 04 12:51:00 RP: do_package changes are amazing, thanks! Feb 04 13:06:04 JaMa: You noticed a speedup? :) Feb 04 13:07:07 JaMa: How bad does the use of RDEPENDS and friends look in the metadata in general? I haven't got a good feel for how many issues we have in that area and I know you have a larger number of layers than I usually use Feb 04 13:07:14 not yet, but with do_packagedata and updated rm_work I have only about 1GB in work directory now :) Feb 04 13:07:37 JaMa: ah, this is the last round? :) Feb 04 13:08:54 RP: I think RDEPENDS and friends are not so bad in layers, I was doing git grep from time to time to fix entries without _PN Feb 04 13:09:30 JaMa: That is good to know. There is a patch in master-next which shows them up as warnings. If you do test with it I'd be interested to know how bad the problem is Feb 04 13:10:09 It seems to be a good time to clean some of that stuff up though Feb 04 13:13:02 I just went through the meta-openembedded repo this morning, only found three recipes that had an issue and sent patches for all of them Feb 04 13:19:04 bluelightning: good, looks like OE-Core is actually worse than others then (and I have patches for it_ Feb 04 13:26:52 bluelightning: and all are already included in in-test branch, I plan to push them after world build is finished Feb 04 13:28:13 JaMa: great, thx Feb 04 13:49:10 RP: I think there are still some issues when some recipe changes PACKAGE_ARCH, I've just moved one package from MACHINE_ARCH to TUNE_PKGARCH and I had to remove shlibs and pkgdata from "old" version manualy to fix do_rootfs (.ipk still RDEPENDing on non-existent libnl packages after .so name changes from last 2 libnl upgrades) Feb 04 13:50:13 RP: I think it's because -c cleansstate removes pkgdata and shlibs from TUNE_PKGARCH, but stale version in MACHINE_ARCH directories is kept Feb 04 13:53:29 JaMa: that makes sense. Not sure how we solve that either :/ Feb 04 14:53:10 Hi everybody, hi otavio. I've seen that here http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=i.MX6Q&fpsp=1&tab=Design_Tools_Tab is present a version of the Vivante SDK based on BSP 1.1 with updated Vivante drivers with vprofiler enabled. Does anybody know if there is any difference with 12.09.01 BSP drivers? Feb 04 14:53:29 I get a lot of Segmentation faults with the old drivers Feb 04 15:04:30 RP: ah it was also caused by stale .ipk file in MACHINE_ARCH feed and do_rootfs using --prefer-arch-to-version :/ Feb 04 15:13:47 JaMa: somehow we need to figure out it changed arch and remove the stale artefacts :/ Feb 04 16:13:29 how do you set up the root filesystem for nfs? after a while, mine loses its mind and I get things like 'ls: command not found' Feb 04 17:27:46 RP: "insane.bbclass: Add pkgvarcheck to check for suboptimal usages of variables" is missing RPROVIDES and RCONFLICTS, is it intentionally? Feb 04 17:28:25 RP: and RREPLACES Feb 04 17:30:47 and RSUGGESTS :), added all 4 and running with all my layers Feb 04 17:31:12 JaMa: no, we should add those. I was just starting off small Feb 04 17:32:23 ok good Feb 04 17:34:24 97 ERRORs, not so bad Feb 04 17:36:57 RP: only tricky part is ttf-fonts, where I had to duplicate a lot of RDEPENDS http://git.openembedded.org/openembedded/commit/recipes/ttf-fonts?id=14c265b107725a8f7534dfb354ad12edf9f7aec3 now the same for pkg_post* Feb 04 17:38:09 JaMa: yes, not ideal but will fix several packages too (e.g. -dbg package will lose the postinstall) Feb 04 17:38:28 true Feb 04 17:39:05 JaMa: you can also indirect it through a variable if that helps Feb 04 17:47:20 RP: I've created update-fonts.bbclass to do postinst/postrm and RDEPENDS Feb 04 17:47:43 JaMa: sounds good Feb 04 19:08:40 how do you set up the root filesystem for nfs? after a while, mine loses its mind and I get things like 'ls: command not found Feb 04 19:25:57 Host distribution "Ubuntu 12.04.2 LTS" has not been validated with this version of the build system Feb 04 19:26:09 I think this should be fixed Feb 04 19:26:36 I have been building with 12.04.2 for a while now Feb 04 19:33:56 khem: send a patch ;) Feb 04 20:19:08 Hi all, I'm trying to add a recipe for a python module that I'm modeling after python-epydoc_3.0.1.bb. When I try to build my image I get an error that my package needs /usr/bin/python-native/python Feb 04 20:19:24 I'm not quite sure wy Feb 04 20:19:30 w/wy/why/ Feb 04 20:21:07 err, I should have probably asked in #oe Feb 04 20:33:33 which option turns of the default console? Feb 04 20:35:29 *off Feb 04 20:37:29 jstashluk: you mean /dev/ttyS0? Feb 04 20:38:27 Garibald1: no I meand /dev/tty0, I don't want one to show up on my HDMI. my boot args already enable console=ttymxc1. Feb 04 20:39:16 jstashluk: you can try adding USE_VT = "0" Feb 04 20:39:30 I added that to my local.conf Feb 04 20:40:00 the sysvinit recipe looks at that variable to determine whether or not to add /dev/tty entries to inittab Feb 04 20:40:27 (*disclaimer: I don't know if that's the "right" thing to do; it's simply something that I've done) Feb 04 20:40:28 Garibald1: thanks, I'll try that Feb 04 21:29:04 If anyone has any ideas about my /usr/bin/python-native/python issues, I'd still appreciate some help :) Feb 04 22:47:25 khem: sakoman: ping Feb 04 22:49:21 I know it's old stuff but I confirm the actual tc is miscompiling 2.6.31 kernels. Same kernel boots if built with oe-classic Feb 04 23:18:43 msm: ping **** ENDING LOGGING AT Tue Feb 05 02:59:59 2013