**** BEGIN LOGGING AT Fri Oct 19 03:00:01 2012 Oct 19 07:19:05 ok , screw udev Oct 19 07:53:37 hi all Oct 19 07:59:58 | cp: cannot stat `/home/hrw/HDD/devel/canonical/aarch64/build/tmp-eglibc/sysroots/x86_64-linux/usr/share/gettext/config.rpath': No such file or directory Oct 19 08:00:01 | ERROR: Function failed: do_configure (see /home/hrw/HDD/devel/canonical/aarch64/build/tmp-eglibc/work/x86_64-linux/shadow-native-4.1.4.3-r8/temp/log.do_configure.28177 for further information) Oct 19 08:00:07 ideas? Oct 19 08:01:17 hrw: rebuild gettext-minimal-native Oct 19 08:01:33 hrw: I had the same problem yesterday, not sure what removed it from sysroot Oct 19 08:02:55 thanks - fixed Oct 19 08:41:22 morning all Oct 19 08:41:41 morning Oct 19 08:46:48 hi bluelightning Oct 19 08:49:10 'monrning Oct 19 08:59:28 ech. core-image-lsb-sdk requires gstreamer -> pulseaudio -> libatomics_ops ;( Oct 19 09:34:35 good morning Oct 19 09:35:17 hrw: I remember I had to fight this one as well... mips? Oct 19 09:39:20 florian: aarch64 Oct 19 09:39:36 hrw: oops Oct 19 09:45:40 is there any oe related documentation regarding systemd usage and adding new servfices ? Oct 19 09:54:53 Can somebody help me with shadow-native? It keeps crying about not being able to stat tmp/sysroots/x86-64-linux/usr/share/gettext/config.rpath Oct 19 09:55:01 I tried to build gettext manually and it succeeds Oct 19 09:55:07 But I can't find the file shadow-native asks about Oct 19 10:17:34 hi Oct 19 10:18:00 send the pull request for the stuff that you have done Oct 19 10:18:09 after that you can test these things Oct 19 10:18:18 but we will be delivering with these things Oct 19 10:18:24 so this is less important Oct 19 10:18:45 aahhhh wrong window Oct 19 10:18:48 sorry Oct 19 10:23:00 my question was if a recipe is inherited by update-rc is there any way that we can control its addition in startup scripts (init-rd)? Oct 19 10:24:00 RagBal: rebuild gettext-minimal-native Oct 19 10:24:15 Hmm thanks Oct 19 10:24:29 I copied the 2 missing files from another project, and they work Oct 19 10:24:38 Noor: surely INITSCRIPT* variables? Oct 19 10:24:44 I'll try to remove the files and rebuild gettext-minimal-native Oct 19 10:26:04 bluelightning: thanks for the tip but these variables are already set Oct 19 10:26:27 I have two setups one uses old version of oe-core and one is on latest poky Oct 19 10:26:43 I am facing problem with avahi-daemon Oct 19 10:27:02 JaMa, thanks it worked like a charm! Oct 19 10:27:25 in old setup no symlink are created in rc.d folder Oct 19 10:27:36 I can't see any change made in the recipe Oct 19 10:28:08 but in poky base distro we are getting avahi-daemon in rc.d folders Oct 19 10:29:06 bluelightning: so confused that why it was not added in fs in old setup Oct 19 10:30:06 Noor: don't know I'm afraid, but it sounds like it's doing the expected thing now Oct 19 10:30:38 bluelightning: thanks .... I'll check with kergoth Oct 19 10:52:38 morning all Oct 19 11:24:28 ERROR: files-injection: Recipe file does not have license file information (LIC_FILES_CHKSUM) Oct 19 11:24:43 how to create recipe for files lacking license? Oct 19 11:25:14 I need recipe for adding random files to rootfs without taking care of what user will put there Oct 19 11:26:09 found Oct 19 11:27:05 LICENSE = "CLOSED" is megaignore flag Oct 19 11:28:26 hrw: a bunch of choices: 1) add a license file within the metadata; 2) copy one in in do_configure / do_configure_prepend; 3) point LIC_FILES_CHKSUM to one of the files in COMMON_LICENSE_DIR; 4) set LICENSE = "CLOSED" Oct 19 11:28:39 with the right solution depending on the situation Oct 19 11:28:55 e.g. LICENSE = "CLOSED" for something that is not proprietary would be wrong Oct 19 11:29:32 bluelightning: I have no idea what people may need it for Oct 19 11:29:53 it can be helloworld.c or closed ARM Ltd. internal binary or GPLv2 archive etc Oct 19 12:13:04 pb_: could you review my kernel.bbclass/update-alternatives change? (http://patchwork.openembedded.org/patch/38129/) Oct 19 12:35:19 Hello. Oct 19 12:35:37 I have a qt application that has 2 .pro files. Oct 19 12:36:05 How do I do a "qmake file1.pro some/other/dir/file2.pro" in a bitbake recipe? Oct 19 12:41:09 It seems the answer is: QMAKE_PROFILES="file1.pro some/other/dir/file2.pro" Oct 19 12:42:07 (after inheriting qt4x11) Oct 19 12:43:03 hrw: right, but you shouldn't try to shoehorn all of those things into one recipe Oct 19 12:43:12 hrw: or you're going to have trouble tracking the license of things Oct 19 12:45:02 inch: correct; the default is to build all .pro files in the main source directory, but if you need different behaviour you can set QMAKE_PROFILES Oct 19 12:51:49 Ouch. It seems that with this source, running that "qmake file1.pro other/dir/file2.pro" it only compiles stuff behind file2.pro. Oct 19 12:52:09 The same happens with bitbake and debian wheezy too. Oct 19 12:52:40 Is there a way to run qmake in the "root" dir and then qmake (and make) in the other/dir separately? Oct 19 12:54:06 inch: you could do the latter in do_configure_append() I guess Oct 19 12:54:22 (well, qmake, not make) Oct 19 12:56:12 Can I somehow inherit the stuff that openembedded-core/meta/classes/qmake_base.bbclass does when i write my own do_configure_append? Oct 19 12:58:33 inch: I think the way it's structured you can only really copy parts of it that you need Oct 19 12:58:42 inch: honestly it sounds like broken source to me... Oct 19 12:59:07 Yeah, maybe it is better to patch it somehow. Oct 19 12:59:21 that would certainly be another option Oct 19 16:35:08 khem: morning... any idea why certain patches are being ignored by patchwork? e.g. my meta-oe hddtemp patch Oct 19 16:35:23 bluelightning: hmmm is it very big Oct 19 16:35:44 usually it should not ignore it Oct 19 16:36:19 just over 300k, so not huge Oct 19 16:43:56 bluelightning: hmm 300K ok Oct 19 16:44:01 resend once again Oct 19 16:46:56 khem: done Oct 19 16:59:39 khem: still doesn't seem to have shown up Oct 19 22:21:01 suggestions on how to diagnose "This autoconf log indicates errors, it looked at host include and/or library paths" errors? Oct 19 22:23:55 yes, look at that log, find which line triggered this Oct 19 22:24:24 uh huh, but the log is long. any tips on how to find the offending part quickly? Oct 19 22:25:54 well, just searching for /usr/include found one problem... Oct 19 22:37:11 heh, those errors should really identify the line numbers Oct 19 22:38:19 :) Oct 19 23:31:36 when packaging a library that uses autotools, shouldn't the autotools class automatically find the lib.h that "make install" creates and stick that in lib-dev package? Oct 19 23:40:32 not exactly. any files that do_install puts in ${D}${includedir} will go into the -dev package, regardless of whether it uses autotools or not Oct 19 23:40:45 and the autotools class provides a default do_install that runs make DESTDIR=${D} install Oct 19 23:41:01 so, effectively yes, but i wanted to clarify that the behavior isn't autotools-specific Oct 19 23:41:14 I see, thanks **** ENDING LOGGING AT Sat Oct 20 03:00:02 2012