**** BEGIN LOGGING AT Wed Jul 20 02:59:57 2011 Jul 20 05:08:55 sgw: there Jul 20 05:17:08 sgw MACHINE=qemux86 TCLIBC=uclibc bitbake -ccleanall clutter-box2d-1.6; MACHINE=qemux86 TCLIBC=uclibc bitbake clutter-box2d-1.6 Jul 20 05:17:11 works here Jul 20 05:24:23 khem: hmm, interesting. I was also getting a failure in libc with a clean build Jul 20 05:30:27 you mean uclibc ? Jul 20 05:31:03 make: *** [libc/sysdeps/linux/arm/sync_file_range.o] Error 1 Jul 20 05:31:08 did you pull from branch or from patchwork Jul 20 05:31:13 branch Jul 20 05:31:16 ah Jul 20 05:31:26 I guess you have older patches Jul 20 05:31:46 I would say throw them away and resync Jul 20 05:32:00 oops, I need stable branches to work with, that invalidates all my testing. Jul 20 05:32:06 that error is from previous version of patches Jul 20 05:32:26 Ok so is that branch valid now? Jul 20 05:34:15 khem: I am still having a problem here with clutter-box2d _git for arm Jul 20 05:34:28 hmm Jul 20 05:34:33 patching issue ? Jul 20 05:34:45 Yes, the b2Math.h file is not there. Jul 20 05:34:55 the -1.6 works fine Jul 20 05:35:08 hmm Jul 20 05:35:14 let me retry Jul 20 05:35:17 I tried 1.6 Jul 20 05:35:43 _git fails Jul 20 05:36:06 hrmm I see yes it does Jul 20 05:36:36 Ok so I will wait for an update before I resetart my build here. Jul 20 05:36:50 Let me know when that branch is updated and I will resync it and restart the build Jul 20 05:37:14 khem: any idea why eglibc doesn't build localedef/gconv (do_install_locale fails) and if I ignore that then later gcc-runtime also fails because libm wasn't built too? Jul 20 05:38:03 khem: it fails the same during building SHR with populated tmpdir as well as from scratch without sstate Jul 20 05:38:14 bbl -> wrk Jul 20 05:41:41 sgw: give me 10 mins Jul 20 05:42:06 sgw: it seems I confused it when applying the patch Jul 20 05:42:15 as you said git version does not require the patch Jul 20 05:43:41 Sure just let me know when you have refreshed that branch Jul 20 05:46:38 biab Jul 20 05:50:43 sgw: ok the branch should be updated now Jul 20 05:50:57 sgw: try building core-image-sato Jul 20 05:51:04 when building with uclibc Jul 20 05:51:41 JaMa: do u have your own distro layer Jul 20 05:52:39 JaMa: and if you do then do u override DISTRO_FEATURES Jul 20 05:56:26 JaMa: you might need something like http://git.angstrom-distribution.org/cgi-bin/cgit.cgi/meta-angstrom/commit/?id=f3d25377e81a1111e89467a573dc7e9c332f5a09 Jul 20 06:12:54 khem: thanks will work on it now Jul 20 06:17:48 khem: yes Jul 20 06:17:59 sgw: ok. Jul 20 06:18:20 sgw: are u in CA somewhere Jul 20 06:18:39 khem: will try, thanks Jul 20 06:18:41 khem: which branch did you push? The version I got still has the patch in the _git.bb Jul 20 06:18:46 khem: Eugene, OR Jul 20 06:19:17 khem: never mind, my bad, I did not re-fetch Jul 20 06:22:48 sgw: hrmm Jul 20 06:22:56 oh ok Jul 20 06:23:01 its quite hot here today Jul 20 06:23:09 brain is functioning slowly Jul 20 06:23:17 Still cool here, rained a bit. Jul 20 06:23:57 I wish it rained here a bit too water bills are too high Jul 20 06:24:33 JaMa: how was ur US east coast trip ? Jul 20 06:25:47 khem: great, west coast was better last year, but this was great too :) Jul 20 06:27:31 khem: a lot of parties.. that's why I forgot about DISTRO_FEATURES_LIBC I guess :) Jul 20 06:30:28 biab Jul 20 06:41:20 JaMa: sure and it was around the time when you went away Jul 20 06:41:24 prolly Jul 20 06:41:38 * khem -> sleep() Jul 20 06:41:57 1 Jul 20 09:24:06 morning all Jul 20 10:04:32 zeddii: I am traveling and won't be able to work too much on oe-core these weeks Jul 20 10:05:01 zeddii: but regarding the kernel sysroot, you can confirm that the kernel-abiversion is also not made into sysroot Jul 20 13:26:56 gm all Jul 20 13:28:04 gm Jul 20 17:08:21 pidge: ping Jul 20 17:08:39 sgw: pong Jul 20 17:09:20 pidge, based on some recent email, I have a request Jul 20 17:09:32 sgw: shoot Jul 20 17:09:46 what's the status of getting oe-core target with both uclibc and eglibc variants? Jul 20 17:10:36 pidge: I am also seeing a spat of /home/pokybuild/poky-autobuilder/scripts/yocto-autobuild: line 60: : No such file or directory failures on the autobuilder Jul 20 17:13:06 Tartarus: ping Jul 20 17:13:10 pong Jul 20 17:14:42 sgw: one moment. Jul 20 17:16:59 Tartarus: your local.conf.sample was targeted for poky, not oe-core, correct? Or am I confused, when I looked at cgit, I dont see a .sample file Jul 20 17:20:53 no, that's oe-core Jul 20 17:21:12 trini@sb-u1004lts-64bit:/scratch/oe-core-template/oe-core$ ls meta/conf/local.conf.sample Jul 20 17:21:12 meta/conf/local.conf.sample Jul 20 17:24:58 Tartarus: it does not show up in cgit (but I can't seem to get to it right now), Ok then I declare myself confused and will get your change pulled in. Jul 20 17:25:24 k Jul 20 17:27:00 sgw: link to the spat of yocto-autobuild failures. they're probably nothing, but I want to make sure Jul 20 17:27:18 http://autobuilder.yoctoproject.org:8010/builders/fuzzy-master/builds/339/steps/shell_6/logs/stdio Jul 20 17:28:00 sgw: as for oe-core. I've done the initial looksee at it. Should take me a day to get done or so. Just finding the time at this point Jul 20 17:28:25 Ok just want to make sure it gets on the list, thanks. Jul 20 17:29:40 * pidge looks at her list and cringes. Jul 20 17:35:31 pidge: you don't want to see my list. Jul 20 17:39:22 * dvhart uses a stack instead of a list Jul 20 17:39:35 more realistic ;) Jul 20 17:40:01 dvhart: we'd be in patch gridlock if I tried that! Jul 20 17:40:22 just depends on the size of your stack ;) Jul 20 17:46:02 Tartarus: can you get to git.oe.org? Jul 20 17:48:26 sgw: nope, melo is down :( see #oe Jul 20 17:49:03 duh, will check then Jul 20 18:05:45 sgw: two patches did not get pulled Jul 20 18:06:42 sgw: mesa-xlib: Dont use locales with uclibc 2) mesa-dri: Fix build on uclibc Jul 20 18:06:59 git.oe.org is sown again sadly, its wip Jul 20 18:07:02 to get it up again Jul 20 18:07:15 there is somethin crazy Jul 20 18:07:32 my doubt is its github mirroring Jul 20 18:08:12 khem: there's comments about the mesa patches on the list Jul 20 18:08:22 ah ok Jul 20 18:29:37 someone seen python's do_install failing like this? http://paste.pocoo.org/show/442972/ second run works, but first always fails (for target python, -native is fine) Jul 20 18:29:59 also distutils LIBDIR is broken again (looking into it now) Jul 20 18:31:24 ahh :) distutils LIBDIR is broken _because_ you have to run do_install twice (and everytime you call do_install twice) Jul 20 23:43:52 khem: ping Jul 20 23:51:25 sgw: yes Jul 20 23:52:00 so my mystery clutter patch failing is solved, maybe Jul 20 23:52:26 khem: I still have failures with that clutter-box2d on multiple machines Jul 20 23:52:38 sgw: hmmm Jul 20 23:52:48 1.6 fails ? Jul 20 23:52:57 problem seems to be some thing (not sure what) stripped away ^M from the patch Jul 20 23:52:58 because the patch is not applied to git Jul 20 23:53:06 oh Jul 20 23:53:18 I dont know what that could be Jul 20 23:53:33 So when you test the patch locally it might have ^M in it, but when I get it they are gone Jul 20 23:53:34 some transfers Jul 20 23:53:55 lemme check Jul 20 23:53:55 email, git, ?? but we need to resolve this. Jul 20 23:56:57 sgw: it seem scp Jul 20 23:57:05 wtf Jul 20 23:57:43 I did git format-patch and git am Jul 20 23:57:46 and lost it Jul 21 00:03:23 sgw: whats your email id Jul 21 00:03:33 let me try to experimentally send Jul 21 00:06:22 khem: sgw@linux.intel.com Jul 21 00:08:58 khem, I fetched and used rebase to incorporate your patch, not git am Jul 21 00:08:59 sent Jul 21 00:09:15 sgw: I think its scp which kills it Jul 21 00:09:23 I sent you the patch once again Jul 21 00:09:29 see if its better Jul 21 00:10:36 I do see ^M it in this time with cat -A, but it seems to introduce them into the .bb file also, let me use git am and see what happens Jul 21 00:10:49 k Jul 21 00:10:56 its the original patch Jul 21 00:11:01 the box I used to do this work Jul 21 00:11:07 does not have any email access Jul 21 00:11:21 so I have to copy over the patches before pushing the branch out Jul 21 00:11:24 its weird process Jul 21 00:11:28 this patch does not apply Jul 21 00:11:38 never mind, my bad hang on Jul 21 00:14:39 khem, git am clobbered the ^M with the trailing whitespace warning Jul 21 00:16:45 khem is the ^M clobbered in your git branch when you commit and push it to oe-contrib? Jul 21 00:20:41 sgw: It scp Jul 21 00:21:16 I prepared patched on this machine but its first time probably it has such hosed source file Jul 21 00:21:23 I never have seen a problem before Jul 21 00:22:51 Khem, so what's next, do I try and create it here and check it in since git am is also clobbering it. Jul 21 00:25:25 yeah I guess so Jul 21 00:25:37 stupid eols Jul 21 00:25:42 I just pushed it to stage/master_under_test Jul 21 00:26:00 check it out, looks like it worked. Jul 21 00:26:39 So it sounds like the problem started at your end with scp, for future we know git preserves them once committed, but git am will not! Jul 21 00:28:48 sgw: the patch I send you is using git format-patch Jul 21 00:28:52 and thats it Jul 21 00:28:55 no scp anything Jul 21 00:29:21 so either its format-patch or send-email which is doing it Jul 21 00:30:49 khem: but remember I pulled it from your commited branch, not from email, so something went strange in the commit process, that's what I am trying to understand Jul 21 00:45:28 sgw: I committed it from a different host Jul 21 00:45:44 so it lost the ^M transfering to that other host? Jul 21 00:45:44 so git format-patch; scp ; git am Jul 21 00:45:57 it lost it in git am Jul 21 00:46:08 but when I sent the patch to u directly Jul 21 00:46:11 yup, got it. Jul 21 00:46:22 I sent using git format-patch git send-email Jul 21 00:46:29 from the original machine Jul 21 00:46:36 but you still got it mangled Jul 21 00:46:47 Right, and when I applied it git am ate the ^M Jul 21 00:46:50 so bottomline is dont use DOS editors Jul 21 00:46:57 I got them OK in email. Jul 21 00:47:04 khem: 100% with on that one. **** ENDING LOGGING AT Thu Jul 21 02:59:56 2011