**** BEGIN LOGGING AT Thu Jun 14 02:59:58 2012 Jun 14 04:29:48 hello all Jun 14 07:44:07 hi. someone used a javavm with ARM? Jun 14 08:06:55 which package / recipe does bring libpthread Jun 14 08:17:34 good morning Jun 14 08:17:44 ericbutters: yes Jun 14 08:20:49 mckoan hi. nice :) which one? and how did you installed it? and on what kind of ARM? armv4/5/6/7? Jun 14 08:24:34 rob_w: libpthread is usually shipped with libc on desktop distributions Jun 14 08:24:55 so maybe you should look at stripped down version of libc in oe Jun 14 08:25:06 like uclibc, I guess. Jun 14 08:25:46 sorry , yes its shipped already .. thx Jun 14 08:29:35 ericbutters: java version "1.6.0_24" - IcedTea Runtime Environment (IcedTea6 1.11.1) (6b24-1.11.1) - CACAO (build 1.1.0pre2, compiled mode) Jun 14 08:29:49 ericbutters: I use KaeilOS branch http://www.kaeilos.com/latest-changes Jun 14 08:33:03 ericbutters: CPU: ARM926EJ-S [41069265] revision 5 (ARMv5TEJ) Jun 14 08:46:58 mckoan great, thx for information. Jun 14 10:05:15 hi all Jun 14 10:56:07 hello, is there a preferred way of "disabling" a task? I have a recipe that installs an example application into a SDK, and I don't want it compiled, just installed. How would I best disable do_compile? Jun 14 11:14:05 jonasdn: just define: Jun 14 11:14:07 do_compile() { Jun 14 11:14:09 : Jun 14 11:14:11 } Jun 14 11:14:27 in your recipe Jun 14 11:14:39 ok, thanks, what does the ':' signify? Jun 14 11:21:57 jonasdn: it signifies to the shell that it's non-empty, I don't know much more beyond that Jun 14 11:22:21 (since this is a shell function) Jun 14 11:23:07 ok, thank you! Jun 14 11:26:07 Google says: "Historically, Bourne shells didn't have true and false as built-in commands. true was instead simply aliased to :, and false to something like let 0." Jun 14 11:40:34 morning all Jun 14 12:23:17 hi pb_ Jun 14 12:40:01 hello lautriv Jun 14 13:56:43 ok, so I have a problem I can't really seem to solve. I have a recipe that has BBCLASSEXTEND = "nativesdk", and when I bitbake it with -nativesdk suffix I get warnings about files installed but not shipped in any package. Jun 14 13:56:55 like /usr, /usr/bin, /usr/bin/ Jun 14 13:57:20 when I do FILES_${PN}_virtclass-nativesdk += "${bindir}/*" Jun 14 13:57:22 it doesn't help Jun 14 13:57:57 but if I do FILES_${PN}_virtclass-nativesdk += "/usr/bin/*" Jun 14 13:58:00 it seems to work Jun 14 13:58:07 is that intendid behavior? Jun 14 14:08:17 hi Jun 14 14:08:38 is it correct to write string in PR? Jun 14 14:08:55 Noor: hi, it's allowed yes Jun 14 14:08:57 like PR = "init-dev" Jun 14 14:09:08 hi bluelightning Jun 14 14:09:13 how r u doing Jun 14 14:09:20 Noor: good and you? Jun 14 14:09:26 I am good as well Jun 14 14:10:16 bluelightning: if it is correct then how bitbake compares it to find that PR has been bumped Jun 14 14:10:53 Noor: there's some version comparison code in bb.utils IIRC Jun 14 14:11:14 takes PE, PV, and PR into account Jun 14 14:11:46 bluelightning: thanks for the pointer Jun 14 15:13:54 first epoch is compared then PV and last is PR Jun 14 15:23:34 does anyone know of any python 3.2 recipes available that i could backport to stable-2009? Jun 14 17:03:21 Did a fresh install of bitbake & openembedded on an Ubuntu 12.04 virtual machine (VMWare 4.1.3, on OS X 10.7.3, two cpu's and 4G memory) - when I do it starts parsing recipes, gets to about 3%, and the machine freezes - 100% cpu utilization, no response in the windowing system, and a terminal ssh'ed in from OS X has no command line response... and the the OS X Activity monitor shows both of the cpu's I've allocated running at 100%... and t Jun 14 17:06:08 I reran after rebooting, with , and the last line to appear was: "DEBUG: LOAD /mnt/hgfs/Linux-extras/stuff/openembedded/classes/relocatable.bbclass", where is the dir I am installed in; it is a shared directory through VMWare, it's part of the OS X file system. Jun 14 17:06:14 Any suggestions? Jun 14 17:20:10 Hi. Does anybody have an xf86-video-ati recipe at hand? :-) Jun 14 17:23:16 Hmmmmm. perhaps a permissions issue, as it is a shared dir/tree. Jun 14 18:46:16 does someone know how a keys are mapped/evaluated in linux and how to adapt this , the translations from event device to X11 keys ? Jun 14 19:55:10 should lava-deployment-tool work on lucid? Jun 14 19:55:50 it looks like i need something newer Jun 14 19:57:37 nevermind, the readme says 10.10 not 10.04 ;) Jun 14 20:00:17 msm, they say also min 1GB ram, why ask you in OE ? Jun 14 20:01:26 lautriv: because i asked in the wrong channel Jun 14 20:01:30 whoops ;) Jun 14 20:01:50 i at least joined the right channel, just did not change to it to type ;) Jun 14 20:02:25 msm, np. just wondered how one would install it on an embedded thingie ;) Jun 14 20:03:06 heh, no trying to do that Jun 14 20:51:41 hello. Is it possible to build a toolchain for i686 when building on a x86_64 box? Jun 14 20:53:25 sorvats, Jun 14 20:53:40 that's most likely the job of a toolchian. Jun 14 20:56:31 heh, you are right :) but well, i was talking about oe itself. And it looks possible as vanilla x86_64 gcc can yield i686 binaries, right? Jun 14 20:57:19 I was hoping there was a way to tell oe the architecture of, for example meta-toolchain-qte Jun 14 20:57:40 wouldnt that be possible with oe? Jun 14 20:58:04 sorvats, i686 is a subset of x86_64 so it should work if you enter the right CFlags, never tried/needed myself. Jun 14 21:02:52 ah, so it would be a matter of changing CFLAGS. doesnt look like a safe way hehe. maybe a chroot or something Jun 14 21:21:20 sorvats: hang on, define your needs clearly Jun 14 21:21:41 sorvats: what is the target you are intended to generate code for Jun 14 21:22:07 OE spins out cross toolchain SDKs but it can generate them in variety of different ways Jun 14 21:22:53 so I am not clear what your host machine would be ( machine on which compiler itself will run) and what your target is ( machine on which the resulting code from compiler will run ) Jun 14 21:27:58 ok Jun 14 21:28:34 I need to generate a meta-toolchain-qte which will run on i686 hosts Jun 14 21:29:16 But my OE is installed on a x86_64 host Jun 14 21:30:32 just issuing bitbake meta-toolchain-qte will produce x86_64 binaries for my toolchain Jun 14 21:31:13 khem: is that clear enough? Jun 14 21:32:10 the target architecture of the toolchain is always armv5te Jun 14 21:35:13 ok so build=x86_64 host=i686 target=arm Jun 14 21:35:23 this is a cross-canadian case Jun 14 21:35:28 evening all Jun 14 21:35:30 hi florian Jun 14 21:35:55 sorvats: set SDKMACHINE="i686" in your local.conf and you must be all set Jun 14 21:36:16 hi pb_ Jun 14 21:36:31 khem: ok, tryingthat, thanks! :D Jun 14 21:36:42 hey pb_ Jun 14 21:38:28 hi khem Jun 14 22:03:54 pb_: how is mips processor treating you ? Jun 14 22:04:01 anything interesting Jun 14 22:13:42 mips itself is ok, but we continue to be plagued by annoying hardware problems Jun 14 22:14:10 ddr3 at 1066MHz is a bit of a nightmare, plus we suddenly seem to be having power supply failures all over the place. Jun 14 22:15:30 had one unit today where the +1.5V dc/dc had stopped switching and was basically just outputting +5V all the time. Jun 14 22:15:49 to my slight surprise the DRAM did still work (sort of) under those conditions but it did get rather hot. Jun 14 22:20:46 eeks Jun 14 22:43:43 yeah, indeed, all bad Jun 14 22:58:43 how can i use variables from the system environment inside a .conf and .bb files? **** ENDING LOGGING AT Fri Jun 15 02:59:59 2012