**** BEGIN LOGGING AT Sun Oct 16 02:59:57 2011 Oct 16 03:00:26 khem: tiring boring Oct 16 03:01:35 khem: apologies for not able to meet. I kind of expected this to be a very hectic trip, as it was last minute Oct 16 03:01:59 jetlag => coding :) Oct 16 03:47:35 likewise: yes. heh Oct 16 09:47:01 heyho Oct 16 09:47:13 anyone here has trouble with building glib-2.0-native from oe-core? Oct 16 09:47:24 my build always fails with "libglib-2.0.so: undefined reference to `qsort_r" Oct 16 09:48:06 full log ist here: http://amethyst.openembedded.net/~morphis/log.do_compile Oct 16 10:01:25 have you tried without that ccache? Oct 16 10:02:38 JaMa|Off: no still not Oct 16 10:02:51 but I am going to try without Oct 16 10:06:22 first I will try removing my .ccache directory Oct 16 10:28:30 JaMa|Off: how can I disable ccache? Oct 16 10:31:47 morphis: add CCACHE = "" to your local.conf I think Oct 16 10:31:48 as CCACHE="" bbrenew glib-2.0-native does not solve the problem Oct 16 10:32:07 ok Oct 16 10:32:11 bluelightning: thanks Oct 16 10:32:13 CCACHE isn't in BB_ENV_EXTRAWHITE so it won't be pulled in from the external environment Oct 16 10:32:21 that's why your example didn't work Oct 16 10:33:03 bluelightning: good to know Oct 16 10:33:19 so lets hope it works now Oct 16 10:34:21 no, problem is still there Oct 16 10:39:30 ccache is probably not the cause then Oct 16 10:50:41 and on gentoo it was iirc controlled by PATH Oct 16 10:50:52 ccache in PATH before gcc&co Oct 16 11:11:01 morphis, in what lib qsort_r is supposed to be? Oct 16 11:39:24 morphis, suspend/resume works flawlessly on fso2-console-image Oct 16 11:40:54 oops wrong channel Oct 16 12:21:16 GNUtoo: qsort_r should be part of libc Oct 16 12:24:35 seems like my host libc is too old Oct 16 13:04:33 ok Oct 16 13:04:57 hi gnutoo Oct 16 13:05:05 hi Oct 16 13:16:37 captainigloo, hi Oct 16 13:57:19 GNUtoo: hi Oct 16 13:57:54 captainigloo, did you succeed with java? Oct 16 13:58:12 GNUtoo: build did not work Oct 16 13:58:35 did binaries work? Oct 16 13:58:36 hm? Oct 16 13:58:38 but finally i'm running java with angstrom packages Oct 16 13:59:01 build with oe-core did not work Oct 16 13:59:06 ah ok, so we won't see java in oe-core yet Oct 16 13:59:16 in oe-core? Oct 16 13:59:18 no Oct 16 13:59:20 extra layer Oct 16 13:59:25 yes I mean that Oct 16 13:59:36 by oe-core I meant the broad thing Oct 16 13:59:44 only if i found the time to have something working in my free time ;) Oct 16 13:59:52 ok Oct 16 14:00:29 but as usual when you find a way they don't care it's the right way Oct 16 14:00:40 ok Oct 16 14:01:02 ? Oct 16 14:01:19 woglinde, basically captainigloo had to have java working for a client Oct 16 14:01:25 i mean my boss Oct 16 14:01:35 so he didn't succeed at importing java in an oe-core layer Oct 16 14:01:42 I told him to contact your company Oct 16 14:01:54 but at the end he used angstrom binaries.... Oct 16 14:02:59 s/your company/tarrent/ Oct 16 15:01:12 morphis: what kind of host distro do u have Oct 16 15:02:07 hi khem Oct 16 15:02:13 woglinde: hello Oct 16 15:02:45 I was watching an indian movie shot in Freiburg. Remembering old times :) Oct 16 15:03:05 uhm Oct 16 15:06:38 woglinde: how are you Oct 16 15:06:42 going to ELCE Oct 16 15:16:40 yes I am going to elce Oct 16 15:32:19 khem: debian lenny Oct 16 15:34:36 uhm Oct 16 15:34:39 still lenny? Oct 16 15:34:44 thats sad Oct 16 16:20:08 morphis: where is ur config.log Oct 16 16:20:34 that function is pretty old should be there in lenny Oct 16 16:33:56 khem: sadly I already removed the tmp dir Oct 16 16:34:15 ok Oct 16 16:34:27 woglinde: yes, it's the amethyst and I did talked to mickeyl about updating it to a newer version Oct 16 16:34:41 khem: I will tell you when I am ready again Oct 16 16:45:21 morphis I would did it quick Oct 16 17:39:28 woglinde: for sure Oct 16 17:58:34 khem: full work dir of glib-2.0-native: http://amethyst.openembedded.net/~morphis/glib-2.0-native/ Oct 16 18:18:52 morphis: Dont apply remove.test.for.qsort_r.patch Oct 16 18:19:01 ah Oct 16 18:19:12 didn't looked at the patches Oct 16 18:19:35 morphis: that patch should not be applied to native recipe Oct 16 18:19:46 its ok for target recipe though Oct 16 18:19:48 iirc this patch was executing arm binary on host so should be disabled for target only Oct 16 18:20:12 but worked fine here with native, I guess because I have qsort_r Oct 16 18:20:27 hm Oct 16 18:20:42 so we should patch the glib-2.0 recipe to not apply the patch for the native version Oct 16 18:20:51 yes Oct 16 18:20:57 is it oe-core ? Oct 16 18:21:00 or classic oe Oct 16 18:21:04 oe-core Oct 16 18:21:13 yes oe-core Oct 16 18:21:25 yes test it Oct 16 18:21:30 ok Oct 16 18:21:33 will do Oct 16 18:21:34 and if it works then propose it for oe-core Oct 16 18:25:16 it seems to work Oct 16 18:25:23 compile needs much more time Oct 16 18:25:52 khem: thanks for poiting this out and I will provide a patch for oe-core later Oct 16 18:28:14 khem: it works Oct 16 19:51:47 GNUtoo: re lightmediascanner: I have not pushed 0.4.4 yet. Oct 16 19:52:03 I am doing that right now. So please wait ten minutes. Oct 16 19:54:29 ok Oct 16 19:55:16 pespin, ^^^ Oct 16 20:28:29 03Paul Menzel  07master * r63d3969b0f 10openembedded.git/recipes/lightmediascanner/lightmediascanner_0.4.4.bb: Oct 16 20:28:29 lightmediascanner: Add version 0.4.4 Oct 16 20:28:29 1. The release announcement [1] has some more information about the current state. Oct 16 20:28:29 2. The “official” Web site listed also in the recipe is not current anymore because the credentials have been lost. Oct 16 20:28:29 3. This recipe is tested using `angstrom-2010.x` and `minimal{,-uclibc}` for `MACHINE = "beagleboard"`. Oct 16 20:28:29 [1] http://blog.gustavobarbieri.com.br/2011/08/19/ann-lightmediascanner-0-4-4-released/ Oct 16 20:28:30 Signed-off-by: Paul Menzel Oct 16 20:35:14 GNUtoo: Oct 16 20:36:00 thanks Oct 16 20:36:08 will you port it to oe-core? Oct 16 20:36:37 GNUtoo: No. I have not migrated yet. I am sorry. Oct 16 20:37:09 ok Oct 16 20:37:10 But you should be able to copy it, add the license check sums and remove `PE = "1"`. Oct 16 20:37:17 yes Oct 16 20:37:33 I've to nack my previous patch then Oct 16 20:37:48 pespin, ^^^ Oct 16 20:38:01 pespin, could you do it? Oct 16 20:38:51 GNUtoo, I'll try :P Oct 16 20:39:07 thanks a lot Oct 16 20:42:12 PaulePanter, should I add only "copied from (oe.dev) plus add license check sums and remove PE" ? Oct 16 20:42:49 … and test it. ;-) Oct 16 20:43:03 I do not see any other conflicts, so that should be fine. Oct 16 20:43:48 pespin: If you could configure Git to know your real name that would be nice to. (`git config user.name "name here"`) Oct 16 20:44:01 sure :) Oct 16 20:44:13 And last thing. There was a vote to not call it oe.dev but oe-classic. Oct 16 20:49:52 ok :) Oct 16 20:52:21 pespin: Thanks for doing that. Oct 16 21:05:40 PaulePanter, is it ok to send only with [PATCH] prefix in header? Oct 16 21:05:51 Subject: [PATCH] lightmediascanner: Add version 0.4.4 Oct 16 21:07:01 btw, it builded fine here :) Oct 16 21:25:05 pespin: If it is for meta-oe you should tag it as meta-oe. Oct 16 21:25:43 [PATCH] [meta-oe]. Oct 16 21:26:04 Although looking at the last messages Koen and Andreas do not do that anymore it seems. Oct 16 21:26:16 s/builded/built/ Oct 16 21:26:42 Please use: … Add version 0.4.4 (initial recipe) Oct 16 21:27:10 And add also what distribution you built the recipe with. Oct 16 21:27:16 Sorry and thanks for your work. Oct 16 21:32:58 PaulePanter, git format-patch --subject-prefix="PATCH][meta-oe" -1 then? Oct 16 21:38:08 pespin: I guess so. Oct 16 21:39:47 PaulePanter, sent Oct 17 01:49:02 hello. I know that arago != OE but still they share somehting in common ... I am trying to build ANY arago image, but I always keep getting a task list of around 5000 tasks which fails for example on "mesa" even though me wanting just a console image. Currently I am doing "bitbake -v console-image" Oct 17 01:49:43 since this is a an OE script-set, what could cause this? Why are so many tasks being processed? Oct 17 01:49:54 (I need arago due to TISDK) **** ENDING LOGGING AT Mon Oct 17 02:59:56 2011