**** BEGIN LOGGING AT Fri Jun 03 02:59:57 2011 Jun 03 03:01:19 will do...are you working close to Sunnyvale or down south? Jun 03 07:17:34 hello, I'm trying to use openembedded with codesourcery prebuild toolchain, but I have to be missing something... Is here please anyone, who could help Jun 03 07:17:36 ? Jun 03 07:45:39 jan__: well, you can start with the manual.. but so far no one volunteered to write a section for it. Jun 03 07:46:29 jan__: another source is to search the mailinglist archive Jun 03 07:47:40 I of course started there and even found a few threads in mailing list, but with no success Jun 03 07:49:07 in samle local.conf file, there is several comments about codesourcery toolchain, but uncommenting that lines is not enough, obviously Jun 03 07:49:25 TOOLCHAIN_VENDOR = "-none" TOOLCHAIN_TYPE = "external" TOOLCHAIN_BRAND = "csl" TOOLCHAIN_PATH = "/opt/arm-2011.03" Jun 03 07:51:47 jan__: and how does it fail? Jun 03 07:52:34 jan__: which DISTRO are you building? Jun 03 07:55:48 hello Jun 03 07:56:05 I have tried Angstrom Jun 03 07:56:38 however is there distro needed for compile eg. one package (busybox)... Jun 03 07:57:53 ? Jun 03 07:58:04 I had builded a filesystem using oe with fltk support .... I had tried my own fltk apps....and cross compiled with oe environment..... but my one fltk apps gives me segmentation fault..... on serially it gives me error of .... Unable to handle kernel paging request at virtual address 4015e000 Jun 03 07:58:04 pgd = c70b8000 Jun 03 07:58:07 .................... Jun 03 07:58:09 ...................... Jun 03 08:08:17 jan__: well, what is the error you get. The distro is there to set policy... Jun 03 08:12:54 I just started it again from scratch with angstrom, fortunately no error so far, but I don't understand why it downloads and builds so many different packages (python, perl, zlib, openssl, ...) on command bitbake busybox Jun 03 08:19:50 jan__: well, you can use bitbake -g to look at the depedency graph Jun 03 08:21:46 jan__: the answer is.. for the -native packages one can not generally depend on the system to have things in the right version, e.g. when targeting about 6 years (old RHEL to latest ubuntu) of distro releases.. some GNU makes are broken... Jun 03 08:24:22 ok, I will wait to the end of building and we will see Jun 03 08:25:03 anyway, is it possible to not set the distro in local.conf? Jun 03 08:28:12 morning Jun 03 08:33:57 ERROR: Function 'Fetch failed: http://busybox.net/downloads/fixes-1.18.4/busybox-1.18.4-hush.patch;name=patch01 checksum mismatch.' failed Jun 03 08:34:18 tried three times with deleting downloaded files... Jun 03 08:42:53 ok, solved by replacing checksums in busybox*.bb Jun 03 09:56:29 Hi everyone. Can someone advise me on problems using the SDK built by meta-toolchain-qte? One or more native libraries seem to be missing from the SDK install. My host is x86_64, building for i586 target. Jun 03 09:56:59 The native libdbus-1.so isn't being installed by the SDK. It is being built though. I can workaround the problem by manually copying libdbus-1.so from the meta-toolchain-qte build into the SDK install tree. Is qt4-tools-nativesdk missing code to install this lib (and possibly others)? Jun 03 09:57:52 jch11fg sorry seems it is still buggy again Jun 03 09:58:05 jch11fg have a look at oe-core Jun 03 09:58:16 maybee thats working betting at sdk level Jun 03 09:58:40 Actually, I am using oe-core. Jun 03 09:58:52 Should have mentioned that. :-) Jun 03 09:59:24 jch11fg than write to the oe-core ml Jun 03 09:59:43 ok, will do Jun 03 11:04:13 hi obi Jun 03 11:04:35 hi woglinde! Jun 03 11:11:51 03Koen Kooi  07org.openembedded.dev * re61d3d8990 10openembedded.git/recipes/xorg-lib/libxaw_1.0.9.bb: Jun 03 11:11:51 libxaw: fix depchains Jun 03 11:11:51 xserver-xorg-dev can't be installed without this Jun 03 11:11:51 Signed-off-by: Koen Kooi Jun 03 11:15:56 when I compiled a fltk apps and run on mini6410 ... I m getting Unable to handle kernel paging request at virtual address 40345000 Jun 03 11:16:02 ......................... Jun 03 11:16:22 sanket_ this dont looks like an oe error Jun 03 11:17:40 actually I builded an oe image for mini6410 with fltk..... Jun 03 11:18:01 hm and you mean oe causes a kernel error? Jun 03 11:18:35 woglinde_, I dont know.... just for my one fltk apps I m getting this error.... Jun 03 11:19:56 http://pastebin.mozilla.org/1241404 Jun 03 11:20:40 intressting connMann causes the kernel oops Jun 03 11:21:22 hm sorry I cannt help you this quick Jun 03 11:21:25 maybee florian can Jun 03 11:21:34 I m getting segmentation fault ..... then I try to debug through serial port.... I m getting this error Jun 03 11:21:37 he has most knowledge about the mini 64 Jun 03 11:21:51 why serial? Jun 03 11:21:58 run gdb on the mini itself Jun 03 12:13:00 woglinde_, how compiled a file using debugging symbols.... I had compiled gdb using oe... Jun 03 12:13:39 Hi everybody Jun 03 12:14:01 I am using oe and I use my image bb Jun 03 12:14:14 sanket_: you can install the various -dbg packages on your target Jun 03 12:15:17 I run it as bitbake my-image Jun 03 12:15:30 I want to give parameter to this command Jun 03 12:16:01 can I give parameter such as bitbake my-image 1.0.3? Jun 03 12:19:41 I dont want users to change version in local.conf, I just want to tell them run "bitbake my-image 1.0.3" Jun 03 12:20:47 is it possible to give command arguments when I call bitbake? Jun 03 12:22:06 sanket_ when you compile your programm via oe you can install the -dbg package to get debug symbols Jun 03 12:22:16 ups zecke was faster Jun 03 12:22:37 bitbake follalal-1.34 Jun 03 12:27:35 pb nice uclibc catch Jun 03 12:27:43 sent it upstream? Jun 03 12:28:25 in oe-core? Jun 03 12:28:41 yes in oe-core and to uclibc Jun 03 12:28:56 oh right, yeah. not yet Jun 03 12:29:25 I'm not quite sure how one goes about sending patches to the uclibc h4x0rs. I guess I should look that up. Jun 03 12:30:02 bugtracker Jun 03 12:30:09 carmelo is looking through it Jun 03 12:30:26 pb btw. I put linux scheduler stuff inside uclibc in sync Jun 03 12:30:37 if you are intressted I can send you the patches Jun 03 12:30:54 and I hacked excevpe together Jun 03 12:31:00 very good. don't think I need that right now, so I guess I will wait for your work to trickle down through the usual channels Jun 03 12:31:50 hehe its my work getting systemd running Jun 03 12:32:02 but now it seems there is some fscanf false Jun 03 12:32:07 will figure it out later Jun 03 12:32:51 woglinde_, I had used Jun 03 12:32:55 okay till later Jun 03 12:33:09 sanket yes bitbake gdb Jun 03 12:33:19 copy the ipk over Jun 03 12:33:25 opkg install gdb-bla.ipk Jun 03 12:34:32 woglinde_, I didnt get u about copying .ipk Jun 03 12:36:22 what is the replacement for amend.bbclass in the new openembedded-core? Jun 03 12:37:55 ensc: seems folks are just using bbappend now Jun 03 12:39:00 foerster: this requires a bitbake > 1.10, or? Jun 03 12:40:59 ensc: looks like it Jun 03 12:41:04 woglinde_ I had already build x11-image using oe... Jun 03 12:41:16 grepping in 1.10 doesn't yield any bbappend. grepping in 1.12 does. Jun 03 12:41:32 bad... will have to backport amend.inc then... Jun 03 12:42:02 I like amend, that's what I've been using in my oe.dev setup Jun 03 12:43:35 yes; I used it too very extensively Jun 03 12:44:39 ensc: theoretically, you can just drop amend.bbclass into your own layer and it will work Jun 03 12:45:06 no; it uses FILESPATHPKG which does not exist anymore Jun 03 12:45:21 ah. So much for that idea ;) Jun 03 12:47:12 Consider using bitbake 1.12 though, parsing time is greatly reduced thanks to kergoth. Jun 03 12:50:15 bitbake 1.12 fails with 'ImportError: No module named ply' Jun 03 12:51:14 try installing python-ply (or similar) on your distribution? Jun 03 12:55:27 does not exist for fedora; will backport/forwardport amend.bbclass Jun 03 12:55:59 hi Jay7 Jun 03 12:56:57 can I share the same recipe for normal kernel and kexecboot? Jun 03 12:57:12 the machine is nokia900 Jun 03 12:57:19 (the nokia n900) Jun 03 12:57:35 I'd like to use the meego kernel that is in oe for kexecbooting Jun 03 12:57:50 but I wonder if I should fork that recipe or not Jun 03 13:13:24 Not happening Jun 03 13:13:31 You need to do FP Jun 03 13:33:34 hi, can I create a ipk package repository on a server? Can someone point me in the correct direction. Is it just http or ftp server or is there a specialised daemon? Jun 03 13:34:08 guyvdb_: just http Jun 03 13:59:30 guyvdb_: put files in dir, run 'ipkg-make-index' Jun 03 14:01:15 hrw: thx Jun 03 14:02:07 hey everyone, I'm having trouble defining a task to build a customized rootfs for my mini2440. there is a git repository that offers a an image recipe for the mini2440 which is based on task-base-extended but which adds too many packages that I don't need. therefore I decided to write my own task-mini2440.bb to only install the bare minimum packages I need. Jun 03 14:02:59 niklas I am using a mini2440. I have a custom package... want to see my bb file? Jun 03 14:03:41 niklas: so, what is the problem you are having? Jun 03 14:04:09 just a second, I'm still trying to describe the problem properly ;) Jun 03 14:04:53 however, it seems like the bitbake doesn't resolve the dependencies properly (or more likely: I did something wrong). this is part of my custom task: RDEPENDS_task-mini2440 := "\ task-boot \, - task-boot.bb tells bitbake to include base-files. my final rootfs doesn't have these files though Jun 03 14:05:04 guyvdb_: I'd love to see your bb-file Jun 03 14:05:24 it's hard to say what might be wrong based on just that fragment Jun 03 14:05:35 if you put your whole .bb file in a pastebin then someone might be able to spot the problem Jun 03 14:06:02 I understand. I'll get on it Jun 03 14:06:05 http://pastie.org/2013840 Jun 03 14:06:12 niklas ^ Jun 03 14:11:15 thank you, I'll post my files in a second Jun 03 14:14:04 okay: the image recipe -> http://pastie.org/2013864 Jun 03 14:14:37 and the custom task -> http://pastie.org/2013867 Jun 03 14:16:38 something that also confuses me is that even though I listed sysvinit under IMAGE_INSTALL, no inittab is installed into the rootfs although the ipk of sysvinit lists sysvinit-inittab as depedency Jun 03 14:16:55 do you get any errors from ipk during rootfs construction? Jun 03 14:16:57 check log.do_rootfs Jun 03 14:17:06 s/ipk/opkg/ Jun 03 14:18:33 re Jun 03 14:21:06 http://lists.denx.de/pipermail/u-boot/2011-June/093776.html Jun 03 14:21:30 tartarus ping Jun 03 14:23:20 pb_: I didn't find any errors in the rootfs logs Jun 03 14:24:52 hm, right. I didn't notice anything obviously wrong with your recipe that you posted. Jun 03 14:25:02 Try inspecting the .ipk file for your task, does it have the right Depends set? Jun 03 14:27:44 is 'gitpkgv' supposed to work with oe-core and bitbake 1.12? I get there an 'AttributeError: 'FetchData' object has no attribute 'branches'' error because fetcher's urldata_init() method (which initializes branches[]) is called too late Jun 03 14:29:28 I think I might have an idea what my problem is. lemme test this real quick Jun 03 14:34:40 I removed the ipk files of the task by hand and rebaked the task recipe. now I get a "ERROR: task-mini2440 is listed in PACKAGES multiple times, this leads to packaging errors." when baking. I actually had this earlier but since the task summary said 0 failed I must have forgotten about it Jun 03 14:37:07 the only files where task-mini2440 is mentioned are task-mini2440.bb and mini2440-image.bb according to a find . | xargs grep 'task-mini2440' -sl Jun 03 14:38:57 yeah, it doesn't like you mentioning it in task-mini2440.bb Jun 03 14:39:03 try removing the PACKAGES assignment from there. Jun 03 14:39:08 ah jeez, there seems to be a missing quotation mark in the task file.... Jun 03 14:39:22 it says PROVIDES = "${PACKAGES} when it should say PROVIDES = "${PACKAGES}" Jun 03 14:39:44 yeah, I don't think that will do you any actual harm (due to the way the bitbake parser works), though Jun 03 14:47:53 ah pb Jun 03 14:47:58 pb you sscanf well? Jun 03 14:48:06 +know Jun 03 14:48:46 pb how can be fscanf(f, "%ma %*i %*i %i", &controller, &enabled) != 2 rewritten Jun 03 14:48:51 to work under uclibc? Jun 03 14:48:57 %ma is not handeld Jun 03 14:51:32 ups %ms Jun 03 14:51:34 sorry Jun 03 14:52:51 pb_: I removed the PACKAGES assigment from the task recipe, removed the ipk files by hand and rebuilt the task recipe again. this time there were no errors and the depends in the ipk now say "Depends: task-boot, task-distro-base, task-machine-base, dropbear, task-base-kernel26, task-base-alsa, task-base-vfat" which looks better Jun 03 14:53:12 I'll try to rebuild the rootfs now Jun 03 14:59:50 woglinde: I guess you can change %ms to %as Jun 03 15:00:04 unfornatly not Jun 03 15:00:08 segfaults than Jun 03 15:01:09 hm, that is strange Jun 03 15:02:01 I don't think there are any significant differences between the two. Jun 03 15:03:11 maybe uclibc's scanf is just broken Jun 03 15:04:03 no segfaults with libc too Jun 03 15:04:10 eglibc Jun 03 15:04:44 oh, weird. Jun 03 15:04:59 hm yes Jun 03 15:05:16 fscanf(f, "%as %*i %*i %i", &controller, &enabled) Jun 03 15:05:18 right? Jun 03 15:05:33 yeah, looks about right to me Jun 03 15:05:52 let me try it now Jun 03 15:06:37 hm, works for me Jun 03 15:06:55 int main() { char *controller; int enabled; scanf("%as %*i %*i %i", &controller, &enabled); } Jun 03 15:07:01 doesn't segfault and does set the vars as expected Jun 03 15:07:42 your 'f' isn't null or anything, is it? Jun 03 15:07:47 no Jun 03 15:07:51 thats tested before Jun 03 15:08:07 well, dunno then. if you compile the code fragment I pasted above, does that work for you? Jun 03 15:08:21 you can add a suitable printf at the end to inspect the values of the two variables, obviously Jun 03 15:09:15 weired it works with %ms Jun 03 15:09:21 but not %as Jun 03 15:09:42 what eglibc are you using? Jun 03 15:09:51 I was testing with 2.12.1-0ubuntu10.2 on my desktop Jun 03 15:11:37 2.13-4 Jun 03 15:11:45 hm Jun 03 15:11:55 maybe the eglibc h4x0rs deleted support for %a or something Jun 03 15:11:59 seems unlikely, but you never know Jun 03 15:12:32 infavour of m? Jun 03 15:12:35 yeah Jun 03 15:12:51 %a was always a gnu extension and it's been deprecated for a while because 'a' conflicts with iso c99 Jun 03 15:13:11 yes Jun 03 15:13:21 okay Jun 03 15:13:30 I will just made it oldschool Jun 03 15:13:36 but I dunno, just speculating. khem is closer to those h4x0rs than me, he might know what's going on. Jun 03 15:13:38 whith chararray Jun 03 15:13:43 heh Jun 03 15:13:44 righto Jun 03 15:13:59 03Michael 'Mickey' Lauer  07org.openembedded.dev * r7ae56b81f8 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Jun 03 15:14:07 wow, mickey lives Jun 03 15:14:08 03Michael 'Mickey' Lauer  07org.openembedded.dev * r449f80f331 10openembedded.git/recipes/vala/ (vala_0.12.0.bb vala_0.12.1.bb): vala[-native]: 0.12.0 -> 0.12.1 Jun 03 15:14:13 tadaa ;) Jun 03 15:14:23 hi mickey :-) Jun 03 15:14:27 hey pb_ Jun 03 15:14:43 hey, are you "the daddy" now? Jun 03 15:14:46 mickeyl: hey! Jun 03 15:14:55 mickeyl hehe Jun 03 15:14:57 unfortunately not. the little one is enjoying her time inside mummy... Jun 03 15:15:02 args Jun 03 15:15:03 we're now 9 days over the ETA Jun 03 15:15:09 ah, right Jun 03 15:15:10 tomorrow we'll start applying "force" Jun 03 15:15:35 hm its girl? Jun 03 15:15:42 yes, it's going to be a girl Jun 03 15:15:46 lol Jun 03 15:15:47 * mickeyl dances a jig Jun 03 15:15:49 they always comes late Jun 03 15:15:50 ah, right. good luck! Jun 03 15:15:55 thanks Jun 03 15:16:07 yeah, the statistics they use for figuring out what is "late" are rather crocked Jun 03 15:16:25 indeed. now that i know a bit more about these affairs it looks a lot like heuristics to me Jun 03 15:16:34 they don't know a lot about birth Jun 03 15:16:35 still Jun 03 15:16:44 yes obviously Jun 03 15:16:47 niklas: you there? Did the new root fs build correctly Jun 03 15:17:08 actually the timing is brillant Jun 03 15:17:12 We had a lot of funn with their high tech machines guessing the weight :-) Jun 03 15:17:20 so i can be best man on the marriage of my best friend today Jun 03 15:17:28 and Sabine could join Jun 03 15:17:35 party's continuing in a bit Jun 03 15:17:36 need to move Jun 03 15:17:39 cu guys Jun 03 15:17:40 mickeyl: heh... Jun 03 15:17:46 mickeyl: enjoy! Jun 03 15:17:50 thanks a lot Jun 03 15:17:55 will keep you posted about the "status" ;) Jun 03 15:18:05 okay :) Jun 03 15:18:19 mickeyl he there are some hours left Jun 03 15:18:20 ;) Jun 03 15:21:42 guyvdb_: unfortunately no. I don't get it. I removed all the task packages (the ipk files) and cleaned up with bitbake -c clean task-xyz. then I changed RDEPENDS_task-mini2440 in my task-mini2440.bb to DEPENDS_task-mini2440. shouldn't that force bitbake to build the dependencies as well? Jun 03 15:22:11 but the boot task or any other tasks didn't get rebuilt Jun 03 15:24:51 eeh, now there is no dependency line present in the mini-2440 ipk at all :S Jun 03 15:27:44 DEPENDS_ will do nothing at all. Jun 03 15:27:51 DEPENDS and RDEPENDS_ Jun 03 15:27:53 not DEPENDS_ Jun 03 15:28:06 DEPENDS is build time, RDEPENDS_ is run time Jun 03 15:28:18 (bitbake obeys both) Jun 03 15:30:56 morning kergoth Jun 03 15:31:09 oh okay Jun 03 15:31:44 hey pb_ Jun 03 15:32:58 he kergoth Jun 03 15:33:16 hey Jun 03 15:33:17 * kergoth yawns Jun 03 15:33:51 whats up? Jun 03 15:34:11 i hate imminent deadlines Jun 03 15:34:29 *g* Jun 03 15:34:53 I like the one dilbert stripe from this year calendar about deadlines Jun 03 15:38:32 hii any integrated plymouth with openembedded Jun 03 15:38:53 *any one Jun 03 16:07:02 I have a patch to contribute to the bitbake manual. Jun 03 16:07:26 I read through the entire manual and fixed all the non-controversial errors 9spelling and such). Jun 03 16:07:44 I sent a patch to bitbake-dev a few months back. Jun 03 16:07:53 But it din't appear on the list. Jun 03 16:08:04 I emailed the list admin address, but got no response. Jun 03 16:08:08 try the new bitbake-devel list Jun 03 16:08:13 we're ditching berlios Jun 03 16:08:39 http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/bitbake-devel Jun 03 16:08:46 probably need to subscribe to send to it Jun 03 16:09:21 I'll subscribe and resend. Jun 03 16:09:23 Thanks. Jun 03 16:09:30 cool. thanks for the contribution, we appreciate it Jun 03 16:09:37 we're obviously lacking in doc folk :) Jun 03 16:09:54 kergoth, isn't that a universal axiom? Jun 03 16:11:01 probably. doc & test are usually weak in open source projects Jun 03 16:11:47 OK, I appear to be subscribed. I'll send the patch. Jun 03 17:28:27 I sent a patch to bitbake-devel about an hour ago, but it's held by the moderator. Does the moderator really exist, or does that mean it was rejected as spam? Jun 03 17:28:40 its likely richard is the moderator (RP) Jun 03 17:28:45 he created the list afaik Jun 03 17:29:16 pingswept: are you a member of the bitbake-devel list? Jun 03 17:29:32 kergoth: Thanks. I'll see if he replies here and send him an email if not. Jun 03 17:29:55 morphis: Yes. The moderation was because the message was too large. Jun 03 17:30:02 pingswept: ok Jun 03 17:38:35 re Jun 03 17:47:17 https://gist.github.com/1006774 - this shouldn't even be possible Jun 03 17:47:19 * kergoth mutters Jun 03 17:58:56 gah, this makes NO SENSE Jun 03 18:01:47 hehe Jun 03 18:01:49 close invalod Jun 03 18:01:55 ups invalid Jun 03 18:01:56 *g* Jun 03 18:02:29 if i add ${PV} to TARGET_CFLAGS in this build environment, it blows up bitbake with the above error Jun 03 18:02:45 but we catch every exception raised in the parse function, and add the recipe attribute to it, the above shouldn't be possible Jun 03 18:02:48 :| Jun 03 18:03:19 it must be a PV in one of these recipes doing something crazy, but i don't see how it could be showing up like this Jun 03 18:03:21 hrm Jun 03 18:05:47 aha Jun 03 18:06:11 there was an exception coming from the exception handler! Jun 03 18:06:13 didn't think about that Jun 03 18:07:50 Sad panda, I'm out of coffee. Jun 03 18:09:49 take that, bug! Jun 03 18:10:42 ! Jun 03 18:28:43 ERROR: Failure expanding variable SRCPV, expression was ${@bb.fetch.get_srcrev(d)} which triggered exception RuntimeError: maximum recursion depth exceeded while calling a Python object Jun 03 18:28:55 anyone ever seen this one? not fetch2, stock fetch Jun 03 18:29:03 hm was append_uclibc-linux right or append_uclibc = Jun 03 18:30:22 03Chris Larson  07master * rc9bb37e588 10bitbake.git/lib/bb/exceptions.py: Jun 03 18:30:23 bb.exceptions: don't choke on frames without arguments Jun 03 18:30:23 Signed-off-by: Chris Larson Jun 03 18:30:53 re Jun 03 18:31:04 jo florian Jun 03 18:36:27 damnit Jun 03 18:39:34 does anyone have a mirror of confuse-2.7.tar.gz I can't seem to get it from the default url Jun 03 18:42:02 ericben: ping? Jun 03 18:54:54 03Khem Raj  07master * r688b8dc036 10openembedded.git/recipes/busybox/busybox_1.18.4.bb: Jun 03 18:54:54 busybox_1.18.4.bb: Fix checksums of busybox-1.18.4-hush.patch Jun 03 18:54:54 Signed-off-by: Khem Raj Jun 03 18:57:04 ericben: it looks like busybox regenerated their 1.18.4 patches, any idea why? Jun 03 18:57:36 Tartarus hm arent they doing it every hour? Jun 03 18:58:03 Tartarus would be cool when you could make sqlite 3.7 the default in oe master too Jun 03 19:03:20 Tartarus khem fixed it? Jun 03 19:04:37 hi :) How to include in my overo build? I include php package into my image file, but I dont' get php executables, only configuration? Jun 03 19:07:22 mrmichael look what packages are will be generated Jun 03 19:40:35 is this kernel bb file ok ? http://dpaste.com/550211/ Jun 03 20:38:45 woglinde: hello, can you support me please to get navit running on Freerunner ?, I get the following error: navit:vehicle_new:invalid type 'gpsd' Jun 03 20:39:10 nschle85 you are proably missing the gpsd modul Jun 03 20:39:56 how do i enable it ? Jun 03 20:42:35 In the navit lib dir i can find a libvehicle_gpsd_dbus.so, i should be loaded automatocally i think Jun 03 20:42:37 look if it is split out into a package Jun 03 20:42:58 yes but than you have to alter your config Jun 03 20:43:04 used gpsd_dbus Jun 03 20:43:09 not gpsd Jun 03 20:44:04 do you have an example ? Jun 03 20:47:42 look at the wiki Jun 03 20:47:56 there are all config options described Jun 03 21:03:23 03Koen Kooi  07org.openembedded.dev * r5348fa080f 10openembedded.git/recipes/linux/ (3 files in 2 dirs): Jun 03 21:03:23 linux-omap4 2.6.35.3: backport fixes from L24.13 kernel in meta-texasinstruments Jun 03 21:03:23 Signed-off-by: Koen Kooi Jun 03 21:03:33 03Koen Kooi  07org.openembedded.dev * r1672042214 10openembedded.git/recipes/linux/ (4 files in 2 dirs): Jun 03 21:03:33 linux-omap 2.6.39: sync with meta-texasinstruments again Jun 03 21:03:33 Signed-off-by: Koen Kooi Jun 03 21:12:04 damnit, i'm totally stuck on this bb.fetch recursion issue Jun 03 21:12:05 hrmph Jun 03 21:21:01 aha, whew Jun 03 21:21:16 a rogue := screwing everything up Jun 03 21:21:19 hrmph Jun 03 21:34:07 GNUtoo: hi Jun 03 21:34:25 GNUtoo: it should be possible to use same kernel Jun 03 21:34:28 Jay7, kexecboot works on the n900, I've to integrate it into oe tough Jun 03 21:34:32 how? Jun 03 21:34:43 include linux-kexecboot.inc? Jun 03 21:34:45 GNUtoo: what was problem with fb ui first time? Jun 03 21:35:14 GNUtoo: hm.. better to ask andrea about this Jun 03 21:35:15 basically it's long to boot and I have no feedback Jun 03 21:35:24 that is the drawbacks Jun 03 21:35:29 long to boot means: Jun 03 21:35:37 * it's very fast up to the GUI Jun 03 21:35:48 * right after kexecing the kenrel it take ages Jun 03 21:36:27 do you have same kernel configs for both? Jun 03 21:36:33 yes but Jun 03 21:36:38 in one I used modules Jun 03 21:36:44 in the other no modules were included Jun 03 21:36:48 ah Jun 03 21:37:30 well.. ask ant about better integration way Jun 03 21:37:39 I'm not very anknowledged Jun 03 21:39:32 ok Jun 03 21:41:45 else I could still fork my kernel recipe Jun 03 21:41:57 or use an inc Jun 03 21:44:13 Is there a way to remove a package from DISTRO_FEATURES? I can override it with DISTRO_FEATURES_local, but that results in me forcing the inclusion of stuff I may later want to remove. Jun 03 21:44:45 make your own distro Jun 03 21:45:05 Yeah, that's probably the right solution. Jun 03 21:45:49 I was hoping there was a less intrusive way to do it. All I'm really trying to do is remove tk from Angstrom. Jun 03 21:52:33 uhhh Jun 03 21:52:38 let's say you have: Jun 03 21:53:05 linux-kexecboot.inc->linux-nokia900-meego.inc->linux-nokia900-meego_git.bb Jun 03 21:53:16 how to handle INC_PR? Jun 03 21:54:55 I'll fork it's better Jun 03 21:55:06 else I've the INC_PR issues Jun 03 22:01:15 * Jay7 -> sleep Jun 04 01:00:57 OE takes care of installing shared libs by looking at the binary's SO_NEEDED? That is awesome. :)))) Jun 04 01:01:43 me <3 oe Jun 04 01:09:10 In http://docs.openembedded.org/usermanual/usermanual.html, should the line that says "based on what the machine needs along with what the distro supports" actually be " based on what the distro needs along with what the machine supports" ? **** ENDING LOGGING AT Sat Jun 04 02:59:56 2011