**** BEGIN LOGGING AT Thu Nov 06 02:59:57 2008 Nov 06 03:46:53 Where can I see the source for "oe_runconf"?? Nov 06 06:10:06 * * OE Bug 4807 has been created by  Nov 06 06:10:08 * * libxfce4util-4.4.2-autobuild Nov 06 06:10:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4807 Nov 06 08:07:58 03Koen Kooi  07org.openembedded.dev * r143a57890a 10openembedded.git/packages/totem/totem-2.24.2/gst-detect.diff: totem: add missing patch for 2.24.2 Nov 06 08:09:05 * * OE Bug 4808 has been created by sledz(AT)dresearch.de Nov 06 08:09:07 * * busybox passwd with shadow support Nov 06 08:09:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4808 Nov 06 08:46:07 woglinde: re - have you committed a fix for it? Nov 06 08:56:41 i am trying to use OE and build a MACHINE="i686-generic" with DISTRO="minimal", however it doesn't appear to compile the cross tools? if i use MACHINE="x86" it goes ahead and compiles the cross compilation toolchain. Does anyone have any pointers on where to look to figure out why the toolchain is not being built? Nov 06 09:01:07 bonjour Nov 06 09:23:09 hi Nov 06 09:30:23 morning Nov 06 09:30:29 morning hrw Nov 06 09:32:14 aloril: I did not even know that such distro exists Nov 06 09:32:28 Crofton|irssi: centos? I used 5.1 with poky Nov 06 09:46:27 03Koen Kooi  07org.openembedded.dev * r00ea9a44ec 10openembedded.git/packages/xorg-xserver/xserver-kdrive_1.4.0.90.bb: xserver kdrive 1.4.0.90: bump PR Nov 06 09:47:57 G'morning Nov 06 11:21:37 zecke: hail zecke Nov 06 11:26:20 master Nov 06 15:14:39 kergoth: good morning Nov 06 15:14:54 hey Nov 06 15:33:48 good morning Nov 06 15:46:48 hi mwester Nov 06 15:55:05 * * OE Bug 4809 has been created by  Nov 06 15:55:07 * * portmap-6.0-autobuild Nov 06 15:55:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4809 Nov 06 15:59:07 http://www.emn.fr/x-info/coccinelle/ looks interesting.. if odd Nov 06 16:29:31 03woglinde  07shared/xorg-7.4-update * rda9083a094 10openembedded.git/packages/glib-2.0/ (files/atomic-thumb.patch glib-2.0_2.16.4.bb): Nov 06 16:29:31 glib-2.0: fix atomic-thumb.patch Nov 06 16:29:31 * is only in included in glib-2.0_2.16.4 Nov 06 16:29:31 * missed depenedcy on .la file Nov 06 16:29:31 * bump PR of glib-2.0_2.16.4 Nov 06 16:29:33 03Koen Kooi  07shared/xorg-7.4-update * r143a57890a 10openembedded.git/packages/totem/totem-2.24.2/gst-detect.diff: totem: add missing patch for 2.24.2 Nov 06 16:29:38 03Koen Kooi  07shared/xorg-7.4-update * r00ea9a44ec 10openembedded.git/packages/xorg-xserver/xserver-kdrive_1.4.0.90.bb: xserver kdrive 1.4.0.90: bump PR Nov 06 16:29:41 03Marcin Juszkiewicz  07shared/xorg-7.4-update * rc51ce4a258 10openembedded.git/conf/checksums.ini: checksums.ini: added some entries Nov 06 16:29:44 03Graeme Gregory  07shared/xorg-7.4-update * rce5a7763a9 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into shared/xorg-7.4-update Nov 06 16:29:47 03Graeme Gregory  07shared/xorg-7.4-update * rcd3ad966c9 10openembedded.git/packages/xorg-driver/xf86-video-vmware_10.16.5.bb: xf86-video-vmware_10.16.5.bb : new version for X11R7.4 Nov 06 16:29:50 03Graeme Gregory  07shared/xorg-7.4-update * r300dbe8af6 10openembedded.git/packages/xorg-driver/ (2 files in 2 dirs): Nov 06 16:29:53 xf86-video-vesa_2.0.0.bb : new version for X11R7.4 with some more Nov 06 16:29:59 include path fixes Nov 06 16:30:01 03Graeme Gregory  07shared/xorg-7.4-update * rc554e88ad6 10openembedded.git/packages/xorg-driver/xf86-video-wsfb_0.2.1.bb: xf86-video-wsfb_0.2.1.bb : new version for X11R7.4 Nov 06 16:30:04 03Graeme Gregory  07shared/xorg-7.4-update * rcce43de90c 10openembedded.git/conf/checksums.ini: checksums.ini : even more new checksums for X11R7.4 Nov 06 16:30:07 03Graeme Gregory  07shared/xorg-7.4-update * rffe90dbdc6 10openembedded.git/packages/xorg-driver/xf86-video-xgixp_1.7.99.3.bb: xf86-video-xgixp_1.7.99.3.bb : new version for X11R7.4 Nov 06 17:21:38 anyone feel like testing the latest tslib svn to make sure nothing explodes horribly? Nov 06 17:21:50 * kergoth no longer has the hardware around to test it, until he can get his dad to ship off his old z's Nov 06 17:23:41 blast from the past... :) Nov 06 17:44:47 ping khem Nov 06 18:19:55 <[X]Spot> how can I add my own package in OE ? Nov 06 18:19:59 <[X]Spot> just to add .bb file for it ? Nov 06 18:21:19 <[X]Spot> just to create a simple myPackage_version.bb in /home/bitbake/oe/org.openembedded.dev/packages/myPackage? Nov 06 18:23:30 hmm, how do I check out a new package from OE git? Nov 06 18:23:41 hubar: what do you mean? Nov 06 18:23:45 Is there something like cvs update for git? Nov 06 18:24:05 I wanna see if there are updates for apache2... Nov 06 18:24:08 i suggest you read the git info on the wiki, and other git resources online Nov 06 18:24:24 git pull is probably what you want, which fetches from upstream, then merges the new commits to your local branch Nov 06 18:45:47 greetings! has anyone ever build an oe image for an asus wl-500 router? Nov 06 18:46:20 there is a machine configuration for the wl-500g, but googling doesn't come back with any experiences Nov 06 18:50:53 <[X]Spot> I have Nov 06 18:50:54 <[X]Spot> NOTE: the following files were installed but not shipped in any package: Nov 06 18:50:54 <[X]Spot> NOTE: /usr/local/lib/libv4lconvert.so.0 Nov 06 18:50:55 <[X]Spot> .... Nov 06 18:51:00 <[X]Spot> ant all the packages Nov 06 18:51:12 <[X]Spot> when I triy to build my own package Nov 06 19:21:12 * prpplague pokes kergoth with a sharp stick Nov 06 19:21:23 ow Nov 06 19:21:41 kergoth: any angstrom devrs hanging out here? Nov 06 19:21:55 pretty sure they all are Nov 06 19:22:06 just ask the question and wait :P Nov 06 19:22:44 kergoth: don't have any yet, just wanted to find some place i could ask, as the #angstrom channel is basically useless Nov 06 19:23:29 kergoth: i'd like to get an angstrom build for the M8050 up and running Nov 06 19:23:31 keep in mind the time difference Nov 06 19:23:40 a substantial portion of the OpenEmbedded and angstrom folk are in europe Nov 06 19:23:53 kergoth: yea, figured as much Nov 06 19:24:11 pretty dead in here just now also Nov 06 19:25:57 kergoth: userland and i just don't get along Nov 06 19:26:09 hehe Nov 06 19:26:23 what issue are you hitting? i doubt I'm current enough on the distributionisms to be much help, but.. Nov 06 19:26:30 oh, right.. Nov 06 19:26:35 well either here or angstrom's chan will work Nov 06 19:26:57 hi kergoth prpplague Nov 06 19:26:59 in most ways angstrom is debian, built by oe, with finer package granularity. Nov 06 19:27:06 so keep that in mind when poking around Nov 06 19:27:15 nothing as of yet, i've been using poky, but it looks as though angstrom would fit better for the 8050 Nov 06 19:27:25 chouimat|work: hey bud Nov 06 19:28:20 kergoth: we are willing to give out some free 8050 units for a few folks to do different distros for it Nov 06 19:28:41 cool Nov 06 19:29:22 kergoth: from what i'm seeing, basically the same identical build as the sl-5x00 would be fine Nov 06 19:29:39 prpplague: If you don't mind my curiosity, what is the 8050? Nov 06 19:29:41 should be relatively painless Nov 06 19:29:59 SirManiac: industrial PDA, http://www.elinux.org/M8050 Nov 06 19:30:29 SirManiac: S3C2440 with 128mb ram/ 64mb flash for the base model Nov 06 19:30:56 * prpplague has to go to a meeting, bbi 30 Nov 06 19:30:57 prpplague: thank you :) Nov 06 19:54:05 * * OE Bug 4810 has been created by  Nov 06 19:54:07 * * mpeg2dec-0.4.0b-autobuild Nov 06 19:54:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4810 Nov 06 19:54:10 anyone know what the error ".preinit_array/.init_array/.fini_array support... no" would mean when building glibc? Nov 06 20:45:15 Hello Nov 06 20:46:20 gcc-cross-initial_4.2.4 do_unpack failed, how can i rebuild it whole? Nov 06 20:49:30 remove the bitbake timestamps, should be something like "tmp/stamps/arm*/gcc-cross-initial*" Nov 06 20:50:42 thanks Nov 06 21:07:26 huh, I'm seeing the same error at this guy: http://www.mail-archive.com/openembedded-users@lists.openembedded.org/msg00511.html Nov 06 21:08:02 why would his MACHINE be qemux86? Nov 06 21:10:03 ? Nov 06 21:11:24 jimsheldon: qemux86 is currently one of the stable machine configs for x86... Nov 06 21:11:25 kergoth: the response suggested setting MACHINE to qemux86 Nov 06 21:11:30 yes.. Nov 06 21:11:40 as denix says, its a functional x86 machine Nov 06 21:12:03 ok I just couldn't figure out why someone would want to run uclibc on an x86 system Nov 06 21:12:09 ? Nov 06 21:12:10 jimsheldon: it has nothing to do with qemu specifically, unless you want to run it under qemu :) Nov 06 21:12:16 there are embedded x86 platforms, jimsheldon Nov 06 21:12:19 quite a few of them, in fact Nov 06 21:12:35 the question is - why wouldn't you want to run uclibc on your x86 box Nov 06 21:12:37 true, like the geode Nov 06 21:12:44 heh Nov 06 21:12:57 CosmicPenguin: it's less stable? Nov 06 21:13:05 sez who? Nov 06 21:13:27 jimsheldon: besides Geode, there are i386-i686 embedded parts as well Nov 06 21:14:13 CosmicPenguin: incomplete? Nov 06 21:14:31 ok now that I've cleared that up... my MACHINE is set to wl500g, does that mean there is not a qemu config for me? Nov 06 21:14:45 CosmicPenguin: I'm not saying it's not good. I use it for embedded targets, but wouldn't use it for my desktop Nov 06 21:16:32 jimsheldon: you actually need to cross-compile. see if there are specific tweaks in uclibc for your machine Nov 06 21:17:06 denix: right, ok thanks for the help Nov 06 21:17:54 denix: I've used it on x86 for years, never complained Nov 06 21:17:57 but ymmv Nov 06 21:19:06 CosmicPenguin: plus it's less supported by mainstream distros Nov 06 21:19:17 CosmicPenguin: what's your host distro? Nov 06 21:19:38 I used OE Nov 06 21:20:10 CosmicPenguin: ha, that's clever... :) Nov 06 21:24:15 CosmicPenguin, hi Nov 06 21:24:37 Hi Nov 06 21:27:05 * * OE Bug 4811 has been created by  Nov 06 21:27:07 * * directfb-1.2.3-autobuild Nov 06 21:27:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4811 Nov 06 21:29:40 is there a conference these days? it's been quiet lately... Nov 06 21:33:03 if there is a machine configuration then it seems like a fairly safe bet that someone has built an image at least once. Nov 06 21:33:07 the git logs should show you who checked in that file, and I would have thought that this would give you a good place to start your enquiries. Nov 06 21:35:59 pb_: sorry, I'm not following you Nov 06 21:45:39 hmm Nov 06 21:45:41 | /bin/bash /usr/share/build-1/libtool --silent --mode=compile ccache i486-linux-gcc -march=i486 -isystem/home/wweng/tmp/staging/i486-linux/usr/include -fexpensive-optimizations -fomit-frame-pointer -frename-registers -O2 -DHAVE_CONFIG_H -DLINUX=2 -D_REENTRANT -D_GNU_SOURCE -isystem/home/wweng/tmp/staging/i486-linux/usr/include -I/home/wweng/tmp/work/i486-linux/apr-util-1.2.12-r2/apr-util-1.2.1 Nov 06 21:45:43 2/include -I/home/wweng/tmp/work/i486-linux/apr-util-1.2.12-r2/apr-util-1.2.12/include/private -I/home/wweng/tmp/staging/i486-linux/usr/include -I/home/wweng/tmp/staging/i486-linux/usr/include -o buckets/apr_buckets_pipe.lo -c buckets/apr_buckets_pipe.c && touch buckets/apr_buckets_pipe.lo Nov 06 21:46:05 Does anyone know why it is looking for libtool in /usr/share directory? Nov 06 21:46:26 Not in the one in work/... directory... Nov 06 21:47:10 SirManiac: sorry for the delay, long meeting Nov 06 21:47:35 SirManiac: the hardware is basially identical to the gta-02 with the exception of it having a 320x240 lcd Nov 06 21:48:22 prpplague: ok, just had some basic curiosity, no problem Nov 06 21:49:57 hmm, say if I have package.1.1.2 and package 1.1.9 in mypackage/ directory, how do I get to build the one with lower version #? Nov 06 22:00:26 hubar: you can specify PREFERRED_VERSION_pkgname = "1.1.2" in your config file Nov 06 22:02:25 hmm, so if I am trying to compile apr-util 1.2.7, I should export PREFERRED_VERSION_apr-util="1.2.7"? Nov 06 22:02:33 hubar: or if you need one version always take precedence over another, you can set DEFAULT_PREFERENCE to raise or lower it Nov 06 22:03:50 that's bitbake var, not sure environment will be visible unless you add it to white list Nov 06 22:05:06 ahh. alright! Nov 06 22:18:34 geez, do people even test their packages before submitting??? Nov 06 22:19:20 It is not that hard to test your stuff if you wrote it urself! Nov 06 22:19:27 hubar mostly Nov 06 22:19:37 but you cant test all combinations Nov 06 22:20:01 indeed. thats why tinderboxes are helpful Nov 06 22:20:03 I am so pissed, I spent days on apache2, and spend most of today on apr-util, none of them really passed. Nov 06 22:20:09 You make a lot of assumptions. Nov 06 22:21:02 hubar as I said earlier the apache und apr suckz Nov 06 22:21:10 its shamless abuse auf autotools Nov 06 22:21:18 hmm where can I see whether the latest build of a package succeeds or not? Nov 06 22:21:23 +buildsystem Nov 06 22:21:24 But perhaps a fundamental assumption that is not quite true would be the assumption that just because a given bb recipe built at one time correctly, it will continue to do so as other recipes or even bb or OE changes. Nov 06 22:21:54 woglinde: I have no choice! I have to build apache. (or 2) Nov 06 22:22:17 hubar why Nov 06 22:22:26 why cant you use lighthttp? Nov 06 22:22:35 mwester: thats why i firmly believe that we need proper maintainers of specific packages. to -keep- it functioning, and shepherd its patches upstream, make sure no gross hacks stay around forever, etc Nov 06 22:22:43 mwester: i think itd do a great deal for our stability Nov 06 22:23:00 there's more to it than that Nov 06 22:23:32 woglinde: long story short, we need to run the apache so our customized script can work. Nov 06 22:23:57 Well, if lighthttp takes an apache2 module. Nov 06 22:23:58 customized script? Nov 06 22:24:08 the assumption you mention wouldnt be a problem if someone kept an eye on their areas of responsibility going forward. that coupled with a capable build server running builds on multiple archs, and i think we'd be sitting a lot better Nov 06 22:24:09 what module you need? Nov 06 22:24:39 we have our own module to interact between apache and our scripting engine. Nov 06 22:24:50 The struggle involves conflicting goals -- to freeze a released firmware with a compatible set of packages would make it easy to shepard everything else along. But that does little for the users of those releases, who expect that the latest updates to various apps be updated. Nov 06 22:25:04 hubar hm okay Nov 06 22:25:21 well, a stable branch helps, but most companies end up maintaining their own repositories anyway Nov 06 22:25:35 and we're short on manpower to really give the assignments I'm talking about even for one branch, much less two Nov 06 22:25:38 sadly Nov 06 22:26:04 Right - therein lies the conflict. A stable branch fixes a lot, but manpower expended on maintaining a stable branch is wasted. Nov 06 22:26:37 So the best compromise seems to be to tie the packages to the dev branch, and re-release when it becomes too difficult. Nov 06 22:26:45 That is exactly where SlugOS is right now. Nov 06 22:27:45 i think itd help to have separate repositories, one per package, with the metadata actually in the source tree, and our changes as topic branches. itd be a lot easier to manage deviations from the main repository when maintaining your own more limited reopsitory that way, imo Nov 06 22:27:54 hmm Nov 06 22:28:15 [X]Spot: you are installing the files into the wrong place Nov 06 22:28:25 Why does anyone commit broken codes anyway? Nov 06 22:28:34 hubar: it wasn't broken. Nov 06 22:28:53 In all probability it is now broken because something else somewhere changed, and that package wasn't updated. Nov 06 22:29:00 pb_: he's long since gone Nov 06 22:29:36 hubar: or otherwise it may be broken for your specific configuration Nov 06 22:30:04 pb_ I fixed your glib atomic patch *g* Nov 06 22:30:06 * * OE Bug 4812 has been created by  Nov 06 22:30:08 * * wpa-supplicant-0.5.10-autobuild Nov 06 22:30:10 * * http://bugs.openembedded.net/show_bug.cgi?id=4812 Nov 06 22:30:15 well, I guess. Nov 06 22:30:34 hubar: its not easy to test for every possible target platform, not to mention there could still be the odd case of a package thats inadvertantly affected by what packages were built before it. if it enables/disables features based on a search of the includedir rather than being explicitly set in the .bb like it should be.. Nov 06 22:30:42 hubar maybee on the weekend I have some time to look into apache2 Nov 06 22:31:35 woglinde: thanks! Nov 06 22:32:18 woglinde: since you are looking into glibc, can you please commit #4798, as khem seems busy lately... Nov 06 22:33:33 weird, two identical setup on different hosts. One builds shared-mime-info-native_0.51 without trouble, the other one fails with missing gio from glib Nov 06 22:33:35 | /usr/bin/ld: cannot find -lgio-2.0 Nov 06 22:33:42 Anybody seen this? Nov 06 22:33:58 * stefan_schmidt still works out the trouble with the broken mime-update-database Nov 06 22:37:12 glibio Nov 06 22:37:13 hm Nov 06 22:37:37 denix hm I know glibc guru Nov 06 22:37:42 aeh I am not Nov 06 22:37:58 denix hopefully pb has some time for it Nov 06 22:38:33 stefan libglib2.0-0: /usr/lib/libgio-2.0.so.0 Nov 06 22:39:08 woglinde: yeah, gio should be part of glib Nov 06 22:39:19 woglinde: It works fine on my workstation but not on the buildhost Nov 06 22:39:31 That's what confuses me Nov 06 22:39:53 The systems even have the same host glib version Nov 06 22:40:02 Not that this should matter Nov 06 22:41:05 woglinde: ok, I'll wait. I'm just trying to do little bit of pushing, as bunch of my patches sit in bugzilla w/o resolution :) Nov 06 22:41:23 denix yes Nov 06 22:41:41 denix unfornalty we dont have laibsch anymore for looking into these patches Nov 06 22:45:09 woglinde: #oe irc and oe-dev ml been unusually quiet lately... mass travel/vacations? :) Nov 06 22:50:40 denix hm the winterterm started in germany so we have to do our assigments Nov 06 22:54:33 woglinde: aha, student life... :) Nov 06 23:08:28 we should have a bug weekend again soon Nov 06 23:08:41 Crofton hm yes Nov 06 23:08:43 preferably when I am at home and my wife is out of town .... Nov 06 23:08:48 haha Nov 06 23:14:08 denix do you have the same error building apache? Nov 06 23:14:10 configure: error: Size of "void *" is less than size of "long" Nov 06 23:17:22 woglinde: I haven't tried building apache recently Nov 06 23:17:28 args Nov 06 23:17:29 sorry Nov 06 23:17:32 I meant hubar Nov 06 23:17:46 woglinde: no. Nov 06 23:18:07 hubar whats your error? Nov 06 23:18:13 woglinde: My error is from chartables.c, in "pcre_internal.h". Nov 06 23:18:58 | In file included from pcre.c:470: Nov 06 23:19:00 | chartables.c:21:27: error: pcre_internal.h: No such file or directory Nov 06 23:19:25 My targetting machine is x86. Nov 06 23:19:40 hubar: do you have pcre built? Nov 06 23:19:59 yeah seems the dependency is missing Nov 06 23:20:35 denix, yes, but it is looking for pcre that is in its srclib. Nov 06 23:23:00 hubar: hmm, does it come with it's own copy of pcre? Nov 06 23:24:54 denix: Yeah it seems so Nov 06 23:25:07 denix: apache2/srclib/pcre/ Nov 06 23:28:06 afk, need to have lunch. :) Nov 06 23:43:06 * * OE Bug 4813 has been created by  Nov 06 23:43:08 * * gamin-0.1.8-autobuild Nov 06 23:43:09 * * http://bugs.openembedded.net/show_bug.cgi?id=4813 Nov 07 00:22:16 good nite Nov 07 00:22:36 gn Nov 07 00:23:14 ~lart apache developers not thinking about crosscompiling Nov 07 00:23:14 * ibot pours gasoline all over apache developers not thinking about crosscompiling, ignites the fire, and then enjoys some toasty marshmallows with the glorious blaze Nov 07 00:23:27 haha Nov 07 00:23:30 right ibot Nov 07 01:25:46 * denix is away: I'm not here **** ENDING LOGGING AT Fri Nov 07 02:59:57 2008