**** BEGIN LOGGING AT Mon Nov 07 02:59:57 2011
Nov 07 03:31:06 <mag> hi
Nov 07 06:14:33 <davidlt> sakoman: ping
Nov 07 07:20:06 <tasslehoff> Is "git://www.sakoman.com/git/linux-omap-2.6;protocol=git" the right thing to put in a recipe. I seem to recall something about git:// vs http:// and protocol=git/http that confused me the last time I tried, and its not working for me now.
Nov 07 07:21:50 <tasslehoff> oh, but suddenly it does. nevermind
Nov 07 07:45:09 <snkt> how to compile rdesktop-1.7 in oe.....??
Nov 07 08:38:21 <Ivan___> Greetings to all! Help to collect HelloWorld-image. I face an error at the assembly libtool-native
Nov 07 08:44:04 <mckoan> good morning
Nov 07 09:07:55 <Ivan___> Help to collect HelloWorld-image.
Nov 07 09:18:09 <Jin^eLD> mornin
Nov 07 09:57:07 <Ivan___> Greetings to all! Help to collect HelloWorld-image. I face an error at the assembly libtool-native
Nov 07 10:00:52 <Ivan___> http://paste.org/pastebin/view/40670
Nov 07 10:49:15 <davidlt> sakoman_: ping
Nov 07 12:25:55 <snkt> I have compiled a rdesktop using OE.... but every time I run rdesktop I m getting ...this message.... rdp_out_unistr: iconv_open[ANCI_X3.4-1968 -> UTF-16LLE] fail 0xffffff
Nov 07 12:26:06 <snkt> can anyone help me what is it...?
Nov 07 12:26:10 <snkt> and why??
Nov 07 12:52:09 <tasslehoff> anyone looking into the "ERROR: Nothing PROVIDES 'base-passwd-nativesdk'" that comes when trying to build SDKs?
Nov 07 12:53:10 <JaMa> yes, see ML
Nov 07 13:21:41 <otavio> Is BAD_RECOMMENDATIONS working in oe-core? it seems not working for me
Nov 07 13:22:09 <JaMa> ot working for me too
Nov 07 13:22:13 <JaMa> not
Nov 07 13:23:47 <otavio> ufa
Nov 07 13:23:49 <otavio> heh
Nov 07 13:28:30 <otavio> JaMa: do you found why?
Nov 07 13:32:08 <JaMa> otavio: no
Nov 07 13:32:24 <JaMa> otavio: in the end I've dropped RECOMMENDATIONS by .bbappend in my layer
Nov 07 13:32:26 <JaMa> :/
Nov 07 13:32:37 <otavio> JaMa: right 
Nov 07 13:32:43 <otavio> JaMa: but seems wrong
Nov 07 13:32:45 <otavio> JaMa: heh
Nov 07 13:33:03 <JaMa> otavio: http://git.shr-project.org/git/?p=meta-smartphone.git;a=commit;h=6407d41a32e585a7b1892535191ee7b3e7b56e6a
Nov 07 13:33:26 * JaMa thinks there is more wrong stuff now :/
Nov 07 13:37:23 <otavio> JaMa: I just checked and it seems opkg is overwritting the status and thus filling it with the not installed seems useless
Nov 07 13:57:37 <Crofton> Does this look familiar to anyone: (from oe-classic)
Nov 07 13:57:38 <Crofton> ERROR: Error parsing /home/fperich/Ettus/oe/sources/openembedded/recipes/ccid/ccid_1.3.8.bb: 'module' object has no attribute 'init'                                             | ETA:  --:--:--
Nov 07 13:57:38 <Crofton> > ERROR: Command execution failed: Exited with 1
Nov 07 14:46:02 <KotH> moin
Nov 07 15:18:04 <mckoan> KotH: moin
Nov 07 15:27:39 <doragasu> Hi
Nov 07 15:27:51 <Jin^eLD> a .bbappend can not override the parents recipe pkg_postinst_${PN}, right ?
Nov 07 15:28:17 <doragasu> I'm trying to build Angstrom distro for Sheevaplug, and I can't build the kernel (linux_2.6.37.bb)
Nov 07 15:29:04 <doragasu> I get a 404 error when trying to dowload http://www.angstrom-distribution.org/unstable/sources/patch-2.6.37.6.bz2
Nov 07 15:29:52 <doragasu> Is there anything I can do besides waiting for the file to get online?
Nov 07 15:30:36 <bluelightning> Jin^eLD: yes, it should be able to
Nov 07 15:30:56 <Jin^eLD> bluelightning: so defining my own in the .bbappend will simply overwrite the parents one? right?
Nov 07 15:31:10 <bluelightning> Jin^eLD: that's the way it should work, yes
Nov 07 15:31:33 <Jin^eLD> ok thanks, just wanted to be sure :)
Nov 07 15:31:54 <bluelightning> np
Nov 07 15:33:18 <mckoan> doragasu: try adding KERNELORG_MIRROR = "http://mirror.nexcess.net/kernel.org"
Nov 07 15:33:49 <doragasu> Thanks mckoan, I'm trying it right now
Nov 07 15:36:32 <doragasu> It's working. Again thanks a lot mckoan.
Nov 07 15:37:01 <mckoan> doragasu: you're welcome
Nov 07 15:37:12 <denix> cbrake: yes, wiki is working
Nov 07 16:43:58 <Jin^eLD> hmm, what would prevent bitbake from finding my conf/local.conf?
Nov 07 16:44:08 <Jin^eLD> I believe I set the BBPATH correctly
Nov 07 16:44:32 <Jin^eLD> still, it's not looking where it should...
Nov 07 16:45:18 <pack3754> can you post your BBPATH and the location of your build/conf/local.conf file
Nov 07 16:49:02 <pack3754> hey all: a dependency of task-gstreamer-ti is failing
Nov 07 16:49:18 <pack3754> this is python-numpy_1.4.1.bb
Nov 07 16:49:46 <Jin^eLD> ok one second
Nov 07 16:50:32 <pack3754> It seems like the root of the problem is that it tries to link against my host system libraries and not the ones compiled for arm arch
Nov 07 16:51:14 <pack3754> in the recipe file, there is a comment from koen
Nov 07 16:51:18 <pack3754> # Make the build fail and replace *config.h with proper one
Nov 07 16:51:18 <pack3754> # This is a ugly, ugly hack - Koen
Nov 07 16:52:07 <pack3754> buuuut... i'm not sure which config.h he is referring to (the "proper one" or the one that should be replaced)
Nov 07 16:55:04 <Jin^eLD> pack3754: what am I missing? http://pastebin.mozilla.org/1376665
Nov 07 16:58:30 <pack3754> Jin^eld: That looks valid. How do you know it's not loading the correct local.conf file? And do you know if it is loading a file from a different location instead?
Nov 07 17:01:05 <Jin^eLD> one sec, another paste coming
Nov 07 17:03:13 <Jin^eLD> http://pastebin.mozilla.org/1376666
Nov 07 17:06:42 <Jin^eLD> it seems to find all the layers, which means it is parsing the bblayers.conf
Nov 07 17:14:39 <pack3754> Did you note lines 152-155 of your pastebin?
Nov 07 17:14:54 <Jin^eLD> of course, and that is what confuses me most
Nov 07 17:15:01 <Jin^eLD> this tells me that it ignored my local.conf completely
Nov 07 17:16:30 <Jin^eLD> first of all, the suggested: "meld conf/local.conf conf/local.conf.sample" is a good way to visualise the changes.
Nov 07 17:16:48 <Jin^eLD> would not work because the sample is in a different place anyway
Nov 07 17:16:53 <Jin^eLD> but my local.conf looks OK
Nov 07 17:17:02 <Jin^eLD> and defines MACHINE and DISTRO
Nov 07 17:17:08 <pack3754> hmmm... ok
Nov 07 17:17:19 <Jin^eLD> in the log I do not see that bitbake would try to load my local.conf at all
Nov 07 17:18:45 <pack3754> does it print out a build configuration? like this :
Nov 07 17:18:46 <pack3754> http://pastebin.mozilla.org/1376675
Nov 07 17:20:02 <pack3754> this would show up near the beginning of the bitbake <recipe> output
Nov 07 17:20:07 <Jin^eLD> no
Nov 07 17:20:13 <Jin^eLD> it did not
Nov 07 17:20:31 <Jin^eLD> it looks like it is trying to build pseudo first but does not even come close naer it
Nov 07 17:21:15 <Jin^eLD> can it be a problem with the bitbake wrapper script?
Nov 07 17:22:57 <pack3754> welll... in which directory did you initially checkout the openembedded tree?
Nov 07 17:23:32 <pack3754> inside that directory there should be a "recipes" dir
Nov 07 17:24:20 <Jin^eLD> I have a subdir called OE that contains all the layers, i.e. meta-xx/recipes*
Nov 07 17:25:06 <Jin^eLD> I setup the bblayers.conf accrodingly
Nov 07 17:25:13 <Jin^eLD> and you can see from the log that it does load those layers
Nov 07 17:27:46 <Jin^eLD> brb, gotta go to the store as long as it is open :P
Nov 07 17:27:48 <pack3754> hmm... I will think on your problem
Nov 07 17:28:21 <Jin^eLD> it's probably some env-var that I am missing
Nov 07 17:28:31 <Jin^eLD> I'll through everything again
Nov 07 17:28:35 <Jin^eLD> *go through
Nov 07 17:39:50 <pack3754> Jin^eLD: maybe check the bitbake.conf file
Nov 07 17:39:55 <pack3754> i.e. line 680
Nov 07 17:40:00 <pack3754> DEBUG: CONF conf/bitbake.conf:680: including conf/site.conf
Nov 07 17:40:00 <pack3754> DEBUG: CONF file 'conf/site.conf' not found
Nov 07 18:06:34 <Jin^eLD> re
Nov 07 18:06:53 <Jin^eLD> pack3754: but site conf should not be mandatory?
Nov 07 18:07:06 <Jin^eLD> or is it?
Nov 07 18:07:07 <pack3754> well, there's a similar line about local.conf
Nov 07 18:07:13 <pack3754> just grabbed the wrong line =P sorry
Nov 07 18:08:19 <pack3754> at line 45-46 of your pastebin
Nov 07 18:09:54 <Jin^eLD> well yes, the question is - why is it not found? shouldn't bitbake be looking in ${BBPATH}/conf/local.conf ?
Nov 07 18:10:30 <pack3754> well, that's what I would guess... I'm not using the newere openembedded-meta stuff though
Nov 07 18:10:47 <pack3754> I'm semi-stuck in oe-core land because of gumstix
Nov 07 18:11:05 <pack3754> could be the oe-meta is doing something slick behind the scenes
Nov 07 18:11:07 <Jin^eLD> that is exactly the problem that I have :) the classic stuff works fine, but since it is becoming deprecated I am kind of forced to update to the new stuff
Nov 07 18:11:58 <pack3754> looks like maybe the file "/dss-oe/OE/openembedded-core/meta/conf/bitbake.conf"
Nov 07 18:12:12 <pack3754> is trying to load all your config files from the same directory
Nov 07 18:12:34 <pack3754> i.e. /dss-oe/OE/openembedded-core/meta/conf/
Nov 07 18:12:46 <pack3754> try putting your local.conf there and see if it finds it
Nov 07 18:12:56 <pack3754> (not a fix, just a diagnostic)
Nov 07 18:13:57 <Jin^eLD> good idea
Nov 07 18:28:49 <martinambrose> Anyone built/run an uClibc image lately? On boot sbin/init fails due to missing register_frame_info function?
Nov 07 18:45:08 <Jin^eLD> hmm, it seems that when bitbake.conf is doing an include conf/local.conf, it is looking only in the own directory but not in the directories that are specified in BBPATH.. ?
Nov 07 18:53:45 <cenobyte_> hello
Nov 07 18:54:06 <cenobyte_> I am new to OpenEmbedded and have a strange error
Nov 07 18:55:09 <cenobyte_> anyone here?
Nov 07 18:56:37 <Jin^eLD> cenobyte_: you did not really ask your question :P
Nov 07 18:57:48 <cenobyte_> oh
Nov 07 18:58:03 <cenobyte_> ok sorry I did not know if anyone was here lol
Nov 07 19:00:18 <cenobyte_> Ok back now
Nov 07 19:00:32 <cenobyte_> well I am trying to bitbake a console-image for now
Nov 07 19:01:18 <cenobyte_> and I get the ERROR please set the persistent_dir or 'cache' variable.
Nov 07 19:01:46 <Jin^eLD> sounds like some environment misconfiguratin prior to building
Nov 07 19:01:55 <Jin^eLD> are you trying oe-classic or oe-core?
Nov 07 19:02:06 <cenobyte_> i just followed the wiki
Nov 07 19:02:16 <cenobyte_> how can I check
Nov 07 19:02:18 <cenobyte_> ?
Nov 07 19:02:19 <pack3754> can you post a link to the wiki you used?
Nov 07 19:02:50 <cenobyte_> http://www.openembedded.org/wiki/Getting_started
Nov 07 19:03:59 <Jin^eLD> that's the good old classic setup 
Nov 07 19:04:40 <cenobyte_> is that bad?
Nov 07 19:05:08 <Jin^eLD> not necessarily, depends on what your goals are
Nov 07 19:06:13 <cenobyte_> I am trying to build an OS for the Ben nanonote with fpu kernel and eventually E17
Nov 07 19:06:22 <cenobyte_> similar to Jlime
Nov 07 19:07:17 <Jin^eLD> you could try Koen's scripts and then base your changes on the jlime stuff if your machine is similar
Nov 07 19:07:18 <Jin^eLD> http://www.angstrom-distribution.org/building-angstrom
Nov 07 19:08:17 <cenobyte_> jlime is a supported distro and ben-nanonote is a known MACHINE
Nov 07 19:08:41 <Jin^eLD> ah, jlime is a distro
Nov 07 19:08:48 <Jin^eLD> I thought it was some machine variant or someting 
Nov 07 19:09:49 <cenobyte_> should I still base on angstrom?
Nov 07 19:10:01 <cenobyte_> so far almost no distros use E as a window manager
Nov 07 19:10:27 <Jin^eLD> angstrom is a distro, so if you want another distro I'd guess that you can not use angstrom then
Nov 07 19:10:52 <cenobyte_> Ya I am trying to do my own, but Jlime works well on my machine
Nov 07 19:11:07 <cenobyte_> I want just a few things different 
Nov 07 19:11:14 <Jin^eLD> well, I could point you to a git repo with some setup stuff that works for me, for oe-classic
Nov 07 19:11:28 <Jin^eLD> maybe that'd help you somehow
Nov 07 19:11:37 <cenobyte_> is that why I got the error?
Nov 07 19:11:54 <Jin^eLD> I think you get the error because of some misconfiguration somewhere
Nov 07 19:12:22 <Jin^eLD> it does not start building, right? you get this before it attempts to build anything?
Nov 07 19:12:39 <cenobyte_> yes, exactly
Nov 07 19:13:05 <Jin^eLD> well that sounds like you ar emissing something either in your local.conf or elsewher
Nov 07 19:13:33 <cenobyte_> i did not see a line in the sample conf that pertained to that though...
Nov 07 19:13:52 <Jin^eLD> are you setting TMPDIR somewhere?
Nov 07 19:16:56 <Jin^eLD> pack3754: it works when I put the conf directory right in the top, i.e. dss-oe/conf
Nov 07 19:17:00 <cenobyte_> no that line is commented out
Nov 07 19:17:22 <Jin^eLD> pack3754: so I guess this is something that is either related to the newer bitbake version or to those wrapper scripts
Nov 07 19:17:33 <cenobyte_> should I set the TMP to something
Nov 07 19:17:36 <Jin^eLD> cenobyte_: well, make sure it's not :> 
Nov 07 19:17:48 <Jin^eLD> yes, set it to the directory where your build should land
Nov 07 19:18:32 <cenobyte_> in my case would it be /stuff/build  
Nov 07 19:18:47 <cenobyte_> that's where my conf folder is
Nov 07 19:19:01 <Jin^eLD> it does not really matter as long as its writable
Nov 07 19:19:08 <Jin^eLD> the result of your build will go there
Nov 07 19:19:28 <Jin^eLD> so you can pick any dir you find appropriate
Nov 07 19:19:59 <cenobyte_> ok so that's where the binaries are to go?
Nov 07 19:20:39 <Jin^eLD> this is where stuff will be built, this is where the crosscompiler will go, the resulting binaries, rootfs images, etc
Nov 07 19:20:59 <cenobyte_> Well no wonder my build failed!!!
Nov 07 19:21:02 <cenobyte_> lol
Nov 07 19:22:16 <msm> pb_: ping
Nov 07 19:23:36 <pack3754> Jin^eld: hmm.. must've been a change somewhere along the way
Nov 07 19:23:45 <pack3754> sounds like you've got it working though
Nov 07 19:24:12 <Jin^eLD> pack3754: well, not quite, I figured out why my setup does not work, but that being said I do not want a generic conf directory in the top dir
Nov 07 19:24:25 <Jin^eLD> so now I am trying to make it work the way it did before
Nov 07 19:24:45 <pack3754> maybe just add dss-oe/build to your BBPATH ? 
Nov 07 19:25:15 <Jin^eLD> well that's the funny thing - dss-oe/targets/dss11-devel is in the BBPATH
Nov 07 19:25:19 <cenobyte_> just a sec
Nov 07 19:25:26 <Jin^eLD> but for some reason bitbake will only look for the conf in dss-oe/conf
Nov 07 19:25:52 <pack3754> cenobyte_ : using the classic OE, bitbake should create a 'tmp' directory where all compiled binaries (and a bunch of other stuff) goes
Nov 07 19:26:32 <pack3754> the build directory is for configuration files that control your build (e.g. setting the machine type, distro, etc)
Nov 07 19:27:06 <cenobyte_> I dont see the BBPATH option in the sample .conf file
Nov 07 19:27:12 <cenobyte_> should I add it?
Nov 07 19:27:22 <cenobyte_> I still get the same error
Nov 07 19:27:41 <cenobyte_> to set the PERSISTANT_DIR or CACHE variable
Nov 07 19:27:47 <pack3754> BBPATH should be set in the 'profile' (which you probably added to the end of your .bashrc ? )
Nov 07 19:28:16 <pack3754> is it possible bitbake is trying to create a tmp directory, but doesn't have the permissions?
Nov 07 19:28:51 <cenobyte_> ok just set that
Nov 07 19:28:57 <cenobyte_> it says you have to do it each time
Nov 07 19:29:05 <cenobyte_> on the wiki
Nov 07 19:29:30 <pack3754> cenobyte_ : correct, unless you add it to your .bashrc
Nov 07 19:29:46 <pack3754> .bashrc is executed everytime you open a new terminal session
Nov 07 19:30:01 <cenobyte_> i wont do that until i get a real build
Nov 07 19:30:07 <pack3754> ok
Nov 07 19:30:12 <cenobyte_> lol
Nov 07 19:30:47 <cenobyte_> i set bbpath
Nov 07 19:30:55 <cenobyte_> and still the same error
Nov 07 19:31:56 <cenobyte_> is there a PERSISTENT_DIR section in the conf file i need to add?
Nov 07 19:32:17 <cenobyte_> the manual did not have info on it
Nov 07 19:33:20 <cenobyte_> so no one else has had this error?
Nov 07 19:35:45 <cenobyte_> OK guys and girls check this out
Nov 07 19:35:47 <cenobyte_> https://groups.google.com/group/beagleboard/browse_thread/thread/c8bc66fd63476e75/2117862d9e4a4613?lnk=raot
Nov 07 19:35:50 <cenobyte_> that has the answer
Nov 07 19:37:08 <pack3754> i did not see the answer there.. but you have solved it?
Nov 07 19:38:34 <cenobyte_> you have to add a line to the conf file
Nov 07 19:38:49 <cenobyte_> PERSISTENT_DIR = "/stuff/persistent" 
Nov 07 19:40:47 <Jin^eLD> hmm I never came across this, to be honest
Nov 07 19:41:20 <cenobyte_> well i added the line and I still get the error
Nov 07 19:41:45 <cenobyte_> i mkdir'ed all the things it wanted like oetmp and persistent
Nov 07 19:41:50 <cenobyte_> brb
Nov 07 19:44:23 <cenobyte_> back
Nov 07 19:44:48 <cenobyte_> would I have a better time using oE core?
Nov 07 19:46:50 <Jin^eLD> cenobyte_: that depends.. you could probably get going quickly using the angstrom scripts
Nov 07 19:50:30 <cenobyte_> if i use newer bitbake the error is gone but a new one replaces it
Nov 07 19:50:45 <cenobyte_> fatal- the bbpath variable is not set
Nov 07 19:54:39 <cenobyte_> ok well this has been fun, but I will just have to hack Jlime for now
Nov 07 19:54:45 <cenobyte_> thanks for tryng
Nov 07 19:55:05 <cenobyte_> it is an odd error that all the people who've had it cannot seem to fix
Nov 07 19:58:32 <cenobyte_> bye
Nov 07 20:02:11 <msm> how do I make a recipe only build for -native?
Nov 07 20:02:32 <msm> or… if I'm using BBCLASSEXTEND it seems like the target version is available by default...
Nov 07 20:02:43 <msm> I just want the -native (and possibly -nativesdk)
Nov 07 20:04:04 <Jin^eLD> I just asked myself the same question :P
Nov 07 20:07:17 <msm> did you answer yourself?
Nov 07 20:07:21 <msm> hehe
Nov 07 20:11:10 <Jin^eLD> unfortunately not :)
Nov 07 20:11:23 <Jin^eLD> so please do tell me if you figure this out :)
Nov 07 20:16:52 <Jin^eLD> the only way I see right now is the way it was done in the old times, with a -native recipe that does inherit native
Nov 07 20:18:12 <msm> Jin^eLD: then an extra BBCLASSEXTEND for just nativesdk
Nov 07 20:18:40 <msm> seems like RPROVIDES_virtclass-target = "" would disable it
Nov 07 20:18:48 <msm> not sure of the extra semantics
Nov 07 20:18:54 <msm> wait on another build to finish
Nov 07 20:19:01 <msm> before I try anything
Nov 07 20:19:25 <kergoth> msm: just make a recipe that inherits native directly like they used to, if you want native and not target, as Jin^eLD suggested
Nov 07 20:19:38 <kergoth> also there's no target virtclass :)
Nov 07 20:19:48 <msm> then BBCLASSEXTEND = nativesdk?
Nov 07 20:19:57 <msm> well RPROVIDES=""
Nov 07 20:20:07 <msm> RPROVIDES_virtclass-native = ${PN}
Nov 07 20:20:35 <msm> seems feasible...
Nov 07 20:20:40 <msm> not sure what works best in the end
Nov 07 20:20:42 <kergoth> should work in theory, but better sanity check things. nativesdk recipes are normally generated by target + nativesdk class, whereas this one would be native + nativesdk class
Nov 07 20:21:54 <msm> kergoth: this is all temporary for me too - since the are lots of issues with the so called "source"
Nov 07 20:22:01 <msm> anyways, i'll figure something out
Nov 07 20:22:12 <kergoth> heh, good luck withi t
Nov 07 20:27:54 * kergoth wonders how well a lua based recipe format would work
Nov 07 20:28:26 <Jin^eLD> please, let us catch up with the oe core changes first ;)
Nov 07 21:00:05 <blindvt> %tell woglinde i have fleshed out your uclibc sched_* patch and pushed it as 19dd090a0f68765db87990ef8eda9bf77bb29581 . Thanks!
Nov 07 22:37:32 <msm> ok, i asked this a while ago i i've still not figured this out
Nov 07 22:37:40 <msm> how do I add an secondary package to my sdk?
Nov 07 22:38:32 <msm> i can easily add the primary package TOOLCHAIN_HOST_TASK = recipe
Nov 07 22:38:41 <msm> but I can't do TOOLCHAIN_HOST_TASK = recipe recipe-foo
Nov 07 22:47:59 <msm> ok
Nov 07 22:48:07 <msm> i was just missing a key point
Nov 07 22:48:10 <msm> i needed to do
Nov 07 22:48:40 <msm> pkgname-nativesdk-foo and not pkgname-foo-nativesdk to TOOLCHAIN_HOST_TASK
**** ENDING LOGGING AT Tue Nov 08 02:59:57 2011