**** BEGIN LOGGING AT Fri Jul 02 02:59:57 2010 Jul 02 03:40:22 03Khem Raj  07org.openembedded.dev * rf29ba5a65d 10openembedded.git/recipes/pam/ (libpam-1.1.1/pam-no-innetgr.patch libpam_1.1.1.bb): Jul 02 03:40:22 libpam-1.1.1: Dont use innetgr on uclibc as its not there. Jul 02 03:40:22 Signed-off-by: Khem Raj Jul 02 03:40:32 03Khem Raj  07org.openembedded.dev * r927721f66e 10openembedded.git/recipes/ (9 files in 7 dirs): Jul 02 03:40:32 recipes: Punt use of patch=1 and use striplevel=n instead of pnum Jul 02 03:40:32 Signed-off-by: Khem Raj Jul 02 03:45:57 khem : do you know where is the git access list ? Jul 02 03:46:15 i don't remember if i still have access on it Jul 02 04:02:57 sorry don't mind :) Jul 02 06:42:21 good morning Jul 02 07:00:32 $ svn co http://opkg.googlecode.com/svn/trunk Jul 02 07:00:33 svn: Server sent unexpected return value (502 Bad Gateway) in response to OPTIONS request for 'http://opkg.googlecode.com/svn/trunk' Jul 02 07:00:42 hmm... Jul 02 07:00:56 anyone else see this, or is it my isp? Jul 02 07:01:44 grg: same problem here in Italy Jul 02 07:02:04 I get same issue Jul 02 07:02:11 mckoan, khem, thanks. Must be a google problem then Jul 02 07:04:21 Hi folks, good mornin' Jul 02 07:05:12 khem, can you please commit the patches of mine that you've oked? (except maybe the opkg one). I'm not a committer. Jul 02 07:06:06 I just bitbaked empathy 2.30.2 (bitbaking all the dependancies on the way) and now when I'm trying to run it on the Beagleboard, I get the following error : "empathy: error while loading shared libraries: libdb-5.0.so: cannot open shared object file: No such file or directory" Jul 02 07:07:28 grg: sure Jul 02 07:07:39 khem, thanks! Jul 02 07:07:40 will do it tomorrow morning Jul 02 07:07:53 screwgoth, i don't see an empathy recipe for 2.30.2... Jul 02 07:07:55 now I will have to catch some sleep Jul 02 07:07:59 :) Jul 02 07:08:20 I made it locally ... Jul 02 07:08:30 g'night all Jul 02 07:08:40 cya Jul 02 07:08:55 Basically the only change I made was to remove the telepathy-gabble module, update the md5sum and sha256 sum and remove the patch file since those changes were already there in 2.30.3 Jul 02 07:09:41 screwgoth, it sounds like its added a library dependency in the newer version Jul 02 07:10:00 you probably had the staging files lying around for some reason, so the build succeeded Jul 02 07:10:31 but when you installed the ipk, the extra dependecies aren't in DEPENDS, so they dont get installed Jul 02 07:10:53 is libdb-5.0.so a berkeley db thing? Jul 02 07:11:41 hmm ... this must be a runtime depenadancy , right ? Else I would have faced issued during bitbake Jul 02 07:11:56 yeah .. I think so and I didn't see a recipe for it Jul 02 07:12:55 well, i suppose it might dlopen() in and bail out if it cant... but its more likely that the build just found it in your staging dir because something else that you've built at some point depends on the same library Jul 02 07:13:28 see if you can find it in your staging lib dir Jul 02 07:13:36 ok.. but if it's there in the staging dir, it doesn't get installed on the board ? Jul 02 07:13:44 ok .. checking it out Jul 02 07:14:10 the staging dir is a special area for build time data (header files and libraries) Jul 02 07:16:00 Well, I had added "rm_work" in my local.conf to conserve space , so would the staging directory still exist ? Where are the staging directories ? Jul 02 07:18:02 morning Jul 02 07:19:36 screwgoth, not sure. on either of those Jul 02 07:19:54 screwgoth, in any event, you need to have "db" in your DEPENDS line for your recipe. Jul 02 07:20:11 hallow hrw Jul 02 07:20:14 Well, I found that my tmp/ dir has a pstage dir, which has ipk packages prefixed with staging Jul 02 07:20:27 grg: Hmm... Lemme try a few things out ... thanks for replying Jul 02 07:21:58 screwgoth, i have dirs at tmp/sysroots/mipsel-oe-linux/usr/{lib,include} that looks suspiciously like staging directories Jul 02 07:22:46 rm_work should only remove old directories under tmp/work, not anything in staging Jul 02 07:25:40 Hey grg: I think that makes sense .. I found my libdb-5.0.so in the path you mentioned Jul 02 07:26:37 03Koen Kooi  07org.openembedded.dev * r8d6ce36505 10openembedded.git/recipes/ti/ti-dmai_2.10.00.10.bb: ti-dmai 2.10.00.10: fix damage introduced by 927721f66e7c2c5b5024ce0a17cd31ca8ccbd1d2 Jul 02 07:28:28 well, good luck. Jul 02 07:28:32 i'm off home Jul 02 07:36:17 03Martin Jansa  07org.openembedded.dev * r783112e143 10openembedded.git/recipes/qi/ (qi.inc qi_git.bb): Jul 02 07:36:17 qi: add om-gta0* overrides to build qi only for right cpu Jul 02 07:36:17 Signed-off-by: Martin Jansa Jul 02 09:16:12 hi xora Jul 02 09:16:18 hey woglinde_ Jul 02 09:20:59 hi wog Jul 02 09:22:16 hi blindman Jul 02 09:22:32 hms seems I have to differ btw. blindvt and blindman Jul 02 09:22:33 *sigh* Jul 02 09:22:40 damn irc *g* Jul 02 09:23:14 haha Jul 02 09:35:57 woglinde: hi! Jul 02 09:36:47 jo zecke Jul 02 09:42:32 morning all Jul 02 09:42:37 hi rp Jul 02 09:43:27 rp hm is it difficult to implement staging for java into new staging system? Jul 02 09:43:51 only diffrence for java is the dir where the jar's should be end up Jul 02 09:47:04 woglinde: I've not looked into java to be honest. It shouldn't be too hard though Jul 02 10:14:14 hi zecke, woglinde, rp, all Jul 02 10:15:12 hi pb Jul 02 11:08:43 * broonie stabs coreutils Jul 02 11:08:54 * woglinde stabs the heat Jul 02 11:09:09 come to vienna, twice as humid! :) Jul 02 11:10:47 do_populate_sysroot is failing whcih isn't very helpful Jul 02 11:12:46 h 74° on cpu Jul 02 11:12:48 not so good Jul 02 11:54:49 Depends on what temp scale one uses, I suppose. :) Jul 02 11:57:12 mwester the real one Jul 02 11:57:14 *g* Jul 02 11:57:25 Kelvin? Jul 02 11:57:26 imperial stuff is out dated Jul 02 11:57:27 :p Jul 02 11:57:53 Imperial is out of date, but then so am I. Jul 02 12:05:35 hm I found the old clearcase 3.2 book Jul 02 12:05:36 haha Jul 02 12:06:02 now THAT's out of date! Jul 02 12:06:32 hm gcc-4.4.4 is broken in oe Jul 02 12:07:04 libgcc_so linker script always reports to many files open Jul 02 12:07:06 hi stefan Jul 02 12:07:11 hi woglinde Jul 02 12:07:45 woglinde: The heap space problem seems fixed now with the bigger mem size Jul 02 12:08:07 woglinde: That shows us some other problems :) Jul 02 12:08:30 stefan hm? Jul 02 12:08:37 robert added cacao yesterday Jul 02 12:08:42 didnt test it so far Jul 02 12:08:47 woglinde: ah, cool Jul 02 12:09:01 woglinde: That would help the javavm segfault problem then? Jul 02 12:09:06 no Jul 02 12:09:15 woglinde: oh, I thought Jul 02 12:09:45 he enabled cacao to use with openjdk Jul 02 12:11:09 puch Jul 02 12:11:17 84° now Jul 02 12:11:30 next cooling system will be water Jul 02 12:11:38 Ice cubes work well. Jul 02 12:11:44 For a while... Jul 02 12:12:06 hm but I would need an external fan Jul 02 12:12:55 * mwester has had enough trouble with radiators in his autos, and doesn't wish to deal with those same problems in a computer! Jul 02 12:13:13 I'd prefer a better fan-based cooler, frankly. Jul 02 12:13:48 Unfortunately, my motherboard has space constraints that I did not consider when I purchased it. Jul 02 12:14:19 So overclocking has been something I've not been able to do. Jul 02 12:30:13 hi Jul 02 12:31:49 khem, i've put my current fixup patches that were used for the pastes i showed you yesterday here, fyi: uclibc.org/~aldot/uClibc/nptl/ Jul 02 12:32:02 khem, i will apply them early next week Jul 02 12:33:48 khem, i've done compile-time tests for i486, mips, ppc, arm. It would be great if you could runtime-check it if you find the time :) Jul 02 12:37:44 woglinde: stefan_schmidt: hey guys, if the the jamvm problem is related to maximum heap space then we could just increase the defaults for that VM Jul 02 12:38:06 thebohemian: yeah, woglinde did that already Jul 02 12:38:09 woglinde, stefan_schmidt: this can be tweaked in the jamvm-native/-initial recipe Jul 02 12:38:11 thebohemian: Its 1GB now Jul 02 12:38:30 err, mom Jul 02 12:38:55 thebohemian: the heap problem is fixed. The javavm segafult are something different, no? Jul 02 12:39:13 it does not complain about missing heap space on this one, just segfaults Jul 02 12:39:33 stefan_schmidt: yeah, I think that is how it normally behaves in that situations Jul 02 12:40:11 maybe it does some calloc/malloc() and does not check the return value Jul 02 12:40:14 thebohemian: Hmm, in one case we always have seen the heap space message and in the other case never Jul 02 12:40:25 hm Jul 02 12:40:44 it would be good to find a separated test case for this Jul 02 12:40:57 so we could fix the bug once and for all in jamvm itself Jul 02 12:41:08 thebohemian: sure, I just was not able to narrow it down Jul 02 12:41:23 that would be a dream :) Jul 02 12:41:37 one can make the buffers very small (few kilobytes) and then write a simple java app that over time allocates more and more of the heap Jul 02 12:42:38 MemoryWaster.java: Jul 02 12:42:38 a = new ArrayList(); Jul 02 12:42:38 while(true){ Jul 02 12:42:38 a.add(new byte[1024]); Jul 02 12:42:38 } Jul 02 12:43:20 strange thing is that something like this should actually trigger an OutOfMemoryError Jul 02 12:44:07 thebohemian: sorry, I didn Jul 02 12:44:18 thebohemian: sorry, I didn't listen... but I assume we overcommit memory Jul 02 12:44:40 thebohemian: and then the kernel will just kill you instead of having a NULL coming back from malloc Jul 02 12:46:00 zecke: but that would not make it look like a segfault right? Jul 02 12:46:20 thebohemian: no, dmesg would give you a clue Jul 02 12:46:48 zecke: hey, btw. default stacksize could also be a problem Jul 02 12:47:02 zecke: its 64kib by now Jul 02 12:47:14 thebohemian: oh yeah, thay can certainly generate a segfault Jul 02 12:47:33 zecke: i'll try something Jul 02 12:47:59 and it only happens on ARM? Jul 02 12:48:14 zecke: it happens during build Jul 02 12:48:17 03Lukas Gorris  07org.openembedded.dev * r718898c523 10openembedded.git/recipes/linux/ (linux-qsd8/htcleo/defconfig linux-qsd8_git.bb): [linux-qsd8] update srcrev, defconfig Jul 02 12:48:23 zecke: on x86_64 here Jul 02 12:48:31 zecke it happens at buildtime Jul 02 12:48:34 and randomly Jul 02 12:48:37 stefan_schmidt: ah then use valgrind, it will tell you about the stack overrun Jul 02 12:49:39 zecke: hmm, good idea Jul 02 12:49:57 -X problem is with openjdk -Xmx4096 istn -Xmx4096m Jul 02 12:50:54 and stupid thing, did you try to generate a core file? is the backtrace meaningful or in jit'ed code? Jul 02 12:51:30 was stefan task to look it up Jul 02 12:51:32 okay Jul 02 12:51:35 bye for now Jul 02 12:51:48 look what up? Jul 02 12:52:08 the segfaults Jul 02 12:52:25 from jamvm Jul 02 12:52:36 really? Jul 02 12:52:46 not? Jul 02 12:52:50 I alway thought I get a shiny all working ja thing from tarent? :) Jul 02 12:53:00 anyway, we will figure it out Jul 02 12:53:14 That is something we should be able to track down together Jul 02 12:53:26 till later maybee Jul 02 12:53:29 cu Jul 02 12:53:37 03Lukas Gorris  07org.openembedded.dev * rd9d930c486 10openembedded.git/recipes/xorg-xserver/xserver-xorg-conf/htcleo/xorg.conf: [xorg-xserver] add xorg.conf for machine htcleo Jul 02 12:53:40 I could fix GNU Smalltalk to execute Java code again. :) Jul 02 12:53:49 zecke bahahahaa Jul 02 12:56:51 zecke: so its not the stack. that is also gracefully handled Jul 02 12:57:02 recurse:32 Jul 02 12:57:02 recurse:33 Jul 02 12:57:02 recurse:34 Jul 02 12:57:02 java.lang.StackOverflowError Jul 02 12:57:19 wrote a little app that uses recursion and set the stack size very small Jul 02 12:57:28 thebohemian: ah, that is for javacode, but the vm itself could run over the stack? Jul 02 12:57:42 hm, ... Jul 02 12:57:49 zecke: that it interesting Jul 02 12:57:55 s/it/is Jul 02 12:58:02 thebohemian: is at least the same package failing during build or is it a random crash? Jul 02 12:58:20 zecke: random, of course ;) Jul 02 12:58:47 ulimit -c unlimited, and inspect the core dump? Jul 02 13:00:48 zecke: Just set it to unlimited. Now the time is to wait until it happens again Jul 02 13:01:17 stefan_schmidt: thanks for doing this.am a bit busy with other tasks,right now Jul 02 13:01:19 From all the from scratch builds I try these days I hope I should get it the next days Jul 02 13:01:45 thebohemian: sure, its mostly a thing how we communicate who is in charge Jul 02 13:02:01 and of course having someone like zecke to though in the good ideas :) Jul 02 13:02:51 stefan_schmidt: well, you should also change the proc_pattern so you will find the core file (well find will find it too) Jul 02 13:03:08 thebohemian: which kind of GC is used for Java these days? Jul 02 13:04:31 zecke: I go with find for now. I'm not root on the build boix Jul 02 13:04:33 box Jul 02 13:06:53 stefan_schmidt: one more good idea might be to enforce that jamvm-initial is being build for all kinds of debug flags Jul 02 13:07:17 thebohemian: How would I do this? Jul 02 13:07:51 stefan_schmidt: can one set custom CFLAGS per recipe? Jul 02 13:08:03 stefan_schmidt: if so, that would be the way to go Jul 02 13:08:13 stefan_schmidt: otherwise modify the recipe Jul 02 13:08:36 thebohemian: ah, you mean debug flags for building javavm Jul 02 13:08:40 misread that Jul 02 13:08:55 ok, will try this when this build run thing is through Jul 02 13:09:06 stefan_schmidt: yup, ok Jul 02 13:09:15 started builds on different machines, maybe one will give us a nice coredump Jul 02 13:09:27 will take some time though. Good for a break :) Jul 02 13:23:55 we should have -ggdb3 there right? Jul 02 13:36:44 zecke: yes Jul 02 13:42:53 hi dcordes Jul 02 13:42:55 hi Jul 02 13:43:07 so what's your issue with #oe? Jul 02 13:43:47 trying to build angstrom-gnome-image again. this time with correct angstrom-2008.1 distro instead of minimal. since angstrom uses old gcc by default and I'm building for cortex-a8 htcleo I use following in local.conf Jul 02 13:43:48 the toolchain version? Jul 02 13:43:55 PREFERRED_VERSION_gcc-cross = "4.4.5" Jul 02 13:43:55 PREFERRED_VERSION_gcc-cross-initial = "4.4.5" Jul 02 13:43:55 PREFERRED_VERSION_gcc-cross-intermediate = "4.4.5" Jul 02 13:44:00 ok Jul 02 13:44:46 now I have problem that bitbake wants gcc-cross_csl-arm-2008q3 after it's done with gcc-cross 4.4.5 Jul 02 13:44:58 csl Jul 02 13:45:02 that sound strsange Jul 02 13:45:09 that sounds like code sourcey linux Jul 02 13:45:25 maybe you did something strange Jul 02 13:45:32 like with assume provided etc... Jul 02 13:45:44 pastebin the whole local.conf Jul 02 13:46:37 dcordes: why do you set the PREFERRED_VERSIONs? Jul 02 13:46:39 ok. also noticed that expat build failes Jul 02 13:46:52 zecke: cause I Want recent gcc. do you recommend different way ? Jul 02 13:47:06 he needs it for msm8k/qsd architecture Jul 02 13:47:16 dcordes: well, you will need problem solving skills if you leave the pre-defined and tested path Jul 02 13:47:18 minimal distro select gcc-cross*_4.4.4 by default Jul 02 13:47:28 dcordes: e.g. bitbake -g should be your friend Jul 02 13:47:35 use angstrom or SHR Jul 02 13:47:40 zecke: it worked without problems before Jul 02 13:48:08 I think SHR would be better for a phone Jul 02 13:48:24 dcordes: well use bitbake -g and see who starts to pick the csl recipe Jul 02 13:48:28 also: if you want recent fso/shr there are some .inc for autorev Jul 02 13:48:29 ok Jul 02 13:48:34 good idea Jul 02 13:48:35 GNUtoo|laptop: I don't want shr Jul 02 13:48:40 ah ok Jul 02 13:49:41 after finishing something with desktop env I will build shr Jul 02 13:49:56 ok Jul 02 13:53:14 GNUtoo|laptop: did you try to compile xf86-video-msm for htcream machine in recent revisions ? Jul 02 13:53:58 "gcc-cross" [label="gcc-cross 0:4.3.2+csl-arm-2008q3-39-r1\n/media/lenny/home/luke-g/build/openembedded/recipes/gcc/gcc-cross_csl-arm-2008q3.bb"] Jul 02 13:54:08 dcordes, yes it get built at each image Jul 02 13:59:31 morning kergoth_ Jul 02 13:59:38 hey Jul 02 14:09:42 thebohemian, woglinde: double free corruption in java-initial: http://pastebin.com/XspF6RLb Jul 02 14:10:30 zecke: In your great wisdom, is this more verbose message also something that comes with ulimit -c unlimited? Jul 02 14:10:49 stefan_schmidt: cool! that is in classpath actually Jul 02 14:11:05 zecke: I never seen this before Jul 02 14:11:21 thebohemian: yeah, at least some good debug info to get us somewhere Jul 02 14:19:20 wow, the core file is 2,1GB Jul 02 14:28:12 heh, welcome to java Jul 02 14:28:39 yeah Jul 02 14:28:47 hmmm Jul 02 14:28:51 I hoped I got away from it after some courses at university Jul 02 14:28:56 :) Jul 02 14:29:00 uh oh, sounds like kergoth is thinking again Jul 02 14:29:58 indeed. be afraid Jul 02 14:30:35 dcordes, I'll go Jul 02 14:30:52 lol 2.1G of core Jul 02 14:31:29 hi kergoth_ Jul 02 14:31:37 kergoth_: Changes in bitbake master look nice :) Jul 02 14:31:55 kergoth_: Are you using that heavily? Jul 02 14:32:28 doing builds with it every day Jul 02 14:32:50 kergoth_: cool. I pushed the various bugfixes that poky built up Jul 02 14:32:55 nice Jul 02 14:36:51 kergoth_: I've also been having a long thing about the whole staging thing. I need to write an email about what I'm thinking of but the summary is allowing tasks to define a "setscene" call back which optionally can provide the output of the task. Before the runqueue does its conventional bottom up build it goes top down and looks for those functions and if they return success skips them and their dependencies from the runqueue Jul 02 14:37:38 So do_installl could provide the install output, staging would probide the staging output, write_ipk would provide the .ipk files and so on Jul 02 14:37:56 Totally granular staging packages without too much pain Jul 02 14:38:06 setscene and all that stamp rubbish goes away Jul 02 14:38:43 Combined with task based checksums... :) Jul 02 14:58:52 RP: That sounds like a reasonable plan to me. Jul 02 15:00:50 * kergoth_ tries to figure out how to handle the various COW copies of datasmart objects with regard to the value tree in the varref code -- perhaps we need a versioned value registry Jul 02 15:08:02 does anyone know if there's a recipe in oe-dev or oe-stable/2009 that provides libgudev ? Jul 02 15:08:18 i'm trying to build modemmanager from oe-dev but it's failing on configure because of the absence of that lib Jul 02 15:09:16 and grep isn't showing anything obvious Jul 02 15:16:26 jconnolly: its provided by udev Jul 02 15:16:32 ah yes i see now Jul 02 15:16:35 my udev is old Jul 02 15:16:39 will try building newer udev Jul 02 15:16:41 thx Jul 02 15:16:42 use 151 Jul 02 15:16:45 or later Jul 02 15:17:16 likewise: Dutch are playing good 1:1 :) Jul 02 15:18:56 Glad I didn't recall the game was on until it was tied up, heh Jul 02 15:19:41 * khem also uses bitbake master as primary bitbake Jul 02 15:20:03 Tartarus: yeah, now game is on Jul 02 15:20:16 Tartarus: and I predict dutch gonna make it Jul 02 15:23:47 I dunno, only 'cuz I've got a bad track record when it comes to rooting for winners :) Jul 02 15:24:57 Tartarus: there you go :) Jul 02 15:24:59 yay Jul 02 15:47:13 wow Jul 02 15:47:15 2:1 now Jul 02 15:47:20 NL Jul 02 15:50:20 and it's over! Jul 02 15:50:40 yaya Jul 02 15:52:08 wow Jul 02 16:18:54 gm Jul 02 16:30:51 This is annoying this morning, after doing a git pull and removing my dir, and firing up .... I get a checksum error on sources/gcc-4.3.3.tar.bz2 Jul 02 16:31:15 Expected MD5: 'cc3c5565fdb9ab87a05ddb106ba0bd1f' and Got: 'aff8a9b77b0edd181bb2de1ac817d28e' Jul 02 16:31:16 Expected SHA256: '309f614a3c7fee88edc4928ff17185a19533949a1642ccf776e87d86303704de' and Got: '34ca6a99f9c744abb026078cfceef3d8d891db23b6fd04e1d3242a2ad6e41ba9' Jul 02 16:32:18 robtow: delete the sources from sources and retry Jul 02 16:32:33 khem - Roger. Wilco. Jul 02 16:53:01 is a personal git tree required to submit patches to the mailing list? according to the User_contributions page it is "encouraged" Jul 02 16:53:15 <--proceeding with bitbaking the omap pm kernel :-) Jul 02 16:55:53 oneshel: sure Jul 02 16:56:07 oneshel: if you post to ml many people have eyes on it Jul 02 16:56:28 kergoth: after todays bitbake update its not honoring DISTRO= on cmdline Jul 02 16:56:32 khem: ok I'll post it there, thanks Jul 02 17:01:21 khem, bisect? :) Jul 02 17:03:48 kergoth: in progress :) Jul 02 17:03:59 I wonder if bitbake -k is fixed Jul 02 17:04:01 had to go back to 1.8 Jul 02 17:06:13 277e8e80c3e1101cdadf234ae74803b9a8d45d25 is the first bad commit Jul 02 17:06:14 commit 277e8e80c3e1101cdadf234ae74803b9a8d45d25 Jul 02 17:06:14 Author: Jeff Dike Jul 02 17:06:14 Date: Tue Jun 29 13:33:24 2010 +0100 Jul 02 17:06:15 create cooker object after cleaning the environment Jul 02 17:06:53 kergoth: to reproduce DISTR Jul 02 17:06:54 what's a "cooker object"? Jul 02 17:07:08 DISTRO=minimal MACHINE=qemuarm bitbake console-image Jul 02 17:07:23 and dont set DISTRO elsewhere in env Jul 02 17:07:58 robtow: a bitbake task handler i believe Jul 02 17:09:50 khem - ah, ok. The name actually makes sense :-) Jul 02 17:11:58 hmm it takes like 10 mins for bb parse on my second poor box Jul 02 17:16:19 re Jul 02 17:19:28 RP: if you are around then 277e8e80c3e1101cdadf234ae74803b9a8d45d25 broke something that was working before in bitbake Jul 02 17:26:00 khem could you please try gcc-4.4.4 Jul 02 17:26:06 it seems to be broken Jul 02 17:26:32 wehn liking with libgcc_so.1 linker script the linker always says to many files open Jul 02 17:28:05 kergoth: hmm so now it cleans the environment now it will not clean these vars if they are in BB_ENV_EXTRAWHITE Jul 02 17:28:17 BB_ENV_EXTRAWHITE="MACHINE DISTRO" MACHINE=qemumips DISTRO=minimal bitbake Jul 02 17:28:18 works Jul 02 17:29:55 kergoth: prolly we should add them to BB_ENV_WHITELIST by default Jul 02 17:31:10 Er, it hasn't it been like that for a long time? Jul 02 17:31:20 I know i've been putting extrawhite into my rc file for a long time now Jul 02 17:31:44 yeah but I got rid of any scripts to set vars Jul 02 17:31:50 I only have local.conf Jul 02 17:31:58 no BBPATH? Jul 02 17:32:00 and then use DISTRO and MACHINE on commandlin Jul 02 17:32:05 no bbpath Jul 02 17:32:08 I use bblayers Jul 02 17:32:20 he mickeyl Jul 02 17:32:26 khem 2:1 btw. Jul 02 17:32:30 brazil out Jul 02 17:32:45 woglinde: yeah :) I thought holland would make it Jul 02 17:32:47 and they did Jul 02 17:32:57 likewise must be drunk now Jul 02 17:33:17 woglinde: that error is strange Jul 02 17:33:23 khem indeed Jul 02 17:33:33 first I tought a linker error Jul 02 17:33:35 but wasnt Jul 02 17:33:50 gcc-4.5 works fine Jul 02 17:33:55 gcc-4.3.x worked too Jul 02 17:35:59 woglinde: I dont have any build with gcc 4.4.4 around Jul 02 17:36:47 khem okay Jul 02 17:36:54 but I will see Jul 02 17:36:59 so what in the linker script Jul 02 17:37:00 was only mentioned it Jul 02 17:37:15 is it something new ? Jul 02 17:37:16 there are normal the static gcc_libs Jul 02 17:37:25 for some fucntions Jul 02 17:37:34 like libc.so script Jul 02 17:37:41 khem, hmm, would be nice if you could set BB_ENV_EXTRAWHITE in config files.. should be doable Jul 02 17:38:02 kergoth: may be bitbake.conf is a good place Jul 02 17:38:10 kergoth: what do you think Jul 02 17:38:43 i think you missed my point. i'm pretty sure right now you can't set BB_ENV_EXTRAWHITE in a .conf at all Jul 02 17:38:50 * kergoth_ checks Jul 02 17:38:53 ah Jul 02 17:39:02 yes its an env var Jul 02 17:39:07 rght Jul 02 17:39:28 puhh now we have summer in germany Jul 02 17:39:44 woglinde: same time as California Jul 02 17:40:06 khem I consider to buy some aircondition Jul 02 17:40:26 woglinde go outside by the lake instead :) Jul 02 17:40:29 or to the see Jul 02 17:40:31 sea Jul 02 17:41:21 how many hours do you have to drive to the pacific? Jul 02 17:41:42 woglinde: an hour or so Jul 02 17:41:51 03Frans Meulenbroeks  07org.openembedded.dev * r47283d0669 10openembedded.git/recipes/tgt/tgt_1.0.5.bb: Jul 02 17:41:51 tgt: upgraded to 1.0.6 Jul 02 17:41:51 Signed-off-by: Frans Meulenbroeks Jul 02 17:41:58 woglinde: but pacific is cold all summer as the name says Jul 02 17:42:10 khem hm thats not to long Jul 02 17:42:24 but I bet you have pool for your house Jul 02 17:42:36 woglinde: no but my neighbour has Jul 02 17:43:32 hm no pool Jul 02 17:44:04 with little kids you are better off not having one Jul 02 17:44:55 hm hm Jul 02 17:55:12 hm 180 euro Jul 02 18:03:17 *sigh* Jul 02 18:03:28 my wife is general against it Jul 02 18:17:17 * mwester will sell a pool -- just have to come to Chicago to un-install it and haul it away ;) Jul 02 19:20:23 will it be ghana vs germany in final ? Jul 02 19:23:17 .nl :) Jul 02 19:23:47 or nl vs esp Jul 02 19:51:08 blindvt: it does not build for me with your patches Jul 02 19:51:17 | libc/libc_so.a(err.os): In function `_L_unlock_96': Jul 02 19:51:17 | err.c:(.text.vwarn_work+0x129): undefined reference to `__lll_unlock_wake_private' Jul 02 19:51:17 | collect2: ld returned 1 exit status Jul 02 19:51:17 | make: *** [lib/libc.so] Error 1 Jul 02 19:51:17 | FATAL: oe_runmake failed Jul 02 20:16:38 expat_2.0.1 -r3 do_unpack fails http://tinderbox.openembedded.net/packages/628405/ anybody who came across this ? Jul 02 20:39:27 03Koen Kooi  07org.openembedded.dev * r16dc19fd7c 10openembedded.git/recipes/devicekit/devicekit-power_014.bb: devicekit-power: use a source location that doesn't change, matches the old checksums Jul 02 20:41:56 dcordes: is your host a ubuntu-10.04/i386 ? Jul 02 20:42:25 khem: yes Jul 02 20:44:17 there is a bug in ubuntu Jul 02 20:44:21 gzip Jul 02 20:44:32 I am afraid you might be hitting that Jul 02 20:51:30 khem: gzip 1.3.12 Jul 02 20:51:59 search in launchpad Jul 02 20:52:12 cbrake also hit the same issue few months back Jul 02 20:56:25 I will install oe gzip in my host Jul 02 21:01:13 should we use gzip-native? :) Jul 02 21:21:00 re Jul 02 21:21:04 hm what a match Jul 02 21:29:48 Jay7: that would be nice Jul 02 21:32:43 woglinde: heh both teams lost their nerves at times Jul 02 21:32:57 anyway to build gcc-cross you should have working tar/gzip/bzip2 and native toolchain.. Jul 02 21:38:03 jo ant Jul 02 21:38:53 hey Jul 02 21:39:24 khem: I've seen the x86 patches. great! **** ENDING LOGGING AT Sat Jul 03 02:59:56 2010