**** BEGIN LOGGING AT Sat Feb 24 02:59:58 2007 Feb 24 07:44:58 gm Feb 24 08:55:54 03tmbinc 07org.oe.dreambox * r2f127456... 10/ (1 packages/tuxbox/tuxbox-libtuxtxt.bb): tuxbox-libtuxtxt: also use 20061009 version for dm7020 Feb 24 09:30:41 03tmbinc 07org.oe.dreambox * rc14ef12c... 10/ (1 packages/enigma/enigma-modem.bb packages/ppp/ppp_2.4.3.bb): enigma-modem,ppp: use our pap-secrets, options on dm600pvr,dm7020 Feb 24 09:30:47 03tmbinc 07org.oe.dreambox * r96773728... 10/ (1 packages/images/dreambox-image.bb): dreambox-image: include modem support for dm600pvr, don't include lcd specific plugins on dm600pvr Feb 24 09:32:08 ~ping tmbinc Feb 24 09:32:10 pong tmbinc Feb 24 09:50:34 i've been trying to build generic-uclibc distro for x86 machine, yet it fails with http://pastebin.ca/370198. Is uclibc broken somehow, or is there something wrong with my environment? Feb 24 09:57:25 redguy : try bitbake bootstrap-image Feb 24 09:58:40 Ifaistos: trying.. Feb 24 10:12:33 Ifaistos: same thing... Feb 24 10:13:38 this seems to be the root of the problem: make: i486-linux-uclibc-gcc: Command not found Feb 24 10:15:04 redguy: Sounds like the crosscompiler either isn't built, or isn't in the searchpath Feb 24 10:15:43 redguy : did you do a clean build or continued from where it had stopped ? Feb 24 10:15:50 NAbyss : hi Feb 24 10:16:56 Ifaistos: Ello Feb 24 10:17:00 redguy : post you local.conf. it's strange i am building it as we speak with no problems. when was your OE checkout ? Feb 24 10:24:32 Ifaistos: the build was continued from where it stopped, I updated OE ~ 2 hours ago Feb 24 10:25:59 redguy : try a clean build using bootstrap-image. Feb 24 10:26:16 just in case: my local.conf http://pastebin.ca/370209 Feb 24 10:29:17 redguy : seems ok Feb 24 10:31:42 Ifaistos: should I build with bitbake -c clean bootstrap-image ? Feb 24 10:32:18 redguy : just remove the tmp. its safer that way ;) Feb 24 10:32:57 Ifaistos: did both :-) Feb 24 11:11:51 Ifaistos: still no luck: http://pastebin.ca/370235 Feb 24 11:15:40 hi all Feb 24 11:50:24 redguy : add this to you local.conf and rebuild -> PREFERRED_VERSION_linux-libc-headers = "2.6.11.1" Feb 24 12:00:57 Ifaistos: will that prevent me from building later kernels ? Feb 24 12:01:15 redguy : no. its just for uclibc Feb 24 12:01:45 redguy : uclibc 0.9.28 is old and does not recognize the new kernel-headers Feb 24 12:02:14 Ifaistos: thanks, building once more ;-) Feb 24 12:03:00 redguy : np. You'll get use to it. till you manage to get a stable "base" system :) Feb 24 12:30:26 redguy : let me know if it builds with the last change Feb 24 12:31:36 Ifaistos: seems that uclibc-initial has built Feb 24 12:32:11 erm, did build Feb 24 12:32:55 redguy : good let me know if it finishes the bootstrap-image, so i can update the generic-uclibc.conf Feb 24 12:34:05 Ifaistos, http://daemons.gr/photos/stable-server/ Rock solid openBSD server... (40cm of concrete in the lab...) Feb 24 12:35:05 :) Feb 24 12:35:49 lol Feb 24 12:36:23 All that was on Simulation Lab of Architecture @ NTUA :P Feb 24 12:38:30 experiment with concrete that wend wrong ? :)) Feb 24 12:42:06 Ifaistos, rebuild of the building near the lab.. wend wrong... Feb 24 12:42:25 there was a hole in the wall :P Feb 24 12:53:24 Ifaistos: still no luck... Feb 24 12:55:12 http://pastebin.ca/370312 Feb 24 12:55:14 :( Feb 24 12:57:54 redguy : Add the following to local.conf and do a rebuild -> SRCDATE_uclibc ?= "20061128" Feb 24 12:57:54 PREFERRED_VERSION_uclibc ?= "0.9.28+svn${SRCDATE_uclibc}" Feb 24 12:59:13 Ifaistos: this will bum up the uclibc version, right? Feb 24 12:59:51 redguy : it will use the latest uclibc from cvs Feb 24 13:00:10 Ifaistos: do I still need the linux-libc-headers = 2.6.11.1 then? Feb 24 13:00:17 yes Feb 24 13:00:20 ok Feb 24 13:00:25 don't worry about them Feb 24 13:01:10 ok, another build then ;-) Feb 24 13:15:50 Crofton are you here ? Feb 24 13:16:47 or any1 else here who's familiar with omap ? Feb 24 13:36:30 re Feb 24 13:36:44 hey koen, already in .be? :) Feb 24 13:37:20 Jin^eLD: yes, I'm squatting next to the OE booth Feb 24 13:37:28 cool! Feb 24 13:37:44 Koen what is this evening oe pub? Feb 24 13:38:41 greentux|fosdem: I have no idea Feb 24 13:45:00 Ifaistos: well, uclibc linked ok this time but I'll keep my fingers crossed ;-) Feb 24 13:46:07 The famous OE booth: http://www.flickr.com/photos/koenkooi/400715501/ Feb 24 13:47:31 filled with power supplies :) Feb 24 13:53:54 *looking koen photoes* Feb 24 13:57:15 http://www.flickr.com/photos/koenkooi/391406865/ does gpe-mini-browser will be in neo by default? Feb 24 14:00:42 i mean not dillo for example Feb 24 14:54:43 Ifaistos: yes? Feb 24 14:55:05 tmbinc : hi ! Feb 24 14:55:46 tmbinc : i am working on ppc also, and yesterday talked to a guy that is also doing ppc work on the linkstation Feb 24 14:56:30 tmbinc : we are thinking of "combining forces" to provide better support for ppc/powerpc in OE Feb 24 14:57:02 oh, that's great Feb 24 14:57:35 to be honest, i haven't followed .dev for some time (shame on me..), what are the current main issues? Feb 24 14:58:24 well there is a new insane.bbclass that does Q&A checks and has come with a load of things that need to be taken care of Feb 24 14:58:47 Add soft-float support for ppc targets without fpu Feb 24 14:59:09 possibly work on eglibc which is optimized for ppc Feb 24 14:59:21 support more targets... Feb 24 14:59:27 just to name a few :) Feb 24 15:07:23 OK i'll check the current .dev in the next time, to see what the current state is. need to go now :/ Feb 24 15:19:56 03Shane 07org.oe.dev * r489837fe... 10/ (3 files in 2 dirs): Feb 24 15:19:56 packages/libxml/libxml2-native_2.6.26.bb Feb 24 15:19:56 packages/libxml/libxml2_2.6.26.bb : Enabling C14N support in libxml allows php 5.2 to compile. Feb 24 15:19:56 Enabling this option only slightly (<2Kbytes) increases the size of the libxml package. Feb 24 15:37:43 03Shane 07org.oe.dev * r9142ed23... 10/ (1 packages/php/php-native_4.4.4.bb packages/php/php_4.4.4.bb): packages/php: Add php 4.4.4, php-native 4.4.4 Feb 24 15:37:46 03Shane 07org.oe.dev * rec94ad79... 10/ (1 packages/php/php-native_5.2.0.bb packages/php/php_5.2.0.bb): packages/php:Add php 5.2.0 and php-native 5.2.0 Feb 24 16:09:52 Quiet in here? :) Feb 24 16:10:03 everyone at FOSDEM I guess Feb 24 16:10:11 :) Feb 24 16:10:25 Jin^eLD: yes, me too ;- Feb 24 16:10:40 heh Feb 24 16:10:47 I hope I can come next year :> Feb 24 16:14:17 RP: ak4535 driver works after few fixes; thanks a lot for writing it:) Feb 24 16:14:40 mmp: np. Are there any fixes that should be merged? Feb 24 16:15:09 yes, some should be (I'm afraid that even one fix in soc-dapm.c) Feb 24 16:15:30 but I'll try first to clean the sources from my debugging printk's :) Feb 24 16:15:54 mmp: Right, Once you get it cleaned up, do send me a patch! Feb 24 16:15:56 (and that will take a while, because this is first time I was doing something with ASoC:) Feb 24 16:17:48 * RP back later maybe ;-) Feb 24 16:19:44 RP: bye:) Feb 24 16:55:12 anyone tried to build openmoko-libs today? Feb 24 16:55:16 or yesterday? Feb 24 16:56:55 03ifaistos 07org.oe.dev * ree82640a... 10/ (4 files in 4 dirs): packages/uclibc/uclibc-0.9.28: Add uClibc config for i586 targets Feb 24 16:57:01 03ifaistos 07org.oe.dev * r3f9f00e1... 10/ (4 files in 4 dirs): packages/uclibc/uclibc-cvs: Add uClibc_svn config for i586 targets Feb 24 17:29:07 hi Feb 24 17:44:19 <_Lobo_> How can I reduce this `cat /proc/meminfo | grep '^Cached:'` value? It 33280 kB for me, so half of my RAM... Feb 24 18:59:03 _Lobo_: just load big programs, like OpenOffice, Firefox, some GNOME apps, whatever. Cache is automatically given to any application that needs them. So, it's actually a GOOD sign if your cache is high Feb 24 18:59:51 But when you load this fat programs, your cache will be reduced, just as you asked ;-) Feb 24 19:00:03 <_Lobo_> schurig: yeah, it could be, problem is, that it freeze with MemFree on zero and Cached is not changing :/ Feb 24 19:00:15 <_Lobo_> I collect some more info and it' Feb 24 19:00:22 <_Lobo_> s probably bug in kernel :/ Feb 24 19:01:03 _Lobo_: no, it's not. Memory handling of kernels is bug-free since years. Feb 24 19:01:25 <_Lobo_> schurig: it can be problem of WBSD :/ Feb 24 19:01:49 <_Lobo_> Winbond SD driver :/ Feb 24 19:02:03 The question is, "what is free memory?". Many poeple think that unused memory is free, and in fact it is. But memory that is used for buffers and cache AND can be given to programs as soon as they ask for memory is free in this sense as well. Feb 24 19:02:05 <_Lobo_> so it's Hack&Dev related :/ Feb 24 19:02:38 but becuase it's not unused, but used for Cache & Buffers, it's even better than having large lumps of megabytes lying around totally unused Feb 24 19:03:32 <_Lobo_> schurig: sure, I see the point, it's probably caused of block device :/ Feb 24 19:49:05 OE is passing -Os to something and I don't want that... how to fix? distro angstrom Feb 24 20:33:51 ji huys Feb 24 20:33:54 guys* Feb 24 20:34:00 damn Feb 24 20:34:01 hi* Feb 24 20:34:11 I need some general info on crosscompiling Feb 24 20:35:00 when I'm trying to crosscompile a thing, configure ends up with Feb 24 20:35:11 checking wheter using broken Apple precompiler... configure: error: cannot run test program while cross compiling Feb 24 20:35:18 how to overcome it? Feb 24 20:50:47 olegfink: only by patching the configure.ac file Feb 24 20:50:59 and how does oe do the trick? Feb 24 20:51:06 olegfink: also, there's a possibility to pre-seed configure with result Feb 24 20:51:58 olegfink: basically, configure tries to run things to find out what is true and what not. And when you provide the result in the config cache, it won't run the test Feb 24 20:52:16 and what's the oe approach? Feb 24 20:52:24 pathcing or precaching? Feb 24 20:52:27 olegfink: and in in a crosscompilation env, you can't (cross)compile a program for, say, ARM, and than run this ARM binary on x86 Feb 24 20:52:37 olegfink: both Feb 24 20:52:44 hm Feb 24 20:52:54 so what should I prefer? Feb 24 20:53:01 some developers do it this way, some other Feb 24 20:53:11 I'm trying to get my packaging system to be able to crosscompile Feb 24 20:53:29 but what 'inherit autotools' do? Feb 24 20:53:30 also, if the check is general (e.g. size of an int, it's not 4 on all platform) and many ./configure's try to check this, then it's better to put this in the cache Feb 24 20:53:42 if the check is only for one app, then it's better to patch that app Feb 24 20:54:06 can you show me the code to write the cache? Feb 24 20:54:13 or is it done by habd? Feb 24 20:54:18 olegfink: in OE, you can see pre-computed results for tests in the "site" directory Feb 24 20:55:01 aha Feb 24 20:55:20 but will it make configure skip Apple checking? Feb 24 20:55:27 inherit autotools makes the bitbake read classes/autotools.bbclass Feb 24 20:55:28 I'm interested in sh-common Feb 24 20:56:44 olegfink: if many of your apps try to check for the apple computer, then look inside your configure script and find the name of the variable it tries Feb 24 20:57:06 aha Feb 24 20:57:09 so I see Feb 24 20:57:09 olegfink: then add this variable and your "no" to site/sh-common Feb 24 20:57:38 the method is very different for every problem Feb 24 20:57:47 no magic here, heh? Feb 24 20:57:55 olegfink: re-compile your package, be sure to clean it (bitbake -b XXX.bb -c clean if I remember the syntax, I have a shell script "bbclean XXX.bb" which I normally use) Feb 24 20:58:07 oops stop Feb 24 20:58:11 I don't use OE Feb 24 20:58:13 :D Feb 24 20:58:22 olegfink: but you are in #oe :-) Feb 24 20:58:31 yep Feb 24 20:58:39 'coz I know oe can do it Feb 24 20:58:45 and I wonder how Feb 24 20:58:56 03coredump2 07org.oe.dev * r98da602b... 10/ (5 files in 4 dirs): ntpclient: Add new recipe for ntpclient, courtesy of Colin Hill aka hillct. Closing bug #1894 Feb 24 20:59:00 03coredump2 07org.oe.dev * r12991427... 10/ (6 files in 4 dirs): logrotate: Add new recipe for logrotate, courtesy of Colin Hill aka hillct. Closing bug #1895 Feb 24 20:59:00 olegfink: run ./configure. You'll see "--cache-file=FILE". Use that as a hint. Feb 24 20:59:28 well Feb 24 21:15:17 hi guys, im a real newbie on oe, just trying to setup the env. I seem to be having trouble getting monotone, does anyone know of a package for openSuse 10.2? Feb 24 21:22:51 03coredump2 07org.oe.dev * rb68accb5... 10/ (1 packages/meta/slugos-packages.bb): slugos-packages.bb: Add ntpclient and logrotate to the SlugOS unstable feeds Feb 24 21:23:36 w00t Feb 24 21:26:11 http://monotone.ca/ Feb 24 22:49:32 hi Feb 24 22:49:38 I am trying to build, handhelds kernel 2.6.20 from cvs Feb 24 22:49:47 but it gives me -hh version mismatch Feb 24 22:49:54 how can i solve this? Feb 24 22:49:58 this is what i use for preferred kernel in angstrom.conf Feb 24 22:50:02 PREFERRED_VERSION_linux-handhelds-2.6_aximx50 = "2.6.20-hh0+cvs${DATE}" Feb 24 23:03:47 ertan, fixed, but you won't be able to build -hh0 now anyway. HEAD is -hh1 now Feb 24 23:03:51 03pfalcon 07org.oe.dev * r25ebcc21... 10/ (1 packages/linux/linux-handhelds-2.6_cvs.bb): linux-handhelds-2.6 cvs: CSV HEAD is 2.6.20-hh1 now. Feb 24 23:04:54 oh thank you. i will go with -hh1 then Feb 25 00:06:15 hi *, Feb 25 00:06:50 how do i make bitbake make uImage instead of zImage in kernel package ? Feb 25 00:09:05 I did that, let me look it up Feb 25 00:10:10 in your kernel.bb file, first make sure to add DEPENDS = "uboot-utils" Feb 25 00:10:14 you will need that for creating the uimage Feb 25 00:10:44 then... Feb 25 00:11:22 hmm, well I see that maybe I did not do it correctly :) Feb 25 00:11:43 take a look in packages/linux Feb 25 00:11:55 grep for uImage, you will find some examples Feb 25 00:12:08 better to look there, cuz now I am not sure if I did mine right :) Feb 25 00:13:33 Jin^eLD: oh thx guy, i'm trying Feb 25 00:14:05 Jin^eLD: bitbake get u-boot source too Feb 25 00:14:23 I did not really need to build uboot, uboot utils were enough Feb 25 00:14:33 but maybe it needs the whole source, don't know Feb 25 00:14:44 basically, you see the mkimage stuff in the .bb file? Feb 25 00:14:57 Jin^eLD: yes, i saw it, Feb 25 00:15:02 you'll have to do something similar, using data for your board.. Feb 25 00:18:42 Jin^eLD: uImage(in mydistro/tmp/work/..linux/arch/arm/boot) not created after i bitbake mydisto Feb 25 00:20:44 Jin^eLD: i think uboot-utils packages, only use mkimage to strip ELF header for u-boot binary to u-boot.bin Feb 25 00:23:20 Jin^eLD: when i compile kernel, make uImage (if i want to get uImage). But i don't know how to pass this argument to bitbake when compile Feb 25 00:25:37 uhm Feb 25 00:25:40 well, I am making a vmlinux Feb 25 00:25:49 and then I am converting it to uImage using the mkimage tool Feb 25 00:26:39 maybe you need to go the same way I did... Feb 25 00:26:45 I set KERNEL_IMAGETYPE = "vmlinux" Feb 25 00:26:56 and then in do_deploy() Feb 25 00:27:17 I ran mkimage over it Feb 25 00:27:48 I needed to do objcopy first with some params **** ENDING LOGGING AT Sun Feb 25 02:59:58 2007