**** BEGIN LOGGING AT Thu Apr 21 02:59:58 2011 Apr 21 03:00:09 the 4th argument to do_split_packages seems odd Apr 21 03:03:15 i think the do_split packages make libopencv-dev and libopencv targets, but the FILES_${PN}-dev section doesn't call them lib${PN} Apr 21 03:03:17 right? Apr 21 03:46:24 wb Apr 21 03:47:52 kergoth: are you from phoenix? Apr 21 03:57:01 * Nedlinpopo grew up in tempe Apr 21 03:57:14 ah well, time to go Apr 21 06:47:49 anyone around who understands PACKAGES_ and FILES_ in a bitbake recipe? Apr 21 06:52:55 what do you mean by PACKAGES_? Apr 21 06:53:13 kergoth was talking about it earlier Apr 21 06:53:57 he talked about PACKAGES without underscore.. Apr 21 06:54:08 I'm trying to figure out why the opencv recipe doesn't put all the bits of it into the filesystem Apr 21 06:54:35 right Apr 21 06:54:46 i'm feeling pretty confused right now Apr 21 06:55:30 there's somethings that aren't very explicit going on here, and bitbake mixes terms in the docs. Apr 21 06:55:54 if you build it with -v you'll see what files are not part of any package Apr 21 06:56:15 you mean bitbake -v opencv? Apr 21 06:56:20 and if you check workdir than you can check packages-split directory to see what files are in which package Apr 21 06:56:25 yes Apr 21 06:57:04 yeah, the thing is that in packages-split there were a whole lot of subdirectories that were not mentioned in the recipe Apr 21 06:57:39 FILES_ variables have sane defaults Apr 21 06:58:55 somehow the symlinks for libopencv_core.so ar enot making it in to the package and the sysroot. I can see they're created but they don't make it in to the package Apr 21 06:59:22 i think it's somethign wrong with the FILES_${PN}-dev line Apr 21 06:59:45 but i think that interacts with do_split_packages Apr 21 07:00:16 ./opencv_2.2.bb:FILES_${PN}-dev = "${includedir} ${libdir}/pkgconfig" should probably add to it, not override default Apr 21 07:00:28 but I haven't really check opencv Apr 21 07:01:12 what exactly does that say? Apr 21 07:01:49 that all files in /usr/include and /usr/lib/pkgconfig will go to opencv-dev Apr 21 07:02:26 but each lib* is packaged separately first by that populate_packages_prepend Apr 21 07:02:29 is pkgconfig a lis tof files that's read or literally a file Apr 21 07:03:13 must be literaly a file Apr 21 07:06:35 does the populate_packages_prepend create path variables like ${libdir}? Apr 21 07:08:17 actually the packages-split/opencv-dev doesnt' have any of the libs in it Apr 21 07:08:33 i wonder how the ones that are in the filesytem got in at all Apr 21 07:19:18 JaMa|Zzz: do you see where the opencv recipe builds any packages? is that implicit from the FILES_line or was it some inherited macro? Apr 21 07:30:08 hi oe dev ! Apr 21 07:30:22 im lookin for some help Apr 21 07:30:39 can i post my question here ? Apr 21 07:30:44 uh, hey there! I'm not an OE dev, but give it a shot Apr 21 07:30:58 some of the OE guys are around too Apr 21 07:31:08 i want to chnage the password for the root user Apr 21 07:31:32 i have no idea in which way to do that Apr 21 07:31:32 there's some stuff about that... let me look Apr 21 07:31:41 ok thanks ! Apr 21 07:31:53 do you want to change it or clear it? Apr 21 07:32:49 i want to set a new password, by default there is an an empty passwd ... Apr 21 07:33:23 that's not quite true, by default there's no password, there's some scripting to clear it Apr 21 07:33:47 (no password as in: not allowed to log in) Apr 21 07:36:01 okay, i think you can do it with sed Apr 21 07:36:25 you already have your own recipe right? Apr 21 07:37:18 yes i have a recipe to build the rootfs of my custom distro Apr 21 07:37:44 okay, take a look at the OE/classes/image.bbclass Apr 21 07:38:01 there's a function called zap_root_password Apr 21 07:38:42 you should be able to write your own function that uses a string other than '*' to fill in the root password filed Apr 21 07:38:54 i have seeen the function zap_root_passwd Apr 21 07:39:15 hello Apr 21 07:39:15 well you can copy it, and use it to put somethign else in there Apr 21 07:40:14 should i add the function name in the varaible IMAGE_PREPROCESS_COMMAND ? Apr 21 07:40:21 there might be a clever way with a chroot and passwd, but it's not immediately obvious to me that will work Apr 21 07:41:24 I think so Apr 21 07:41:44 ok, thanks a lot for this clue !! Apr 21 07:41:51 probably by using += also Apr 21 07:41:56 i will try this way Apr 21 07:42:07 thanks Apr 21 07:42:18 hope it works Apr 21 07:42:34 * Nedlinpopo is new to OE, but not new to linux Apr 21 07:42:41 I just had one of the weirdest bug Apr 21 07:42:56 was it an actual insect? Apr 21 07:43:08 i was wondering if it was related to OE, or to my distro Apr 21 07:43:20 what distro? Apr 21 07:43:46 something which does not begin with deb :) Apr 21 07:44:31 well, there's only about 50 gazillion of those. Apr 21 07:44:35 :P Apr 21 07:45:12 there's a grep in do_rig_locales in glibc.inc Apr 21 07:45:57 for some reason I don't understand, when grep fails to find an occurence in the source file, the do_rig_locales function just exits Apr 21 07:46:21 I think that is beyond my OE knowledge Apr 21 07:46:35 if you hang around for a while, someone else may be able to help Apr 21 07:46:39 The only solution which worked was to add || true at the end of the line Apr 21 07:47:32 I sure hope so, this stuff makes me feel stupid :/ Apr 21 07:48:29 it seems to me that bitbake is still growing, and so this relatively minor thing may just be a minor regression Apr 21 07:48:40 but i only speak for myself Apr 21 07:48:56 yeah especially with the poky merge :) Apr 21 07:48:59 * Nedlinpopo is merely a confused user Apr 21 08:11:52 okay sleepytime Apr 21 08:33:50 @Nedlinpopo : i've tried to set the root password qiththe function "zap_root_passwd" but the file /etc/passwd is unchanged Apr 21 08:39:31 look, many images contain #zap root password for release images Apr 21 08:39:33 ROOTFS_POSTPROCESS_COMMAND += '${@base_conditional("DISTRO_TYPE", Apr 21 08:39:35 "release", "zap_root_password; ", "",d)}' Apr 21 08:43:37 my disto is tag as debug Apr 21 09:08:44 i cant set the root user password Apr 21 09:09:55 i'm able to chnage the file /etc/password in order to add the root password, after the task do_rootfs and before the task do_build Apr 21 09:10:20 but it change nothing in my target Apr 21 09:10:27 anyone ? Apr 21 09:18:52 http://cgit.openembedded.org/cgit.cgi/openembedded/tree/recipes/base-passwd/files/root-home.patch?h=org.openembedded.dev Apr 21 09:20:24 th eonly issue I can remember is http://patchwork.openembedded.org/patch/2096/ Apr 21 09:20:34 it should just work ;) Apr 21 12:02:15 there is somebody here who can explain me how to set a password for the root user please ? Apr 21 12:09:38 passwd root Apr 21 12:13:55 oops, i want to set the password at the do_rootfs stage Apr 21 13:58:11 03Eyal Reizer  07master * r79ea0f7541 10openembedded.git/recipes/ti/ti-wifi-utils_git.bb: Apr 21 13:58:11 ti-wifi-utils: update to a newer git commit Apr 21 13:58:11 * Change to a newer calibrator version supporting new wl12xx chips. Apr 21 13:58:11 * Add a PR number Apr 21 13:58:11 Signed-off-by: Eyal Reizer Apr 21 13:58:11 Acked-by: Koen Kooi Apr 21 13:58:11 Signed-off-by: Denys Dmytriyenko Apr 21 13:58:22 03Eyal Reizer  07master * r3b3e1c91e9 10openembedded.git/recipes/hostap/ (4 files in 2 dirs): (log message trimmed) Apr 21 13:58:22 hostap-daemon: build from git sources Apr 21 13:58:22 * Builds a new version 0.8.x which is not released yet as a zip file. Apr 21 13:58:22 This version has new features like WIFI-direct used in new mac80211 Apr 21 13:58:23 versions Apr 21 13:58:23 * As the recipe is very similar to the 0.7.3 version, I have created Apr 21 13:58:24 a common include file shared between the 0.7.3 version and the Apr 21 13:58:32 03Eyal Reizer  07master * r20a0d0c566 10openembedded.git/recipes/wpa-supplicant/ (7 files in 2 dirs): (log message trimmed) Apr 21 13:58:32 wpa-supplicant: Build from git sources Apr 21 13:58:32 * Builds a new version 0.8.x which is not released yet as a zip file. Apr 21 13:58:32 This version has new features like WIFI-direct used in new mac80211 Apr 21 13:58:32 versions Apr 21 13:58:32 * Add PV Apr 21 13:58:34 * Update defconfig Apr 21 14:09:20 hi all Apr 21 14:09:52 I still have problems with dbus not installing init scripts links in /etc/rc*.d/ Apr 21 14:10:06 in dbus.inc there are these values. Apr 21 14:10:14 INITSCRIPT_NAME = "dbus-1" INITSCRIPT_PARAMS = "start 02 5 2 . stop 20 0 1 6 ." Apr 21 14:10:22 are they correct? Apr 21 14:10:42 in my image I've only /etc/init.d/dbus-1 script Apr 21 14:10:50 but no links for runlevels Apr 21 14:21:41 mrAlmond: they're fine on my system here Apr 21 14:22:37 mrAlmond: lrwxrwxrwx 1 root root 16 Jan 13 14:47 /etc/rc2.d/S02dbus-1 -> ../init.d/dbus-1 Apr 21 14:23:30 foerster : This is very strange...I've also tried to clean and rebuild it but with no luck Apr 21 14:24:12 btw..can you explain me the sequence in INITSCRIPT_PARAMS ? Apr 21 14:24:23 are you looking in rootfs, or what? Apr 21 14:24:39 this happens on first boot i think b/c of the adduser stuff in pkg_postinst Apr 21 14:24:50 I'm looking at the final image Apr 21 14:24:57 bah, not adduser stuff, mkdir stuf Apr 21 14:24:59 btw...I'm using cramfs image Apr 21 14:25:07 that is read-only Apr 21 14:25:32 I've added cramfs recipe because it was missing Apr 21 14:25:55 well, some things require to be run at boot up time Apr 21 14:26:00 this appears to be of that case Apr 21 14:26:23 so you say that dbus links are created during the first boot? Apr 21 14:26:34 see dbus.inc, line 87: http://cgit.openembedded.org/cgit.cgi/openembedded/tree/recipes/dbus/dbus.inc#n87 Apr 21 14:28:41 the postinst exits if it's been run on the build machine because $D is non-empty then. Apr 21 14:29:43 yes...it says # can't do adduser stuff offline Apr 21 14:31:00 I had problems with dbus so I have to add manually the things that it's doing in that postinst script Apr 21 14:31:27 but I don't understand where it should create links Apr 21 14:31:30 at run time Apr 21 14:31:49 could it be : chmod u+s /usr/libexec/dbus-daemon-launch-helper ? Apr 21 14:31:56 but it's not launched Apr 21 14:33:11 FYI -you'll normally see an /etc/rcS.d/S40configure that will run opkg-cl configure, which i believe will run the postinst. S40configure deletes itself after it's run once. Apr 21 14:33:35 dbus works fine here, so I'm not sure what exactly you're after. Apr 21 14:34:55 mine is S98configure Apr 21 14:35:08 I've seen that it runs opkg-cl Apr 21 14:35:26 but I wasn't aware of what it was doing Apr 21 14:36:06 so I'm having all these problems because of the read-only filesystem Apr 21 14:36:07 afaik, any postinst that does "exit 1" on the host will then end up running on the target with that Apr 21 14:36:29 fully read-only fs isn't the normal use-case (at least, that's what I had seen before). Apr 21 14:36:36 ok tnx for the hint Apr 21 14:36:47 I have to use cramfs...no other ways... :-( Apr 21 14:36:58 so I mounted /tmp in tmpfs Apr 21 14:37:14 and I made some writable files links to /tmp/ Apr 21 14:38:03 maybe try unionfs on /etc or something? Apr 21 14:38:12 and run opkg-cl configure every boot? Apr 21 14:38:24 I think so... Apr 21 14:38:34 because S98configure is never removed Apr 21 14:38:53 yes, b/c you cannot delete it since it's RO fs Apr 21 14:39:33 is there a standard way to tell yocto to use unionfs on the /etc Apr 21 14:40:14 no idea, i'm still using oe.dev Apr 21 14:41:01 ok...thank you for the help Apr 21 14:41:05 np Apr 21 14:57:20 where should I find the postinst code in the rootfs...the one that should be executed when "opkg-cl configure" is run? Apr 21 14:57:41 because actually "opkg-cl configure" does nothing Apr 21 15:05:18 /usr/lib/opkg/info/.postinst Apr 21 15:06:51 whoops...that dir is empty in my rootfs Apr 21 15:07:26 no...I don't have that dir at all... Apr 21 15:07:36 getting beyond my depth of knowledge now. Perhaps something to do with ONLINE_PACKAGE_MANAGEMENT variable or something. Apr 21 15:08:18 ok Apr 21 15:10:52 I've found the problem... my .postinst files are under /var/volatile/lib/opkg/info/ Apr 21 15:11:04 and here there is dbus-1 in which links are created Apr 21 15:11:22 but at runtime this directory is mounted in tmpfs Apr 21 15:11:44 so these files are "removed" because that dir is re-mounted Apr 21 15:11:56 ah Apr 21 15:12:42 I don't know why they are there...they should be placed as you said in /usr/lib/opkg/info/ Apr 21 15:25:33 a small kernel related question (sorry) Apr 21 15:25:48 I boot with "mem=100M" Apr 21 15:26:07 but if I do "cat /proc/meminfo" Apr 21 15:26:38 I have a memtotal of only 45M Apr 21 15:26:52 so, I missing >50M Apr 21 15:26:54 :) Apr 21 15:27:45 if I look at the messages being printed, I see Apr 21 15:27:53 Memory: 50MB = 50MB total Apr 21 15:27:53 Memory: 45628k/45628k available, 56772k reserved, 0K highmem Apr 21 15:28:07 so Apr 21 15:28:28 for some reason 56M is being reserved by the kernel when it's booting Apr 21 15:28:35 anyone knows why? Apr 21 15:30:51 tdebrouw, pastebin the whole dmesg, and tell machine/disrto/image combo Apr 21 15:35:41 http://pastebin.com/QG0ae2Kn Apr 21 15:36:21 machine is a ti8168evm (cortexA8) Apr 21 15:36:29 distro arago -> overlay on oe Apr 21 15:36:44 image -> console - nothing fancy Apr 21 15:37:19 I just wonder why 56M is reserved (see line 83) Apr 21 15:38:02 errrr Apr 21 15:38:18 please ignore me for a few secs Apr 21 15:38:25 let me check something first Apr 21 15:40:26 yep Apr 21 15:40:40 i'm deeply sorry :-/ Apr 21 15:41:11 it seems it's used for the framebuffers (line 71) - I missed that one. Apr 21 15:45:05 np Apr 21 15:45:11 but your board has 2GB??? Apr 21 15:45:11 thats quite some framebuffer :-D Apr 21 15:45:33 ah 2GiB Apr 21 15:45:43 DRAM: 2 GiB Apr 21 15:45:52 maybe it means 512M Apr 21 15:47:25 03Martin Jansa  07master * r9b8912fa3a 10openembedded.git/recipes/freesmartphone/libfreesmartphone-glib_git.bb: Apr 21 15:47:25 libfreesmartphone-glib: bump SRCREV and PV to fix build against newer fso-specs Apr 21 15:47:25 Signed-off-by: Martin Jansa Apr 21 15:47:30 my board had 2 GB :) Giga Bytes Apr 21 15:47:49 but I can only use 100M atm Apr 21 15:47:52 :( Apr 21 15:48:45 640k is enough for anyway Apr 21 15:49:22 *anyone? Apr 21 15:49:44 tdebrouw: aye, typo central here Apr 21 16:11:12 http://maps.google.com/maps?f=q&source=s_q&hl=en&geocode=&q=2390+Lawrence+St,+Eugene,+OR+97405&aq=0&sll=37.0625,-95.677068&sspn=35.631106,70.224609&ie=UTF8&hq=&hnear=2390+Lawrence+St,+Eugene,+Oregon+97405&ll=44.033017,-123.098261&spn=0.000987,0.002143&t=h&z=19&layer=c&cbll=44.033224,-123.098408&panoid=7RdQZSe2KtXdxS05sRyU8w&cbp=12,337.37,,0,0 Apr 21 16:11:30 wrong window, sorry Apr 21 18:20:20 oh.. meta-gnome Apr 21 18:21:01 * Jay7 fears that something like current x11-image with gpe for zaurus will require about 3-4 repos Apr 21 18:21:09 well, layers Apr 21 18:25:35 Jay7: I am also taking the yocto induced layer movement with a grain of salt. But there are lots of pros as well Apr 21 18:26:03 hiya dcordes Apr 21 18:26:34 dcordes: yeah.. flexibility vs usability Apr 21 18:27:17 Jay7: bon voyage Apr 21 18:27:18 Jay7: I don't see a problem with 3-4 layers, personally. as you say, gives you flexibility. we just need solid tooling to improve the usability Apr 21 18:28:04 kergoth: we need strict and transparent policy Apr 21 18:28:12 mostly updating policy Apr 21 18:28:21 kergoth: if you think about committing changes that affect various layers, tooling becomes difficult. Apr 21 18:28:46 s/Jay7/ka6sox-farfarawa/ Apr 21 18:29:13 I don't see it being particularly difficult Apr 21 18:32:04 kergoth: I am trying to think of situations where a script can lead a way when it comes to adding a change that affects various layers. with the one-metadata-soup you just fix everything in your single working branch, run a test build and are set.. Apr 21 18:33:05 and I'm also a bit worried about compatibility of distros which are now outsourced Apr 21 18:49:22 gm Apr 21 18:49:29 hey likewise Apr 21 18:49:33 long time Apr 21 18:49:39 how r u Apr 21 18:49:54 busy with hardware bringup (sam9g45 board is finished) Apr 21 18:50:04 so now back in OE land :-) Apr 21 18:50:16 nice Apr 21 18:50:26 and you? Apr 21 18:50:26 oe landscape has changed a bit Apr 21 18:50:37 I am same usual occasionally hacking on oe Apr 21 18:53:17 likewise: are you looking into oe-core/yocto Apr 21 18:53:26 khem: yes Apr 21 18:53:31 nice Apr 21 18:53:47 only for non-critical developments (i.e. my own projects) Apr 21 18:54:18 I see Apr 21 18:54:45 I should get my git and workflow in sync with the pull model. Apr 21 18:55:41 likewise: yes, there are some helpful tools in oe-core Apr 21 18:56:01 I need to do a wiki writeup actually that can help folks in workflow Apr 21 18:56:07 he khem Apr 21 18:56:11 woglinde: hi Apr 21 19:03:47 he jama Apr 21 19:05:21 he Apr 21 19:07:55 khem: I bet many would welcome such wiki page on oe-core based workflow Apr 21 19:08:11 khem: including me Apr 21 19:09:02 dcordes: ok, I will make it a priority Apr 21 19:09:19 after playing soccer with my son this evening Apr 21 19:09:40 khem, I stopped garnet after it built everything in 2days. Apr 21 19:09:51 iirc it was over 5days before. Apr 21 19:09:56 khem: cool thanks a lot ! Apr 21 19:10:06 ka6sox-farfarawa: ok 2 days for 4 arches ? Apr 21 19:10:35 yes Apr 21 19:10:38 ka6sox-farfarawa: for 5 days I built tonnes of images Apr 21 19:10:49 are you using same bake.sh ? Apr 21 19:10:55 or oe-core based Apr 21 19:10:59 yes bake.sh Apr 21 19:11:04 oh wow Apr 21 19:11:09 thats impressive Apr 21 19:11:10 it built all of them in 2days. Apr 21 19:11:17 good job Apr 21 19:11:23 ya, so thats stopped now... Apr 21 19:11:26 * khem hands an eBeer to ka6sox-farfarawa Apr 21 19:11:38 ok power it up Apr 21 19:11:41 I need to know what repos to pull oe-core Apr 21 19:11:44 I can setup oe-core on it Apr 21 19:12:04 kk...go to the poky stuff and put the right bits in there please? Apr 21 19:12:14 its alongside your stuff. Apr 21 19:12:17 let me write a new script there which will do it Apr 21 19:12:30 its up Apr 21 19:12:33 khem I already played soccer today Apr 21 19:12:34 ok Apr 21 19:12:41 woglinde: time difference :) Apr 21 19:12:51 now beert Apr 21 19:12:57 ups beer Apr 21 19:13:06 okay I was going to use the poky builds of just core as I want to start working on getting stats back online Apr 21 19:13:17 and has to watch germany next topmodel Apr 21 19:13:31 woglinde, they are too skinny Apr 21 19:13:40 except for Claudia Apr 21 19:13:49 woglinde: heh married men watch with care Apr 21 19:14:03 khem, true Apr 21 19:14:23 khem my wife is watching Apr 21 19:14:27 so I have too Apr 21 19:14:53 woglinde: heh she is watching what you are watching :) Apr 21 19:15:12 so be careful when you admire anyone of them Apr 21 19:15:23 lol Apr 21 19:15:24 no Apr 21 19:15:37 whole family is watching Apr 21 19:15:55 oh it seems like a family show :) Apr 21 19:16:09 is it a reality tv Apr 21 19:16:48 like Duetschland suckt den superstar Apr 21 19:17:22 ieehks Apr 21 19:17:24 this suckz Apr 21 19:17:39 I still remeber that lame Daniel Apr 21 19:17:41 model show is a bit diffrent Apr 21 19:17:49 and a lot better than the us series Apr 21 19:17:55 woglinde: totally yeah :) Apr 21 19:18:14 are they beaties with brains Apr 21 19:18:15 khem he is now millionaire with solar producing power plants Apr 21 19:18:22 can they program kernel Apr 21 19:18:28 no Apr 21 19:18:37 but one which won, was study maths Apr 21 19:18:44 thats close Apr 21 19:18:57 red hair Apr 21 19:19:35 woglinde: is it like America's Next Top Model Apr 21 19:19:52 I don't like the show since Bruce Darnell left Apr 21 19:20:21 dcordes: did he pay you half of his salary to watch it :) Apr 21 19:20:30 * khem never watch reality tv Apr 21 19:21:26 khem I said its better Apr 21 19:21:37 khem: heh that would have been nices. I 'm not a fan either but sometimes I have to watch it Apr 21 19:21:39 hmm sure Apr 21 19:22:03 German models are better then Amis ;) Apr 21 19:23:02 true...at least Claudia had some meat on her bonse. Apr 21 19:23:05 bones Apr 21 19:23:19 Heidi too Apr 21 19:27:42 * khem needs food Apr 21 19:28:39 hehe Apr 21 19:28:43 barbeque Apr 21 19:28:53 on satarday Apr 21 19:32:25 shot in the dark here, anyone have an idea what might be the origin of an intermittent error we get here with xserver-xorg 1.7.4 against omapfb in linaro 2.6.35? error: http://pastebin.com/fsY6a3nT, non-error: http://pastebin.com/L2dh0BtR Apr 21 19:32:35 maybe should ask in #linaro or #linux-omap Apr 21 19:43:17 03Stanislav Brabec  07master * rfed15638c6 10openembedded.git/recipes/tcltk/tcl_8.5.8.bb: Apr 21 19:43:17 tcl: Removed orphan reference to tcl_sysroot introduced by 7c4aeaf. Apr 21 19:43:17 * See http://lists.linuxtogo.org/pipermail/openembedded-devel/2011-April/032126.html Apr 21 19:43:17 Signed-off-by: Stanislav Brabec Apr 21 20:11:13 a newbie question, when i do 'opkg update', how is it decided to pick up wich opkg config files under /etc/opkg? Apr 21 20:11:54 there are quite a lot conf files under /etc, and i want to ignore most if not all of them, instead want to install from a local directory /ipkg instead Apr 21 20:13:08 i have already added FEED_DEPLOYDIR_BASE_URI "file:///ipk" to local.conf Apr 21 20:14:06 ? Apr 21 20:14:26 that was extremely unclear. you mentioned both build time and runtime, and where you're installing to and from Apr 21 20:14:40 basically i want to opkg from a local directory instead of http, Apr 21 20:14:43 installing packages from another directory is just a matter of adding an src to opkg.conf Apr 21 20:14:45 one line. Apr 21 20:14:59 src foo file:/path/to/my/ipk/repository Apr 21 20:15:07 iirc, anyway. Apr 21 20:15:55 when i do opkg update what conf file it will pick up, i saw "opkg.conf", mach-feed.conf, noarch-feed.conf, arch.conf,etc Apr 21 20:16:38 basically tehre are 20 feed.conf under /etc, i don't know how opkg is to pick them up, if there is one opkg.conf that it's easier to understand Apr 21 20:17:09 and...i found no doc on this either Apr 21 20:19:25 xxiao thats angstroem anyway Apr 21 20:20:25 this is true, this question has nothing to do with oe and everything to do with angstrom and opkg Apr 21 20:27:52 does angstrom has an irc? Apr 21 20:27:54 thanks guys Apr 21 20:28:12 i will bug koen then Apr 21 20:28:39 #angstrom, iirc Apr 21 20:29:05 i wouldn't recommend bugging koen (or anyone else) directly. ask the question in the right irc channel or mailing list, thereby giving others the opportunity to respond as well Apr 21 20:34:23 koen's watching tv anyway Apr 21 20:36:02 NL top model? Apr 21 20:37:39 he didn't say Apr 21 21:17:23 I have a recipe that produces a couple packages, is it possible to prevent do_runstrip from processing just one of the packages? Apr 21 21:24:54 hi all Can any one tell me what package creates the /dev folder when we build teh filesystem image Apr 21 21:26:09 files/device_table_something.txt if I recall correctly Apr 21 21:28:34 thanks Apr 21 21:44:55 Hi all I am getting the following error while building the filesystem can any one tell me what causes this error "ERROR: ld.so: object 'libfakeroot.so' from LD_PRELOAD cannot be preloaded: ignored. in do_rootfs" Apr 21 22:02:08 atl19711: I dont think so Apr 21 22:02:28 you can try to redefine PACKAGE_STRIP Apr 21 22:02:45 like its done in gcc-package-cross.inc but that may not be all Apr 21 22:03:04 sreddy: fakeroot is not working it seems Apr 21 22:03:12 whats your host distro Apr 21 22:03:18 and is it with classic oe ? Apr 21 22:05:13 khem: what is classic oe Apr 21 22:06:35 khem: I was hoping I could insert a function into PACKAGEFUNCS or maybe override runstrip that exist to not strip some specific files, thought I would ask Apr 21 22:10:04 khem : we are using arago distro Apr 21 22:13:54 khem: thanks, gcc-package-cross.inc looks useful Apr 21 22:24:03 khem: host is ubuntu 10.04 and distro which we are using is arago Apr 21 22:28:44 khem: I'm having the fakeroot issues .. running Ubuntu 10.04 Apr 21 22:36:29 03Stanislav Brabec  07master * r2ebaf11c15 10openembedded.git/recipes/quilt/ (quilt-package.inc quilt_0.48.bb): Apr 21 22:36:29 quilt: Do not check host system when building non-native package. Apr 21 22:36:30 Signed-off-by: Stanislav Brabec Apr 21 22:58:21 well, in /udev we have unnecessary amount of cruft :/ Apr 21 22:59:01 that would obliuge to have an override for each version for a machine... Apr 21 22:59:11 let see if I can fix it Apr 21 23:36:31 hey guys, question for you: Apr 21 23:37:50 on Ubuntu 11.04 perl-native fails to build because they moved some of the core libs and the perl automake stuf ignores what gcc says it's search path is. I've got a patch that just adds those 2 libraries to the search path. Apr 21 23:38:17 It *probably* can't hurt any other platform, but i don't have any others to test on at the moment Apr 21 23:38:26 do you guys want the patch? Apr 21 23:55:35 03Andrea Adami  07org.openembedded.dev * r6ddb7f900a 10openembedded.git/recipes/udev/ (99 files in 29 dirs): Apr 21 23:55:35 udev: remove stale patches and reorder directory structure. Apr 21 23:55:35 * many files were committed as version-specific while most are common Apr 21 23:55:35 * allow easier overrides of the defaults once for all recipes Apr 21 23:55:35 * bump INC_PR Apr 21 23:55:36 Signed-off-by: Andrea Adami Apr 21 23:56:36 jeez..this patch was for another branch Apr 22 00:05:23 03Andrea Adami  07org.openembedded.dev * r2c4042a2eb 10openembedded.git/recipes/udev/ (99 files in 29 dirs): Apr 22 00:05:23 Revert "udev: remove stale patches and reorder directory structure." Apr 22 00:05:23 This reverts commit 6ddb7f900a19e51f70ecc01214f8ecb13ae98cac. Apr 22 00:08:36 03Andrea Adami  07org.openembedded.dev * rfcb6e4993d 10openembedded.git/recipes/udev/ (99 files in 29 dirs): (log message trimmed) Apr 22 00:08:36 udev: remove stale patches and reorder directory structure. Apr 22 00:08:36 * second try Apr 22 00:08:36 * many files were committed as version-specific while most are common Apr 22 00:08:36 * allow easier overrides of the defaults in /udev once for all recipes Apr 22 00:08:36 * prepare for removal of obsolete versions Apr 22 00:08:37 * bump INC_PR **** ENDING LOGGING AT Fri Apr 22 02:59:57 2011