**** BEGIN LOGGING AT Thu Sep 10 02:59:57 2009 Sep 10 03:01:07 hi can someone please help me out with bitbaking ortp? Sep 10 03:50:17 whitefox: what is the error Sep 10 04:43:02 03Graham Gower  07org.openembedded.dev * rec1461ce16 10openembedded.git/ (3 files in 3 dirs): (log message trimmed) Sep 10 04:43:02 glibc: Build glibc 2.9 with glibc-ports 2.9 Sep 10 04:43:02 This combination of glibc-2.9 with glibc-ports-2.7 fails to build on mipsel. Sep 10 04:43:02 The reason we have been using glibc-ports 2.7 was that glibc-ports 2.9 was Sep 10 04:43:04 not tagged or we didn't see it for various reasons. Sep 10 04:43:06 Build glibc 2.9 with glibc-port 2.9 and drop the three ARM patches as Sep 10 04:43:08 they have been applied to the upstream glibc-ports tree: Sep 10 05:36:58 khem: I am still getting the glib.h not found error Sep 10 05:37:13 khem: i have glib-2.0 and glib-1.2 added to ortp's dependency Sep 10 05:37:50 I have both glibc and glib-2.0 in DEPENDS Sep 10 05:46:46 whitefox : how are you calling bitbake to build the recipe? Sep 10 05:47:09 with '-b ortp.bb'? Sep 10 07:10:17 mornink Sep 10 07:25:51 good morning Sep 10 07:28:22 hello Sep 10 07:29:06 goodbye Sep 10 07:29:08 :P Sep 10 07:40:20 was there anyone else with sheeva plug here? I guess hrw has one? Sep 10 07:40:31 * czr is trying to find some distributor within EU for them Sep 10 07:58:12 czr: AFAIK only DigiKey have them Sep 10 08:06:15 hmm, according to this globalscane and ionics should have them too. http://www.plugcomputer.org/index.php/resources/plug-providers Sep 10 08:06:24 at least the globalscale webshop lists them properly Sep 10 08:06:34 but it's in the US, and I'm in EU, so the customs would bite me. Sep 10 08:06:50 that's why I'm asking Sep 10 08:07:18 can't find the plugs on digikey though Sep 10 08:07:44 czr: sorry I was mistaken with Beagle Sep 10 08:07:52 ah, np Sep 10 08:08:16 czr: Sheeva are sold only by Globalscale Sep 10 08:08:23 * czr nods Sep 10 08:08:44 czr: where are you from? Sep 10 08:09:09 finland. Sep 10 08:11:05 m4t: i did try -b but still does not work Sep 10 08:11:15 i am having problem with bitbaking ortp Sep 10 08:11:19 could someone please help me? Sep 10 08:11:25 I am getting error Sep 10 08:24:04 hi all Sep 10 08:24:34 installed vsftpd, how can I share all of the rootfs? Sep 10 08:30:42 sx AFAIK it shoud work, ensure you have not set this #chroot_list_enable=YES Sep 10 08:31:05 see config file /etc/vsftpd.conf Sep 10 08:48:24 mckoan, netstat -a | grep ftp returns this Sep 10 08:48:30 tcp 0 0 (null):ftp (null):* LISTEN Sep 10 08:48:53 mckoan, the (null) must be * Sep 10 09:02:54 Hi, I have few patches for latest xorg stuff for OE including xserver_xorg_git, which could be usefull now with forthcoming release of 1.7.. is someone interested? Sep 10 09:04:48 sx on my Ubuntu desktop is Sep 10 09:04:58 tcp 0 0 *:ftp *:* LISTEN Sep 10 09:05:23 do you need my /etc/vsftpd.conf ? Sep 10 09:05:47 mckoan, may be can help Sep 10 09:06:13 or I think, in angstrom the structure is different, for example there is no inetd.conf file Sep 10 09:06:32 sx may well be Sep 10 09:06:53 hi could someone please help me with bitbaking ortp? Sep 10 09:07:19 whitefox: perhaps, but only if you explain what problem you are having Sep 10 09:07:46 this is the standard vsftpd.conf coming with Ubuntu 8.04 http://pastebin.com/m441fc82c happy hacking ;-) Sep 10 09:08:44 thanks mckoan Sep 10 09:09:16 JaMa: please send your patches to the mailing list Sep 10 09:10:28 florian: good morning Sep 10 09:10:43 sx: you're welcome :-D Sep 10 09:12:56 pb__: I've also patch for bumping xorg revisions based on latest releases in http://xorg.freedesktop.org/releases/individual/, should I send it too to contrib directory? Sep 10 09:15:21 I am getting glib.h not found Sep 10 09:15:30 when I try bitbaking ortp Sep 10 09:15:42 hi pb Sep 10 09:15:49 I don't understand why and I aready have glib-2.0 in DEPENDS Sep 10 09:15:55 could someone please help me ? Sep 10 09:18:10 do you have glib.h in your staging directory? Sep 10 09:18:12 whats ortp? Sep 10 09:18:21 if no, something has gone wrong with your glib-2.0 build. Sep 10 09:18:28 pb__ how do I check? Sep 10 09:18:35 if yes, either ortp doesn't actually want glib 2.0 (maybe it wants 1.2), or it is using the wrong cflags. Sep 10 09:18:37 pb__ I also tried bitbake glib 1.2 and glib 2 Sep 10 09:18:40 and it is ok Sep 10 09:18:44 pb__ hm then he had an eraler error Sep 10 09:18:53 whitefox: just look inside tmp/staging, see if you can find a file with that name. Sep 10 09:18:57 I will look at the recipe Sep 10 09:19:00 i tried to put both 2.0 and 1.2 seperately in the DEPENDS line Sep 10 09:19:34 woglinde: I'm not quite sure what an eraler error would be. Sep 10 09:20:12 i have glib-2.0 inside tmp/staging/armv7a-angstrom-linux-gnueabi/usr/lib/ Sep 10 09:20:31 woglinde it is an opensource rtp library Sep 10 09:20:41 that's a good start. do you have the corresponding header files in the include directory? Sep 10 09:21:17 no Sep 10 09:21:22 it only have glibconfig.h Sep 10 09:21:33 I think you're looking in the wrong include directory Sep 10 09:21:44 you need to look in tmp/staging/armv7a-angstrom-linux-gnueabi/usr/include Sep 10 09:21:49 not tmp/staging/armv7a-angstrom-linux-gnueabi/usr/lib/glib-2.0/include/ Sep 10 09:22:25 inside that dir, I have glib-2.0 and glib-1.2 Sep 10 09:22:38 both have the glib.h file Sep 10 09:22:40 hm new version is otu anyway Sep 10 09:22:52 we should update the recipe Sep 10 09:24:58 hm Sep 10 09:25:06 the ubuntu deps only has libc Sep 10 09:26:46 hm ah Sep 10 09:26:56 there are the requirements Sep 10 09:27:06 so the recipe is wrong anyway Sep 10 09:27:13 how to fix this problem? Sep 10 09:27:19 i have the same problem with centos Sep 10 09:27:22 wonder why ubuntu dont have them Sep 10 09:27:35 args Sep 10 09:27:42 seems I am not awake Sep 10 09:28:07 I looked at the linphone one Sep 10 09:31:24 *sigh* another not so fine configure.ac Sep 10 09:31:33 own test for lipthread Sep 10 09:32:30 hm there is no need for glib.h Sep 10 09:32:41 could you pleas pastebin the complete error Sep 10 09:36:52 whitefox still there? Sep 10 09:39:09 whitefox: okay, so it sounds like ortp is failing to add the right include paths to its CFLAGS. Sep 10 09:39:29 pb I grep ortp Sep 10 09:39:37 there is now sign for glib.h include Sep 10 09:47:23 mckoan, it works!!! thanks so much Sep 10 09:48:19 <_whitefox> woglinde: how can I fix that? Sep 10 09:48:22 <_whitefox> i need it badly Sep 10 09:48:53 whitefox as I said Sep 10 09:48:58 pastebin the whole error Sep 10 09:49:04 bitbake first Sep 10 09:50:17 ~pastebin Sep 10 09:50:18 [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste Sep 10 10:03:04 <_whitefox> woglinde: http://pastebin.com/m2e6e9922 Sep 10 10:03:32 <_whitefox> woglinde: please take a look Sep 10 10:05:42 _whitefox uh Sep 10 10:05:48 does someone know where the ${OECMAKE_C_COMPILER} variable is defined (and how it is defined) ? Sep 10 10:05:49 you are using the stone age version Sep 10 10:06:03 whitefox are you on stable branch? Sep 10 10:07:06 please use at least version 0.13.1 Sep 10 10:07:38 <_whitefox> woglinde: where do I modify to update to the latest? Sep 10 10:08:15 <_whitefox> when I do git pull, I get org.openembedded.dev -> origin/org.openembedded.dev Sep 10 10:08:21 <_whitefox> i thought it is in dev Sep 10 10:08:40 yes it is Sep 10 10:08:52 and version 0.13.1 is the latest recipe Sep 10 10:09:01 <_whitefox> 0.13.1 for ortp? Sep 10 10:09:16 yes Sep 10 10:09:21 <_whitefox> woglinde: how do I sync up with the latest recipe? Sep 10 10:09:26 dont know why its building 0.7.1 Sep 10 10:09:33 03Sebastian Krzyszkowiak  07shr/import * r99f2dd4a72 10openembedded.git/recipes/ffalarms/ffalarms_0.2.2.bb: ffalarms: move openmoko-alsa-scenarios to RSUGGESTS. Fixes building of images for other machines than om-gta01/02 Sep 10 10:09:34 which has the glib depends line Sep 10 10:09:44 03Michael  07shr/import * r0778286a10 10openembedded.git/recipes/gtk-webcore/ (files/ua-iphone.patch midori_0.1.2.bb): ua-iphone.patch: add forgotten blank. Sep 10 10:09:45 03Thomas Zimmermann  07shr/import * rbbd086156c 10openembedded.git/ (4 files in 3 dirs): Fixed reciepes for midori 0.1.8, webkit-efl and reciepe for webkit-gtk 1.1.4 Sep 10 10:09:45 03Sebastian Krzyszkowiak  07shr/import * r2f01e26717 10openembedded.git/ (2 files in 2 dirs): packagekit: bump to 0.5.2 Sep 10 10:09:57 hm Sep 10 10:10:02 <_whitefox> how can I fix that? Sep 10 10:10:32 hm maybee angstroem is setting the version Sep 10 10:10:34 let me see Sep 10 10:10:58 ah yes Sep 10 10:11:05 okay edit your local.conf Sep 10 10:11:20 and put PREFFERED_VERSION_orts = "0.13.1" in it Sep 10 10:12:35 okay lunch now Sep 10 10:13:17 <_whitefox> i am lost Sep 10 10:14:41 ????? Sep 10 10:15:04 <_whitefox> i am doing it now Sep 10 10:15:13 <_whitefox> pls wait Sep 10 10:15:17 nope Sep 10 10:15:22 I have to go Sep 10 10:15:23 till later Sep 10 10:16:26 <_whitefox> there is no such line in the local.con, i just add it there right? Sep 10 10:17:02 <_whitefox> pleaset take alook at pastebin.com/made0alf Sep 10 10:17:43 <_whitefox> is that the way to do it? Sep 10 10:25:23 ortp-0.13.1 builds for me in dev Sep 10 10:25:33 do a git status and look for the line Sep 10 10:25:47 # On branch org.openembedded.dev Sep 10 10:29:39 <_whitefox> can someone please help me out with not being able to bitbake ortp? Sep 10 10:37:39 _whitefox, you really should be running .dev .... Sep 10 10:38:06 your description of the problem implies you are running stable Sep 10 10:38:50 also, 0.7.1 builds on stbale for me Sep 10 10:42:24 _whitefox, try bitbake glib-2.0, then rebuild ortp Sep 10 10:51:02 bb -c stage myrecipe fails with "Task do_stage does not exist" but it clearly has a do_stage() what could be wrong? Sep 10 10:58:05 Stecchino: do_stage() isn't a task. Sep 10 10:58:11 try "-c populate_staging" Sep 10 10:58:41 pb__: should that be run during a rebuild? Sep 10 10:58:56 what do you mean by a rebuild, exactly? Sep 10 10:59:03 if you cleaned and then built again, yes, it should. Sep 10 10:59:06 -c rebuild Sep 10 10:59:33 yes, that seems to have the same effect Sep 10 10:59:46 specifically I'm struggeling with dbus-binding-generator that is supposed to be installed by dbus-glib Sep 10 11:00:26 had an old version that fails to generate the bindings for networkmanager 0.7 but the new one just in't installed it seems Sep 10 11:03:41 pb__: figured out it's in the -dev package Sep 10 11:04:02 to hasty while reading the bb Sep 10 11:04:04 ah right Sep 10 11:06:20 pb__: it's in the FILES_${PN}-dev, yet that is not installed in populate_staging. do i have to run ipkg install dbus-glib-dev on the host? Sep 10 11:08:54 Stecchino: no, that wouldn't help since the generated binary won't run on the host. Sep 10 11:09:00 I think you should probably be looking at dbus-glib-native Sep 10 11:09:29 I thought so, but I don't see where dbus-binding-generator is build then Sep 10 11:09:56 hm, yes, it does seem to be missing altogether. Sep 10 11:10:00 I wonder where your old one came from. Sep 10 11:10:48 oh, maybe the regular "make install" in the native package installs it. Sep 10 11:10:51 re Sep 10 11:11:05 probably easiest to try it and see what happens Sep 10 11:12:06 man, I should properly eat before trying to figure out a problem. Just to edgy when hungry. Thanks for the help pb__ Sep 10 11:22:32 good morning Sep 10 11:26:42 i get the error "evdev.c(EVIOCGBIT): Suspicious buffer size 511, limiting output to 64 bytes". how can i avoid it? Sep 10 11:30:17 did you see http://userweb.kernel.org/~dtor/eviocgbit-bug.html? Sep 10 11:36:03 yes i did - it says the problem lies with the application Sep 10 11:36:27 <_whitefox> pb__ Sep 10 11:36:38 <_whitefox> I am still having problem with bitbaking ortp Sep 10 11:36:41 <_whitefox> can someone please help me? Sep 10 11:36:55 but the issue here is the same kernel some how works with android; im not sure how it works Sep 10 11:37:18 i mean the same version of the kernel is working for someone else not for me Sep 10 11:37:33 <_whitefox> woglinde: i am still stuck Sep 10 11:37:46 _whitefox, I have posted several suggestions Sep 10 11:37:53 _whitefox: did you act on what I suggested before? Sep 10 11:37:54 try those before asking for help again Sep 10 11:38:54 pb_, do you know much about ortp? it seems to have gine from the linphone site? Sep 10 11:39:21 no, I don't know much about it Sep 10 11:39:31 I haven't used linphone in years. Sep 10 11:39:42 ok, I just did a little extra digging Sep 10 11:40:16 <_whitefox> pb__ yes Sep 10 11:40:22 <_whitefox> i added glib-2.0 in DEPENDS Sep 10 11:40:38 whitefox: okay, so it sounds like ortp is failing to add the right include paths to its CFLAGS. Sep 10 11:40:49 <_whitefox> pb__ oh that one Sep 10 11:40:58 <_whitefox> how do I add teh include path? Sep 10 11:41:11 pkg-config --cflags glib-2.0, or some such Sep 10 11:41:18 <_whitefox> where do I add it? Sep 10 11:41:32 probably in the configure script, or the makefile Sep 10 11:41:36 it depends on how ortp is built Sep 10 11:43:09 <_whitefox> pb__ where can i locate ortp's make file? Sep 10 11:43:52 er, in its source directory I imagine Sep 10 11:44:21 <_whitefox> pb__ is it under /tmp/staging/armv7a-angstrom-linux-gnuebi? Sep 10 11:44:40 no, it'd be in tmp/work/.../ Sep 10 11:45:07 <_whitefox> I get the Make file now Sep 10 11:45:11 pb_, ortp builds for me in stable and dev Sep 10 11:45:27 oh, right. that is a bit odd. Sep 10 11:45:32 are you using the same version as _whitefox? Sep 10 11:45:36 I suspect glib-2.0 was built by something else Sep 10 11:45:47 _whitefox, I can get an aswer to what he isusing Sep 10 11:46:12 I would bitbake glib-2.0, then bitbake ortp Sep 10 11:46:16 and undo all edits Sep 10 11:46:44 once people start hacking at metadata it gets very difficult to fix things Sep 10 11:47:53 if the goal is use linphone, the version in OE looks old Sep 10 11:50:37 <_whitefox> Crofton|work: no I want to use openbts Sep 10 11:50:40 03Thomas Zimmermann  07shr/import * rc9f0d00516 10openembedded.git/ (2 files in 2 dirs): Update eve to revision 41151, any newer version needs webkit-efl > 0.1 Sep 10 11:50:43 <_whitefox> but I am already using the latest Sep 10 11:50:46 <_whitefox> git pull Sep 10 11:51:12 _whitefox, run git status Sep 10 11:51:37 I think you are running stable, and should be running dev to use the openbts stuff I sent out Sep 10 11:51:57 <_whitefox> where do I change to use the dev? Sep 10 11:52:24 <_whitefox> Crofton: git status shows branch stable/2009 Sep 10 11:52:39 ok Sep 10 11:52:56 well, you need to checkout the dev branch Sep 10 11:53:07 <_whitefox> just do git checkout? Sep 10 11:53:16 and go over your setup to change the need vars in local.conf and friends Sep 10 11:53:27 I forget the exact syntax Sep 10 11:55:45 <_whitefox> is it the DISTRO? Sep 10 11:55:55 <_whitefox> instead of using angstrom-2008.1? Sep 10 11:58:04 review: http://wiki.openembedded.net/index.php/Getting_started Sep 10 12:01:56 <_whitefox> crofton: that is the page I always refer to, but I can't find any reference to how to change between dev or stable branch Sep 10 12:02:08 <_whitefox> it just says it is better to use stable for most usage Sep 10 12:02:42 openbts is not most usage Sep 10 12:02:47 <_whitefox> yes Sep 10 12:02:50 infortunately Sep 10 12:03:09 when they have the source fully online, I will invest some time in ot Sep 10 12:03:11 <_whitefox> but how to change my OE from stable to dev? Sep 10 12:03:52 <_whitefox> does anyone know how to change OE from stable to dev? Sep 10 12:03:53 _whitefox, based on what you know, the best thing is to reinstall using the getting start instruction Sep 10 12:05:17 whitefox yeah check out new Sep 10 12:05:23 even git is so slow Sep 10 12:05:35 or is it faster after the switching? Sep 10 12:05:55 <_whitefox> yes I am using the getting start instruction on the site you just pointed me to Sep 10 12:06:00 <_whitefox> it is already build with that Sep 10 12:09:52 <_whitefox> woglinde: what do you mean by checking out new? Sep 10 12:10:07 <_whitefox> woglinde: i just got this new env set up this morning Sep 10 12:10:53 whitefox what is git branch -a telling you? Sep 10 12:11:05 03Thomas Zimmermann  07shr/import * rb16d35a15b 10openembedded.git/recipes/gtk-webcore/midori_0.1.8.bb: Fix midori reciepe Sep 10 12:11:54 <_whitefox> woglinde: it says stable/2009 Sep 10 12:12:40 so why do you thin stable is dev? Sep 10 12:14:26 <_whitefox> when I do git pull, it says openembedded.dev Sep 10 12:14:29 <_whitefox> so i thought... Sep 10 12:14:44 <_whitefox> how can I change that? what is the command that I can use to pull from dev? Sep 10 12:16:10 read getting started Sep 10 12:16:15 and the gitphrasebook Sep 10 12:16:36 learning curve has to be high Sep 10 12:18:42 jo raster Sep 10 12:19:08 woglinde: woggie! Sep 10 12:19:10 yo yo Sep 10 12:23:47 leave people, leave, the big pixel is coming... Sep 10 12:26:16 stefan hu Sep 10 12:46:27 lalalaaalaaa Sep 10 12:47:33 * stefan_schmidt sneaks behind woglinde .... buuuuuh Sep 10 12:50:14 * czr is too busy bisecting to leave Sep 10 12:50:55 hmm, big pixel does not react on his new nickname. yo raster Sep 10 12:51:03 is it just me or do more than half of builds between proper releases fail with biset? Sep 10 12:51:05 bisect even. Sep 10 12:51:48 * stefan_schmidt does not expect OE being bisec friendly. Sep 10 12:52:53 ah yes. wasn't talking about OE, but kernel, sry :-) Sep 10 12:54:14 heh Sep 10 12:55:18 trying to track down the commit that causes by hair to turn gray. Sep 10 12:55:24 s/by/my/. Sep 10 13:07:07 czr hehe Sep 10 13:07:23 but kernel bisect should be working fine Sep 10 13:07:41 yes. except that I get kernels that don't even build half of the time :-) Sep 10 13:07:51 so it's pretty difficult to test whether they're "good" or "bad" :-) Sep 10 13:07:58 but I found git bisect skip, my new friend. Sep 10 13:08:13 plus I have out-of-tree patches to apply after each bisect step. Sep 10 13:08:20 hm that could automate Sep 10 13:08:27 skip when kernel build fails Sep 10 13:09:01 I guess so. doing everything by hand now though. 9 iterations to go! woohoo! Sep 10 13:09:15 hm and skip if the commit is not a driver/foo commit Sep 10 13:10:25 gm Sep 10 13:10:40 that's a bit too risky. given that arm and mach bases keep changing quite often Sep 10 13:11:00 what's the proper way to remove psplash from an angstrom image? Using the technique recommended in the angstrom-2008.1.conf file (BAD_RECOMMENDATIONS) doesn't appear to work Sep 10 13:12:35 hrw|gone: 0024-Xserver-... doesn't like to apply Sep 10 13:20:18 03Koen Kooi  07org.openembedded.dev * r3a8e482ccc 10openembedded.git/recipes/quake/quake2_svn.bb: quake2: fix QA errors Sep 10 13:22:02 bye Sep 10 14:01:53 hi all Sep 10 14:10:54 hi gremlin[it] Sep 10 14:11:01 hi mickeyl ! Sep 10 14:11:03 florian: when will anonymous git work again? Sep 10 14:11:48 mickeyl: should work again... fixed it half an hour ago i believe Sep 10 14:12:27 ah, cool, thanks Sep 10 14:24:03 hi gremlin[it] :-) Sep 10 14:31:58 hi mckoan ! Sep 10 15:00:49 hello Sep 10 15:02:15 I want to run linux on an arm920T device but i have no idea how to do this Sep 10 15:27:17 anyone working in backporting the qt improvements to stable? Sep 10 15:34:15 03ghost  07org.openembedded.dreambox * ref8b83ccdf 10openembedded.git/packages/gstreamer/ (2 files in 2 dirs): gst-plugins-base_0.10.24.bb: replace own hotfix with fix from gstreamer team Sep 10 15:34:16 03ghost  07org.openembedded.dreambox * r193ec78216 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git://git.openembedded.net/openembedded into org.openembedded.dreambox Sep 10 15:34:17 03ghost  07org.openembedded.dreambox * r83eb51ead4 10openembedded.git/packages/gstreamer/ (7 files in 4 dirs): packages/gstreamer: update plugins-good and plugins-bad, move patches to right folder, mpegpsdemux speedup Sep 10 15:34:18 03ghost  07org.openembedded.dreambox * r0038144fc7 10openembedded.git/packages/dreambox/dreambox-dvb-modules.bb: (log message trimmed) Sep 10 15:34:21 dreambox-dvb-modules.bb: update dm800 and dm8000 drivers Sep 10 15:34:23 fix 1080p24/30 synchronisation problems via AV Receiver Sep 10 15:34:25 fix 720p24 output via dvi Sep 10 15:34:27 fix loss of dvi(hdmi) audio on framerate change Sep 10 15:34:29 another small deinterlacer auto handling fix Sep 10 15:34:31 improved locking times for alps-bsbe1-c/d01a tuner Sep 10 15:34:33 03Felix Domke  07org.openembedded.dreambox * r965c6de181 10openembedded.git/: Sep 10 15:34:36 Merge branch 'org.openembedded.dreambox' of git://git.opendreambox.org/git/ghost/opendreambox-1.5 into org.openembedded.dreambox Sep 10 15:34:58 Conflicts: Sep 10 15:34:58 packages/gstreamer/gst-plugins-bad_0.10.14.bb Sep 10 15:34:58 03ghost  07org.openembedded.dreambox * r8b63e1bb9e 10openembedded.git/packages/gstreamer/gst-plugin-dvbmediasink.bb: gst-plugin-dvbmediasink.bb: push srcdate Sep 10 15:35:00 03ghost  07org.openembedded.dreambox * r69fbca2a85 10openembedded.git/: Merge branch 'org.openembedded.dreambox' of git://git.openembedded.net/openembedded into org.openembedded.dreambox Sep 10 16:01:39 kergoth: you were right, the RP fixes I need have already been committed in poky Sep 10 16:02:26 03Felix Domke  07org.openembedded.dreambox * r3b1e6e131f 10openembedded.git/packages/gstreamer/gst-plugins-bad_0.10.14.bb: gst-plugins-bad: add PR Sep 10 16:03:13 RP: when/if you happen to have 5 mins pls sync packaged-staging.bbclass with poky Sep 10 16:06:58 RP: atually in OE we are at rev 4861 + couple of patches (and the one about OELAYOUT_ABI is different) Sep 10 16:07:18 03Felix Domke  07org.openembedded.dreambox * r2992f11abf 10openembedded.git/packages/gstreamer/gst-plugins-bad_0.10.14.bb: gst-plugins-bad: fix git merge breakage Sep 10 16:19:49 mickeyl: what changes has FSO made to Vala? Sep 10 16:20:16 mickeyl: would "error: D-Bus methods must almost throw DBus.Error" be one of them? Sep 10 16:23:04 * RP wonders whether anyone has played with the canadian classes much Sep 10 16:23:19 looks rather strange in places :/ Sep 10 16:24:41 :-) Sep 10 16:24:56 kergoth: did you look at the git autopv stuff? Sep 10 16:25:06 haven't had the time yet Sep 10 16:25:24 kergoth: oh ok Sep 10 16:25:26 * kergoth has quite a bit on his plate at work this sprint Sep 10 16:31:37 RP: ah, you're there, not yet frozen on silicon... Sep 10 16:32:00 cbrake: yes, that's among it Sep 10 16:32:17 cbrake: these changes are submitted for upstream inclusion Sep 10 16:32:28 so you can expect them to show up soon Sep 10 16:33:17 mickeyl: great, thanks Sep 10 16:33:48 np, sorry if you have to change your programs, but proper error reporting was completely missing in Vala until our patches Sep 10 16:33:57 (dbus error reporting, that is) Sep 10 16:34:06 mickeyl: you're forcing me to clean up my code !! Sep 10 16:34:12 hehe Sep 10 16:34:14 you'll notice mor Sep 10 16:34:14 e Sep 10 16:34:23 default constructors no longer autogenerated Sep 10 16:34:28 struct members all public Sep 10 16:34:34 and more fun :) Sep 10 16:34:47 vala 0.7.x is much more strict Sep 10 16:34:49 (which is a good thing) Sep 10 16:38:21 mickeyl: should that be: "D-Bus methods must _always_ throw DBus.Error" instead of almost? Sep 10 16:38:54 * broonie wonders how you almost throw an exception :) Sep 10 16:38:54 yes, ptitjes has spelling problems :D Sep 10 16:39:10 :-) Sep 10 16:39:27 grrrrrrrrrrrrr Sep 10 16:39:29 :p Sep 10 16:39:30 hehe Sep 10 16:40:03 mickeyl: heh I believe you have the committer rights on that Sep 10 16:40:06 :p Sep 10 16:40:34 indeed, i'm just scared to mess up the repository with our dozens of branches ;) Sep 10 16:41:34 héhé lazy man Sep 10 16:41:51 I leave ciao Sep 10 16:41:54 cu Sep 10 16:41:54 I call that word dyslexia -- I think our typing skills are trained to the point where we type if words in instead of letters. So instread of typing the wrong letter, we type the wrong word Sep 10 16:42:13 yep Sep 10 16:42:27 ^^^^^ (case in point, my last sentence) :-\ Sep 10 16:42:35 LOL Sep 10 16:42:38 heh Sep 10 16:46:15 mickeyl: origin/origin/org.openembedded.dev that branch seems bogus Sep 10 16:49:19 re Sep 10 16:50:23 khem: heh, indeed Sep 10 16:50:34 some folks need to remember that when pushing like that Sep 10 16:50:36 foo:bar Sep 10 16:50:44 jojojo Sep 10 16:50:44 bar is not supposed to include origin Sep 10 16:50:57 i'll delete that branch Sep 10 16:52:20 done Sep 10 16:52:32 mickeyl is git now faster? Sep 10 16:52:40 than what/when? Sep 10 16:52:45 12k Sep 10 16:53:18 hmm, you speak in riddles Sep 10 16:53:33 please formulate a sentence :) Sep 10 16:53:53 hm last times when a used pull after a longer time I only get a update speed of 12kb Sep 10 16:54:04 ah Sep 10 16:54:16 well, that always depends Sep 10 16:54:24 OSUSL has good bandwidth Sep 10 16:54:30 but we're on a virtual machine Sep 10 16:54:35 so I dont want know how long it will be last when I woulf checkout from scratch Sep 10 16:54:36 so at times it can be crowded Sep 10 16:54:43 okay Sep 10 16:55:45 03Koen Kooi  07org.openembedded.dev * rdd5c666d21 10openembedded.git/recipes/ogre/freeimage_3.11.0.bb: freeimage: fix QA errors Sep 10 17:03:00 woglinde: where are you? Sep 10 17:04:31 cbrake germany berlin Sep 10 17:04:53 2 mbit download Sep 10 17:49:55 pb_: ping Sep 10 17:50:13 jo kergoth Sep 10 17:50:31 hey woglinde Sep 10 17:50:34 how's it going? Sep 10 17:50:53 mickeyl: there is org.openembedded.stable and then stable/2009 Sep 10 17:51:10 I think only one is enough Sep 10 17:51:33 we should really drop the silly org.openembedded. thing at some point. move dev to master, .. Sep 10 17:51:39 just monotone remnants, no need for it Sep 10 17:51:40 woglinde: do you still use woglinge/gettext Sep 10 17:51:46 yeah I was to suggest that next Sep 10 17:51:59 moving to master is better in git terminology Sep 10 17:52:09 khem at least there was one package which had problem with gnome stublib Sep 10 17:52:18 but cannt remember which Sep 10 17:52:21 khem: and saves us some typing :) Sep 10 17:52:31 khem ah the branch Sep 10 17:52:34 sorry Sep 10 17:52:54 khem there some patches not merged Sep 10 17:53:03 will this making top point Sep 10 17:54:06 okay lunch Sep 10 17:54:14 kergoth: agree though I have bash completions so doesnt matter Sep 10 17:54:18 hehe Sep 10 17:54:23 well, git completions can sometimes be slow Sep 10 17:54:35 if they're trying to figure out what exactly they should complete Sep 10 17:54:43 i.e. git log can accept a path, or a committish Sep 10 17:54:51 right Sep 10 17:55:35 i really want to make the zsh completion slightly smarter Sep 10 17:55:48 so itd know that completion after the -- is always a path Sep 10 17:55:52 would really speed that up Sep 10 17:57:50 * khem is content with bash Sep 10 18:16:19 pb_: http://gist.github.com/184705 - i think this re-raise is appropriate, so callers of exec_func know what exceptions they need to catch. does this patch seem sane to you? Sep 10 18:16:24 or, for that matter, anyone else Sep 10 18:16:25 mickeyl Sep 10 18:16:27 ? Sep 10 18:17:42 we need to do something to hide these ugly tracebacks from users who don't need to see it. maybe we should just silence the traceback if debugging isn't enabled Sep 10 18:18:24 * kergoth wades through bitbake code to see if there's a more appropriate place Sep 10 18:19:23 oh! right, just need to tweak better_compile and better_exec instead Sep 10 18:19:25 perhaps Sep 10 18:54:41 * kergoth plays with sys.excepthook Sep 10 19:06:56 argh...glibc_2.9.bb, do_package) Sep 10 19:06:57 NOTE: preparing tree for binary locale generation Sep 10 19:06:57 NOTE: generating locale de_DE (UTF-8) Sep 10 19:06:57 FATAL: kernel too old Sep 10 19:12:36 dont use glibc Sep 10 19:16:08 ant__, the recent checkin messed something up with binarylocale generatoin Sep 10 19:16:12 turn them off :) Sep 10 19:16:31 of course, if you ned that stuff, it is annoying Sep 10 19:29:31 Crofton|work: must have been that glibc-ports 2.9 Sep 10 19:30:19 anyone run into this issue with package_ipk and locales? Sep 10 19:30:20 ERROR: Exception:exceptions.OSError Message:[Errno 2] No such file or directory: '/home/clarson/oe/projects/34871/tmp/work/i686-mv-linux/hello-2.3-r0/install/hello-locale-th' Sep 10 19:30:32 multiple locales before that one worked fine Sep 10 19:30:53 this one gives a traceback and aborts the task Sep 10 19:31:41 hmmm Sep 10 19:53:25 03Khem Raj  07org.openembedded.dev * r5a881d0d78 10openembedded.git/ (7 files in 2 dirs): (log message trimmed) Sep 10 19:53:25 cherokee: Add new version 0.99.24 Sep 10 19:53:25 * Use INC_PR Sep 10 19:53:25 * Add new recipe for 0.99.24 Sep 10 19:53:25 * Add --with-mysql configure option to point Sep 10 19:53:27 to staging area. Otherwise with build machines Sep 10 19:53:29 having local installation of mysql-dev, cherokee Sep 10 20:11:30 there are some packages like binutils which provide there own md5.sum do we use that Sep 10 20:11:57 khem: what do you mean? Sep 10 20:12:11 ftp://mirrors.kernel.org/sources.redhat.com/binutils/snapshots/md5.sum Sep 10 20:12:41 just use this file to authenticate the tar file Sep 10 20:13:04 you'd have to put those in checksums.ini Sep 10 20:13:11 don't know of any converter or anything Sep 10 20:13:46 hmm it would make possible to use weekly snapshots Sep 10 20:14:03 otherwise its a pain Sep 10 20:14:26 write a converter, prepend to do_fetch :) Sep 10 20:14:31 can i disable checksum for certain recipes Sep 10 20:15:03 well, i know you can set OE_STRICT_CHECKSUMS = "" in a recipe to make it not abort if they don't match, i'm sure you could set that other var to disable it entirely for it Sep 10 20:15:31 OE_STRICT_CHECKSUMS is build wide I thought Sep 10 20:16:25 it's just a variable. Sep 10 20:16:25 do_fetch obeys it Sep 10 20:16:25 it doesn't care where it gets it from Sep 10 20:16:47 oh so I can override it for just one recipe Sep 10 20:16:50 thats cool Sep 10 20:18:18 hmmm but if file name remains same and the remote file contents changed and I have a local tar downloaded fetcher is not going to know if this source was updates at source so I should redownload it Sep 10 20:18:34 so anyway idea of using snapshot is bad Sep 10 20:19:20 has anyone had issues with python 2.6? Sep 10 20:19:48 jeaneus: not me. I infact have it working on ARM/thumb mode Sep 10 20:21:25 i am getting an error with bitbake-- http://pastebin.com/d239f1c3f Sep 10 20:22:20 yikes Sep 10 20:23:47 re Sep 10 20:24:11 any ideas? I was trying to build libgles-omap3_3.00.00.06.bb for a zoom2, but I get that error... Sep 10 20:24:51 jeaneus which error? Sep 10 20:25:08 kergoth: this short diff is the leap btw poky and oe wrt packaged-staging.bbclass http://fr.pastebin.ca/1561430 Sep 10 20:25:47 woglinde: -- http://pastebin.com/d239f1c3f Sep 10 20:25:54 ant__: 404 Sep 10 20:26:00 ^_^ Sep 10 20:26:28 jeaneus please paste complete output Sep 10 20:27:31 I think its a case of use before definition Sep 10 20:34:17 woglinde: I think I found the issue.. Let me try something first.. Sep 10 20:38:26 re pb Sep 10 20:39:09 jeaneus: os.popen Sep 10 20:56:30 hi woglinde Sep 10 20:56:36 * pb____ returns from the local beer festival Sep 10 20:57:08 nice Sep 10 20:57:11 pb haha Sep 10 20:57:24 and your wife didnt hurt you? Sep 10 20:57:32 no, my wife was there too Sep 10 20:57:41 hehe okay Sep 10 20:57:44 and my daughter, for that matter, though she didn't actually have any beer Sep 10 20:57:56 okay, i have a patch to use sys.excepthook to hide the python tracebacks from the uncaught exceptions raised by python tasks, etc, unless you ran with bitbake -D Sep 10 20:58:18 kergoth: sounds nice Sep 10 20:58:31 * kergoth thinks its a lot less ugly that way Sep 10 20:58:44 besides, what good is the traceback.. no shit, it ran exec_func... Sep 10 20:58:48 :) Sep 10 20:59:26 heh, right. it'd be nice to find a way to massage the traceback spew into a form that's usable for debugging. Sep 10 21:00:09 the traceback is some sort of linked list or something, we should be able to change what subset of that stack is displayed Sep 10 21:00:45 not sure where that massaging would be best located, though. perhaps within better_exec or exec_func, massage then push back up Sep 10 21:01:31 sakoman_, what is the latest version of u-boot-omap3? i have "2009.08" (from git's master), but a recipe i have used to add "rc2" Sep 10 21:03:50 mrc3_: I'm still using 2009.06 so you are ahead of me :-) Sep 10 21:04:37 Been working on other things and haven't had a real need to move to the latest (whatever it is!) Sep 10 21:07:24 sakoman_, i see. so where would i get this lastname of "rc2" from? Sep 10 21:07:33 i'm testing x-load/u-boot/kernel on overo and beagleboard Sep 10 21:07:55 but don't know what to put in the recipe for "rc" Sep 10 21:08:22 rc2 just meant tha it was a release candidate, if what you are building is final, then there is no rc-anything! Sep 10 21:08:47 so you could just remove it completely Sep 10 21:09:14 sakoman_, will do! don't know where that rc ever came from Sep 10 21:09:15 thanks! Sep 10 21:21:43 khem: os.popen? Sep 10 21:22:19 sakoman_, fwiw, i'll try to get sakoman.net/git/ out of the x-load recipe Sep 10 21:22:22 i have one for poky now: http://git.pokylinux.org/cgit.cgi/poky-contrib/commit/?h=mrchapp/beaglemm&id=5fba64cc1e030ce32bd9b48f148e900a0025f0bd Sep 10 21:22:58 mrc3: thanks :-) Sep 10 21:25:13 has anyone encountered an issue when building gpe-bootstrap where "strip" reports an "unrecognized format" error? Sep 10 21:30:26 kg4ysn: most probably your install is messed Sep 10 21:30:46 hm new directfb Sep 10 21:30:50 are u using coreutils-native or not Sep 10 21:31:57 khem: i think so. should i try removing it? Sep 10 21:32:11 woglinde: no u should use it Sep 10 21:32:31 ah Sep 10 21:33:33 hm I should consider sending the tslib configure patch upstream Sep 10 21:34:36 woglinde: yes thats always a good deed Sep 10 21:35:50 what patch is that? Sep 10 21:36:14 directfb-1.4.1/ts_lib_autotools.patch Sep 10 21:36:25 ah, cool Sep 10 21:36:30 kergoth did you follow the qt/tslib problem? Sep 10 21:37:04 as someone already posted on the thread, the linking back to libts from the modules is already done in the tslib svn Sep 10 21:37:10 just the OE recipe(s) need updating, probably Sep 10 21:37:28 hehe Sep 10 21:37:30 do it Sep 10 21:37:32 *g* Sep 10 21:37:52 * kergoth doesn't even use a device that has a touchscreen anymore Sep 10 21:38:09 hm Sep 10 21:38:31 by the way you could turn tslib into git repo Sep 10 21:38:31 well, i might have a Z in the closet Sep 10 21:38:35 i should do something with one Sep 10 21:38:39 yeah, not a bad idea Sep 10 21:38:44 i'm using git-svn for it anyway :P Sep 10 21:38:48 lol Sep 10 21:38:52 (hence the signed svn commits) Sep 10 21:38:58 berlios has git support Sep 10 21:39:09 they do? huh Sep 10 21:39:31 yeah since half a year or so Sep 10 21:39:56 cool Sep 10 21:40:02 yupp Sep 10 21:40:18 but I dont know if they offer automatic transition Sep 10 21:40:31 i'm pretty good at conversions now Sep 10 21:40:50 i did like 5 grafts and a pile of filter-branches to get the bitbake git repo where i wanted it Sep 10 21:49:02 khem gcc-4.4.1 for minimal or micro? Sep 10 21:49:06 or both Sep 10 21:51:18 I tested it on minimal Sep 10 21:51:30 probably its same for micro too Sep 10 21:51:45 woglinde: takes 4 days to run gcc tests on osk Sep 10 21:51:51 lol Sep 10 21:52:09 so I need something faster Sep 10 21:52:23 how do we connect beagle to LAN Sep 10 21:52:50 usb/ethernet think? Sep 10 21:53:18 hmm usb ports I have plenty but no ethernet Sep 10 21:53:28 or buy an iegd plaform v2 if it will come out? Sep 10 21:53:32 or sheeva plug Sep 10 21:54:24 or use qemu :-} Sep 10 21:54:29 hihi Sep 10 21:54:41 i should do something with my sheevaplug Sep 10 21:54:43 qemu does not catch unaligned accesses Sep 10 21:54:46 it's collected a fair bit of dust since i bought it Sep 10 21:54:47 is qemu scoping with multiple cpu? Sep 10 21:56:05 kergoth: donate it to me if you do not use it :) Sep 10 21:56:46 heh, i dont often use the linux box i have at home.. or rather i do, but usually remotely, its off in another room.. just need to take the time to hook the thing up and see about setting up nfsroot and all Sep 10 21:56:52 havent gotten around to it Sep 10 21:58:08 pb____: Did you see my gcc results I posted Sep 10 21:58:34 khem: no Sep 10 21:59:34 pb____: ok have a look and if you think gcc 4.4.1 is ok for micro let me know Sep 10 21:59:58 I tested weirdest combination i.e. arm/thumb/uclibc Sep 10 22:06:37 it does look as though visibility is a bit broken. it'd be nice to know what's going on there, and/or whether those tests passed with gcc 4.2.x. Sep 10 22:06:41 other than that it looks fine to me, yeah. Sep 10 22:06:59 I intend to find about visibility Sep 10 22:07:16 I think it could be because I was using binutils 2.18 Sep 10 22:07:31 but then I saw that these tests only check the asm output Sep 10 22:07:40 so binutils is out of blame game Sep 10 22:09:37 ah nice disko now runs on beagle Sep 10 22:10:45 uh and they have support for gstreamer_ti cool Sep 10 22:13:55 args Sep 10 22:14:02 glibc compile now gives me Sep 10 22:14:03 FATAL: kernel too old Sep 10 22:18:11 use --enable-kernel=2.6.0 Sep 10 22:18:43 yes yes Sep 10 22:19:51 or simply use eglibc :) Sep 10 22:21:38 ah right Sep 10 22:21:39 hm Sep 10 22:22:03 ah, that's what I did and now I see what bites dth :-) Sep 10 22:39:32 pb____: I know now why the visibility tests were failing Sep 10 22:39:44 because I did not have cross-objdump in my path Sep 10 22:40:02 so it did not find arm-oe-linux-gnueabi-objdump Sep 10 22:40:55 pb____: I think this need to be passed when configuring gcc with GCC_OBJDUMP_FOR_TARGET Sep 10 22:43:40 ah right, okay Sep 10 22:48:31 hm uclibc/locales dont works Sep 10 22:57:42 pb____: I have seemingly fixed it in gcc cross recipe I will test it locally by just running visibility tests Sep 10 22:58:00 woglinde: totally kaput Sep 10 22:58:04 khem yes Sep 10 22:58:16 woglinde: ugh Sep 10 22:58:16 khem could you look at uclibc-initial again Sep 10 22:58:24 it compiles the whole uclibc Sep 10 22:58:29 instead of libc.so Sep 10 22:59:19 I think some install hooks make this Sep 10 23:00:40 hmm Sep 10 23:00:45 lemme look Sep 10 23:03:38 how do I get past the glibc_2.9.bb error? I tried using 2.7 but did not help.... Sep 10 23:04:26 use newer kernel headers Sep 10 23:04:28 or eglibc Sep 10 23:04:53 woglinde: do_stage function is defined twice once in uclibc.inc and second time in uclibc-initial Sep 10 23:05:00 which one should bitbake use Sep 10 23:05:10 I think it choses first Sep 10 23:05:23 that ends up building whole stuff Sep 10 23:05:52 * khem pokes kergoth Sep 10 23:06:36 khem uh Sep 10 23:06:39 nice catch Sep 10 23:06:45 hm Sep 10 23:06:54 let me see Sep 10 23:06:57 the run files Sep 10 23:07:02 that depends on the parse order. if i set FOO="bar" and then after that, FOO="baz", FOO will be baz. do_stage is no different Sep 10 23:07:05 heh Sep 10 23:07:54 run.do_stage file tells me the one in uclibc-initial is picked up which is what I expected Sep 10 23:08:22 khem nope Sep 10 23:08:32 aeh I meant yes uclibc-inital one is called Sep 10 23:09:53 so either install_dev or install_runtime changed Sep 10 23:13:34 03Henning Heinold  07org.openembedded.dev * rb76c22a5c6 10openembedded.git/ (3 files in 3 dirs): directfb: update to version 1.4.2 Sep 10 23:14:47 wahahah Sep 10 23:14:56 why the disko people didnt see my recipes Sep 10 23:18:17 where can I set eglibc? Sep 10 23:18:51 jeaneus: which distro are you using Sep 10 23:19:03 angstrom Sep 10 23:19:11 if you use minimal its already used as default Sep 10 23:19:27 for angstrom I think you need to set LIBC=eglibc' Sep 10 23:19:51 where can I set that? Sep 10 23:19:56 LIBC="eglibc" in your local.conf Sep 10 23:20:08 okay thanks.... Sep 10 23:20:09 or may be LIBC_local = "eglibc" Sep 10 23:22:53 minimal is Angstrom without nay user support .... Sep 10 23:22:58 er any Sep 10 23:35:08 Crofton|work: I use minimal and micro all the time now Sep 10 23:35:08 it works pretty well Sep 10 23:35:23 I plan to bitbake world one day using them Sep 10 23:35:29 I am not far Sep 10 23:35:38 I already have x11-image building Sep 10 23:35:59 * kergoth used to do a world build every week or two back before he left, but that'd be a tad difficult now Sep 10 23:36:22 kergoth: I think it would be good to revive that Sep 10 23:36:47 and then start adding recipes as new packages get released Sep 10 23:36:57 iow be on top of everything :) Sep 10 23:38:10 it's a good way to keep yourself busy Sep 10 23:38:14 always a ton of build failures to fix Sep 10 23:38:16 hehe Sep 10 23:41:30 yeah Sep 10 23:44:43 args Sep 10 23:45:01 ansgtroem still has the git-native not strip problem Sep 11 00:08:58 pb____: So I will make gcc 4.4.1 default on micro is that ok ? Sep 11 00:11:40 args Sep 11 00:11:53 khem what was the fix for udev and uclibc ppoll? Sep 11 00:12:06 cbrake posted it to ml Sep 11 00:12:21 there is no fix in library itself Sep 11 00:12:23 glibc uses a wrapper Sep 11 00:12:30 which is not desired in uclibc Sep 11 00:12:51 * m4t 's fix was to revert to 124 w/ the gentoo patch Sep 11 00:13:19 m4t yes I could do this too Sep 11 00:13:28 but I thought there was a better fix Sep 11 00:13:34 there might be Sep 11 00:13:42 i saw someone change the syntax of how it was called Sep 11 00:13:47 different args. Sep 11 00:14:02 http://patchwork.openembedded.org/patch/1008/ Sep 11 00:14:04 also, there is a newer version of udev on kernel.org that calls the function differently as well Sep 11 00:15:04 hrm i thought it had to do with uclibc Sep 11 00:16:17 quickfix now is use older udev Sep 11 00:16:33 woglinde i can pastebin you the patch Sep 11 00:16:33 otherwise I have to rebuild uclibc Sep 11 00:16:34 sec. patch Sep 11 00:16:41 uclibc will implement the function if you provide a kernel with syscall wired Sep 11 00:16:52 khem I know the script Sep 11 00:16:55 I fixed it Sep 11 00:17:16 if you remeber Sep 11 00:18:09 khem hmm okay, so how does uclibc determine compile time if its implemented? kernel headers? Sep 11 00:18:50 m4t uclibc needs kernelheader to compiles Sep 11 00:19:01 woglinde thanks Sep 11 00:19:03 and there is a script which greps all syscalls from the headers Sep 11 00:19:05 woglinde: install -d ${CROSS_DIR}/${TARGET_SYS} Sep 11 00:19:11 this is the problem Sep 11 00:19:23 khem uclibc-inital? Sep 11 00:19:25 it invoked the compile because that the deps of install Sep 11 00:19:27 yes Sep 11 00:19:32 looking at 0.9.30.1 Sep 11 00:20:03 m4t: it uses the linux-libc-headers to determine what syscalls are wired Sep 11 00:20:28 so its important that you use same or newer kernel which you used for linux-libc-headers Sep 11 00:20:32 khem????? Sep 11 00:20:50 that has nothing to do with uclibc compilation Sep 11 00:21:19 woglinde: why not Sep 11 00:21:36 let me verify Sep 11 00:21:48 install is only wraper for cp and mkdir Sep 11 00:22:12 on unix Sep 11 00:22:34 on windows it might be mdir and copy Sep 11 00:24:24 hmm another thing i see Sep 11 00:24:27 pregen install_dev Sep 11 00:24:34 we only need install_headers here Sep 11 00:25:36 woglinde: http://khem.pastey.net/125007 Sep 11 00:26:02 this one should fix the problem Sep 11 00:26:08 khem and I dont like the compiling in do_stage Sep 11 00:26:14 do_compile should compile Sep 11 00:26:18 do_stage should copy Sep 11 00:26:44 woglinde: there is nothing to compile actually Sep 11 00:26:53 it just does dummy stuff Sep 11 00:27:15 khem the .o stuff is compiled by mysteries wizrads? Sep 11 00:27:53 well I can change that if you like Sep 11 00:28:01 jepp Sep 11 00:28:04 please do Sep 11 00:28:18 otherwise the seperation makes no sense Sep 11 00:31:52 woglinde: how about this http://khem.pastey.net/125008 Sep 11 00:32:37 args Sep 11 00:32:47 do_compile again in do_stage Sep 11 00:33:27 and where is the header_install gone? Sep 11 00:33:36 or dont need gcc the headers? Sep 11 00:34:54 http://khem.pastey.net/125009 Sep 11 00:34:58 complete diff :) Sep 11 00:35:48 hm Sep 11 00:35:58 is the install_header needed before compiling? Sep 11 00:36:18 yes Sep 11 00:36:34 hm did you say me this? Sep 11 00:37:24 guess I should look what install_headers really does Sep 11 00:37:43 it installs the headers which are needed to build gcc Sep 11 00:38:00 the order here does not matter Sep 11 00:38:05 where do you want to put them Sep 11 00:38:16 they should be installed into staging area Sep 11 00:38:17 so it has to go to do_stage Sep 11 00:38:23 *g* Sep 11 00:38:35 I think we now already have cleaned it Sep 11 00:38:37 oh well let me do that Sep 11 00:39:01 do we really need CC="${CC}" for install_headers? Sep 11 00:39:33 I think some shells did not pass it on Sep 11 00:39:38 I will leave it in there for now Sep 11 00:41:46 I will also clean it up for othe recipes Sep 11 00:42:13 many thanks Sep 11 00:42:15 for this Sep 11 00:48:19 woglinde: thanks for being so adamant :) Sep 11 00:48:41 I like QS Sep 11 00:48:49 and the QS in this caes wasnt hard Sep 11 00:48:59 to realize Sep 11 00:49:19 hm why hal depends on the kernel Sep 11 00:49:21 for buildtime Sep 11 00:49:26 yeah with kids clinging around the chair you forget all those :) Sep 11 00:50:15 woglinde: I will commit it and you can test it for some versions if you have time Sep 11 00:50:59 khem my son learned some days ago how firefox works Sep 11 00:51:28 now he always wanted to browse to the site of "Sendung mit der Maus" Sep 11 00:54:10 hehe Sep 11 00:57:03 03Khem Raj  07org.openembedded.dev * rfae7226fa0 10openembedded.git/recipes/uclibc/ (6 files): (log message trimmed) Sep 11 00:57:03 uclibc-initial: Move the compile stuff to do_compile Sep 11 00:57:03 * Currently all work is done in do_stage. On Henning's Sep 11 00:57:03 suggestion I refactored it. Sep 11 00:57:03 * For 0.9.30 onwards installing pregen and install_dev Sep 11 00:57:05 built almost whole uclibc. We only needed libc headers Sep 11 00:57:09 so replaced with install_headers Sep 11 01:34:39 good nite Sep 11 01:34:40 03Henning Heinold  07org.openembedded.dev * rbcc5de7935 10openembedded.git/recipes/eject/eject_2.1.5.bb: eject: inherit gettext to honour USE_NLS Sep 11 01:34:41 03Henning Heinold  07org.openembedded.dev * ra74a29fb40 10openembedded.git/recipes/disko/ (4 files in 2 dirs): Sep 11 01:34:41 disko: update to newer git rev Sep 11 01:34:41 * stole some input form the disko-oe-git, but rework some stuff Sep 11 02:22:01 any particular reason that curl/libcurl are using gnutls instead of openssl? Sep 11 02:26:30 03Khem Raj  07org.openembedded.dev * r045f4e9106 10openembedded.git/conf/distro/include/sane-toolchain.inc: Sep 11 02:26:30 sane-toolchain.inc: Default to gcc 4.4.1 Sep 11 02:26:30 * ppc405 has also got tested using 4.4.1 so bump this too. Sep 11 02:26:30 * ARMv7 is also switched to use 4.4.1. Sep 11 02:26:30 Signed-off-by: Khem Raj Sep 11 02:26:32 Acked-by: Michael 'Mickey' Lauer Sep 11 02:26:36 Acked-by: Phil Blundell Sep 11 02:26:38 03Khem Raj  07org.openembedded.dev * rad856cde59 10openembedded.git/recipes/binutils/ (binutils-cross.inc binutils_cvs.bb): Sep 11 02:26:41 binutils-cross: Do not install libibery.a in wrong place Sep 11 02:26:43 * This interferes with canadian cross build of binutils Sep 11 02:26:45 because it finds cross libiberty.a where as it needs the Sep 11 02:26:47 target one. Sep 11 02:26:49 Signed-off-by: Khem Raj Sep 11 02:26:51 03Khem Raj  07org.openembedded.dev * r01407868c7 10openembedded.git/recipes/gcc/gcc-configure-cross.inc: Sep 11 02:26:54 gcc-configure-cross.inc: Export more binutils Sep 11 02:26:56 * Export AS_FOR_TARGET OBJDUMP_FOR_TARGET Sep 11 02:26:58 OBJCOPY_FOR_TARGET STRIP_FOR_TARGET Sep 11 02:27:00 Signed-off-by: Khem Raj Sep 11 02:27:02 03Khem Raj  07org.openembedded.dev * rd3fc551eb3 10openembedded.git/recipes/images/console-image.bb: Sep 11 02:27:11 console-image.bb: Require console-base-image Sep 11 02:27:13 * Get rid of code duplication. Sep 11 02:27:15 Signed-off-by: Khem Raj Sep 11 02:27:17 03Khem Raj  07org.openembedded.dev * r2aaa2e297f 10openembedded.git/recipes/gcc/ (gcc-4.4.1.inc gcc-configure-common.inc): Sep 11 02:27:20 gcc-configure-common: Dont fiddle with NATIVE_SYSTEM_HEADER_DIR Sep 11 02:27:22 * This does not work on canadian cross build of target gcc. Sep 11 02:27:26 Signed-off-by: Khem Raj Sep 11 02:27:28 03Khem Raj  07org.openembedded.dev * r8050c52cb5 10openembedded.git/recipes/images/native-sdk-image.bb: Sep 11 02:27:31 native-sdk-image.bb: Use IMAGE_BASENAME = "native-sdk-image" Sep 11 02:27:33 Signed-off-by: Khem Raj Sep 11 02:27:35 03Khem Raj  07org.openembedded.dev * r50d8ec53a4 10openembedded.git/recipes/ (images/native-sdk-console-image.bb tasks/task-native-sdk.bb): Sep 11 02:27:42 images: Rename native-sdk-console-image.bb as native-sdk-image.bb Sep 11 02:27:44 tasks: Rename task-native-sdk.bb as task-sdk-native.bb Sep 11 02:27:46 Signed-off-by: Khem Raj **** ENDING LOGGING AT Fri Sep 11 03:00:00 2009