**** BEGIN LOGGING AT Fri Jul 23 02:59:56 2010 Jul 23 05:18:15 03David Lanzendörfer  07org.openembedded.dev * r2149b28874 10openembedded.git/recipes/maxima/maxima_5.21.1.bb: Jul 23 05:18:15 maxima: add recipe, tool for algebraic calculation Jul 23 05:18:15 Signed-off-by: Martin Jansa Jul 23 05:18:17 03Martin Jansa  07org.openembedded.dev * r95cc60009e 10openembedded.git/recipes/tasks/task-shr-feed.bb: task-shr-feed: add maxima Jul 23 05:43:30 btw rm_work is not removing some files ie 2.5G work/armv5te-angstrom-linux-gnueabi/webkit-gtk-1.3.1+svnr60413-r60413-r1/.libs Jul 23 05:50:41 JaMa: there was a patch for that Jul 23 06:08:09 03Khem Raj  07org.openembedded.dev * r5ef0077927 10openembedded.git/ (6 files in 3 dirs): Jul 23 06:08:09 distro: Add new feature to indicate thumb-interworking. Jul 23 06:08:09 * This would be used to decide if we can use BX in return Jul 23 06:08:09 instructions in uclibc. Jul 23 06:08:09 Signed-off-by: Khem Raj Jul 23 06:22:14 03Pramod Poudel  07org.openembedded.dev * r9353e56c0e 10openembedded.git/recipes/opencv/opencv-dsp-acceleration_svn.bb: Jul 23 06:22:14 opencv-dsp-acceleration: Revision changed to 128 Jul 23 06:22:14 *IMGLIB_PATH is changed to path where other TI tools are found. Jul 23 06:22:14 Signed-off-by: Pramod Poudel Jul 23 06:22:14 Signed-off-by: Koen Kooi Jul 23 06:57:01 morning Jul 23 06:58:22 good morning Jul 23 07:10:34 hi lrg, hi hrw Jul 23 07:10:45 hey florian, hrw Jul 23 07:21:57 hi people Jul 23 07:22:43 hi Jul 23 07:22:50 hey JaMa hrw Jul 23 07:26:43 hi hrw Jul 23 07:26:49 hi jama Jul 23 07:26:52 hi pb Jul 23 07:28:01 03Khem Raj  07org.openembedded.dev * rd9375a9283 10openembedded.git/recipes/yaffs2/ (yaffs2-utils.inc yaffs2-utils_cvs.bb): Jul 23 07:28:01 yaffs2: Remove legacy staging (from poky) Jul 23 07:28:01 Signed-off-by: Khem Raj Jul 23 07:28:05 03Khem Raj  07org.openembedded.dev * r4c5fadef88 10openembedded.git/recipes/ (zaurus-utils/encdec-updater-native.bb zbedic/libbedic_1.1.bb): Jul 23 07:28:05 libbedic_1.1.bb, encdec-updater-native.bb: Remove legacy staging Jul 23 07:28:05 Signed-off-by: Khem Raj Jul 23 07:29:31 he khem Jul 23 07:29:51 khem did you bump gcc-4.5? Jul 23 07:30:29 5.5h on battery after whole night in sleep mode is nice Jul 23 07:30:38 hrw yes Jul 23 07:32:45 woglinde: he did Jul 23 07:36:03 cool Jul 23 07:58:38 is it possible to build a x86_64 rootfs with oe Jul 23 08:00:21 nikon should be Jul 23 08:00:32 but is not tested widly Jul 23 08:07:36 re hrw Jul 23 08:09:09 re Jul 23 08:14:33 jo mickeyl Jul 23 08:49:45 kgilmer are you on tour? Jul 23 08:50:26 good morning Jul 23 08:50:32 hi mckoan Jul 23 09:26:53 anybody know a working configuration machine based on PXA with OE, looks like all are broken. Jul 23 09:27:32 hm why? Jul 23 09:27:46 woglinde_: errors Jul 23 09:28:27 03Nicolas Boichat  07org.openembedded.dev * r08b3fcb51d 10openembedded.git/recipes/libpcap/ (libpcap-1.1.1/ieee80215-arphrd.patch libpcap_1.1.1.bb): (log message trimmed) Jul 23 09:28:27 libpcap 1.1.1 recipe. Jul 23 09:28:27 Notes about patches from 1.0.0: Jul 23 09:28:27 - ieee80215-arphrd.patch: ported to 1.1.1 Jul 23 09:28:27 - aclocal.patch: does not seem to be needed (unsure) Jul 23 09:28:41 woglinde_: I tried MACHINE=cm-x270 and akita Jul 23 09:29:17 gnu-config_git.bb do_patch failed Jul 23 09:29:21 03Nicolas Boichat  07org.openembedded.dev * rac8bb5b111 10openembedded.git/recipes/tcpdump/ (5 files in 2 dirs): Jul 23 09:29:22 tcpdump-4.1.1 Jul 23 09:29:22 Ported most patches from 4.0.0, except no-ipv6-tcpdump4.patch (fixed upstream). Jul 23 09:29:22 Signed-off-by: Koen Kooi Jul 23 09:29:55 moreover today we face to TMPDIR has changed ABI (4 to 5) Jul 23 09:29:59 doh Jul 23 09:30:09 *sighÜ why nobody yet converted libpng to bblass extended Jul 23 09:31:29 mckoan: I built for armv5te two days ago before remove-cross commits Jul 23 09:31:53 pxa255 Jul 23 09:31:57 koen reported a multi machine failure this morning Jul 23 09:32:33 yes **** BEGIN LOGGING AT Fri Jul 23 15:16:14 2010 Jul 23 15:16:27 khem: lsmod shows no interface modules loaded Jul 23 15:17:57 ls /sbin/ifconfig returns nothing Jul 23 15:19:54 it is the minimal distro so maybe networking is not included Jul 23 15:20:51 and my attempts to add packages to a distro have met with failure so if anyone can provide advice or a link I would be grateful Jul 23 15:21:12 maybe I just need to add net-tools or networkmanager Jul 23 15:21:40 boot a livecd to find out what module to include for the nic Jul 23 15:30:40 you don't need networkmanager, you just need the kernel support for your nic and an entry in /etc/network/interfaces (which i'm pretty sure already defaults to dhcp for eth0, so you just need the device to exist) Jul 23 15:44:41 whoa Jul 23 15:44:52 sh: stage-manager-ipkg: command not found Jul 23 15:44:53 ERROR: Couldn't install the staging package to a temp directory Jul 23 15:44:58 what the heck causes THAT!? Jul 23 15:45:15 I was trying to bitbake -cconfigure virtual/kernel Jul 23 15:46:09 I won the digikey twirvia thursday... wonder what the prize is Jul 23 15:48:05 can someone try a test commit on git? Jul 23 16:10:59 kergoth: don't I need 'iface' at least? Jul 23 16:12:10 does anyone know how to make a kernel recipe copy the uimage to flash when doing an opkg upgrade? i tried to add a pkg_postinst_kernel, but it didn't seem to work. Jul 23 16:16:13 re Jul 23 16:16:58 karmakop, huh? Jul 23 16:27:15 anyone know of a functional equivalent of gtk-recordmydesktop in oe? Jul 23 16:34:10 ERROR: QA Issue with staging: libssh2.pc failed sanity test (tmpdir) in path /ninja/build/tmp/sysroots/armv7a-angstrom-linux-gnueabi/usr/lib/pkgconfig; if i delete libssh2.pc; i'm able to finish the staging... Jul 23 16:34:29 libssh2.pc looked ok Jul 23 16:34:37 no Jul 23 16:34:47 /ninja/build/tmp/sysroots/armv7a-angstrom-linux-g Jul 23 16:35:01 use inherit pkgconfig Jul 23 16:35:04 in recipe Jul 23 16:35:12 jconnolly: can you use x11vnc to view display remotely, and then record it on the desktop? Jul 23 16:35:23 same error with and without inheriting pkgconfig Jul 23 16:35:33 cbrake: interesting, hadn't though of that, that works Jul 23 16:36:07 should matter if i inherit autotools first? (inherit autotools pkgconfig) Jul 23 16:39:43 nope Jul 23 16:40:41 warflyr hm Jul 23 16:40:47 is git now back again? Jul 23 16:41:19 i havent tried git today o.0 Jul 23 16:41:38 thought of rebuilding any more packages is terrifying Jul 23 16:48:16 is nobody committing or is CIA broken? Jul 23 16:58:09 I did an Angstrom build, but didn't get pciutils. I then added "pci" to MACHINE_FEATURES, but while bitbake base-image built the package, it still isn't included in the root tarball Jul 23 16:58:22 what do I need to do to get base-image to pick it up? Jul 23 16:58:48 ka6sox the git reops is borken Jul 23 16:59:04 its somehow restettet to 3 weeks ago Jul 23 17:14:54 wtf Jul 23 17:14:55 fatal: bad tree object 8761827020c5f2fb4b92e9375048f1188125f31d Jul 23 17:14:56 error: waitpid (async) failed Jul 23 17:14:56 fatal: git-upload-pack: aborting due to possible repository corruption on the remote side. Jul 23 17:14:56 remote: Total 0 (delta 0), reused 0 (delta 0) Jul 23 17:15:02 * kergoth_ gets that trying to pull now -- weird.. that's a new one Jul 23 17:15:25 * kergoth_ checks git fsck Jul 23 17:17:21 kergoth: git is whacked I guess Jul 23 17:17:25 kergoth didnt you read what I wrote? Jul 23 17:17:29 I could not pull Jul 23 17:17:35 there was some problem on the server again Jul 23 17:17:52 which resete git to backup from 3 weeks ago Jul 23 17:17:52 how many hours ago was that? i didn't read the full scrollback, no Jul 23 17:18:06 18:58 < woglinde_> ka6sox the git reops is borken Jul 23 17:18:06 18:59 < woglinde_> its somehow restettet to 3 weeks ago Jul 23 17:18:36 I think something worse than that is happening, git seems to think its corruption, not just an old repo Jul 23 17:18:42 old repo would be fine, just push to update it again Jul 23 17:19:12 hm thats what jama told as he tried to pull Jul 23 17:19:24 ka6sox any chance to fix it? Jul 23 17:20:02 * kergoth_ tests something Jul 23 17:20:42 woglinde_, a fresh clone doesn't even work. Jul 23 17:21:00 as far as i know, you could clone 3 weeks ago :P Jul 23 17:21:48 hm okay Jul 23 17:22:20 so we have to wait if ka6sox fixes it again Jul 23 17:33:17 khem ping Jul 23 17:33:37 ping ping khem Jul 23 17:33:39 hms Jul 23 17:49:13 nobody has any advice for me? how do I get "bitbake base-image" to include a package that wasn't present the first time I ran it? Jul 23 17:49:31 the image always rebuilds from current packages. Jul 23 17:49:43 if you added a package to IMAGE_INSTALL, it'll take effect right away Jul 23 17:49:51 if you added it to a task, like task-base, you'd have to clean and rebuild the task Jul 23 17:49:57 woglinde_: hello Jul 23 17:50:04 RP: around ? Jul 23 17:50:16 RP: the patch I posted is easier Jul 23 17:50:24 kergoth_: I added "pci" to MACHINE_FEATURES, and now pciutils is built (good), but not included when I re-run task-base Jul 23 17:50:26 having trouble with in my openembedded repo this morning Jul 23 17:50:41 I dont have fast machine so turnaround time is a bit larger Jul 23 17:50:42 hollisb, no recipes magically rebuild when variables change Jul 23 17:50:46 hollisb, they obey the stamps Jul 23 17:50:47 error message says ...remote: aborting due to possible repository corruption on the remote side. Jul 23 17:50:47 error: waitpid (async) failed Jul 23 17:50:47 fatal: error in sideband demultiplexe Jul 23 17:51:05 hollisb, if task-base was previously built, it doesn't know to rebuild when MACHINE_FEATURES changes. do a bitbake -c clean task-base, then bitbake base-image Jul 23 17:51:39 * Jay7 is back Jul 23 17:51:40 kergoth_: ok. I hope that will be quick, i.e. not rebuilding all the packages from source? Jul 23 17:51:41 does bitbake -c clean; clean all sources? Jul 23 17:51:46 ? Jul 23 17:51:51 -c clean cleans *one* recipe, the one you request Jul 23 17:51:55 which is a task -- it builds nothing Jul 23 17:52:05 great, thanks Jul 23 17:52:09 warflyr, it doesn't clean out files that were downloaded Jul 23 17:52:13 but the unpacked, patched, etc stuff yes Jul 23 17:52:17 hollisb, np Jul 23 17:52:20 ah Jul 23 17:52:51 hollisb, task recipes just emit an empty binary package that depends upon other packages, pretty much a no-op Jul 23 17:53:22 kergoth_: bitbake talks so much about detecting obsolete caches that I thought it automatically did the right thing when changing configuration files Jul 23 17:53:27 like Linux does with .config Jul 23 17:54:01 the cache just holds specific vars so it doesn't need to reparse all the recipes up front every time Jul 23 17:54:21 it can't determine what to build without parsing every recipe that exists, to get depends, rdepends, providers, etc Jul 23 17:54:47 ok Jul 23 17:55:04 In the long term, we'll likely be moving away from the stamp concept Jul 23 17:55:21 in favor of hashes, tracking input and output files and metadata, so that changing anything thats used will automatically re-run the affected tasks Jul 23 17:55:23 but we're not quite there yet Jul 23 18:04:02 re Jul 23 18:04:11 khem do you now have some time for me? Jul 23 18:04:40 re jama and jay7 Jul 23 18:05:17 re Jul 23 18:07:39 woglinde_: sure Jul 23 18:07:41 whats up Jul 23 18:08:34 khem we have some problems at linking time where uidiv and uidivmod is not found Jul 23 18:09:11 maybee the problem comes from mixing march armv5t code with linker for armv7a Jul 23 18:09:25 woglinde_: hmmm those are in libgcc Jul 23 18:09:43 woglinde_: which package is it Jul 23 18:10:05 khem ti-codec-engine Jul 23 18:10:10 but it is bit complicate Jul 23 18:10:55 woglinde_: ok Jul 23 18:11:10 woglinde_: these functions should be there in libgcc Jul 23 18:11:14 are you linking in libgcc Jul 23 18:11:19 yes Jul 23 18:11:32 linker is opening it Jul 23 18:11:41 the files in questions are ar archivs Jul 23 18:12:06 ok Jul 23 18:12:19 can you readelf -s libgcc for me Jul 23 18:12:34 libgcc and libgcc_s? Jul 23 18:13:01 khem do you have build for beagleboard at hand Jul 23 18:13:11 try bitbake ti-codec-engine Jul 23 18:13:12 woglinde_: its building :) Jul 23 18:13:15 ugh Jul 23 18:13:27 I am trying to fix this burning issue with multimachine builds Jul 23 18:13:55 hm where is libcc cross now? Jul 23 18:15:16 /devel/arm/oetmp-ang/sysroots/i686-linux/usr/lib/gcc/arm-angstrom-linux-uclibceabi/4.5.1/libgcc.a Jul 23 18:15:28 or or Jul 23 18:15:29 /devel/arm/oetmp-ang/sysroots/armv7a-angstrom-linux-uclibceabi/lib/libgcc_s.so Jul 23 18:16:09 khem Jul 23 18:16:10 16: 00000000 0 FUNC GLOBAL HIDDEN 1 __aeabi_uidiv Jul 23 18:16:10 17: 000000a8 32 FUNC GLOBAL HIDDEN 1 __aeabi_uidivmod Jul 23 18:16:15 /devel/arm/oetmp-ang/sysroots/i686-linux/usr/lib/gcc/arm-angstrom-linux-uclibceabi/4.5.1/libgcc.a Jul 23 18:16:19 woglinde_: in natve sysroot usr/arm-oe-linux-uclibceabi/lib Jul 23 18:16:45 hidden hmmm Jul 23 18:16:58 because of neon? Jul 23 18:18:40 let me check something Jul 23 18:18:54 all other linking is working Jul 23 18:19:03 but that special case not Jul 23 18:19:23 its not hidden for me Jul 23 18:19:37 $ readelf -s libgcc_s.so.1|grep __aeabi_uidiv Jul 23 18:19:38 134: 00002718 0 FUNC GLOBAL DEFAULT 10 __aeabi_uidiv@@GCC_3.5 Jul 23 18:19:38 169: 000027c0 32 FUNC GLOBAL DEFAULT 10 __aeabi_uidivmod@@GCC_3.5 Jul 23 18:19:38 732: 00002718 0 FUNC GLOBAL DEFAULT 10 __aeabi_uidiv Jul 23 18:19:38 767: 000027c0 32 FUNC GLOBAL DEFAULT 10 __aeabi_uidivmod Jul 23 18:19:39 * kergoth_ gets those errors too Jul 23 18:19:52 but I suspect the ar archive are the problem beacuse they are compiled with -march=armv5t with any float options Jul 23 18:20:14 its uclibc here with gcc-4.5 Jul 23 18:20:22 and binutils 2.20 Jul 23 18:20:41 woglinde_: yeah seems so Jul 23 18:25:29 Engine.c:(.text+0x520): undefined reference to `__aeabi_uidiv' Jul 23 18:25:57 khem in which file shoud the be? Jul 23 18:26:30 hi Jul 23 18:26:31 http://pastebin.com/x5rJuSAp Jul 23 18:26:36 remote git corruption on oe Jul 23 18:26:37 woglinde_: are you having multiple machines Jul 23 18:26:42 in same sandbox Jul 23 18:26:45 793: 00002790 32 FUNC GLOBAL DEFAULT 10 __aeabi_idivmod Jul 23 18:26:46 794: 00003fa4 0 FUNC GLOBAL DEFAULT 10 __aeabi_ldivmod Jul 23 18:26:55 what should we do? Jul 23 18:27:00 hm maybee its using libgcc.a? Jul 23 18:27:02 GNUtoo|laptop: wait and pray Jul 23 18:27:07 ouch Jul 23 18:27:24 someone will try to fix that? Jul 23 18:27:40 I hope ka6sox is working on it Jul 23 18:27:54 17:48 < ka6sox> can someone try a test commit on git? Jul 23 18:27:54 18:48 < ka6sox> is nobody committing or is CIA broken? Jul 23 18:28:33 ah ok Jul 23 18:29:00 btw we should mirror sms sentry Jul 23 18:29:10 or recreate a tarball with the ipk (python) Jul 23 18:32:18 hm Jul 23 18:32:37 okay removing the section entry from the linkerscript works now Jul 23 18:35:13 woglinde_: how ? Jul 23 18:35:19 * khem goes to lunch Jul 23 18:36:52 khem they still have the chitty linkerscript files Jul 23 18:37:32 with COPY entry Jul 23 18:40:06 is there a variable equiv to this? ${D}/usr/share Jul 23 18:40:33 ${D}${datadir} Jul 23 18:40:37 read conf/bitbake.conf Jul 23 18:40:37 ahh datadir Jul 23 18:40:54 ahh, i was going through the manual o.0 Jul 23 18:41:07 :) Jul 23 18:42:53 if i dont inherit a class from a recipe, and dont define the function do_unpack(), will do_unpack be excuted? (considering i have a .tgz in SRCUI) Jul 23 18:46:21 do i need to inherit base? Jul 23 18:48:45 hmm Jul 23 18:49:30 I'm trying to figure out why the overo_hdq kernel module's not getting built into a package or anything on this system... bitbake -cconfigure virtual/kernel and then viewing the defconfig seems to indicate that it should be a module Jul 23 18:49:52 but bitbake -ccompile virtual/kernel isn't building it. I'm a little lost (oe-wise) in figuring this out, can anyone assit? Jul 23 18:50:53 bitbake -c menuconfig... does the module still appear to be enabled? Jul 23 18:52:06 checking. what kind of magic is being performed with this "virtual" marker? Jul 23 18:52:31 menucconfig wants to do a GUI configure, not the ncurses one Jul 23 18:53:17 yeah... i still havent figured that one out yet either o.0 Jul 23 18:53:41 well Jul 23 18:53:59 ive gone into the unpacked source directory and did make menuconfig Jul 23 18:54:09 yeah that is what I gneerally do Jul 23 18:54:39 bitbake -cconfigure virtual/kernel, then go to tmp/work/blah/linux, cp defconfig to git/.config and make ARCH=arm menuconfig Jul 23 18:54:50 make my changes, then copy it back to defconfig Jul 23 18:55:00 but in this case I didn't have to make any changes, it was already selected Jul 23 18:55:21 and yes, bitbake -c menuconfig is showing that HDQ is selected as a module Jul 23 18:55:44 is there a way to tell bitbake to just build the damnt hing again, ignore the fact that it's already been comiled once? Jul 23 18:56:10 do you see the module in tmp/deploy/glibc/ipk/armv7a Jul 23 18:56:23 i think you can -c rebuild Jul 23 18:57:09 there are no kernel modules in armv7a, they all seem to be in overo/ Jul 23 18:57:18 ahh yeah thats right Jul 23 18:57:29 there are a ton of them there, but nothing that would indicate w1 or overo_hdq Jul 23 18:58:45 and looking for overo_hdq.ko in the oe tmp/ tree doesn't show anything either Jul 23 18:59:08 it seems happy with -crebuild, trying that Jul 23 18:59:36 heh, it says it rebuilt in like 30 seconds Jul 23 18:59:43 my pontiac isn't that new Jul 23 19:01:34 well, it doesnt rebuild the entire kernel Jul 23 19:01:42 err doesnt rebuild all the sources Jul 23 19:02:08 just any changes you made and packages a new uimage Jul 23 19:05:56 tzanger look into your machine config MACHINE_FEATURES and task-base recipe Jul 23 19:15:44 warflyr: ok. I did a -cclean and -ccompile and there doesn't seem to be ANY w1 support whatsoever in there Jul 23 19:15:52 find . -name '*.ko' doesn't show anything w1 Jul 23 19:16:39 hmm Jul 23 19:16:53 hmm MACHINE is just set to 'overo', how would I track down what the MACHINE_FEATURES are getting set to based on that? Jul 23 19:17:10 oe/conf/machine/overo.conf Jul 23 19:17:19 ok that's a pretty serious 'duh' moment. :-) Jul 23 19:18:18 hmm, nothing specifically mentioning w1 or 1wire, but I'm also not sure that that's a valid machine feature Jul 23 19:18:29 still, the kernel should have the module built if the config says to Jul 23 19:19:08 yeah, you could create the machine feature in the task-base, then add that feature to the MACHINE_FEATURES Jul 23 19:19:22 yeah, i agree Jul 23 19:19:48 you could try to make the module manually by going into the source Jul 23 19:19:54 now hang on.. did a -c meuconfig again and this time it wasn't there.. turned it on and am rebuilding Jul 23 19:20:15 warflyr: I know how to do that outside of OE... is there any magic in setting the cross compilers and stuff inside of OE? Jul 23 19:20:38 ahh, where are you getting your defconfig from? Jul 23 19:20:43 tzanger o.O Jul 23 19:21:30 hmm, menuconfig doesn't seem to be saving it for the -ccompile Jul 23 19:21:37 warflyr: it's just whatever was default Jul 23 19:21:51 tzanger copy the right .config file to right defconfig Jul 23 19:22:02 woglinde_: I'm about to do that :-) Jul 23 19:22:41 ./linux-omap-psp-2.6.32/overo Jul 23 19:22:48 I hope you are using 32 Jul 23 19:22:58 no, it's an older system Jul 23 19:23:02 uh Jul 23 19:23:11 I have a newer oe ready but I don't want to try and switch everything over right atm Jul 23 19:23:28 tzanger ive had better luck with the kernel using the gumstix repo for the overo Jul 23 19:23:45 do find -name "*overo*" in recipes/linux Jul 23 19:23:48 ups Jul 23 19:24:00 put the -name "*overo*" to the end Jul 23 19:25:19 this is the gumstix repo Jul 23 19:26:11 woglinde_: what am I looking for? Jul 23 19:27:31 tzanger look for: oe//recipes/linux/linux-omap3/overo/defconfig Jul 23 19:28:26 maybe use bitbake linux-omap3 rather than the virtual package, though i dont believe there will be a different Jul 23 19:29:16 warflyr: I'm looking at recipes/linux/linux-omap3-2.6.30/defconig, replaced it with the one I added HDQ in to Jul 23 19:30:24 -cclean and -ccompile again... fun fun :-) Jul 23 19:30:49 I appreciate the help guys... it's taking me a little longer than I epected to get my 'sea legs' with oe Jul 23 19:36:42 theres alot to it, i know very little so far o.0 Jul 23 19:40:08 khem still there? Jul 23 19:41:00 I'm trying to build Angstrom image for Beagle board using bitbake base-image, but it fails with when configuring ipkg-collateral, the error info is as following: Jul 23 19:41:15 | Configuring ipkg-collateral. Jul 23 19:41:17 | Collected errors: Jul 23 19:41:19 | * check_data_file_clashes: Package ipkg wants to install file Jul 23 19:41:21 /home/robincao/gl/angstrom-setup-scripts/build/tmp-angstrom_2008_1/rootfs/base-image/usr/bin/update-alternatives Jul 23 19:41:22 | But that file is already provided by package * opkg-nogpg-nocurl Jul 23 19:41:25 | * opkg_install_cmd: Cannot install package ipkg. Jul 23 19:42:27 how can i get install to recursively copy dirs/files? (or at least not error out when it needs to copy a recursive dirctory, then i could run additional install to get the other files) Jul 23 19:42:46 Any pointers on how i can fixing this error? i'm a newbie about building system. Jul 23 19:43:34 damn google search Jul 23 19:44:14 anyone knows how to search "(COPY) :" Jul 23 19:53:54 woglinde: I am back now Jul 23 19:53:55 whats up Jul 23 19:57:04 do you remember xdc.meta (COPY) : {*(xdc.meta)} Jul 23 19:57:13 in a linkerscript SECTIONS Jul 23 19:57:24 was this the correct syntax? Jul 23 19:57:48 for newer binutils Jul 23 19:58:14 I cannt find any other example for DSECT INFO COPY Jul 23 20:06:26 woglinde: yes I do Jul 23 20:06:44 syntax seems correct Jul 23 20:07:04 how do you get these nice qualifiers @debian/developer/otavio Jul 23 20:08:06 khem: I think it's related to freenode spnsorship Jul 23 20:09:24 ah Jul 23 20:09:35 khem but there is no linker options which can merge this output with the default linkerscript? Jul 23 20:10:06 woglinde: in what way Jul 23 20:10:23 khem think of a seperate file Jul 23 20:10:28 woglinde: its upto you how you want to merge input sections into segment layour Jul 23 20:10:34 woglinde: I dont think so Jul 23 20:10:36 which has one INPUT { } Jul 23 20:10:47 there can only be one linker script Jul 23 20:11:01 I wonder why it worked with binutils 2.18 Jul 23 20:11:19 ka6sox: any news about git repo? Jul 23 20:11:29 hm the input section in a seperate file worked Jul 23 20:11:42 woglinde: hmmm Jul 23 20:11:48 INPUT( bla ) Jul 23 20:13:44 woglinde: multiple -T options should accumlate Jul 23 20:14:25 woglinde: so ld -T 1.ld -T 2.ld should add both together Jul 23 20:14:28 but it cant merge Jul 23 20:14:41 so it will append the content of second file to first Jul 23 20:14:54 woglinde: are you using multiple -T options Jul 23 20:15:13 khem no it works without -T options Jul 23 20:15:24 well try to add -T Jul 23 20:15:27 -T overrides default linkerscript what is not wanted Jul 23 20:15:44 woglinde: ok so here is what you can do Jul 23 20:15:53 dump the default linker script to some script Jul 23 20:15:56 JaMa, its broken? Jul 23 20:15:58 and then use both linker scripts Jul 23 20:15:59 the goal is to save the space for this section Jul 23 20:16:07 ka6sox: yes we cant pull or push Jul 23 20:16:19 khem both with -T? Jul 23 20:16:20 and messages are scary like corrupt SHA Jul 23 20:16:25 woglinde1: yes Jul 23 20:16:30 hm okay Jul 23 20:16:43 http://pastebin.com/2a5gFZ9P o.0 files arnt being put in the package as id expect them to... any obvious problems? Jul 23 20:16:46 woglinde1: I think -T is to override the default linker script not append to it Jul 23 20:16:59 ka6sox: yes.. http://git.openembedded.org/cgit.cgi/openembedded/tree/recipes looks empty Jul 23 20:17:02 khem -Wl,map=foo.txt right? Jul 23 20:17:09 shr@opmbuild:~/openembedded$ git pull Jul 23 20:17:09 remote: Total 0 (delta 0), reused 0 (delta 0) Jul 23 20:17:09 remote: aborting due to possible repository corruption on the remote side. Jul 23 20:17:09 error: waitpid (async) failed Jul 23 20:17:09 fatal: error in sideband demultiplexer Jul 23 20:17:13 ls --verbose Jul 23 20:17:16 ld Jul 23 20:17:27 ka6sox please use a backup Jul 23 20:17:32 we can push it up again Jul 23 20:17:33 ka6sox: the same error for new git clone Jul 23 20:17:55 ka6sox seems the only solution for it Jul 23 20:18:14 okay guys...1 at a time... Jul 23 20:18:31 *g* Jul 23 20:18:37 ka6sox: and on my 2nd box it worked for a while and now is able to say it's up-to-date Jul 23 20:18:38 woglinde1: who has the latest? Jul 23 20:18:40 bitbake@jama ~/oe $ git pull Jul 23 20:18:43 Already up-to-date. Jul 23 20:18:55 ka6sox thats not the problem Jul 23 20:18:56 commit 570b4127ed3618be84d7952ec09b4fbb094f5dca Jul 23 20:18:57 ka6sox: btw. its only git this time other servers are up Jul 23 20:19:00 is lates Jul 23 20:19:02 :) Jul 23 20:19:17 do we still have disk failures? Jul 23 20:19:24 or is it a late outcome? Jul 23 20:19:44 woglinde, yes, but it goes RO...doesnt' lose data (unless its in the middle of an exchange) Jul 23 20:20:11 okay...lets back up a minute. Jul 23 20:20:19 is that related to this ? fatal: error in sideband demultiplexer Jul 23 20:20:50 dcordes yes Jul 23 20:21:16 so we dont have stable hw at the moment Jul 23 20:21:20 not so good Jul 23 20:21:23 woglinde: any ETA on when we can pull again ? Jul 23 20:21:31 but we dont pay something Jul 23 20:21:33 *g* Jul 23 20:21:45 dcordes depends on ka6sox Jul 23 20:21:45 woglinde: when it worked after reboot for a while, it did restore right history (jumped back 3 weeks to today) Jul 23 20:21:53 ok guys Jul 23 20:22:01 jama yes you told me Jul 23 20:22:04 one by one, explain to me what is up with git Jul 23 20:22:20 crofton the git database is corrupted Jul 23 20:22:24 ok Jul 23 20:22:32 how old is th backup? Jul 23 20:22:44 suggested solution from me, install safe backup Jul 23 20:22:48 and push it up Jul 23 20:22:55 woglinde: no.. first it jumped 3 weeks back, and then it wasn't available for a while and then it got history right (this second part I didn't say to you yet IIRC) Jul 23 20:23:02 that will be good for dev, but what about branches? Jul 23 20:23:03 I should have one of the latest pulls Jul 23 20:23:25 crofton there wasn much activity Jul 23 20:23:47 yeah Jul 23 20:23:52 cgit looks ok atm Jul 23 20:23:57 or did some changed the mail stuff lately so no user branches git commits where shown Jul 23 20:24:02 is git online? Jul 23 20:24:11 its online Jul 23 20:24:15 but corrupted Jul 23 20:24:16 ok Jul 23 20:24:16 Crofton: try "tree" and 2nd level directory like recipes -> looks empty which is bad Jul 23 20:24:32 JaMa, good point Jul 23 20:25:07 can some please try clone bitbake? Jul 23 20:25:14 is there git magic we can do on the server? Jul 23 20:25:32 I dont think sp Jul 23 20:25:35 ups so Jul 23 20:26:14 thnk we likely need to restore from backup, or clone a fresh one up there from one of ours Jul 23 20:26:21 ok Jul 23 20:26:26 do we have a backup? Jul 23 20:26:38 $ git clone git://git.openembedded.net/bitbake Jul 23 20:27:00 Receiving objects: 8% (1043/12499), 212.00 KiB | 11 KiB/s Jul 23 20:27:22 dcordes_lib, bitbake != oe Jul 23 20:27:54 kergoth_: I am thinking on introducing a new var called TARGET_SUB_ARCH Jul 23 20:28:12 kergoth: TARGET_ARCH=arm TARGET_SUB_ARCH=armv7a Jul 23 20:28:25 or TARGET_ARCH=arm TARGET_SUB_ARCH=armv5te Jul 23 20:30:06 kergoth_: orly? can some please try clone bitbake? Jul 23 20:30:40 woglinde warflyr thank you both... I've now got the kenrel module there. Jul 23 20:30:59 kergoth, there is an issue with branchs Jul 23 20:31:15 khem hm Jul 23 20:31:20 tzanger nice job Jul 23 20:31:21 although I think will have up to date copies of the important ones Jul 23 20:32:04 woglinde: Checking out files: 100% (75/75), done. Jul 23 20:32:49 Crofton: how so? anyone who's cloned has all the branches under the remote. you don't have to have created local ones for them all.. Jul 23 20:32:56 you just hve to have it as a remote and have fetched recently Jul 23 20:33:14 yeah Jul 23 20:33:29 and if no one has a good copy, it is likely dead :) Jul 23 20:33:47 I am trying to be sensitive Jul 23 20:34:18 kergoth, is it worht bothering with the backup restore, or do you have something you can push that is very recent? Jul 23 20:34:39 okay so problem is only oe Jul 23 20:34:44 not bitbake too Jul 23 20:34:58 I assume that pulling OE will fail ? Jul 23 20:35:09 yes Jul 23 20:35:14 ok Jul 23 20:35:16 or for branches? Jul 23 20:35:33 as kergoth says, people can repush their branches Jul 23 20:35:41 yupp Jul 23 20:35:43 I am mainly concerned with dev and stable Jul 23 20:35:54 I have latest stable Jul 23 20:36:02 people have been active in stable, so I believe we have a backup Jul 23 20:36:07 exactly :) Jul 23 20:36:28 khem it worked Jul 23 20:36:30 thanks Jul 23 20:37:12 ok guys, I need to go edit a paper so I can upload it before I start drinking beer tonight Jul 23 20:37:23 do we feel like there is a plan? Jul 23 20:37:40 ka6sox, is looking for a backup atm Jul 23 20:37:55 good Jul 23 20:37:58 when that is restored, kergoth and woglinde can push there repos Jul 23 20:38:15 worst case he can reset to empty and go from there Jul 23 20:38:27 he is trying to check with cbrake to see where backups go Jul 23 20:38:51 hehe it has one outcome I dont need to beg to delete my old gettext branch Jul 23 20:38:58 I'll try to keep an eye on the channel, but I need to get the paper submitted tonight Jul 23 20:38:59 heh Jul 23 20:38:59 woglinde: ok glad something worked in this times :) Jul 23 20:39:01 yeah Jul 23 20:39:44 * khem wishes to delete all those bogus branch merge messages Jul 23 20:40:53 32 xdc.meta 0000014c 000000a8 000000a8 00072f7c 2**2 Jul 23 20:40:53 CONTENTS, READONLY Jul 23 20:41:58 hm oh objdump shows something about eabi Jul 23 20:42:01 private flags = 5000002: [Version5 EABI] [has entry point] Jul 23 20:44:49 woglinde: I have latest oe.dev too, but I'm leaving tomorrow, so just for confirmation I've pushed my branch to gitorious http://gitorious.org/~jama/angstrom/jama-shr-experimental/ (my rebased stuff is starting with Local modifications commit HEAD-17) Jul 23 20:46:13 * florian spots a bad bug in the armv4t sdk toolchains Jul 23 20:46:44 jama I think nobody will work an shr until you come back Jul 23 20:46:54 florian uh our? Jul 23 20:47:08 or cs? Jul 23 20:48:36 it looks like lib1funcs.asm gets compiled in a way that we get armv5 instructions Jul 23 20:48:45 woglinde: I mean that's just oe.dev where it's also last commit I was able to push before it went wrong Jul 23 20:50:04 this results in a toolchain producing armv4t binaries with armv5 instructions Jul 23 20:50:38 the Git server is hozed so lets wait till I get the backup installed and people can push. Jul 23 20:50:58 florian that might be the reason jama saw opendjdk cacao with illegal instrctuion Jul 23 20:51:51 woglinde: indeed... Jul 23 20:52:02 jama uh your branch is more up to date Jul 23 20:52:34 a student at the office ran into this today producing binaries for an armv4t machine Jul 23 20:52:49 let me see if i can find out why this happens Jul 23 20:56:02 crappy connection Jul 23 20:57:46 Does someone here happen to know offhand where these __ARM_ARCH... thingies get defined? Jul 23 20:57:55 re Jul 23 20:58:13 woglinde_: wb Jul 23 20:58:30 florian no Jul 23 20:58:36 in which dir is it? Jul 23 20:59:08 oh Jul 23 20:59:19 sorry... talking about gcc :) Jul 23 20:59:41 kergoth_, go ahead and push Jul 23 21:00:14 florian yes Jul 23 21:00:32 I know I would try to lookup it in the svn webfrontend Jul 23 21:00:46 it gets used in gcc/config/arm Jul 23 21:01:16 the backup was from 3hrs before the crash. Jul 23 21:01:45 03Michael 'Mickey' Lauer  07org.openembedded.dev * r369a4f9a5e 10openembedded.git/recipes/freesmartphone/fsogsmd_git.bb: fsogsmd: temporarily disable palm pre support, until upstream has been sorted out Jul 23 21:01:55 args Jul 23 21:02:07 mickeyl pushed Jul 23 21:02:26 gut its git Jul 23 21:02:30 args but Jul 23 21:03:13 thanks ka6sox Jul 23 21:03:24 thank cbrake for good backups :D Jul 23 21:03:25 i fixed our git tree Jul 23 21:03:40 mickeyl? Jul 23 21:03:42 if anything is missing, please holler Jul 23 21:03:52 it found a bunch of dangling commits Jul 23 21:04:18 woglinde_, you see if you are up todate. Jul 23 21:04:27 mickeyl jama had the latest Jul 23 21:04:29 http://gitorious.org/~jama/angstrom/jama-shr-experimental/ Jul 23 21:04:41 good. ~praise distributed SMCs Jul 23 21:04:45 SCMs, even Jul 23 21:05:09 good recovery. Jul 23 21:05:21 git fsck followed by git gc did it Jul 23 21:05:32 okay unless there is anything else I'm going to work. Jul 23 21:05:47 thanks. Jul 23 21:06:02 mickeyl will you push jama's stuff? Jul 23 21:06:29 is it lost? Jul 23 21:06:41 -> http://gitorious.org/~jama/angstrom/jama-shr-experimental/ Jul 23 21:07:08 anyone can push that, feel free to Jul 23 21:07:18 he I am lazy Jul 23 21:07:24 heh i fixed the tree Jul 23 21:07:24 and dont have fast internet Jul 23 21:07:26 now it's your turn :D Jul 23 21:09:28 thanks for the help mickey|tv Jul 23 21:09:43 np Jul 23 21:09:50 are we back to where we were Jul 23 21:09:54 i'm not sure which of these have been pushed already Jul 23 21:09:57 florian which instruction is it? Jul 23 21:10:04 mickey|tv: did Stefan Raab gave you an idea how to fix it :) Jul 23 21:10:06 khem: apparantly we're missing 20 hours Jul 23 21:10:13 wow, I actually got to use my backups -- that is a first :-) Jul 23 21:10:33 khem no Jul 23 21:10:35 but pushing jamas backup would us bring there Jul 23 21:10:36 cbrake: yeah backups are like insurance you want to have them but not use them Jul 23 21:10:38 ya, then mickey|tv brought it uptodate..excellent. Jul 23 21:10:53 florian hm which operation is it? Jul 23 21:11:43 khem, I've seen people keep write-only backups Jul 23 21:12:12 fatal: http://gitorious.org/~jama/angstrom/jama-shr-experimental//info/refs not found: did you run git update-server-info on the server? Jul 23 21:12:43 ok, git:// works Jul 23 21:13:08 I have all lost commits in my local tree Jul 23 21:13:14 will they be gone if I rebase Jul 23 21:13:37 no idea, better save that tree and then rebase Jul 23 21:13:37 khem sure? Jul 23 21:13:37 khem until gc is run you didnt lost something Jul 23 21:13:55 ups Jul 23 21:13:57 woglinde_: not sure... its around line 459 Jul 23 21:14:01 args Jul 23 21:14:06 its too late Jul 23 21:14:09 sorry khem Jul 23 21:14:16 rebasing should apply the lost ones which you could then push Jul 23 21:14:18 worth a try Jul 23 21:14:24 else we reclone from you Jul 23 21:14:27 if you have all barnches Jul 23 21:14:41 ok which means I can revert one of my own before I push :) Jul 23 21:15:13 worst thing is we need to wait one week for JaMa to apply his stuff again ;) Jul 23 21:15:38 they are gone with rebase Jul 23 21:15:48 so lets see Jul 23 21:16:15 we could pick up patch for patch Jul 23 21:16:29 from webinterface Jul 23 21:16:33 hm Jul 23 21:16:46 woglinde_: patchwork you mean ? Jul 23 21:17:05 actually let me fetch from JaMa's tree Jul 23 21:17:07 github Jul 23 21:17:22 ups gitrooius Jul 23 21:17:47 ok, so Jul 23 21:17:47 http://gitorious.org/~jama/angstrom/jama-shr-experimental/ Jul 23 21:17:53 i have cloned jamas tree Jul 23 21:17:58 then rebased on top of our current Jul 23 21:18:01 what i have now is: Jul 23 21:18:10 * khem git remote add Jul 23 21:19:59 I somehow get the impression that this might be an issue fr someone with more gcc knowledge Jul 23 21:20:26 pb_ might have an idea... Jul 23 21:22:41 http://pastebin.ca/1907154 Jul 23 21:22:44 does that look right? Jul 23 21:23:51 mickey|tv: it has lot more Jul 23 21:23:54 than it was Jul 23 21:24:01 mickey|tv: I will push what there was Jul 23 21:24:03 already Jul 23 21:24:13 give me some minutes Jul 23 21:24:19 * khem is cherry picking Jul 23 21:24:48 florian hm its a thumb operation? Jul 23 21:26:31 woglinde: it is in a block starting with "#if __ARM_ARCH__ >= 5" which must be wrong for armv4t Jul 23 21:27:39 woglinde: its the clz operation: http://infocenter.arm.com/help/topic/com.arm.doc.dui0068b/CIHJGJED.html Jul 23 21:28:05 yes Jul 23 21:28:11 clz is armv5 only Jul 23 21:28:19 yes Jul 23 21:28:22 #else /* __ARM_ARCH__ < 5 || defined (__OPTIMIZE_SIZE__) */ Jul 23 21:28:23 583 #if __ARM_ARCH__ >= 5 Jul 23 21:28:33 thats looks not good Jul 23 21:28:36 I found out what's wrong... Jul 23 21:29:30 #if __ARM_ARCH__ >= 5 && ! defined (__OPTIMIZE_SIZE__) Jul 23 21:29:32 in the same dir in arm.c starting from line 1204: " If the user did not specify a processor, choose one for them..." Jul 23 21:29:37 args no Jul 23 21:29:53 hm Jul 23 21:29:56 hehe Jul 23 21:29:57 it runs into this code... we do not select a cpu for the toolchain obviously Jul 23 21:30:23 this means it selects the default which is armv6 Jul 23 21:30:25 plain gcc Jul 23 21:30:31 without -march Jul 23 21:30:32 ah Jul 23 21:31:42 right... we are not compiling native here so i guess we need to do this in a different way. Jul 23 21:33:13 ok now I think we should be where we were Jul 23 21:33:17 03Pramod Poudel  07org.openembedded.dev * r80b63ab66c 10openembedded.git/recipes/opencv/opencv-dsp-acceleration_svn.bb: Jul 23 21:33:17 opencv-dsp-acceleration: Revision changed to 128 Jul 23 21:33:17 *IMGLIB_PATH is changed to path where other TI tools are found. Jul 23 21:33:17 Signed-off-by: Pramod Poudel Jul 23 21:33:17 Signed-off-by: Koen Kooi Jul 23 21:33:19 03Khem Raj  07org.openembedded.dev * r517c18f42c 10openembedded.git/recipes/gcc/ (gcc-4.5.inc gcc-4.5/gcc-pr43698-arm-rev-instr.patch): Jul 23 21:33:19 gcc-4.5: Fix rev instruction generation on armv6+ (backport PR 43698) Jul 23 21:33:26 03Khem Raj  07org.openembedded.dev * rc93277191e 10openembedded.git/recipes/yaffs2/ (3 files): Jul 23 21:33:27 yaffs2: Remove legacy staging (from poky) Jul 23 21:33:27 * yaffs2-util remove native recipe in favor of BBBLASSEXTEND. Jul 23 21:33:27 Signed-off-by: Khem Raj Jul 23 21:33:27 03Khem Raj  07org.openembedded.dev * r797299e771 10openembedded.git/recipes/uclibc/ (uclibc-git/fix-thumb1-PIE.patch uclibc_git.bb): Jul 23 21:33:42 hm fix-thumb1-PIE.patch Jul 23 21:33:55 only for armv4t khem? Jul 23 21:33:56 I have pushed all changes from jama's tree except his local changes + 1 commit from me which I wanted to revert anyway Jul 23 21:33:57 03Khem Raj  07org.openembedded.dev * r324341274d 10openembedded.git/recipes/quilt/ (quilt-native.inc quilt-native_0.48.bb): Jul 23 21:33:58 quilt-native: Dont configure with --without-date Jul 23 21:33:58 * This ends up quilt installing its little perl Jul 23 21:33:58 handler for date which does not understand -r Jul 23 21:33:58 option used in quilt refresh Jul 23 21:34:08 03Nicolas Boichat  07org.openembedded.dev * r7b9e14891f 10openembedded.git/recipes/libpcap/ (libpcap-1.1.1/ieee80215-arphrd.patch libpcap_1.1.1.bb): (log message trimmed) Jul 23 21:34:08 libpcap 1.1.1 recipe. Jul 23 21:34:08 Notes about patches from 1.0.0: Jul 23 21:34:08 - ieee80215-arphrd.patch: ported to 1.1.1 Jul 23 21:34:08 - aclocal.patch: does not seem to be needed (unsure) Jul 23 21:34:25 woglinde: that patch is not needed anymore I have bumped SRCREV for uclibc Jul 23 21:34:37 hm okay Jul 23 21:34:58 ah not exactly... linux-gnueabi gets us arm10tdmi as a default which armv5 Jul 23 21:35:10 more important was the ld.so can not treat protected symbols as they should be in ELF Jul 23 21:35:18 which gets NPTL working well Jul 23 21:35:21 on arm Jul 23 21:35:32 so far only arm and x86 have the fix Jul 23 21:35:37 others need to follow Jul 23 21:54:19 florian: which machine Jul 23 21:54:25 are you building for Jul 23 21:54:55 khem: some armv4t, simone in this case Jul 23 21:55:01 florian: ok Jul 23 21:55:06 I think you are right Jul 23 21:55:13 we need to pass this info to gcc configure Jul 23 21:55:22 with --with-arch --with-tune Jul 23 21:55:24 options Jul 23 21:55:28 need to set accordingly Jul 23 21:55:42 I had a patch to do that sometime ago Jul 23 21:55:45 which I cant find Jul 23 21:58:50 let me try what happens if i pass the options Jul 23 22:21:37 oops Jul 23 22:23:00 khem: looks like it isn't trivial... need to take a look tomorrow after getting some sleep Jul 23 22:33:29 good night Jul 23 23:54:56 hi Jul 23 23:55:00 anyone around? Jul 23 23:55:06 I am trying to get some files from TI Jul 23 23:55:24 specifically the C6000 stuff Jul 23 23:55:29 except all their links are down :/ Jul 23 23:55:36 anyone else seeing the same results? Jul 24 01:55:49 good evening Jul 24 01:57:36 I just updated and got an error: "TMPDIR has changed ABI (4 to 5) and you need to rebuilt, revert or adjust it" Jul 24 01:57:49 Anyone know how to fix it? Jul 24 02:02:48 yes. rm -rf tmp Jul 24 02:05:07 ;-) Jul 24 02:05:27 is it possible to revert it? Jul 24 02:05:38 will there b a problem? Jul 24 02:53:08 kergoth: hey Jul 24 02:53:57 kergoth: what would you prefer armv5te-oe-linux-gnueabi-gcc or armv5te/arm-oe-linux-gnueabi Jul 24 02:54:22 I think I like the latter better, though neither are very pretty ;) Jul 24 02:54:56 kergoth: /bin/armv5te/ Jul 24 02:55:04 and it presents other problems Jul 24 02:55:27 so it will be /usr/bin/armv5te/bin/ blah blah Jul 24 02:55:44 and it gets uglier that other tools Jul 24 02:56:17 armv5te-oe-linux-gnueabi-gcc will stay in </usr/bin Jul 24 02:56:29 which is more like how gcc is designed Jul 24 02:56:44 to have multiple gcc coexist Jul 24 02:56:50 ah. hmm Jul 24 02:57:03 kergoth: I am inclined to do it how gcc does it Jul 24 02:57:13 that makes sense, for consistency Jul 24 02:57:55 so you will have armv5te-oe-linux-gnueabi-gcc and armv7-oe-linux-gnueabi-gcc live side by side Jul 24 02:58:09 and names are more expressive :) Jul 24 02:58:22 sounds reasonable Jul 24 02:58:26 but are against the notion of multilib Jul 24 02:59:14 kergoth: so we should have this defined in machine.conf Jul 24 02:59:24 and bitbake.conf should use that value Jul 24 02:59:35 TARGET_ARCH will be the highlevel arch as it is now **** ENDING LOGGING AT Sat Jul 24 02:59:57 2010