**** BEGIN LOGGING AT Thu Oct 11 02:59:58 2012 Oct 11 08:00:47 what is the difference between oe-core denzil branch and git://git.yoctoproject.org/yocto denzil branch ? Oct 11 08:03:48 rob_w: the later does not exists? Oct 11 08:04:39 http://git.yoctoproject.org/cgit/cgit.cgi/poky/log/?h=denzil Oct 11 08:04:41 ? Oct 11 08:08:02 rob_w: that's different repo then what you said first.. Oct 11 08:09:08 oh ok Oct 11 08:09:33 rob_w: and it's basically the same, poky repo is merging multiple repositories (oe-core, meta-yocto, bitbake) together while with oe-core you 1 checkout for each layer Oct 11 08:10:36 ok thx Oct 11 08:10:56 something like git submodules but created with combo-layer tool Oct 11 08:11:37 so if i take poky i dont need oe-core anymore Oct 11 08:22:46 o.O Oct 11 08:22:54 poky includes oe-core Oct 11 08:23:45 rob_w: and do you need poky or oe-core + bitbake? :) Oct 11 08:24:35 well Oct 11 08:25:12 i do need a a chain which is stable or frozen as possible Oct 11 08:25:43 i got myself a repo config which pulls poky and some meta layers together to get a easy to setup bsp toolchain for our boards Oct 11 08:26:36 and i am a long year oe-classic user and just make my baby steps into oe-core, so i am still not familiar with the sorting etc Oct 11 08:27:25 if you're maintaining other layers with some script then I don't see any advantage of poky instead of just oe-core + bitbake Oct 11 08:27:52 unless you want to use poky.conf as DISTRO (oe-core could be used without DISTRO) Oct 11 08:30:42 yes i want to be on poky.conf .. as i understood its kinda frozen in the denzil release , right ? Oct 11 08:59:22 does one know where qt4-x11-free-gles recipes live as of oe-core ? Oct 11 09:36:03 what is supposed to provide the keymap ? Oct 11 09:36:53 if i understand well the startup scripts, for a 2.6 kernel, it will look at /etc/keymap-2.6.map Oct 11 09:36:56 afournier: recipes-bsp/keymaps/keymaps_1.0.bb Oct 11 09:37:15 JaMa: i meant the .map keymap Oct 11 09:37:46 * JaMa too Oct 11 09:38:22 i don't see any files except GPLv2.patch keymap.sh Oct 11 09:38:36 BSP layer should have .bbappend Oct 11 09:38:48 ok Oct 11 09:38:53 meta-nokia and meta-handheld has Oct 11 09:39:02 ok :p Oct 11 09:39:14 it's meant to be overloaded Oct 11 09:39:16 good Oct 11 11:53:42 When I run bitbake helloworld-image I get LOTS of errors, starting with Oct 11 11:53:45 ERROR: Error in executing: /home/john/Joc/OEBASE/openembedded/recipes/prismstumbler/prismstumbler_0.7.3+0.7.4pre1.bb Oct 11 11:53:45 ERROR: Exception: Message:list indices must be integers, not NoneType Oct 11 11:53:45 ERROR: Printing the environment of the function Oct 11 11:53:45 Oct 11 11:53:54 What am I doing wrong? Oct 11 12:11:42 jmd: probably too new bitbake for old oe-classic Oct 11 12:12:52 is current bitbake to build rootfs without internet connection ? Oct 11 12:12:59 *able Oct 11 12:13:19 because 3 months ago i have tryed that but without success Oct 11 12:16:03 CruX|: if you have done -c fetchall first it should be able to yes Oct 11 12:17:11 bluelightning: before 3 months I have failed on git update cmd Oct 11 12:17:23 so wondering if now it is working Oct 11 12:17:31 CruX|: did you also set BB_NO_NETWORK = "1" / Oct 11 12:17:33 ? Oct 11 12:17:39 CruX|: if there is some recipe with tag param then it still does not work Oct 11 12:17:49 right, was just about to point that out Oct 11 12:18:09 (or any recipe with SRCREV pointing to something that is not a full sha1 hash in the case of git) Oct 11 12:18:19 ah :/ Oct 11 12:36:59 thanks for helping jmd out - which version of bitbake would be recommended? Oct 11 12:37:45 or where would one be able to get non-oe-classic openembedded files from (which git branch on git.openembedded.org, whatever) Oct 11 12:37:46 tia Oct 11 12:44:19 lkcl: current work happens in OE-Core which is in the openembedded-core repository Oct 11 12:44:27 overview: http://www.openembedded.org/wiki/OpenEmbedded-Core Oct 11 12:45:40 so if you're working on something new, we would strongly recommend basing it on OE-Core Oct 11 13:00:07 when i runqemu, it sets the ip address thru the kernel cmdline ip=192.168.7.2::192.168.7.1:255.255.255.0, but at the end, when it finishes booting, eth0 is down and the ip address is not set, i can still "ifconfig eth0 192.168.7.2" etc. but i would like to know what is missing, so it the ip address could be set automatically Oct 11 14:16:09 i built a test root filesystm for panda board. Now the root of the file system does not have a /linuxrc file. what boot arguments should i give Oct 11 15:59:29 does anyone know if an AMD 16-core is better than an Intel quad core for OE builds? Oct 11 16:01:40 * mwester usually runs out of memory before running out of cores, so /me suspects that the answer depends on more than just the number of cores. Oct 11 16:02:41 I have an older 6-core xeon with 12GB RAM, which does pretty well (I can max out the CPU), but customer is asking what he should buy Oct 11 16:03:57 My *general* (not-OE-specific) experience with many-core build systems is that for supposedly-equivalent AMD vs Intel CPUs, it takes 1.5 to 2 AMDs to match the Intel (disclaimer: all figures come from VMware environments, so some of the difference may be virtualization rather than build factors). Oct 11 16:05:34 there's also the issue of resource usage by bitbake. it definitely won't keep that many cores pegged for a majority of the build, from what i've seen, anyway Oct 11 16:05:37 * kergoth shrugs Oct 11 16:06:31 I also underscore for my customers that they need to ignore the "hyperthreading" pseudo-cores on both AMD and Intel, in order to compare them; my build experience is that an Intel hyperthreaded core is actually only worth 1/10 of a real core... Oct 11 16:32:43 kergoth: here, with BB_NUMBER_THREADS=4 I see a parallelization of 2.5 (build takes Oct 11 16:33:47 and 1446.35812 (stime) + 5795.033021 (utime) Oct 11 16:38:19 another project gives a rate of nearly 3: http://www.cvg.de/people/ensc/metrics.txt.gz Oct 11 16:40:23 first time played with hob - looks nice Oct 11 17:08:51 hi Oct 11 17:10:25 hi chouimat **** ENDING LOGGING AT Fri Oct 12 03:00:01 2012