**** BEGIN LOGGING AT Wed Jan 11 02:59:57 2012 **** BEGIN LOGGING AT Wed Jan 11 05:01:37 2012 Jan 11 08:23:50 good morning Jan 11 10:36:19 03Lianhao Lu  07master * r9d8f45407c 10bitbake.git/ (4 files in 3 dirs): (log message trimmed) Jan 11 10:36:19 bitbake: Automatically start local PR service. Jan 11 10:36:19 [YOCTO #1126] Jan 11 10:36:19 A local PR service will be started and stopped automatically along Jan 11 10:36:19 with the bitbake invocation/ternimation. Jan 11 10:36:19 This local PR service will be started only and if only when the Jan 11 10:36:19 PRSERV_HOST is set to 'localhost' and PRSERV_PORT is set to '0'. Jan 11 10:36:29 03Lianhao Lu  07master * r379567ee87 10bitbake.git/ (4 files in 2 dirs): (log message trimmed) Jan 11 10:36:29 bitbake/PRservice: Added no_hist mode and export/import. Jan 11 10:36:29 [YOCTO #1556] Jan 11 10:36:29 1. Added the package_arch into the index to the DB table. Because the Jan 11 10:36:30 change in PACKAGE_ARCH will results in different checksum, and it is Jan 11 10:36:30 better to have seperate PR value domains for differnt PACKAGE_ARCH of Jan 11 10:36:31 the same pakcage. Jan 11 11:41:10 florian: hi. I'f you're wondering about some night watch mail: that was your kernel with tworaz rootfs Jan 11 11:41:32 hi cs_nbp Jan 11 11:41:35 florian: thought might be important since it shows kbd error is not in the kernel Jan 11 11:42:07 cs_nbp: amazing... maybe my X is bad. need to switch to the later userland Jan 11 11:42:54 florian: isn't as bad as that - as far as I can say yours runs smoother and touchscreen is more responsive Jan 11 11:43:02 florian: only kbd missing Jan 11 11:43:20 florian: btw do I need to start dbus to see a pcmcia wifi card? Jan 11 11:44:05 cs_nbp: no, I don't think so Jan 11 11:44:20 florian: ah then it must be the driver ;) Jan 11 11:44:24 well bits like networkmanager might use it Jan 11 11:47:47 lets see if the repo is still messed up Jan 11 11:52:01 utterly broken still-.- Jan 11 11:52:39 I'm getting loads of ERROR: Nothing RPROVIDES 'modutils-depmod' Jan 11 12:01:03 * starts from complete scratch -.- Jan 11 12:07:14 btw don't buy dell printers, we just got 20 2350 that rattle like a truck and got alpha firmware that keps crashing Jan 11 12:11:01 perhaps they should use oe too :D Jan 11 12:24:01 cs_nbp: about your 'modutils-depmod', seems a blast from the past (I was bit long ago :) Jan 11 12:24:10 see bottom http://lists.linuxtogo.org/pipermail/openembedded-core/2012-January/015914.html Jan 11 12:26:23 grep in your layers and find the offending recipe Jan 11 12:27:27 nvm ant, I deleted everything and started from 0 Jan 11 12:27:32 just to be sure Jan 11 12:27:51 a lot of stuff was moved around and changed in one repo Jan 11 12:27:58 not enough if the metadata is broken :/ Jan 11 12:28:04 and stuff stopped working Jan 11 12:28:21 that's true Jan 11 12:28:40 but I wanted to be sure that it's not due to my tree Jan 11 12:28:43 ;) Jan 11 12:48:32 i want to include a tightvnc in my minimal-image.bb file unfortunately i get http://pastebin.com/tHGpvh4u Jan 11 12:49:09 but it works when i include vnc or x11vnc Jan 11 12:52:02 that is indeed unfortunate Jan 11 12:52:09 where are you getting the tightvnc recipe from? Jan 11 12:53:03 pb_, from vnc folder Jan 11 12:53:55 in recipes Jan 11 12:54:18 oh, right, legacy oe Jan 11 12:54:34 hm. well, there's nothing very obviously wrong with that. what happens if you try to build it directly with bitbake -b? Jan 11 12:54:57 it builds Jan 11 12:55:48 oh, I see Jan 11 12:55:58 it doesn't actually provide "tightvnc" Jan 11 12:55:58 heh Jan 11 12:56:18 so you need to either fix it so it does, or depend on something that it does provide. Jan 11 12:57:12 how ? Jan 11 12:57:39 how do i use RDEPENDS ? Jan 11 13:13:45 everything working out now it seems (not getting linux-dummy-rc1 kernel anymore :D) Jan 11 13:18:58 pb_, Jan 11 13:21:54 hi otavio, just read your pm Jan 11 13:47:21 how do i include a particular package in my recipe even after getting this error http://pastebin.com/tHGpvh4u Jan 11 13:53:20 brb Jan 11 13:57:22 florian: is there something online you can recommend to read that would help me optimize the kernel config? Jan 11 13:58:44 cs_nbp: not really... the point is that we should get built in whatever we need in any case... filesystem, usb, keyboard... Jan 11 13:59:53 florian: yes that's clear. Jan 11 14:00:40 florian:btw kernel building is working again Jan 11 14:09:53 cs_nbp: ok... will update later Jan 11 15:06:37 is there already an easy way to put GL/gl.h into a build? mesa-common-dev seems unknown Jan 11 15:10:58 cs_nbp: libgl-dev ? Jan 11 15:11:19 foerster: thanx ill try that Jan 11 15:12:05 mesa-2_7.10-r11.0/packages-split$ find . -name gl.h Jan 11 15:12:05 ./libgl-dev/usr/include/GL/gl.h Jan 11 15:12:32 i'm using an old oe-classic setup still, but that's where it is for me. Jan 11 15:12:58 ok im on oe-core but will find it;) Jan 11 15:13:14 cs_nbp: virtual/libgl Jan 11 15:13:58 cs_nbp: that's mesa or mesa-dri depending on distro config, but if you're looking for GL/gl.h needed ie for qemu-native build then you would need mesa-native which is not available in OE Jan 11 15:14:32 JaMa: webkit-gtk_svn needs it Jan 11 15:16:00 cs_nbp: there is patch for it on ML already.. Jan 11 15:16:26 JaMa: yeh need to look on ml more often it seems Jan 11 15:21:04 JaMa: JaMa: can you remember what the subject was? can only find something for webkit-efl_svn Jan 11 15:22:18 sup cs_nbp Jan 11 15:22:46 muriani hayZ Jan 11 15:23:46 cs_nbp: maybe it was on poky or yocto ML, but it was sent by Robert Day and I had some replies IIRC Jan 11 15:24:24 got florian's kernel and rootfs booting last night Jan 11 15:24:26 cs_nbp: ah no http://lists.linuxtogo.org/pipermail/openembedded-core/2012-January/015755.html Jan 11 15:24:29 played around in lxde a little Jan 11 15:24:33 it's sloooooow. Jan 11 15:24:54 X is, at least Jan 11 15:25:25 and my USB cable works Jan 11 15:25:33 muriani, cs_nbp: Did you ever see the Psion Linux stuff? :-) Jan 11 15:25:33 thats good news Jan 11 15:25:36 cs_nbp: and original thread https://lists.yoctoproject.org/pipermail/poky/2012-January/007401.html Jan 11 15:25:47 onoz i looked in the devell ml al the time Jan 11 15:25:49 florian: yes, yes I did. Jan 11 15:26:06 i use it on a daily basis Jan 11 15:26:07 It's why the nbp has had CE on it for so long :) Jan 11 15:26:14 muriani: not really fast... but a faster at least Jan 11 15:26:16 like now for irc :~D Jan 11 15:26:30 JaMa: thanks alot Jan 11 15:27:31 well 'fastness' of your kernel image was enough for me, florian Jan 11 15:28:07 started on atari 800xl so i dont expect so much :D Jan 11 15:29:05 well apart the 12 minutes waiting for ubi stuff to finish -.- Jan 11 15:31:38 hahaha Jan 11 15:31:38 yeah Jan 11 15:31:54 it's like waiting for an old C64 game to load Jan 11 15:33:03 hey. what's the status on oe-meta? i'd like to assess the impact of moving to oe-meta. are there any docs on this? Jan 11 15:33:19 oe-core / meta-oe, you mean :) Jan 11 15:33:30 bah if psionlx had wpa and emacs i prolly never would have thought of installing somehting else ;) Jan 11 15:34:36 alright, let's run a build. Jan 11 15:34:49 oh crap I need to add the rootdelay in boost, right? Jan 11 15:35:16 muriani y Jan 11 15:35:42 kergoth, yes :) Jan 11 15:36:10 is there a page on how to build a simple console image with it? Jan 11 15:36:16 what to checkout etc Jan 11 15:36:25 (i.e. git clone) Jan 11 15:36:48 CMoH|office, there are lots of different setups using the layers. different ways of pulling them in. angstrom has one, shr has one, poky does its own thing, but they're all relatively similar Jan 11 15:36:58 CMoH|office, something like http://sakrah.homelinux.org/blog/2011/03/using-openembedded-core-to-build-angstrom-for-qemu/ may be helpful for you Jan 11 15:37:07 thanks kergoth Jan 11 15:37:21 CMoH|office, alternatively: git clone --recursive https://github.com/kergoth/oe-core-template; cd oe-core-template; make Jan 11 15:37:26 or any number of others :) Jan 11 15:42:08 cs_nbp: oh, rootdelay is already there. Jan 11 15:42:34 root=/dev/sda2 rw rootdelay=1 console=tty0 init=/sbin/init quiet Jan 11 15:45:24 muriani: thats why you gotta put rootwait Jan 11 15:47:59 oh, missed that Jan 11 15:48:57 forgot my coffee this morning, lol Jan 11 15:51:10 so basically one should use bitbake master for oe-core? Jan 11 15:51:40 5y Jan 11 15:58:33 I wonder why my OE rebuids tar-replacement-native from time to time, I see no apparent reasons for it Jan 11 15:58:40 seems just to happen randomly Jan 11 15:59:06 Jin^eLD, if your host's tar is < version 1.23, it will build and use its own Jan 11 15:59:13 so depends on your distro Jan 11 15:59:20 mine gets built on some of my boxes, but not others Jan 11 15:59:36 hm, pcmcia (non-cardbus) network adapter for $40. Jan 11 15:59:37 it's 1.23 exactly Jan 11 15:59:41 seems a little steep Jan 11 16:07:03 Ah, ebay, $12 shipped. Jan 11 16:17:20 I added PREMIRRORS = "(ftp|http?)$://.*/.*/ ftp://myownftpserver.com/" to my distro conf file, but I get the error "ERROR: Error parsing /home/kobe/svn/test/openembedded/recipes/tasks/task-opie-games.bb: too many values to unpack". What is the correct way to specify a mirror? OpenEmbedded should look at my own server first, preferably. Jan 11 16:17:37 Note: I'm using OE classic Jan 11 16:27:06 Is it possible to set a server as a mirror for all packages (preferably by changing one of .conf files and not the .bbclass files)? Jan 11 16:28:41 kobe_: are you sure that $ is correct in your mirror statement above? Jan 11 16:28:49 kobe_: you may wish to make use of own-mirrors.bbclass Jan 11 16:29:52 I copied the line from one the existing .conf files, but it seems they use different methods too (and in many cases the PREMIRRORS line is commented out). Jan 11 16:30:00 kobe_: all you need to do is add this to your local.conf: Jan 11 16:30:06 INHERIT += "own-mirrors" Jan 11 16:30:16 SOURCE_MIRROR_URL = "http://my.server.url" Jan 11 16:32:01 (or ftp, or whatever) Jan 11 16:32:14 ah, okay. I see. Thanks, I'll try that. Jan 11 16:37:18 omg sry for nab question, but how do I correctly extract a patch from a html? I keep getting 'Patch format detection failed, no matter what I try. yes I know what a diff looks like. or is it my fault to try w git am? Jan 11 16:38:15 cs_nbp: where are you geetting the patch from? Jan 11 16:38:31 muriani: oe core mailing list Jan 11 16:39:06 mailing list mail isn't html. are you trying to get it out of a web mailing list archive? Jan 11 16:40:00 @bluelightning: Thanks, it works!! Jan 11 16:40:11 kobe_: np :) Jan 11 16:41:40 yes, i'm in the archive Jan 11 16:44:30 cs_nbp: is it in http://patchwork.openembedded.org/project/oe-core/list/ somewhere? Jan 11 16:44:39 because there has the mbox and patch downloads Jan 11 16:44:41 that work with git am Jan 11 16:44:50 or should, at least Jan 11 16:47:09 muriani: page seems good but got to verify it's the same patch Jan 11 16:58:30 muriani: txh, found it some pages further Jan 11 17:05:46 kergoth: ok I got the patch, put it openembedded-core/ and did a git am nameof.patch, and then only get ''patch does not apply" Jan 11 17:06:41 kergoth: got it from patchwork.openembedded.org Jan 11 17:06:47 cs_nbp, probably applies but with offsets. git am uses git apply, which is more picky than 'patch' Jan 11 17:07:01 i'd try with patch, or at least patch with --dry-run, to see whats really going on Jan 11 17:09:23 kergoth: maybe a stupid question, but can patch handle *.patch files? patch --dry-run file.patch is just sitting there doing nothing it seems Jan 11 17:09:41 that's not what patch's syntax is Jan 11 17:09:43 read the man page Jan 11 17:09:52 k;) Jan 11 17:09:55 patch --dry-run kergoth: tries to change files and get's rejected, all 'Hunks' failed Jan 11 17:18:18 kergoth: manually checking GNUmakefile.am, looks like the patch was already applied Jan 11 17:21:07 kergoth my webkit file revision is 101488, so the patch indeed doesn't apply Jan 11 17:21:38 kergoth: but webkit still fails, even being revision 101488 Jan 11 17:21:52 * kergoth shrugs, doesn't know anything about webkit builds Jan 11 17:35:24 figured it out, needed virtual/libgl like JaMa suggested (txhbtw) Jan 11 17:36:38 cs_nbp: always search ml Jan 11 17:36:46 there is lot of good info out there Jan 11 17:37:03 many times issues you run into are already dealt with Jan 11 17:39:12 khem: i will;) lesson learned Jan 11 18:04:21 -.- Jan 11 18:34:44 dint find anything in ml: fetching of ca-certificates failed because of checksums, I edited the .bb to reflect the different checksums. now it fetches, but i get ERROR: 'Unpack failure for URL' Jan 11 18:36:40 there's one post in ml about FILESEXTRAPATH but can't find any of this in the related dirs Jan 11 18:38:16 look at the rest of the error, or the do_unpack log. Jan 11 18:38:22 that isn't telling you *why* it failed tou npack Jan 11 18:38:29 that error message, that is Jan 11 18:40:06 -.- it says stdin: not in gzip format Jan 11 18:41:02 so the download file may be broken in repo? Jan 11 18:41:23 because from the outside itappears as .tar.gz Jan 11 18:41:25 your download is corrupt, yeah. rm -rf it and the .done file next to it in your download dir Jan 11 18:41:43 its common enough. sometimes the upstream .tar.gz actually hands back an html file with an error message or similar Jan 11 18:41:49 and get back the old checksums i guess Jan 11 18:41:52 the unpacking code doesn't call file to check the file type or anything Jan 11 18:42:00 k Jan 11 18:42:02 probably, yeah. could be you need to google for the archive to get the real file Jan 11 18:42:18 you could always cat the file thats in your download dir, see if thats what happened, or if it was just a truncated download or something Jan 11 18:43:02 yyyupp got a html Jan 11 18:43:39 should really have a troubleshooting cookbook of sorts, with common problems and debugging tips for those problems, and possible solutions Jan 11 18:43:47 like a more fleshed out faq Jan 11 18:48:56 yeah when i ran into all i could try to make a compendium ;) Jan 11 18:49:07 hehe Jan 11 18:49:51 yeh when i don't forget like w upower-.- Jan 11 18:50:35 thats one of the things i've been wanting to write up forever, but never seem to get around to. along with some sort of historical / conceptual overview to help people grasp the big picture, which seems to be a common problem Jan 11 18:50:37 heh Jan 11 18:53:00 autoreconf failed, unknown warning category 'cross' gonna look on ml Jan 11 18:54:39 re Jan 11 18:54:39 that message isn't the failure Jan 11 18:54:42 it's harmless Jan 11 18:54:43 look further Jan 11 18:54:51 note: warning != error Jan 11 18:55:25 hm Jan 11 18:55:27 WARNING: QA Issue: package libiconv contains bad RPATH $ORIGIN/../lib:/devel/arm/git/setup-scripts/build/tmp-angstrom_2010_x-uclibc/work/armv7a-vfp-d16-angstrom-linux-uclibceabi/libiconv-1.14-r1/image/usr/lib in file /devel/arm/git/setup-scripts/build/tmp-angstrom_2010_x-uclibc/work/armv7a-vfp-d16-angstrom-linux-uclibceabi/libiconv-1.14-r1/packages-split/libiconv/usr/bin/iconv Jan 11 18:55:37 looks like recipe needs some trimming Jan 11 18:58:33 hm and I wonder why khem dont hits Jan 11 18:58:33 kergoth: well... in oe we have quite some places where warnings and errors are mixed in a confusing way Jan 11 18:58:36 | ../../gl/sys/time.h:396:50: error: 'restrict' has a previous declaration as 'timeval* restrict' Jan 11 18:58:43 florian, heh, true. though its improving Jan 11 18:58:44 with gnutls and uclibc Jan 11 18:58:47 we're too verbose all around Jan 11 18:59:15 libtoolize: ERROR: Function do_configure failed Jan 11 18:59:42 makes ERROR:autoreconf failed Jan 11 18:59:43 again, thats not the real error Jan 11 18:59:47 you're not very good at parsing output Jan 11 19:00:00 im a beginner at this Jan 11 19:00:07 i'd suggest just putting the whole thing on a pastebin so we can analyze it for you :) Jan 11 19:00:34 well those were the only 2 error or ERROR strings in that file Jan 11 19:02:18 what should I look for instead? or not so easy to say? Jan 11 19:03:09 we'd really need to see it Jan 11 19:03:24 what you look for isn't as easy as a particular string to find in your text editor Jan 11 19:04:51 hi Jan 11 19:05:53 i'm trying to pull all the oe stuff, but i'm having trouble with meta smartphone Jan 11 19:06:08 has anybody had issues with this? is it down for some reason? Jan 11 19:08:46 'having trouble' isn't specific enough Jan 11 19:10:29 http://pastebin.com/pnvFmEER Jan 11 19:10:52 but wasn't there something w upower the last days? Jan 11 19:11:17 libupower-glib/Makefile.am:62: HAVE_INTROSPECTION does not appear in AM_CONDITIONAL Jan 11 19:11:17 looks like a likely candidate for the actual problem Jan 11 19:11:25 * kergoth wanders off to run errands Jan 11 19:11:38 yeeh i chased him away -.-;) Jan 11 19:13:51 cs_nbp: yes, that's the one Jan 11 19:14:17 I found another one: libusb1 depends on libxml2 Jan 11 19:15:57 hay im cs_nbp Jan 11 19:15:59 -.- Jan 11 19:16:12 battery of nbp needs reloading Jan 11 19:16:31 florian: btw libusb from libxml is a good one Jan 11 19:25:47 florian: what do I have to set instead of (0x0) to make the kernel look for rootfs on sd? thought bootparameters would do that -.- Jan 11 19:26:40 sry mean cf Jan 11 19:26:40 root=/dev/mmcblk0p1 or whatever sd is on nbp doesn't work? Jan 11 19:26:46 oh Jan 11 19:26:54 and i have /dev/sda2 Jan 11 19:26:55 as opposed to looking on nand? Jan 11 19:26:57 right Jan 11 19:27:00 and get kernel panic Jan 11 19:27:03 cs_nbp: yes indeed... which kernel do you use now and how are the parameters set? Jan 11 19:27:49 florian: self compiled kernel, in boost-bootcode /dev/sda2 ext2 - oh, that means ext4 don't work -.- Jan 11 19:28:59 cs_nbp: right... and I'm not 100% sure if there was ext2 in kernel but iirc yes. Jan 11 19:29:03 florian: think you can only mount ext2 as ext4, not the other way round.,. Jan 11 19:29:28 florian: in default config yes, in yours it's ext4 Jan 11 19:29:35 cs_nbp: yes... the ext2 driver does not know how to handle it Jan 11 19:29:46 florian: so for that upower stuff we simply need to wait? Jan 11 19:30:07 not the best idea ever since ext2 is more likely to get corrupted and the ext4 driver handles ext2 as well Jan 11 19:30:39 so change to ext4 for all builds that we do? Jan 11 19:30:49 ubifs not possible? Jan 11 19:32:08 just thought since booting from cf Jan 11 19:32:22 and ubifs is said to be flash friendly Jan 11 19:33:31 florian: ah yeh i should remind you of something you said Jan 11 19:33:46 build a kernel w/o ubifs warning for wince haters Jan 11 19:39:47 cs_nbp: oh right Jan 11 19:40:03 just want to send the latest patch to the list Jan 11 19:41:40 np wasn't hurrying Jan 11 19:44:42 * Jay7 is considering another round of job hunting.. Jan 11 19:46:01 hi jay7 Jan 11 19:46:09 long time not seen Jan 11 19:46:09 hi woglinde Jan 11 19:46:17 I'm always here :) Jan 11 19:46:21 just silent Jan 11 19:49:10 * florian shoots git-send-email Jan 11 19:50:34 Trying to put as much as possible into the subject line is not the best idea ever. Jan 11 19:51:42 florian it isnt Jan 11 19:52:25 the secretaries who tell us they can't work usually put the whole email in the subject Jan 11 19:53:21 git-send-email tries to do the same Jan 11 19:54:27 cs_nbp: our secretaries don't know what is subject line used for Jan 11 19:54:43 just sending emails with empty subject Jan 11 19:55:07 yes they did this once but now a subject is required or we won't send the mail Jan 11 19:55:15 Jay7: excessive content in subjects is like IP over DNS ;) Jan 11 19:55:23 hehe Jan 11 19:58:44 should I activate braille console? -.- Jan 11 19:58:53 do it! Jan 11 20:02:44 forian: while you wrestle w git-send-mail, here is the boot information of the default config: (0x0) Compressed ROM boot loader base address, (0x0) Compressed ROM boot loader BSS address,(root=1f04 mem=32M), Kernel command line type:Use bootloader kernel arguments if available Jan 11 20:15:26 cs_nbp: looks weird... why 32M? and root might want to become something like mmcblk0p1 or sda2 rather Jan 11 20:16:28 florian: I'm new to this so I thought it had to do w the 32m of flash Jan 11 20:17:00 cs_nbp: its about ram... Jan 11 20:20:27 florian: so I should put 128M rather than 32 Jan 11 20:21:09 or i'm shooting the wrong bird Jan 11 20:29:59 this webkit gl issue is a known one? Jan 11 20:30:15 florian: yes theres a patch, waita sec Jan 11 20:30:37 cs_nbp: yes... or just nothing Jan 11 20:31:08 florian: http://patchwork.openembedded.org/patch/18029/ , but that doesn't solve it, you have to bitbake virtual/libgl Jan 11 20:31:20 florian: it's looking for GL/gl.h Jan 11 20:31:46 florian: yes i now understand it's a classic boot parameters line Jan 11 20:31:56 florian:_ like eg in grub Jan 11 20:32:28 my webkit was up to date so when you refreshed today you don't need the patch Jan 11 20:34:40 your kernel and tworaz rootfs are not starting anymore, even not with a fresh reflash of everything Jan 11 20:35:05 can't open /dev/console no such file or directory Jan 11 20:35:07 cs_nbp: Not sure if this is related... for me it looks like a missing dependency Jan 11 20:35:09 -.- Jan 11 20:35:22 cs_nbp: unpack the filesystem as root Jan 11 20:35:32 florian: well i bitbaked virtual/libgl and webkit built fine then Jan 11 20:35:45 florian: I always do Jan 11 20:36:19 florian: tar xfvz rootfs.tar.gz --numeric-owner Jan 11 20:36:28 florian: as root Jan 11 20:37:36 florian: problem solved - there's a SD card in there again-.- geezus Jan 11 20:39:09 cs_nbp: ah... it mounts the sd but there was no useful fs to boot into on it? Jan 11 20:39:19 florian: yes -.- Jan 11 20:39:41 florian: rootfs was on CF card and /dev/console was there, I was quite confused Jan 11 20:40:05 florian: dumb thing is the dummy sd looks just like a real sd card Jan 11 20:40:19 florian: so you never really look there Jan 11 20:40:30 florian: ill mark the dummy pink Jan 11 20:41:22 cs_nbp: the script searching for rootfs is rather primitive. maybe we should test for /sbin/init on them Jan 11 20:41:28 florian: is the bootdisk the only info the kernel really needs or should i put /dev/console (or a different cli) and other information there also Jan 11 20:42:07 florian: you mean look on one drive after the other for /init? Jan 11 20:42:18 florian and only then try to boot Jan 11 20:42:30 right Jan 11 20:43:05 florian: btw i hope it doesn't beep everytime i type florian:, I switched my irc sound off Jan 11 20:43:18 hello, is there a popular distribution which is based on oe-core which can be used for pandaboard ES ? i am using SHR for my phone but i need something with KDE libs support for pandaboard ES. i would like to setup a small streaming client for my living room Jan 11 20:43:45 cs_nbp: I never had it enabled ;) Jan 11 20:44:49 florian: anyways there's still something wrong w upower http://pastebin.com/pnvFmEER Jan 11 20:45:19 florian: l85++ Jan 11 20:47:21 cs_nbp: add gobject-introspection-native to depends Jan 11 20:50:31 florian: ah yes, i heard that once already in the last days-.- Jan 11 20:51:44 * florian sends another patch Jan 11 20:52:30 florian: patchmaster F, eh Jan 11 20:52:44 this reminds me to a large pile of oe-classic additions and fixes in my queue Jan 11 20:53:07 you mean they could be converted? :D Jan 11 20:53:14 but these I can push myself Jan 11 20:53:17 no Jan 11 20:53:20 i know Jan 11 20:53:51 I haven't started converting... not yet... after Embedded World maybe Jan 11 20:54:00 florian: i loved oe-classic ... now all is splitted Jan 11 20:54:23 btw did you find out why your kbd isn't working in lxde? Jan 11 20:54:42 muriani said for him it is working Jan 11 20:54:47 with your image Jan 11 20:55:07 nschle85: I still love it :-) Jan 11 20:56:23 florian: i miss it :-) but the future is OE CORE Jan 11 20:58:37 florian: do you know a good distribution for pandaboard ES ? i need to setup a stream client, connected to TV, software is XMBC or a kde program Jan 11 20:58:42 nschle85: I'm still using it... with oe-core what I currently do is too much effort. Jan 11 20:59:09 nschle85: Sounds like a job for Angstrom Jan 11 20:59:35 florian: is angstrom core based ? Jan 11 21:03:11 florian, don't think Angstrom installs the accelerated drivers though, Jan 11 21:03:30 So might be a bit hard to get the fullscreen playback etc Jan 11 21:05:26 phdeswer: maybe... but I remember it used to have this stuff for some omap3 Jan 11 21:07:46 -ccleanall virtual/kernel is all I need before rebuilding the kernel, right Jan 11 21:09:21 phdeswer: florian: on desktop i like ubuntu, but there i need a small dedicated distribution... ok ill investigate angstrom ... Jan 11 21:10:25 nschle85, for small it should work. Did it not so long ago myself. But still need to figure out how to get the binary drivers on the board Jan 11 21:11:56 phdeswer: binary drivers ? Jan 11 21:12:28 also there's this http://iso.qa.ubuntu.com/qatracker/milestones/163/builds/5326/testcases/696/results if you like ubuntu Jan 11 21:12:34 didn't test that myself Jan 11 21:13:45 sorry the site is this http://iso.qa.ubuntu.com/qatracker/milestones/205/builds Jan 11 21:14:25 nschle85, afaik the pandaboard has some binary drivers for accelerated display and 3D Jan 11 21:15:28 phdeswer: hmm ok they are currently packaged in ubuntu 11.10 for pandaboard Jan 11 21:15:56 indeed. Jan 11 21:15:59 Btw http://rsalveti.wordpress.com/2012/01/06/hw-video-decode-and-xbmc-ubuntu-linaro/ Jan 11 21:16:18 i arrived at * extract_archive: Cannot create symlink from ./var/run to 'volatile/run': File exists. again :D Jan 11 21:16:46 phdeswer: thought all software is open :-( Jan 11 21:19:12 nschle85, unfortunately not. http://www.omappedia.com/wiki/PandaBoard_FAQ#What_will_Graphics.2FGFX_support_be_like.3F_Will_drivers_be_released.3F Jan 11 21:20:35 cs_nbp: I will be there pretty soon too Jan 11 21:31:42 florian: i removed all findable references and browed ml, dint find a solution Jan 11 21:32:33 florian: i mean i added rm -rf {D}{localstatedir}/run/* to the install{} function of some recipes Jan 11 21:32:35 * florian has ~10 tasks left Jan 11 21:32:51 florian me only 3 or 4 Jan 11 21:33:58 I wonder if putting gtk+ recipes under a recipes-gnome directory is a good idea Jan 11 21:35:21 perhaps if you want to build glade -.- Jan 11 21:56:54 florian: see any light at the end of the tunnel? can't find errors in the recipes, can it be something IN the actual packages themselves? Jan 11 21:57:10 florian: that get installed? Jan 11 21:57:28 florian: and perhaps write to var/run Jan 11 22:10:11 cs_nbp: one sec... just have to fix another download Jan 11 22:14:09 florian: don't let that get a habit ;) Jan 11 22:14:53 btw what does the red led signify? it's always-on since the screen became white in an auto-suspend attempt Jan 11 23:00:42 cs_nbp: sorry, it di not fail here Jan 11 23:00:48 eh did Jan 11 23:01:29 classic fail? var/run? Jan 11 23:01:34 * florian *hides* Jan 11 23:01:47 cs_nbp: no failure here... Jan 11 23:02:00 florian: ah -.- here only half failure Jan 11 23:02:35 florian: i tested my kernel, then slid out the sd card again and it looked for ext2 filesystem, wonder where it gets that idea from Jan 11 23:03:11 florian: can I specify that on bkernel bootparameter line too? Jan 11 23:03:21 cs_nbp: you added "rootwait"? Jan 11 23:03:30 florian: yes Jan 11 23:03:40 florian: btw at next checkout its kernel 3.2 Jan 11 23:03:56 cs_nbp: it waits till the specified root device appears Jan 11 23:04:10 florian: yes but there is no ext2 device Jan 11 23:04:16 florian: only ext4 Jan 11 23:04:50 ext2 is fallback Jan 11 23:04:56 ah ok Jan 11 23:05:11 i specified nothing, neither in bootloader nor in kernel Jan 11 23:05:24 so it takes that Jan 11 23:05:51 florian: btw does that mean you solved the var/run thingy? what is it? Jan 11 23:06:38 cuz im sd carding back and forth to try to resolve dependencies for my wifi driver Jan 11 23:10:35 cs_nbp: sort of... I did not run into it for some reason Jan 11 23:10:54 *looks suspiciously at florian Jan 11 23:13:01 I updated all repositories before I started and changed distribution configuration back to Jlime Jan 11 23:15:24 k ill try that too, did that at 15h today Jan 11 23:15:44 i got jlime too Jan 11 23:15:51 yeah, I'm the same Jan 11 23:16:08 updated, distro=Jlime Jan 11 23:16:23 may be time to wipe it all and start fresh... again. Jan 11 23:16:43 updated all n ow, just new recipe for kernel 3.2 Jan 11 23:16:50 muriani: Depends on when you intend to go to bed and how fast your machine is ;) Jan 11 23:16:52 the new 3.2.0 recipe in meta-netbookpro boots nice and quick though Jan 11 23:16:59 florian: it's not bad. Jan 11 23:17:23 45min for a full build? Jan 11 23:17:29 haven't really timed it. Jan 11 23:17:50 muriani: for the lxde image? this would be pretty good Jan 11 23:18:07 yeah, I'm building on a pretty nice box Jan 11 23:18:49 but I die at do_rootfs Jan 11 23:19:21 same here Jan 11 23:19:25 strange Jan 11 23:19:32 but for tworaz and florian it workd Jan 11 23:19:50 so for the nabs it doesn't work? oe haz nab detect? Jan 11 23:20:07 I found 2GB PCMCIA harddisks at the office :) Jan 11 23:20:20 heh Jan 11 23:20:26 under a pile of dust and dead rats? Jan 11 23:20:33 niice Jan 11 23:20:44 whoo, it seems my pcmcia wlan Jan 11 23:20:48 (prism2) Jan 11 23:21:13 sometimes oe is really evil... it might be an idea to check latest oe core changes if there is any suspicious change. this might speed up things a little bit Jan 11 23:21:16 nothing here to connect to Jan 11 23:21:30 but at least it works Jan 11 23:21:40 florian: you don't know by chance where menuconfig saves the .config? I can only see it's {D}/.config Jan 11 23:21:45 maybe I'll use this for workign from the downstairs tonight Jan 11 23:21:50 cs_nbp: well... piles of dust and scary things like iPAQ boxes Jan 11 23:22:03 florian: ...the dead rats Jan 11 23:22:30 cs_nbp: it should be in ${S} where menuconfig runs Jan 11 23:22:30 ...if I can get ssh installed Jan 11 23:22:41 isn't it? Jan 11 23:22:55 not in this rootfs? Jan 11 23:23:20 omg bash and less but not ssh? (or emacs?) Jan 11 23:23:51 cs_nbp: the rats are one floor above at the attic... you know the dec rats? Jan 11 23:24:07 muriani: dropbear maybe? Jan 11 23:24:35 florian: yeh http://www.cedmagic.com/history/dec-rainbow.jpg Jan 11 23:25:14 :-) Jan 11 23:25:32 well at work we have a sparcstation Jan 11 23:25:45 prevents a door from closing Jan 11 23:26:38 we have a wohole collection of classic non-x86 workstations Jan 11 23:27:02 florian: i run it from oe/setup-scripts/ but theres notten in deh Jan 11 23:27:15 even not a .config Jan 11 23:27:36 oops... never tried that one Jan 11 23:27:57 yes and i got the impression i can make one kernel and then it forgets the config Jan 11 23:28:26 cd .. Jan 11 23:29:19 cs_nbp: bitbake -c devshell linux-netbookpro Jan 11 23:29:47 argh... why this does open a new terminal these days? Jan 11 23:30:08 a very clever idea if you are connected to a remote server over a thin line Jan 11 23:30:08 florian: you mean menuconfig? Jan 11 23:30:30 i use nxclient for that -.- simple ssh not enough Jan 11 23:31:15 yes... but opening xterm you end uo in X11 forwarding! Jan 11 23:31:37 florian: yeah that sux big time Jan 11 23:31:56 aaand i run into var/volatile fail again Jan 11 23:32:13 maybe, just maybe, is there a packe missing on my system? Jan 11 23:32:31 do I perhaps need a header locally that isn't there Jan 11 23:32:47 only way i can explain that tworaz and florian are not seeing it Jan 11 23:33:14 cuz repos should be identicall for all 4 Jan 11 23:33:38 uh... 6k/s from linuxtogo.org - I need to take a look why it is that slow Jan 11 23:33:53 thought that was normal Jan 11 23:34:15 ls Jan 11 23:34:35 not really Jan 11 23:35:02 this time it even seems to be not guilty at all... it seems my local connection is that slow Jan 11 23:36:33 well this devshell put me somewhere in tmp/work , but still theres no .config Jan 11 23:36:42 theres Kconfig but thats lame :p Jan 11 23:36:51 Jin|away: did you get to test the locale fix Jan 11 23:37:54 florian: yeah there's no dropbear here either Jan 11 23:38:00 thjat's server though, right? Jan 11 23:38:05 I need ssh client :P Jan 11 23:38:41 do_configure_prepend() { Jan 11 23:38:42 install -m 0644 ${S}/arch/arm/configs/netbookpro_defconfig ${S}/.config Jan 11 23:38:42 } Jan 11 23:38:47 cs_nbp: that's the working directory... you should have one there after do_configure Jan 11 23:38:49 thats whats in the kernel bb Jan 11 23:39:02 well if I can get this on the network at home tonight, I'll see if I can opkg it and get online Jan 11 23:39:12 muriani: ok... we might want to add them to the image :) Jan 11 23:39:18 yeah :) Jan 11 23:41:30 sup now i found 2 configs Jan 11 23:42:15 guess the one here is used build/tmp-angstrom_2010_x/sysroots/netbookpro/kernel/.config Jan 11 23:42:31 and the one here is the 'mod' build/tmp-angstrom_2010_x/work/netbookpro-oe-linux-gnueabi/linux-netbookpro-3.1.8-r1/git/.config Jan 11 23:43:00 well that one's for 3.1.8, you're building a 3.2.0 kernel now right? Jan 11 23:43:29 ah yeh so that config wasn't even saved it seems Jan 11 23:43:41 or locatedb too old Jan 11 23:44:10 I use find in that case Jan 11 23:44:52 next time Jan 11 23:46:32 cs_nbp: I'm pretty sure you get this error because you are building angstrom and I build Jlime Jan 11 23:47:02 I build Jlime Jan 11 23:47:06 same error. Jan 11 23:47:17 MACHINE = "netbookpro" Jan 11 23:47:17 DISTRO = "Jlime" Jan 11 23:47:24 BUT Jan 11 23:47:48 i did setup-scripts http://sakrah.homelinux.org/blog/2011/03/using-openembedded-core-to-build-angstrom-for-qemu/ Jan 11 23:47:56 and muriani too Jan 11 23:48:00 srey muriani Jan 11 23:48:04 *shame* Jan 11 23:48:10 yeah, because suppoesdly that fixes EVERYTHING Jan 11 23:48:11 lol Jan 11 23:48:22 yeh it did, we found a complete new error Jan 11 23:48:25 I guess that could be the only other thing we have in common. Jan 11 23:48:38 so k start from scratch with oe-core tree i guess Jan 11 23:48:40 barf Jan 11 23:48:44 I've got no problem with going back to pulling my own trees Jan 11 23:49:02 setup-scripts pulled a bunch of crap I don't need anyway Jan 11 23:49:11 yeh Jan 11 23:49:23 like recipes-mangochutney Jan 11 23:49:31 ;) Jan 11 23:50:06 but that means doenloads be gone Jan 11 23:50:20 or can i copy that over Jan 11 23:50:45 shold be able to copy it over Jan 11 23:50:51 *copies over Jan 11 23:52:05 florian: so yeh you win setup-scripts optimises stuff for angstrom Jan 11 23:52:18 florian: so maybe distro was overridden Jan 11 23:52:30 florian: how mmany buld steps you had approx Jan 11 23:52:42 it says jlime in the bitbake stats when it starts Jan 11 23:52:43 florian: i had 4696 Jan 11 23:53:24 cs_nbp: 4008 Jan 11 23:54:15 florian: yeh thought so Jan 11 23:54:21 obvious discrepance Jan 11 23:54:29 even if one of us is wrong by 200 Jan 11 23:55:06 bah whut y only 1.5MB/s Jan 11 23:55:50 florian: ah but now i remember needing meta-oe too then Jan 11 23:55:54 cs_nbp: angstrom adds some bits to every image... avahi, dropbear... Jan 11 23:56:39 cs_nbp: it depends on meta-oe and oe-core IIRC Jan 11 23:56:54 i mean for tworaz stuff Jan 11 23:56:58 right Jan 11 23:57:00 for tworaz Jan 11 23:57:12 i put meta-oe NEXT to oe-core, right Jan 11 23:57:30 as long as the bblayers are config'd right Jan 11 23:57:46 meta-oe has meta-otherstuff under it Jan 11 23:57:54 iirc Jan 11 23:58:02 thats prolly why i get tempted to put it there Jan 11 23:58:36 yeah I did all my git clones next to each other Jan 11 23:58:55 anyhoo folks, time to head home. Might be able to hop on later to bang at this some more Jan 11 23:59:14 well ur in the us right? Jan 11 23:59:18 yeah Jan 11 23:59:25 here its somewhat after going to bed time ^^ Jan 11 23:59:29 CST Jan 11 23:59:32 heh Jan 11 23:59:36 CET Jan 11 23:59:49 so ur 7 hours behind Jan 11 23:59:49 yeah, but I've got a woman and kids and such at home :) Jan 12 00:00:02 i got a woman and no kids yet Jan 12 00:00:30 muriani: same here with one difference... I'm at home already :) Jan 12 00:00:39 florian: nice Jan 12 00:00:53 if I have some free time I'll poke at it some more Jan 12 00:00:54 muriani: CET as well :) Jan 12 00:01:04 see y'all Jan 12 00:01:10 muriani: cu Jan 12 00:01:18 cyaz Jan 12 00:05:54 all networks are dead slow today Jan 12 00:06:01 * florian -> sleep Jan 12 00:07:52 ok gn8 Jan 12 00:08:06 ill dip on another run and goto bed too Jan 12 00:36:03 k gn8 all Jan 12 00:36:11 or good morning, depends ;) **** ENDING LOGGING AT Thu Jan 12 02:59:57 2012