**** BEGIN LOGGING AT Wed Sep 12 02:59:59 2012 Sep 12 04:32:02 why would this inittab line fail? `hm:2345:respawn:/usr/bin/health_monitor`. as root, /usr/bin/health_monitor runs just fine. Sep 12 04:32:10 it does write a log file, would that kill it? Sep 12 06:54:35 Have anyone tried adding BBCLASSEXTEND = "nativesdk" to bash_3.2.48 in oe-core and compiled it as a nativesdk package? I can't see why it won't compile. The compile process complains about "./mkbuiltins: error while loading shared libraries: __vdso_gettimeofday: invalid mode for dlopen(): Invalid argument" I don't understand why that can't be loaded. Any ideas? Sep 12 07:26:43 morning all Sep 12 08:21:43 morning all Sep 12 08:22:05 morning Sep 12 08:50:40 grml. how do I add a do_devshell override in kernel.bbclass? i've tried what feels like a dozen ways but fail each time :) Sep 12 10:19:16 ant_work, are the logs rotating correctly? Sep 12 10:20:05 ka6sox: yes, since 28-Aug-2012. thanks! Sep 12 10:20:31 ant_work, good to hear...I couldn't see the issue but another SA found it... Sep 12 10:44:15 eh, OE gcc has -nostdlib++ option. need to add it ;( Sep 12 10:45:39 how was the wedding? Sep 12 10:46:18 i'm just promoted to being best-man for an old friend next month Sep 12 10:46:47 good morning Sep 12 10:49:19 hey florian Sep 12 10:49:27 did things calm for you a bit or is it still hectic? Sep 12 10:49:34 mickey_office: great it was - broke most of traditions Sep 12 10:49:43 hrw: excellent, sounds good :) Sep 12 10:49:54 mickey_office: not really... its going to take a while Sep 12 10:50:07 hrw: your wedding?! Sep 12 10:50:40 florian: friend Sep 12 10:50:58 hrw: :-) Sep 12 11:04:09 adding new architecture into OE can be painful Sep 12 11:04:27 all those recipes for packages with old config.{guess,sub} files Sep 12 11:12:15 does someone tried to add do_configure_prepend() which would update those two files? Sep 12 11:34:48 hrw: don't we run autoreconf for that reason? Sep 12 11:35:28 zecke: have to check db and ncurses recipes then Sep 12 11:36:46 db has own one Sep 12 11:37:15 same with ncurses Sep 12 11:40:48 ncurses is just a pain, someone should update it to work with a newer autoconf Sep 12 11:41:02 I don't know why db is not getting autoreconf'd, maybe it's just a bug in the recipe Sep 12 11:43:20 for some reason it has STRIP=true and then oe_runconf Sep 12 11:43:36 rmeoved do_configure, rebuilding Sep 12 11:47:25 when i do bitbake openjdk-6 it just stops at Parsing Reciepes and ETA does not show any value.. ? what might be the reason.? Sep 12 11:48:25 vlrk1: give it 15 minutes Sep 12 11:48:37 vlrk1: slow network? Sep 12 11:51:47 hrw: thx Sep 12 11:52:04 it took already 20 min Sep 12 11:59:53 one do_configure_prepend() in bbappend sorted problem out for now Sep 12 12:01:02 hrw: We've slowly been killing off the custom do_configures but still one or two left... Sep 12 12:08:16 hrw: still no change it showing parsing recipes with %0 only Sep 12 12:08:25 hi all Sep 12 12:08:39 can see like any packet movement in wireshark Sep 12 12:08:44 *can i see Sep 12 12:08:46 ? Sep 12 12:52:26 how to do right staging in oe-core? Sep 12 12:53:47 silvio_: install the file in the appropriate place in do_install and the file should be staged Sep 12 13:04:15 ~curse rpm-native Sep 12 13:04:19 May you be reincarnated as a Windows XP administrator, rpm-native ! Sep 12 13:04:59 4 years ago I said 'from now on, I do not know anything about rpm'. I hoped that will never have to deal with it ;) Sep 12 13:05:07 | /home/hrw/HDD/devel/canonical/aarch64/build/tmp-eglibc/work/x86_64-linux/rpm-native-5.4.9-r46/rpm-5.4.9/rpmdb/.libs/librpmdb.so: undefined reference to `db3Free' Sep 12 13:05:37 ideas? Sep 12 13:05:56 bluelightning, i try but the packages-split have empty dir, the package is qjson Sep 12 13:06:50 silvio_: well, packaging and staging are two different things Sep 12 13:07:31 hrw: I can answer why it is built even if you don't select it - we require rpmdeps to calculate per-file dependencies Sep 12 13:07:45 bluelightning: I know - checked that already Sep 12 13:07:46 hrw: but I don't know why you would see that specific error though Sep 12 13:07:53 bluelightning: but why it fails is other thing Sep 12 13:19:02 tmp-eglibc/sysroots/x86_64-linux/usr/lib/libdb-__EDIT_DB_VERSION_MAJOR__.so Sep 12 13:19:02 tmp-eglibc/sysroots/x86_64-linux/usr/lib/libdb-__EDIT_DB_VERSION_MAJOR__.__EDIT_DB_VERSION_MINOR__.so Sep 12 13:19:05 thats why Sep 12 13:20:15 ??? Sep 12 13:22:02 did rebuilt and works fine Sep 12 13:23:51 looks like some kind of artifact of how db is versioned within its own configure.ac Sep 12 13:26:00 hrw: this is because you removed the do_configure in the db recipe isn't it? Sep 12 13:26:19 I suspect that was there because of this problem Sep 12 13:28:25 maybe Sep 12 13:28:40 now I have do_configure_prepend in bbappend Sep 12 13:28:41 the db source provides a script which makes the appropriate substitutions for a release, but we do not run that, so running autoreconf as we would without that do_configure will result in the default unmodified __EDIT_DB_VERSION_* coming through from dist/configure.ac Sep 12 14:40:58 how to mark recipe as non-working? Sep 12 14:42:19 hrw: depends on what the criteria are... COMPATIBLE_MACHINE / COMPATIBLE_HOST? Sep 12 14:43:31 hrw: or alternatively you can just raise bb.parse.SkipPackage("some reason") in an anonymous python function based on some more advanced check Sep 12 14:43:35 hrw: "git rm" Sep 12 14:43:37 :-} Sep 12 14:43:45 heh Sep 12 14:44:33 pb_: easiest indeed Sep 12 14:44:48 bluelightning: I want to remove one recipe for arch Sep 12 14:46:55 COMPATIBLE_MACHINE="fwfwe" works Sep 12 14:47:56 hrw: some kind of regex in COMPATIBLE_HOST would seem to be indicated, but I can't find any examples where we just exclude some archs rather than just specify a list of the compatible ones Sep 12 14:48:49 bluelightning: anyway it is private branch so far Sep 12 14:49:53 hrw: sure, I'm giving the answer as it if were ultimately going to be integrated into OE-Core :) Sep 12 14:51:05 I know Sep 12 15:25:24 my patches to remove the poky/owl references from the pimlico stuff isn't merged yet, if anyone is looking for a patch to merge Sep 12 16:29:26 alright, well bitbake-layers show-overlayed doesn't list the bbappend in my layer, but I gather that is proper since I've appended the bb and not created my own. my append is shown by show-appends. in general the various outputs of bitbake-layers appear sane. bitbake-selftest returns OK. Also, bitdoc is missing a ) on line 23 Sep 12 18:25:22 denim, pb_, kergoth: thanks for all the help. I think I've gotten it. this may be hacky but here goes: On a lark I switched from FILESEXTRAPATHS_prepend in my bbappend file to FILESPATH_prepend for the directory containing my version of the defconfig. That directory is first in FILESPATH now and it appears to be using my defconfig. Sep 12 19:08:51 is psyco being replaced as it's dead? Sep 12 19:12:38 when its dead what should replace it? Sep 12 19:15:48 woglinde: they are pointing to pypy **** ENDING LOGGING AT Thu Sep 13 03:00:00 2012