**** BEGIN LOGGING AT Mon Jul 12 02:59:57 2010 Jul 12 04:52:09 03Vladimir Sorokin  07org.openembedded.dev * reab31c26dc 10openembedded.git/recipes/conntrack-tools/ (conntrack-tools_0.9.14.bb files/init): Jul 12 04:52:09 conntrack-tools: fix restart command in init script Jul 12 04:52:09 Signed-off-by: Vladimir Sorokin Jul 12 04:52:09 Signed-off-by: Roman I Khimov Jul 12 04:52:12 03Roman I Khimov  07org.openembedded.dev * rac8b14020e 10openembedded.git/recipes/corosync/corosync_1.2.3.bb: Jul 12 04:52:12 corosync: update to version 1.2.6 Jul 12 04:52:12 * bug fix release, considered as safe upgrade Jul 12 04:52:12 Signed-off-by: Roman I Khimov Jul 12 04:52:13 03Vladimir Sorokin  07org.openembedded.dev * r9d4bc378ba 10openembedded.git/recipes/corosync/ (corosync_1.2.6.bb files/init): Jul 12 04:52:22 03Roman I Khimov  07org.openembedded.dev * rf65a2a31d3 10openembedded.git/recipes/pacemaker/ (4 files in 2 dirs): Jul 12 04:52:22 pacemaker: update to version 1.0.9.1 Jul 12 04:52:22 * bugfix release, considered as safe upgrade Jul 12 04:52:22 Signed-off-by: Roman I Khimov Jul 12 06:08:03 gm Jul 12 06:32:17 grg, no Jul 12 06:35:29 eFfeM_work, ? Jul 12 06:35:50 (04:25:00 AM) grg: 'bitbake gcc-cross' builds perl too? Jul 12 06:35:54 19:25 < grg> 'bitbake gcc-cross' builds perl too? Jul 12 06:35:59 the answer is "no" Jul 12 06:36:02 oh. Well, it in fact does. Jul 12 06:36:31 grg: who pulls it in Jul 12 06:36:55 khem, not a clue. The dependency graph is incomprehensible to me. Jul 12 06:36:56 i have a build here that uses a very small image; it has a gcc-cross dir in work, but no perl dir Jul 12 06:37:28 try rm tmp; bitbake gcc-cross and see what you get Jul 12 06:37:49 eFfeM_work, well i have been testing various gcc versions and tried a gcc-cross target. e.g. mv tmp tmp-foo; bitbake gcc-cross Jul 12 06:37:59 and it definitely tries to build perl Jul 12 06:38:04 grg: bitbake -g -u depexp gcc-cross Jul 12 06:38:09 thinking of it, if you have a different gcc than I have it might be added (i used 4.1.2 in this test) Jul 12 06:38:20 this is with gcc 4.3.4 Jul 12 06:39:13 yeah, did a quick grep, found this: Jul 12 06:39:14 gcc-4.3.1/gcc-4.3.1-SYSROOT_CFLAGS_FOR_TARGET.patch Jul 12 06:40:22 don't see a depends or so though Jul 12 06:41:07 eFfeM_work: gcc build depends on host perl Jul 12 06:41:19 ah ok, ofc Jul 12 06:41:21 so if it builds perl-native thats ok Jul 12 06:41:34 its building both perl-native and perl Jul 12 06:41:41 but target perl I dont know Jul 12 06:41:46 and the dependency explorer doesn't mention either Jul 12 06:41:47 why its being pulled in Jul 12 06:42:45 grg: hmmm strange Jul 12 06:43:10 grg: if you are using git master then it should generate 3 dot files with -g Jul 12 06:43:21 and look at the pn one Jul 12 06:43:22 "gcc-cross-initial.do_package_stage_all" -> "perl.do_package_stage" Jul 12 06:43:36 oh Jul 12 06:47:09 grg: is it on .dev Jul 12 06:47:23 khem, yep Jul 12 06:48:25 grg: do you have run.do_package_stage_all.xxxx file in gcc-cross-initial build tree Jul 12 06:50:02 khem, http://pastebin.com/uXHBnRwn Jul 12 06:52:26 grg: I dont have this task at all Jul 12 06:53:00 khem, ? wtf have i done? Jul 12 06:53:36 grg: are you using package staging Jul 12 06:53:55 khem, i have no idea what that is or how i would know if im using it Jul 12 06:54:39 grg: r u using angstrom Jul 12 06:55:07 khem, minimal Jul 12 06:55:25 err Jul 12 06:55:28 hmmm Jul 12 06:55:32 I use minimal too Jul 12 06:55:37 :P Jul 12 06:56:33 but whatever you pasted doesnt refer perl Jul 12 06:56:44 so its strange why bitbake thinks otherwise Jul 12 06:56:55 what does your local.conf look like Jul 12 06:57:28 khem, http://pastebin.com/LtVyS23t Jul 12 06:59:06 INHERIT = "package_ipk" Jul 12 06:59:48 you dont need that Jul 12 07:02:25 morgen Jul 12 07:02:37 hrw: hey Jul 12 07:03:07 grg: I dont have this task for qemumipsel Jul 12 07:03:17 grg: but I am using gcc 4.5 Jul 12 07:04:55 khem, my 4.5 tmp dir doesn't have it either Jul 12 07:05:17 hmm ok some sanity returns Jul 12 07:05:30 grg: that means it special to 4.3.x Jul 12 07:05:53 khem, or i've done something weird to gcc recipes Jul 12 07:06:18 did u muck with it Jul 12 07:06:21 hi hrw Jul 12 07:07:38 khem, gcc_4.3.4.bb currently looks like this: http://pastebin.com/ZyPd3LE2 Jul 12 07:08:08 Q: I need to build two different u-boots for the same machine (for nand and sdcard), how is that best done ? Jul 12 07:08:30 I need one build with UBOOT_MACHINE_calamari = "MPC8536DS_SDCARD_config" and one with UBOOT_MACHINE_calamari = "MPC8536DS_NAND_config" Jul 12 07:08:52 but these use the same dir etc Jul 12 07:09:05 grg: why would one need ibad patch for target gcc I wonder Jul 12 07:09:25 what would be the best way/ two recipes? build twice in the recipe ? Jul 12 07:09:32 khem, sorry, that's gcc-cross_4.3.4.bb Jul 12 07:09:47 eFfeM_work: check openmoko uboot recipe? Jul 12 07:09:59 eFfeM_work: two sepate recipes I would think Jul 12 07:10:00 eFfeM_work: they built 3-6 versions on one run Jul 12 07:11:50 heh a for loop Jul 12 07:11:53 \nice Jul 12 07:12:37 UBOOT_MACHINES = "SDCARD NAND" and then Jul 12 07:13:00 for mach in ${UBOOT_MACHINES};do oe_runmake ${mach}_config Jul 12 07:13:05 and stuff Jul 12 07:13:31 may be UBOOT_CONFIG would have been a better name Jul 12 07:13:54 khem, hrw thanks, i was not aware that openmoko already solved this Jul 12 07:14:36 actually I would have preferred if they went for one u-boot image that could be put in nand and on sd, but didn't see an easy patch to achieve that Jul 12 07:14:56 grg: you can temporarily revert your changes to this recipe and see Jul 12 07:15:34 khem, yes i'll keep playing. Don't have time for it right now however Jul 12 07:15:45 sleep time here Jul 12 07:15:55 4am yet? Jul 12 07:16:01 g night Jul 12 07:16:08 cya Jul 12 07:16:15 heh no is midnight Jul 12 07:16:33 nite khem (9.15 am here) Jul 12 07:43:15 I built the angstrom-gnome-image successfully using bitbake... However, when I try to boot the beagleboard from the SD card, it is stuck at "INIT: Version 2.86 booting" Jul 12 07:43:23 The x11-image on the same build environment works fine Jul 12 07:43:30 Am I missing something here Jul 12 07:43:31 ? Jul 12 08:38:34 hi pb Jul 12 08:38:37 screwgoth: this seems more a distro or machine issue than an oe issue Jul 12 08:38:52 eFfeM_work regards your gcc-patches Jul 12 08:38:59 yes Jul 12 08:39:05 you hardcoded some statical path from your layout Jul 12 08:39:13 ouch Jul 12 08:39:36 hmm... maybe ... Lemme try thr angstrom and beagleboard channels .... Jul 12 08:39:59 hello Jul 12 08:41:00 woglinde: thanks for the pointer, i see it, will fix it, saw it before, then repaired it, but apparently it ended up in it again after fixing the crtstuff part Jul 12 08:41:29 probably generated the patch after testing the work and so it crept in, sry about that Jul 12 08:41:51 effem no prob :) Jul 12 08:42:41 feel free to ack with a remark that the ack is after this is fixed (will try to commit it this evening @ home Jul 12 08:43:04 hm its already in Jul 12 08:43:16 ? Jul 12 08:44:58 woglinde, I assume you are talking about the INCLUDES fix. according to cgit it is not in: http://cgit.openembedded.org/cgit.cgi/openembedded/tree/recipes/gcc/gcc-4.1.2.inc Jul 12 08:45:56 hm Jul 12 08:46:47 when is it added according to your git log ? Jul 12 08:47:57 hm Jul 12 08:48:09 maybee I missread the email Jul 12 08:49:15 it is a proposed patch, but as it related to the toolchain I posted for review/ack Jul 12 08:49:36 okay sorry Jul 12 08:49:48 is no problem :-) Jul 12 08:50:26 woglinde: bei Gelegenheit, pls look at root/recipes/dietlibc/dietlibc-0.32/diethome.patch Jul 12 08:51:27 + DIETHOME=$(PWD) bin... there is a space before bin- Jul 12 08:51:46 now it fails, how could it work before? Jul 12 08:52:34 ant fix it Jul 12 08:52:38 dont know either Jul 12 08:52:40 *g* Jul 12 08:52:51 I built it too in the past ^_^ Jul 12 09:12:26 morning Jul 12 09:12:31 hi jay7 Jul 12 09:20:39 03Sebastian Krzyszkowiak  07org.openembedded.dev * r03b0bfff03 10openembedded.git/ (3 files in 2 dirs): Jul 12 09:20:39 shr-splash: don't use PROVIDES, as it's not used when resolving runtime deps nor image content; use SHR_SPLASH_THEME variable instead Jul 12 09:20:39 Signed-off-by: Martin Jansa Jul 12 09:34:19 kergoth_: ping Jul 12 10:23:34 hi, do you have any problems with giotorious ? since 2 days i cannot clone webkit-efl Jul 12 10:24:35 I've seen reports here about this Jul 12 10:24:52 Gnutoo has complaining afaik Jul 12 10:29:56 dcordes: hi Jul 12 10:30:14 I've got the exact sime error like someone else pastebined it Jul 12 10:30:21 but I can't get a solution Jul 12 10:30:23 http://pastebin.com/XfzKVwGC Jul 12 10:30:27 here is the error Jul 12 10:30:34 do you know why this occurs? Jul 12 10:34:05 leviathan: lsof receipe somehow doesn't unpack losf_source.tar, i've unpack it manually to oe/tmp/work/armv6-novfp-oe-linux-gnueabi/lsof-4.83-r0 and it works for me Jul 12 10:34:23 zenob: oh, ok, thx Jul 12 10:34:38 shouldn't we fix this recipe? Jul 12 10:34:47 if it breaks building for everyone Jul 12 10:35:08 its a clear indicator, that the recipe itself is broken Jul 12 10:35:10 *thought* Jul 12 10:35:47 leviathan: hi there Jul 12 10:35:56 leviathan: I think I've also seen lsof trouble recently. Jul 12 10:36:27 so we should really fix this issue Jul 12 10:36:49 but I don't have a log handy unfortunately Jul 12 10:37:02 you can check tinderbox for dcordes or de-nl_dreamteam builds Jul 12 10:37:03 its an elementary package, so no permission to be broken ;-) Jul 12 10:37:17 ok Jul 12 10:37:24 lol, dreamteam? Jul 12 10:37:30 for htcdream? Jul 12 10:37:44 no general htc-linux team Jul 12 10:37:50 ah, ok Jul 12 10:37:57 and its called dreamteam just so Jul 12 10:37:58 ok Jul 12 10:38:00 :-) Jul 12 10:38:10 but regarding dream.. I wanted to talk to you. I am fighting the androidism monster now, too. on htcleo / hd2 Jul 12 10:38:11 the other would be a funny coincident Jul 12 10:38:18 and I wanted to ask you for some help Jul 12 10:38:20 dcordes: ok Jul 12 10:38:36 I recently got an htc universal Jul 12 10:38:40 for development work Jul 12 10:38:42 he has two Jul 12 10:38:46 so he gave me one Jul 12 10:38:50 to get SHR running there Jul 12 10:38:51 leviathan: we have the full spectrum of androidism. touchscreen, sensors etc Jul 12 10:39:03 ah unviersal is a nice device Jul 12 10:39:03 dcordes: of course Jul 12 10:39:06 vga and good keyboard Jul 12 10:39:08 jep Jul 12 10:39:33 you can find builds on the nostalgy htc-linux.org section Jul 12 10:39:34 :) Jul 12 10:39:52 'old website for pxa devices' Jul 12 10:40:16 if you have fso/shr going it will be nice to upload some builds on new section then Jul 12 10:40:27 yes Jul 12 10:40:37 I thought I will just upload it to my repo first Jul 12 10:40:40 ftp.o2s.ch Jul 12 10:40:51 then anyone can "mirror" it Jul 12 10:41:04 nice Jul 12 10:41:08 now it seems to build Jul 12 10:41:17 so the htcuniversal shr-image will be done soon Jul 12 10:41:18 :-) Jul 12 10:42:32 I bet oe has mega obsolete kernel for uni Jul 12 10:42:51 let me know then I will mirror it Jul 12 10:43:10 I hope they won't look to confused in korea when I go through customs with that amount of oe devices xD Jul 12 10:43:39 ah, thats an htcuniversal with shr Jul 12 10:43:43 and htcdream with shr Jul 12 10:43:48 two fso2 freerunner Jul 12 10:43:51 lol vacation ? Jul 12 10:44:01 one with mokotouch and one with shr Jul 12 10:44:05 dcordes: nope Jul 12 10:44:10 negotiations Jul 12 10:44:14 ^^ Jul 12 10:44:48 lol you old busienssman Jul 12 10:45:00 after I've written my examina, around next month Jul 12 10:45:19 I still need to look for exact data Jul 12 10:45:26 dcordes: yea Jul 12 10:45:56 my comps planing a branding of an shr phone under samsung Jul 12 10:46:05 so I'll need to talk to this guys there Jul 12 10:46:10 who have the full control Jul 12 10:46:16 present them our architecture Jul 12 10:46:38 guess initial we will just have mickey, me and gnutoo under NDA working on basic stuff Jul 12 10:46:47 and then push a developer phone release Jul 12 10:46:51 limited edition Jul 12 10:46:55 for other devs Jul 12 10:47:02 count me in I need a job :) Jul 12 10:47:09 yea Jul 12 10:47:22 would be nice, if they would even sponser us Jul 12 10:47:30 sometimes it comes better then you hope Jul 12 10:47:42 last meeting I earned 900.- in 45minutes Jul 12 10:47:58 could be such a beautiful happening again Jul 12 10:47:59 ^^ Jul 12 10:50:25 I'm frankly, thats the reason why they are so generous I guess Jul 12 10:51:56 and I also always facturate the exact work time and hold my own failures back, so also count effective work time, not learning and so Jul 12 10:52:12 and you need to be friendly :-) Jul 12 10:52:43 dcordes: so you get jobs Jul 12 10:52:47 and about contacts Jul 12 10:52:49 ^^ Jul 12 10:52:59 e.g. me, btw, I'll look Jul 12 10:53:15 dos1: what about dos2.3 Jul 12 10:53:20 :-) Jul 12 10:56:01 dcordes: hey Jul 12 10:56:12 leviathan: that's awesome! I'm not a real programmer as you know but if you need any help let me know Jul 12 10:56:19 Jay7: hi Jul 12 10:56:33 dcordes: as I understand you have some htc device? Jul 12 10:56:39 Jay7: plenty Jul 12 10:57:06 Jay7: I also do stuff with htc devices I don't have. http://htc-linux.org Jul 12 10:57:06 can we test kexecboot on it? :) Jul 12 10:57:50 Jay7: on most hacked htc devices we boot from WinCE using the HaRET bootloader Jul 12 10:57:56 ah.. Jul 12 10:58:16 but what about real android devices? Jul 12 10:58:31 well bootloader is not correct. it is the Handhelds Reverse Engineering Tool (which also servers as Linux bootloader in WinCE) Jul 12 10:58:36 is it possible to build own kernel? Jul 12 10:58:44 Jay7: yes Jul 12 10:58:56 Jay7: there is even one very well supported 'android device' in OE Jul 12 10:59:12 Jay7: htcdream (aka G1). ask mickey|bbl gnutoo and leviathan about it Jul 12 10:59:41 dcordes: so, htcdream is booting linux kernel directly and this kernel can be rebuilt from some sources? Jul 12 10:59:57 Jay7: yesh Jul 12 11:00:10 then we can try to port kexecboot there :) Jul 12 11:00:39 can kexecboot be useful on htcdream? Jul 12 11:01:35 Jay7: I think the bootloader on these devices is quite rich Jul 12 11:01:44 leviathan: you can even boot from mmc directly, right ? Jul 12 11:01:54 Jay7: configuration might be complicated though.. don't know Jul 12 11:02:03 Jay7: involves some nand flashing Jul 12 11:02:04 dcordes: yes Jul 12 11:02:08 Jay7: iirc Jul 12 11:02:11 ah, ok then Jul 12 11:03:26 so it might be of use Jul 12 11:03:58 Jay7: in the newer devices we plan to install u-boot as sources are available for qsd8250 soc Jul 12 11:06:40 well.. I'll look for next target then ;) Jul 12 11:07:45 it would be nice to get kexec boot on htcleo until there is uboot Jul 12 11:07:57 cause A) uboot has downside of removing the WinCE option Jul 12 11:08:00 and B) it might take a while Jul 12 11:08:18 HaRET is rather limited as opposed to the android bootloader Jul 12 11:11:00 Jay7: so what is minimum need to use kexecboot? does it work with small initrd ? Jul 12 11:13:52 dcordes the minimum for arm and qemu was all kernel + kexecboot roun 800k Jul 12 11:13:52 dcordes: minimum need is ability to boot linux kerenl with kexec enabled and working kexec tool :) Jul 12 11:14:27 true, kernel + initramfs with kexecboot and kexec inside is about 800k now Jul 12 11:14:38 thanks to ant__ and lzma :) Jul 12 11:14:54 Jay7: ok that's pretty doable I guess Jul 12 11:15:02 I disbaled all printf and debug stuff Jul 12 11:15:05 *g* Jul 12 11:15:05 right now working framebuffer is required too Jul 12 11:15:13 ah yes Jul 12 11:15:35 but I will implement text-mode UI Jul 12 11:15:55 we have patches against old version and this is sponsored by CELF Jul 12 11:16:12 I would like to try this on htcleo. I can load like 10M initrd and modify kernel in any way. the device has a ts (that doesn't work with tslib) and 6 buttons that work has keypad input Jul 12 11:16:28 dcordes: that would be great :) Jul 12 11:16:30 dcordes sounds perfect Jul 12 11:16:56 I'm reworking GUI code now so testing on other devices will be very useful for us Jul 12 11:17:16 atm I hope we are supporting 1,2,4,16,18,24,32 bpp :) Jul 12 11:17:17 Jay7: could I like mount mmc parition with vfat and mount and boot ext2 image files from it chosing them from some menu ? Jul 12 11:17:28 oh and device has wvga display Jul 12 11:17:47 jay7 will the gui be smaller? Jul 12 11:17:59 woglinde: it will be faster :) Jul 12 11:18:02 not sure about size Jul 12 11:18:02 hehe Jul 12 11:18:20 but I'll provide configure knobs for disabling unused bpp modes Jul 12 11:18:28 this can do it smaller Jul 12 11:18:36 sounds good Jul 12 11:18:38 03Xerxes Rånby  07org.openembedded.dev * rf683643681 10openembedded.git/recipes/llvm/llvm-common.bb: llvm-common: unbreak build NATIVE_INSTALL_WORKS = "1". Jul 12 11:18:43 pictures still xpm? Jul 12 11:19:03 dcordes: you can load kernel + initrd from vfat partition Jul 12 11:19:04 Jay7: cause that's the biggest use I see now. most people want android on the device but I'm also planning to add some OE and other GNU rootfs once I got the ts working. so people could chose rootfs and maybe kernel from kexec after starting it from haret Jul 12 11:20:09 dcordes: we can test it at least :) Jul 12 11:20:40 right now we are supporting kexec'ing kernel + initrd from partition Jul 12 11:21:06 so kernel will take this partition as rootfs by default but you can override it in boot.cfg file Jul 12 11:21:23 woglinde: yes :( Jul 12 11:21:26 since people can't format sd card, we need to use what they have. currently it works like so: they boot WinCE, load zImage in haret with cmdline to use initrd, initrd mounts vfat mmc, initrd loopmounts rootfs.ext2 from mounted vfat mmc, initrd 'newroot's to mounted rootfs.ext2 Jul 12 11:21:46 omg.. :) Jul 12 11:22:07 :) Jul 12 11:22:07 ieehks yes Jul 12 11:22:11 dcordes: can people split SD on two partitions (vfat + ext2/3)? Jul 12 11:22:22 Jay7: no, they can't. :D Jul 12 11:22:38 well.. there was some question about loop mounts Jul 12 11:22:43 Jay7: many of them don't even know that google OS uses a linux kernel Jul 12 11:23:06 all the above runs automatically once they start the HaRET executable Jul 12 11:23:09 reasonable.. Jul 12 11:23:50 I distributed some sets with ubuntu.ext2 Jul 12 11:24:31 well.. haret can boot kexecboot kernel + initramfs, then kexecboot can look for kernels/images, loopmount it (not supported yet) and kexec kernel from there Jul 12 11:24:31 so they know there's other user space than android running on linux Jul 12 11:24:49 sounds good Jul 12 11:25:06 so I can do what we have now, with a nice menu plus the ability to chose between kernels Jul 12 11:25:36 which is very useful cause android makes many dependencies on kernel level that are not compatible with normal GNU/Linux distro Jul 12 11:25:44 multiple kernels per partitions is not supported yet but this is sponsored work too Jul 12 11:26:09 at least I see ways to do it :) Jul 12 11:26:59 dcordes: welcome to #kexecboot then :) Jul 12 11:27:33 since there are quite some hacked linux htcs that I think will never see a direct bootloader, this has much potential Jul 12 11:53:29 03Martin Jansa  07org.openembedded.dev * r41c81a78c3 10openembedded.git/conf/distro/include/preferred-shr-versions.inc: Jul 12 11:53:29 shr: switch to a newer glib/gtk combo as Angsrom does Jul 12 11:53:29 Signed-off-by: Martin Jansa Jul 12 13:07:24 hi JaMa Jul 12 13:07:34 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * re2e1cf695f 10openembedded.git/recipes/tasks/task-shr-minimal.bb: (log message trimmed) Jul 12 13:07:34 task-shr-minimal: fix case when binary locales generation is disabled Jul 12 13:07:34 2 issues were present: Jul 12 13:07:34 *disabled was a string and so,when binary locales generation was disabled Jul 12 13:07:34 it never executed the return statement Jul 12 13:07:44 JaMa, ^^^ Jul 12 13:07:48 GNUtoo|laptop: thanks for fix Jul 12 13:08:22 np Jul 12 13:56:45 hi kergoth Jul 12 13:57:52 hey Jul 12 14:01:53 03Xerxes Rånby  07org.openembedded.dev * r605dee9cbb 10openembedded.git/recipes/llvm/ (llvm2.7/arm_ppc.patch llvm2.7_2.7.bb): llvm2.7: workaround llvm pr5201 / icedtea6 pr399 for ARM and PPC Jul 12 14:08:57 http://nokia.taleo.net/careersection/10120/maemo/moresearch.ftl Jul 12 14:09:12 Jobs on meego Jul 12 14:10:22 Education Level Master's Degree (±18 years) Jul 12 14:10:24 haha Jul 12 14:38:47 03Xerxes Rånby  07org.openembedded.dev * rd25de801a4 10openembedded.git/recipes/llvm/ (llvm2.7/BX_to_BLX.patch llvm2.7/MOVLRPC.patch llvm2.7_2.7.bb): llvm2.7: Make ARMv4 compatible llvm pr7608 Jul 12 15:15:12 hrw: sorry, but can you approve me to join OE group on linkedin? Jul 12 15:15:26 sure Jul 12 15:18:19 thanks Jul 12 15:18:47 Jay7: there are few more managers of that group btw ;D Jul 12 15:19:05 I've seen but you are active right now :) Jul 12 15:19:37 btw, OE on ohloh is still unmanaged Jul 12 15:20:00 heh.. when I was there last time... Jul 12 15:21:39 btw - what to manage there? Jul 12 15:21:58 hrw: click 'I'm manager' iirc Jul 12 15:22:32 ah... applied Jul 12 15:22:44 looks like I forgot to do that befoe Jul 12 15:24:07 I do not like ohloh way of working. I have many 'projects' listed just because they are 'linux kernel under other name' Jul 12 15:24:28 hehe.. Jul 12 15:25:32 I even have 'FSO Framework' listed but just two simple patches Jul 12 15:27:54 hrw: now you will get ubuntu and linaro stacks there ;) Jul 12 15:28:32 rather not Jul 12 15:28:40 ohloh do not track them Jul 12 15:35:09 hi Jul 12 15:35:24 is there someone with commit right to opkg svn? Jul 12 15:35:42 i have a little bugfix for it Jul 12 15:41:30 ~lart developers which forces me to rewrite a mac adress to long Jul 12 15:41:31 * ibot wallops developers which forces me to rewrite a mac adress to long with a main rotation server that needs rehubbing. It won't take long Jul 12 15:44:38 ibot: but it has to be long... Jul 12 15:44:39 So we want to be stupid, right, has to be long... ? Fine, pushups, right now. OK, side straddle hops, right now. OK, run in place, right now. I didn't say stop! Jul 12 15:45:35 haha Jul 12 15:51:00 heh.. now I can write my own psplash.. Jul 12 16:08:53 hi (i hope it won't disconnect me now) Jul 12 16:09:00 is there someone with opkg svn commit right? Jul 12 16:09:14 i have little patch which fixed list_upgradable with libopkg Jul 12 16:52:55 hm Jul 12 16:52:58 had a question for hrw|gone too Jul 12 17:42:18 good morning Jul 12 17:45:22 on thursday I cloned git://git.openembedded.org/openembedded. trying 2 build "gpe-image" I got couple errors Jul 12 17:49:28 1. when parsing bb files I got "Can't build bluez-cups-backend", "bluez-gstreamer-plugin", bluez-utils-alsa" and "bluez-libs-3.36" Jul 12 17:54:26 any idea what to do about this? Jul 12 17:56:03 anyone out there tried to build apr or apr-util recently? Jul 12 18:03:39 dos1: send the patch to opkg-devel@googlegroups.com Jul 12 18:04:31 khem: i've sent it with google groups interface, did it came? Jul 12 18:05:12 http://groups.google.com/group/opkg-devel/browse_thread/thread/fa80833b53eafe2c Jul 12 18:17:28 hm.. something is wrong on LTG Jul 12 18:17:42 high storage io? Jul 12 18:23:25 2. python-native_2.6.4.bb is erroring out Jul 12 18:23:42 here not Jul 12 18:24:32 so whats the error igork Jul 12 18:24:47 florian: check git.ltg.org machine please Jul 12 18:25:06 when this is not backup or something like :) Jul 12 18:27:19 Jay7: ? Jul 12 18:27:47 git/ssh is hanging Jul 12 18:28:13 looking like too high disk i/o Jul 12 18:28:33 re pb Jul 12 18:29:12 Does anybody know if there is a way to put the entire kernel source in a user.collection, so that I may compile a copy of the kernel with modified source? So far, I have only seen how to keep a separate copy of the the kernel config in the user.collection. Jul 12 18:29:21 hi woglinde Jul 12 18:31:30 hm I dont understand the question Jul 12 18:31:43 you can make a local git repo Jul 12 18:31:53 an let generate the kernel-source from there Jul 12 18:34:10 woglinde: That makes sense. I'm a little noob when it comes to that though. I downloaded the git repo linux-omap3-2.6 , but I'm not sure how I would go about compiling the kernel from this source. Jul 12 18:34:46 Jay7: grumps... will take a look Jul 12 18:35:06 florian: thank you Jul 12 18:36:09 bkinman as other kernel Jul 12 18:36:10 s Jul 12 18:36:49 looks like there are some users building thins with narcissus Jul 12 18:37:04 hehe.. Jul 12 18:37:12 btw, is #linuxtogo alive? Jul 12 18:38:23 its not really crowded but i'm there most of the time Jul 12 18:40:05 florian: cool Jul 12 18:40:10 jo khem Jul 12 18:40:14 florian: does it have precompiled stuff Jul 12 18:40:21 woglinde: viva espania Jul 12 18:40:35 khem netherlands are not good losers Jul 12 18:41:00 they played to hard Jul 12 18:41:25 yeah they did Jul 12 18:41:28 dutch people are evil Jul 12 18:41:40 aim no Jul 12 18:42:02 im dutch, and pretty evil Jul 12 18:42:05 woglinde: robben missed a chance again which my son would have converted :) Jul 12 18:42:17 methril: won't you get a salary boost now that Spain is the world soccer champion? :-) Jul 12 18:42:32 khem nope that was overtime Jul 12 18:42:41 but i agree tho, they really played aggresive Jul 12 18:42:42 and robben was running a lot Jul 12 18:42:46 ive never seen them play like that Jul 12 18:42:50 so he had no strength left Jul 12 18:43:17 woglinde: it was not in overtime it was regular time when he spilled 1:1 with Iker Jul 12 18:43:35 khem hm than it was maybee another scene Jul 12 18:43:50 but spain hat enough chances too Jul 12 18:44:05 woglinde: http://theoriginalwinger.com/2010-07-11-arjen-robbens-missed-chance-against-spain-iker-casillas-save Jul 12 18:44:11 and didnt made it Jul 12 18:44:13 this one Jul 12 18:44:32 you dont win world cup if you dont convert those Jul 12 18:46:33 hm ah this one Jul 12 18:47:22 woglinde: the first issue is that there is no buildable bluez* packets Jul 12 18:47:23 yes Jul 12 18:47:30 igork wrong Jul 12 18:47:53 igork and you didnt mentioned what version of oe you use Jul 12 18:48:05 woglinde: How do i know which version of the 2.6 kernel im compiling? I downloaded linux-omap3-2.6 repo, but i have know idea which minor version it is? Jul 12 18:48:44 woglinde: Im using bitbake 1.8, not sure about OE version Jul 12 18:49:07 how do I check that? Jul 12 18:50:57 git branch Jul 12 18:51:13 and the one with * is what you are on Jul 12 18:52:03 bkinman look in the main Makefile Jul 12 18:52:32 woglinde: thanks! Jul 12 18:54:20 greetings! does OE have any standards for packages adding/removing cron entries? Jul 12 18:54:41 vixie-cron appears to not have /etc/cron.d, so I can't just stick a file in there Jul 12 18:58:45 I think I posted it: git://git.openembedded.org/openembedded Jul 12 19:01:55 Is there a way to make bitbake spit out which commands it is running as it runs? It appears that --verbose doesn't do this. Jul 12 19:03:35 bkinman: bitbake --help Jul 12 19:03:41 there is -D Jul 12 19:03:53 and you can have more Ds to dump more Jul 12 19:19:49 woglinde, khem: is this what were u looking for? Jul 12 19:30:05 khem: so, It seems Ds doesnt actually spit out the commands that are being run. Particularly, what i was looking for was the make command that it was using on the kernel when i type bitbake linux-omap3-2.6.33. Jul 12 19:30:31 bkinman: those are in temp/log.do_* Jul 12 19:30:48 khem: excellent! thanks! Jul 12 19:31:13 bkinman: you might have to add V=1 to the linux make invocation to make it spit out complete commands Jul 12 19:40:09 ive created a bb file for autossh (needs ./configure, make, make install) (i put EXTRA_OECONF = "--prefix=${D}${prefix}" in the bb file); when im building the package it appears it attempts to install on the root FS (permission denied errors): | mkdir: cannot create directory `/usr/share/examples': Permission denied Jul 12 19:40:38 same issue occurs without having EXTRA_OECONF in the bb file Jul 12 19:44:38 also wondering how i can make bitbake copy recipes/autossh/files/autossh to oeroot/etc/init.d Jul 12 19:45:24 warflyr: add file://autossh to SRC_URI Jul 12 19:45:44 then you have to install it into ${sysconfdir}/ Jul 12 19:45:58 in say do_install_append () function in bb file Jul 12 19:46:07 you dont need to set --prefix Jul 12 19:46:19 its already taken care if you inherit autotools Jul 12 19:46:49 yeah thats what i figured; thats why im confused on why its trying to touch the system root Jul 12 19:59:38 just to be sure Jul 12 20:00:06 can I build multiple machines with the same TMPDIR? Jul 12 20:00:17 *for multiple Jul 12 20:00:22 yes Jul 12 20:03:26 woglinde: any idea? Jul 12 20:04:05 igork whats the error now? Jul 12 20:07:18 for issue 1 or 2? Jul 12 20:07:41 igork first issue again Jul 12 20:08:04 or maybee your local.conf file is wrong Jul 12 20:08:55 http://tinderbox.openembedded.net/public/logs/task/6494621.txt Jul 12 20:09:07 quilt native, MACHINE=qemux86 Jul 12 20:09:26 ah.. that my host problem Jul 12 20:10:11 yes Jul 12 20:10:23 GNU patch 2.6.1 Jul 12 20:10:23 this paste describes my problem: http://pastebin.com/9gjpnUpq Jul 12 20:10:24 hehe.. Jul 12 20:11:22 warflyr seems the autotools files on autossh are borken Jul 12 20:11:37 "Missing or un-buildable dependancy Jul 12 20:11:58 woglinde_, ahh, i'll try to create a patch; thanks Jul 12 20:12:07 "bluez-cups-backend" Jul 12 20:12:23 igork please pastebin your local.conf Jul 12 20:13:46 warflyr: it seems that its not cross compilation friendly Jul 12 20:16:34 ok, here is the error: http://paste.pocoo.org/show/236717 Jul 12 20:17:06 igork thats not tje local.conf Jul 12 20:17:59 binutils-2.16? Jul 12 20:18:02 what the hell Jul 12 20:18:15 and the local.conf - http://paste.pocoo.org/show/236719 Jul 12 20:18:23 ah 2.18 Jul 12 20:18:53 ieehkssss Jul 12 20:18:56 2144 again Jul 12 20:19:02 aeh 2440 Jul 12 20:19:23 khem do you suggest i do a automake and autoconf to generate new makefiles/configure; or should it require just a few edits to the current auto files? Jul 12 20:19:38 warflyr: I dont know Jul 12 20:19:50 IgorK are you using the this olde broken overlay? Jul 12 20:19:54 warflyr: you have to see if it assumes /usr/bin etc. Jul 12 20:20:24 dont get it why all 2440 users has such problems with oe Jul 12 20:21:09 igork comment TARGET_OS = "linux" Jul 12 20:21:18 no need to enable this manually Jul 12 20:21:50 no, I clone'd git://git.openembedded.org/openembedded Im not using mini2440 anymore Jul 12 20:22:35 igork why you need ASSUME_PROVIDED += "linux-libc-headers-native"? Jul 12 20:22:40 is 2.6.31 to old? Jul 12 20:24:26 besides that the local.conf file looks right Jul 12 20:24:29 khem Makefile.in uses $(TARGET) $(bindir) $(prefix) $(datadir) $(mandir); they are set with prefix=@prefix@, bindir=@bindir@, etc; do these names seem out of convention? Jul 12 20:24:53 ieehks Makefile.in Jul 12 20:24:57 no Makefile.am Jul 12 20:25:00 ? Jul 12 20:25:24 seems the developer have no clue to use autotools the right way Jul 12 20:25:34 lol; yeah no .am :) Jul 12 20:25:49 I suggest write it from sratch Jul 12 20:25:54 all other will be pain Jul 12 20:26:11 start with autoscan Jul 12 20:26:16 warflyr: you might have to rewrite do_configure Jul 12 20:26:24 or autoproject Jul 12 20:26:33 khem no thats pain Jul 12 20:26:34 because autotools one does autoreconf Jul 12 20:26:52 a complete an sane rewrite is better Jul 12 20:27:37 * khem nods Jul 12 20:28:06 woglinde: Im using "local.conf" from here: http://www.electronics.diycinema.co.uk/wp-content/uploads/2009/11/local.conf.txt Jul 12 20:28:18 when I think of autotools I think of gcc and friends and I wouldnt dare to rewrite that Jul 12 20:28:18 *sigh* Jul 12 20:28:24 igork as I said Jul 12 20:28:33 you used older stuff Jul 12 20:28:54 they uncomment TARGET = linux which is not needed Jul 12 20:28:59 the assumed libc-headers Jul 12 20:29:09 which will not come out of the box in oe Jul 12 20:29:21 please forgot about them Jul 12 20:29:35 they are missleading a lot of people Jul 12 20:29:46 hm yes Jul 12 20:30:05 now I remember and know who is failing all the 2440 people Jul 12 20:30:24 hi bluelightning Jul 12 20:30:37 hi woglinde_ Jul 12 20:31:38 hm, so I need to comment out TARGET = linux and remove libc-headers? Jul 12 20:31:49 igork yes Jul 12 20:32:14 ok, will try that, thank you Jul 12 20:37:29 Anybody here compiled linux-omap3-2.6 by hand? Jul 12 20:37:46 yes Jul 12 20:38:03 make menuconfig ARCH=arm Jul 12 20:41:24 woglinde: "bluez-utils-3.x has been replaced by bluez4". Do you know what this is? Jul 12 20:42:04 bkinman i did, but found i had a much more functioning kernel when i just used bitbake o.0 Jul 12 20:42:13 igork only warnings Jul 12 20:42:45 warflyr: Yeah =( The problem is that I need to modify the kernel source... So i'm not exactly sure about the proper way to go about that from within the OE context. Jul 12 20:43:20 bkinman hm I have anotherway Jul 12 20:43:31 tar the dir exclude .git Jul 12 20:43:38 make your kernel recipe Jul 12 20:43:42 md5sum it Jul 12 20:43:55 put right md5sum and sha256sum into the recpie Jul 12 20:43:59 build it Jul 12 20:44:27 woglinde_: That is actually a pretty good idea. Jul 12 20:44:51 fine Jul 12 20:44:58 look at the other recipes Jul 12 20:45:20 use file:// in your SRC_URI for the kernel.tar.bz2 Jul 12 20:46:02 "Angstrom does not support libiconv glibc builtin iconv will be used" Jul 12 20:46:18 it's another warning Jul 12 20:46:37 yes we know Jul 12 20:46:42 as a developer I don't like warnings :) Jul 12 20:46:58 igork haha lol Jul 12 20:47:08 ive learned to like warnings... so much better than errors Jul 12 20:47:13 please look howmuch warnings a glibc build spoils Jul 12 20:47:14 or gcc Jul 12 20:47:30 even the kernel... Jul 12 20:47:50 I usually set the maximum warnigs settings in Visual Studio or gcc... Jul 12 20:48:28 IgorK: important thing is do you take any action on it Jul 12 20:48:33 yes they are problem is its possible they indicate a real problem Jul 12 20:49:03 igork aeh Jul 12 20:49:10 we might have noticed it Jul 12 20:49:43 khem: depends on the warning itself Jul 12 20:56:33 igork is it working now? Jul 12 20:59:13 we will see it starts downloading and compiling... Jul 12 20:59:24 so no error Jul 12 20:59:25 fine Jul 12 20:59:46 Ill wait till it finish and report back for the issue # 2 Jul 12 21:04:00 jo ant Jul 12 21:04:58 hello Jul 12 21:07:37 I have a recipe for a 2.6.28 linux kernel... but there is no recipe for 2.6.28... how can I create a new recipe for 2.6.33? Jul 12 21:08:04 aszpain what machine/board you have? Jul 12 21:09:06 nope the build still fails for python-native Jul 12 21:09:50 what error? Jul 12 21:10:16 hold on let me check the messages Jul 12 21:12:08 aszpain i havent tested a kernel from this repository, but it has kernel 2.6.32 maybe 33: taskit-overlay Jul 12 21:12:18 IGEPv2 Jul 12 21:12:34 is there any 2.6.33 recipe for IGEPv2? Jul 12 21:13:23 khem ping Jul 12 21:13:50 woglinde: whats up Jul 12 21:14:18 khem could please look up why db3 is need when gcc-4.5 and uclibc-git is build? Jul 12 21:14:38 hmmm Jul 12 21:14:43 I looked at the dep-graph Jul 12 21:14:50 and saw none directly dep Jul 12 21:14:54 its just in Jul 12 21:14:58 out of nowhere Jul 12 21:15:08 woglinde: check task deps Jul 12 21:15:40 how does anyone check if there is some 2.6.33 kernel version for its recipe? Jul 12 21:15:54 what the hell Jul 12 21:15:55 "pkgconfig-native.do_package_stage_all" -> "db3.do_package_stage" Jul 12 21:16:22 wierd Jul 12 21:17:01 thats he successor error Jul 12 21:17:04 didnt find the root yet Jul 12 21:17:08 I feel I will die of cold someone turned on A/C right over my head Jul 12 21:18:40 hm I have a fear Jul 12 21:18:42 perl-native Jul 12 21:18:57 woglinde: try grg's patch Jul 12 21:19:17 http://patchwork.openembedded.org/patch/2336/ Jul 12 21:19:44 ???? Jul 12 21:19:49 thats da dep error Jul 12 21:19:52 not a sed error Jul 12 21:20:08 well.. about quilt native configure error Jul 12 21:20:08 seems the A/C is doing wired stuff in you Jul 12 21:20:13 heh Jul 12 21:20:28 sed does something important read the patch Jul 12 21:20:33 this was fixed by changes in configure.ac Jul 12 21:20:53 Jay7: what error do you get Jul 12 21:21:05 khem: http://tinderbox.openembedded.net/public/logs/task/6494621.txt Jul 12 21:21:23 [jay@is ~]$ patch --version Jul 12 21:21:25 GNU patch 2.6.1 Jul 12 21:21:48 qemux86 Jul 12 21:22:04 OMG Linus has done a special ranking about defconfig creators ;) Jul 12 21:22:32 (ARM) Jul 12 21:22:40 heh Jul 12 21:22:51 TI will be on top Jul 12 21:22:58 iow omap Jul 12 21:23:06 you bet ;) Jul 12 21:24:47 khem: not sure that is error in quilt configure or on my host.. Jul 12 21:25:02 git rebase --onto $1-1 $1 Jul 12 21:25:03 woglinde i have IGEPv2... do u know if there is some 2.6.33 recipe? Jul 12 21:25:09 guess what that does ? Jul 12 21:25:28 it punts out $1 which is a SHA from your local commits Jul 12 21:25:49 some experiments and I adore git a bit more now Jul 12 21:25:50 aszpain try the psp recipe in oe-dev Jul 12 21:25:59 aeh its -32 only Jul 12 21:26:19 Jay7: whats your build box Jul 12 21:26:41 khem: ALT Linux :) Russian own distro Jul 12 21:26:50 ugh Jul 12 21:27:05 Does it have a vodka emulator :) Jul 12 21:27:06 but what is output of your patch --version? Jul 12 21:27:14 khemy can you please do a bitbake -g pkgconfig-native and lookup if its need db3? Jul 12 21:27:19 with uclibc-git Jul 12 21:27:24 khem: developers are still testing it ;) Jul 12 21:27:42 Jay7: testing or tasting ? Jul 12 21:27:57 damn A/C Jul 12 21:28:02 kinda :) Jul 12 21:28:18 Jay7: then you can add recipe to OE Jul 12 21:28:28 patch --version Jul 12 21:28:28 patch 2.6 Jul 12 21:28:38 hm ah this looks strange a bit "perl.do_configure" -> "db3.do_populate_sysroot" Jul 12 21:28:54 aha.. Jul 12 21:28:59 no GNU prefix Jul 12 21:29:18 well.. I'm first encountered this error then.. Jul 12 21:29:40 Jay7: let me see the quilt configure Jul 12 21:29:58 line 5899 Jul 12 21:30:07 patch_version=$2 Jul 12 21:30:13 I've changed it for $3 Jul 12 21:30:47 but it will break anything :) Jul 12 21:31:17 we can strinp GNU from beginning Jul 12 21:31:22 *strip Jul 12 21:31:30 Jay7: you need to fix patch on ALT Jul 12 21:31:44 hm.. Jul 12 21:31:53 here on fedora and debian/ubuntu its always the second string not third Jul 12 21:31:57 hm ah Jul 12 21:32:03 perl depends on virtual/db Jul 12 21:32:15 why the hell db3 got selected Jul 12 21:32:17 khem: have you 2.6.1 on fedora or debian? Jul 12 21:32:28 Jay7: no hmm Jul 12 21:32:32 its 2.6 max Jul 12 21:33:05 Jay7: ok on F12 its 2.6.1 Jul 12 21:33:10 $ patch -v Jul 12 21:33:10 patch 2.6.1 Jul 12 21:33:14 hms Jul 12 21:33:18 right Jul 12 21:33:24 db3 is selected Jul 12 21:33:26 hms Jul 12 21:33:27 well, ok Jul 12 21:33:33 woglinde: perl is pulling it in Jul 12 21:33:40 and target gcc needs perl Jul 12 21:33:55 actually lets say it needs perl to build itself Jul 12 21:34:19 khem the question is why db3 is pulled in Jul 12 21:34:25 instead of db5 Jul 12 21:34:29 ugh Jul 12 21:34:35 if you have a concrete kernel version of uImage-2.6.X and a concrete kernel version for your modules modules-2.6.X... it doesnt matter from where you get a rootfs right? Jul 12 21:34:54 aszpain: in threory yes Jul 12 21:34:58 in practice may not be Jul 12 21:35:08 so rootfs are... somehow... versionless right? Jul 12 21:35:16 there could be some features that assumed in kernel Jul 12 21:35:35 aszpain: yes thats the intent but its not the case 100% of time Jul 12 21:35:41 yes, lately udev added more requirements Jul 12 21:35:52 e.g. Jul 12 21:36:04 so you always need a tool that makes uImage, modules.tar and rootfs at once yes? Jul 12 21:36:25 at least uImage and modules Jul 12 21:36:48 woglinde: which version of gcc is it ? Jul 12 21:36:50 ieehks Jul 12 21:36:54 okay Jul 12 21:37:05 * woglinde slaps himself Jul 12 21:37:48 woglinde: ? Jul 12 21:37:52 arre you ok Jul 12 21:38:14 I set it myself Jul 12 21:38:17 in local.conf Jul 12 21:38:20 *sigh* Jul 12 21:38:36 yeah local.conf bites ass sometimes Jul 12 21:38:43 khem: about woglinde ... ;) he cannot remember about http://cgit.openembedded.org/cgit.cgi/openembedded/tree/recipes/dietlibc/dietlibc-0.32/diethome.patch Jul 12 21:38:48 too hot in Berlin Jul 12 21:38:58 woglinde: let's aks khem Jul 12 21:39:35 ant I can remember Jul 12 21:39:42 DIETHOME=$(PWD) bin-$... khem: some magic code could rebuild the path Jul 12 21:39:44 but dont know why it worked Jul 12 21:39:58 he he Jul 12 21:40:03 ant I thought you fixed it already Jul 12 21:40:18 how diffeernet is dietlibc from klibc Jul 12 21:40:18 I was battling with darwin bootloader... ;) will do later Jul 12 21:40:27 is it hard to make a recipe from one concrete kernel version to another one? what do I use for base? Jul 12 21:40:47 khem: size was very comparable Jul 12 21:41:09 aszpain: if you are using already ported machine then its as easy as adding a defconfig Jul 12 21:44:30 khem: about dietlibc, static-built kexec did not work when I tried it. klibc has a 'magic' syscall patch...too much for me :D Jul 12 21:45:54 khem ah another question Jul 12 21:46:22 where do enable now lstat64 and stuff Jul 12 21:46:32 for uclibc Jul 12 21:47:05 add largefile to DISTRO_FEATURES Jul 12 21:47:24 woglinde: or use minimal-uclibc I already added it there Jul 12 22:00:24 khem where? the defconfig thats is used to make the images? Jul 12 22:09:00 aszpain: surf in recipes/linux Jul 12 22:09:12 what machine are u interested in Jul 12 22:13:18 khem: iirc he was asking about IGEP2 Jul 12 22:13:40 aka beagle-clone iirc Jul 12 22:14:50 khem I found it Jul 12 22:15:02 "libtimedate-perl-native" -> "perl-module-time-local Jul 12 22:15:20 the damn PACKAGES_DYNAMIC bug Jul 12 22:16:13 "quilt-native" -> "libtimedate-perl-native" [style=dashed] Jul 12 22:16:17 hms Jul 12 22:16:48 woglinde: I added that dependency Jul 12 22:16:57 to quilt-native Jul 12 22:17:21 sure and now its borked Jul 12 22:17:28 "libtimedate-perl-native" -> "perl-module-time-local Jul 12 22:17:45 woglinde: whats wrong with that Jul 12 22:18:00 or did it just unearthed a bug elsewhere Jul 12 22:18:05 perl-module-time-local is TARGET!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Jul 12 22:18:09 noz native Jul 12 22:19:39 libtimedate-perl-native is what I added Jul 12 22:19:52 RDEPENDS_${PN} += "perl-module-carp perl-module-exporter perl-module-strict perl-module-time-local" Jul 12 22:19:55 and its sane thing to ask for a native recipe to depend on a native package Jul 12 22:20:13 quilt is on of the first packages build Jul 12 22:20:28 hrmm Jul 12 22:20:34 right Jul 12 22:21:06 makes sense to revert this part Jul 12 22:21:10 hm why the native.bbclasse dont resets the RDEPENDS Jul 12 22:21:39 its inherited afterwards Jul 12 22:21:53 thats okay Jul 12 22:22:00 and itds not BBCLASSEXTEND Jul 12 22:22:13 hm Jul 12 22:22:16 I will fix it Jul 12 22:22:26 for the recipe Jul 12 22:22:27 you can add a check Jul 12 22:22:29 for sanity Jul 12 22:22:44 make sure that some native recipe is not accidently wanting something on target Jul 12 22:24:50 cost me half the evening Jul 12 22:24:53 hms Jul 12 22:24:59 one beer please Jul 12 22:25:02 german beer Jul 12 22:25:04 *g* Jul 12 22:27:03 * khem send woglinge an eBeer Jul 12 22:27:45 khem will you reverted it? Jul 12 22:28:06 because you cannt build perl without quilt normaly Jul 12 22:28:16 woglinde: yes Jul 12 22:28:26 I has just now tested a patch Jul 12 22:30:54 woglinde: try it now Jul 12 22:31:01 03Khem Raj  07org.openembedded.dev * r1f44f5cb6d 10openembedded.git/recipes/quilt/ (quilt-native.inc quilt-native_0.48.bb): Jul 12 22:31:01 quilt-native: Remove RDEPENDS on libtimedate-perl-native Jul 12 22:31:01 * quilt-native is build quite early. Its not sensible to depend Jul 12 22:31:01 on natve packages which inturn uses quilt to patch. Jul 12 22:31:01 * Thanks to woglinge for finding it. Jul 12 22:33:57 03Henning Heinold  07org.openembedded.dev * rf496207eb0 10openembedded.git/recipes/perl/libtimedate-perl_1.20.bb: Jul 12 22:33:57 libtimedate-perl: don't depend on target packages Jul 12 22:33:57 * bump PR Jul 12 22:39:50 thanks khem Jul 12 22:43:21 woglinde: atleast we fixed one more bug :) Jul 12 22:43:31 woglinde: Here is the error after changes - http://paste.pocoo.org/show/236783 Jul 12 22:44:07 sorry for the long delay, I tried to generate the file with the errors ;-) Jul 12 22:45:22 why the hell it still will use bluez3 Jul 12 22:45:31 ansgtroem set it to bluez4 Jul 12 22:51:21 hi. I'm seeing an odd issue with do_configure of gtk+-2.20.0 Jul 12 22:51:38 which one? Jul 12 22:51:46 the libtool that is produced from the configure script is broken and wants to invoke /usr/bin/nm Jul 12 22:52:02 I narrowed the issue down to this, so far... Jul 12 22:52:18 when I run configure with: ./configure --host=arm-oe-linux-gnueabi --disable-glibtest gio_can_sniff=yes Jul 12 22:52:27 zub: may be you need to set something like NM=.. Jul 12 22:52:27 I get broken libtool Jul 12 22:52:38 when I run it with: ./configure --disable-glibtest --host arm-angstrom-linux-gnueabi gio_can_sniff=yes Jul 12 22:52:43 I get working libtool Jul 12 22:52:47 anyone else seeing and issue building APR? Jul 12 22:52:59 khem: it would probably help, but why is it failing in the first place? :-( Jul 12 22:53:19 and why on gtk+ and not on other recipes Jul 12 22:54:14 zub: whats your distro Jul 12 22:54:45 DISTRO="shr" Jul 12 22:55:04 and if you mean the native :) then gentoo Jul 12 22:55:44 I had to do some binutils hack a`la http://patchwork.openembedded.org/patch/2178/ Jul 12 22:56:07 sh@t.. seems I can't continue building on ALT Linux further.. Jul 12 22:56:46 to get binutils-cross built Jul 12 22:56:57 or I should replace patch with my own version.. Jul 12 22:57:12 so maybe there is some fallout from that hack, but why/how would it affect this specific thing? Jul 12 22:57:34 I'm trying to abuse configure script to find out the exact cause Jul 12 22:59:41 hm.. that is interesting how quilt can be build on FreeBSD.. Jul 12 23:01:01 ah.. gnu patch from ports Jul 12 23:04:26 if I run .../temp/run.do_configure by hand, does it behave the same way as -c configure (provided all deps were already built) Jul 12 23:05:41 s/$/?/; Jul 12 23:24:21 khem: can quilt be in ASSUME_PROVIDED? Jul 12 23:26:42 jay7 I think so Jul 12 23:30:26 * Jay7 have asked about correcting patch --version output in distro ML.. Jul 12 23:30:58 well.. will look tomorrow.. Jul 12 23:33:28 * Jay7 -> sleep Jul 12 23:33:46 wwoglinde: I have no idea, All changes u suggested I made Jul 12 23:34:25 But more interestingly why its failing to compile python-native? Jul 12 23:47:11 anybody else? khem? Jul 13 00:01:55 ok, I figured out what's happening... uf Jul 13 00:02:42 anybody willing to build gtk+-2.20.0-r8.0 so that I can see what's happeing on some other box, pls? Jul 13 00:05:21 or, don't :-/ Jul 13 00:07:24 or do :) Jul 13 00:07:31 in configure output I see: Jul 13 00:07:39 checking for BSD- or MS-compatible name lister (nm)... arm-oe-linux-gnueabi-nm Jul 13 00:07:39 checking the name lister (arm-oe-linux-gnueabi-nm) interface... BSD nm Jul 13 00:08:04 and then after N lines of other stuff, NM is detected again, this time: Jul 13 00:08:26 checking for nm... /usr/bin/nm Jul 13 00:09:15 in the second test, NM is already set (or in fact it might have got there from the env.), and there's a test to see if it contains a slash. it does not and it's killed by /usr/bin/nm Jul 13 00:10:17 and that second NM makes it to arm-oe-linux-gnueabi-libtool: Jul 13 00:10:31 $ grep NM= gtk+-2.20.0/arm-oe-linux-gnueabi-libtool Jul 13 00:10:32 NM="/usr/bin/nm" Jul 13 00:11:34 this in turn fails when libtool is run, as my /usr/bin/nm doesn't like the arch/format/whatever. If I change it to arm-oe-linux-gnueabi-nm then compilation succeeds Jul 13 00:20:51 hm, NM does come from the env: export NM="arm-oe-linux-gnueabi-nm" Jul 13 00:21:40 and the second NM test in configure (why is it there? it's a bit different from the forst one - and it's broken w.r.t cross-compilation) kills it off, as it doesn't contain a / Jul 13 00:24:45 replacing the value of NM with full path (->the string now contains a slash) makes the second NM test keep the from-env NM value and in the end I get correct libtool Jul 13 00:25:00 I mean the NM in env with full path Jul 13 00:38:18 zub: paste the configure.in somewhere Jul 13 00:41:05 btw. the multitarget binutils hack has to do with this too - without multitarget my native nm is still used, but it doesn't complain Jul 13 00:41:10 ok Jul 13 00:42:10 zub: did you try 2.20.1 Jul 13 00:42:16 does it happen same ? Jul 13 00:42:26 will do, but I wanted to have what SHR does Jul 13 00:43:21 hmmm nevermind I think its wrong with 2.20.1 too Jul 13 00:43:29 I see this in my libtool file Jul 13 00:43:35 NM="/usr/bin/nm" Jul 13 00:43:43 but I dont use gentoo Jul 13 00:45:06 http://linux.fjfi.cvut.cz/~zub/OE/gtk+-2.20.0-r8.0/ Jul 13 00:45:12 here are the logs Jul 13 00:45:17 well log and .in Jul 13 00:45:45 well, I looked into the generated configure - and as I said - fo rosme reason NM is tested twice Jul 13 00:45:59 the second test is broken, and finds /usr/bin/nm Jul 13 00:46:08 I don't know where does the second test originate form though Jul 13 00:46:43 OE regenerates configure script Jul 13 00:46:51 and I think it might be messed up Jul 13 00:47:00 some packages want certain version of autotools Jul 13 00:47:06 brb Jul 13 00:50:00 02:45:45 < zub> well, I looked into the generated configure - and as I said - fo rosme reason NM is tested twice Jul 13 00:50:05 I mean in the regenerated configure Jul 13 00:50:27 but I think it was the same with the configure script that is in the tarball... but let me try to be sure Jul 13 00:57:08 it does: configure:17312: found /usr/bin/nm Jul 13 00:57:31 but only if I do export NM="arm-angstrom-linux-gnueabi-nm" as OE does Jul 13 00:58:53 (as I said: the seconf test for NM leaves the already-found NM alone if it does contain a slash, and without NM from env, the first test sets NM to full path of OE NM, so it has a slash in it) Jul 13 01:11:11 I mean... arm-angstrom-linux-gnueabi-nm when I was configuring by hand for my angstrom toolchain, OE sets it to arm-oe-linux-gnueabi-nm Jul 13 01:11:19 anyway, 3am here, I'm off to bed Jul 13 01:11:32 thanks for feedback Jul 13 01:17:43 khem, i'm not sure if it was your quilt-native change or woglinde's libtimedate-perl change, but my perl dependency has disappeared now :) **** BEGIN LOGGING AT Tue Jul 13 15:10:14 2010 Jul 13 15:40:44 all, services should be back up and running again! Jul 13 15:41:12 ~praise timtimred Jul 13 15:41:13 All hail timtimred! Jul 13 15:42:04 if there are any issues with crashed db tables etc just scream here Jul 13 15:42:04 03Tom Hayward  07org.openembedded.dev * rcb007c1e0d 10openembedded.git/recipes/python/python-2.6-manifest.inc: python 2.6: generated new python-2.6-manifest.inc for python-numbers Jul 13 15:42:04 03Tom Hayward  07org.openembedded.dev * rf7f8bfbd02 10openembedded.git/recipes/python/python-django_1.2.1.bb: python-django: added recipe for 1.2.1, based on 1.0.2 recipe. Jul 13 16:00:35 ~seen morphis Jul 13 16:00:37 morphis <~morphis@p5489B196.dip.t-dialin.net> was last seen on IRC in channel #openmoko-cdevel, 11d 21h 11m 30s ago, saying: 'heyho'. Jul 13 16:06:23 im trying to compile PHP, though its failing to configure due to missing zlib (checking for gzgets in -lz... no); the directory for zlib is specified and libz.so.1 (and others) do exist in this folder, the existing zlib is version 1.2.3 Jul 13 16:07:04 commit message generator: http://whatthecommit.com/ Jul 13 16:09:48 "REALLY F*KING FIXED" thats an understatement Jul 13 16:10:43 dont expect this much help for php here Jul 13 16:11:06 yeah i kinda figured o.0 Jul 13 16:11:50 could i build it with my local/native libs? Jul 13 16:12:18 dont know Jul 13 16:12:28 if host-arch == target-arch maybee Jul 13 16:12:48 yeah not the case o.0 Jul 13 16:13:35 im just going to disable that support and see if i have any issues in my applicatiob Jul 13 16:23:18 is the oe repo back online? Jul 13 16:23:20 hi btw Jul 13 16:24:15 yes, it is Jul 13 16:26:56 jo gnutoo and mickeyl Jul 13 16:27:06 hi woglinde Jul 13 16:35:44 thanks Jul 13 16:35:46 hi woglinde Jul 13 16:36:21 woglinde, are you interested in a not-written-yet bug app that permit to use navit with java system? Jul 13 16:37:15 gnutoo we already wrote an java app which is using navit Jul 13 16:37:33 its gpl but for the customer only Jul 13 16:37:43 ok Jul 13 16:38:04 I meant something that output a file that navit can use Jul 13 16:38:11 like java->navit Jul 13 16:38:26 my app won't be using navit Jul 13 16:38:29 o.O? Jul 13 16:38:31 but navit will use the app output Jul 13 16:38:40 because for now I've to kill java-cdc Jul 13 16:38:46 to get serial port ownership Jul 13 16:39:02 btw I've an autotool issue: http://pastebin.com/dQZ0ga6b Jul 13 16:40:41 HAVE_INTROSPECTION does not appear in AM_CONDITIONAL Jul 13 16:40:44 sure Jul 13 16:40:49 configure.ac is wrong Jul 13 16:42:08 GNUtoo|laptop: ptitjes is back, perhaps he committed something to libgee git. Jul 13 16:42:15 GNUtoo|laptop: try reverting to the rev before Jul 13 16:42:24 ok Jul 13 16:42:38 but I think fixing the issue is better Jul 13 16:42:50 woglinde, seem to have an idea Jul 13 16:43:12 HAVE_INTROSPECTION needs to be define in configure.ac Jul 13 16:43:55 it is Jul 13 16:43:58 in autogen.sh Jul 13 16:44:00 I'll look Jul 13 16:44:08 REQUIRED_M4MACROS=introspection.m4 gnome-autogen.sh "$@" || exit 1 Jul 13 16:44:21 s/it is/maybe it is Jul 13 16:44:48 looks like Jul 13 16:44:58 but what's gnome-autogen.sh Jul 13 16:44:59 ? Jul 13 16:45:12 seems like someone which dont have a clue about autotools Jul 13 16:45:21 lol ok Jul 13 16:45:40 maybe as most free software project devs? Jul 13 16:45:49 introspection.m4 has to be in m4 dir Jul 13 16:45:57 ok Jul 13 16:46:03 or other dir and configure.ac should point to this dir Jul 13 16:46:05 I'll try to fix it Jul 13 16:46:25 or it should be find in system-m4 dir Jul 13 16:46:53 * GNUtoo|laptop fetches his autotool manual Jul 13 16:52:34 ok hmmm Jul 13 16:52:41 I don't find this introspection.m4 Jul 13 16:52:44 I'll look Jul 13 16:53:46 maybe a gnome .m4? Jul 13 16:53:58 maybe Jul 13 16:54:03 where are thoses located? Jul 13 16:54:14 I'm grepping sysroot right now Jul 13 16:55:03 hm seems to come from glib Jul 13 16:55:07 ok Jul 13 16:55:15 grep gave nothing Jul 13 16:56:48 http://trac.openstreetmap.ch/trac/memphis/browser/trunk/m4/introspection.m4?rev=206 Jul 13 16:57:33 yes I saw that Jul 13 16:57:38 but it's not in glib's workdir Jul 13 16:57:45 so should I use that? Jul 13 16:57:49 the openstreetmap one? Jul 13 16:58:03 it would be a bit counter-nature Jul 13 16:58:24 use a random introspection.m4(not the official one) from openstreetmap Jul 13 16:58:32 I'll try to findout its provenance Jul 13 16:58:32 no the only copied it from somewhere too Jul 13 16:59:50 some gnome things seem to have it Jul 13 16:59:56 gtk+-native Jul 13 17:00:25 /usr/share/aclocal/introspection.m4gobject-introspection Jul 13 17:00:39 in debian its package gobject-introspection Jul 13 17:01:06 dont know if we ship it Jul 13 17:01:35 http://live.gnome.org/GObjectIntrospection Jul 13 17:02:35 thanks a lot Jul 13 17:08:45 hmmm in the recipe dir in glib dir there are a lot of: Jul 13 17:08:46 no-introspect.patch Jul 13 17:08:57 so should I add this m4? Jul 13 17:09:04 dont know Jul 13 17:09:05 should I add an introspect recipe and depend on it? Jul 13 17:09:19 I'll ask #introspect in the gnome server Jul 13 17:11:02 ask on the ml Jul 13 17:11:05 our Jul 13 17:11:19 that will be long Jul 13 17:11:24 I've not internet all the time Jul 13 17:11:30 but I'll do that Jul 13 17:11:38 and forget again about my build Jul 13 17:12:03 I only have internet in a wifi cafee Jul 13 17:15:38 if i creat a recipe for a php 5.3.x version and put it in my own build/recipes folder; how can i tell bitbake to use this rather than the 5.2.6 thats in the 2009/stable branch? Jul 13 17:15:56 warflyr read about overlay Jul 13 17:16:12 or just put in the stable dir Jul 13 17:16:46 in the stable directory, is the greatest version the package selected? Jul 13 17:17:14 yes Jul 13 17:17:38 if no other recipe has a higher DEFAULT_PREFRENCES variable has Jul 13 17:17:58 ah Jul 13 17:20:02 thanks a lot woglinde Jul 13 17:24:17 03Roger Monk  07org.openembedded.dev * rb520cd2983 10openembedded.git/recipes/u-boot/ (2 files in 2 dirs): Jul 13 17:24:17 u-boot_git: Change default console for OMAP3EVM Jul 13 17:24:17 * Change ttyS2 -> ttyS0 Jul 13 17:24:17 * bump PR Jul 13 17:24:17 Signed-off-by: Roger Monk Jul 13 17:25:47 hmmm, thatll break me lol Jul 13 17:26:02 well, i use 2009/stable so not really Jul 13 17:30:31 yay just finished building my first custom image with a few custom packages Jul 13 17:41:56 hm.. how to cancel git add? Jul 13 17:48:35 git reset HEAD Jul 13 17:48:40 with maybe a --soft or --hard Jul 13 17:48:45 git reset help should have an example tho Jul 13 17:48:58 git add -i allows this Jul 13 17:49:17 but now other question - how to cancel git mv.. Jul 13 17:49:41 well.. git reset is easier Jul 13 18:02:51 good morning Jul 13 18:03:54 woglinde: after continuing the build here is where I stuck: http://paste.pocoo.org/show/237117 Jul 13 18:04:09 hi all Jul 13 18:04:14 any idea? do you want me to post "local.conf"? Jul 13 18:05:42 and I can't continue any further :-( Jul 13 18:20:16 woglinde, anybody? Jul 13 18:22:46 IgorK: what are you trying to build? Jul 13 18:23:08 IgorK: It seems as if you are trying to buildsomething that needs bluez 3 Jul 13 18:23:15 IgorK: Do you have angstrom as distro? Jul 13 18:26:02 yes, hang on, let me re-post my local.conf.... Jul 13 18:27:20 here it is - http://paste.pocoo.org/show/237132 Jul 13 18:28:09 stefan: Im trying to build gpe for the mini2440 Jul 13 18:30:23 IgorK: The local.conf looks ok. The problem I suspect is that gpe somehow depends on bluez3 while angstrom as distro only supports bluez4 Jul 13 18:30:44 IgorK: You wre trying just gpe-image? Jul 13 18:31:09 yes "bitbake gpe-image" Jul 13 18:32:07 IgorK: hmm, I did not build it for ages I fear. No idea what drags it in. Jul 13 18:32:20 I want to build gtk+/wxwidgets to make my program. Jul 13 18:32:57 maybe u can suggest some other image that will put gtk+/wxwidgets in it? Jul 13 18:33:19 hmm Jul 13 18:33:43 IgorK: you could try gnome-image or take an x11-image and add your gtk+ and wxwidgets Jul 13 18:34:43 there is gnome-image? didn't know about it...:) learn something new everyday Jul 13 18:35:25 did u try to build gnome-image recently? Jul 13 18:37:13 IgorK: no :) Jul 13 18:37:30 IgorK: But I think the angstrom devs do it form time to time Jul 13 18:37:41 ok, I guess I will try it now, thank u Jul 13 18:37:47 IgorK: np Jul 13 18:38:34 and it looks like gpe is dead :) its more an issue between different gpe companies.... Jul 13 18:39:03 IgorK: may be, don't know. Jul 13 18:39:26 and licensing Jul 13 18:40:17 ok, Im off to try building gnome-image, hoping that it will run on mini2440 :) Jul 13 18:48:46 I have compiled a .ko module and now I want that module to be loaded at boot time... how? I have angstrong distro... Jul 13 18:50:24 stefan_schmidt: I received: "Nothing PROVIDES 'gnome-image'" Jul 13 18:51:05 aszpain: Why does it not autoload? Jul 13 18:51:20 aszpain: om-gta02.conf has an example for it: Jul 13 18:51:22 module_autoload_snd-soc-neo1973-gta02-wm8753 = "snd-soc-neo1973-gta02-wm8753" Jul 13 18:51:27 IgorK|building: hmm, let me check Jul 13 18:52:03 IgorK|building: my bad. Its angstrom-gnome-image Jul 13 18:57:38 re Jul 13 18:57:42 hi stefan Jul 13 18:57:55 hi woglinde Jul 13 18:59:14 it does not autoload because its a compiled one Jul 13 18:59:49 I mean if you copy just the .ko file its impossible to the kernel to know that he must autoload it right? Jul 13 19:01:46 aszpain: With the right MODULE_ALIAS and build in the same tree it should still be possible. Jul 13 19:02:06 Anyway, OE has something for this as it is used in the machine file I pointed out Jul 13 19:02:17 listen the name is hid-sunplus.ko Jul 13 19:02:24 whats the alias? Jul 13 19:04:47 if I had compiled that module in the kernel.. getting the uImage file ... do u think I got those problems? does it autoload when is compiled in the kernel? Jul 13 19:05:17 stefan_schmidt: it looks it will start building I will come back if there will be an issue... Jul 13 19:45:11 if I had compiled that module in the kernel.. getting the uImage file ... do u think I got those problems? does it autoload when is compiled in the kernel? Jul 13 19:55:03 stuff compiled into the kernel will be used automatically Jul 13 19:56:14 anyone want to help me with angstrom? on boot X (gpe-dm) doesn't start. trying to start the init script manually doesn't work, trying to start X manually does Jul 13 20:15:35 if I had compiled that module in the kernel.. getting the uImage file ... do u think I got those problems? does it autoload when is compiled in the kernel? Jul 13 20:20:09 Kevin`: you probably want #angstrom for that Jul 13 20:27:52 pb__: hasn't been anybody there for days, so I thought i'd try :) Jul 13 20:28:58 aszpain: as I just said, modules compiled into the kernel will be used automatically Jul 13 20:29:57 Kevin`: ah, heh. in that case, check syslog or run "gpe-dm 0-n" by hand Jul 13 20:30:01 er, without that 0 Jul 13 20:30:05 * pb__ stabs his keyboard Jul 13 20:33:11 checking Jul 13 20:37:27 what posesses it to use the orinoco driver by default.. Jul 13 20:41:36 for what? orinoco is wifi, gpe-dm shouldn't care about it at all Jul 13 20:41:56 or did you mean angstrom in general? Jul 13 20:42:09 pb__: http://nelson.kwzs.be/~kevin/gpe-log1.txt Jul 13 20:42:12 in general, I mean Jul 13 20:42:34 there's some funnyness there because of screen being funny, but useful information Jul 13 20:45:24 /etc/X11/xorg.conf is empty Jul 13 20:47:28 kevin what recipe did u used? Jul 13 20:47:38 bitbake x11-image? Jul 13 20:49:01 pb__ as i didn't see other objections or suggestions on COMPATIBLE_TARGET_SYS I was planning to go ahead and push it (unless you come up with objections) Jul 13 20:49:32 aszpain: x11-gpe-image Jul 13 20:59:27 aszpain ? Jul 13 21:00:30 fwiw, gpe-dm -n does much the same thing. same errors Jul 13 21:04:50 03Dmitry Eremin-Solenikov  07org.openembedded.dev * r1a35b60635 10openembedded.git/: Merge branch 'org.openembedded.dev' of git://git.openembedded.org/openembedded into org.openembedded.dev Jul 13 21:04:52 03Dmitry Eremin-Solenikov  07org.openembedded.dev * re1a3091c24 10openembedded.git/recipes/u-boot/u-boot_2010.06.bb: Jul 13 21:04:52 u-boot: add a recipe for fresh release of u-boot. Also make it default for tqm8540 board Jul 13 21:04:52 Signed-off-by: Dmitry Eremin-Solenikov Jul 13 21:04:52 03Dmitry Eremin-Solenikov  07org.openembedded.dev * r3be79bbe6a 10openembedded.git/recipes/linux/ (linux-2.6.34/tqm8540/defconfig linux_2.6.34.bb): Jul 13 21:04:53 linux-2.6.34: add tqm8540 configuration Jul 13 21:05:00 03Dmitry Eremin-Solenikov  07org.openembedded.dev * r7aec284bbe 10openembedded.git/recipes/u-boot/ (u-boot_2010.03.bb u-boot_2010.06.bb): Jul 13 21:05:00 u-boot: switch lite5200 to current u-boot and drop u-boot 2010.03 which had no other users Jul 13 21:05:00 Signed-off-by: Dmitry Eremin-Solenikov Jul 13 21:31:05 pb__: hey Jul 13 22:01:14 bitbake x11-image gets angstrong with some GUI interview right? why then its not showed up when booting? hmmm Jul 13 22:08:10 was that to me? Jul 13 22:08:20 or are you having a similar problem to me Jul 13 22:14:31 no Jul 13 22:14:36 well yes and no Jul 13 22:14:42 u are tryng to boot from NFS right? Jul 13 22:15:13 no, sd Jul 13 22:15:29 My problem is that I want to boot normally and see the x window environment in my HDMI capable TV but I only see the output of the console.. Jul 13 22:15:40 booting from nfs would kind of be interestig since my only networking is usb client or wifi Jul 13 22:15:44 whats ur problem? Jul 13 22:16:11 my problem is that when angstrom boots up the display manager doesn't start properly, it just goes to a console Jul 13 22:16:54 X itself works if started manually, and of course framebuffer works, bieng necessary for the console and splash stuff Jul 13 22:17:03 a console with RAMFS by default? Jul 13 22:17:22 what happens if u make ls -la Jul 13 22:17:29 ramfs? no, i'm not using an initrd Jul 13 22:17:39 if you are, it sounds like it isn't working properly, or isn't finding root Jul 13 22:17:40 well it happens to me Jul 13 22:17:49 does it say anything on the console? Jul 13 22:17:55 hmm no Jul 13 22:18:08 it only show me the login prompt Jul 13 22:18:13 via console Jul 13 22:18:24 but after feeezing like for 15 secs Jul 13 22:18:27 ls -la shows the contents of /home/root, bieng .ash_history and .ssh/ Jul 13 22:18:50 what is it doing before freezing Jul 13 22:19:03 I think its checking somehting that is not found and then shows me the console Jul 13 22:19:07 before? normal booting Jul 13 22:19:13 dmesg Jul 13 22:19:28 do you have serial? Jul 13 22:19:33 ye Jul 13 22:19:35 yes Jul 13 22:19:46 Same outpuyt Jul 13 22:19:49 as HDMI Jul 13 22:19:56 have the entire boot show there instead of to vga Jul 13 22:19:59 log it, and pastebin Jul 13 22:20:09 what device are you using anywy Jul 13 22:20:15 igepv2 Jul 13 22:20:31 oh, I wanted to get one of those, but didn't :) Jul 13 22:20:40 what did u get finally? Jul 13 22:20:44 nothing Jul 13 22:20:45 :) Jul 13 22:20:49 lol Jul 13 22:20:50 thats bad Jul 13 22:20:52 :) Jul 13 22:20:57 i'm waiting on a pandora though Jul 13 22:21:16 that should be fun, although it will be obsolete before I get it Jul 13 22:21:29 still an upgrade from my current system Jul 13 22:21:44 I bought that swindle of board called BeagleBoard REV C.... Jul 13 22:21:56 but then I switch to IGEPv2 Jul 13 22:22:40 gtg bye Jul 14 00:24:11 03Martin Jansa  07org.openembedded.dev * rc4bebcea12 10openembedded.git/recipes/shr/ (2 files in 2 dirs): Jul 14 00:24:11 libphone-ui: bump SRCREV, remove applied patch Jul 14 00:24:11 Signed-off-by: Martin Jansa Jul 14 00:24:16 03Sebastian Krzyszkowiak  07org.openembedded.dev * r1542ab2397 10openembedded.git/recipes/tasks/task-shr-feed.bb: Jul 14 00:24:16 task-shr-feed: add missing shr-splash themes to feed Jul 14 00:24:16 Signed-off-by: Martin Jansa Jul 14 00:24:16 03Martin Jansa  07org.openembedded.dev * rc7c2d15ab6 10openembedded.git/recipes/linux/ (3 files in 3 dirs): Jul 14 00:24:17 linux-openmoko-2.6.34: bump SRCREV, try to disable UBI debug and reenable background thread (seems to work now) Jul 14 00:24:23 03Martin Jansa  07org.openembedded.dev * r09d68a75d0 10openembedded.git/recipes/freesmartphone/ (4 files): Jul 14 00:25:16 freesmartphone: bump cornucopia SRCREV, remove aplied patch from fsogsmd and use FSO_CORNUCOPIA_SRCREV again in mdbus2 Jul 14 00:25:17 Signed-off-by: Martin Jansa Jul 14 01:35:15 khem: I _think_, quilt symlink + patch.bz2 is broken Jul 14 01:51:06 khem: Yeah, seems so Jul 14 02:34:48 Tartarus: hmmm which recipe should I try Jul 14 02:42:27 Dunno what's upstream, but you can make on fit the problem easily enough :) Jul 14 02:43:11 actually, hmm, wait Jul 14 02:43:14 maybe it's local Jul 14 02:43:17 * Tartarus tries more Jul 14 02:56:58 * Tartarus definitely has something going funny, maybe it's collections / amend.inc related **** ENDING LOGGING AT Wed Jul 14 02:59:56 2010