**** BEGIN LOGGING AT Sat Apr 10 02:59:57 2010 Apr 10 03:01:14 :-D Tab completion has become a victim of OE's increasing popularity, it seems. Apr 10 06:50:49 good morning Apr 10 07:07:49 hi all i am doing a bitbake meta-toolchain-qte and gdbm keeps on failing to compile with different results every time. This time the log file shoes something like this Apr 10 07:07:50 http://pastebin.com/uEmVXHG2 Apr 10 07:08:04 please help as to how should i resolve this Apr 10 07:08:33 do i need to bitbake as root user Apr 10 07:08:34 ?? Apr 10 07:10:31 please help me out, i am doing this for 5 days now :'( Apr 10 07:13:54 oh, i think i am not supposed to use bitbake as root Apr 10 07:14:16 but then kindly tell me as to how can i resolve the problem, anyone ???? Apr 10 07:47:12 i know i am not a developer of the community, but as a human being just please help me out Apr 10 07:54:40 Spyzer: dont ever run bitbake as root Apr 10 07:54:55 Spyzer: and are you be any chance running on a NTFS or FAT32 drive? Apr 10 07:55:17 yes i am Apr 10 07:55:31 you hace to use a unix filesystem Im afraid Apr 10 07:55:34 so ext2/3/4 Apr 10 07:55:59 windows file systems dont have a 1:1 mapping with unix permissions Apr 10 07:56:04 or filenames Apr 10 07:56:35 i can't believe it this was an error, thanks a lot XorA|gone without your help and experience it would have been impossible for me to detect the error Apr 10 07:56:49 thanks a loooooooottttttt :) Apr 10 07:56:54 Spyzer: no probs Apr 10 07:57:07 Spyzer: happy bitbaking Apr 10 07:57:23 :) Apr 10 07:57:57 * XorA|gone returns to AFK again Apr 10 10:32:30 03Roman I Khimov  07org.openembedded.dev * r790089554f 10openembedded.git/recipes/perl/ (perl-rdepends_5.10.1.inc perl_5.10.1.bb): Apr 10 10:32:31 perl 5.10.1: update perl-rdepends Apr 10 10:32:31 Signed-off-by: Roman I Khimov Apr 10 13:56:55 03Mike Westerhof  07org.openembedded.dev * r7d3e822bef 10openembedded.git/conf/distro/include/preferred-slugos-versions.inc: SlugOS: preferred-slugos-versions.inc - older atk required due to autoconf Apr 10 14:04:08 * mwester should actually try to see if SlugOS images will still boot, after all the disruptions in the past few weeks. :( Apr 10 14:04:59 Not sure why anyone would want pam modules in the rootfs on an embedded device, but apparently that's the default behavior for OE-generated images anymore! Apr 10 14:05:50 Unless everyone is switching to an all-busybox rootfs? Does that work now? Apr 10 14:19:07 mwester: the idea is that PAM should be a distro feature, I seem to recall I did some more work on Sudo and Shadow to make sure that was the case but not sure if they found there way back into OE dev. Apr 10 14:20:04 mwester: that said, Angstrom (not OE ;-)) defaults to enabling the PAM feature and on a lot of modern devices like the OMAP3's etc. it's actually needed as the userspace is so 'desktop' these days. Apr 10 14:20:25 Ok. It seems to be pulled in via e2fsprogs which pulls in the entire util-linux-ng at this point in time. Apr 10 14:21:08 I've noticed that the small-flash devices seem to have been left behind by OE recently; a bit sad. :( Apr 10 14:22:16 mwester: do you mean OE or Angstrom? I would say that may be true of Angstrom but not raw OE Apr 10 14:22:55 SlugOS needs a very small initial in-flash image; the user's install process will set up a rootfs on an external device -- the external device can have the full set of "stuff" installed, but the initial flash image needs to be limited to <5MB in size. Apr 10 14:23:03 DJWillis: I mean OE . Apr 10 14:23:32 And indeed, it is "raw" OE that is pulling in the full e2fsprogs and linux utils packages now. Apr 10 14:24:22 mwester: so is that not the aim of things like the kexec images and the initrd stuff (you could adapt that for a small flash easilly enough). That said, I don't have that may flash constrained devices so I can't speak from experance. Apr 10 14:24:40 Ok, then I will assume that to be a bug. Apr 10 14:25:14 And if I find time to figure out what's wrong, I'll expect my patch to be greeted by great joy. :p Apr 10 14:25:56 * mwester is a little skeptical of that after perusing the commit logs associated with the e2fsprogs and linux utils changes. Apr 10 14:26:04 mwester: I would say that is PAM is pulled in and it's NOT in the distro features for what your using that is a bug, pure and simple ;-) Apr 10 14:26:30 mwester: that said, I am not an OE dev just a hacker who sometimes gets patches in. Apr 10 14:26:39 pam is put in by policykit Apr 10 14:26:57 I don't have policykit in the rootfs. Apr 10 14:27:30 there are some other packages that depend on pam, i tried to remove those depencies for SHR but i failed Apr 10 14:30:09 Yeah, no "USE" flags (I think that's what they're called -- I have had similar trouble where I needed to change the configure flags for packages to remove dependencies. Apr 10 14:31:32 I suspect it might be easier just to remove the unnecessary files after the rootfs is built, rather than trying to build the packages without the dependencies. Apr 10 14:34:31 Actually, I think maybe I should try a busy-box-only rootfs, once again. Perhaps the latest busybox versions will let me do that, in which case all of this just "goes away". :) Apr 10 14:36:55 03Michael Lippautz  07org.openembedded.dev * r0c33cc6d11 10openembedded.git/recipes/libxslt/libxslt_1.1.21.bb: Apr 10 14:36:55 libxslt: Convert old version to new staging Apr 10 14:36:55 Signed-off-by: Michael Lippautz Apr 10 15:01:59 03Thomas Zimmermann  07org.openembedded.dev * r83c92ae8d1 10openembedded.git/recipes/omnewrotate/omnewrotate_0.5.4.bb: Apr 10 15:01:59 omnewrotate: make recipe valid Apr 10 15:01:59 *it's an svn recipe that had rev=HEAD in SRCURI Apr 10 15:01:59 Signed-off-by: Thomas Zimmermann Apr 10 15:44:30 03Michael Lippautz  07org.openembedded.dev * r216fd1ee6a 10openembedded.git/recipes/libxslt/libxslt.inc: Apr 10 15:44:30 libxslt: Inherit from binconfig to add support for xslt-config Apr 10 15:44:30 Signed-off-by: Michael Lippautz Apr 10 19:54:32 03Michael Lippautz  07org.openembedded.dev * r7e760b239a 10openembedded.git/recipes/smpeg/smpeg_svn.bb: Apr 10 19:54:32 smpeg: Installing smpeg.m4 inoto staging aclocal Apr 10 19:54:32 * Fixes bug #4989 Apr 10 19:54:32 Signed-off-by: Michael Lippautz Apr 10 19:54:38 03Michael Lippautz  07org.openembedded.dev * ra7838e202b 10openembedded.git/recipes/zbar/zbar_0.8.bb: Apr 10 19:54:38 zbar: Convert to new staging Apr 10 19:54:38 Signed-off-by: Michael Lippautz Apr 10 19:58:29 hi Apr 10 19:58:39 whom do I ask if I want someone who knows OpenEmbedded to make a package out of my software? Apr 10 20:00:53 hi , how can i do bitbake for some glibc version ? Apr 10 20:01:13 when i do bitbake , i got glibc version not compatible with my OS Apr 10 20:02:06 mira, you could do it yourself. Apr 10 20:03:30 hrm, relatively unlikely Apr 10 20:03:37 I did look at OE in the past but didn't quite get it Apr 10 20:07:21 it’s “just” a shell… Apr 10 20:07:42 a small yet powerful one, easy to cross-build… shouldn't be too hard Apr 10 20:08:43 (it's mksh, which replaced /system/bin/sh in android-x86 a month or so ago already) Apr 10 20:41:16 i'm having a problem cross-compiling Csound with a recipe that i've written. it's sort of a hack job and compiles, but strangely not for the target architecture Apr 10 20:41:33 might anyone be willing to help me troubleshoot my bbfile? Apr 10 21:16:37 maybe if i ask nicely, like "pretty please?" Apr 10 21:32:42 03Klaus Kurzmann  07org.openembedded.dev * r50247d6aa2 10openembedded.git/recipes/shr/shr-e-gadgets_git.bb: Apr 10 21:32:42 shr-e-gadgets_git.bb: bump SRCREV Apr 10 21:32:42 Signed-off-by: Klaus Kurzmann Apr 10 21:32:49 03Klaus Kurzmann  07org.openembedded.dev * r292f07f054 10openembedded.git/recipes/shr/e-wm-config-illume2-shr_git.bb: Apr 10 21:32:50 e-wm-config-illume2-shr_git.bb: bump SRCREV Apr 10 21:32:50 Signed-off-by: Klaus Kurzmann Apr 10 22:40:21 LANman247: Or perhaps just give a link to your recipe, so people can look at it without having to ask you for it, and wait for your response before they can. ;-) Apr 10 22:43:58 heh good idea! actually, i've figured out a bit more since i posted that, and am getting an error during the compilation stage: "| /stuff/build/tmp/work/armv7a-angstrom-linux-gnueabi/csound-1.0-r0/temp/run.do_compile.4715: line 320: /stuff/build/tmp/staging/i686-linux/usr/bin/scons: No such file or directory | FATAL: scons build execution failed." Apr 10 22:46:03 here's a link to my recipe http://pastebin.com/z2LpA0gD Apr 11 01:09:42 is there anyone that can help me with my recipe? Apr 11 01:22:46 You picked a poor time for OE; the OE folks are seldom about on the weekends. Apr 11 01:23:28 What is not working with your recipe? Apr 11 01:28:38 well, i'm having a problem getting scons to work Apr 11 01:28:51 (and i figured as much with them being away :P) Apr 11 01:29:10 i've spent about 12 hours trying to get this to compile today Apr 11 01:29:19 ok, let me dive deeper into the problem Apr 11 01:29:46 here's my recipe http://pastebin.com/z2LpA0gD Apr 11 01:30:25 and i'm getting an error when bitbake tries to call scons: "| /stuff/build/tmp/work/armv7a-angstrom-linux-gnueabi/csound-1.0-r0/temp/run.do_compile.4715: line 320: /stuff/build/tmp/staging/i686-linux/usr/bin/scons: No such file or directory | FATAL: scons build execution failed." Apr 11 01:30:57 just for reference, i'm trying to port Csound over to the Gumstix Overo COM Apr 11 01:36:44 i read somewhere that scons needed to be built in bitbake in order to use it for a recipe, even though i have it installed on my Ubuntu system already. so, i tried to build python, and i get this error with readLine "| NOTE: make DESTDIR=/stuff/build/tmp/work/armv7a-angstrom-linux-gnueabi/readline-5.2-r6/image install | make: *** No rule to make target `install'. Stop. | FATAL: oe_runmake failed" Apr 11 01:38:12 needless to say, for someone that's not at a pro at this like some of the people here, a large crater has been forming on my desk from me banging me head on it Apr 11 01:39:37 arse scons is broken again :( Apr 11 01:41:21 :( Apr 11 01:41:42 man its python as well, that rules me out of fixing Apr 11 01:41:52 did python-scons-native actually get built? Apr 11 01:42:38 what's the best way to check? "bitbake python-scons-native" ? Apr 11 01:43:12 yeah Apr 11 01:43:18 alright, trying... Apr 11 01:43:40 its been about 1000 years since I used scons Apr 11 01:44:38 at least you've had experience with it before. today was my first day, and so far scons is winning :P Apr 11 01:45:06 I have a horrible feeling that was my experience Apr 11 01:45:29 BTW we are all quite here as its 2:45am :-) Apr 11 01:45:47 or 3:45am for a good 80% of OE people Apr 11 01:46:13 lol ah yes, europe/UK? Apr 11 01:47:07 I'm here too,I was awaken Apr 11 01:47:09 Im in UK :-) Apr 11 01:47:21 and I've a big kenrel issue Apr 11 01:49:48 hey there GNUtoo good morning :) Apr 11 01:50:18 hi Apr 11 01:51:35 ooops forgot I was on a proxy and rebooted the proxy :-) Apr 11 01:51:41 :P Apr 11 01:52:08 python-scons-native wasn't built, but now it is Apr 11 01:52:18 XorA, btw what could be the cause of "Unable to handle kernel paging request at virtual address" ? no more free ram? Apr 11 01:52:41 GNUtoo|oeee: no idea, I thought any kernel address badness caused that Apr 11 01:53:00 ok Apr 11 01:53:09 wow, this actually seems like it's compiling now Apr 11 01:53:34 LANman247: scons.bbclass must be slightly broken then, it does DEPENDS+="python-scons-native" so it should in theory build first Apr 11 01:55:02 ah alright, i see Apr 11 01:55:18 * XorA is too tired to figure out why, send an email to the list :-) Apr 11 01:55:26 lol Apr 11 01:55:29 * XorA is just waiting for hot water so he can shower then sleep Apr 11 01:55:46 * LANman247 doesn't blame XorA Apr 11 01:56:01 * GNUtoo|oeee tried sleep but had a kenrel oops Apr 11 01:56:24 GNUtoo|oeee: if you oops in your sleep does anyone reboot you? Apr 11 01:57:07 lol Apr 11 01:57:17 haha Apr 11 01:57:49 alright, i tried a compile but scons pooped out on me Apr 11 01:57:55 | CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include/fltk-1.1 Apr 11 01:57:55 | cc1plus: internal compiler error: in add_path, at c-incpath.c:425 Apr 11 01:57:56 | Please submit a full bug report, Apr 11 01:57:56 | with preprocessed source if appropriate. Apr 11 01:57:56 | See for instructions. Apr 11 01:57:56 | scons: *** [Opcodes/ampmidid.os] Error 1 Apr 11 01:57:56 | scons: building terminated because of errors. Apr 11 01:57:57 | FATAL: scons build execution failed. Apr 11 02:00:20 our gcc is "fixed" to issue that error if someone hardcodes an include from /usr/include Apr 11 02:00:28 you need to fix the package source Apr 11 02:02:36 gmm alright, i'll have to try to figure that out Apr 11 02:02:39 *hmm Apr 11 02:09:42 one small question if anyone's still around: for an scons recipe, how would i pass a variable, as in "scons usePortMIDI=1" ? Apr 11 02:10:16 sorry no idea, Im off now, good luck Apr 11 02:10:39 alright, cheers and g'night **** ENDING LOGGING AT Sun Apr 11 02:59:56 2010