**** BEGIN LOGGING AT Sun Mar 14 02:59:56 2010 Mar 14 10:10:46 anyone having any experience with building an arm uclibc toolchain? Mar 14 10:10:57 I'm trying to build one, but the process bails out on configuration of libiconv (1.12). Mar 14 10:11:13 automake (1.11.1) returns following error: cp: cannot stat `..../usr/share/gettext/config.rpath': No such file or directory Mar 14 11:05:42 | Cannot load module /home/filip.zyzniewski/pda/oe/build_dir/tmp/work/i686-linux/gtk+-native-2.18.6- Mar 14 11:05:45 r8.4/gtk+-2.18.6/modules/input/im-xim.la: /home/filip.zyzniewski/pda/oe/build_dir/tmp/work/i686-linu Mar 14 11:05:48 x/gtk+-native-2.18.6-r8.4/gtk+-2.18.6/modules/input/.libs/im-xim.so: undefined symbol: g_dgettext Mar 14 11:05:57 sorry for multiple lines Mar 14 11:27:50 has AUTOTOOLS_STAGE_PKGCONFIG still any use? I cannot find a class that uses it Mar 14 11:48:59 damn I am stuck Mar 14 11:59:35 filip: no idea about gtk+, maybe try to build gettext first if it is not build Mar 14 12:04:55 eFfeM1: this symbol iws in libglib Mar 14 12:05:53 filip.zyzniewski@tefnet ~ $ readelf -s /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libglib-2.0.so.0 | grep g_dgettext | head -n 1 1328: 00058950 83 FUNC GLOBAL DEFAULT 11 g_dgettext Mar 14 12:06:33 filip.zyzniewski@tefnet ~ $ ldd /home/filip.zyzniewski/pda/oe/build_dir/tmp/work/i686-linux/gtk+-native-2.18.6-r8.4/gtk+-2.18.6/modules/input/.libs/im-xim.so | grep libglib libglib-2.0.so.0 => /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libglib-2.0.so.0 (0xb6ebe000) Mar 14 12:06:47 so I have no idea why the complaints... Mar 14 12:09:59 filip: is glib loaded ? Mar 14 12:10:26 i mean is it in the libs that are used to build it Mar 14 12:10:35 * eFfeM1 is not too familiar with gtk+ Mar 14 12:10:38 yes Mar 14 12:10:48 ldd shows it Mar 14 12:13:37 filip: yeah, ubt is there a -Lglib when building Mar 14 12:14:24 s/ubt/but/ Mar 14 12:15:14 (...) -lresolv /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libpangoft2-1.0.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libpango-1.0.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libfontconfig.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libfreetype.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libexpat.so /home/filip Mar 14 12:15:22 whoopz Mar 14 12:15:30 line ends like this Mar 14 12:16:11 linking one Mar 14 12:16:11 does not look too good Mar 14 12:16:45 well it's just 'bitbake console-image' from a clean org.openembedded.dev tree Mar 14 12:17:14 strange, Mar 14 12:17:24 check makefile and configure things Mar 14 12:17:28 -lresolv /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libpangoft2-1.0.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libpango-1.0.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libfontconfig.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libfreetype.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libexpat.so /home/filip.zyzni Mar 14 12:17:43 -lresolv /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libpangoft2-1.0.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libpango-1.0.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libfontconfig.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libfreetype.so /home/filip.zyzniewski/pda/oe/build_dir/tmp/staging/i686-linux/usr/lib/libexpat.so /home/filip.zyzni Mar 14 12:18:09 damn! Mar 14 12:18:35 sorry, I had it in my clipboard and wiped my touchscreen :/ Mar 14 12:20:38 any expertise on pkg_config around ? Mar 14 12:20:47 i have an issue with paths Mar 14 12:21:16 why would console-image need native gtk+ is beyond me... Mar 14 12:22:51 try bitbake -g console-image and look in the depends.dot that it generates Mar 14 12:23:33 but i suspect it is something like: console-image wants bluez4 which wants gst-plugins which wants gt+ Mar 14 12:26:53 filip, it should be possible to upgrade console-image to a full X machine :) Mar 14 12:27:03 um, how can I trigger do_clean on a virtual:native:/home/filip.zyzniewski/pda/oe/openembedded/recipes/gtk+/gtk+_2.18.6.bb ? Mar 14 12:27:11 just because it gets built does not mean it is installed Mar 14 12:27:21 this virtual:native confuses me Mar 14 12:28:11 bitbake -b virtual:native:/home/filip.zyzniewski/pda/oe/openembedded/recipes/gtk+/gtk+_2.18.6.bb -c clean Mar 14 12:28:19 this complains about sanity Mar 14 12:30:36 nevermind... Mar 14 13:28:23 03Michael 'Mickey' Lauer  07org.openembedded.dev * r20a01a4f6e 10openembedded.git/recipes/freesmartphone/zhone2_git.bb: zhone2: add FSO dependencies Mar 14 13:30:26 03Michael 'Mickey' Lauer  07org.openembedded.dev * r6fedad5a85 10openembedded.git/ (2 files in 2 dirs): android-image-utils-native: new recipe; image utilities to build kernels for android machines Mar 14 14:00:13 03Michael 'Mickey' Lauer  07org.openembedded.dev * r6615eb9944 10openembedded.git/recipes/android/android-image-utils-native_git.bb: android-image-utils-native: fix deploying fastboot binary Mar 14 14:06:46 * RP yawns and returns from his trip away Mar 14 14:08:45 RP wb Mar 14 14:09:32 and, while you are here, do you happen to know if AUTOTOOLS_STAGE_PKGCONFIG still makes any sense, could not find a place in either bitbake or classes/* that uses it Mar 14 14:12:17 yo RP Mar 14 14:12:35 03Michael 'Mickey' Lauer  07org.openembedded.dev * re0a5c66df5 10openembedded.git/recipes/linux/linux-leviathan_git.bb: linux-leviathan: depend on android-image-utils and prepare kernel image that can be booted via fastboot Mar 14 14:12:38 eFfeM1: No, that variable is dead Mar 14 14:13:03 (thankfully) Mar 14 14:28:14 03Michael 'Mickey' Lauer  07org.openembedded.dev * ra2da6e643c 10openembedded.git/recipes/linux/linux-leviathan_git.bb: linux-leviathan: remove bogus ' characters that disturb mkbootimg Mar 14 14:28:48 hej Mar 14 14:29:14 mickeyl: could you maybe look at http://patchwork.openembedded.org/patch/1631/ , and what additionally has to be done to support natve extend for distutils ? Mar 14 14:30:20 hmm, i don't know whether i have time for something like that soon Mar 14 14:30:50 ah ok :/ Mar 14 14:34:02 the motivation to do it is a bit low as well Mar 14 14:34:07 as things are working fine without Mar 14 14:34:08 :) Mar 14 14:34:24 and it took me a lot of time to get distutils to behave at all Mar 14 14:34:37 plus i'm hardly using it anymore Mar 14 14:34:51 my lion share goes into Vala development Mar 14 14:35:23 aye, distutils is working perfect indeed ;) Mar 14 14:35:33 so would staging (i think ..) Mar 14 14:40:54 RP, ok thanks Mar 14 14:49:31 hi Mar 14 14:49:51 | checking whether btowc(EOF) is correct... configure: error: in `/home/embedded/oetmp_eeepc701/work/i686-angstrom-linux/fastjar-0.98-r0/fastjar-0.98': => | configure: error: cannot run test program while cross compiling Mar 14 14:49:57 that's in fastjar Mar 14 14:50:29 I'll look config.log Mar 14 14:50:57 not much details in there Mar 14 14:51:05 it says the same thing Mar 14 14:52:13 too bad woglinde is not there Mar 14 15:11:31 ok, I traced it to the m4 Mar 14 15:11:42 m4/btowc.m4 Mar 14 15:13:15 GNUtoo|oeee: it tries to do a test to dynamiccaly find a var, but ofc you cannot do so while crosscompling Mar 14 15:13:33 ok Mar 14 15:13:50 I know I shouldn't be able to run test while cross-compilings Mar 14 15:14:16 the thing I'm looking into right now is how to prevent it from testing Mar 14 15:14:28 the problem is this one: Mar 14 15:14:29 AC_TRY_RUN([ Mar 14 15:14:44 if test $LOCALE_FR != none; then Mar 14 15:15:01 if test $LOCALE_FR != none; then makes it run the AC_TRY_RUN([ Mar 14 15:15:41 ( http://pastebin.com/tDV7gFt6 ) Mar 14 15:16:05 GNUtoo|oeee: not sure about AC_TRY_RUN but AC_RUN_IFELSE has an additional [] clause for cross compiling Mar 14 15:16:16 ok Mar 14 15:16:55 - Macro: AC_TRY_RUN (program, [action-if-true], [action-if-false], [action-if-cross-compiling]) Mar 14 15:16:59 AC_TRY_RUN also has it Mar 14 15:18:22 ok thanks Mar 14 15:19:11 wow thanks a lot Mar 14 15:19:20 []) is the action when cross-compiling Mar 14 15:19:25 :) Mar 14 15:20:46 I'll look if wchar.h is present and its content and put yes if so Mar 14 15:20:53 all manually Mar 14 15:26:36 ok tested with devshell Mar 14 15:54:07 maybe site system should be used instead of patches? Mar 14 15:54:36 I guess that depends if those settings would be valid for all packages or not? Mar 14 15:55:27 ah ok Mar 14 15:55:38 I guess so because it tests standard libc features Mar 14 15:56:07 btw I never digged into the site system Mar 14 15:56:32 I only know that it cashes values for the autotools in order to prevent tests to be run Mar 14 15:56:46 I think you can define default autoconf values there, so that configure thinks the test has been run already and just takes the value that was set there Mar 14 15:56:48 yes, exactly Mar 14 15:56:51 (it contain the result of theses tests) Mar 14 15:57:04 ok Mar 14 16:29:22 hi it's keeping doing that: Mar 14 16:29:24 | configure: error: "The rhino jar /home/embedded/oetmp_eeepc701/staging/java/rhino.jar was not found." Mar 14 16:29:31 At the beginning I had: Mar 14 16:29:53 STAGING_DIR_JAVA = "${STAGING_DIR}/java" and STAGING_DATADIR_JAVA ?= ${STAGING_DIR_JAVA} Mar 14 16:30:00 I just removed it because of this error Mar 14 16:30:04 but the error persist Mar 14 16:30:18 I'll activate it again Mar 14 16:30:40 and yes I cleaned before retrying Mar 14 16:31:25 of course I could move the jar Mar 14 16:31:26 but... Mar 14 16:32:25 mmm --with-rhino=${STAGING_DATADIR_JAVA}/rhino.jar Mar 14 16:32:31 is there a jalimo channel? Mar 14 16:33:00 seem that nobody is there...I'll hack the recipe Mar 14 17:06:46 quick question: do_split_packages can't handle symlinks, can it ? Mar 14 17:47:23 ouch...compilation failure Mar 14 17:47:30 nvm, found a git log entry stating that it doesnt cover symlinks ;) Mar 14 17:47:32 I'll look but I bet I won't be able to handle it Mar 14 17:48:09 ah ok Mar 14 17:48:14 maybe I will Mar 14 17:48:14 openjdk-ecj/build/linux-i586/j2sdk-image/jre/lib/i586/server/libjvm.so Mar 14 17:48:25 was not found Mar 14 17:51:28 hm, any idea what is wroing if a do_distribute_sources and it fails in SRC_DISTRIBUTECOMMAND ? it tries to ln . to ./. guess ${SRC} is not set properly (if I look at the conf file) but no idea why/how Mar 14 18:13:18 nevermind, overlooked an earlier error Mar 14 18:21:52 anyone here who tried the new archos 5IT layover? Mar 14 18:27:02 anybody having this http://bugs.openembedded.net/show_bug.cgi?id=5413 problem? Mar 14 18:27:59 did not even got that far Mar 14 19:01:07 03Klaus Kurzmann  07org.openembedded.dev * r5a77cf7d7d 10openembedded.git/conf/distro/include/sane-srcrevs-fso.inc: Mar 14 19:01:07 sane-srcrev-fso.inc: bump rev of libframeworkd-glib Mar 14 19:01:07 Signed-off-by: Klaus Kurzmann Mar 14 19:04:28 MindFly, they have an irc channel,maybe you would have more specific help there Mar 14 19:05:04 maybe it's #openpma, not shure...they changed too much irc channel Mar 14 19:05:13 like #openpma -> #x04 -> #openpma Mar 14 19:05:26 I'm not shure about the last -> Mar 14 19:19:14 btw I'm trying to do a mpd client while waiting for java to build Mar 14 19:19:55 and I've this python error that apear only on machine rootfs built with oe Mar 14 19:20:10 socket.gaierror: [Errno 3] invalid value for ai_flags. Mar 14 19:20:43 but unfortunately I was just asked to do something in real life and I'll have to reboot under a distro which support scanning+printing Mar 14 19:20:56 sorry I'll be back soon Mar 14 20:01:53 03Michael 'Mickey' Lauer  07org.openembedded.dev * r64eae2b3c9 10openembedded.git/recipes/efl1/ (3 files in 2 dirs): Mar 14 20:01:53 ecore: add patch that fixes the ecore-fb touchscreen initialization Mar 14 20:01:53 With this patch, we have a working touchscreen in framebuffer mode again Mar 14 20:01:54 03Michael 'Mickey' Lauer  07org.openembedded.dev * r2d6f65949c 10openembedded.git/recipes/tslib/ (tslib/htc-msm7/ts.conf tslib/htc-msm7/tslib.sh tslib_1.0.bb): tslib: add configuration for htc-msm7 devices Mar 14 20:01:54 03Michael 'Mickey' Lauer  07org.openembedded.dev * re32f710d17 10openembedded.git/recipes/pointercal/ (files/htc-msm7/pointercal pointercal_0.0.bb): pointercal: add data for htc-msm7 devices Mar 14 20:01:55 03Michael 'Mickey' Lauer  07org.openembedded.dev * rf86ca9a341 10openembedded.git/recipes/efl1/elementary_svn.bb: elementary: packaging updates Mar 14 20:04:19 mmm Mar 14 20:04:19 http://pastebin.com/1XAsmvLp Mar 14 20:05:11 I'll go and when I come back I'll look where the 'X_ShmAttach' identifier is declared Mar 14 20:38:34 back Mar 14 20:40:13 hi again Mar 14 20:40:27 i'll take a look in openpma Mar 14 20:40:57 the archos openembedded overlay patch seems to be broken Mar 14 21:38:10 can I do a git pull of the man branch when my tree is on a side branch (without nuking the side branch or going into a merge) Mar 14 21:41:29 mmm don't know Mar 14 21:41:41 ah ok Mar 14 21:41:47 you mean git pull Mar 14 21:41:58 without touching your branch Mar 14 21:41:59 maybe Mar 14 21:42:07 mmm Mar 14 21:42:13 I never looked at it Mar 14 21:42:28 s/looked at it/payed attention if it did it Mar 14 21:42:48 btw can I change the status? **** ENDING LOGGING AT Mon Mar 15 02:59:56 2010