**** BEGIN LOGGING AT Fri Nov 27 02:59:58 2009 Nov 27 06:35:10 03Thomas Zimmermann  07org.openembedded.dev * rf7d4b310b5 10openembedded.git/ (2 files in 2 dirs): Nov 27 06:35:10 babiloo: new recipe Nov 27 06:35:10 Signed-off-by: Klaus Kurzmann Nov 27 06:40:43 03Benjamin Schieder  07org.openembedded.dev * rb0005a48b5 10openembedded.git/recipes/fltk/fltkcocktailbar_svn.bb: Nov 27 06:40:43 add package fltkcocktailbar Nov 27 06:40:43 Signed-off-by: Klaus Kurzmann Nov 27 06:45:32 morning Nov 27 06:46:35 03Klaus Kurzmann  07org.openembedded.dev * r5be7660163 10openembedded.git/recipes/tasks/task-shr-feed.bb: Nov 27 06:46:35 task-shr-feed: add fltkcocktailbar and babiloo-efl Nov 27 06:46:35 Signed-off-by: Klaus Kurzmann Nov 27 08:01:07 Hi there. I get an ERROR with gcc-cross-4.3.3-r9.1: http://pastebin.ca/1689228 Please help! :) Nov 27 08:09:21 Anyone here...? Nov 27 08:09:49 Hasse__: looks like the error is in libobjc? Are you building with ObjectiveC support? Nov 27 08:10:29 DJWillis: ehmm... guess not...? Nov 27 08:10:51 DJWillis: what option would do that? Nov 27 08:12:07 Hasse__: it may be there by default, just wanted to check it was nothing you had done to the recipe. Nov 27 08:13:10 DJWillis: I don't see why I would need that - we use C++ only :) Nov 27 08:13:24 Hasse__: I assume you have tried bitbake -cclean gcc-cross-4.3.3 && bitbake gcc-cross-4.3.3? Nov 27 08:18:55 03Graeme Gregory  07org.openembedded.dev * r71e2c5d143 10openembedded.git/recipes/gcc/ (gcc-4.3.3.inc gcc-4.3.3/arm-gcc-objective.patch): Nov 27 08:18:55 gcc-4.3.3.inc : replace objc patch with undamaged one from gcc-patches Nov 27 08:18:55 There is a one hunk peice of damage in the patch originally committed, as Nov 27 08:18:55 compared to the patch posted on gcc-patches. Replacing with the original Nov 27 08:18:55 patch enables platforms like arm oabi to compile again. Nov 27 08:19:35 * XorA|gone suspects he pushed that fix at just the right time Nov 27 08:20:20 Hasse__: that bug is fixed Nov 27 08:22:15 DJWillis, XorA|gone: Oki thanks, I'll investigate my system for branch updates etc. Nov 27 08:22:32 Hasse__: I pushed the fix only 1 minute ago Nov 27 08:22:57 XorA|gone: :D .. ah, okay - it sounded like days ago ;) Nov 27 08:23:09 XorA|gone: ahhh, well timed push Nov 27 08:23:39 XorA|gone: double thanksThen; Nov 27 08:23:41 DJWillis: I found the bu yesterday afternoon, but needed to locate/test a fix Nov 27 08:23:57 DJWillis: it seems wherever koen got that patch did a manual merge and fscked it up Nov 27 08:25:00 XorA|gone: Doh!, always the way Nov 27 08:25:44 new patch also has original email header so its properly acknowledged Nov 27 08:40:06 I've a problem with do_fetch. it fails to download this SRC_URI = "http://bundles.openttdcoop.org/opengfx/releases/opengfx-0.1.1.zip" Nov 27 08:40:58 It's because of .zip or because the filename after downlowding is: download.php?file=opengfx%2Freleases%2Fopengfx-0.1.1.zip ? Nov 27 08:59:05 Heinervdm: it doesnt handle when http servers rename files Nov 27 09:06:02 XorA|gone: can i workaround this somehow? Nov 27 09:08:59 ok, found a way to work around. tried to download with that filename and that works Nov 27 09:09:19 I was going to say find somewhere that doesnt have the file renaming Nov 27 09:29:57 hello all together Nov 27 09:31:29 i am a newbie and i have a general question to openembedded Nov 27 09:33:43 yesterday i was installing and configuring openembedded and bitbake. After i had done so, i followed the Getting Startet. I tried to execute the command bitbake nano. It starts do work and running the tasks. Nov 27 09:34:34 but at the command "NOTE: Running task 293 of 435 (ID: 7, /home/The_Collar/oe/openembedded/recipes/glibc/glibc_2.9.bb, do_compile)" it works and works and works all over the night Nov 27 09:36:50 and now it works nearly 24hours but it does not go on! How much time will it take in general to execute a complet bitbake nano comand? Nov 27 09:39:38 what's your CPU? how much RAM do you have? Nov 27 09:39:41 good morning Nov 27 09:39:50 building glibc is slow Nov 27 09:40:00 i use a pentium 4 (3ghz) with 512mb RAM Nov 27 09:41:18 the first time you run bitbake , it takes a lot of time because it needs to build the toolchain. Maybe you are a bit low on RAM. I think maybe it's normal for it to take more than 24 hours Nov 27 09:42:13 I use a Pentium D with 1 GB RAM and I can't remember exactly how much time did I need, but I think it was between 12 and 18 hours Nov 27 09:42:24 to bitbake nano Nov 27 09:42:28 ok! If it's normal then I will let it work Nov 27 09:43:34 I only wonder about the long time it takes and I supposed my pc has a problem Nov 27 09:45:13 thanks for the answer! I will let it work, if necessary all over the weekend - then it should be finished! Nov 27 09:49:54 goodbye! Nov 27 09:56:10 hi heinervdm Nov 27 10:09:51 morning Nov 27 10:11:25 question about kernel_do_stage() in kernel.bbclass: I'd need to add "include/mtd". Nobody else ever needed mtd headers? Nov 27 10:11:45 we ended up staging own copy in the recipe :/ Nov 27 10:30:39 hi Nov 27 10:30:59 What is the difference between mesa/mesa-dri/mesa-xlib ? Nov 27 10:34:17 mesa-dri is Mesa with the dri backend enabled Nov 27 10:34:23 mesa-xlib is Mesa with the xlib backend enabled Nov 27 10:34:29 mesa itself is, as far as I can tell, historical cruft Nov 27 10:36:47 03Klaus Kurzmann  07org.openembedded.dev * r7b2a48f169 10openembedded.git/recipes/shr/phoneuid_git.bb: Nov 27 10:36:47 phoneuid: conform with the recent changes in upstreams Makefiles Nov 27 10:36:47 Signed-off-by: Klaus Kurzmann Nov 27 10:45:34 03Thomas Zimmermann  07org.openembedded.dev * r6372f6f812 10openembedded.git/ (5 files in 3 dirs): Nov 27 10:45:34 openttd: add version 0.7.4-RC1 and sound and graphic files from openttdcoop.org Nov 27 10:45:34 * Can't reuse .inc because openttd won't build with --target configure option Nov 27 10:45:34 * Sounds and Graphics from openttdcoop.org are GPL and CC licensed files to replace the Nov 27 10:45:36 propertarie files needed by this game Nov 27 10:45:38 Signed-off-by: Klaus Kurzmann Nov 27 11:02:39 morning all Nov 27 11:02:43 hi rp Nov 27 11:02:59 ant_work: mtd headers are added by mtdutils iirc Nov 27 11:04:45 hi RP Nov 27 11:06:09 morning Nov 27 11:06:33 hi hrw Nov 27 11:06:45 hi hrw Nov 27 11:08:10 Getting a new bug now, where it fails to find the md5 files :( is this fixed upstreams or totally new? Nov 27 11:08:25 NOTE: Task failed: Unknown fetch Error: [Errno 2] No such file or directory: '/mnt/1TB/openembedded/sources/vim-7.0.tar.bz2.md5' Nov 27 11:32:21 re hi Nov 27 11:44:38 May anyone help me with a simple doubt? If I will use a 2.6.29 kernel version, do i need the "wifi" flag in the machine features? This doubt is because i get a conflict with madwifi and this kernel version Nov 27 11:46:23 rvfo: What kind of conflict? Nov 27 11:49:28 florian: when bitbake tried to compile madwifi, I get a configure error, in the log was written to set PCI modules in the .config, and I set the flag "pci" in the MACHINE_FEATURES Nov 27 11:57:49 rvfo: sounds like a bug, can you paste the error somewhere? Nov 27 11:58:23 florian: just a minute i will look for Nov 27 12:03:56 forian: i lost the pastebin link, i will to compile again and post, thanks for attention Nov 27 12:14:00 RP: yes, it seems mtd-abi.h and mtd-user.h appear staged by mtd-utils. I'll have to DEPEND on them. Nov 27 12:15:03 but finally, why not staging in kernel.bbclass? Nov 27 12:56:35 RP: I see now nandlogical already depends from mtd-utils. Probably the issue was just with the klibc static version and some -nostdinc flags... Nov 27 12:57:07 yes it was like that Nov 27 13:12:02 03Klaus Kurzmann  07org.openembedded.dev * rd6aa55b485 10openembedded.git/recipes/shr/phonefsod_git.bb: Nov 27 13:12:02 phonefsod: conform with latest autotools configuration changes in upstream Nov 27 13:12:02 Signed-off-by: Klaus Kurzmann Nov 27 13:18:43 hmm... whom has one to beg pardon for when one added a branch by accident :c Nov 27 13:34:22 bye Nov 27 13:35:25 mrmoku, I think mickey|zzZZzz can delete branches Nov 27 13:37:24 In his sleep! Nov 27 13:45:52 Crofton|work: ok, thanks :) Nov 27 14:14:41 03Sebastian Spaeth  07org.openembedded.dev * r5b7e82c4c2 10openembedded.git/recipes/bootchart-lite/bootchart-lite_svn.bb: Nov 27 14:14:41 bootchart-lite: Use mkdir rather than install in postinst. Nov 27 14:14:41 * install is part of coreutils which might not be installed. Nov 27 14:14:41 * As this is a -lite image it might not make sense to pull in coreutils just to create a dir. Nov 27 14:14:41 Signed-off-by: Sebastian Spaeth Nov 27 14:14:46 03Sebastian Spaeth  07org.openembedded.dev * rafe9a3e394 10openembedded.git/: Merge branch 'org.openembedded.dev' of ssh+git://git@git.openembedded.net/openembedded into org.openembedded.dev Nov 27 14:14:47 03Sebastian Spaeth  07org.openembedded.dev * rea07ac9bb2 10openembedded.git/recipes/openmoko-3rdparty/babiloo-efl_2.0.9-3.bb: Nov 27 14:14:50 babiloo-efl: add tar-based recipe in addion to _bzr variant. Nov 27 14:14:52 Signed-off-by: Sebastian Spaeth Nov 27 14:18:32 03Sebastian Spaeth  07org.openembedded.dev * r3310e71744 10openembedded.git/recipes/openmoko-3rdparty/babiloo-efl_bzr.bb: Nov 27 14:18:32 babiloo-efl: don't prefer _bzr version over releases Nov 27 14:18:32 Signed-off-by: Sebastian Spaeth Nov 27 14:35:25 good morning Nov 27 14:35:39 gm Nov 27 14:36:18 I'm getting a build failure for python-pygobject on my x86_64 machine: http://www.sakoman.net:8000/public/logs/task/145519.txt Nov 27 14:36:58 works fine on my other non-64 bit machines Nov 27 14:37:04 any ideas? Nov 27 14:37:32 hmm, missing header files Nov 27 14:39:37 Crofton|work: well, it thinks so. but they are actually there Nov 27 14:39:55 it built on my few day old checkout Nov 27 14:39:56 pulling Nov 27 14:40:05 has the bb file changed altely? Nov 27 14:40:29 hmmm . . . seems that on 32 bit machines it generates a -I for the glib-2.0 include directory but does not on 64 bit machines Nov 27 14:40:58 so perhaps packageconfig for glib-2.0 is broken on 64 bit Nov 27 14:41:41 morning Nov 27 14:42:32 gm rsalveti Nov 27 14:43:06 Crofton|work: I don't see any suspicious recent changes Nov 27 14:43:48 Crofton|work: I ran into this last night, was too tired to deal with it. So I tried a clean build while I slept. Same issue :-( Nov 27 14:47:54 I'll poke and see if I ahve the same issue Nov 27 14:47:56 Crofton|work: well forget the glib-2.0 pc theory. the pc files are identical on 32 and 64 bit Nov 27 14:48:32 so it must be something in the config process for python-pygobject Nov 27 14:49:23 cause the failing command on 64 bit is clearly missing the -I for glib-2.0 Nov 27 14:50:34 mine is rebuild gcc-cross atm Nov 27 14:51:20 Crofton|work: sorry Nov 27 14:51:36 heh Nov 27 14:52:13 Crofton|work: I wont be around much tonight and tomorrow, prod eV members to vote :-) Nov 27 14:52:27 prod Nov 27 14:53:23 Hey, let people exercise their democratic right to apathy! Nov 27 14:54:20 well, I don't think we get hung up on quorum issues Nov 27 14:54:36 Apathy was not an option on the ballot :-D Nov 27 14:55:36 03Sebastian Spaeth  07org.openembedded.dev * red569d68e0 10openembedded.git/recipes/tasks/task-shr-feed.bb: Nov 27 14:55:36 task-shr-feed: add babiloo-efl, remove hermining. Nov 27 14:55:36 * Also sort some entries alphabetically. This list is a mess. Nov 27 14:55:36 Signed-off-by: Sebastian Spaeth Nov 27 14:58:49 Crofton|work: I suspect your build will work Nov 27 14:59:11 the error happens in the girepository subdirectory build Nov 27 15:00:08 and OE repo currently doesn't have a gobject-introspection recipe Nov 27 15:00:19 so you will never enter that subdirectory Nov 27 15:00:49 I've added that recipe so I am taking a path that hasn't typcally been taken before Nov 27 15:01:11 so I guess I'm on my own :-( Nov 27 15:02:38 heh Nov 27 15:06:45 Interesting, it generates GLIB_CFLAGS properly, but just doesn't include them in the INCLUDES definition Nov 27 15:06:59 so perhaps a bit of sed magic . . . Nov 27 15:13:06 good morning **** BEGIN LOGGING AT Fri Nov 27 15:32:52 2009 Nov 27 15:44:45 03Sebastian Spaeth  07org.openembedded.dev * ra5e4a8b36a 10openembedded.git/recipes/openmoko-3rdparty/babiloo-efl_2.0.9-3.bb: Nov 27 15:44:45 babiloo: upstream changed SRC_URI Nov 27 15:44:45 Signed-off-by: Sebastian Spaeth Nov 27 15:56:23 I have an important question about sending e-mail containing patches using the company e-mail Nov 27 15:57:27 ok Nov 27 15:57:56 my fantastique e-mail server doens't send perfectly the body Nov 27 15:58:12 that is bad ,,,,' Nov 27 15:58:16 some ">" or other character can be changed Nov 27 15:58:27 any suggestion? Nov 27 15:58:33 best thing to do is send them to yourself and see if you can apply them Nov 27 15:58:37 I'm getting crazy to explain them Nov 27 15:58:53 I did it, and patches don't apply Nov 27 15:59:10 maybe you can get yourself a gmail account and send them from there. Nov 27 15:59:37 pb_: it is an idea, but we want company visibility of our work Nov 27 16:00:12 I'm not quite sure what you mean by "company visibility". Nov 27 16:00:51 pb_: we work on pxa270 to improve a product and we want to contribute offically to kernel Nov 27 16:01:11 then we'll work on davinci , .... so on Nov 27 16:01:59 if this is the kernel, then I imagine the upstream h4x0rz would rather that you publish your changes in a git tree anyway. Nov 27 16:02:12 so you might be better off doing that rather than worrying about making applyable patches. Nov 27 16:02:41 We have a person good to write and send patches Nov 27 16:04:37 pb_: Upstream kernel wants patches via e-mail. Nov 27 16:04:49 recalcati: Traditional approach is to use a webmail account (eg, gmail) Nov 27 16:05:02 recalcati: Or other third party SMTP server to send patches. Nov 27 16:05:03 what good is an MTA that trashes all emails? Nov 27 16:05:24 recalcati: eg, at Wolfson we have a separate domain which bypasses Exchange for all public Linxu work. Nov 27 16:05:25 ah, good idea, you could check the dns blacklists to find an open relay and then use that. Nov 27 16:05:39 you can register your company on GHS and send as though you were sending from company email server Nov 27 16:05:43 pb_: and promptly get blacklisted :) Nov 27 16:06:09 Google Hosted Services doesnt have to be your only MTA but it can be one of them Nov 27 16:06:48 recalcati, use .com address on signoff Nov 27 16:07:07 Note also that you can set the e-mail address for patches that get applied separately to the sending user so you can do things like send from a personal account with the patch author set to be an account on the fucked server. Nov 27 16:07:23 that will persist in the log entries long after the email address the patch came from Nov 27 16:07:49 Crofton|work: git records teh patch author and in the kernel we're fairly careful to apply patches properly. Nov 27 16:08:04 Crofton|work: So you need to get *both* right. Nov 27 16:08:07 urg Nov 27 16:08:36 Crofton|work: but "From: blah" as teh first line of the log body allows sending patches with an author different to the sender. Nov 27 16:09:52 Thanks, it seems the betst solution is the "at Wolfson we have a separate domain which bypasses Exchange for all public Linxu work." Nov 27 16:10:28 That's the ideal; folks like Marvell do use the GMail approach which does also work effectively. Nov 27 16:10:58 but marvell doesn't appear inside the commit message Nov 27 16:12:53 broonie: "Note also that you can set the e-mail address for patches that get applied separately to the sending user so you can do things like send from a personal account with the patch author set to be an account on the fucked server" what does it mean? can explain me the sintax ? Nov 27 16:13:02 syntax Nov 27 16:13:06 16:08 < broonie> Crofton|work: but "From: blah" as teh first line of the log body allows sending patches with an author different to the sender. Nov 27 16:13:16 git format-patch will DTRT for you. Nov 27 16:15:25 broonie: but, it is correct way to do it? can I have problems? Nov 27 16:15:45 Yes, that's perfectly normal - standard tools like git am work with it. Nov 27 16:16:00 If you want to test stuff I'd suggest mailing it to yourself and trying to apply with git am. Nov 27 16:16:01 broonie: thx Nov 27 16:16:17 If the resulting log looks OK everything is working as expected. Nov 27 16:17:53 broonie: I'll do it using my private account, but, anyway, I have to register myself to kernel to obtain a machine number. I'll do it with my company e-mail and than if with it I'll not able to send patches, I'll use a private one with "From: myself_private_account" Nov 27 16:31:44 thx to all Nov 27 16:32:35 bye bye Nov 27 16:36:05 * * OE Bug 5344 has been created by tdomhan(AT)gmail.com Nov 27 16:36:07 * * missing checksums for v2.6.14-gitcurrent.patch and full.patch Nov 27 16:36:09 * * http://bugs.openembedded.org/show_bug.cgi?id=5344 Nov 27 17:06:15 Is there an easy way to tell (in a bb) if an arch is ARM and ideally if it's ARMv5>? Just cleaning up a bb that can make use of a lot of ARM opt. code patches if you can be sure of the arch at build time. Nov 27 18:10:17 03Liam Girdwood  07org.openembedded.dev * r0e0de60f44 10openembedded.git/ (8 files in 3 dirs): Nov 27 18:10:17 pulseaudio: add version 0.9.21 Nov 27 18:10:17 Signed-off-by: Liam Girdwood Nov 27 18:30:57 eV members remember to vote! Nov 27 18:32:46 vote early, vote often! Nov 27 18:33:13 dont let the eV fail in a pile of apathy! Nov 27 18:35:39 Crofton|work -- you sound like a Chicago citizen! Nov 27 18:36:02 heh Nov 27 20:00:57 re Nov 27 20:04:04 Any objections to give SDL a bit of TLC and update to 1.2.14 and also enable FB and X11 while I am at it rather than lots of seperate SDL's (also, setup TSLIB support while I am at it). Nov 27 20:14:37 DJWillis: you probably ought to use your new DISTRO_FEATURES skills to make directfb, x11 and tslib conditional :-) Nov 27 20:14:45 other than that it sounds like a fine idea to me Nov 27 20:18:40 pb___: that's just what I have in mind ;), the X11 bit I am unsure of but the rest is getting there. 1.2.14 is just test building now but I can't see why half the patches are needed anymore. Nov 27 20:31:38 gehe gleich zum essen Nov 27 20:31:52 stimmt Nov 27 20:32:05 da lohnt sich nichts mehr anzufangen :-) Nov 27 20:32:26 und du ? Nov 27 20:40:34 pb___: (or others) is ok considered ok to move a package from a platform arch (i.e. ARMv7) to a machine arch? It just strikes me that is the RIGHT way to sort libSDL and build the touchscreen support based on the machine features. Nov 27 20:42:03 DJWillis: most of the features in question are going to be distro ones, not machine ones. If there are any DISTROs that want to split it per machine then yes, making the libsdl package be machine specific would be an ok solution for those DISTROs. Nov 27 20:42:20 but, in the general case, it is not very desirable to make it machine specific. Nov 27 20:42:59 pb_: snag is 'touchscreen' is machine level, hmmm, not sure the best way to cleanly do this, may just stick my head in the sand and claim it never happened ;-) Nov 27 20:43:27 DJWillis: are there any DISTROs which need both a touchscreen-enabled and a non touchscreen enabled libsdl? Nov 27 20:44:48 hi, does someone use mpi with boost Nov 27 20:44:48 ? Nov 27 20:44:59 pb_: I can't think of any, the current setup with it off by default is not great to be honest. Maybe just turn it on and let people shout, it can't go wrong without it but it then has to drag in tslib. Nov 27 20:45:02 I guess not Nov 27 20:45:17 but maybe with a cluster of sheevaplug? Nov 27 20:45:35 or high-end power5 architecture Nov 27 20:45:47 DJWillis: if enabling touchscreen support doesn't break anything then I would expect most heterogenous DISTROs can just turn it on and accept that they will get a useless copy of tslib installed on non touchscreen machines. Nov 27 20:45:51 tslib isn't very big, after all Nov 27 20:46:20 and, if any such heterogenous DISTROs have a problem with that, they can do the work to make touchscreen support be per-MACHINE for them. :-} Nov 27 20:46:31 pb_: very true Nov 27 20:46:41 indeed it's a machine feature Nov 27 20:46:46 moreover some machine use evdev Nov 27 20:46:50 some other have tslib Nov 27 20:47:09 some don't have touchscreen at all and have restricted space Nov 27 20:47:38 gnutoo: hmm, yep, but if space is that tight are they going to carry SDL ;-) but good point on tslib/evdev. Hmmm Nov 27 20:49:58 lol I didn't read the beginning Nov 27 20:52:03 gnutoo: anything you want to chip in? Trying to work out the easiest way for me to get 1.2.14 in OE ;) Nov 27 20:52:40 I'll try wesnoth Nov 27 20:52:50 not yet Nov 27 20:53:01 I'll add options if I needed them for wesnoth Nov 27 20:54:13 can I disable mpi with boost? Nov 27 20:54:23 or is there an mpi lib in oe Nov 27 20:54:33 find + grep don't find anything Nov 27 20:54:39 only compile patches Nov 27 20:56:17 florian: How are you? I talked with you earlier today, about a conflict with the flag "wifi" and the kernel version, only now i can repeat the error, may you help me? Nov 27 20:56:53 roliveira: yes sure, i'll try Nov 27 20:57:28 florian: the complete error is here: http://pastebin.com/m238ed054 Nov 27 20:59:42 florian: i set the flag pci in the MACHINE_FEATURE, but it didn't solve Nov 27 21:03:05 roliveira: I see... it looks like PCI is turned off in the kernel configuration. Nov 27 21:04:56 indeed it checks for kenrel config... Nov 27 21:05:41 florian: I thought that to, maybe, would I set this manually? And another strange thing, in the distro conf I set kernel version equal = 2.6.29 and this error refers to 2.6.25 version Nov 27 21:06:45 florian: gntoo: Is it possible to do automatically? Nov 27 21:07:51 roliveira, there are several ways to do it: Nov 27 21:08:12 *change the .config with make ARCH=arm menuconfig,save,copy back into the defconfig in the oe tree Nov 27 21:08:32 *use a menuconfig task(I don't remember the exact name of the task) Nov 27 21:09:08 *look if the kernel recipe checks for the pci machine feature,if not implement that and >> to the config Nov 27 21:09:44 but can't you use ath5k that is FLOSS? Nov 27 21:11:41 gnutoo: he is building for i486 :) Nov 27 21:12:03 gnutoo: i just put in MACHINE_FEATURES ".. pci.. " and the kernel version = 2.6.29 and executed the command "bitbake task-base" Nov 27 21:12:16 gnutoo: and this happened Nov 27 21:12:33 ok Nov 27 21:12:49 gnutoo: yes I have an old geode NSC Nov 27 21:12:56 roliveira: you don't do this. usually you have a kernel recipe that supports your device and comes with a defconfig file that matches the needs of your device. Nov 27 21:13:21 roliveira: which kernel does it build actually? Nov 27 21:13:21 so remove the ARCH=arm,sorry for the error Nov 27 21:15:32 florian: I set to the 2.6.29 and in the error is wrote 2.6.25. The problem is that I need a newer version than 2.6.25, because I need to use a camera, and I read that old version than 2.6.28 doesn't support this Nov 27 21:26:18 roliveira: this means you bild the wrong kernel anyway. do you build for a device that was in oe already of for something you added? Nov 27 21:32:23 yes, i was trying to build one device with a newer kernel version, because I need the drivers to video capture and I only found this way to do that. Because my hardware have a webcam, nowaday I'm using a simple debian distro with kernel 2.6.28, but I'm needing space in the hardware Nov 27 21:35:21 florian: and this is the reason to use OE and another things, I didn't know OE before yesterday and I liked Nov 27 21:36:23 roliveira: ok right, you need to make sure the newer kernel is set to be compatible to your device and comes with a matching defconfig Nov 27 21:38:24 florian: I'll look for this, if It isn't compatible I will write the .config, thanks so much Nov 27 21:38:57 roliveira: that's quite easy. look at the older kernel bb file and see how it is done. the defconfig you might be able to copy from the old kernel. this goes to the subdir $new-kernel-$version/$machine/ Nov 27 21:39:41 florian: thanks a lot, i will do that Nov 27 21:39:53 roliveira: otherwise the old gernel gets picked because no new one is available. Nov 27 21:39:57 roliveira: yw Nov 27 21:49:19 can I disable MPI in boost 1.40 ? Nov 27 21:49:40 or where can I find an mpi lib in oe tree....I didn't find one Nov 27 22:05:47 hm what a mess Nov 27 22:05:58 webkit-gtk needs over 6 gig to build Nov 27 22:07:08 ... and a lot of cpu cycles Nov 27 22:11:00 Don't forget to vote guys Nov 27 22:12:04 hm Nov 27 22:12:15 seems I am on private or so Nov 27 22:12:20 args not Nov 27 22:12:24 where to vote? Nov 27 22:13:58 jo ant Nov 27 22:14:19 Crofton|work: oh right Nov 27 22:14:31 :-) Nov 27 22:14:55 hey EU devs Nov 27 22:15:06 woglinde: see the oe-members list, xora sent out a ballot Nov 27 22:15:09 amys still digesting... Nov 27 22:15:43 yeah seems I have to subscribe Nov 27 22:16:11 hrm... didn't we subscribe all members? Nov 27 22:17:03 florian, better check against the member list on the website Nov 27 22:17:07 seems I am no Nov 27 22:17:08 t Nov 27 22:17:36 Crofton|work: Yes I'm working on this already... Nov 27 22:18:39 hi spaetz Nov 27 22:19:08 hi Nov 27 22:20:07 RP: ping? Nov 27 22:20:43 RP: about that mtd headers, it happens that compiling with klcc only mtd.h is in the path Nov 27 22:22:26 so it's just matter of adding -I${STAGING_DIR_HOST}/usr/include/mtd Nov 27 22:22:35 or? Nov 27 22:23:17 seems bettaer than adding own copy of the same headers :/ Nov 27 22:23:42 still don't understand why mtd-utils are deploying these same headers, though... Nov 27 22:24:25 these are sanitized kernel headers, which imho should be staged in kernel.bbclass and not by mtd-utils Nov 27 22:24:30 just my 2cents Nov 27 22:26:16 probably I'm missing something... Nov 27 22:49:14 Crofton|work: done Nov 27 22:57:49 thanks florian Nov 27 22:58:16 woglinde: yw Nov 27 23:30:32 hurray Nov 27 23:30:41 "localmodconfig" Nov 27 23:30:58 thats what I wanted latley for the kernel Nov 27 23:57:33 *sigh* building beagle demo image shows so many bad packages Nov 28 00:29:50 Yeah, like sane-backends. :( Nov 28 00:32:02 mwester there are much more Nov 28 00:32:11 even packages which get compiled Nov 28 00:32:16 and dont get packaged Nov 28 00:32:37 because of FILES_ arent correct Nov 28 00:39:39 * mwester hasn't encountered any of those! Nov 28 01:00:54 The dbus recipe seems to be broken now. Is anyone working on fixing it? Nov 28 01:02:04 whats broken? Nov 28 01:06:22 I do not know. http://pastebin.com/pastebin.php?dl=d6aaed2d8 Nov 28 01:10:54 Right now, I just know the image build seems to fail with dbus. Nov 28 01:12:58 woglinde: http://pastebin.com/pastebin.php?dl=d6aaed2d8 Nov 28 01:13:19 That's the entire build log, but the last few lines are probably most interesting. Nov 28 01:14:10 Everything seems to depend on dbus, and therefore everything is breaking. Nov 28 01:14:23 I can't even build base-image now. Nov 28 01:16:17 re Nov 28 01:21:23 wb Nov 28 01:25:56 * RobotGuy sighs Nov 28 02:12:44 03Mario Domenech Goulart  07org.openembedded.dev * rf200c0f560 10openembedded.git/recipes/chicken/chicken-4.2.0/translator-cross.patch: Nov 28 02:12:44 chicken, chicken-cross: added translator-cross.patch for chicken 4.2.0 Nov 28 02:12:44 Fixes path to the translator (chicken) in the compiler driver (csc) Nov 28 02:12:44 when cross compiling. Nov 28 02:12:46 Signed-off-by: Mario Domenech Goulart Nov 28 02:12:48 03Mario Domenech Goulart  07org.openembedded.dev * r0ad83bc173 10openembedded.git/recipes/chicken/ (3 files): Nov 28 02:12:51 chicken: Added recipes for chicken 4.2.10 Nov 28 02:12:53 Signed-off-by: Mario Domenech Goulart Nov 28 02:12:55 03Mario Domenech Goulart  07org.openembedded.dev * rf0ae52610d 10openembedded.git/recipes/chicken/ (chicken-cross_4.2.0.bb chicken.inc chicken_4.2.0.bb): Nov 28 02:12:58 chicken: path and dependencies fixes for chicken recipes Nov 28 02:13:00 Removed gcc-cross-sdk as dependency. Nov 28 02:13:02 Signed-off-by: Mario Domenech Goulart **** ENDING LOGGING AT Sat Nov 28 02:59:56 2009