**** BEGIN LOGGING AT Thu Jul 10 02:59:57 2008 Jul 10 03:28:57 I have some builds failing in do_patch() but my temp/run.XXXX file doesn't have a do_patch() function and there is nothing is in the temp/log.XXXX file. Jul 10 03:29:24 er, run.staging_helper.XXXX Jul 10 06:46:38 morning Jul 10 08:36:17 03koen 07org.oe.dev * r778242e4... 10/ (1 packages/ffmpeg/ffmpeg_git.bb): ffmpeg git: adjust staging Jul 10 08:36:22 03koen 07org.oe.dev * r731f64c9... 10/ (1 packages/ffmpeg/omapfbplay_git.bb): omapfbplay: bump SRCREV to move frame display to a seperate thread Jul 10 08:41:03 03tmbinc 07org.oe.dreambox * r42ce97d1... 10/ (7 files in 3 dirs): enigma2-defaultservices: add reasonable defaults for Astra 19.2E and Hotbird Jul 10 08:41:07 03tmbinc 07org.oe.dreambox * r16f378c2... 10/ (1 packages/libdreamdvd/libdreamdvd.bb): libdreamdvd: update Jul 10 08:41:11 03tmbinc 07org.oe.dreambox * r2c8cd8f8... 10/ (1 packages/python/python-pyopenssl_0.6.bb): python-pyopenssl: backport from .dev Jul 10 08:41:14 03tmbinc 07org.oe.dreambox * r0198e80a... 10/ (1 packages/liboil/liboil_0.3.14.bb): liboil: update Jul 10 08:41:18 03tmbinc 07org.oe.dreambox * ra70e7d0f... 10/ (5 files in 2 dirs): gstreamer: update Jul 10 08:41:22 03tmbinc 07org.oe.dreambox * ra1ddba68... 10/ (4 files in 3 dirs): libmatroska: add Jul 10 08:41:25 03tmbinc 07org.oe.dreambox * r685f006a... 10/ (1 packages/libebml packages/libebml/libebml_0.7.8.bb): libeml: add package Jul 10 09:50:52 i can use the oe git repo instead of the monotone tree yeah? Jul 10 11:06:38 03koen 07org.oe.dev * r60c9d8f7... 10/ (4 files in 3 dirs): omapfbplay: make a static binary Jul 10 12:04:27 pigeon: yes, the OE git tree seems to be syncronized very well, so that is what I use most of the time. Jul 10 12:04:45 excellent, thanks. Jul 10 12:14:01 03koen 07org.oe.dev * rc1a81a9f... 10/ (1 packages/images/beagleboard-demo-image.bb): beagleboard demo image: add mplayer and omapfbplay Jul 10 12:15:25 cd Jul 10 12:15:31 doh Jul 10 12:24:11 has anyone ever used rxtx with cacoa/jamvm on an ARM system? Jul 10 12:57:37 cbrake: I am sure we used that in earlier variants of the freedroidz software (http://wiki.evolvis.org/freedroidz/index.php/Main_Page) Jul 10 12:58:18 cbrake: it communicated through a bluetooth serial connections which was established first Jul 10 12:58:35 What is the difference between site.conf and local.conf? It is referenced seemingly interchangably in this: http://www.kernel-labs.org/files/openembedded-guide/openembedded-guide-6.html Jul 10 13:01:05 piroko: I assume that if you have multiple people working on OE site.conf is a way to share commons settings while every user may customize even more through local.conf Jul 10 13:01:27 Ah. I'm just trying to make an overlay Jul 10 13:01:48 And I think bitbake wants to eat my children Jul 10 13:01:52 piroko: then put the site.conf into that overlay Jul 10 13:03:13 Well I think I have it set up correctly, and then bitbake goes and uses the non-overlay bb file to do its thing Jul 10 13:03:38 piroko: AFAIR you need to put your overlay before the regular OE files Jul 10 13:03:47 BBFILES = "${HOME}/devel/jalimo/jalimo-trunk/oe-overlay/packages/*/*.bb \ Jul 10 13:03:47 ${HOME}/oe/org.openembedded.dev/packages/*/*.bb \ Jul 10 13:03:47 " Jul 10 13:04:00 And what file does that go in? Jul 10 13:04:03 this is how I set my overlay Jul 10 13:05:26 build/conf/local.conf? Jul 10 13:09:35 Well putting that in my local.conf worked. Thank you Jul 10 13:09:53 I just still feel like I'm doing it wrong though. Bitbake looks at so many env variables and files... Jul 10 13:19:03 piroko: all of the conf files create one global variable space -- its pretty simple Jul 10 13:19:48 cbrake: Yeah, but the environment variables I set were getting clobbered even when I used += Jul 10 13:20:04 piroko: check in: org.openembedded.dev/conf/bitbake.conf Jul 10 13:20:14 piroko: there you can see the order conf files are included Jul 10 13:20:44 piroko: so if later included file uses '= Jul 10 13:21:03 piroko: so if later included file uses '=' to set a variable, it will clobber anything set earlier Jul 10 13:21:15 piroko: that is why you see ?= a lot Jul 10 13:23:16 Ahh Jul 10 13:23:18 Thanks :) Jul 10 13:29:30 piroko: have you solved the double free problems? Jul 10 13:29:50 diego_at_work: Nope Jul 10 13:30:08 Works great in ubuntu! :D :P Jul 10 13:32:06 * * OE Bug 4424 has been created by diegor(AT)tiscali.it Jul 10 13:32:08 * * Build error task-java Jul 10 13:32:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4424 Jul 10 13:53:37 piroko: nice ^^ Jul 10 13:55:13 Yeah, but I don't use Ubuntu. Did it from live-cd chroot Jul 10 14:00:42 piroko: defining the overlay in site.conf is fine if all users then define BBFILES += "" Jul 10 14:02:05 since you need to mention all directories in BBPATH additionally bitbake can access site.conf in the overlay Jul 10 14:03:19 not sure how others do it but I have multiple OE configuration dirs in a shared directory for all things OpenEmbedded Jul 10 14:03:31 e.g. for the beagleboard, openmoko and the nokia internet tablets Jul 10 14:03:48 I use a small shell function to set all the needed variables: http://pastebin.com/m35d178fc Jul 10 14:04:39 Awesome. Thanks :) Jul 10 14:04:42 I added that to my .bashrc Jul 10 14:05:11 so if I enter 'oe-env beagle' I can start building stuff for the beagleboard Jul 10 14:06:31 the script always adds the jalimo overlay but I am sure you can abstract that away and replace it with your needs :) Jul 10 14:06:43 Yep yep! Jul 10 14:07:18 piroko: I have a checkout of bitbake inside OE_HOME. this is also put into the path Jul 10 14:07:37 the bitbake-1.8 branch to be specific Jul 10 14:07:54 (using bitbake trunk/ is not recommended) Jul 10 14:09:33 piroko: I am sorry one needs to be a guru to use OpenEmbedded properly but you shouldn't give up. If you grok it once it will save your time afterwards :) Jul 10 14:11:23 rschuster: Hehe. It's alright. It has saved me time already. Just needs time for me to grok it all Jul 10 14:11:35 SIASL ftw Jul 10 14:25:47 anyone been able to run Kdrive on omapzoom or similar? Jul 10 14:56:12 03thebohemian 07org.oe.dev * ra846a4c2... 10/ (4 files in 3 dirs): cacao-native 0.99.1: New recipe. Jul 10 14:56:17 03thebohemian 07org.oe.dev * r26d45c37... 10/ (4 files in 3 dirs): disapproval of revision 'a846a4c23e94c3e771fd7aaf448e8bb4e11f4506' Jul 10 14:56:22 03thebohemian 07org.oe.dev * rbad7ee44... 10/ (3 files in 2 dirs): Jul 10 14:56:22 cacao-native 0.99.1: New recipe. Jul 10 14:56:22 * add support for new configure options in cacao-native.inc Jul 10 14:56:27 03thebohemian 07org.oe.dev * rf25e95db... 10/ (1 conf/distro/chinook-compat.conf): chinook-compat: Set dbus-glib version preference to 0.7.4. Jul 10 15:25:04 RP: hi, I am seeing a problem when building target gcc recipe gcc_4.3.1.bb. It is adding -I Jul 10 15:25:18 why do so many machines use IMAGE_FSTYPES=ext2 - wouldn't they be better off using something journaled like ext3? Jul 10 15:25:39 and this path is getting included even for the utilities that gcc builds for build machine like genmodes Jul 10 15:25:50 and the build gets hosed up Jul 10 15:26:21 I was hunting where this include gets into the build Jul 10 15:27:29 any idea on this would help Jul 10 15:51:27 khem: did you check the temp/ directory run.* scripts in the tmp/work/ dir? Jul 10 15:58:21 likewise: I did Jul 10 15:58:31 likewise: I know the reason but I do not know the cause Jul 10 16:00:28 khem: ok. there are little ppl using 4.3.1 here, except koen for the beagle (texas' omap armv7) Jul 10 16:00:48 khem: and some of the other Beagle folks Jul 10 16:01:19 had to reboot... so does anyone know why you would want to use ext2 vs ext3 fs which adds journaling? Jul 10 16:01:20 khem: I can try to use gcc-4.3.1 for my targets and look along with your findings, but not this week. Jul 10 16:01:50 tharvey: many targets probably used a ext2 ramdisk. Yes, you'ld typically want ext3 on your disk-like devices. Jul 10 16:03:32 likewise: it works for cross Jul 10 16:03:47 03thebohemian 07org.oe.dev * r23077a14... 10/ (5 files in 3 dirs): classpath: Added patch fixing build on newer GCCs (and prevents a bug). Jul 10 16:03:48 likewise: I think there should be same problem on 4.2.x as well Jul 10 16:03:59 likewise: if you build the target compiler Jul 10 16:04:07 likewise, you mean many targets are probably outdated if they specify ext2? Jul 10 16:05:31 tharvey: well, unless there is a specific reason to use ext2 instead of 3, yes I suspect it due to outdation. Ask the machine maintainer. Jul 10 16:05:54 cool - thanks Jul 10 16:06:46 khem: ok, I'll try to build a target compiler for avr32 in a few minutes when my build completes. Jul 10 16:07:01 ext3 the journal doesnt move and you have to assume the flash storage levels the writes correctly Jul 10 16:07:38 likewise: avr32 might have its own problems but nevertheless its good if you can reproduce this Jul 10 16:08:17 I reckon not many people might build target compiler for embedded device Jul 10 16:09:50 khem: no, indeed. especially on #oe I have not seen (m) any people really do this. Jul 10 16:29:01 03jluebbe 07org.oe.dev * r478afd23... 10/ (4 files in 4 dirs): fso-image: use e-wm instead of matchbox Jul 10 16:31:43 How do you get a {PV} to expand properly in a bb file? Jul 10 16:32:21 I just made my own custom bb file based off of libusb's, and the {PV} part in SRC_URI is getting translated literally Jul 10 16:32:51 try ${PV} Jul 10 16:33:20 Doh Jul 10 16:40:14 Well crap. This package I am trying to build uses rpath Jul 10 16:40:42 So I get a nice QA issue with it :( Jul 10 16:42:16 Is there a flag I can put in the .bb file to disable rpath? Jul 10 16:43:34 try EXTRA_OECONF += "--disable-rpath" Jul 10 16:43:42 Yeah, just saw that. Trying it now. Jul 10 16:43:54 Thank you :) Jul 10 16:43:58 if the package support it through libtool Jul 10 16:45:15 I am setting up the local conf and I am confused about something. Eventually I want to make a i386 generic with no math coprocessor 2.6 kernel. Does the minimal.conf distro work? I know there are some that are more "loved" than others Jul 10 16:45:54 woglinde: Fail :( Jul 10 16:49:05 does oe compile a new kernel for the image? Jul 10 16:50:00 for the official manual...does everyone get Not Found? Jul 10 16:50:03 lol Jul 10 16:50:28 javatexan you can fetch via monotone and compile yourself Jul 10 16:51:02 whats that ... a new kernel? Jul 10 16:51:30 ??? Jul 10 16:51:34 I unfortunately installed OE and bitbake via synaptic so all the documentation is off a little Jul 10 16:51:36 I meant the manual Jul 10 16:51:50 oh...okay the manual.....right Jul 10 16:52:22 * khem leaves for work Jul 10 16:53:55 woglinde: Is there a different way to disable rpath so oe will be happy? Jul 10 16:54:32 khem: cya Jul 10 17:08:07 woglinde: how do I list the repos in monotone...ie how do I find the manual repo to download Jul 10 17:14:39 Jeez... There's a bunch of stuff in this configure script that uses rpath... Jul 10 18:21:06 * * OE Bug 4425 has been created by likewise(AT)gmx.net Jul 10 18:21:08 * * avr32 linux-2.6.25.10.atmel.2 Jul 10 18:21:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4425 Jul 10 18:43:55 03koen 07org.oe.dev * r5e7b17fa... 10/ (1 packages/ffmpeg/omapfbplay_git.bb): omapfbplay: fix depends Jul 10 18:44:00 03koen 07org.oe.dev * rd07d8aef... 10/ (1 conf/distro/include/angstrom-glibc.inc): angstrom-glibc: don't use -Os on armv7, csl2007q3 breaks horribly with it Jul 10 18:46:47 okay...how do i get the manual down to compile...the suspense is killing me.. ;) Jul 10 18:50:34 ??? Jul 10 18:50:44 is it already down Jul 10 18:51:12 or do i have to pull it down from another project/repo/etc.... Im confused Jul 10 18:52:21 !seen laibsch Jul 10 18:52:22 woglinde, laibsch is right here! Jul 10 18:52:40 laibsch ping Jul 10 18:55:23 http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=shortlog;h=refs/heads/org.openembedded.documentation Jul 10 18:55:32 seems you can use git now Jul 10 18:56:45 i dont mind using monotone, I just am not sure what to get Jul 10 18:57:51 or how....i am new to oe and monotone and bitbake...so I am drowning in manuals...the only doc I have right now is the wiki which doesnt seem to say anything about compiling the manual...sorry for the trouble Jul 10 18:58:09 yeah is the new one now Jul 10 18:58:15 but you can read getting started Jul 10 18:58:25 or fetch the git tree I posted Jul 10 18:58:34 and compile the manual yourself Jul 10 18:58:35 i did the getting started Jul 10 18:58:41 okay Jul 10 18:59:00 I hope laibsch will be awake some time now, so he can fix the wiki Jul 10 18:59:09 lol Jul 10 18:59:12 i hear ya Jul 10 18:59:25 the getting started is great... Jul 10 19:00:55 and as I said you can use git now Jul 10 19:01:01 instead of montone Jul 10 19:01:10 for readonly Jul 10 19:02:50 woglinde: what is wrong with the wiki? Jul 10 19:03:42 cbrake the manual link isnt right Jul 10 19:03:58 http://wiki.openembedded.net/user-manual?q=user-manual Jul 10 19:04:35 woglinde: hmm, is there a way to get to the old site ... Jul 10 19:05:08 cbrake not really Jul 10 19:05:08 woglinde: I can't find anything as oe.linuxtogo.org redirects to the new wiki as well Jul 10 19:05:25 woglinde: I guess we should probably just generate a pdf for now and link to it Jul 10 19:11:14 I am trying mtn --db=/home/javatexan/oestuff/manual pull monotone.openembedded.org org.openembedded.documentation but it says that /manual is a dir not a db. How do i make the new db here...sorry...am i close? Jul 10 19:15:00 the wiki works fine for me Jul 10 19:15:37 I don't know why the whole website has been repointed to the wiki Jul 10 19:18:56 laibsch but the manual link istn working Jul 10 19:19:03 can you fix it? Jul 10 19:20:17 re Jul 10 19:20:46 woglinde: ask mickey|openezx Jul 10 19:22:27 flo_lap: Do you know anything about oe.linuxtogo.org completely redirecting to amethyst? Jul 10 19:22:33 seems a few things got broken Jul 10 19:22:44 hi flo Jul 10 19:23:01 laibsch hm oh I thought you mangened it Jul 10 19:23:18 Laibsch: I have no idea and did not make any changes. Jul 10 19:24:49 flo_lap: but maybe you can take a peek Jul 10 19:25:05 woglinde: I co-manage the wiki, not the whole website Jul 10 19:31:55 * cbrake looks at making a pdf of the manual ... Jul 10 19:34:15 I don't see the need for a hurry Jul 10 19:34:23 we should eventually fix things up right Jul 10 19:34:48 Laibsch: I would really like a pdf of the manual generated on a regular basis anyway Jul 10 19:35:04 Laibsch: as that is what I prefer to use Jul 10 19:35:53 600 mtn revs to go -- guess I'll go find something else to do for a bit ... Jul 10 19:36:12 cbrake hm why didnt you use the git tree? Jul 10 19:36:21 should be much faster Jul 10 19:36:28 woglinde: I want to make sure the documentation branch is sync'd Jul 10 19:37:02 woglinde: plus, I have some things I need to push. I use git for all my dev work. Jul 10 19:37:21 woglinde: does anyone push to the git tree yet, or does it all have to go through mtn? Jul 10 19:38:03 cbrake doc repo isnt updated this much Jul 10 19:51:04 03thebohemian 07org.oe.dev * reed451c4... 10/ (1 packages/jamvm/jamvm-initial_1.5.0.bb): jamvm-initial: Fixed installation directory. Jul 10 19:51:09 03thebohemian 07org.oe.dev * rd16b0517... 10/ (1 packages/cacao/cacao-native.inc): cacao-native.inc: Fix setting include dir. Jul 10 19:51:14 03thebohemian 07org.oe.dev * rc3c2fc52... 10/ (1 packages/classpath/classpath-native_0.97.2.bb): classpath-native 0.97.2: New recipe. Jul 10 19:54:39 Laibsch: can you enable uploading *.pdf files to the wiki? Jul 10 19:55:02 Laibsch: otherwise I could just put it on my server I suppose Jul 10 19:55:51 if we link it officially, we should be sure it gets updated from time to time Jul 10 19:55:59 preferably in an automatic fashion Jul 10 20:03:44 Laibsch: http://wiki.openembedded.net/index.php/Documentation Jul 10 20:04:23 Laibsch: I'll set up a cron to update the PDF daily Jul 10 20:04:35 Laibsch: I may as well do HTML docs while I'm at it Jul 10 20:23:10 03koen 07org.oe.dev * r7131f281... 10/ (1 packages/ffmpeg/omapfbplay_git.bb): omapfbplay: update DEPENDS Jul 10 21:01:30 nite Jul 10 21:15:29 cbrake: you rock..thank you Jul 10 21:16:26 javatexan: you're welcome -- feel free to improve the wiki :-) Jul 10 21:17:02 well I was going to add how to mtn the manual and compile it, but I never figured out how.... Jul 10 21:17:38 it says i am not specifying a branch to clone Jul 10 21:18:05 i tried mtn clone monotone.openembedded.org org.openembedded.documentation Jul 10 21:18:14 in a manual directory Jul 10 21:19:21 javatexan: I ended up checking it out of the git repo Jul 10 21:20:22 thats cool....git, mtn are both new to me...been using svn and cvs for a long time. Just switched over to OE from slind yesterday...still getting feet wet. Jul 10 21:22:30 and of course, I installed OE, bitbake, etc via synaptic, so all the docs are off a little on location of things Jul 10 21:22:55 i am slowly treading water now....I'll get to the shallows soon Jul 10 21:22:58 i hope Jul 10 21:26:17 javatexan: its worth it :-) Jul 10 21:27:37 I am sure it is...I can see myself having a lot of fun with it, once i figure out how,why,when,where.... Unfortunately this first project is for a job...so its a little stressful to be patient. Jul 10 21:29:09 just to clarify....OE using bitbake and the build/conf/local.conf builds an ?image? that i can test in Qemu...or not quite yet. Jul 10 21:48:00 I'm having trouble trying to get a machine to mount nfsroot - http://www.pastebin.org/50376 shows the console - I 'can' mount via std nfs-client just fine so it wouldn't appear to be a server issue - any ideas? Jul 10 21:51:42 should I be able to simply nfs export build/tmp/rootfs and nfsroot to that? Jul 10 21:56:09 tharvey: don't you have to disable locking or something of that sort? And don't forget to disable the root squashing. Jul 10 21:57:07 ah... root squash - never understood too well what that does and I definitely don't have that set on the server Jul 10 21:57:25 as for locking... you mean '-o nolock' right? how would I instruct the kernel to use that I wonder? Jul 10 21:58:01 No, I thought it was something on the server side. Jul 10 21:58:22 (it's been a long time since I've done any of that!) Jul 10 21:59:45 ah... no_root_squash means do not dissallow root - I always mistook squash to mean something more literal Jul 10 22:00:26 * mwester envisions garden vegetables with ethernet ports... Jul 10 22:01:01 * tharvey envisions some sort of nifty file compression Jul 10 22:01:31 Yeah, it's a lousy name. "supress" might have been better, IMO. Jul 10 22:01:41 er, suppress I mean. Jul 10 22:02:50 hmmm... that didn't do it. From googling around it would seem that the only server-side opts I need are rw,no_root_squash Jul 10 22:03:16 experience tells me that for nfs-client if you don't specify '-o nolock' you need portmapper running... but I'm not sure how that relates to kernel mounting rootfs Jul 10 22:03:47 I just checked here and my notes indicate that the kernel defaults to no locking. Jul 10 22:03:58 Just so that you don't need a portmapper. Jul 10 22:04:11 is using build/tmp/rootfs appropriate? I notice that build/tmp/rootfs/dev/console is an empty regular file - NOT a device node!? Jul 10 22:04:14 I presume you've checked the DHCP or network ettings, right? Jul 10 22:04:42 Ah, no that's not a good rootfs. Jul 10 22:04:49 yes... IP settings are definately correct Jul 10 22:05:05 thats what I thought... it it supposed to be a good rootfs? Jul 10 22:05:16 No. Jul 10 22:06:06 You need to generate a rootfs.tar.gz when you build your image; that can be "untarred" by root and should have the correct permissions and file types. Jul 10 22:06:46 The problem is that /tmp/.../rootfs is built by an unpriveleged user, and thus cannot have files owned by root, and cannot create device nodes. Jul 10 22:07:25 It uses "fakeroot" to do that, and the tar file image is created from fakeroot, and thus has the correct device nodes and permissions. Jul 10 22:08:24 ah... then hopefully thats my issue (although if it was I would think I would get an error different than a timeout error... hmmm) Jul 10 22:09:02 mwester, that makes sense about build/tmp/rootfs then - I never quite understood why it wasn't put in build/tmp//rootfs - guess its b/c it really is a scratch directory Jul 10 22:22:55 ok, so I've now got a 'valid' root fs hopefully (extracted from base-image-.tar.bz2 as root to export dir and my export is with options rw,no_root_squash - now I'm getting different issues when trying to boot via nfs: http://www.pastebin.org/50387 Jul 10 22:33:40 tharvey: looks like you're missing some services on the server, no? Jul 10 22:34:42 Is rpcbind running? Do you have a firewall running on that server? Jul 10 22:35:10 well... I can mount via std nfs-client if I boot the same kernel on the same machine without nfsroot Jul 10 22:36:11 no firewall on server... no rpcbind running Jul 10 22:36:43 mountd and nfsd are running... do I need rpcbind? I apparently don't need it for std nfs-client Jul 10 22:44:19 hmmm... I just booted poky 3.1 via nfs just fine from its image on the same server, so its not my server config Jul 10 22:49:07 must be missing something in my kernel config thats in poky-3.1 kernel Jul 10 22:58:08 so if your board is a 386 with no coproc, should I start with x86 as the machine or is something else closer Jul 10 23:00:53 javatexan, x86.conf defines its ARCH as i486 by the way... Jul 10 23:02:23 well the final board is a SOC RISC pretending to be 486....it has no coproc though. I can run the x86 created in qemu hopefully until I get ready to make the new noncoproc kernel Jul 10 23:02:39 so I guess x86 should work....I hope Jul 10 23:13:12 okay it is not using the config file that I changed....it is complaining that Distro "none" not found Jul 10 23:14:16 are you creating your own distro? Jul 10 23:15:03 on my machine /home/javatexan/oestuff/build/conf /usr/src/oe/org.openembedded.dev and /usr/bin/bitbake are the major players Jul 10 23:15:27 no...just trying the x86 with linux with distro minimal Jul 10 23:15:45 in the config file in /home/javatexan/oestuff/build/conf Jul 10 23:15:54 so the only file you touched is your local build/conf? Jul 10 23:16:05 yes...and set the BBPATH Jul 10 23:16:07 does your BBPATH include the oestuff dir? Jul 10 23:16:17 and did you export it? Jul 10 23:16:47 doh...mispelling.... Jul 10 23:17:18 i was just trying the example in Getting Started Jul 10 23:17:27 bitbake nano Jul 10 23:18:01 it is doing something now though...thanks Jul 10 23:18:19 its been my experience that the OE getting started guide is lacking and possibly out of date - if you want to learn OE definitely start by reading the poky manual in its entirety Jul 10 23:18:26 computers are so picky with spelling.... :) Jul 10 23:19:02 ok...thanks tharvey...will do...just wanted to make sure everything was 'alive' after install Jul 10 23:21:12 mwester, you must have been talking about CONFIG_LOCKD with regards to locking and NFS - it is set as selecting NFS file system selects it Jul 10 23:21:49 I'm comparing my kernel config with that of pinky-3.1.1's and I can't see any diff in the way I've config'd NFS options - yet pinky boots with nfsroot and my OE built kernel/machine doesn't :( Jul 10 23:23:25 tharvey: did you double-check built-in versus modules? Jul 10 23:25:31 yes... all built-in Jul 10 23:29:40 okay....so now that ive "bitbake nano" all i have in ~/oestuff/build is conf/ and tmp/ What happened? it said it ran 255 tasks of which 0 failed Jul 10 23:29:45 ah... it 'is' now mounting rootfs - just that network was getting taken down due to config after mounting root Jul 10 23:29:56 now to figure out exactly what I changed to make it start working... Jul 10 23:30:17 javatexan, build/tmp/deploy ... Jul 10 23:31:00 personally I think putting deploy in $TMPDIR is strange... should put it in build I would think Jul 10 23:31:42 javatexan, so if your using glibc somewhere down from build/tmp/deploy/glibc/ipk you should see a nano*.ipk Jul 10 23:41:48 all i have under tmp/ is abi_version cache saved_tmpdir Jul 10 23:42:04 i know why..nvm..let me try something Jul 10 23:47:23 heh... damn typo bit me in the arse too... Jul 10 23:54:09 mwester, got nfsroot working - thanks for your help. In the end I was booting a kernel that somehow was not setting the MAC addr of my eth adapter properly and I failed to notice - in the end it would seem that no_squash_root is 'not' needed on the server Jul 10 23:56:33 so does OE have a nifty way to prepare an nfsroot directory, or should I create my own nfsroot.bb which requires base-image.bb (or other image) then extracts the necessary rootfs? Jul 11 00:10:02 tharvey, I just unpack one of the tar.gx imager files into the nfsroot directory Jul 11 00:11:20 ya... I was just wondering if there was some IMAGE_FSTYPES type or something else built in that I should be using which did that for you automatically Jul 11 00:16:07 its it wrong to create a machine or package name that has an '_' in it? seems like that might be confusing bitbake into thinking everything past the _ is a machine/arch Jul 11 00:22:41 s/its is/is it/ Jul 11 00:54:29 hmm, i'm trying to build an angstrom image, i've getting a fail for do_configure with linux-handhelds-2.6-2.6.21-hh20-r20, the log says oe/tmp/deploy/uclibc/images/hx4700/ does not exist, but i've set ANGSTROM_MODE to glibc, any suggestions? Jul 11 01:04:05 * * OE Bug 4426 has been created by ka6sox(AT)nslu2-linux.org Jul 11 01:04:07 * * Security Vulnerabilities in Bind {8,9} Jul 11 01:04:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4426 Jul 11 02:13:45 ah, google might have solved my problem... **** ENDING LOGGING AT Fri Jul 11 02:59:56 2008