**** BEGIN LOGGING AT Tue Mar 15 02:59:57 2011 Mar 15 05:56:10 03Khem Raj  07master * rf3ef166dca 10openembedded.git/recipes/uclibc/ (3 files in 2 dirs): Mar 15 05:56:10 uclibc_git: Upgrade to latest 0.9.32 snapshot Mar 15 05:56:10 Signed-off-by: Khem Raj Mar 15 06:00:02 Hi Mar 15 06:00:17 I need to clear one thing Mar 15 06:00:39 is there a way to build console kit without X11 Mar 15 06:01:17 I saw that it is dependent on X11 .... just confirming whether we can build it without X11 Mar 15 07:01:01 gm Mar 15 07:11:58 03Steffen Sledz  07org.openembedded.dev * re6596c3d07 10openembedded.git/recipes/linux/ (linux-2.6.24/cfi-amic.patch linux_2.6.24.bb): Mar 15 07:11:58 linux-2.6.24: add support for AMIC flashes (e.g. A29L160AT) Mar 15 07:11:58 Similar patch is on its way to the mainline kernel: Mar 15 07:11:58 http://thread.gmane.org/gmane.linux.kernel/1111144 Mar 15 07:11:58 Signed-off-by: Steffen Sledz Mar 15 07:12:09 03Steffen Sledz  07org.openembedded.dev * r0eec980ccf 10openembedded.git/recipes/net-snmp/net-snmp_svn.bb: (log message trimmed) Mar 15 07:12:09 net-snmp: disable libnl use Mar 15 07:12:09 Fixes compile error Mar 15 07:12:09 | ./.libs/libnetsnmpmibs.so: undefined reference to `nl_handle_destroy' Mar 15 07:12:10 | ./.libs/libnetsnmpmibs.so: undefined reference to `nl_handle_alloc' Mar 15 07:12:10 which occurred sometimes depending on build order (not in clean package Mar 15 07:12:11 only builds). Mar 15 07:58:00 03Paul Menzel  07master * rd2c2c71e3e 10openembedded.git/recipes/avahi/mango-lassi_git.bb: Mar 15 07:58:01 mango-lassi: Typo. s/ESCRIPTION/DESCRIPTION/ Mar 15 07:58:01 • Also bump PR after discussion on list openembedded-devel [1]. Mar 15 07:58:01 [1] http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-March/030940.html Mar 15 07:58:01 Signed-off-by: Paul Menzel Mar 15 08:31:13 when excute bitbake x11-image, Mar 15 08:31:15 I got error like this:| * satisfy_dependencies_for: Cannot satisfy the following dependencies for angstrom-gpe-task-base: Mar 15 08:31:17 | * libc6 (>= 2.12) * Mar 15 08:31:19 | * opkg_install_cmd: Cannot install package angstrom-gpe-task-base. Mar 15 08:31:21 but my libc6 is neraly the lastest! Mar 15 08:31:34 NOTE: Running task 5911 of 5913 (ID: 14, /samba/stuff/openembedded/recipes/images/x11-image.bb, do_rootfs) Mar 15 08:31:36 error int this task! Mar 15 08:32:48 thats something else Mar 15 08:35:11 is there a way to build console kit without X11?? Mar 15 08:36:06 But i cant get the package of the rootfilesystem Mar 15 08:36:30 tdh2002 anyway I would use eglibc as libc Mar 15 08:36:37 hi ant Mar 15 08:40:13 the i must modify bb file , change libc to eglibc? Mar 15 08:40:26 local.conf Mar 15 08:47:05 LIBC = "eglibc" Mar 15 08:47:44 gm pb Mar 15 08:48:06 Morning Mar 15 08:48:17 GM game master good morning Mar 15 08:51:48 good morning Mar 15 08:51:58 hi mckoan Mar 15 08:53:58 hi woglinde Mar 15 08:59:58 7 hours ago Linus Torvalds Linux 2.6.38 master v2.6.38 Mar 15 09:00:00 :) Mar 15 09:00:08 jupp Mar 15 09:02:39 morning folks Mar 15 09:03:28 eFfeM: you'll see some serious clearings in recipes/linux very soon ;) Mar 15 09:07:11 hi blidman Mar 15 09:07:51 ant_work: seems a good plan Mar 15 09:08:09 ant_work: hope you won't get burned down for it Mar 15 09:08:14 eFfeM: in the dubios cases -> obsoleted Mar 15 09:08:21 sure Mar 15 09:08:47 eFfeM: that guy has now his linux-omap series to lay with ;) Mar 15 09:08:55 s/lay/play/ :p Mar 15 09:09:06 Mar 15 09:10:10 well got some flac from other people too, then again also got some appreciated messages Mar 15 09:11:11 btw I like the oe-core version retention policy proposal, hope we can do the same for the non oe-core recipes Mar 15 09:11:42 hopefully Mar 15 09:11:46 woglinde: any known issue in navit? I would like to bump SRCREV/PV Mar 15 09:12:05 eFfeM: you see, I'm now in that situation with mtd.utils Mar 15 09:12:25 1.3.1 is working on our devices, 1.4.x not Mar 15 09:13:22 so until debugged, I'll have to pin the old version Mar 15 09:13:35 jama lot of bugfixes Mar 15 09:13:51 I know I've it with AUTOREV :) Mar 15 09:14:30 and it works ok on my devices.. so just asking if there is something seriously broken what I'm not using or doesn't happen here Mar 15 09:17:02 03Martin Jansa  07master * refc6527ff3 10openembedded.git/recipes/navit/navit_svn.bb: Mar 15 09:17:02 navit: bump SRCREV and PV Mar 15 09:17:02 * since r3705 it's development for 0.5.0 version Mar 15 09:17:02 Signed-off-by: Martin Jansa Mar 15 09:17:13 03Martin Jansa  07master * rec0f190332 10openembedded.git/recipes/freesmartphone/ (4 files): Mar 15 09:17:13 freesmartphone: bump cornucopia, msmcomm, fso-specs, libfso-glib SRCREV Mar 15 09:17:13 * fsoaudiod doesn't compile with old SRCREV because it didn't exist in that rev Mar 15 09:17:13 Signed-off-by: Martin Jansa Mar 15 09:17:36 :) Mar 15 09:19:05 ant_work: understood Mar 15 09:19:21 time for dayjob.. bye Mar 15 09:19:43 i have no problem with it, but some recipes seem to accrue lots of versions where no one knows why we have them Mar 15 09:19:46 eFfeM: btw, that one is a typical case Mar 15 09:19:55 someone updated mtd-utils Mar 15 09:20:15 ant git log Mar 15 09:20:23 the recipe needs BBCLASSEXTEND btw Mar 15 09:20:34 I have ready 1.4.2 but is still broken on Z... Mar 15 09:20:40 I can't commit it Mar 15 09:21:03 woglinde: heavy changes btw 1.3 and 1.4 Mar 15 09:21:12 ant_work: seems git bisect is our friend :\ Mar 15 09:21:27 I need confirmation it fails/works on other devices Mar 15 09:21:38 it fails on all 3 Zaurus I have Mar 15 09:21:58 it = flash_eraseall -j Mar 15 09:22:15 flash_erase and flash_eraseall have been unified Mar 15 09:22:28 rally new code, now ioctls... Mar 15 09:22:45 I tried old and new kernel w/out any success Mar 15 09:29:10 hi risca Mar 15 09:29:22 hello, Mar 15 09:30:08 i still have problems starting an oe build. ERROR: Error parsing /openembedded/recipes/images/bootstrap-image.bb: global name 'runtime_mapping_rename' is not defined Mar 15 09:30:55 Cyberian whats your bitbake version? Mar 15 09:31:13 1.12.0 Mar 15 09:31:33 hm Mar 15 09:31:40 try git master version Mar 15 09:32:14 i had a complete build till last week untiil i updated my openembedded git repository Mar 15 09:32:15 03Martin Jansa  07master * rb1fbaf62cd 10openembedded.git/recipes/ffmpeg/ffmpeg_git.bb: Mar 15 09:32:15 ffmpeg_git: bump PR Mar 15 09:32:15 * because SRC_URI change from a96b9ad7525812574d7b719241472fc0e9e95b83 resets LOCALCOUNT in SRCPV Mar 15 09:32:15 Signed-off-by: Martin Jansa Mar 15 09:33:25 cyberian oe dev is on going work Mar 15 09:34:22 which is latest stable so I can fall back to it Mar 15 09:35:38 2011.03-maintenance Mar 15 09:36:30 Good morning woglinde Mar 15 10:00:06 finally, after switching to 2011.03-maintenance build goes on Mar 15 10:21:14 and now I'm back to my initial problem ;) http://pastebin.com/SEewxk3i autoconf213 does not configure Mar 15 10:21:45 why the hell you have autoconf213? Mar 15 10:21:54 what distro are you using= Mar 15 10:21:55 ? Mar 15 10:22:20 pastebin please your local.conf Mar 15 10:23:37 building fennec? :) Mar 15 10:25:01 Cyberian: you need this patch http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-March/030754.html Mar 15 10:27:52 hi woglinde Mar 15 10:32:58 Hi ..... Is there any way to build consolekit without X11 package? Mar 15 10:33:31 noor edit the recipe yourself= Mar 15 10:33:32 ? Mar 15 10:33:38 yes i tryt to build fennec. or can you suggest another browser that supports java script? Mar 15 10:34:25 I was just looking whether it is possible or not ... if it is then I'll change the recipe Mar 15 10:34:29 Thanks woglinde Mar 15 10:35:13 noor I dont know it Mar 15 10:35:19 perhaps read the website Mar 15 10:35:23 or study the code Mar 15 11:07:29 I'm trying to get powermanagement running for a very beaglelike board. I have a kernel with powermanagement support, and I have programmed the tps with some scripts. I miss pm-suspend and friends though. Is there a recipe I need to add to get those commands? Mar 15 11:17:11 hi, can anyone suggest how to decrease amount of hdd space that OE needs? >20gb is not acceptable for a minimal image Mar 15 11:19:16 ? Mar 15 11:19:27 inherit rm_work Mar 15 11:19:38 it's already here Mar 15 11:19:49 than 20 gig is enough Mar 15 11:20:47 woglinde: no :) Mar 15 11:20:55 console-image build eats more Mar 15 11:21:41 ? Mar 15 11:21:45 you have to build it in few steps like "bitbake -k gcc-cross; bitbake -k virtual/kernel; bitbake -k console-image" Mar 15 11:21:52 I tested it enough Mar 15 11:21:58 20 gig was okay for me Mar 15 11:22:05 because rm_work is executed near the end for runqueue Mar 15 11:22:22 hm depens on your BBTHREADS yes Mar 15 11:22:53 https://spreadsheets.google.com/ccc?key=0AhY8lmfkCabTdGZRSGZ0eXg5cWhkV2k1c0Jvd0lvVnc&authkey=CLeK150N Mar 15 11:23:14 hm.. well, 20Gb is minimal Mar 15 11:23:26 with BB_SCHEDULER=completion Mar 15 11:23:51 ah.. Mar 15 11:24:19 there was multiple images build Mar 15 11:24:20 sorry Mar 15 11:24:27 console-image, x11-image, opie-image Mar 15 11:24:45 * anarsoul 'll try with BBTHREADS=1 Mar 15 11:25:18 ? Mar 15 11:25:20 no Mar 15 11:25:32 than rm stuff will run at the ends Mar 15 11:25:34 with more Mar 15 11:25:46 there is a chance some rm works btw. Mar 15 11:25:53 but do what jay7 said Mar 15 11:26:03 firs run something big Mar 15 11:26:07 jama said :) Mar 15 11:26:09 like kernel compile Mar 15 11:26:12 ups sorry Mar 15 11:26:28 anarsoul: and set BB_SCHEDULER=completion in local.conf Mar 15 11:26:37 ok Mar 15 11:27:11 build will be slightly slower but rm_work should be called earlier Mar 15 11:27:45 what about BB_NUMBER_THREADS? Mar 15 11:27:45 you may see this difference on chart4 vs chart3 Mar 15 11:28:08 how many processors you have? Mar 15 11:28:15 imho this shouldn't affect space Mar 15 11:28:18 2 cores Mar 15 11:28:24 may be very little chance Mar 15 11:30:12 Are any of the people that worked on valgrind about (Vladimir/Roman/Khem)? I'd like to understand what the recipes/valgrind/valgrind-3.6.0/fix-link-tool.patch was attempting to do before I submit a patch that removes it. Mar 15 11:42:34 hi Mar 15 11:42:42 hi obi Mar 15 11:54:52 gm again Mar 15 11:55:02 re effem Mar 15 12:08:23 I seem to have another version conflict now: "Perl lib version (5.10.1) doesn't match executable version (v5.12.3)". I thought the $PATH variable is going to be set accordingly to build process Mar 15 12:08:52 hm I thought that was fixed Mar 15 12:09:12 in some places yes.. but probably not all (like ancient fennec..) Mar 15 12:09:29 i can set a symbolic link from original perl location but that would only be a workaround Mar 15 12:09:58 Cyberian: check which file is executed and if there is shebang !/usr/bin/perl instead env perl Mar 15 12:33:39 hi flo_lap Mar 15 12:40:24 hm, 2.6.38 out Mar 15 12:40:54 ynezz: who will be first with adding it? Mar 15 12:41:09 JaMa|Off: I've already some freash defconfigs for Z family, a bit less modular Mar 15 12:41:21 ant_work: ok Mar 15 12:41:28 waiting someone adds 2.6.38 ;) Mar 15 12:41:43 * JaMa|Off has pretty fresh rev in 2.6.38-rc8.. Mar 15 12:42:01 JaMa|Off, could you make that a bit more clear for me? As i see this happens because during compile of autoconf aclocal-1.11 is called which calls perl. Mar 15 12:42:03 * ynezz rebased yesterday Mar 15 12:42:14 if not there I'll do in a few hours Mar 15 12:42:30 btw, linux-rp will be obsoleted after that Mar 15 12:42:41 (we really hope) Mar 15 12:42:44 Cyberian: in some fine there is #!/usr/bin/perl which is bad (calls perl on your host instead perl-native) Mar 15 12:43:02 except offline charging :( Mar 15 12:43:24 little things... 8) Mar 15 12:43:35 * JaMa|Off cannot test 2.6.38 on device now.. so also waiting for someone else to do it :) Mar 15 12:43:45 it boots Mar 15 12:43:58 well, rc-8 did Mar 15 12:44:34 yes, at least for the aclocal-1.11 on my system. but this also exists in ./tmp/sysroots/x86_64/usr/bin which I guess should be called and starts with "#! /usr/bin/env perl" Mar 15 12:45:17 JaMa|Off: I had a moment of panic yesterday...then I checked spitz and it lacks config_blk_dev_sd=y :) Mar 15 12:46:15 ant_work: ./spitz/defconfig:CONFIG_BLK_DEV_SD=m Mar 15 12:46:17 I guess somewhere on my way to the compile command setting the PATH variable is missing Mar 15 12:46:31 JaMa|Off: yep = no boot from CF / HDD Mar 15 12:47:26 I cloned that and now we have the whole family, with minimal diffs btw each other Mar 15 12:47:46 we'll have to improve USB* support Mar 15 12:47:47 hrw: we'll see :) Mar 15 12:48:25 ant_work: you mean linux-kexecboot config or linux? Mar 15 12:48:32 linux Mar 15 12:49:28 ant_work: hmm I have it there Mar 15 12:49:44 ant_work: ah only as module.. Mar 15 12:50:31 ant_work: but still booting 2.6.37.2 just fine :) Mar 15 12:50:38 ant_work: from SD Mar 15 12:51:14 l-k has been refined too (other models) Mar 15 12:51:32 do you think sound should be compiled inkernel? Mar 15 12:51:52 or left as module? Mar 15 12:52:16 well no opinion on that /me didn't use sound from spitz for few years Mar 15 12:53:04 It's easier in kernel becasue there's nothing to autoload it. Mar 15 12:53:18 how can i trace back where and when perl was called during build? Mar 15 12:53:58 I see in linux-rp there is e.g. module_autoload_snd-soc-spitz_spitz = "snd-soc-spitz" Mar 15 12:54:27 and e.g. RPROVIDES_kernel-base += "hostap-modules" Mar 15 12:55:57 I'd prefer to compile in kernel the drivers for the devices physically present on the board Mar 15 12:56:26 broonie: yes, otherwise you need more per-machine customization Mar 15 13:02:33 how can i trace back where and when perl was called during build? Mar 15 13:04:30 hi all! Mar 15 13:09:50 hi flo_lap Mar 15 13:15:38 Jay7: it does not clean sources from tmp/work/armv5te-angstrom-linux-gnueabi/gcc-cross-* Mar 15 13:15:40 why? Mar 15 13:16:40 there's INHERIT += "rm_work" in local.conf Mar 15 13:16:48 same for eglibc and binutils Mar 15 13:18:27 anarsoul: that is question for some core-dev's :) Mar 15 13:19:44 it does for me Mar 15 13:20:04 shr@opmbuild:~$ ls shr-unstable/tmp/work/armv4t-oe-linux-gnueabi/gcc-cross-4.5-r33.1+svnr170880/ Mar 15 13:20:07 temp Mar 15 13:20:17 what distro? Mar 15 13:20:23 ah, shr Mar 15 13:22:09 I'm on Angstrom-2008.1 from the latest dev, with linux-omap-psp 2.6.32. I want to try linux-omap 2.6.38 to see if it solves some of my power management woes, and wonder how I should go about this. Just change kernel-version "somewhere", or has the time come to upgrade DISTRO as well? Mar 15 13:23:08 Angstrom-2008.1 is not in latest dev.. so yes or change branch Mar 15 13:24:13 JaMa|Off: ah. went away recently? I just fetched and haven't started compiling yet :) Mar 15 13:24:24 yes Mar 15 13:25:12 I would ask koen, 2.6.38 might not even boot... Mar 15 13:25:44 woglinde is using 2.6.37.y on beagle, so this might work Mar 15 13:26:11 but dunno which recipe or patches... Mar 15 13:26:58 noticed after todays pull that beagleboard is in COMPATIBLE_MACHINE for .38, though I don't know how much that says Mar 15 13:27:39 there's no .38 in oe Mar 15 13:27:42 (yet) Mar 15 13:29:39 the linux-omap_2.6.38.bb is just pretending? It says: PV = "2.6.37+2.6.38-rc7", and I don't know what that means :) Mar 15 13:29:50 ah, sorry then Mar 15 13:37:19 to get most out of beagle Mar 15 13:37:25 you need the psp .32 kernel Mar 15 13:37:36 .37 the dsp codecs dont work well Mar 15 13:38:19 he just needs to test some pm stuff Mar 15 13:40:42 I don't use the dsp, but I need sleep/suspend/wake. I have added power scripts and an i2c driver that implements pm_power_off, but I have no idea how to test & use it. Mar 15 13:41:41 don't have much of the fun stuff from http://elinux.org/OMAP_Power_Management. I can do "echo mem > /sys/power/state", but then things just die for good Mar 15 14:05:10 tasslehoff: alias pm_suspend="echo mem > /sys/power/state" :) Mar 15 14:06:22 tasslehoff: isn't pmtools recipe you needed? Mar 15 14:06:35 Jay7: perhaps :) tried that, and it says "[ 4007.675750] PM: Syncing filesystems ... done. " and no more. can't quite figure out if I've made it work or not... Mar 15 14:08:01 Jay7: that could very well be. I actually dismissed it because of the name of the recipe. seemed so old :) Mar 15 14:10:30 hm, it definitely does not do rm_work for some reason Mar 15 14:10:34 how to debug why? Mar 15 14:17:00 anarsoul check your local.conf again Mar 15 14:17:26 INHERIT += "rm_work" Mar 15 14:18:28 woglinde: cat local.conf | grep 'INHERIT +=' Mar 15 14:18:29 INHERIT += "shr-mirrors rm_work" Mar 15 14:19:10 put it in two Mar 15 14:19:16 woglinde: same effect Mar 15 14:19:21 hm Mar 15 14:19:27 which bitbake version? Mar 15 14:19:40 bitbake --version Mar 15 14:19:40 BitBake Build Tool Core version 1.10.2, bitbake version 1.10.2 Mar 15 14:19:50 try 1.12 Mar 15 14:19:57 or git master Mar 15 14:25:19 will try 1.12.0 Mar 15 14:28:18 03Koen Kooi  07org.openembedded.dev * rde0a4b75c3 10openembedded.git/recipes/policykit/policykit-gnome_0.100.bb: Mar 15 14:28:18 policykit-gnome: append to EXTRA_OECONF instead of overriding it Mar 15 14:28:18 This is needed to disable introspection Mar 15 14:28:18 Signed-off-by: Koen Kooi Mar 15 14:31:07 i tried to track down where my perl libraries got mixed up but no success yet. Only that calling perl from command line gives me correct version (the one required for openembedded), aclocal-1.11 starts with "#! /usr/bin/env perl" Mar 15 14:38:48 Cyberian: there's nothing wrong with /usr/bin/env perl. that'll find perl in the PATH, and OE has its sysroots in its path Mar 15 14:38:59 or..? Mar 15 14:39:24 as far as i see yes Mar 15 14:39:54 i went though some of the steps in autoconf makefile to reproduce all manually Mar 15 14:40:25 and i get same errors. somewhere on the way it uses perl from /usr/bin Mar 15 14:46:17 morning kergoth Mar 15 14:46:46 hey Mar 15 14:46:52 hm, any uk folks know a good place to buy 1.8" hard disks? I think I have dropped my ipod one time too many. Mar 15 14:54:20 03Koen Kooi  07org.openembedded.dev * r8a9ce9d369 10openembedded.git/conf/distro/angstrom-2010.x.conf: Mar 15 14:54:20 angstrom v2010.x: blacklist bash-noemu Mar 15 14:54:20 It doesn't build and has the dreaded replaces/conflicts combo Mar 15 14:54:20 Signed-off-by: Koen Kooi Mar 15 14:54:42 pb__: It's probably cheaper/easier to buy a new iPod. Mar 15 14:57:11 pb__: are you a reseller? computer2000.co.uk Mar 15 15:03:02 OK, I've got two different problems going here, but I suspect they're both the same config problem. I'm trying to compile an Angstrom image, my local.conf is targeting 2011.3-maintenance. Without any INHERIT += "package_*" lines, it fails to inherit rootfs_{IMAGE_PKGTYPE}.bbclass. With either package_ipk or package_deb selected, it fails to parse uclibc_git.bb, with a python traceback. Google is not being helpful for me. What's going Mar 15 15:04:19 what's a good way to work with the release tag + some local changes? release-2011.03 is a moving target at the moment, isn't it? Mar 15 15:04:39 it's not, it's released Mar 15 15:04:50 work with that branch Mar 15 15:05:02 not with the tag Mar 15 15:05:12 ynezz: ah. cool. is there a branch as well? Mar 15 15:06:08 I mean: there is a branch as well? :) Mar 15 15:07:13 2011.03-maintenance? Mar 15 15:07:31 ah, so I'm mistaken :) it's a maintenance branch Mar 15 15:08:55 http://openembedded.org/index.php/Release-2011.03 Mar 15 15:09:53 ynezz: I was on that page, but didn't read the "has been completed" part. thanks :) Mar 15 15:11:53 minifig404: what machine, distro, bitbake version? Mar 15 15:12:28 minifig404: btw how you can target 2011.3-maintenance branch in the local.conf? Mar 15 15:13:02 machine, beagleboard, host distro ubuntu 10.04 LTS (required for other tools), bitbake version is 1.12 Mar 15 15:13:15 distro in oe Mar 15 15:13:37 angstrom-2008.1? Mar 15 15:13:43 distro in local.conf is "angstrom-2011.3-maintenance" Mar 15 15:14:04 That was auto-selected by angstrom setup script. Mar 15 15:15:17 impossible Mar 15 15:15:42 can you pastebin your local.conf please? Mar 15 15:15:42 ...Can you elaborate on that please? Mar 15 15:15:52 Give me a minute... Mar 15 15:16:04 there's angstrom-2008.1 and angstrom-2010.x as far as I know Mar 15 15:16:54 I seem to recall a website announcement that earlier this month, 2008.1 was sent to maintenance. Mar 15 15:19:10 http://pastebin.com/R9EvUXGc Mar 15 15:19:39 you've edited that file? Mar 15 15:19:59 Yes; I changed INHERIT Mar 15 15:20:07 I didn't touch the DISTRO. Mar 15 15:20:19 It'd be great if that was the problem. Mar 15 15:23:40 it is problem Mar 15 15:23:50 angstrom-2008.1 should be there instead Mar 15 15:23:57 OK. Mar 15 15:26:32 The recipe parse step is working much better, but I'm getting a lot of warnings about unsupported packages... libiconv, bluez, fso-apm. I'm assuming those are expected. Thank you for your help. Mar 15 15:27:15 I finally found the error in my build regarding Perl Versions. Mar 15 15:28:26 and are stuck at a new one -.- Mar 15 15:45:41 Hey, is the bitbake-dev list dead or just slowly moderated? I sent a patch to it last week, and it hasn't appeared. Mar 15 15:46:56 which address did you use? Mar 15 15:49:32 I sent mail to bitbake-dev@lists.berlios.de yesterday and have not yet seen it on the list Mar 15 15:50:05 I used bitbake-dev@lists.berlios.de Mar 15 15:50:26 hmm, ok, let me send a test mail as well Mar 15 15:50:45 The last message in the web archive is 8 days old-- a bigger gap than is usual. Mar 15 15:56:35 woglinde: same problem with bitbake-1.12.0 Mar 15 15:57:32 it does not do rm_work Mar 15 16:09:11 or it's going to do rm_work later? Mar 15 16:09:37 i.e. perl-native compilation is finished, m4-native compilation is started, but rm_work for perl-native is not performed Mar 15 16:09:42 is it some kind of bug? Mar 15 16:11:18 btw, when it does rm_work? Mar 15 16:11:27 when it _should_ do it? Mar 15 16:12:31 it will do it near the end of runqueue Mar 15 16:13:11 if you kill current build and run just "bitbake perl-native" you should see it done almost immediatelly Mar 15 16:13:39 JaMa|Off: but it's stupid :\ Mar 15 16:14:27 JaMa|Off: it increases free space requirements Mar 15 16:14:50 then fix it and send patch.. last time I asked about it I got answer that it's hard to know that workdir is no longer needed (just from dependency tree) Mar 15 16:15:28 and as Jay7 said 12:24:09 < Jay7> anarsoul: and set BB_SCHEDULER=completion in local.conf Mar 15 16:15:35 already done Mar 15 16:16:05 it should advise scheduler to do it a bit sooner, but still not soon enough to be optimal Mar 15 16:17:29 JaMa|Off: I'm not a oe developer, I'm just a user, and I don't understand its internals Mar 15 16:17:54 and I don't want to, I have another things to hack Mar 15 16:18:34 neither do I to so low level.. that's why I just build in multiple steps (each consuming only as much work space I have available) Mar 15 16:19:21 and my /tmp partition where my workdir is, is just 10G so I know the pain :) Mar 15 16:19:37 I there a filed bug about this issue? Mar 15 16:20:19 don't know.. I just asked here and got reasonable answer Mar 15 17:29:58 yeah, it's not Mar 15 17:29:59 ga Mar 15 18:02:58 /msg denix at this point we wait for mickey to reply :) Mar 15 18:10:31 Crofton: no problem :) Mar 15 18:19:20 on what? Mar 15 18:24:15 we need to set at ime for the meeting Mar 15 18:25:21 right. florian's response though is also missing Mar 15 18:25:58 so i take it you're all in rather for weekdays than for weekends? Mar 15 18:26:00 mickeyl: good evening Mar 15 18:26:05 hey pb_, good evening Mar 15 18:27:48 yeah Mar 15 18:27:55 denix, is not a morning person :) Mar 15 18:28:03 heh Mar 15 18:39:03 I'm not a morning person, but sometimes there are too many work meetings scheduled for mornings... :) Mar 15 19:32:20 It ain't easy moving from one set of irc clients to another, heh Mar 15 20:03:03 Tartarus, koen fixed my policykit-gnome problem another way Mar 15 20:03:17 So it wasn't the EXTRA_OECONF bit? Mar 15 20:03:22 well Mar 15 20:04:04 he change the = to a += Mar 15 20:04:14 I ahd disbled instrospection directly Mar 15 20:07:37 yeah Mar 15 20:07:44 thinko in the original recipe I think Mar 15 20:07:54 since i'm 99% sure I just copied the old one and fixed up checksums Mar 15 20:07:57 but maybe I didn't Mar 15 20:08:02 long long, um... week? :) Mar 15 20:08:37 :) Mar 15 20:28:42 Can anyone help me figure out why bitbake insists on including things from tmp/sysroots/i686-linux/usr/include/? Mar 15 20:29:30 atlanta800: sure Mar 15 20:29:32 What recipe? Mar 15 20:29:49 I'm working on a omniORBpy recipe and I have a native and non-native version of it, so there's duplicate headers, and the ones in the native versions aren't patched (which they shouldn't be) Mar 15 20:31:50 Recipe: http://pastebin.com/UyjtRBc2 and the error: http://pastebin.com/wcfqLQjj Mar 15 20:33:22 "-I/home/ossie/OE/build/tmp/sysroots/i686-linux/usr/include" just plain shouldn't be there Mar 15 20:38:18 hm Mar 15 20:38:45 Well, there's a few things Mar 15 20:39:03 BBCLASSEXTEND? Mar 15 20:39:12 That too Mar 15 20:39:16 And no new do_stages Mar 15 20:39:17 :) Mar 15 20:39:18 But.. Mar 15 20:39:50 The first two exports shouldn't be needed in do_compile Mar 15 20:40:25 yeah, I think they were leftover from some other problems I've been having, figured I'd clean everything up when I got it working Mar 15 20:40:29 Second, peek at do_configure Mar 15 20:40:32 BBLCASSEXTEND? Mar 15 20:40:35 to see where it's pulling in the host one from Mar 15 20:40:45 how do I do that? Mar 15 20:40:52 I am having a serious flashback about he long doubel stuff Mar 15 20:41:36 @Crofton|work: yeah, there's a patch for it, but the native i686 files don't get patched Mar 15 20:41:51 I think 686 should have long double Mar 15 20:41:55 but arm does not Mar 15 20:41:58 correct Mar 15 20:42:33 so the 686 files have them, but the arm files get it patched out, but the recipe seems to want to pull the 686 headers Mar 15 20:42:39 I thought that got fixed in omniorb Mar 15 20:42:44 how do I peek at do_configure? Mar 15 20:44:35 omniorb-native doesn't apply the rm_LongDouble.patch Mar 15 20:45:06 go into the work directry and look at the config.log Mar 15 20:46:51 and see if you can figure out what test is adding the i686 path Mar 15 20:51:09 running 'grep "\/i686-linux\/usr\/include" config.log' turns up nothing Mar 15 20:51:31 and looking over it manually doesn't turn anything out of the ordinary Mar 15 20:52:23 if you look in the Makefile, does that give you any clues Mar 15 20:52:40 03Tom Rini  07master * r8e8b299294 10openembedded.git/recipes/guile/guile-native.inc: Mar 15 20:52:40 guile-native: Put gettext-native back into DEPENDS Mar 15 20:52:40 guile must have gettext to build. Mar 15 20:52:40 Signed-off-by: Tom Rini Mar 15 20:54:08 unfortunately there is no Makefile, just a GNUmakefile, and I'm unfamiliar with how it works, it includes dir.mk files but I've already investigated those Mar 15 20:54:49 not using autotools? Mar 15 20:56:11 No? I'm not very acquainted with such things. Mar 15 20:56:22 I knew I gave up on omnipy for a reason :) Mar 15 20:56:35 I think it ussed autotools Mar 15 20:56:45 you are including the autotools class Mar 15 20:57:03 I've actually gotten it to work by severely hacking it in the past, but didn't document my work in the past, plus I'm looking for a proper solution now Mar 15 20:57:33 03Koen Kooi  07org.openembedded.dev * r5696a64b1c 10openembedded.git/recipes/opencv/opencv_2.2.bb: Mar 15 20:57:33 opencv 2.2: bump SRCREV to where the patch has been applied upstream Mar 15 20:57:33 Signed-off-by: Koen Kooi Mar 15 20:57:44 atlanta800: GNUmakefile is just a different filename for Makefile.. it's not special Mar 15 20:58:12 kergoth: gotchya Mar 15 20:58:12 the one difference of course is that bsd's make won't even see it, so its clearly indicating that it requires gnu make, but it's quite common to use gnu make specific functionality in Makefile as well.. Mar 15 20:58:51 well I've got to run to class, I'll have to check back later Mar 15 21:17:05 Tartarus: do you have any comments or concerns about Ben's patches to adjust UBI image naming? Mar 15 21:17:58 testing them now Mar 15 21:18:02 have you gotten a build done? Mar 15 21:18:19 (and any chance you've been able to toss the image on hw?) Mar 15 21:19:44 Tartarus: wouldn't be able to test in on hw, sorry Mar 15 21:20:40 Tartarus: as of build test, I'm still working on it - requires switching several components back and forth, due to angstrom-2008.1 removal... Mar 15 21:22:04 ppl Mar 15 21:22:26 I need someone with Intel 64-bit arch and 64-bit Linux :) Mar 15 21:22:35 please show me output of readelf -h `which readelf` Mar 15 21:22:44 Machine field Mar 15 21:24:44 Jay7: http://pastebin.com/mx9FHRNM Mar 15 21:25:49 well, Advanced Micro Devices X86-64 as well Mar 15 21:30:20 yes, it's either intel 80386 or amd x86-64 :) Mar 15 21:30:28 no matter what cpu Mar 15 21:34:31 load average: 2511.07, 2512.97, 2515.37 Mar 15 21:34:32 hehe.. Mar 15 21:34:42 eeks Mar 15 21:34:47 solaris on sparc's :) Mar 15 21:34:58 still reachable :) Mar 15 21:35:31 amazing... how many cpus does ist have actually? Mar 15 21:36:10 4 cpu Mar 15 21:41:22 denix: k, i kicked off a helloworld-image build but as I was telling kergoth I'm continuing to fail to learn my lesson and did it on my work laptop Mar 15 21:41:31 someone saw this? Mar 15 21:41:33 i'll at least keep a good eye on the actual image logs Mar 15 21:41:33 http://trac.cross-lfs.org/ Mar 15 21:42:43 hm they can build on bsd and solaris okay Mar 15 21:43:25 I know we reference LFS stuff in some recipes Mar 15 21:43:37 and yeah, it's just here's the configure arguments and so forth, with the why Mar 15 21:43:46 unless i'm missing their higher level bits Mar 15 22:02:38 And of course I commit a mysql fixup in the branch I have those changes in Mar 15 22:02:44 so they better complete so I can push all 3 :) Mar 15 22:06:23 :) Mar 15 22:09:33 good evening Mar 15 22:09:47 hi ericben Mar 15 22:09:59 hi woglinde Mar 15 22:14:04 hi GNUtoo|laptop Mar 15 22:27:03 ericben, hi Mar 15 22:31:31 03Andrea Adami  07org.openembedded.dev * r01aa5fd0af 10openembedded.git/recipes/linux/linux-kexecboot/ (6 files in 6 dirs): Mar 15 22:31:31 linux-kexecboot_git: update Zaurus defconfigs and prepare for 2.6.38 Mar 15 22:31:31 Signed-off-by: Andrea Adami Mar 15 22:31:44 03Andrea Adami  07org.openembedded.dev * r54e9249e6a 10openembedded.git/recipes/linux/linux-2.6.37/ (4 files in 4 dirs): Mar 15 22:31:45 linux-2.6.37: add more Zaurus defconfigs. Mar 15 22:31:45 * adding akita c7x0 and poodle Mar 15 22:31:45 * editing spitz to allow boot from CF/HDD Mar 15 22:31:45 Signed-off-by: Andrea Adami Mar 15 22:36:44 Debian GNU/Linux 6.0 (squeeze) /Advanced Micro Devices X86-64 Mar 15 22:36:45 AMD Phenom(tm) II X6 1055T Processor @2.8GHz, 6 cores /4Gb Mar 15 22:36:52 got this by automated way Mar 15 22:37:12 lsb_release + lot of awk'ing over /proc :) Mar 15 22:49:53 http://www.hy-research.com/scalex9-volunteer.html Mar 15 22:54:29 We need a larger OE poster on that side of the Atlantic ocean ;-) Mar 15 22:56:53 ah, florian is hiding behind flo_lap... :) Mar 15 22:58:22 oops sorry :-) Mar 15 22:58:48 yes... I took the chance to return to the console for a while. Mar 15 22:59:05 And to answer the most important emails Mar 15 23:10:24 florian: congratulations, btw! Mar 15 23:13:34 denix: ah... my mail made it :-) many thanks! Mar 15 23:20:52 mickeyl: do you know what's up with the openmoko.org mail server? it seems to have been offline for a while. Mar 15 23:25:43 pb_: unfortunately the situation is unclear. the two admins have been informed, but they're incommunicado atm. I'll try to reach Harald who has the general access for all the infrastructure. Mar 15 23:25:49 (LaForge, that is) Mar 15 23:25:53 congrats florian!! Mar 15 23:28:11 mickeyl: many thanks! Mar 15 23:28:37 florian: heh thats first time probably it will grow bigger Mar 15 23:28:54 florian: its expensive to print things here Mar 15 23:29:17 03Andrea Adami  07org.openembedded.dev * r7725ee2508 10openembedded.git/recipes/linux/ (8 files in 7 dirs): Mar 15 23:29:17 linux-kexecboot: remove 2.6.36 version used only for Zaurus devel. Mar 15 23:29:17 Signed-off-by: Andrea Adami Mar 15 23:31:39 03Andrea Adami  07org.openembedded.dev * ra1f31cb141 10openembedded.git/recipes/linux/ (8 files in 7 dirs): Mar 15 23:31:40 linux-kexecboot: remove Zaurus support in 2.6.37. Mar 15 23:31:40 Signed-off-by: Andrea Adami Mar 15 23:33:35 mickeyl: righto, fair enough. I just suddenly noticed loads of openmoko.org mail in the lists.oe.org queue. Mar 15 23:35:56 hiya pb_ Mar 15 23:37:25 congrats Florian Mar 15 23:38:11 * khem wants to congrat Florian too but is unaware of context Mar 15 23:38:44 :-) Mar 15 23:38:48 yeah Mar 15 23:38:56 khem: you said my words :) Mar 15 23:39:31 I got a son about a week ago Mar 15 23:39:35 ah Mar 15 23:39:45 well, now contgrats :) Mar 15 23:39:49 Congrats ! welcome to club Mar 15 23:40:07 its good stuff. Mar 15 23:40:46 khem: its my third kid :-) Mar 15 23:42:04 oh hmmm Mar 15 23:44:00 florian: what size poster do u have printed for OE Mar 15 23:49:03 khem: i think it is A0 sized Mar 15 23:49:56 florian: third one? you are the man! I have 2 sons and I'm overwhelmed... :) Mar 15 23:50:13 46.8 x 33.1 in Mar 15 23:55:36 denix: yep... this will be fun, but i guess we'll get used to it Mar 15 23:57:44 i remember there is a well known kernel hacker with five kids Mar 15 23:58:10 (no i'm not going to compete with him ;-) Mar 15 23:58:16 it's impossible Mar 15 23:59:13 I can do something useful at home only after kid is gone to bed :) Mar 15 23:59:26 right Mar 15 23:59:28 even better after wife gone to bed too :) Mar 16 00:00:04 that's why i work at the office even if it is not strictly necessary Mar 16 00:00:21 * Jay7 have office now too because of this Mar 16 00:01:22 oh, how familiar... :) Mar 16 00:02:19 OE "kindergarten" :) Mar 16 00:03:03 good one :) Mar 16 00:03:54 03Andrea Adami  07org.openembedded.dev * r07aeb2dd3b 10openembedded.git/recipes/linux/ (7 files in 7 dirs): Mar 16 00:03:55 linux-kexecboot: add 2.6.38 defconfigs for Zaurus family. Mar 16 00:03:55 * Keep DEFAULT_PREFERENCE = "-1" until we have 2.6.38 as default. Mar 16 00:03:55 Signed-off-by: Andrea Adami Mar 16 00:04:06 florian, me too..9yr olds don't understand "daddy's working" Mar 16 00:06:34 ka6sox-home: it seems to take more than 30 years to learn this :-) Mar 16 00:07:04 * florian need some sleep Mar 16 00:07:11 good night Mar 16 00:07:15 9yrs is old enough to start working in a coal mine... that should garner an understanding of "daddy's working" Mar 16 00:08:06 hmm... its a good time to start learning C :-) Mar 16 00:08:25 gn Mar 16 00:09:06 my son is making and animating 3D figures and using Blender... Mar 16 00:09:09 its SCARY. Mar 16 00:09:14 grg, yup agreed. Mar 16 00:14:53 03Andrea Adami  07org.openembedded.dev * r497bf2d5bf 10openembedded.git/recipes/linux/linux_2.6.38.bb: Mar 16 00:14:53 linux: commit initial 2.6.38 recipe. Mar 16 00:14:53 Signed-off-by: Andrea Adami Mar 16 00:49:22 03Andrea Adami  07org.openembedded.dev * rb12825f762 10openembedded.git/recipes/linux/ (5 files in 5 dirs): Mar 16 00:49:23 linux: add 2.6.38 initial defconfigs for c7x0, poodle, akita, spitz. Mar 16 00:49:23 * collie and tosa following soon Mar 16 00:49:23 Signed-off-by: Andrea Adami Mar 16 00:59:31 03Yuri Bushmelev  07master * rce869976ec 10openembedded.git/contrib/testing/testbuilder/ (6 files in 2 dirs): contrib/testing/testbuilder: Update to 0.6 Mar 16 01:03:38 well.. Mar 16 01:03:41 * Jay7 -> sleep Mar 16 01:11:46 nite Jay7 Mar 16 01:13:23 khem working mostly on bitbake? Mar 16 01:13:35 I've got a q about the exit code for bitbake Mar 16 01:13:43 namely what makes it return a non-zero value Mar 16 01:14:52 http://pastebin.com/J86EJCCM is our buildbot output, but it's returning a non-zero val on building our image... is it because of the initial Angstrom does not support... stuff? Mar 16 01:15:14 if I build smaller targets, like openjdk-6, it will return 0, fine Mar 16 01:16:52 but I'm wondering if something like do_rootfs is failing and triggering the non-zero exit code Mar 16 01:22:47 jconnolly: qa errors ? Mar 16 01:22:59 jconnolly: what return value does it return Mar 16 01:23:00 that's what I was thinking khem Mar 16 01:23:03 let me see Mar 16 01:23:28 nothing returning qa errors in that particular target Mar 16 01:24:02 perhaps ERROR: Multiple .bb files are due to be built which each provide virtual/psplash (/opt/buildbot/slave/build_R21_incremental/build/oe-buglabs/recipes/psplash/psplash_svn.bb /opt/buildbot/slave/build_R21_incremental/build/oe-buglabs/recipes/psplash/psplash-angstrom_svn.bb). Mar 16 01:24:24 let me see if that's common to other non-zero exit codes, and also see what the return value is for me Mar 16 01:24:32 exit code 1 Mar 16 01:28:25 let's say, for example, that I bitbake foo, which is a dependency for bar, if I bitbake foo and it has qa issues, then I bitbake far, should I expect a non-zero exit code for bitbaking bar? Mar 16 01:28:48 (btw, bitbaking foo with qa-issues, still returns a 0 exit code) Mar 16 01:30:01 the strange thing is that i see: NOTE: Tasks Summary: Attempted 12272 tasks of which 4294 didn't need to be rerun and 0 failed. Mar 16 01:30:16 but looking at the exit code: program finished with exit code 1 Mar 16 01:30:21 (bitbake) Mar 16 01:31:14 The exit status from bitbake is known to be questionable; not only does it vary based on bitbake version IIRC, but there is some disparity in terms of opinions when it comes to determining what that value should be in various cases. Mar 16 01:31:20 jconnolly: that multiple .bb is a real problem Mar 16 01:31:23 Best off to ignore the exit status altogether. Mar 16 01:31:27 is it because we're using Angstrom 2008 that makes this so arcane? Mar 16 01:31:31 you need to solve that even if the build succeeded Mar 16 01:31:41 mwester-laptop: eh Mar 16 01:31:52 ok, let me see in our logs if that's common to all non-zero exit codes Mar 16 01:31:54 This is not specific to Angstrom; in general this is an issue. Mar 16 01:32:22 indeed, it seems to be Mar 16 01:32:31 at least for the last few days Mar 16 01:32:42 It's been a lot longer than that, actually. :) Mar 16 01:33:19 khem: indeed, seems to be common for the last few eeks Mar 16 01:33:21 *weeks Mar 16 01:33:26 But I can see the other points of view -- there are those who would choose to ignore QA warnings, and would not expect a non-zero exit in the face of those. Mar 16 01:33:38 should I define a PREFFERED_PROFIDER somewhere? Mar 16 01:33:44 jconnolly: yes Mar 16 01:33:52 well usually its more than that Mar 16 01:34:13 jconnolly: problem is that there is a contention and these two recipes provided something which is not common Mar 16 01:34:20 so the other gets pulled in Mar 16 01:34:36 jconnolly: your best bet is to do bitbake -g Mar 16 01:34:46 yes, I'm familiar with that, but not in identifying what is not common. Mar 16 01:34:52 then analyse the .dot files to track deps thats pulling both in Mar 16 01:35:12 ok generate those .dots and upload somewhere I can have a quick look Mar 16 01:35:50 khem: that's a giant nightmare if I look at the graphviz represenation... Mar 16 01:36:02 is there a way that you use to analyze the .dots ? Mar 16 01:36:06 just curious ;D Mar 16 01:37:13 jconnolly: I open them in vi and look for stuff Mar 16 01:37:30 ;) You could use a real text editor, too. Mar 16 01:37:38 on my machine nothing can open them in graphics Mar 16 01:37:44 good to know. tho how you use vi to bounce back/forth with dependencies is enviable Mar 16 01:37:49 mwester-laptop: which one ? Mar 16 01:37:55 mwester-laptop: are you on emacs camp Mar 16 01:38:00 uh oh Mar 16 01:38:06 vi/emacs war Mar 16 01:38:08 * jconnolly ducks Mar 16 01:38:14 vi is to text editors as grub is to an operating system. Mar 16 01:38:24 mwester-laptop: I am equally proficient in both Mar 16 01:38:33 mwester-laptop: I just use whatever is there Mar 16 01:38:53 That's enviable. I'm just competent enough with vi to get emacs installed. Mar 16 01:38:54 mwester-laptop: and emacs is like OS Mar 16 01:38:57 It is. Mar 16 01:39:00 :) a real one I think Mar 16 01:39:16 That's why I use emacs on Windows -- it's better. Mar 16 01:39:30 you can read email newsgroups probably do skype these days Mar 16 01:39:34 its insanity Mar 16 01:39:44 I will generate a .dot file for my target image and get back to you, khem, thanks for responding Mar 16 01:39:49 ugh, emacs on windows Mar 16 01:39:56 I stick to simple stuff, though. Like writing Perl scripts to parse .dot files.... Mar 16 01:40:05 * khem use notepad++ on windows Mar 16 01:40:43 * jconnolly would probably use eclipse Mar 16 01:41:09 eclipse will eclipse your resources like its name Mar 16 01:41:26 indeed, but I bought an i7 + 16GB ram recently Mar 16 01:41:34 try to eclipse that, eclipse Mar 16 01:41:40 http://pastebin.com/fB0kjALL Mar 16 01:41:45 ($999 usd) Mar 16 01:41:49 ^ jconnolly Mar 16 01:42:26 mwester-laptop: nice, I'll have a look Mar 16 01:42:50 jconnolly: I dont have a 3 year old laptop which is moderately powerful in current context Mar 16 01:43:05 redirect the output to xyz.html, and point your browser at it, jconnolly -- easier than grepping or searching. Mar 16 01:44:06 * mwester-laptop notices the pastebin got truncated Mar 16 01:45:18 http://pastebin.com/e5XnURUY Mar 16 01:45:22 That's better. Mar 16 02:08:53 03Ben Gardiner  07master * rcfde49e8d0 10openembedded.git/classes/image.bbclass: (log message trimmed) Mar 16 02:08:53 image.bbclass: add link to .rootfs. or ..img Mar 16 02:08:53 The current image link-creation code will unconditionaly create a link Mar 16 02:08:53 from . pointing to the .rootfs. output. Mar 16 02:08:53 This is not compatible with the UBIFS images produced which have Mar 16 02:08:53 ..img extension since they are not considered to be valid rootfs Mar 16 02:08:54 images when they are not included in a UBI container. Mar 16 02:09:02 03Ben Gardiner  07master * r7308e68fc2 10openembedded.git/conf/bitbake.conf: (log message trimmed) Mar 16 02:09:02 bitbake.conf: use .ubifs.img extension in IMAGE_CMD_ubi Mar 16 02:09:02 The current IMAGE_CMD_ubi creates an interim image with a .rootfs.ubifs Mar 16 02:09:02 extension. Mar 16 02:09:02 The ubifs image created is not considered a valid rootfs without a UBI Mar 16 02:09:03 container. Mar 16 02:09:04 Change the filename of the iterim ubifs image used by IMAGE_CMD_ubi to Mar 16 02:09:05 03Tom Rini  07master * r3eafc1796e 10openembedded.git/recipes/mysql/mysql5-native_5.1.40.bb: Mar 16 02:09:05 mysql5-native: Drop PSTAGING_DISABLED Mar 16 02:09:05 In 1775a986a1c063198ddacea0575623c0ef8c6de3 we dropped the rm_work hack Mar 16 02:09:06 as not being applicable anymore. That hack was also why I had added Mar 16 02:09:06 PSTAGING_DISABLED but have confirmed it's not needed now. Mar 16 02:09:07 Signed-off-by: Tom Rini Mar 16 02:28:14 So, I almost hate to admit this Mar 16 02:28:22 but I'm getting fond of bitbake -u depexp -g Mar 16 02:28:25 rather than fiddling with the dot file Mar 16 02:30:12 But maybe I need to take some remedial crotchty old guy tech guy classes, I also finally gave up on screen+something for IRC and switched to proxies + pidgin Mar 16 02:46:39 Gah! GUIs! Mar 16 02:47:03 Real computers don't have GUIs. Mar 16 02:50:50 Does anyone have a Sheevaplug? And can anyone confirm that the generic linux kernel recipe is, in fact, the correct one for OE to be selecting when building a kernel for that machine? Mar 16 02:51:03 (It used to be linux-kirkwood) **** ENDING LOGGING AT Wed Mar 16 02:59:57 2011