**** BEGIN LOGGING AT Thu Feb 17 02:59:58 2011 Feb 17 04:55:04 morning Feb 17 04:55:13 What could be the main problems of such error: Feb 17 04:55:14 | Collected errors: Feb 17 04:55:14 | * opkg_install_cmd: Cannot install package recess-sqlite. Feb 17 05:12:36 fixed Feb 17 06:37:25 03Khem Raj  07master * r4677d73b1a 10openembedded.git/recipes/gcc/ (11 files in 3 dirs): (log message trimmed) Feb 17 06:37:25 gcc-4.5: Bring latest from linaro 4.5 and bump svn SRCREV for upstream Feb 17 06:37:25 gcc-4.5-linaro-r99474.patch is disabled because it ends up Feb 17 06:37:25 bombing x86 cross-gcc build see Feb 17 06:37:25 http://lists.linaro.org/pipermail/linaro-dev/2011-February/002614.html Feb 17 06:37:26 gcc-4.5-linaro-r99474.patch is disabled because it causes GCC crash Feb 17 06:37:27 for XScale see **** BEGIN LOGGING AT Thu Feb 17 07:32:20 2011 Feb 17 07:37:29 good morning Feb 17 07:43:40 Tartarus, khem : I'm trying to solve kaeilos libtool issue, could you please explain what's wrong? Feb 17 08:16:27 morning Feb 17 08:16:40 Jay7: gm Feb 17 08:17:09 morning Feb 17 08:36:56 yo dudes Feb 17 08:38:10 yo Feb 17 08:38:39 can somebody verify git is functioning normally? Feb 17 08:41:14 XorA|gone, can you please verify you can edit the wiki? Feb 17 08:47:01 ka6sox: I can with no captcha now :-D Feb 17 08:48:17 ka6sox: git is ok Feb 17 08:49:29 XorA|gone, perfect Feb 17 08:49:37 Jay7, thanks...jsut verifying Feb 17 08:50:54 git is aptly named :-D Feb 17 08:51:17 yes, it is a Git. Feb 17 08:55:20 okay...my work here is done...time to save the universe. Feb 17 08:57:55 ka6sox: I think you would work better as a villian :-D Feb 17 09:01:08 XorA|gone, why? Feb 17 09:04:11 x at the end of you name :-D Feb 17 09:04:17 aha... Feb 17 09:04:31 does that make you a GOOD guy with an X to start? Feb 17 09:04:59 Im zero :-) Feb 17 09:05:32 he he...okay bbiab...server is back up and healthy... Feb 17 09:05:38 (for now) Feb 17 09:05:56 hehe, if only they stayed that way Feb 17 09:06:06 but then I guess I wouldnt make money in that case Feb 17 09:11:02 gm Feb 17 09:12:48 03Koen Kooi  07org.openembedded.dev * r0812019bd7 10openembedded.git/recipes/gnome/libsoup-2.4_2.33.6.bb: Feb 17 09:12:48 libsoup-2.4 2.33.6: lower default pref Feb 17 09:12:48 * it's a release from the unstable series Feb 17 09:12:48 * it needs a too new glib-2.0 Feb 17 09:12:48 Signed-off-by: Koen Kooi Feb 17 09:13:18 XorA|gone: can you check how big lego plate is needed to be usable for base of pandaboard case? 13x13cm size I think but I do not know how much it is in lego dots Feb 17 09:13:50 hrw: remind me next week, Im in Japan Feb 17 09:14:06 XorA|gone: ok Feb 17 09:19:02 03Koen Kooi  07org.openembedded.dev * r71c14f008a 10openembedded.git/conf/distro/include/angstrom-2010-preferred-versions.inc: Feb 17 09:19:03 angstrom next: move to glib 2.28.0 Feb 17 09:19:03 Signed-off-by: Koen Kooi Feb 17 09:20:30 03Koen Kooi  07org.openembedded.dev * r438e74c76d 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: Feb 17 09:20:30 angstrom 2008: prefer a libsoup from the stable series Feb 17 09:20:30 Signed-off-by: Koen Kooi Feb 17 09:21:44 http://geekandpoke.typepad.com/.a/6a00d8341d3df553ef0147e2876bcc970b-pi Feb 17 09:24:39 hrw: so this did happen to you? :P Feb 17 09:26:27 hrw: LOL Feb 17 09:26:49 hrw: what a good news! congratulations! Feb 17 09:27:57 no, thats not mine Feb 17 09:28:05 just got picture link Feb 17 09:30:14 * ant_work sees Poky fixed and reenabled PARALLEL_MAKE for gettext, ncurses, pango, eds-dbus and openssl Feb 17 09:31:14 * ant_work even scrubs a patch for perl-native (in testing) Feb 17 09:40:35 how could I deal with ERROR: QA Issue with db-bin errors? How can I patch them? Feb 17 09:41:11 FATAL: QA run found fatal errors. Please consider fixing them. but how? where? Feb 17 09:41:31 /recipes/db/db_5.0.21.bb ? Feb 17 09:47:41 ant_work: let us know how that worked :-) Feb 17 09:49:17 he Feb 17 09:49:51 are we supposed to cherry-pick from Poky? Should we just wait for oe-core to be setup? Feb 17 09:50:07 I'll have to reread the ML Feb 17 09:54:32 it's too long Feb 17 09:54:43 I haven't read it for a day and I'm lost now Feb 17 09:54:49 he he Feb 17 09:55:04 high traffic... Feb 17 09:55:08 for curiosity, I read poky and yocto archives too Feb 17 09:55:26 s/read/'ve read/ Feb 17 09:55:44 too many lists for an human ! Feb 17 09:56:21 ynezz, anyway, the sentence is Feb 17 09:56:23 "We all agreed to populate an openembedded-core repo starting with Poky" Feb 17 09:56:29 ant_work: I think both paths should be taken Feb 17 09:56:57 ant_work: feel free to cherry-pick Feb 17 09:57:00 ant_work: we don't want our current meta data to die on us while we transition (or "try") openembedded-core Feb 17 09:57:07 ant_work: having some recipes in sync is good Feb 17 09:57:23 likewise: exactly, the transition will take some time Feb 17 09:57:41 we could/should sync between poky and oe in this time Feb 17 09:57:51 I hope the designed dev (Khem?) will push from Poky both in oe-core and in master Feb 17 09:58:14 oh, so there's oe-core already? Feb 17 09:58:21 ant_work: We still need to figure out all the details Feb 17 09:58:41 fine but the gap is enlarging Feb 17 09:58:44 ant_work: it should be khem and RP doing the initial push into oe-core and then starting to work on it Feb 17 09:58:47 while we talk Feb 17 09:59:11 ant_work: sure it is, no way we could stop any of the two projects working Feb 17 09:59:38 ynezz: the repo is not there yet. Should come in the next days Feb 17 10:05:24 about the gap...can OE swallow a Poky-recipe as-it-is today? Feb 17 10:49:10 hello! Feb 17 10:49:21 I had some problems with g_ether, so I tried updating OE. Now I don't see g_ether.ko anywhere. Feb 17 10:49:31 Even weirder - /lib/modules has two sub-directories 2.6.36 and 2.6.34. uname -r prints 2.6.34 Feb 17 10:49:45 (using beagleboard demo image) Feb 17 11:11:33 hello. I have probleb to compile libnl2-1. compiler exit with error: Feb 17 11:12:56 | route/sch/netem.c: In function 'rtnl_netem_set_delay_distribution': Feb 17 11:12:56 | route/sch/netem.c:867: error: 'NAME_MAX' undeclared (first use in this function) Feb 17 11:12:56 | route/sch/netem.c:867: error: (Each undeclared identifier is reported only once Feb 17 11:12:56 | route/sch/netem.c:867: error: for each function it appears in.) Feb 17 11:12:56 | route/sch/netem.c:867: warning: unused variable 'name' Feb 17 11:12:57 ~pastebin Feb 17 11:12:57 | make[2]: *** [route/sch/netem.lo] Error 1 Feb 17 11:12:57 Feb 17 11:12:57 [~pastebin] A "pastebin" is a web-based service where you should paste anything over 3 lines so you don't flood the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://bin.cakephp.org/ , http://asterisk.pastey.net/ , or install pastebinit with yum or aptitude. Feb 17 11:20:54 tsahee, hi Feb 17 11:21:35 can you just opkg install it for a start? Feb 17 11:21:58 then maybe add it to the modules to be shipped with the images in the beagleboard.conf Feb 17 11:25:35 hi obi Feb 17 11:29:26 Please help. I have problem to compile libnl2_2.0.bb: Feb 17 11:29:26 | route/sch/netem.c: In function 'rtnl_netem_set_delay_distribution': Feb 17 11:29:26 | route/sch/netem.c:867: error: 'NAME_MAX' undeclared (first use in this function) Feb 17 11:29:40 google Feb 17 11:30:05 yourself Feb 17 11:30:13 where NAME_MAX should come from Feb 17 11:30:20 tnx Feb 17 11:30:33 anyway you didnt tell us which config you are used Feb 17 11:31:57 last time git corrected Feb 17 11:32:44 yes. distro angstrom-2008.1 mashine at91sam9263 and build base-image Feb 17 11:38:38 NAME_MAX is defined in linux/limits.h Feb 17 11:38:46 seems your kernel dont has it Feb 17 11:38:52 or kernel-headers Feb 17 11:48:52 no limits.h include NAME_MAX (=255) Feb 17 11:48:52 i'm see it in: Feb 17 11:48:52 /build/tmp/work/at91sam9263ek-angstrom-linux-gnueabi/linux-2.6.28-r15/linux-2.6.28/include/linux Feb 17 11:49:38 okay Feb 17 11:49:44 then libnl misses the include Feb 17 11:49:54 but limits.h isnt included directly Feb 17 11:52:13 most likely the compiler header files without starting. Feb 17 11:52:50 where it can be corrected? Feb 17 11:53:34 hm try toinclude it directly Feb 17 11:53:46 uh yeah .28 is old Feb 17 12:07:04 GNUtoo|laptop: thanks (sorry for the long delay, was away..) I managed to install a package names usb-gadget. Feb 17 12:07:25 It installs a binary named usb-gadget which I'm not sure what it does Feb 17 12:08:19 I tried looking for a package named kernel-module- *gadget* or things like that but didn't find anything Feb 17 12:21:55 tsahee, try kernel-module-g-ether Feb 17 12:22:20 opkg list | grep ether Feb 17 12:22:23 if it's not there Feb 17 12:22:48 you should modify /etc/opkg/*-feed.conf Feb 17 12:22:55 to point to your repository Feb 17 12:23:05 and then opkg update Feb 17 12:23:10 and then opkg install it Feb 17 12:32:45 I have an ubuntu machine used to build and the beagleboard itself running openembedded. The ubuntu machine doesn't have opkg, the beagleboard doesn't have network (that's what I'm trying to do.. :)) Feb 17 12:33:10 opkg list | grep ether on the beagleboard returned no results Feb 17 12:33:21 so did find tmp/deploy/glibc/ipk/ -name *ether* Feb 17 12:35:27 I'm also trying to look for any recipie that will get me there.. no luck yet Feb 17 12:53:59 Hi guys, I'm getting internal compiler error when bitbaking uclibc, could anyone have a look at this with me? http://pastebin.com/k8fntKXE Feb 17 13:11:51 sedrik: change -mtune=xscale to -mtune=armv5te Feb 17 13:12:05 afaik iwmmxt stuff is broken since gcc 4.4 Feb 17 13:12:20 or use older gcc Feb 17 13:12:57 aa ok Feb 17 13:24:21 anarsoul: hmm, is conf/machine/include/tune-scale.inc the correct file to edit? from how I follow the structure that is where mtune is set Feb 17 13:24:35 or is it someway I can do it in my local conf? Feb 17 13:25:15 sedrik: ask someone who's responsible for toolchain stuff Feb 17 13:41:25 03Eric Bénard  07org.openembedded.dev * r870374197f 10openembedded.git/recipes/nodejs/ (files/libev-cross-cc_0.4.0.patch nodejs_0.4.0.bb): (log message trimmed) Feb 17 13:41:25 nodejs: add 0.4.0 Feb 17 13:41:25 compile tested for armv5, previous node-cross-cc_0.2.6.patch Feb 17 13:41:25 doesn't seems anymore necessary unless I made a mistake, AJ may Feb 17 13:41:25 you please check this point ? Feb 17 13:41:26 Signed-off-by: Eric Bénard Feb 17 13:41:27 Cc: AJ ONeal Feb 17 13:41:30 03Eric Bénard  07org.openembedded.dev * r1bedd87f85 10openembedded.git/recipes/util-linux-ng/ (6 files in 2 dirs): Feb 17 13:41:30 util-linux-ng: add 2.18 Feb 17 13:41:30 from the changelong, the most important update over 2.17 Feb 17 13:41:30 is a memory leak fix in mount Feb 17 13:41:30 Signed-off-by: Eric Bénard Feb 17 13:41:31 Acked-by: Khem Raj Feb 17 13:41:36 03Eric Bénard  07org.openembedded.dev * rfe604206c1 10openembedded.git/recipes/busybox/ (9 files in 3 dirs): Feb 17 13:41:36 busybox: add 1.18.3 and its fixes Feb 17 13:41:36 - 1.18.3 is 1.18.2 + the fixes already presents in OE Feb 17 13:41:36 - additional fixes from http://busybox.net/downloads/fixes-1.18.3/ Feb 17 13:41:37 were added to the recipe Feb 17 13:41:37 Signed-off-by: Eric Bénard Feb 17 13:41:38 Acked-by: Khem Raj Feb 17 13:41:41 03Eric Bénard  07org.openembedded.dev * r0e58dc7404 10openembedded.git/recipes/gnokii/gnokii_0.6.30.bb: Feb 17 13:41:42 gnokii: add latest version Feb 17 13:41:42 - build tested for armv5 with PARALLEL_MAKE="-j8" so I removed Feb 17 13:41:42 PARALLEL_MAKE = "" which was set in previous recipes. Feb 17 13:41:42 Signed-off-by: Eric Bénard Feb 17 13:41:43 03Eric Bénard  07org.openembedded.dev * rde84e4fc39 10openembedded.git/recipes/gnokii/ (gnokii/configure.patch gnokii_0.6.14.bb gnokii_0.6.4.bb): Feb 17 13:41:43 gnokii: remove old recipes Feb 17 13:41:43 Signed-off-by: Eric Bénard Feb 17 13:42:05 03Eric Bénard  07org.openembedded.dev * r53a8358f0a 10openembedded.git/conf/distro/include/angstrom-2010-preferred-versions.inc: Feb 17 13:42:05 angstrom-2010: prefer busybox-1.18.3 Feb 17 13:42:05 this is the latest stable version Feb 17 13:42:05 Signed-off-by: Eric Bénard Feb 17 13:42:05 Cc: Koen Kooi Feb 17 13:42:05 Acked-by: Koen Kooi Feb 17 13:42:10 03Eric Bénard  07org.openembedded.dev * r8f999f569c 10openembedded.git/conf/distro/include/preferred-slugos-versions.inc: Feb 17 13:42:11 slugos: prefer busybox 1.18.3 Feb 17 13:42:11 Signed-off-by: Eric Bénard Feb 17 13:42:11 Cc: Mike Westerhof Feb 17 13:42:11 Acked-by: Khem Raj Feb 17 13:42:11 Acked-by: Mike Westerhof Feb 17 13:42:30 gah! :p I just started a full build... Guess I'll kill it and start again! :D Feb 17 13:59:12 Hey folks -- does anybody here care about xinetd ? Feb 17 14:28:21 hi ericben Feb 17 14:29:31 ping me when you have some time Feb 17 14:31:06 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r4abe105964 10openembedded.git/recipes/pingus/pingus_0.7.2.bb: (log message trimmed) Feb 17 14:31:06 pingus 0.7.2: fix newer boost(1.45.0) detection (break older boosts) Feb 17 14:31:06 Without that fix we have: Feb 17 14:31:06 Fatal Errors: Feb 17 14:31:06 * library 'boost_signals' not found Feb 17 14:31:07 FATAL: scons build execution failed. Feb 17 14:31:07 Note that this commit is incompatible with older(than 1.45.0) boost. Feb 17 14:31:14 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r41d7e2a6ea 10openembedded.git/recipes/linux/ (linux-bug/defconfig linux-bug_2.6.27.2.bb): (log message trimmed) Feb 17 14:31:14 linux-bug: fix undefined reference to `__gnu_mcount_nc' issue and bump PR Feb 17 14:31:14 Without that fix we have: Feb 17 14:31:14 LD .tmp_vmlinux1 Feb 17 14:31:14 init/built-in.o: In function `run_init_process': Feb 17 14:31:15 calibrate.c:(.text+0x10): undefined reference to `__gnu_mcount_nc' Feb 17 14:31:16 init/built-in.o: In function `init_post': Feb 17 14:33:11 hi GNUtoo|laptop Feb 17 14:33:32 I looked for you at fosdem Feb 17 14:33:42 you weren't there right? Feb 17 14:35:14 GNUtoo|laptop: unfortunatly no. To much work actually Feb 17 14:35:21 ah ok Feb 17 14:35:34 can I PM you briefly? Feb 17 14:36:22 if you can't then ping me when you can Feb 17 14:40:05 I did not get enough time to talk with the OE people that stopped by the stand :( Feb 17 14:40:33 I went at the stand for some time but not for all the Fosdem Feb 17 14:40:50 I also went with tuxbrain people and one person from #openmoko-cdevel Feb 17 14:41:15 went with means eat with etc...and attend to some talk with them Feb 17 14:41:22 I also met an enlightenment person Feb 17 14:41:56 yeah, very hectic Feb 17 14:41:56 I even took the enlightenment photo group picture, it was like an xkcd story Feb 17 14:42:29 basically I was the only one with a camera, so I took the picture, and gave the microsd to someone to copy it, Feb 17 14:42:39 also, anytime you want to work the stand, let people know Feb 17 14:42:46 maybe because I had CHDK....his computer didn't recognize it Feb 17 14:42:47 we always need more breaks :) Feb 17 14:43:01 so I tried on my laptop that went out of battery Feb 17 14:43:19 then I gave to someone who had an eeepc....that didn't seem to have a GUI Feb 17 14:43:32 altough it ran X but you could only see a shell on it Feb 17 14:43:49 so he had a hard time figuring out which was the right pictures Feb 17 14:44:04 he forgett to | head or |tail the sort Feb 17 14:44:22 so after heading or tailing it he finally copied the pictures on the usb key..... Feb 17 14:45:15 heh Feb 17 14:46:17 how many programmers do you need for taking a picture? ....a lot.... Feb 17 14:49:44 Crofton, better than not be able to stop for long at the OE stand to talk to the people behind it Feb 17 14:50:29 fosdem is crazy :) Feb 17 14:51:16 Crofton|work, try it as organizer... That's even crazier. Feb 17 14:51:59 I'll bet Feb 17 14:57:37 GNUtoo|laptop: you can PM , latency can be high because I'm between the keyboard and the soldering iron (and the coffe machine also ;-) Feb 17 14:58:10 Crofton|work, well this year I managed to run past the stands quickly. That's an improvement. Cost me money though. Saw the pandaboard and ordered one last week :D Feb 17 14:58:26 cool Feb 17 14:58:53 lol ok Feb 17 14:59:39 Philippe, next year I will have a talk for the embedded room Feb 17 14:59:50 revolving around hanging an fpga off an omap3 Feb 17 14:59:52 hmm Feb 17 14:59:59 unless I have to go to WSR Feb 17 15:00:12 Crofton|work, Would be great. Let's see when Fosdem comes closer. Feb 17 15:00:13 we'll see Feb 17 15:00:53 At least I should hang around IRC and do more stuff again. Fosdem made me realize how Nokia has killed my creativity the last two years Feb 17 15:01:48 yeah Feb 17 15:02:05 morning Feb 17 15:03:56 hey chouimat Feb 17 15:04:10 geiseri: wow you rediscovered irc? Feb 17 15:04:36 lol, im getting problems with bitbake so i am desperate :) Feb 17 15:06:35 for some reason my kernel package is no longer detecting module deps Feb 17 15:09:20 Whats the simplest way of letting OE build both a vmlinux and a zImage from the same build? Feb 17 15:09:43 Well, vmlinux is built for a zImage, always :) Feb 17 15:09:56 But if you mean in general two separate image types, that's a bit harder Feb 17 15:16:33 Yes, I'd like to have both zImage and vmlinux in the deploy directory from the same build. Feb 17 15:24:56 custom machine and such? Feb 17 15:27:22 * Tartarus goes trolling for acks on upgrading db Feb 17 15:35:05 khem: hi Feb 17 15:39:17 Yepp Feb 17 15:40:44 OK, add some lines to do_deploy Feb 17 15:42:07 In which receipt you mean? Feb 17 15:42:26 in the linux recipe for your custom machine Feb 17 15:42:31 or if you used amend.inc, in there Feb 17 15:42:57 Good idea, thanx Feb 17 15:48:10 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r7296ce2b1e 10openembedded.git/conf/machine/bug20.conf: Feb 17 15:48:10 bug20: add MACHINE_EXTRA_RRECOMMENDS Feb 17 15:48:10 Signed-off-by: Denis 'GNUtoo' Carikli Feb 17 15:48:10 Acked-by: Stefan Schmidt Feb 17 15:48:15 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r6fa8906e6e 10openembedded.git/recipes/hostap/ (hostap-daemon-0.7.3/defconfig hostap-daemon_0.7.3.bb): Feb 17 15:48:15 hostapd 0.7.3: convert to libnl2 Feb 17 15:48:15 Signed-off-by: Denis 'GNUtoo' Carikli Feb 17 15:48:15 Acked-by: Stefan Schmidt Feb 17 15:48:21 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * rf3645c50cd 10openembedded.git/recipes/tasks/task-bug.bb: Feb 17 15:48:21 task-bug: make it compatible with bug 1.x,machine arch ${PN} Feb 17 15:48:21 Some packages like bug2v4l2 are specific to bug 2.0, so we need Feb 17 15:48:21 to machine arch the package depending on them. Feb 17 15:48:21 Signed-off-by: Denis 'GNUtoo' Carikli Feb 17 15:48:21 Acked-by: Stefan Schmidt Feb 17 15:48:23 03Denis 'GNUtoo' Carikli  07org.openembedded.dev * r1cf4010dfc 10openembedded.git/recipes/buglabs-osgi/com.buglabs.bug.base.bb: Feb 17 15:48:23 com.buglabs.bug.base: add support for bug 1.x Feb 17 15:48:23 Without that fix, it fails to fetch the now gone hirop branch Feb 17 15:48:23 Signed-off-by: Denis 'GNUtoo' Carikli Feb 17 15:48:24 Acked-by: Stefan Schmidt Feb 17 15:51:57 morning kergoth Feb 17 15:56:25 hi pb_ Feb 17 16:03:52 hi, could someone teach me how to solve this error? ERROR: QA Issue with db-bin : http://pastebin.com/B9mDq8ta Feb 17 16:04:12 I'm reading insane.bbclass Feb 17 16:04:23 but I am a C/C++ programmer Feb 17 16:04:57 I don't understand what package_qa_check_rpath is trying to do there Feb 17 16:05:20 mckoan: well, Tartarus just posted some patches for this on the ml... Feb 17 16:07:09 hi mickey|office Feb 17 16:14:09 Tartarus: looks like patchwork is no longer used, how do I get your patch to test it before ack-ing? Feb 17 16:18:12 hmmm Feb 17 16:18:35 usually people do that: Feb 17 16:18:45 DEFAULT_PREFERENCE = "-1" Feb 17 16:18:56 DEFAULT_PREFERENCE_foodistro = "1" Feb 17 16:19:14 is that on purpose because here's the side effect Feb 17 16:19:24 let's say you have version 4.75 with that Feb 17 16:19:34 then you remove DEFAULT_PREFERENCE = "-1" of version 4.85 Feb 17 16:19:44 and 4.75 get pinned the same for foodistro Feb 17 16:19:51 because the default preference is 0 Feb 17 16:19:59 4.85 get 0 as preference Feb 17 16:20:04 4.75 get 1 Feb 17 16:20:12 and 4.75 get chosen Feb 17 16:36:03 GNUtoo|laptop: yes, that's the expected behaviour. if you don't want that then you should set DEFAULT_PREFERENCE_foodistro = "0". Feb 17 16:36:59 pb_, I know it's expected Feb 17 16:37:11 I just point out that people do that practice blindly Feb 17 16:37:49 I'm not sure they really want the thing I described Feb 17 16:38:02 but just want to unlock a version Feb 17 16:45:54 is there a code translating kernel module dependencies into the package manager dependencies? Feb 17 16:46:42 filip i think its kernel.bbclass Feb 17 16:46:58 def parse_depmod(): Feb 17 16:47:08 im looking at it now cause it seems not to be working right now Feb 17 16:48:04 aaah, so I am not the only one Feb 17 16:48:29 and here I thought I broke something Feb 17 16:48:33 well, from what i can tell its only us two :) Feb 17 16:48:50 i upgraded my kernel package to 2.6.37, so i figured that did it Feb 17 16:49:00 but im not sure Feb 17 16:50:21 mckoan, patchwork moved Feb 17 16:50:33 http://patches.openembedded.org/patch/667/ Feb 17 16:51:57 but i'm gonna push them both in a minute, now :) Feb 17 16:53:09 these do, btw, fix up all the QA errors I could hit on DISTRO=kaeilos Feb 17 16:53:32 03Tom Rini  07org.openembedded.dev * r87a92185e8 10openembedded.git/recipes/db/ (11 files in 4 dirs): Feb 17 16:53:32 db: Bring in 5.1.19 from poky Feb 17 16:53:32 Based on e77207ccdf15f34651a086099ed4816712e3656b in poky, bring in Feb 17 16:53:32 5.1.19 (no RPATH issues when using libtool 2.2) and drop un-pinned Feb 17 16:53:32 4.3.29 and 5.0.21 as this is backwards compatible. Feb 17 16:53:33 Acked-by: Marco Cavallini Feb 17 16:53:34 Signed-off-by: Tom Rini Feb 17 16:53:40 03Tom Rini  07org.openembedded.dev * r280b1f1df8 10openembedded.git/recipes/gettext/gettext_0.18.bb: Feb 17 16:53:40 gettext: Pass --disable-rpath on target Feb 17 16:53:40 We disable rpath stuff as this casues problems with libtool 2.2 Feb 17 16:53:40 and 2.4 is fine. Feb 17 16:53:40 Acked-by: Marco Cavallini Feb 17 16:53:41 Signed-off-by: Tom Rini Feb 17 16:53:41 on: BBPKGS='minimal-image qt4e-demo-image qt4-x11-demo-image native-sdk-image nas-server-image console-image meta-toolchain' DISTRO=kaeilos MACHINE=at91sam9263ek bitbake Feb 17 16:55:29 Tartarus: many thanks, I would not be able to make a patch so complex Feb 17 16:56:09 Tartarus: I wonder why do not you also had with Angstrom Feb 17 16:56:11 I'm still strugling to work out why I can't include libnet-dev in my SDK. Has anyone else actually done it? Feb 17 16:56:49 Tartarus: does not occured Feb 17 16:56:54 geiseri: what are your findings? Feb 17 16:57:11 mckoan, 2010.x is libtool 2.4 Feb 17 16:57:18 I forget why my 2008.1 stuff failed, heh Feb 17 16:57:49 OK, i did see it on 2008.1 Feb 17 16:57:55 And why I don't see them on minimal is libtool 2.4 Feb 17 16:58:03 filip im not sure yet, i think its not calling the populate_packages_prepend, but i need to figure out how to trace the script... Feb 17 16:59:20 Tartarus: I testsed it last night with angstrom-2008 and it worked Feb 17 16:59:37 geiseri: I am using linux-jlime-jornada7xx_2.6.34+git Feb 17 16:59:45 geiseri: so the problem is common to both Feb 17 17:00:49 mine is 2.6.37 from kernel.org with my own bitbake script Feb 17 17:02:41 it looks like it is calculating shared libs and not the depmod... Feb 17 17:03:05 i wonder if its my recipe Feb 17 17:11:07 geiseri: I wouldn't be affected in such case Feb 17 17:11:13 evening Feb 17 17:11:47 Jay7: hi Feb 17 17:12:11 have a nice rest of the day Feb 17 17:19:21 geiseri: let me know if I can do anything to get you closer to a patch :) Feb 17 17:19:44 i just did an update, but im not sure if there were any changes... im still learning git. Feb 17 17:29:20 So, um Feb 17 17:29:27 sqlite3-native needs tcl-native Feb 17 17:30:23 seems like someone should turn off tcl support for sqlite3 native then Feb 17 17:30:24 NOTE: package lzma-native-4.65-r5.1: task SRC_DISTRIBUTECOMMAND: Failed Feb 17 17:30:51 fray, good point Feb 17 17:31:22 I could MAYBE see someone want it on a device, if they have tcl support on the device.. but on the native side that seems strange Feb 17 17:31:45 i'll poke that after i finish testing openssl 1.0.0d update, heh Feb 17 17:32:29 hummm Feb 17 17:34:39 gm all Feb 17 17:34:54 temp swings like 20 degrees in a day here Feb 17 17:35:32 otavio: can you try gcc-cross-sdk now Feb 17 17:35:41 otavio: with the latest gcc push I did Feb 17 17:35:47 khem: sure. Feb 17 17:35:53 fray, so, it looks like we can't just drop that, it's needed for making sqlite3.h Feb 17 17:36:00 But... what we have now is wrong for adding the de Feb 17 17:36:01 p Feb 17 17:36:09 otavio: with that I am able to build i586-generic/minimal gcc-cross-sdk successfully Feb 17 17:36:30 khem: testing Feb 17 17:39:41 Tartarus: pass --disable-tcl thru extra oeconf to sqlite3-native Feb 17 17:40:31 khem, maybe? That comments in git imply that it isn't an option Feb 17 17:43:03 Good evening everybody Feb 17 17:43:48 I'm new to oe and are playing around with a Toradex Colibri board Feb 17 17:44:07 I've been looking around the web for some starting points regarding this board, but information is quite spare Feb 17 17:44:22 Does anybody of you perhaps have some good hints for this one? Feb 17 17:45:12 fray, khem, yeah, no: Feb 17 17:45:18 SQLite does not require TCL to run, but a TCL installation is required Feb 17 17:45:19 by the makefiles. Feb 17 17:45:33 So we should pass --disable-tcl Feb 17 17:45:37 but we still need tcl-native to build Feb 17 17:45:56 Ahh I didn't realize tcl was used in the install process Feb 17 17:46:12 not just install but generating files Feb 17 17:46:43 ajb_oe: IMAGE_INSTALL_append = " libnet-dev" ? Feb 17 17:47:26 ajb_oe: and make sure that its the name of the package and not recipe Feb 17 17:48:18 khem: I'm building an SDK so I have added libnet-dev to the RDEPENDS_${PN} of my task-toolchain-target.bb Feb 17 17:48:18 Feb 17 17:50:12 khem: bitbake comlains that it can't satisfy that dependancy but not why Feb 17 17:50:29 AFAICT libnet should have a libnet-dev package Feb 17 17:54:29 Nobody? A simple "rtfm" would be reaction enough ;-) Feb 17 17:56:20 ajb_oe: do bitbake libnet Feb 17 17:56:29 ajb_oe: and then see in the ipk/ dir Feb 17 17:56:36 what the actualy name of ipk is Feb 17 17:56:42 thats what you have to use Feb 17 17:56:50 there may be lib renaming that might have happened Feb 17 17:57:01 03Tom Rini  07org.openembedded.dev * ra2c3af2d60 10openembedded.git/recipes/sqlite/ (sqlite3.inc sqlite3_3.6.23.1.bb sqlite3_3.7.5.bb): Feb 17 17:57:01 sqlite3-native: Fixup DEPENDS Feb 17 17:57:01 The way things were before we were missing tcl-native. In addition, Feb 17 17:57:01 we can clean up the recipes slightly since sqlite3.inc is where we Feb 17 17:57:01 set BBCLASSEXTEND = "native" Feb 17 17:57:01 Signed-off-by: Tom Rini Feb 17 17:57:19 lackS, I don't think any of us have heard of the board. Can you perhaps give us the connection to OE? Feb 17 17:57:44 There's a good deal of random embedded hardware knowledge int eh channel Feb 17 17:58:58 Tartarus: Strange that you didn't hear of it. On your success stories site, Toradex is listed with the colibri board for using oe Feb 17 17:59:06 heh Feb 17 18:00:42 khem: building Feb 17 18:02:07 khem: failed. Same error Feb 17 18:02:16 _s.so.1.backup; else true; fi && mv ./libgcc_s.so.1.tmp ./libgcc_s.so.1 && ln -s libgcc_s.so.1 ./libgcc_s.so Feb 17 18:02:19 | /home/otavio/hacking/embedded-linux/tmp/sysroots/x86_64-linux/usr/i586/i586-oe-linux/bin/ld: cannot find /lib/libc_nonshared.a Feb 17 18:02:22 | collect2: ld returned 1 exit status Feb 17 18:02:44 khem: I did a clean only on gcc-cross-sdk. Is required me to clean anything else? Feb 17 18:07:03 Some pre-defined machines seem to be running on a pxa270... maybe I'll start with these ones Feb 17 18:07:14 Thanks for your help anyways... :-) Feb 17 18:15:59 lackS: there is any Toradex baords support here Feb 17 18:16:45 khem: where do the ipkg files end up, I can find the pkgdata for it: ./staging-pkg/pkgdata/i686-oe-linux/runtime/libnet-dev Feb 17 18:16:50 NOTE: The checksums for '/home/koan/devel/sources/gettext-0.18.tar.gz' did not match. Feb 17 18:17:48 khem: ahh found it: ./staging-pkg/deploy/ipk/i686/libnet-dev_1.1.2.1-r4_i686.ipk Feb 17 18:18:00 so it seems the name is correct... Feb 17 18:25:13 khem, ping Feb 17 18:26:01 I needed to delete and re-download a fresh copy of gettext-0.18.tar.gz and gettext-0.18.tar.gz.md5 Feb 17 18:26:21 build continues... Feb 17 18:26:26 why does bitbake keep downloading binutils from the redhat CVS server? It's currently down for me so I can no longer build (although I'd of thought binutils should be cached?) Feb 17 18:26:34 supper's ready Feb 17 18:27:00 (and seconds out) ;-) Feb 17 18:30:10 ajb_oe, did you tell it to use a not released version of binutils? Feb 17 18:30:53 Tartarus: don't think so - although the initial bringup was done by a consultant - I'll check Feb 17 18:33:36 Tartarus: I can't see anything in the .conf to indicate special binutils, what would I be looking for to tell? Feb 17 18:33:57 BINUTILS Feb 17 18:34:00 in caps Feb 17 18:36:35 Nope, no mention of that in any of the configs Feb 17 18:36:47 what distro are you using? Feb 17 18:36:57 geiseri: I think that parse_depmod returns a proper dictionary Feb 17 18:37:22 geiseri: I noticed it when orinoco_cs.ko did not have orinoco.ko in dependencies, and: 'kernel/drivers/net/wireless/orinoco/orinoco_cs.ko': ['kernel/drivers/net/wireless/orinoco/orinoco.ko', Feb 17 18:37:31 geiseri: so I think this is something later Feb 17 18:38:03 Just what Phil setup for us: http://pastebin.com/WJy7Tk0z Feb 17 18:46:08 PREFERRED_BINUTILS_VERSION isn't set anywhere? Feb 17 18:48:34 03Tom Rini  07org.openembedded.dev * r72a088f7e1 10openembedded.git/recipes/gnutls/gnutls.inc: Feb 17 18:48:34 gnutls.inc: Add --disable-rpath Feb 17 18:48:34 Signed-off-by: Tom Rini Feb 17 18:48:37 03Tom Rini  07org.openembedded.dev * rba84fd99c2 10openembedded.git/ (12 files in 3 dirs): Feb 17 18:48:37 openssl: Update from 1.0.0c to 1.0.0d Feb 17 18:48:37 Security fixes only so just bumping everyone that pinned up. Feb 17 18:48:37 Signed-off-by: Tom Rini Feb 17 18:51:03 Tartarus: no, I guess it defaults to CVS then if nothing set? Feb 17 18:51:38 no Feb 17 18:51:50 what you pasted defaults to sane-toolchain.inc which sets 2.21 Feb 17 18:51:59 odd Feb 17 18:52:58 ahh, in the logs: Feb 17 18:53:01 NOTE: preferred version 2.20.1 of binutils-cross-sdk not available (for item binutils-cross-sdk) Feb 17 18:53:37 Well, there you go Feb 17 18:53:59 time to resync to HEAD? Feb 17 18:54:09 or just grab that one recipe, yes Feb 17 18:54:12 since it's a 3 liner Feb 17 18:55:20 geiseri: I think I have it tracked down Feb 17 19:03:33 filip, what did you find? Feb 17 19:03:42 mine is still rebuilding Feb 17 19:09:06 03Eric BENARD  07org.openembedded.dev * r2471584764 10openembedded.git/conf/bitbake.conf: (log message trimmed) Feb 17 19:09:07 bitbake.conf: export LC_ALL = "C" Feb 17 19:09:07 * thus the build system language won't create troubles during builds Feb 17 19:09:07 (example : perl is hacking some generated files' content but only Feb 17 19:09:07 supports english outputs). Feb 17 19:09:07 * this problem is only met when using bitbake master Feb 17 19:09:07 Signed-off-by: Eric Bénard Feb 17 19:10:01 hm.. Feb 17 19:10:13 ericben: was this Ack'ed by someone else? Feb 17 19:10:35 I'm just unsure about this Feb 17 19:10:44 That was me Feb 17 19:10:48 * Tartarus kicks pw-am Feb 17 19:10:50 ah, well Feb 17 19:10:56 And, poky has it too, via RP Feb 17 19:11:05 ok, np then Feb 17 19:11:23 * Tartarus kicks tbird to cc oe-devel Feb 17 19:11:53 So.... here's a big fixer to the RPATH issues Feb 17 19:12:04 tbird? Tim Bird? Feb 17 19:12:05 Back in 2007, poky dropped -rpath stuff from TARGET_LDFLAGS Feb 17 19:12:09 thunderbird, heh Feb 17 19:12:12 ah :)) Feb 17 19:12:35 Which is to say, just general having --sysroot in CC makes a bunch of that not needed Feb 17 19:12:47 hm Feb 17 19:12:48 Ah-ha, ok, CIA cut off my S-O-B line Feb 17 19:12:53 * Tartarus still cc's oe-devel Feb 17 19:12:56 that works with older libtool? Feb 17 19:13:15 geiseri: I've posted the description on the mailing list Feb 17 19:13:29 oh cool Feb 17 19:14:00 sh#t.. I need new monitor.. Feb 17 19:14:06 woglinde_, it's not libtool sysroot related Feb 17 19:14:11 dead pixels area is too large :( Feb 17 19:14:14 geiseri: http://thread.gmane.org/gmane.comp.handhelds.openembedded/42795 Feb 17 19:14:17 btw guys Feb 17 19:14:21 any thoughts on this? Feb 17 19:14:24 seems pretty serious Feb 17 19:14:26 woglinde, that said, I'm gonna kicka big build on kaeilos Feb 17 19:14:32 And open to suggestions of other stuff to build Feb 17 19:17:15 filip, i cannot see anything that should have changed it in the commits to kernel.bbclass lately that would have broken it in the last few days Feb 17 19:23:26 otavio: did you use minimal as distro or micro Feb 17 19:23:54 otavio: I would say build from scratch Feb 17 19:24:38 * Tartarus curses whoever didn't pull the glib-mkenums patch forward, just a bit Feb 17 19:24:45 * Tartarus tests putting it back Feb 17 19:25:12 Tartarus: now I am here Feb 17 19:26:57 khem: :-( Feb 17 19:26:57 khem: minimal, as you suggested Feb 17 19:26:57 khem: really? OK. Feb 17 19:27:02 filip: you should only compare file and not whole path Feb 17 19:27:11 khem: :-( Feb 17 19:27:13 khem: minimal, as you suggested Feb 17 19:27:16 khem: really? OK. Feb 17 19:27:22 gm khem Feb 17 19:27:24 otavio: well I was able to build gcc-cross-sdk on minimal Feb 17 19:27:34 so I dont know what the problem in your case could be Feb 17 19:27:43 woglinde_: hey Feb 17 19:28:03 khem, nm for the moment but i'll want your comments on something i'm pulling back from poky Feb 17 19:28:44 sure Feb 17 19:30:09 khem: I? Feb 17 19:30:14 khem: it's not my code :) Feb 17 19:30:42 khem: I have never contributed to kernel.bbclass Feb 17 19:30:57 there always beginning Feb 17 19:31:37 03Tom Rini  07org.openembedded.dev * rb3ad287d0f 10openembedded.git/recipes/glib-2.0/ (2 files in 2 dirs): Feb 17 19:31:37 glib-2.0 2.28.0: Bring back glib-mkenums-interpreter.patch Feb 17 19:31:37 Signed-off-by: Tom Rini Feb 17 19:31:51 khem: do you think stripping /lib/modules/version out of file in get_dependencies should do? Feb 17 19:32:07 khem: I don't like changing code I am not familiar with Feb 17 19:33:30 filip, can you pm me the patch? i can test :) Feb 17 19:33:42 geiseri: don't have one yet :) Feb 17 19:34:38 okay, just offering, im interested in _not_ debugging this... i'd rather get back to my development. Feb 17 19:35:12 geiseri: same here :P Feb 17 19:35:55 filip, heheh Feb 17 19:36:11 who commited the patch to strip the version out? Feb 17 19:36:16 * chouimat bangs head on desk Feb 17 19:36:29 * geiseri doesn't know how to do a svn blame with git Feb 17 19:36:40 git blame? Feb 17 19:37:12 Crofton|work: but on what :P? Feb 17 19:37:21 Crofton|work: last commit to kernel.bbclass was on jan 27 Feb 17 19:37:32 Crofton|work: I did a good build a couple of days ago Feb 17 19:37:36 yeah this worked last week Feb 17 19:38:18 * khem bangs desk on chouimat's head Feb 17 19:38:30 khem: hey! Feb 17 19:38:51 how is your head now Feb 17 19:39:06 feels lighter :) Feb 17 19:39:06 bleeding Feb 17 19:40:11 I am liking the build profile effort that yocto folks are doing Feb 17 19:40:25 they have like 40-core machines to build it seems Feb 17 19:40:40 Does build profile really matter on such giants Feb 17 19:40:40 *g* Feb 17 19:40:48 cool ... and I'm stuff with a 8 core one :( Feb 17 19:40:57 I have a p4 Feb 17 19:40:58 s/stuff/stuck Feb 17 19:41:30 i at least have a core2 Feb 17 19:41:41 I have a pentium m :P Feb 17 19:41:41 but its a laptop, so the drive and ram gets in the way Feb 17 19:42:41 filip, im looking in other classes that have changed lately... it has to be sucking something in that is causing it to strip the kernel version Feb 17 19:42:53 * chouimat is rebuilding is macpro ... Feb 17 19:42:57 geiseri: I have a patch, testing now Feb 17 19:43:15 in what class? Feb 17 19:43:43 kernel.bbclass. but it doesn't work yet :P Feb 17 19:44:09 there was a chance in utils.bbclass, from the 14th Feb 17 19:44:18 explode_deps Feb 17 19:44:27 and I still have no idea why it happened in the first place Feb 17 19:44:28 err 11th Feb 17 20:10:13 geiseri: sent the patch to the ML Feb 17 20:19:03 khem: starting a full build Feb 17 20:24:09 all, in a list of BBPKGS='minimal-image qt4e-demo-image qt4-x11-demo-image native-sdk-image nas-server-image console-image meta-toolchain' what else should I add to make sure some pretty big change doesn't break? Feb 17 20:27:05 angstrom-gnome-image? ;) Feb 17 20:27:35 should that work on kaeilos? Feb 17 20:27:43 I'll toss that into angstrom-2008.1 Feb 17 20:27:45 not sure :) Feb 17 20:31:22 OK, new branch, trini/simplify-target-flags Feb 17 20:31:31 Gonna toss a small subset of the matrix at that Feb 17 20:31:39 along w/ angstrom-gnome-image on angstrom Feb 17 20:32:06 Tartarus: is this gonna get rid of global opt flags Feb 17 20:32:24 nope, sec Feb 17 20:32:37 * Tartarus waits on cgit Feb 17 20:32:46 | tclsh ./tool/mksqlite3h.tcl . >sqlite3.h Feb 17 20:32:47 | /bin/sh: tclsh: command not found Feb 17 20:32:47 | make: *** [sqlite3.h] Error 127 Feb 17 20:32:52 otavio, fixed Feb 17 20:32:54 know issue with current oe? Feb 17 20:32:59 Tartarus: really? grr Feb 17 20:33:06 rebasing my tree Feb 17 20:35:33 hm did I checked the latest sqlite in? Feb 17 20:36:06 unrelated Feb 17 20:36:12 khem, can you get us some text on SCALE for the news section of the website? Feb 17 20:36:41 http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?id=a2c3af2d608b1b713018d688b00e03873a538993 is fix Feb 17 20:37:08 khem: http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?h=trini/simplify-target-flags Feb 17 20:39:32 gm Feb 17 20:39:39 gM? Feb 17 20:39:45 ~ugt Feb 17 20:39:46 from memory, ugt is Universal Greeting Time. Created in #mipslinux, it is a rule that states that whenever somebody enters an IRC channel it is always morning, and it is always late when the person leaves. The local time of any other people in the channel, including the greeter, is irrelevant. http://www.total-knowledge.com/~ilya/mips/ugt.html Feb 17 20:39:48 good morning Feb 17 20:40:01 nice one :) Feb 17 20:40:12 Crofton|work: ok Feb 17 20:40:45 I just archived all the old news and added cebit :) Feb 17 20:42:02 khem, assuming my mini (that's not so mini) build test works, would you ack that? Feb 17 20:43:32 Tartarus: that will break pre libtool 2.4 distros Feb 17 20:43:39 Tartarus: like angstrom 2008 Feb 17 20:43:49 khem, I don't think so Feb 17 20:43:53 i'm testing them Feb 17 20:43:55 but poky did that in 2007 Feb 17 20:44:08 Tartarus: poky is subset Feb 17 20:44:24 there could be a list of recipes that may fallout Feb 17 20:44:33 we'll see, yes Feb 17 20:44:37 i'm building a lotta stuff Feb 17 20:44:42 so if angstrom 2008 build the known targets Feb 17 20:44:49 then I will be ok with it Feb 17 20:44:54 k Feb 17 20:45:47 beagleboard-linuxtag2010-demo-image.bb Feb 17 20:45:55 try that one with angstrom 2008 Feb 17 20:46:46 heh Feb 17 20:46:59 i'll do that by hand Feb 17 20:47:11 got angstrom-gnome-image in the list already Feb 17 20:48:26 ok nice Feb 17 20:50:05 03Khem Raj  07org.openembedded.dev * r8610b9795e 10openembedded.git/contrib/testing/testscript.sh: (log message trimmed) Feb 17 20:50:05 contrib/testing/testscript.sh: Dont define BBMASK Feb 17 20:50:05 * BBMASK is set in bitbake.conf by default and it Feb 17 20:50:05 skips nonworking and obsolete recipes which is Feb 17 20:50:05 what we want emptying it here could cause redundant Feb 17 20:50:05 parse of the graveyards Feb 17 20:50:05 Signed-off-by: Khem Raj Feb 17 20:50:39 khem, thanks for the patch ^^ Feb 17 20:50:46 filip, cool what was it? Feb 17 20:55:57 what is responsible for uploading firmwares to the kernel ATM? Feb 17 20:56:12 geiseri: no idea :P Feb 17 20:56:24 =) Feb 17 20:57:47 ah, udev Feb 17 20:59:08 nice, it's just killing the machine Feb 17 21:10:29 hi Feb 17 21:10:41 hi ericben Feb 17 21:10:51 is anyone here aware of a recipe providing a tool for bluez4 pairing ? Feb 17 21:10:53 hi woglinde_ Feb 17 21:11:40 ericben: ui based? or do you just want to know how to pair Feb 17 21:11:40 the only thing I found is using python ... which is quite heavy Feb 17 21:11:51 eFfeM: command line based Feb 17 21:12:24 hm, let me look Feb 17 21:12:30 last time I did that it was with ... bluez 1 or 2 :) Feb 17 21:12:53 that was simple, now it seems eveyrthing has been done to get this action complicated Feb 17 21:14:16 florian ping? Feb 17 21:14:41 ericben: I seem to recall creating a file with the pin, but apparently I did not write down details Feb 17 21:15:16 eFfeM: that was with bluez3 I think Feb 17 21:15:28 now that's far more complicated Feb 17 21:15:44 Crofton, Crofton|work ping? Feb 17 21:15:51 you need a "simple-agent" (simple by its name, not by the way to build or run it ;-) Feb 17 21:16:09 pong Feb 17 21:16:20 btw trying to compile a dbus tool with oe's sdk fails, it seems our dbus headers are missing soomething Feb 17 21:16:44 ericben: could well be it has been a while since I did this Feb 17 21:17:03 eFfeM: don't worry, thanks foro searching Feb 17 21:19:12 did you get the info on SCALE? Feb 17 21:19:25 Crofton|work, ^^^ Feb 17 21:19:50 ericben: from http://www.gentoo.org/doc/en/bluetooth-guide.xml: Feb 17 21:19:51 You can also use passkey-agent (found in bluez-utils) from the command line. Feb 17 21:20:00 pin was indeed bluez3 Feb 17 21:20:54 eFfeM: I think this doc is for bluez3, there is no more hcid in bluez4 Feb 17 21:21:17 not yet Feb 17 21:22:39 ericben some more from the doc: Feb 17 21:22:40 Note: Beginning with >=bluez-libs-3.x and >=bluez-utils-3.x, pin helpers have been replaced by passkey agents. There are a few different graphical passkey agents available to help manage your PIN, such as bluez-gnome and kdebluetooth. You can also use passkey-agent (found in bluez-utils) from the command line. Feb 17 21:23:11 ok first I fix dbus headers in the sdk, then I manage to compile http://git.kernel.org/?p=bluetooth/bluez.git;a=blob_plain;f=test/agent.c;hb=HEAD then that should work Feb 17 21:23:23 doc is indeed for 3, but this note was added for 4 Feb 17 21:24:11 ERROR: Error Method already seen: git_drop_tag_prefix in' gitpkgv.bbclass' now in 'gitver.bbclass' Feb 17 21:24:12 eFfeM: unfortunatly, there is no more bluez-utils : http://www.bluez.org/download/ that was for v3 Feb 17 21:24:19 any thoughts on how to make this go away? Feb 17 21:24:50 ericben: ah ok Feb 17 21:26:28 eFfeM: I've spent several hours on bluez4 today ;) Feb 17 21:26:51 once I managed to get this stupid hci chipset to talk to my cpu Feb 17 21:28:01 ericben: then follow the suggestions from our openmoko friends (simple-agent): http://wiki.openmoko.org/wiki/Manually_using_Bluetooth#Pair Feb 17 21:28:05 btw what is the way to solve QA error like this : RROR: QA Issue with gnutls-xx: package gnutls-xx contains bad RPATH ? Feb 17 21:28:26 eFfeM: yes been there, simple-agent is the python script I would like to avoid Feb 17 21:28:38 and agent.c is the one I'm actually compiling Feb 17 21:28:48 ericben which libtool? Feb 17 21:28:53 ericben: not fully sure, on the RPATH, iirc this is often a libtool issue Feb 17 21:29:05 woglinde_: angstrom 2008.1, so I think 2.2 Feb 17 21:29:10 yes Feb 17 21:29:12 ericben: ah ok Feb 17 21:29:23 ericben watch my latest commits Feb 17 21:29:33 one or two sed's are needed Feb 17 21:29:47 depends if g++ is involved or not Feb 17 21:30:37 woglinde_: thanks, found the one for flac Feb 17 21:34:03 So, ericben Feb 17 21:34:11 gnutls RPATH can be fixed w/ --disable-rpath Feb 17 21:34:14 i forget if i fixed that or not Feb 17 21:34:23 no Feb 17 21:34:29 not in all cases Feb 17 21:34:29 But the other reason for the branch I just made is I think that it will fix 90% of the RPATH bits Feb 17 21:34:39 libiconv flac libsdl-mixer Feb 17 21:36:17 Tartarus: ok I'm trying woglinde_ 's way now, where --disable-rpath should be added ? Feb 17 21:36:31 EXTRA_OECONF Feb 17 21:39:25 ericben, top of tree? Feb 17 21:39:33 like i say, i think i pushed gnutls Feb 17 21:40:33 Tartarus: yes actual head Feb 17 21:41:10 I confirm there is your patch with --disable-rpath Feb 17 21:41:32 hm anyone knows if git can merge per directory? Feb 17 21:42:15 or if I can diff btw branches and dirs Feb 17 21:42:25 ok I 'm restarting from a fresh tmp as this one seems quite in a bad state Feb 17 21:42:50 odd, yeah Feb 17 21:42:53 woglinde_: what do you mean by diff btw branches and dirs ? Feb 17 21:42:54 I know it fixed it for me Feb 17 21:43:04 Tartarus: for angstrom 2008 ? Feb 17 21:43:23 ericben I want to diff the cypto/aes dir of two branches Feb 17 21:43:41 hm I could use grep Feb 17 21:43:46 thats an Idea Feb 17 21:44:04 git diff branch branch works Feb 17 21:44:35 woglinde_: on one branch you can do git diff otherbranch directory ? Feb 17 21:44:51 ah I will use checkout and copy Feb 17 21:44:54 thats easiest Feb 17 21:47:21 So, all Feb 17 21:47:33 1h12m or so until I update testing-release-2011.03 Feb 17 21:47:47 Anything you need/want in for that, please hurry or poke me to delay updating the branch Feb 17 21:51:42 hmm anyone know how gitver and gitpkgv realte? the classes? Feb 17 21:52:54 building oe from phone... lol Feb 17 21:53:48 Tartarus: Iplan to push 2 or 3 of my machines this weekend. Do you think that will make it in the release ? Feb 17 21:54:37 03Khem Raj  07master * ra44300e54c 10openembedded.git/recipes/libpng/libpng.inc: Feb 17 21:54:37 libpng.inc: Package library in libpng and provide libpng3 Feb 17 21:54:37 Signed-off-by: Khem Raj Feb 17 21:54:48 Tartarus: (not in tonight's one, but for the 2011.03) Feb 17 21:55:28 ericben, per my email to the ML the other day, I do not want to branch or tag any sooner than the 24th Feb 17 21:55:36 It may be later, but it will not be sooner Feb 17 22:00:22 hmmm, got an rpath error all the time, digging at it more.. Feb 17 22:00:52 Tartarus: 24th of februrary or march ? Feb 17 22:00:57 feb Feb 17 22:01:09 to make the release ? Feb 17 22:01:29 ok so I have to push my machines this weekend Feb 17 22:01:40 to be safe, yes Feb 17 22:01:58 for now I have to sleep :) bye Feb 17 22:10:00 FWIW I have setup a patchwork project for bitbake Feb 17 22:15:22 ok, actual -rpath problem in acl Feb 17 22:16:10 *g* Feb 17 22:23:42 mwester, around? Feb 17 22:23:57 I'm looking at your vfork-gunzip patch again Feb 17 22:24:44 I think it should be ok to use gnu gunzip, the man page says you can set the GZIP env var to pass options Feb 17 22:26:02 grg: is it on target Feb 17 22:26:06 or on build host Feb 17 22:26:49 khem, that patch is to vfork "busybox gunzip". For targets. Feb 17 22:26:59 ok Feb 17 22:27:09 busybox gunzip and gnu gunzip have slightly different semantics Feb 17 22:27:15 does busybox have gunzip app Feb 17 22:27:23 yeah Feb 17 22:27:42 gnu gunzip is more verbose, unless the --quiet option is given Feb 17 22:36:53 i swear gmail is getting slower by the day... Feb 17 22:37:19 testing a fix for acl and rpath now and waiting for that to finish so i can test one for libtiff Feb 17 22:38:49 grg: yeah gmail.com was using switch gears from company where my wife works and they stopped buying so slowdown :) Feb 17 22:38:54 good nite Feb 17 22:54:29 ok solved my problem with libtiff, gnutls, etc and external toolchain. Problem is that configure decided it wanted to pass -isystem=xxx/usr/include Feb 17 23:00:30 it would appear that you "really" don't want to specify sysroot path again manually. more details: http://gcc.gnu.org/ml/gcc-help/2008-09/msg00047.html Feb 17 23:02:27 so g++ found an included *.h, which included another .h, and I suspect because of that description g++ searched in a weird order and decided it couldn't find that 2nd file Feb 17 23:03:06 I just eliminated the -isystem so it relied checked sysroot for the correct file and it proceeded just fine Feb 17 23:05:23 khem: hello, i have a problem building shr-lite-image: shr-lite-image.bb, do_qa_staging No such file or directory: '/home/nschle85/shr-build/shr-unstable/tmp/sysroots/armv4t-oe-linux-gnueabi/usr/lib/libpng.la' Feb 17 23:48:39 grg: ping. Feb 17 23:48:44 yo Feb 17 23:49:17 You were looking for me earlier? Feb 17 23:49:34 i have a revised opkg-vfork-gunzip patch Feb 17 23:49:39 let me find a pastebin... Feb 17 23:49:43 :) Feb 17 23:50:16 mwester, http://pastebin.com/V0jH3SQH Feb 17 23:50:31 mwester: for me linux-ixp4xx_2.6.37.bb was selected for nslu2be/slugos/slugos-image Feb 17 23:50:35 mwester: is that intended Feb 17 23:51:31 I'm hoping setenv("GZIP", "--quiet", 0); will do what's needed for gnu gunzip Feb 17 23:51:41 seems to work Feb 17 23:51:43 grg: cool solution. I like it! Feb 17 23:52:23 khem: yes, that's the new kernel for SlugOS. I hope it's new enough ;) Feb 17 23:52:31 btw, where are you setting OPKG_USE_VFORK? Feb 17 23:52:37 mwester: hmmm ok Feb 17 23:52:56 mwester: if I see log it failed to fetch but that might be a blip in my internet connection Feb 17 23:53:10 because it built fine for ucslugc Feb 17 23:53:17 grg: The idea is that affected users will set it themselves, either in their bash env, or just "OPKG_USE_VFORK=1 opkg install " Feb 17 23:53:43 khem: I tested LE, BE, and BE ucslugc with it. Feb 17 23:53:53 mwester: cool Feb 17 23:54:12 mwester: Did xinetd issue got solved Feb 17 23:54:13 (But not exhaustively - there may be some problems with some USB devices -- I have only a limited set of them) Feb 17 23:54:42 khem: No. I'm travelling again, and haven't been able to post a patch to the ML. Feb 17 23:54:52 ok Feb 17 23:55:01 now xinetd is fetchable I can take a look Feb 17 23:55:09 mwester: was the issue with BE only ? Feb 17 23:55:14 Oh yes. Feb 17 23:55:35 hmm ok Feb 17 23:55:38 let me see Feb 17 23:55:55 config.sub does a huge lookup, and can't match the be-oe part of the string for the toolchain, etc. Feb 17 23:56:08 ah Feb 17 23:56:14 er, I mean the armeb-oe part. Feb 17 23:56:33 It will find armeb, but apparently we added the -oe at some point, not too far back? Feb 18 00:06:53 mwester, i've committed that patch. So be aware that the next opkg SRCREV bump will require removing the slugos patch. Feb 18 00:07:21 I'm looking forward to it. :) And then we can delete the patch and the old hack recipe. Feb 18 00:07:34 feel free to bump then Feb 18 00:08:12 * Tartarus kicks acl Feb 18 00:08:29 * Tartarus hopes he sees how other distros are avoiding rpath, hopefully... Feb 18 00:15:54 Tartarus: debian uses, sometimes, chrpath Feb 18 00:17:46 not anymore Feb 18 00:17:52 that was related to doing --enable-lib64 Feb 18 00:21:40 Hi all I want to use the headers of liboil-0.3.17-r1 which are in " .... sysroots/armv7a-none-linux-gnueabi/usr/include/liboil-0.3/ " for compiling my gstreamer.bb file and i added the following to my recipe DEPENDS += " virtual/kernel liboil " and EXTRA_OECONF += " --with-liboil=${STAGING_DIR_TARGET}/usr/include/liboil-0.3 " but still my recipe cannot find the headers while compiling . Can someone help me Feb 18 00:23:15 i was able to compile the same when i add the CFLAGS_append = " ${STAGING_DIR_TARGET}/usr/include/liboil-0.3 " Feb 18 00:24:16 khem : can you suggest what i am missing Feb 18 01:07:44 Tartarus: I'll run my testbuilder now on testing-release branch Feb 18 01:10:09 * Jay7 -> sleep Feb 18 01:10:58 WARNING: oestats: exception , value HTTP Error 500: Internal Server Error Feb 18 01:11:00 :( Feb 18 01:12:12 Jay7, you broke it! Feb 18 01:12:31 * Jay7 hides Feb 18 02:05:44 * otavio has created a new development chroot to test if it fixes gcc-cross-sdk build problem Feb 18 02:11:50 OK, got tiff and acl fixed locally Feb 18 02:11:58 testing w/ the bigger fix to fix a the rest I hope Feb 18 02:13:13 Tartarus: hey did you see my comment earlier? solved another problem with external toolchain. major choke when it passes -isystem Feb 18 02:14:38 if you use sysroot, lots of c++ stuff tends to die if you also include system paths manually on the commandline. fix was to strip all the isystem lines that were duplicates of sysroot/usr/include Feb 18 02:14:56 Ah good Feb 18 02:15:03 I'm killing the -isystem now in my branch Feb 18 02:15:35 ya that took forever to solve. I thought I was chasing down a libtool or sysroot general problem. turned out the header says #include_next which means "search the rest of the path" in such a way that the manual isystem was a major problem Feb 18 02:15:47 almost surprised it didn't come up as an issue normally but I guess the sysroot isnt as critical normally Feb 18 02:16:38 I'll have more info as I investigate further tomorrow, but ya killing isystem is a safe start. Main offenders are anything with c++ especially gnutls (has an example cpp file), libtiff, forget what else Feb 18 02:17:56 basically through a huge chain of includes it goes looking for wchar.h, and based on how includes are processed it decided to skip usr/include where that header actually lived Feb 18 02:22:45 isystem should not be needed with proper sysroot now Feb 18 02:27:46 ERROR: QA Issue with libxml-parser-perl: package libxml-parser-perl contains bad RPATH /home/balister/oe/tmp/sysroots/armv7a-angstrom-linux-gnueabi/usr/lib in file /home/balister/oe/tmp/work/armv7a-angstrom-linux-gnueabi/libxml-parser-perl-2.36-r4/packages-split/libxml-parser-perl/usr/lib/perl5/auto/XML/Parser/Expat/Expat.so Feb 18 02:27:52 have you guys run across this one Feb 18 02:29:05 key Jefro Feb 18 02:29:12 congrats on the new job! **** ENDING LOGGING AT Fri Feb 18 02:59:57 2011