**** BEGIN LOGGING AT Thu Jul 21 02:59:56 2011 Jul 21 03:01:06 yeah Jul 21 03:01:09 thats what I thought Jul 21 03:01:16 clean build Jul 21 03:01:22 current dev Jul 21 03:01:31 with local overlat, but not much there Jul 21 03:01:38 and it used to work until I updated Jul 21 03:03:23 ok goner to bed Jul 21 03:03:27 back at this tomorrow Jul 21 03:05:28 the git server has been moved....it is on different hardware. Jul 21 07:10:50 good morning Jul 21 07:50:46 Can someone help out on a build error. I am building Angstrom branch 2011.03-maintenance console-image. Portmap is failing on oe_install. It is a permission error (install: cannot change ownership of ...) http://www.pastie.org/2247266 Jul 21 07:52:09 opps got disconnected... question above ^ Jul 21 07:54:58 It is the line "oe_runmake install DESTDIR=${D}" in the recipie that is failing. Jul 21 08:17:51 khem ping Jul 21 08:22:55 uh oh. Jul 21 08:23:30 ka6sox? Jul 21 08:23:33 sleep Jul 21 08:23:34 now Jul 21 08:24:19 just as soon as I repartition this device on the fly...if it works this time I can sleep. Jul 21 08:24:46 hm what hw problems melo has again? Jul 21 08:24:49 by the way? Jul 21 08:25:07 its not happy...keeps crashing, and its software. not hardware. Jul 21 08:39:14 hi bluelightning Jul 21 08:39:25 ka6sox software? Jul 21 08:39:31 what the hell Jul 21 08:39:37 the VM is old and crufty Jul 21 08:39:43 so i am replacing it. Jul 21 08:39:52 with a shiny fresh one Jul 21 08:51:09 yeah Jul 21 09:12:58 morning woglinde_, ka6sox, all Jul 21 09:13:11 howdy bluelightning Jul 21 09:19:41 morning all Jul 21 09:21:40 hi pb_ Jul 21 09:26:36 hi pb Jul 21 12:17:29 is git working? Jul 21 12:19:20 with new dns Jul 21 12:26:28 yes Jul 21 12:26:33 I changed it last night :) Jul 21 12:27:17 http://pastebin.com/EYgdwCTT Jul 21 12:27:57 lotta weirdness Jul 21 12:28:06 had to build autoconf-native manually Jul 21 12:40:25 hi all Jul 21 12:41:05 I just bought a toradex colibri tegra2 + iris baseboard. is anyone else using oe-classic (or oe-core) with this setup? Jul 21 12:54:20 ok, bitbake autoconf-native; bitbake bison-native works Jul 21 12:59:39 but fails later, on another native package, betting I also need libtool-native bui;t Jul 21 13:00:45 o.O Jul 21 13:00:54 yes Jul 21 13:02:27 seems like the native depends are not getting built early enough Jul 21 13:02:53 Crofton|work, I spend one day solving those issues :) Jul 21 13:03:29 did anything get pushed? Jul 21 13:10:36 ? Jul 21 13:11:30 Crofton, I did this way: send a long bitbake -k something ; waited until end: saw packages... bitbake missing native... restart Jul 21 13:13:46 I['m looking for the proper fix Jul 21 13:13:51 this is ridiculous Jul 21 13:14:06 Crofton, you're saviour than me :) Jul 21 13:17:43 nothing obvious in the metadata log Jul 21 13:17:48 I wonder if it is bitbake Jul 21 13:18:32 is anyone working on tegra2 ? Jul 21 13:18:36 any clues what version of bitbake is know to work with oe-classic? Jul 21 13:19:31 master works for me Jul 21 13:19:38 bother Jul 21 13:20:02 now you made me choke on my coffee Jul 21 13:20:14 why do I have -native issues? Jul 21 13:20:46 i had them too :) Jul 21 13:20:59 they are new Jul 21 13:21:04 this has worked for years Jul 21 13:21:04 (two weeks ago when doing a setup on a new workstation Jul 21 13:21:14 I had 2 weeks ago Jul 21 13:21:19 I hadn't pulled in like a month or so Jul 21 13:21:45 and I workaround'd them ny doing what I said Jul 21 13:22:07 Maybe Yocto's work in OE :D (joking!!!) Jul 21 13:22:12 I know, but this must be fixed Jul 21 13:22:32 I'm trying to think how that could happen, but the only route is via bitbake Jul 21 13:24:48 JaMa, do you use a lot of assume provided for native stuff? Jul 21 13:25:21 Crofton|work: http://paste.pocoo.org/show/443516/ not really surprising that it doesn't work correctly.. but this is broken for months.. I can find the post where I reported it first.. but it's really long.. Jul 21 13:26:26 bitbake 1.12.0 may be better Jul 21 13:27:21 anyone use toradex colibri tegra or tegra ? any info on bsp for oe-core ? Jul 21 13:27:58 alx I would use it Jul 21 13:28:00 I haven't heard of anyone with a tegra Jul 21 13:28:02 its based .32 kernel Jul 21 13:28:09 which is really old for tegra now Jul 21 13:28:46 I wonder anyway why they came back to oe Jul 21 13:28:53 woglinde, I just bought a toradex colibri tegra2 + iris baseboard. Jul 21 13:28:57 abodonned it some years agao Jul 21 13:29:04 may we talk in que a bit ? Jul 21 13:38:40 RP_, bitbake master is broken with .dev, looks like 1.12.0 is ok though Jul 21 13:55:30 Hi to all Jul 21 13:57:11 I have a problem to build angstrom-x-image.bb, I can ask here? Jul 21 13:58:10 ~ask Jul 21 13:58:11 sure Jul 21 13:58:11 Questions in the channel should be specific, informative, complete, concise, and on-topic. Don't ask if you can ask a question first. Don't ask if a person is there; just ask what you intended to ask them. Better questions more frequently yield better answers. We are all here voluntarily or against our will. Jul 21 13:58:30 alx_, it looks like bitbake 1.12.0 works ok Jul 21 13:59:11 Crofton|work, cool :) good to know if I've to restart a workstatio Jul 21 13:59:13 n Jul 21 14:01:01 task do_rootfs fails with : opkg_install_cmd: Cannot install package task-xfce46-extras. | * resolve_conffiles: Existing conffile /home/marco/oe/tmp-angstrom_2008_1/rootfs/angstrom-x-image/etc/device_table is different from the conffile in the new package.... How I can resolve this? Jul 21 14:05:35 is there some docs on "What is oe-core ? How to use it ? oe-core for dummies ?" Jul 21 14:06:30 alx_: am writing something atm Jul 21 14:06:42 in the mean time there is this: https://wiki.yoctoproject.org/wiki/OpenEmbedded-Core Jul 21 14:08:33 bluelightning, let's understand :) Jul 21 14:45:09 alx_ for oe-core is git clone git://git.angstrom-distribution.org/setup-scripts too Jul 21 14:45:14 for a quick setup Jul 21 14:54:18 woglinde, I'm using it already Jul 21 14:54:31 trying to build a qemuarm console-image for test Jul 21 14:54:33 :D Jul 21 14:54:49 hi JaMa Jul 21 14:55:04 hi Jul 21 14:55:11 | checking for linux/if_alg.h... no for connman Jul 21 14:55:13 what's that? Jul 21 14:55:14 jo jama Jul 21 14:55:28 | configure: error: User-space algorithm header files are required Jul 21 14:55:31 GNUtoo|laptop: check last 2 patches I've sent to oe-devel ML today Jul 21 14:55:36 ok Jul 21 14:55:38 jama is it possible for the navit package to switch the navit.xml to /etc/navit? Jul 21 14:55:38 then: Jul 21 14:55:50 should we update dfu-util-native in oe Jul 21 14:55:50 GNUtoo|laptop: cannot push it to git as my dns still sees wrong server Jul 21 14:55:56 ok Jul 21 14:56:11 thanks a lot for fixing the libc issue yesterday Jul 21 14:56:21 and for closing the bugreport Jul 21 14:56:30 GNUtoo|laptop: yes.. I have work around for libusb with the _svn version we have.. but to test 0.4 and fix it properly would be better Jul 21 14:56:33 I was blocked by that during the whole period where you were offline Jul 21 14:56:47 there is a 0.4? Jul 21 14:56:57 I tought 0.3 was the last version Jul 21 14:57:17 GNUtoo|laptop: "dfu-util-native added -lusb-1.0 -lpthread -lusbpath -lrt to libusb.pc in x86_64 sysroot" Jul 21 14:57:18 I have 0.3 locally in oe.dev and it succedded only for the non-native version Jul 21 14:57:28 GNUtoo|laptop: stefan announced 0.4 yesterday iirc Jul 21 14:57:33 ah ok nice Jul 21 14:57:41 and biggest change is using newer libusb Jul 21 14:58:02 wow nice Jul 21 14:59:31 GNUtoo|laptop: http://dfu-util.gnumonks.org/releases/ Jul 21 15:00:04 ok thanks Jul 21 15:01:35 ah same here for the dns : git.openembedded.net[0: 140.211.169.165]: errno=Connection refused Jul 21 15:01:48 I guess I've to wait Jul 21 15:02:01 or to point to github Jul 21 15:02:07 as it was said in the ml Jul 21 15:02:46 oh, really? NOTE: multiple providers are available for runtime libcanberra-alsa (eglibc, eglibc-nativesdk, glibc-nativesdk, glibc, libcanberra) Jul 21 15:04:29 I've that too Jul 21 15:17:13 openembedded.net points at the wrong place Jul 21 15:17:17 use .org Jul 21 15:18:04 need to get mickeyl to drop the .net dns for .org services Jul 21 15:18:15 I think we have cleaned out all the docs that refer to them Jul 21 15:22:26 | path2devnum.c:(.text+0x265): undefined reference to `usb_get_busses' Jul 21 15:22:33 JaMa, ^^^ => could you do it? Jul 21 15:22:51 what? dfu-util-native? Jul 21 15:25:56 yes Jul 21 15:26:10 anyway I've to go Jul 21 15:26:45 GNUtoo|laptop: use that work around or upgrade to 0.4 as I said before Jul 21 15:27:39 JaMa, but that's what I was doing Jul 21 15:27:44 that is with the 0.4 upgrade Jul 21 15:27:48 ah wait a sec Jul 21 15:27:55 no it cannot be Jul 21 15:27:58 I've no checksums Jul 21 15:28:13 maybe the version are locked Jul 21 15:28:16 but I must go Jul 21 15:28:49 looks like the same issue as with _svn version (incomplete -l flags for libusb) Jul 21 16:41:54 oe-core git-native seems to be hosed Jul 21 16:42:04 and perl is whats troubling it Jul 21 16:42:22 Can't locate ExtUtils/Command.pm in @INC (@INC contains: ....... Jul 21 16:42:28 anyone seeing it Jul 21 16:46:30 I think RP__ has Jul 21 16:46:38 but that's a sstate relocation thing Jul 21 16:50:42 got bit by that yesterday, had to cleansstate it, indeed Jul 21 16:57:45 Tartarus: hmm yes I am building from sstate Jul 21 16:58:05 for record I have never been able to build from sstate peacefully Jul 21 16:58:11 when I nuke tmpdir Jul 21 16:58:17 kind of makes it useless to me Jul 21 16:58:46 should I cleansstate git-native or perl-native Jul 21 16:59:15 khem: ultimately the solution if you're experiencing problems with sstate is to try to fix them Jul 21 16:59:38 thats obvious Jul 21 17:02:18 the trouble is everyone's approach right now (mine included) is to just delete sstate-cache or do -c cleansstate when problems appear Jul 21 17:02:45 yes since there are more burning fires elsewhere Jul 21 17:08:35 there are a lot of these sstate reuse issues lurking, I'm sure Jul 21 17:42:25 If people run into issues, I'd suggest at least documenting the failures in the bugzilla if possible Jul 21 17:43:04 * RP__ knows of a couple of sstate issues in there which are in various states of being looked t Jul 21 17:43:05 at Jul 21 17:52:09 Bah, x86_64 doesn't actually work as an OVERRIDE, does it.. Jul 21 17:54:59 heh Jul 21 18:01:57 # BUILD_CC_ARCH=${@base_contains('OVERRIDES', 'x86_64', '-m64', '', d)} Jul 21 18:01:57 does what I need, heh Jul 21 18:06:49 heh Jul 21 18:13:08 Tartarus, I solved last nights problem by falling back to bitbake 1.12.0 Jul 21 18:13:20 Ah Jul 21 18:16:23 the -native depends seem borked with buitbake/master and dev Jul 21 18:16:58 We should use amd64 Jul 21 18:17:03 for denoting x86_64 Jul 21 18:17:16 autotools understand amd64 equally wel Jul 21 18:17:17 l Jul 21 20:15:03 Tartarus: there ? Jul 21 20:15:08 pong Jul 21 20:15:12 pooop, sorry Jul 21 20:33:58 pb_: did the automatic update of the github mirrors get broken by the change in host key for git.oe.org? Jul 21 20:34:10 or, no, i think i just need to wait for the sync, nevermind Jul 21 20:34:18 juggling too many windows, getting mixed up Jul 21 20:35:02 hmm Jul 21 20:54:53 khem ping Jul 21 20:56:17 how do I set uclibc-git version in oe-core? Jul 21 21:12:23 woglinde_: pong Jul 21 21:13:14 khem how do I choose uclibc-git with oe-core? Jul 21 21:13:37 woglinde_: are you using angstrom ? Jul 21 21:14:46 yes Jul 21 21:14:55 then use DISTRO = "angstrom-bleeding" in conf/local.conf Jul 21 21:14:59 editing site.conf didnt hellp Jul 21 21:15:01 help Jul 21 21:15:14 setting preffered version Jul 21 21:15:15 hm Jul 21 21:15:27 I am only wanted uclibc-bleeding Jul 21 21:15:58 PREFERRED_VERSION_uclibc ?= "0.9.32+0.9.33%" Jul 21 21:16:04 PREFERRED_VERSION_uclibc-initial ?= "0.9.32+0.9.33%" Jul 21 21:16:08 that will help Jul 21 21:16:11 uhm Jul 21 21:16:12 okay Jul 21 21:16:20 thanks Jul 21 21:18:26 woglinde_: btw dont use ?= but = Jul 21 21:33:50 hm gcc dont poackages gcov Jul 21 21:33:52 /usr/lib/arm-angstrom-linux-uclibceabi/4.5.4/libgcov.a Jul 21 21:34:05 QA Issue: No GNU_HASH in the elf binary: '/devel/arm/git/setup-scripts/build/tmp-angstrom_2010_x-uclibc/work/armv7a-vfp-angstrom-linux-uclibceabi/libgcc-4.5-r39.1+svnr175127/packages-split/libgcc/lib/libgcc_s.so.1' Jul 21 21:34:28 khem fix it! Jul 21 21:34:29 *g* Jul 21 21:35:32 yeah I know that issue Jul 21 21:36:01 only way to fix it _nicely_ is to configure toolchain to pass hash-style implicitly Jul 21 21:36:21 but thats not preferred in oe Jul 21 21:36:25 hm cannt believe it that the intel guys didnt fixed it Jul 21 21:36:30 folks want to use the option Jul 21 21:36:52 dunno Jul 21 22:43:44 could someone advise me on how to check out from openembedded master GIT repository? Jul 21 22:43:59 mrmichael: you can use the official or a mirror Jul 21 22:44:05 git clone git://github.com/openembedded/openembedded/tree/master/recipes/php doesn't work Jul 21 22:44:31 specifically, I want the latest PHP recipe Jul 21 22:44:52 mrmichael: git clone git://github.com/openembedded/openembedded.git Jul 21 22:45:03 ok Jul 21 22:45:07 will that give me PHP only? Jul 21 22:45:10 if gst-plugin-pulse_0.9.7.bb depends on pulseaudio and pulseaudio PR is bumped, would gst-plugin-pulse be automatically rebuilt? Jul 21 22:45:42 running it now ;) Jul 21 22:45:49 how much will this pull down? Jul 21 22:45:56 mrmichael: everything Jul 21 22:46:14 is that like...gigs? Jul 21 22:46:43 oh hey...this aint bad Jul 21 22:46:53 10 % at 10 meg Jul 21 22:46:55 :) Jul 21 22:47:57 i'm more of an SVN guy...this GIT stuff is a little strange to me Jul 21 22:48:02 I get a kernel panic because I it says unable to open console, I'm booting kernel from tftp and rootfs off NFS.. http://pastebin.com/R77njZUV Jul 21 22:49:09 this won't fill up my 200 meg hard disk, nah ;) Jul 21 22:49:36 thanks octavio. Jul 21 22:50:50 mrmichael: heh, no problem Jul 21 22:51:05 mrmichael: if you're starting a new development consider using oe-core, not oe Jul 21 22:53:11 ok. Jul 21 22:54:50 what would cause an unable to open an initial console. Jul 21 23:01:23 jkridner not in classic Jul 21 23:01:30 with oe. Jul 21 23:01:38 ups oe-core perhaps Jul 21 23:07:14 what kind of file permissions should I have set on my NFS fs on the server? Jul 22 00:41:33 I have an embedded linux system trying to boot off of NFS, the system loads the kernel, the network seems to come up and mount the NFS but I get a console unable to open http://pastebin.com/NiFYGPZR I have looked into the inittab to make sure I have the correct settings for the console, but I'm not sure what to change if it even needs changed Jul 22 00:46:07 off topic. this isn't a general embedded linux channel, it's specifically for the openembedded project (http://openembedded.org/) Jul 22 00:47:23 oh sorry Jul 22 00:53:49 tlab: Did you extract your nfsroot tarball as root? Jul 22 01:16:31 Tartarus, yes Jul 22 01:24:35 I wonder if something is messed up with my nfs, but I can load the kernel from nfs... Jul 22 01:32:07 kernel came over tftp, most likely Jul 22 01:32:29 ok Jul 22 01:32:35 so, you don't have dhcp configured Jul 22 01:32:44 kernel worked over tftp and nfs... I figured it out tho, was a setting in /etc/exports Jul 22 01:32:45 so you need to pass in the right kernel cmdline args to get the nfs server, etc Jul 22 01:32:55 ip= is missing Jul 22 01:34:03 had fsid=0 Jul 22 01:34:22 I still do get a weird error, but it continues to boot Jul 22 01:35:36 hello every body: am computer telecommunication student. and i wanna to start a embedded system formation. so please can you tell me what language shall i study and what i need else to be ready? Jul 22 01:36:07 right Jul 22 01:36:09 i mean on the boot args Jul 22 01:36:14 you need ip=.. stuff Jul 22 01:36:16 see your pastebin Jul 22 01:36:25 154-158 Jul 22 01:36:27 no IP info Jul 22 01:36:31 since you don't have dhcp providing it Jul 22 01:36:35 and you didn't pass it via ip= Jul 22 01:38:56 in bootargs I'm just giving it an ip at the moment, I know I need to add more... think that will fix that error? Jul 22 01:45:31 google up the syntax of the argument Jul 22 01:45:38 you need to pass most things Jul 22 01:45:45 afk, dinner, sorry Jul 22 01:45:57 http://lxr.linux.no/linux+v2.6.32/Documentation/filesystems/nfsroot.txt **** ENDING LOGGING AT Fri Jul 22 02:59:57 2011