**** BEGIN LOGGING AT Thu Oct 30 02:59:57 2008 Oct 30 03:34:28 hmm guys Oct 30 03:34:36 what do u think of disablign the ldconfig on boot? Oct 30 04:54:19 hoi hoi Oct 30 09:00:50 hi Oct 30 09:21:12 03Koen Kooi  07org.openembedded.dev * r6314775271 10openembedded.git/packages/xfce/ (exo/exo-no-tests.patch exo_0.3.4.bb): Oct 30 09:21:12 exo: fix build of python modules and disable running of tests at Oct 30 09:21:12 buildtime, my buildmachine can't run target code Oct 30 09:32:47 03Koen Kooi  07org.openembedded.dev * rbdec70c94c 10openembedded.git/packages/dsplink/ti-lpm-module_2.21.bb: ti-lpm-module: not using oe_runmake allows us to get rid of all the sed magic Oct 30 09:32:47 hi woglinde Oct 30 09:33:09 morning pb Oct 30 09:33:55 pb hm do you know if I can build om2008 image norw right of oe? Oct 30 09:34:14 or do I need the om git repo Oct 30 09:34:39 You can build fso-image out of OE and get a working binary, but I don't think om2008 is merged. Oct 30 09:34:55 hm okay Oct 30 09:35:00 even for fso-image I think there might be bits in the OM tree which are more uptodate than OE. I'm not exactly sure about that though, mickey|shower would be your resource. Oct 30 09:35:33 hm okay I need the qtopia stuff anyway Oct 30 09:35:38 okay Oct 30 09:35:49 probably best use the om tree then. Oct 30 09:35:57 I guess you could ask in #openmoko-cdevel, the h4x0rs in there should know Oct 30 09:37:04 03Koen Kooi  07org.openembedded.dev * r9f0a59964a 10openembedded.git/packages/dsplink/codec-engine_2.21.bb: codec-engine 2.21: apply fix for ti-lpm-module build Oct 30 09:41:45 h, Oct 30 09:41:47 hm Oct 30 09:43:22 ah Oct 30 09:43:23 now Oct 30 09:43:26 *g* Oct 30 09:44:32 heh Oct 30 09:44:36 woglinde: you aren't making a lot of sense Oct 30 09:45:18 pb yes yes I wanted to clone the om repo and missed a /git/ Oct 30 09:51:29 Is it possible to configure bitbake to run clean after each sucessfull build of bitbake file? Or are files in tmp/work needed for other builds? Oct 30 09:51:46 JaMa: see rm_work Oct 30 09:52:00 pb__: thanks Oct 30 10:17:28 florian_: good morning Oct 30 10:19:28 good morning Oct 30 10:20:22 hi florian Oct 30 10:30:31 morning all Oct 30 10:31:04 hi thesing Oct 30 10:32:54 03Koen Kooi  07org.openembedded.dev * r22e85416d9 10openembedded.git/packages/linux/ (2 files in 2 dirs): linux-omap git: add patch for fix scaling for the overlay Oct 30 10:35:17 morning oe devs Oct 30 10:35:24 i've got an issue with devshell not setting up the PATH environment variable properly Oct 30 10:35:31 args what are om doing there Oct 30 10:35:38 pulling all kernel Oct 30 10:35:40 s Oct 30 10:35:52 as git Oct 30 10:37:13 mattface can you explain in a better way what your problem is? Oct 30 10:37:40 yep, gimme sec. Oct 30 10:40:28 devshell is starting a new gnome-terminal in the source folder of the package im working with, thats perfect, but the command it is supposed to exec (bash --rcfile $TERMRCFILE) doesnt seem to do to add anything to the PATH Oct 30 10:40:38 incidentally $TERMRCFILE is empty Oct 30 10:41:16 whats devshell? Oct 30 10:41:37 03Koen Kooi  07org.openembedded.dev * r1320d90a7a 10openembedded.git/classes/package.bbclass: package bblass: fix ldconfig_postinst_fragment to not return an error if no /etc/ld.so.conf is present Oct 30 10:41:39 and how is it related to openembedded? Oct 30 10:41:42 oh dear :) Oct 30 10:42:00 it's a class in openembedded, right? Oct 30 10:42:31 hm okay I never used it Oct 30 10:42:41 kk Oct 30 10:43:12 and almost none of the devs here use it Oct 30 10:43:31 hmm Oct 30 10:44:42 it's kinda perfect for what I need Oct 30 10:45:58 mattface: I think mickey|shower is the man you need. Oct 30 10:46:05 let's hope he gets out of the shower soon, he's been in there for about 22 hours. Oct 30 10:46:32 pb *g* Oct 30 10:46:41 :) Oct 30 10:47:18 hm he is awake Oct 30 10:54:33 hey Oct 30 10:54:41 anyone have problems qith qemu builds of late? Oct 30 10:55:56 stefan was denn? Oct 30 10:56:14 ups Oct 30 10:56:39 03Koen Kooi  07org.openembedded.dev * r39430f1c3b 10openembedded.git/contrib/angstrom/build-feeds.sh: angstrom feed builder: add xf86-input-tslib Oct 30 11:11:04 03Koen Kooi  07org.openembedded.dev * rdb3b632021 10openembedded.git/packages/dsplink/ (3 files in 2 dirs): ti-cmemk-module, codec-engine: add patch to fix builds with >=2.6.26 and class_device_* API change Oct 30 11:17:42 hi rschuster Oct 30 11:18:40 hi florian Oct 30 11:39:30 hmm Oct 30 11:39:32 question Oct 30 11:39:34 how do i disable a .bb Oct 30 11:39:40 but keep it there in the tree Oct 30 11:39:52 basically the new qemu_svn packages dont compile - compiler errors Oct 30 11:40:00 (cant alloc registers) Oct 30 11:40:06 the old 9.1 works fine Oct 30 11:40:50 raster: you could use BBMASK Oct 30 11:41:17 i dont want the packages completley removed Oct 30 11:41:39 just want it to use qemu-native_0.9.1.bb & qemu_0.9.1.bb Oct 30 11:41:55 instead of qemu-native_svn.bb & qemu_svn.bb Oct 30 11:42:08 disabling in the svnrevs doesnt do that Oct 30 11:42:41 oe seems to "refer an _svn to a _0.9.1 Oct 30 11:43:05 either use BBMASK, like thesing said, or set PREFERRED_VERSION_qemu-native. Oct 30 11:47:08 hi, if I want to change the config parameters for a kernel build, what would be the best way ? Oct 30 11:48:57 hmmm how do i use BBMASK? Oct 30 11:49:33 no docs on it in the tree :( Oct 30 11:52:53 BBMASK = "(qemu-*_svn*)" Oct 30 11:52:54 ? Oct 30 11:54:02 raster, edit the .bb file and change the PREFERRED line Oct 30 11:55:39 there isnt one Oct 30 11:55:45 thats the thing Oct 30 11:55:54 this is a global oe problem Oct 30 11:56:07 every machine and distro is affected Oct 30 11:56:26 there isnt any specific qemu preferred version Oct 30 11:56:41 i can start adding it into every .inc file i can find Oct 30 11:56:45 but thats "wrong" Oct 30 11:58:47 or it just seems wrong Oct 30 11:59:16 hmm Oct 30 11:59:21 seems putting it in local.conf works Oct 30 11:59:25 its a bit better :) Oct 30 11:59:47 raster, google for openembedded user manual the read section 8.5 Oct 30 12:00:24 DEFAULT_PREFERENCE is probably what you want to use in the recipe (not PREFERRED as my memory told me) Oct 30 12:01:24 http://www.openembedded.info/user-manual/index.html Oct 30 12:01:29 hmm Oct 30 12:01:32 hope this is still actual Oct 30 12:01:35 doesnt seem to cover that there Oct 30 12:01:37 jupp Oct 30 12:02:21 seems same problem with uclibc-0.9.29 and uclibc_svn Oct 30 12:02:47 pdf is at http://oe.linuxtogo.org/~koen/usermanual.pdf Oct 30 12:03:00 8.25 it seesm talks about that Oct 30 12:03:04 and it is section 8.25, not 8.5 Oct 30 12:03:43 zecke pasted "ssh key for woglinde" at http://paste.lisp.org/display/69444 Oct 30 12:03:44 bye the way you can build the manual yourself by checking out the doc repos Oct 30 12:03:49 * eFfeM thinks that having a bb file for the latest version of svn is quite risky, better use a date based extract Oct 30 12:04:05 DEFAULT_PREFERENCE_qemu-native = "0.9.1" Oct 30 12:04:20 woglinde: did not know that Oct 30 12:04:54 eFfeM: there is an SVNREV in sane-srcrevs Oct 30 12:04:56 raster no Oct 30 12:04:59 for qemu Oct 30 12:05:07 but whatever svnrev was chosen doesnt work Oct 30 12:05:08 raster, not sure if that will work, I only did this inside specific bb files Oct 30 12:05:17 problem is my prior wqorking tree worked on 0.9.1 Oct 30 12:05:22 so i know that works Oct 30 12:05:26 DEFAULT_PREFERENCE takes argmuents from -1 to 100 not versions Oct 30 12:05:28 yeah Oct 30 12:05:32 i get the specific .bb's Oct 30 12:05:44 the problem is i'm clearing up more general problems and updates for oe Oct 30 12:05:51 si need to have this work "generally" Oct 30 12:06:29 ooh Oct 30 12:06:31 inside the .bb Oct 30 12:06:33 i see Oct 30 12:06:38 packages/glibc/glibc_2.3.2+cvs20040726.bb:DEFAULT_PREFERENCE_sh3 = "-99" Oct 30 12:06:44 hmm Oct 30 12:09:27 raster not everything is perfect (and I guess that since it is still an int it will just work), but feel free to fix and/or write a bug report Oct 30 12:10:30 sure Oct 30 12:10:34 i just have 20 patches to merge in Oct 30 12:10:46 and severalk maintainers to hunt down and "coordinate fixes" with Oct 30 12:10:53 and i now have a non-buildign tree Oct 30 12:11:00 so i'm a bit miffed and want a quick fix Oct 30 12:11:18 oe sucks so much time compared to upstream dev whihc is much faster - i am just trying to minimise time Oct 30 12:11:37 i understand, but I do not have a quick fix Oct 30 12:13:41 sure Oct 30 12:13:42 I just took it as a fact of life. it all depends on who made the bb files and how they tested etc. And not everyone is equally rigid (and yes, I've made my mistakes too, so I tend to be forgiving) Oct 30 12:13:52 the DEFAULT_PREFERENCE seems to work Oct 30 12:13:55 maybe an expert can step up with a fix Oct 30 12:14:07 this is a deep compiler issue/bug Oct 30 12:14:14 cant alloc a register Oct 30 12:14:22 i just want a workign tree again! :) Oct 30 12:14:34 Oct 30 12:14:34 btw Oct 30 12:14:38 ~seen koen Oct 30 12:14:40 koen was last seen on IRC in channel #oe, 322d 2h 8m 17s ago, saying: 'I forgot to commit that portion yesterday'. Oct 30 12:14:53 zecke pasted "other key for woglinde" at http://paste.lisp.org/display/69446 Oct 30 12:14:54 is koen gone here ? Oct 30 12:15:03 whats the down-low with Oct 30 12:15:04 git push origin yourname/testbranch Oct 30 12:15:14 he learned me all those things. I've been away far too long! Oct 30 12:15:22 ie pushing my own branch to oe's git server to people can test/review the patches i've made Oct 30 12:15:57 ie Oct 30 12:16:04 git push origin raster/head Oct 30 12:16:06 for example Oct 30 12:16:42 so i can have my own branch that i can merge with head and keep changes in safely and trickle-feed them back to the main dev branch as the maintainers "approve" Oct 30 12:16:43 ? Oct 30 12:18:49 raster, can't help you with this, I'm not a git wiz Oct 30 12:19:48 raster: try? Oct 30 12:20:23 zecke: is it ok if i do it Oct 30 12:20:29 it is git Oct 30 12:20:38 if i mess somehting up u can always revert/backout etc. Oct 30 12:20:46 just the "guide" says "dangerous" Oct 30 12:20:57 so i'm actually paying attention to the policies Oct 30 12:21:03 raster: --force is dangerous and should be forbidden on org.openembedded.dev :) Oct 30 12:21:08 i am temtped tho to just rabidly commit fixes :) Oct 30 12:21:19 nah Oct 30 12:21:22 not goign to force Oct 30 12:21:29 git-push origin raster/backout looks safe and innocent Oct 30 12:21:36 cool Oct 30 12:21:43 i'll put my changes there for now Oct 30 12:21:47 at leats theiy're published Oct 30 12:21:52 that way i can refer to them easily for review Oct 30 12:22:04 and feed them back to origin as appropriate Oct 30 12:24:54 error: src refspec raster/backout does not match any. Oct 30 12:24:55 hmmm Oct 30 12:25:10 raster: you don't have a local branch called that way? Oct 30 12:25:21 local branch is just "raster" Oct 30 12:25:30 i dont intend to keep multiple branches of my own Oct 30 12:25:33 raster: git-push origin raster:raster/backout? let me take a look Oct 30 12:25:37 just one (me!) Oct 30 12:26:01 git push origin raster:raster/backout Oct 30 12:26:01 ? Oct 30 12:26:09 raster: try that :) Oct 30 12:26:37 nup Oct 30 12:26:39 same error Oct 30 12:26:40 you should get a "mini-me" Oct 30 12:26:46 hahaha Oct 30 12:27:05 you really have a raster branch? Oct 30 12:27:10 yup Oct 30 12:27:13 git-checkout -b raster/raster? Oct 30 12:27:24 and then push that? Oct 30 12:27:53 11:26PM ~/oe/openembedded > git branch Oct 30 12:27:53 org.openembedded.dev Oct 30 12:27:53 * raster Oct 30 12:28:33 workaround is to create raster/raster and push that, I need to get food now Oct 30 12:28:38 is that unhappiness/bad? Oct 30 12:28:44 hmmm Oct 30 12:28:52 ok Oct 30 12:28:59 and that will copy my existing commits over - right? Oct 30 12:29:40 git checkout -b raster/raster will create a branch from your current branch Oct 30 12:29:49 ugh Oct 30 12:29:50 :) Oct 30 12:29:59 ugh Oct 30 12:30:02 that doesnt work Oct 30 12:30:22 11:27PM ~/oe/openembedded > git-checkout -b raster/raster Oct 30 12:30:22 error: unable to resolve reference refs/heads/raster/raster: Not a directory Oct 30 12:30:27 this is not going well Oct 30 12:30:59 03Koen Kooi  07org.openembedded.dev * r3f9ca3c829 10openembedded.git/packages/dsplink/ti-paths.inc: dsplink: add omap3evm support Oct 30 12:32:05 re Oct 30 12:44:26 hi zecke Oct 30 12:49:08 when I do a bitbake virtual/kernel it complains that the preferred version 2.,5 of glibc is not available), yet bitbake glibc-2.5 builds Oct 30 13:06:50 03Koen Kooi  07org.openembedded.dev * r4bbfd7ad9f 10openembedded.git/packages/dropbear/dropbear.inc: Oct 30 13:06:50 dropbear: fix do_install for all distros that aren't named 'openmoko' Oct 30 13:06:50 * IMPORTANT: test -e foo && bar will make do_install exit with an error if 'foo' doesn't exist, so use the if [ -e foo ] ; then pattern instead Oct 30 13:08:49 hmm Oct 30 13:08:59 i hope my push didnt do anything bad.. i hope it went to my branch... Oct 30 13:20:04 did someone built aspell lately? Oct 30 13:23:09 otavio I am not Oct 30 13:26:43 woglinde: libtool is broken; I'll probably add a patch to fix libtool Oct 30 13:27:33 otavio libtool or aspell? Oct 30 13:27:44 woglinde: libtool script in aspell source Oct 30 13:27:59 hehe Oct 30 13:28:34 there are some errors in the newer libtool but no in our proposes Oct 30 13:29:31 03Koen Kooi  07org.openembedded.dev * r26f6477a2d 10openembedded.git/contrib/weekly-changelog-report.py: weekly changelog: add in bugstats from the mtn2cl script Oct 30 13:31:58 humm it could be a failure due the shell Oct 30 13:32:02 let me check Oct 30 13:32:08 it has libtool 2.2.2 on it Oct 30 13:32:13 it shouldn't fail Oct 30 13:33:27 hm# Oct 30 13:33:38 in oe we use 2.2.4 with ansgtroem Oct 30 13:33:40 morning Oct 30 13:33:43 this can cause problems Oct 30 13:33:47 hi vivijim Oct 30 13:36:39 I had issues with libvncserver using libtool 2.2; so I've reverted it locally to use libtool 1.5 Oct 30 13:36:52 * otavio is updating it Oct 30 13:57:37 03Koen Kooi  07org.openembedded.dev * rb6220af33a 10openembedded.git/classes/kernel.bbclass: Oct 30 13:57:37 kernel bbclass: start adding support for >=2.6.27 where headers moved Oct 30 13:57:37 * still lack a symlink for mach/, dunno how to do that properly since the information is dynamically generated with makefiles Oct 30 14:50:40 hi kergoth Oct 30 14:50:51 hey Oct 30 14:53:16 hi kgilmer Oct 30 14:53:22 hey woglinde Oct 30 14:58:24 lo Oct 30 14:58:38 otavio OE enviroment working :D Oct 30 15:15:59 hey guys its possyble to use distcc with bitbake? Oct 30 15:19:40 alo: yes, but it's not likely to result in a statistically significant improvement overall, because i doubt cpu is the main bottleneck for most builds. maybe glibc and stuff.. :) Oct 30 15:21:28 hum.. my problem is that i try to buil all the enviroment in the ps3 to gain speed becose the other option is my laptop celeron 500Mhz ^^ Oct 30 15:21:37 iwas try* Oct 30 15:23:06 all work but when i do bitbake nano the bitbake says something os ppc64 for the ps3 architecture..and i dont saw nothing about over there.. Oct 30 15:23:23 in any case im reading about distcc Oct 30 15:26:38 it should be able to make use of distcc, but i expect you may have to set up the crosscompilers manually.. haven't looked at it in years, maybe someone's improved it Oct 30 15:26:39 * kergoth shrugs Oct 30 15:35:49 morning kergoth Oct 30 15:37:24 re pb Oct 30 15:55:59 some one with OE in the ps3? Oct 30 15:55:59 xd Oct 30 16:33:16 <_diego_> alo, try icecc Oct 30 16:34:59 hi, is there a way to rebuild all packages that depend on some package? Oct 30 16:35:04 using bitbake Oct 30 16:35:38 exodos hm notr really Oct 30 16:37:27 exodos: I'm afraid that you'll have to do it manually... Oct 30 16:37:43 you could add it Oct 30 16:37:45 pretty easily Oct 30 16:38:31 hmmmm Oct 30 16:38:35 * kergoth looks into it Oct 30 16:38:39 I manually added new libdrm-2.4.0 to my package list (because of xf86-driver-intel), now I have to rebuild all packages depending on it Oct 30 16:39:07 but I have no clue how Oct 30 16:39:40 oh, hmm.. nevermind, i was thinking the other way around Oct 30 16:39:47 hm khem has smonething doen for gcc and libc Oct 30 16:39:52 but I am not sure Oct 30 16:40:07 could makea little task to output the package name of any package that deps on it, and go from there Oct 30 16:40:28 kergoth: how can i do that? Oct 30 16:40:32 im newbie with bitbake Oct 30 16:56:50 hey i i wanna compile only the htcuniversal angstrom how much space i need in my /tmp folder??? Oct 30 16:56:55 if* Oct 30 16:57:49 if i want to compile only one image full for a architecture how much space i need? Oct 30 17:02:59 kergoth, exodos: there is something like that: BB_STAMP_POLICY. I didn't really play with it, but if set to the right value bitbake will rebuild all packages that depend on X if you change X Oct 30 17:03:24 But I dont know if its possible for just one package. Oct 30 17:04:05 ALoGeNo hm up to 10gig if you do not use rm_work Oct 30 17:04:49 and how i do that? Oct 30 17:04:56 :)) Oct 30 17:05:14 i have all working 2 days testing.. Oct 30 17:05:15 xD Oct 30 17:06:08 woglinde you think that is a hard work to configure bitbake to build with distcc or something like that? Oct 30 17:10:31 ALoGeNo we have ice support Oct 30 17:11:07 ALoGeNo hm I suggest you read the oe manual Oct 30 17:11:32 this will answer a lot of questions Oct 30 17:11:34 http://bec-systems.com/oe/html/index.html Oct 30 17:11:51 oh.. very nice that link you are so kind :) Oct 30 17:13:05 oh isaw clicked lines.. i need read more there i guess :) Oct 30 17:13:22 tnx in any case :) Oct 30 17:13:29 no porb Oct 30 17:33:42 re robert Oct 30 17:54:01 Hi. I have following misconfiguration error while bitbake: Oct 30 17:54:01 Error, Your PACKAGE_ARCHS field contains duplicates. Perhaps you set EXTRA_PACKAGE_ARCHS twice accidently through some tune file? Oct 30 17:54:01 I updated from mandriva 2008.1 to 2009.0, with same local.conf all was working.. Oct 30 17:54:36 Kalemas we changed the package archs from machine to tuning files Oct 30 17:54:51 and not all machine.conf's are updated yet Oct 30 17:55:07 kalemas which machine for you are compiling? Oct 30 17:55:48 HI. I have following misconfiguration error while bitbake: Oct 30 17:55:48 Error, Your PACKAGE_ARCHS field contains duplicates. Perhaps you set EXTRA_PACKAGE_ARCHS twice accidently through some tune file? Oct 30 17:55:48 I updated form mandriva 2008.1 to 2009.0, earlier all was working. Oct 30 17:56:20 HI. I have following misconfiguration error while bitbake: Oct 30 17:56:20 Error, Your PACKAGE_ARCHS field contains duplicates. Perhaps you set EXTRA_PACKAGE_ARCHS twice accidently through some tune file? Oct 30 17:56:20 I updated form mandriva 2008.1 to 2009.0, earlier all was working. Oct 30 17:56:20 hi guys, is there any list of all supported features? (DISTRO/MACHINE_FEATURES) Oct 30 17:56:28 zecke: ^ Oct 30 17:56:29 dfs Oct 30 17:56:30 vivijim nope Oct 30 17:56:49 kalemas did you read my answers? Oct 30 17:56:57 oh Oct 30 17:57:05 seems some problems with irc client Oct 30 17:57:50 woglinde: :( I think it would be useful to avoid duplications like wifi/wireless touchscreen/ts Oct 30 17:59:08 Hi. I have following misconfiguration error while bitbake: Oct 30 17:59:08 Error, Your PACKAGE_ARCHS field contains duplicates. Perhaps you set EXTRA_PACKAGE_ARCHS twice accidently through some tune file? Oct 30 17:59:08 I updated from mandriva 2008.1 to 2009.0, earlier all was working. Oct 30 17:59:40 Hi. I have following misconfiguration error while bitbake: Oct 30 17:59:40 Error, Your PACKAGE_ARCHS field contains duplicates. Perhaps you set EXTRA_PACKAGE_ARCHS twice accidently through some tune file? Oct 30 17:59:40 I updated from mandriva 2008.1 to 2009.0, earlier all was working. Oct 30 17:59:42 asfd Oct 30 17:59:42 asd Oct 30 17:59:42 a Oct 30 17:59:43 Kalemas2 whats wrong? Oct 30 17:59:43 sd Oct 30 17:59:45 asd Oct 30 17:59:47 asd Oct 30 17:59:49 a Oct 30 17:59:51 sd Oct 30 18:00:11 Kalemas: are you drunk? :P Oct 30 18:00:15 Hi. I have following misconfiguration error while bitbake: Oct 30 18:00:15 Error, Your PACKAGE_ARCHS field contains duplicates. Perhaps you set EXTRA_PACKAGE_ARCHS twice accidently through some tune file? Oct 30 18:00:15 I updated from mandriva 2008.1 to 2009.0, earlier all was working. Oct 30 18:00:20 oh no Oct 30 18:02:19 vivijim: no , I don't think such a list exists. it would be good to have it Oct 30 18:03:48 Hi. I have following misconfiguration error while bitbake: Oct 30 18:03:48 Error, Your PACKAGE_ARCHS field contains duplicates. Perhaps you set EXTRA_PACKAGE_ARCHS twice accidently through some tune file? Oct 30 18:03:48 I updated from mandriva 2008.1 to 2009.0, earlier all was working. Oct 30 18:04:52 Hi. I have following misconfiguration error while bitbake: Oct 30 18:04:52 Error, Your PACKAGE_ARCHS field contains duplicates. Perhaps you set EXTRA_PACKAGE_ARCHS twice accidently through some tune file? Oct 30 18:04:52 I updated from mandriva 2008.1 to 2009.0, earlier all was working. Oct 30 18:06:12 hm will he stop Oct 30 18:06:17 or is he just trolling Oct 30 18:06:30 Kalemas: could you repeat that one more time? Oct 30 18:07:25 Sorry! client was not displayed anything but status messages. Oct 30 18:08:18 Kalemas2 solution is Oct 30 18:08:25 edit your machine conf Oct 30 18:08:33 and remove the package-arch Oct 30 18:08:49 they are now handled by the tune files Oct 30 18:09:03 and not all machine.confs are fixed yet Oct 30 18:11:05 hm Oct 30 18:18:35 jfyi at this moment the complete feature list is: ['acpi', 'alsa', 'apex', 'apm', 'bluetooth', 'connex', 'ethernet', 'ext2', 'fat', 'gps', 'i2c', 'irda', 'iwmmxt', 'kernel24', 'kernel26', 'keyboard', 'lcd', 'mmc', 'nfs', 'pci', 'pcmcia', 'phone', 'ppp', 'raid', 'redboot', 'robostix', 'screen', 'serial', 'smbfs', 'sound', 'spi', 'touchscreen', 'uboot', 'usbgadget', 'usbhost', 'ushbost', 'verdex', 'vfat', 'wifi', 'x86'] Oct 30 18:20:02 woglinde: what i need to remove? "armv4 armv4t armv5te iwmmxt " Oct 30 18:20:20 kalemas2 all that are in the tune file too Oct 30 18:25:54 so bye for now Oct 30 18:28:47 woglinde: ok. it should be : PACKAGE_EXTRA_ARCHS = "iwmmxt"? Oct 30 18:29:09 :^( Oct 30 18:52:27 did someone packaged sun jre and flash? Oct 30 19:36:22 hmm Oct 30 19:39:27 hey to build a single image rootfs is with task-base no? Oct 30 19:45:55 re kergoth Oct 30 19:47:10 hey Oct 30 19:47:59 hm Oct 30 19:48:16 grr, i can reliably make this irc client crash :P Oct 30 19:57:31 03Otavio Salvador  07org.openembedded.dev * rb7983d0f73 10openembedded.git/packages/gnumeric/gnumeric_1.8.3.bb: gnumeric: do not overwrite PACKAGES_DYNAMIC to allow locales to be used Oct 30 19:57:32 03Otavio Salvador  07org.openembedded.dev * ra8744762b0 10openembedded.git/packages/mozilla/fennec/mozconfig: fennec: use system libraries while building Oct 30 20:02:47 03Koen Kooi  07org.openembedded.dev * rc5a381985f 10openembedded.git/packages/mplayer/mplayer_svn.bb: mplayer: enable more inputs Oct 30 20:04:31 03Koen Kooi  07org.openembedded.dev * r6b982d8fbd 10openembedded.git/packages/linux/linux-omap_git.bb: linux-omap git: bump PR for staging changes Oct 30 20:12:38 hmm, http://comments.gmane.org/gmane.comp.version-control.git/99516 Oct 30 20:33:29 re florian Oct 30 20:35:45 re Oct 30 21:02:57 anyone familiar with angstrom? Oct 30 21:03:10 i cant use the 'make' command.... Oct 30 21:03:24 is there another command that angstrom uses ? Oct 30 21:03:28 ???? Oct 30 21:03:33 bitbake Oct 30 21:03:35 and their website is down Oct 30 21:03:44 ..... bitbake? Oct 30 21:03:49 hm yes linuxtogo seems to have problems Oct 30 21:04:03 Fl1pFl0p: for what and from where you trying to use 'make'? Oct 30 21:04:19 make mekaiser Oct 30 21:04:21 *g* Oct 30 21:04:26 i extracted a driver that needs to be made Oct 30 21:04:42 florian could you please look at the server? Oct 30 21:04:49 extracted? Oct 30 21:04:57 it was tar balled Oct 30 21:05:20 i edited the 'Makefile' and now i need to 'make' Oct 30 21:05:29 ohh.. Oct 30 21:05:30 hm I thought you were one of these guys who mounts images an exchange binaries Oct 30 21:06:15 any ideas? Oct 30 21:06:46 yes Oct 30 21:06:51 Fl1pFl0p: you can use make on target device only.. but you needed full toolchain then (compiler, linker, etc) Oct 30 21:07:06 write your .bb recipe and provide a patch for it in the openembedded bugtracker Oct 30 21:11:32 woglinde: should be back Oct 30 21:12:08 YEA Oct 30 21:12:11 thanks Oct 30 21:12:13 Fl1pFl0p I suggest you read the oe-manual Oct 30 21:12:28 http://bec-systems.com/oe/html/index.html Oct 30 21:12:58 will do Oct 30 21:16:10 site crash again? Oct 30 21:16:54 nope Oct 30 21:17:03 hm Oct 30 21:17:06 args chache Oct 30 21:17:07 yes Oct 30 21:17:12 crashed again Oct 30 21:18:57 somethign floods the webserver with requests Oct 30 21:19:05 hm Oct 30 21:19:28 DoS? Oct 30 21:19:40 hi ant Oct 30 21:19:52 ant_: :) Oct 30 21:20:05 hi Oct 30 21:21:06 looks like the yahoo bot might be guilty... but I wonder why this happens, the webserver works pretty well for quite some time now. Oct 30 21:21:24 florian hehe time for failtoban Oct 30 21:21:46 florian are we using apache or lighty? Oct 30 21:21:57 woglinde: apache Oct 30 21:22:01 *shrug* Oct 30 21:22:25 * florian votes for an additional server and much of free time to install it better than the first one Oct 30 21:24:34 * florian tries again Oct 30 21:24:46 * Jay7 can help with installation when needed Oct 30 21:26:44 re zecke Oct 30 21:27:28 jay7 you are not alone *g* Oct 30 21:27:45 :) Oct 30 21:29:20 2008 packages are not showing up on webserver now Oct 30 21:32:45 HI. I have following error during bitbake: Oct 30 21:32:46 | checking for i686-linux-gcc... ccache gcc Oct 30 21:32:46 | checking for C compiler default output file name... configure: error: C compiler cannot create executables Oct 30 21:33:05 hello woglinde Oct 30 21:34:29 hi khem Oct 30 21:34:35 Kalemas2: your compiler is hosed, look at the config.log Oct 30 21:34:51 compiler or ccache Oct 30 21:34:55 Kalemas2: there should be some compile or link failure pretty earyl Oct 30 21:35:02 compiler Oct 30 21:35:13 khem could you look at the uclibc hash-style option and find a new test Oct 30 21:35:26 with newer binutils the test fails badly Oct 30 21:35:32 woglinde: hmm Oct 30 21:36:06 which test are you talking about Oct 30 21:36:34 I am thinking that this hash-style thing should be controlled in compiler Oct 30 21:37:06 khem mom Oct 30 21:38:19 khem its check_ld in Rules.mak Oct 30 21:38:32 if $(LD) $(1) -o /dev/null -b binary /dev/null > /dev/null 2>&1; Oct 30 21:39:03 Hello. I've got a question about RMK patches. At ARM Linux homepage is said: "As of 2.6.0-rmk2, sufficient patches have been merged into mainline to make the release of -rmk kernels unnecessary. Therefore, there will be no further -rmk patch releases." Can somebody tell me why is "2.6.3-rmk0-hh0-xda0" branch in Xanadux CVS? It isn't clear to me. Oct 30 21:40:19 woglinde: I still do not understand :) Oct 30 21:40:25 Do you have LDSO_GNU_HASH_SUPPORT Oct 30 21:40:28 set in .config Oct 30 21:40:51 khem nope the build process is testing it Oct 30 21:41:13 yes it tests it what error do you get Oct 30 21:41:16 ifeq ($(LDSO_GNU_HASH_SUPPORT),y) Oct 30 21:41:26 LDFLAGS_GNUHASH:=$(call check_ld,--hash-style=gnu) Oct 30 21:41:27 so Oct 30 21:41:33 and if you have newer binutils Oct 30 21:41:35 this fails Oct 30 21:41:50 hmmm thats odd Oct 30 21:41:51 when you have hash-style Oct 30 21:42:06 because of $(LD) $(1) -o /dev/null -b binary /dev/null Oct 30 21:42:08 fails Oct 30 21:42:37 this looks ok Oct 30 21:42:45 where ${1) is -Wl,--hash-style=gnu Oct 30 21:42:55 it will only fail if linker does not accept --hash-style option any longer Oct 30 21:43:04 NOPE!!!!!!!!!!!!!!!! Oct 30 21:43:14 are not you listining me Oct 30 21:43:23 it fails always with newer binutils Oct 30 21:44:14 yes its my fault I didnt report it to the bugtracker Oct 30 21:44:26 and do not have the error message anymore Oct 30 21:50:16 http://googleblog.blogspot.com/2008/10/what-we-learned-from-1-million.html is pretty sad Oct 30 21:51:39 woglinde: I am listening to you Oct 30 21:51:56 woglinde: My point is that if test is failing thats binutils problem not uclibc's Oct 30 21:53:26 woglinde: I am pretty sure that ld is messed up Oct 30 21:54:16 all the check does is it supplies the option and then tries to link /dev/null to a output file which is also /dev/null and then checks for ld's return code Oct 30 21:54:47 if ld is returning non zero on a valid option then its ld's problem Oct 30 21:54:51 hm Oct 30 21:54:56 this is the error Oct 30 21:54:57 ld: warning: cannot find entry symbol _start; not setting start address Oct 30 21:55:21 GNU ld (GNU Binutils for Debian) 2.18.0.20080103 Oct 30 21:55:29 warning should not alter the return code Oct 30 21:55:58 hm okay Oct 30 21:55:59 right Oct 30 21:56:12 sorry Oct 30 21:56:13 yes Oct 30 21:56:28 We could add -e 0x8000 or somesuch and silence the warning Oct 30 21:56:38 aborting at /devel/arm/oetmp2/work/arm-oabi-angstrom-linux-uclibc/binutils-cross-2.18.50.0.7-r4/binutils-2.18.50.0.7/bfd/elf.c line 369 in bfd_elf_get_elf_syms Oct 30 21:56:47 this is the real error for arm Oct 30 21:56:50 thats the real problem Oct 30 21:57:16 I think this is not fixed either in binutils 2.19 Oct 30 21:57:23 could be Oct 30 21:57:36 didnt see a bugreport this Oct 30 21:57:48 and with last binutils 2.18 for linux Oct 30 21:57:51 it wasnt fixed Oct 30 22:04:57 woglinde: file a PR with binuitls bugz Oct 30 22:10:41 khem yes I will test it with 2.19 Oct 30 22:10:44 and then file a bug Oct 30 22:15:07 03Koen Kooi  07org.openembedded.dev * rb694b15ff4 10openembedded.git/packages/bluez/bluez-gnome_0.26.bb: bluez-gnome: package icons and mime types Oct 30 22:17:32 hey please.. that is normal? NOTE: Running task 44 of 2011 , i mean.. my laptop goes to take one week to build that ^^' Oct 30 22:18:01 with task-base what im building? Oct 30 22:18:30 i mean what is the result of the compilation? Oct 30 22:18:44 ALoGeNo: You should let it go Oct 30 22:18:47 rootfs full for my pda model? Oct 30 22:18:56 ALoGeNo: it will barf if it finds errors Oct 30 22:19:09 How feeble is your laptop Oct 30 22:19:38 wowo man my conecction is 20KB/s any times and mi laptop celeron 500Mhz heh.. Oct 30 22:19:51 any way to do more quickly? Oct 30 22:19:56 ALoGeNo: yeah it might take a while Oct 30 22:20:03 What do u want Oct 30 22:20:09 just playing around ? Oct 30 22:20:37 nono np, the problem is the HD of the laptop is like motorcycles Oct 30 22:20:38 xDD Oct 30 22:20:39 you can make it go more quickly by upgrading that ancient machine :P Oct 30 22:20:42 ^^' Oct 30 22:21:10 yah man.. but the crysis cruch is mortal.. Oct 30 22:21:26 * khem agrees with kergoth's advise Oct 30 22:22:00 okey guys anyone with powerpc? Oct 30 22:22:35 the ps3 gives to my one error ppc64-linux whatever.. :/ Oct 30 22:22:53 nothing.. you see? and google the same.. Oct 30 22:23:08 i must use the anciente but strong laptop ;) Oct 30 22:25:14 hey if the same repo and enviroment that i was used in the ps3 works in the laptop like a charm, why can be that in the ps3 nothing, is something there in the repo precompiled for x86? Oct 30 22:26:01 nope Oct 30 22:26:05 nothing precompiled Oct 30 22:26:23 i was looking in google and i was saw one guy with powermac that have the same herror.. Oct 30 22:26:27 ALoGeNo, you might remove your tmp and check conf and vars Oct 30 22:26:30 you can google aroung if some provides the tar.gz from building oe sdk Oct 30 22:27:04 but oe sdk isnt the crosscompiler and git? Oct 30 22:27:20 im using my own cross compiler buil in the ps3 Oct 30 22:27:42 why don't you use the toolchain OE builds? Oct 30 22:27:43 arm-linux-gnu Oct 30 22:27:56 ALoGeNo: ps3 is ppc64 right ? Oct 30 22:28:00 becose my conecction is very slow.. Oct 30 22:28:08 yes Oct 30 22:28:20 in that case I dont know if oe has any ppc64 usecases yet Oct 30 22:28:59 but the ppl with powermac or cell procesors? Oct 30 22:29:14 cant to be envolved in your projects? Oct 30 22:29:25 that cant be :P Oct 30 22:29:26 ALoGeNo, the ps3 is the host, not target, right? Oct 30 22:29:41 yeah thats it Oct 30 22:30:27 bitbake is telling me that no have info of some thing.. let my paste the falut to you guys please.. :) Oct 30 22:30:55 pastebin Oct 30 22:32:07 woglinde: it seems the ld problem could be because of ld/5692 Oct 30 22:34:11 hey take a look please http://pastebin.com/m7c4d9e70 Oct 30 22:34:46 if i dont stop the task that follow to the end.. but stop for the error.. Oct 30 22:35:16 khem ah thanks Oct 30 22:37:06 khem hm is maked as fixed Oct 30 22:38:54 you see some clue? khem ? Oct 30 22:42:14 why is telling me ERROR: Information not available for target 'ppc64-linux' Oct 30 22:42:29 if the target is htcuniversal.. i can understand.. Oct 30 22:42:34 cant* Oct 30 22:46:13 ALoGeNo: run uname -a on the machine. Oct 30 22:47:26 ok Oct 30 22:47:40 pi@ps3:~/stuff/build$ uname -a Oct 30 22:47:41 Linux ps3 2.6.16-1-ps3pf #1 SMP Sun Dec 31 17:35:13 JST 2006 ppc64 GNU/Linux Oct 30 22:47:41 pi@ps3:~/stuff/build$ Oct 30 22:48:20 woglinde: I think they fixed something but the original problem still remains :) Oct 30 22:48:24 reopen it Oct 30 22:48:58 ALoGeNo: well then. the ps3 is a ppc64 architecture machine. so of course thats the target it's trying to build some things for Oct 30 22:49:11 ALoGeNo: and whatever package is failing there, must not support that arch Oct 30 22:49:29 yep looks like that.. Oct 30 22:49:44 what package is it? Oct 30 22:50:18 i told you that i just buil all in the ps3 copy that in usb HD and put in the laptop and runn like speedy gonzales :) Oct 30 22:50:56 i dunno what pakage is :( Oct 30 22:51:23 for that i ask before if the git downloaded have something only for x86 Oct 30 22:51:35 hm insane.bbclass needs to be fixed too for ppc64 Oct 30 22:51:42 becose that is weird Oct 30 22:52:10 oohh.. you know.. you knoow.. heh.. what can i do? Oct 30 22:52:21 algoneo hm I have a better idea maybee Oct 30 22:52:32 try to setup icecc Oct 30 22:52:32 yo command :) Oct 30 22:52:42 and control it from the laptop Oct 30 22:52:55 look in the oe manual or wiki how to achieve this Oct 30 22:53:07 ohh, i see where its failing Oct 30 22:53:14 i got bit by those exact errors when trying to build on Mac OS X Oct 30 22:53:24 its the siteinfo stuff in the class Oct 30 22:53:30 hehe Oct 30 22:53:39 yes this needs to be fixed too Oct 30 22:53:44 which is laid out weirdly, imo it should slpit arch vs os info Oct 30 22:53:49 since theres a lot of duplication in that table Oct 30 22:53:50 x Oct 30 22:53:51 d Oct 30 22:54:17 im a good betatester Oct 30 22:54:18 xD Oct 30 22:54:44 you now can to say that it works in the ps3 you must to see how work the machine :) Oct 30 22:54:53 compile like the devil Oct 30 22:56:39 okey if you go to fix that in a few days i will try with icecc thanks for explain all the stuff guys :) Oct 30 22:57:45 woglinde but if deactive the insane stuff in config that will run? Oct 30 22:58:32 or it needs more changes to do? Oct 30 22:59:33 ALoGeNo yes Oct 30 22:59:42 oki doki Oct 30 22:59:53 or you add the right entries Oct 30 23:00:01 but I do not know them for ppc64 Oct 30 23:00:17 oh you think that ill know? Oct 30 23:00:34 hehe.. icecc better Oct 30 23:00:35 XD Oct 30 23:01:19 ^^ thanks a lot for your help teachers :) Oct 30 23:02:11 ALoGeNo hm ps3 is buildmachine is not bad Oct 30 23:02:22 so let us try to get there Oct 30 23:03:13 oh np, i go learn to use icecc takeit easy guys :) Oct 30 23:03:51 but if you need to test something put my nick and so :) Oct 30 23:04:54 and happy hallowen >:-)= Oct 30 23:17:55 hmm... $ su - root \ su: must be suid to work properly \ /bin/su -> /bin/busybox Oct 30 23:18:14 ant hehe Oct 30 23:18:24 seems only x11-image..doh Oct 30 23:18:39 I think this was discussed Oct 30 23:18:55 already chmodded Oct 30 23:19:01 and agreed that you dont will setuid busybox Oct 30 23:19:16 (I missed the discussion) Oct 30 23:19:29 hehe.. Oct 30 23:20:11 ah, opie-mage has real su Oct 30 23:20:19 it seems Oct 30 23:22:04 nope, bulls@it, it's busybox Oct 30 23:22:12 * ant_ reading ML Oct 30 23:26:10 morning raster Oct 30 23:26:20 moo Oct 30 23:26:29 didi break oe? Oct 30 23:26:30 :) Oct 30 23:26:39 heh, hey raster Oct 30 23:26:53 hey hey Oct 30 23:27:05 kergoth: does the branch i pushed bother anyone? Oct 30 23:27:15 heh, dont ask me, I'm still half out of the game Oct 30 23:27:17 raster until now not Oct 30 23:27:18 i'm pushing to my own branch before i pollue origin... Oct 30 23:27:21 just starting to get back into it Oct 30 23:27:28 bbl, home Oct 30 23:27:28 ok Oct 30 23:27:30 fair enough Oct 30 23:34:08 khem still around? Oct 30 23:34:17 woglinde: yes Oct 30 23:34:18 binutils 2.19 has this bug either Oct 30 23:34:45 so its there upstream too right ? Oct 30 23:35:01 khem pardon? Oct 30 23:35:35 you can reproduce this bug in 2.19 right Oct 30 23:36:00 2.19 is not that old so I was assuming that this bug is there in binutils cvs too Oct 30 23:36:16 2.19 is the latest Oct 30 23:36:19 from this week Oct 30 23:36:20 haha Oct 30 23:36:43 but it seems the bug is triggered only with hash-style Oct 31 02:01:09 hmmm you guys really need to get your wiki info on git and your actual git tree to work Oct 31 02:01:17 branching etc. doesnt work as explained Oct 31 02:01:20 thngs are disabled/blocked Oct 31 02:01:34 most things tend to error out Oct 31 02:22:40 * Crofton|work wonders what the wiki/documentation team is up to .... **** ENDING LOGGING AT Fri Oct 31 02:59:57 2008