**** BEGIN LOGGING AT Thu Apr 15 02:59:56 2010 Apr 15 04:37:53 zecke, since you seem to appreciate security update patches: http://patchwork.openembedded.org/patch/1913/ Apr 15 04:37:56 fyi Apr 15 04:39:03 * zenlinuxPDX heads to bed early, trying to transform into a morning person :-/ Apr 15 05:11:52 03Martin Jansa  07org.openembedded.dev * rf2aac8bf86 10openembedded.git/recipes/openmoko-3rdparty/ventura_svn.bb: Apr 15 05:11:52 ventura: bump SRCREV for usability fixes Apr 15 05:11:52 Signed-off-by: Martin Jansa Apr 15 05:54:34 mckoan|away: please check and ack my stable update Apr 15 06:22:23 goodmorning Apr 15 06:28:54 Hi All Apr 15 06:29:00 Goodmorning Apr 15 07:34:44 good morning Apr 15 07:37:52 morning Apr 15 08:54:55 03Koen Kooi  07org.openembedded.dev * r67cae2cf06 10openembedded.git/recipes/xorg-lib/ (4 files in 2 dirs): pixman git + 0.18.0: add patches to fix TLS behaviour Apr 15 08:55:02 03Koen Kooi  07org.openembedded.dev * rafe441d2c6 10openembedded.git/classes/ (gnome.bbclass gnomebase.bbclass recipe_sanity.bbclass): gnome: split out gnomebase.bbclass that doesn't have gconf and gmime depends Apr 15 08:55:02 03Koen Kooi  07org.openembedded.dev * re01d04f18d 10openembedded.git/recipes/gnome/ (3 files): gnome-common, gnome-doc-utils: use gnomebase.bbclass Apr 15 09:00:11 03Scott Garman  07org.openembedded.dev * rf994e036d7 10openembedded.git/recipes/libnids/ (libnids-1.24/configure.patch libnids_1.24.bb): Apr 15 09:00:11 libnids: Upgrade to 1.24 Apr 15 09:00:11 * Addresses CVE-2010-0751 Apr 15 09:00:11 * Use new-style checksums Apr 15 09:00:11 Signed-off-by: Scott Garman Apr 15 09:00:12 Signed-off-by: Holger Hans Peter Freyther Apr 15 09:16:39 hi, is there a newbie channel or can i ask newbie questions here Apr 15 09:19:29 hello, anyone? Apr 15 09:20:45 arimeh: just ask Apr 15 09:21:26 basically i'm looking for a good startup documentation Apr 15 09:22:11 the ones on free-electrons, kernel-labs and uv-ac.de don't work Apr 15 09:22:24 i've checked out org.openembedded.stable Apr 15 09:22:28 bitbake cannot parse it Apr 15 09:22:43 so in essence i can't get started and im no bitbake expert Apr 15 09:22:59 i just need to get it going by following some instructions Apr 15 09:23:40 nothing works as expected and all the documentation is out-of-date (no surprise :)) Apr 15 09:23:57 03Koen Kooi  07org.openembedded.dev * r8e8609c78d 10openembedded.git/recipes/gnome/gail_1.20.0.bb: gail: convert to new style staging Apr 15 09:24:08 03Koen Kooi  07org.openembedded.dev * r7f2ff32784 10openembedded.git/recipes/gnome/ (52 files): gnome: convert a lot of recipes to new style staging Apr 15 09:25:24 arimeh: which platform? Apr 15 09:25:46 angstrom for instance (from free-electrons doc) Apr 15 09:26:03 basically i need someone tell me the following, i get the basic idea Apr 15 09:26:16 1. download this specific version of bitbake Apr 15 09:26:27 2. checkout this branch with git Apr 15 09:26:38 3. select this as machine and that as distro Apr 15 09:26:43 bitbake this image Apr 15 09:26:48 and it builds! Apr 15 09:27:01 is there such configuration please?? Apr 15 09:27:19 arimeh: I mean which CPU Apr 15 09:27:35 ARM Apr 15 09:27:37 arimeh: offical doc is this: http://wiki.openembedded.net/index.php/Getting_started Apr 15 09:27:57 arimeh: my favourite (of coures) is this: http://www.kaeilos.com/download Apr 15 09:28:08 i know this page Apr 15 09:28:50 arimeh: bitbake must be bitbake-1.8.18 Apr 15 09:30:24 so i try all the instruction with my ubuntu 9.10 and you guarantee that it works? :) Apr 15 09:30:32 i mean kaeilos Apr 15 09:30:45 arimeh: yes I do Apr 15 09:31:04 ok let me try it out, thanks mckoan Apr 15 09:31:27 arimeh: you're welcome :-) Apr 15 09:52:05 03Koen Kooi  07org.openembedded.dev * r636141f6fa 10openembedded.git/classes/kernel.bbclass: kernel bbclass: read version from ${B}, remove obsolete hostap logic that breaks the build (from poky) Apr 15 09:52:05 03Koen Kooi  07org.openembedded.dev * r8f025e984f 10openembedded.git/recipes/gnome/gnome-dvb-daemon_0.1.17.bb: gnome-dvb-daemon: add 0.1.17 Apr 15 09:52:06 03Koen Kooi  07org.openembedded.dev * raa9b5a7221 10openembedded.git/classes/kernel.bbclass: kernel bbclass: don't depend on c library (from poky) Apr 15 10:54:08 hrw: hi, did you ever blog about your ELC-E talk on license compliance? Apr 15 10:54:41 hi Apr 15 10:54:53 zecke: I did not had such talk Apr 15 10:55:12 zecke: you probably want Harald Walte keynote talk from SCW'09 Apr 15 11:00:50 03Henning Heinold  07org.openembedded.dev * re8360af5f8 10openembedded.git/recipes/phoneme/ (2 files in 2 dirs): phoneme-advanced-foundation: add build 160 Apr 15 11:02:49 03Koen Kooi  07org.openembedded.dev * r8cc2abb679 10openembedded.git/recipes/python/python-pygobject_2.20.0.bb: python-pygobject 2.20: convert to new style staging Apr 15 11:45:11 hi, i know its ot but is there any gpl/lgpl support channel arround? Apr 15 11:50:02 hrw: well, but you talked about the distribute source class at ELC-E? Apr 15 11:50:14 nik0n: did you try the software freedom law center? Apr 15 11:50:34 zecke: http://marcin.juszkiewicz.com.pl/download/presentations/2009-SCW-Poky_Linux_and_OpenEmbedded_based_environment.pdf Apr 15 11:54:28 03Koen Kooi  07org.openembedded.dev * rb9a6bb07f5 10openembedded.git/recipes/gnome/gnome-dvb-daemon_0.1.17.bb: gnome-db-daemon: fix src_uri name Apr 15 12:04:55 zecke: present? Apr 15 12:05:11 http://hrw.pastebin.com/2XAQ29KV Apr 15 12:05:46 zecke: bitbake master 1c6b31c649474b4c2b63ef9d9311e61de20bc8c2 Apr 15 12:06:19 checking with 1.10 Apr 15 12:13:07 03Koen Kooi  07org.openembedded.dev * r7e7f9fc958 10openembedded.git/recipes/xorg-lib/ (6 files): xorg-lib: switch to new style staging Apr 15 12:15:43 03Koen Kooi  07org.openembedded.dev * r7d73b5d57a 10openembedded.git/recipes/xorg-app/ (4 files): xorg-app: convert to new style staging Apr 15 12:39:43 Hi, I sent cm-x300 patches to the mailinglist yesterday, I there any extra action required from me, or do I just have to wait until someone pushes them upstream? Apr 15 12:40:05 you do not have r/w? Apr 15 12:40:39 I guess only read? Apr 15 12:40:54 ok Apr 15 12:41:34 will merge Apr 15 12:42:46 03Siebren Hakkesteegt  07org.openembedded.dev * re60f963759 10openembedded.git/conf/machine/cm-x300.conf: Apr 15 12:42:46 cm-x300: add machine configuration Apr 15 12:42:46 Signed-off-by: Marcin Juszkiewicz Apr 15 12:42:47 03Siebren Hakkesteegt  07org.openembedded.dev * rd1b29786ad 10openembedded.git/recipes/linux/ (3 files in 2 dirs): Apr 15 12:42:47 linux 2.6.31: add cm-x300 support Apr 15 12:42:48 Signed-off-by: Marcin Juszkiewicz Apr 15 12:42:59 o.k. thanks :) Apr 15 12:46:03 np Apr 15 12:46:12 something more to merge from patchwork? Apr 15 12:50:22 never ever toshiba laptops Apr 15 12:51:52 ;) Apr 15 12:52:28 thats crap Apr 15 12:52:44 if I login remotely via wifi to it connection hangs Apr 15 12:52:55 ethernet is totally borked Apr 15 13:39:48 hi. I want to test my wxwidgets application on a cheap simpad... which one do you suggest which is supported by linux ? Apr 15 13:54:34 hi, how can I distribute _all_ the kernel modules from virtual/kernel in an image? Right now only a small subset of them is copied in lib/modules/, while all them are in the modules-...-machine.tar.gz archive Apr 15 13:54:57 ao2: add kernel-modules Apr 15 13:55:46 JaMa|Away, where exactly? to machine conf? Apr 15 13:56:10 let me grep for kernel-modules :) Apr 15 13:56:11 image where you want them or some task used there seems better to me.. Apr 15 13:56:20 03Koen Kooi  07org.openembedded.dev * ra25f07ca61 10openembedded.git/recipes/libxml/libxml2.inc: Apr 15 13:56:20 libxml2: fix regression where -native wouldn't build python stuff needed for xml2po anymore Apr 15 13:56:21 * also fix 1 bug in do_install where it pokes stuff outside of ${D} Apr 15 13:56:21 * binconfig hack is still there, though Apr 15 13:56:36 but I had exactly oposite problem.. Apr 15 13:57:24 upgrading all those modules built by defconfig ie because kernel PR bump is long Apr 15 14:00:36 I can see some machines setting MACHINE_EXTRA_RDEPENDS += " kernel-modules ", but doing that in an image or a task is cleaner indeed Apr 15 14:26:21 fwiw, I can add arbitrary service methods for you in fso-glib, of you need 'em Apr 15 14:26:32 (or you can add them yourself ;) Apr 15 14:26:48 err, wrong channel Apr 15 14:30:05 RP: ping Apr 15 14:32:02 mickeyl: he's in SC at lfcollab, just so you're not waiting forever for pingback :-) Apr 15 14:34:20 hehe, k Apr 15 14:34:25 nothing important anyways Apr 15 14:34:28 thanx Apr 15 14:43:35 hrw: that error is because master requires 2.6 at the moment. 1.10 should be fine with 2.5. Apr 15 14:43:53 mickeyl, zecke, hrw: thoughts on http://github.com/kergoth/OpenEmbedded/commits/oelib ? Apr 15 14:44:37 kergoth: would be nice it if would check for 2.6 first Apr 15 14:44:44 hrw: i thought it did.. Apr 15 14:45:37 hrw: http://git.openembedded.org/cgit.cgi/bitbake/commit/?id=85c000cc7db47cfd489cf282c58db2dab0d87908 .. maybe I got the hex version value wrong, just did a replacement on it Apr 15 14:46:46 20605f0 = my 2.6.5 (not default) Apr 15 14:47:34 hmm. in pythoncli your check works Apr 15 14:47:37 so 20600f0 should be fine for 2.6.0, so it should've errored with 2.5? Apr 15 14:47:39 weird Apr 15 14:47:41 hmm Apr 15 14:47:49 wait, i think i see Apr 15 14:47:54 i bet we need to check it before it imports the modules Apr 15 14:49:35 yep, thats it Apr 15 14:50:08 hrw: pull, should check properly now Apr 15 14:50:16 hrw: sorry about the hassle. Apr 15 14:50:23 no problem, moved to 1.10 Apr 15 14:50:42 somone wants to upgrade libpng? Apr 15 14:50:58 * kergoth 's been trying to clean up the codebase in master some, figured taking advantage of current features and prepping for 3.0 eventually wasn't a bad idea Apr 15 14:51:57 another question, sorry, I have 'touchscreen' in MACHINE_FEATURES but it looks like task-base-touchscreen does not get added, I mean in the image I don't get tslib-tests and tslib-calibrate as I would expect by reading task-base.bb Apr 15 14:52:48 ao2: generally adding to MACHINE_FEATURES means the task for that feature gets created, but it has to be in both MACHINE and DISTRO to be automatically included via the main task-base package, and therefore in the images by default. Apr 15 14:52:56 ao2: haven't looked at that one specifically, but that's the general behavior Apr 15 14:53:24 so either add to distro features, or IMAGE_INSTALL_append = " task-base-touchscreen" or something Apr 15 14:54:41 kergoth, ok, thanks for the hint. I thought those "machine-tasks" were added automatically via task-base Apr 15 14:55:27 machine defines what the hardware is capable of, distro defines the things the target distro / software intends to support Apr 15 14:55:28 np Apr 15 15:00:57 03Koen Kooi  07org.openembedded.dev * rb7b9e9109a 10openembedded.git/recipes/gpsd/gpsd.inc: gpsd: fix dbg packaging Apr 15 15:06:09 ao2: take a brief look at task-base.bb to see how it's matched Apr 15 15:06:46 any problem with xorg-headers-native-1.0 out there? Apr 15 15:16:09 mickeyl, I see it now, the logic defines the task-base-* tasks (their dependencies), without actually adding them. Right now I am adding task-base-touchscreen to fso-console-image just as a quick test. Apr 15 15:16:45 I see also that some images add those task-base-* deriving from MACHINE_FEATURES again Apr 15 15:17:08 hrw: there, if sys.version_info < (2, 6, 0): is less ugly. tuples compare item by item by default, iirc, so all is good :) Apr 15 15:17:29 anybody know how does the new xorg-lib switch to new style staging works? Apr 15 15:18:09 looks like xorg build is broken sice then Apr 15 15:19:17 :( Apr 15 15:21:10 * mckoan is unable to find any RFC about new staging method, do you have a link please? Apr 15 15:23:08 new staging generically was certainly discussed on the mailing list. I don't know about xorg specifically. Apr 15 15:23:10 what goes wrong with it? Apr 15 15:24:51 pb__: http://pastebin.com/BDkDy6D1 Apr 15 15:25:44 14:13 < CIA-2> Koen Kooi org.openembedded.dev * r7e7f9fc958 openembedded.git/recipes/xorg-lib/ (6 files): xorg-lib: switch to new style staging Apr 15 15:26:23 "No rule to make target `install'" is a pretty bad failure. It sounds like the package didn't unpack and/or configure correctly. Apr 15 15:28:05 pb__: this is the patch pushed a few hours ago http://pastebin.com/qs4BaEgX Apr 15 15:29:21 oh, I see. that does look pretty broken. I guess you should take it up with koen. Apr 15 15:30:36 pb__: :-( Apr 15 15:32:19 that patch you pasted was backwards, right? Apr 15 15:32:40 if what you pasted was the actual patch then it is even weirder. Apr 15 15:33:01 but, if it's backwards, I think you are just suffering from a bad interaction between native.bbclass and the staging expectations of that package. Apr 15 15:33:29 I just pull-ed OE and I got it Apr 15 15:34:58 grr Apr 15 15:35:17 kergoth: mm? Apr 15 15:35:21 kergoth: ? Apr 15 15:35:50 just dealing with python version bits. master uses multiple things from 2.6 now, which is fine, but need to add another version check, i think Apr 15 15:36:16 running bitbake as python2.5 errors out with a message about the version, but running setup.py as python2.5 just explodes, since it imports the bb module directly to get __version__ Apr 15 15:36:17 heh Apr 15 15:36:24 oops. Apr 15 15:37:00 kergoth: so the problem is'n caused by xorg-headers-native patch? Apr 15 15:37:15 ? no, i'm talking about something entirely unrelated to your issue Apr 15 15:37:37 kergoth: uh, sorry. so I continue writing a mail to ML Apr 15 15:37:42 heh Apr 15 15:44:14 03Koen Kooi  07org.openembedded.dev * rc9bbe810c5 10openembedded.git/recipes/ti/ti-xdctools.inc: ti-xdctools: add PRIVATE_LIBS list for internal libs Apr 15 15:44:15 03Koen Kooi  07org.openembedded.dev * rbd1de467fc 10openembedded.git/recipes/gnome/at-spi_1.30.0.bb: at-spi: add 1.30.0 Apr 15 15:44:17 03Koen Kooi  07org.openembedded.dev * r49c62b9ca0 10openembedded.git/recipes/sylpheed/claws-mail_3.6.1.bb: claws-mail: bump PR Apr 15 15:44:18 03Koen Kooi  07org.openembedded.dev * rb5ec44fec6 10openembedded.git/recipes/gnome/ (gedit-plugins_2.30.0.bb gedit_2.30.0.bb): gedit(-plugins): add 2.30.0 Apr 15 15:44:19 03Koen Kooi  07org.openembedded.dev * rfaa565a164 10openembedded.git/recipes/abiword/abiword_2.8.3.bb: abiword 2.8.3: bump PR Apr 15 15:44:27 03Koen Kooi  07org.openembedded.dev * r1fec6b9ba5 10openembedded.git/recipes/gtksourceview/ (gtksourceview2_2.10.0.bb gtksourceview2_2.9.1.bb): gtksourceview2: add 2.10.0, convert 2.9.1 to new style staging Apr 15 15:57:16 pb__: http://github.com/kergoth/OpenEmbedded/commit/1150aca45c1e30c4307fb8dd6d16dddfc68557b8, http://github.com/kergoth/OpenEmbedded/commit/4034da072000d6347bec0c236622cb3c145ef323 - thoughts? Apr 15 16:03:08 03Koen Kooi  07org.openembedded.dev * ra470efeb85 10openembedded.git/recipes/xorg-lib/xorg-headers-native.bb: xorg-headers-native: flag with NATIVE_INSTALL_WORKS = "1" Apr 15 16:04:10 * kergoth grumbles Apr 15 16:05:46 that's my job! Apr 15 16:06:13 :) Apr 15 16:08:26 * mckoan hopes you could appreciate it ;-) http://kom.kaeilos.com/documentation/use-kom-wizard Apr 15 16:12:59 mckoan: nice Apr 15 16:13:11 kergoth: thx Apr 15 16:14:29 have a nice rest of the day Apr 15 16:27:26 hmm, thinking we should change the task events in bitbake. currently they format the mesage themselves and the uis just display it.. but that's silly. the task should carry the information in the object, let the UI decide how to display it Apr 15 16:33:56 * mwester-laptop agrees Apr 15 16:35:11 The alternative is to post-process the output logs with regexs -- which I have attempted to do. Apr 15 16:36:51 kergoth: *nod* Apr 15 16:36:56 who did the touchscreen0 udev rule? Apr 15 18:06:47 So, anyone have any thoughts on moving some of our python utility functions into an 'oe' python package, rather than all over the metadata? Apr 15 18:14:09 03Roman I Khimov  07org.openembedded.dev * re460fae9dd 10openembedded.git/recipes/clamav/ (7 files in 2 dirs): Apr 15 18:14:09 clamav: remove versions <0.95 Apr 15 18:14:09 See removal.txt and Apr 15 18:14:09 c1aab90e5e removal.txt: proposal to remove ClamAV versions < 0.95 in April Apr 15 18:14:09 for reasoning. Apr 15 18:14:10 Signed-off-by: Roman I Khimov Apr 15 18:25:29 hello Apr 15 18:27:07 I have a uBoot question - currently I am interupting boot of an Overo Gumstix, and using "mm" to enter a hex value into a specific location - how do I modify things & build with OE so that gets done automagicly by uBoot? Apr 15 18:34:06 robtow: have you looked into creating your own u-boot script? Apr 15 18:34:57 robtow: http://www.gumstix.net/wiki/index.php?title=U-Boot#boot.scr Apr 15 18:35:36 robtow: or you could create modify/patch the u-boot sources where the environment is set for your board Apr 15 18:35:43 oneshel - that is exactly what I need to do. I have run successfully; but don't know what to modify in the recipe tree. Apr 15 18:37:13 oneshel - thanks for the llink - looks like I could make a boot.scr with a command Apr 15 18:39:20 hmmm. Apr 15 18:41:54 robtow: I'm trying to find where the environment gets set in the u-boot sources... Apr 15 18:42:21 oneshel... thanks. I'm poking about, too. Apr 15 18:42:24 :-) Apr 15 18:43:44 a-ha include/configs/omap3_overo.h Apr 15 18:43:58 #define CONFIG_EXTRA_ENV_SETTINGS Apr 15 18:44:14 and #define CONFIG_BOOTCOMMAND Apr 15 18:44:30 oneshel - I'll look at that! Apr 15 18:47:26 Hmmmm. Apr 15 18:47:38 I'm not finding omap3_overo.h Apr 15 18:49:28 robtow: stupid question, are you using the overo or the vertex gumstix board? Apr 15 18:49:36 *verdex Apr 15 18:50:06 Overo. Apr 15 18:50:21 Not stupid q at alll Apr 15 18:51:30 I diid a git pull yyesterday; and today diid both bitbake x-load and bitbake u-boot-omap3 with machine set to overo Apr 15 18:52:36 robtow: have you modified a bitbake recipe before? Apr 15 18:53:22 oneshel - some easy stuff, yes. Apr 15 18:53:57 I just diid a recursive search through "recipes" for omap3_overo.h, didn't see it. Apr 15 18:54:23 robtow: ah, that file won't be under the recipes area Apr 15 18:54:55 it's in the u-boot sources, which is outside the openembedded files Apr 15 18:54:58 That would explain it! LOL Apr 15 18:55:16 http://git.denx.de/?p=u-boot.git;a=summary Apr 15 18:55:18 Apr 15 18:55:24 that's the main u-boot git repository Apr 15 18:56:09 Would it be easierr to write a boot.scr, and put a command in that, as I need simply to write one value into one memory location? Apr 15 18:56:21 I'm sure there is a decent "getting started" guide out there somewhere, but basically what you want is to create a patch and add it to the .bb file for your board's version of u-boot Apr 15 18:56:39 maybe, will your board always use a microSD card? Apr 15 18:57:51 oneshel - yes; I will have a 16G or 32G micro SD card attached. Apr 15 18:58:10 ah great, yeah that's probably your best bet then Apr 15 18:58:48 Ah, ok. Apr 15 19:00:15 Thank-you for the helpful advice! Apr 15 19:00:43 robtow: no problem :) Apr 15 20:11:50 hey guys... noob q from me Apr 15 20:12:05 i checked out stable/2009 and i want to build an image for dns323 (NAS device) Apr 15 20:12:15 usually i only work with opendreambox Apr 15 20:12:29 how do i know which distro to use? Apr 15 20:12:37 for the nas thingie Apr 15 20:50:48 angstrom ? Apr 15 21:03:44 fraxinas, angstrom is probably a safe bet unless there is good documentation on using another distro for that device Apr 15 21:04:22 i haven't really seen any Apr 15 21:04:28 however, there already is a machine conf file Apr 15 21:04:43 i am just not sure right now how i even start ;) Apr 15 21:04:56 seems qemu is having some kind of trouble with my amd64 host Apr 15 21:05:14 /proc/sys/vm/mmap_min_addr is not 0. This will cause problems with qemu so please fix the value (as root). Apr 15 21:05:16 what host linux distro are you running? Apr 15 21:05:25 yeah, just do that it says Apr 15 21:05:35 Linux fox 2.6.32-ARCH #1 SMP PREEMPT Mon Mar 15 20:44:03 CET 2010 x86_64 AMD Phenom(tm) 9550 Quad-Core Processor AuthenticAMD GNU/Linux Apr 15 21:05:45 is that really a good idea? okay i'll try Apr 15 21:05:55 it's necessary for emulators like quemu Apr 15 21:06:02 qemu, even Apr 15 21:06:20 did Apr 15 21:06:34 i hope it doesn't break too many other things ^^ Apr 15 21:06:50 shouldn't Apr 15 21:06:51 opendreambox builds without that :) Apr 15 21:07:04 technically it might open you up to some kernel NULL pointer attacks more easily Apr 15 21:07:06 thx for your assistance btw Apr 15 21:07:09 np Apr 15 21:07:13 fraxinas, are you using current openembedded.dev or something else? Apr 15 21:07:24 qemu doesn't need that change and we don't check for it for a little while now Apr 15 21:07:27 opendreambox-1.6 Apr 15 21:07:33 Ah Apr 15 21:07:42 Now that's also important information to know, to help you out Apr 15 21:08:06 oh i checked out stable/20009 Apr 15 21:08:09 specifically for this Apr 15 21:08:29 uh-oh Apr 15 21:08:40 whoops, sorry for the misinfo Apr 15 21:09:00 I don't think I've ever used qemu without it requiring that change Apr 15 21:09:10 lots of ERROR: Could not inherit file classes/rootfs_${IMAGE_PKGTYPE}.bbclass while parsing Apr 15 21:09:25 huh. moving bits into a python package appears to have caused a 3-6% decrease in overall build time from a partial task-boot. did 4 builds before and 4 builds after, and it consistently dropped the time between 2 and 4 minutes out of the total of 55-59. Apr 15 21:09:32 is stable/2009 a good choice to go with at all? Apr 15 21:26:27 trying with org.openembedded.dev Apr 15 21:33:37 ~seen ant Apr 15 21:33:41 ant was last seen on IRC in channel #oe, 408d 21h 26m 6s ago, saying: 'nite all'. Apr 15 21:33:52 ~seen ant_ Apr 15 21:33:53 ant_ was last seen on IRC in channel #oe, 445d 22h 23m 9s ago, saying: 'yes but then it sneaks out on framebuffer :-/'. Apr 15 21:34:37 so, guys, i started out with an entirely new clone of both oe and bb Apr 15 21:34:51 and when first running bitbake i get a load of errors Apr 15 21:35:02 you're going to have to be more specific. Apr 15 21:35:09 "load of errors" is not particularly useful Apr 15 21:35:49 http://developer.elitedvb.net/pastebot/paste Apr 15 21:35:58 yeah yeah i was about to paste it :) Apr 15 21:36:04 pushy pushy Apr 15 21:37:15 okay i see something that might be a possible problem Apr 15 21:37:26 it uses my dreambox 1.8 bitbake instead of the new one Apr 15 21:37:28 gotta fix paths Apr 15 21:38:34 that url is where i'd go if i wanted to paste something, not the link to yours Apr 15 21:38:40 unless i'm missing something. Apr 15 21:41:29 yeah i was indeed Apr 15 21:41:39 it's late ;) http://developer.elitedvb.net/pastebot/196 was it Apr 15 21:41:48 but it was the path issue i guess Apr 15 21:42:03 k Apr 15 21:42:11 oh no actually still the same :( Apr 15 21:42:22 IMAGE_PKGTYPE is set by your distro Apr 15 21:42:27 so its likely your DISTRO is set incorrectly Apr 15 21:43:06 humm! might be Apr 15 21:43:19 DISTRO = "angstrom" Apr 15 21:43:19 look in openembedded/conf/distro/ Apr 15 21:43:27 make sure you're selecting one that has a config file there Apr 15 21:43:36 case sensitive? Apr 15 21:43:42 you probably don't want "angstrom", but instead a specific angstrom version as you can see in teh conf directory Apr 15 21:43:53 yes, everything in the metadata is case sensitive Apr 15 21:43:58 ah it's angstrom-2008.1.conf Apr 15 21:44:10 right, there you go, so you want DISTRO = "angstrom-2008.1" Apr 15 21:45:37 okay cool Apr 15 21:45:55 now, i'm doing bitbake linux-orion to begin with Apr 15 21:46:06 how do i know the name of the actual image package to build? Apr 15 21:46:14 that depends on what image you want. Apr 15 21:46:17 look in recipes/images/ Apr 15 21:46:17 in our distro it's bitbake dreambox-image Apr 15 21:46:40 any image can be built with any distro, just some include sets of packages more appropriate for a given device, some are more generic Apr 15 21:46:58 image just controls what packages go into the image, and some of that is still controlled by the machine and distro config, not the image itself Apr 15 21:47:19 okay so i go for something that sounds about right Apr 15 21:47:23 kind of like nas-server-image.bb :) Apr 15 21:47:48 I'm assuming you're using an existing MACHINE, so it'll know to build the right kernel and all Apr 15 21:48:20 yeah it exists actually Apr 15 21:48:37 good, should be fine then :) Apr 15 21:49:07 yay it's being a good toolchain and bakes me a kernel atm Apr 15 21:51:27 ant!!!!!!!!!!! Apr 15 21:51:33 kergoth is the debug output supposed to look so ugly? http://developer.elitedvb.net/pastebot/197 Apr 15 21:51:33 hello Apr 15 21:51:48 ant I got now kernel with kexecbootmenu loading Apr 15 21:52:02 but I dont know how to specify bootcponfig Apr 15 21:52:02 with the download progresses? seems to be missing carriage returns Apr 15 21:52:28 woglinde_: in /boot/boot.cfg in the rootfs Apr 15 21:53:20 hm on the device right? Apr 15 21:53:24 kexecboot-cfg_0.1.bb Apr 15 21:53:38 I cann use oe ipk Apr 15 21:54:24 ant so kexecboot is readying /boot/boot.cfg on what ever partition it will find it Apr 15 21:54:34 yes Apr 15 21:54:56 hm why I cannt configure somethin inside kexecboot initramfs itself Apr 15 21:55:03 this suckz a little biit Apr 15 21:55:14 where would ypu save it? in nad? Apr 15 21:55:20 *nand Apr 15 21:55:26 inside the initramfs Apr 15 21:55:48 I dont need to save something Apr 15 21:56:03 in that case you could just hardcode it Apr 15 21:56:04 hm and I cannt see which partions it finds at all Apr 15 21:56:24 and I saw no vfat support Apr 15 21:56:27 :( Apr 15 21:56:35 fraxinas: looks okay to me. clearly wget is rather more verbose than we need, so that's something we could fix with our fetchers, and when you use BB_NUMBER_THREADS > 1, the output gets interleaved. hopefully we'll get some better UIs in bitbake master, since it has a proper backend/frontend split Apr 15 21:57:10 woglinde_: vfat+jffs2 patch is in fstype.c in klibc -> kexecboot sources Apr 15 21:57:29 ant do I have more output if I compile kexecboot with DEBUG= Apr 15 21:57:31 ? Apr 15 21:57:39 yes, do you have serial? Apr 15 21:58:03 unfornatly not Apr 15 21:58:45 I only could setup netconsole Apr 15 21:58:48 via usbnet Apr 15 21:59:02 ant where are the vfat patches? Apr 15 21:59:41 in klibc Apr 15 21:59:52 why are they in klibc? Apr 15 21:59:54 and in kexecboot Apr 15 22:00:21 initially thesing committed klibc for kexecboot project ;) Apr 15 22:00:38 yes that I learned Apr 15 22:00:45 I only see it for klibc Apr 15 22:00:50 not kexecboot Apr 15 22:01:07 iirc fstype.c is already patched in kexecboot Apr 15 22:01:13 mom Apr 15 22:02:02 http://git.linuxtogo.org/?p=groups/kexecboot/kexecboot.git;a=blob;f=fstype/fstype.c Apr 15 22:02:32 but hey..I guess you're the first one testing vfat :D Apr 15 22:02:37 not ah Apr 15 22:02:43 hm Apr 15 22:03:11 yes, uses romfs code Apr 15 22:03:12 hm yes the mmc is on vfat on the n8x0 Apr 15 22:03:38 okay now how do I hardcompile the config Apr 15 22:03:40 woglinde_: patch was kept minimal and rejected upstream :] Apr 15 22:03:56 in kexecboot or klibc? Apr 15 22:04:00 klibc Apr 15 22:04:14 okay Apr 15 22:04:18 I made them aware we exist Apr 15 22:04:22 ;) Apr 15 22:04:37 okay kergoth :) i am glad that it actually all works so easily Apr 15 22:04:52 hm disable cursor patch would be nice on kexecboot kernels Apr 15 22:05:12 yes, we have somethin gsimilar Apr 15 22:06:47 woglinde_: we still have a black cursor: just one char but is there ... Apr 15 22:07:46 on n810 its blinking Apr 15 22:09:13 but disappears in kexecboot-menu Apr 15 22:09:19 nope Apr 15 22:09:20 at least on Zaurus Apr 15 22:09:24 argh Apr 15 22:09:34 is just over boot logo Apr 15 22:10:10 in the top-left corner, during kernel loading Apr 15 22:10:23 then is killed Apr 15 22:10:48 your patch would probably eradicate the issue Apr 15 22:11:53 hm I dont have one Apr 15 22:12:01 only read about it Apr 15 22:12:04 ok, we postponed too... Apr 15 22:12:58 see in util.c Apr 15 22:13:13 . * Deactivate/Activate terminal input */ Apr 15 22:13:44 but here is userland... Apr 15 22:20:04 woglinde_: how small in the vfat partition? Apr 15 22:20:10 DPRINTF("Skipping little partition with size (%dk) < 200k\n", blocks); Apr 15 22:21:20 ant 2 gig Apr 15 22:21:29 /dev/mmcblk0p1 1.8G 780.1M 1.1G 42% /media/mmc2 Apr 15 22:21:31 ^_^ Apr 15 22:24:24 woglinde_: the point is each kernel could have a different cmdline requirements so boot.cfg must reside in /boot. Remember the ordering issue too (sda->sdb->sdc): the first mounted is sda Apr 15 22:24:31 ant I am now building an more featurefull initramfs Apr 15 22:24:37 to see what the kernel get Apr 15 22:25:06 we were just needing a full-fledged bootloader Apr 15 22:25:17 replacement Apr 15 22:26:04 If I ever get meta-toolchain to build, what will I end up with? I have the same desire as philip and marc at: http://groups.google.com/group/beagleboard/browse_thread/thread/b11d9c53d6f09536/33c537f7d08f1952?#33c537f7d08f1952 Apr 15 22:29:05 ah..this is new gtk+-2.20.0/modules/input/.libs/im-xim.so: undefined symbol: g_realloc_n Apr 15 22:31:24 gtk+-native-2.20.0-r8.0/gtk+-2.20.0/modules/input/im-xim.la does not export GTK+ IM module API Apr 15 22:31:52 /dev/mmcblk0p1 1.8G 780.1M 1.1G 42% /media/mmc2  Apr 15 22:31:54 args Apr 15 22:31:58 cg1 use native-sdk Apr 15 22:34:12 uh..what's that? CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include/python2.6 Apr 15 22:34:22 zapping all... Apr 15 22:34:50 which modul? Apr 15 22:35:11 now is libxml2-2.7.7-r7.0 Apr 15 22:35:30 libxml2-2.7.7-r7.0 Apr 15 22:35:35 hm Apr 15 22:35:36 bah Apr 15 22:35:37 | cc1: internal compiler error: in add_path, at incpath.c:429 Apr 15 22:35:45 lookinto configure.log Apr 15 22:35:50 args config.log Apr 15 22:35:54 woglinde: I forced gcc 4.4.2 Apr 15 22:36:00 like armv4 uses Apr 15 22:36:10 and binutiuls 2.20 Apr 15 22:36:24 zaurus? Apr 15 22:36:29 but probably need something more. Angstrom does not include sane-toolchain.. Apr 15 22:36:29 woglinde_: thanks. That sounds more correct. But, for curiousity, what is meta-toolchain? Apr 15 22:37:23 cg1 was an Idea of sdk Apr 15 22:52:23 woglinde_: I'm building now minimal. I only need linux-kexecboot for armv4 and armv5te Apr 15 22:52:52 iirc now the kernel does not pull *libc anymore Apr 15 22:52:58 let see Apr 15 22:54:15 woglinde_: btw 2.6.34-rc4 is out Apr 15 22:56:41 I don't suppose anyone here is well-versed with the ruby recipe/build? Apr 15 22:57:58 Trying to build a package with the socket library included Apr 15 23:00:36 zenlinuxPDX: I see a short log: http://cgit.openembedded.org/cgit.cgi/openembedded/log/?qt=grep&q=ruby Apr 15 23:02:41 Here's the mkmf log from the socket module's directory: http://pastebin.com/H0drstUu Apr 15 23:02:47 looks like I'm missing a dev library **** ENDING LOGGING AT Fri Apr 16 03:00:02 2010