**** BEGIN LOGGING AT Tue Mar 09 02:59:58 2010 Mar 09 04:57:07 DJWillis: ping? Mar 09 07:08:36 hrw|gone: Laibsch mickey|zzZZzz: pushed the sanity branch yesterday evening, feel free to have a look at it Mar 09 07:14:41 (& others ofc too, & gm ) Mar 09 07:17:49 good morning Mar 09 07:37:49 siji: XorA told that the feed is fixed so after an opkg update you should be ok Mar 09 07:38:01 hi Mar 09 07:38:10 yes am proceeding with it now Mar 09 07:38:16 just restarted Mar 09 07:38:26 till now everything is o Mar 09 07:38:27 ok Mar 09 07:38:30 thanks :) Mar 09 07:38:35 DJWillis: now? Mar 09 07:38:42 siji: great, have fun Mar 09 07:39:05 DJWillis: I made a fresh rebuild and I'm way past git-native, and in the the log you pasted I can't seem to find the error message :} Mar 09 07:45:41 03Petr Štetiar  07org.openembedded.dev * r396964ae64 10openembedded.git/recipes/angstrom/ (angstrom-led-config.bb angstrom-led-config/ts72xx/leds): Mar 09 07:45:41 angstrom-led-config: Add LED setup for the ts72xx and bump PR Mar 09 07:45:41 Signed-off-by: Petr Štetiar Mar 09 07:45:41 Signed-off-by: Koen Kooi Mar 09 07:50:15 hi everyone Mar 09 07:53:06 my first post here. i am getting the following error http://pastebin.com/4Zb2gGum when i run bitbake console-image for my omap board. Any ideas?? Mar 09 08:07:18 any idea on what went wrong?? i been not able to use my OE for a very long time becos of this Mar 09 08:08:06 hgs, not sure, that is definitely a strange message Mar 09 08:09:08 a wrong python version? Mar 09 08:14:20 XorA: thanks. which python version is required ? i have 2.5 and 2.6. Mar 09 08:14:41 2.6 is preferred but it should work with 2.5 Mar 09 08:14:44 I beleive Mar 09 08:15:48 hgs: you will need to add print to the stage manager and check its input Mar 09 08:18:28 hm, somebody knows how I tell bitbake not to unpack a jar-file I included in the SRC_URI of the recipe? Mar 09 08:22:49 good morning Mar 09 08:23:19 XorA: i tried selecting python version to use by setting #!/usr/bin/env python2.6 and #!/usr/bin/env python2.5. both didnt help to solve the issue Mar 09 08:25:32 zecke: iam newbie to the build systems, so dont be hard on me..any way i will try to figure out what you meant Mar 09 08:26:54 XorA: forgot to mention that i made the selection of python version in the "bitbake" script Mar 09 08:28:16 also iam using bitbake version 1.8.12 Mar 09 08:30:33 hgs: I'm using bitbake/1.8 bitbake/master on one host with python2.5 without any problem.. Mar 09 08:32:54 hgs: but I've seen similar error when I had some broken line in cache Mar 09 08:37:53 JaMa: could you provide me details on how to fix it? Mar 09 08:38:34 hgs: first you can try without tmp dir.. to make sure it's the same issue Mar 09 08:39:34 and then I don't remember exactly what was wrong, but with a bit of help from strace it was easy to find which file was read before this issue and when I removed single irregularity there.. it started to work as before Mar 09 08:40:00 hgs: that has nothing to do with hard. Mar 09 08:42:19 zecke: :) Mar 09 09:11:08 hm, puzzled, I get a bitbake AttributeError: 'NoneType' object has no attribute 'startswith' when I change distro from angstrom-2008.1 to something else Mar 09 09:11:16 see http://www.pastebin.ca/1829655 Mar 09 09:11:31 anyone an idea what is wrong, i didn't change anything else Mar 09 09:11:59 guess some var is missing Mar 09 09:18:06 eFfeM, somebody assume a string and does "foo".startswith("fo") but foo is not a string but None. The backtrace should tell you which var is None instread of a string Mar 09 09:31:24 good morning Mar 09 09:31:46 blindvt` seems it wants to construct a path to bitbake.conf Mar 09 09:31:52 hi florian Mar 09 09:31:55 see the pastebin Mar 09 09:32:48 morning Mar 09 09:32:52 hi hwr Mar 09 09:32:54 hrw Mar 09 09:33:10 and the only change I made is DISTRO (scratch head) Mar 09 09:35:27 eFfeM: I have few changes which can interest you Mar 09 09:36:32 eFfeM: http://marcin.juszkiewicz.com.pl/download/diffs/oe/ Mar 09 09:43:05 hrw, looks ok, but don't have time right now Mar 09 09:43:20 and can't get things to build here see above Mar 09 09:44:27 morphis: jar == zip and bitbake knows how to unpack jar files Mar 09 09:44:51 hrw: jepp, but I don't want that bitake unpack this file! Mar 09 09:45:29 I am doing this now by append *.nounpack to the file and copy it with a do_unpack_append into the right dir Mar 09 09:45:38 but thats hackisch Mar 09 09:46:02 a SRC_URI = "file://test-file.jar;nounpack" would be nice Mar 09 09:57:43 morphis: I did osgi.jar.nounpack Mar 09 09:58:21 Hi, how can I commit "Machine configs" files in dev branch? Mar 09 09:58:59 Whether review is required or not? Mar 09 10:01:17 03Koen Kooi  07org.openembedded.dev * rc7bf24bc24 10openembedded.git/recipes/webkit/ (3 files in 2 dirs): webkit-gtk: bump SRCREV Mar 09 10:01:26 03Koen Kooi  07org.openembedded.dev * r50f9577409 10openembedded.git/recipes/gnome/libsoup-2.4_2.29.91.bb: libsoup-2.4: add 2.29.91 Mar 09 10:01:26 03Koen Kooi  07org.openembedded.dev * r108b984b95 10openembedded.git/recipes/powervr-drivers/kernel-module-bc_git.bb: kernel-module-bc: recipe to build WTBU bc_cat.ko Mar 09 10:01:27 03Koen Kooi  07org.openembedded.dev * r69339bac7c 10openembedded.git/recipes/hal/consolekit_0.4.1.bb: consolekit 0.4.1: sync packaging with 0.3.x Mar 09 10:01:50 bala_holems: push your patches (as git formatted patches) to the mailing list for review and queueing if you do not have commit. Mar 09 10:04:19 In Commit Policy 4th points said " Machine configs are less sensitive but machine maintainers should be consulted where present and known". Whether review required or not? Mar 09 10:04:29 bala_holems: send to ML Mar 09 10:04:43 bala_holems: unless you have r/w then commit new machine Mar 09 10:06:41 hrw: How can I get r/w permission? Mar 09 10:07:09 you can buy it, it costs 500EUR/month Mar 09 10:07:14 as in other projects - submit code/patches etc. show that you are good dev etc Mar 09 10:07:24 hrw: ML means commit request mailing list?... Mar 09 10:07:51 ~curse subversion Mar 09 10:07:52 May the fleas of a thousand camels infest your most sensitive regions, subversion ! Mar 09 10:09:19 bala_holems: ML means OEML means OpenEmbedded devel mailing list Mar 09 10:11:03 eFfeM: http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?h=eFfeM/sanity&id=b8af3ee2981b54212d8a0d0ea3b5755125a337ca can be done by using BPN instead of PN probably Mar 09 10:11:51 hrw, ah ok didn't know about BPN Mar 09 10:12:07 would change but somehow can' Mar 09 10:12:20 eFfeM: checksums patch reverted bison recipes Mar 09 10:12:23 t get things to setup on this system Mar 09 10:12:59 hrw, will look for that one too, might have picked up a git mv or so when making the patch Mar 09 10:13:02 eFfeM: and IIRC in few places where you removed all "older" versions and only git recipe stayed, then with removed D_P it went backwards as SRCPV was pointing to some older revision than latest removed release Mar 09 10:13:29 yeah, SRCPV is an issue of attentiontoo Mar 09 10:13:37 and SRCREV Mar 09 10:13:50 hrw : ok, I like to add new machine in OE, so whether review required? I have seen in Commit Policy 4th points said " Machine configs are less sensitive but machine maintainers should be consulted where present and known". Mar 09 10:13:54 s/SRCPV/SRCREV/ Mar 09 10:13:57 :) Mar 09 10:14:13 :-) Mar 09 10:14:48 bala_holems: one question: do you have r/w access to OE? Mar 09 10:15:44 bala_holems: if yes then push else follow ML review scheme Mar 09 10:15:45 fi Mar 09 10:16:58 hrw : I don't have r/w access. If I need to get, what procedure? Mar 09 10:18:32 bala_holems: we already told you what is the procedure Mar 09 10:18:57 morning Mar 09 10:20:10 hrw : sorry, what I'm asking, to get commit rights procedure? Mar 09 10:21:59 bala_holems: so far you did not shown any code (as for me you are "Unnamed developer") so the only one procedure for your code is 'how to send code for review' procedure. Mar 09 10:22:33 bala_holems: 'how to get write access' procedure is open for those which submitted code by first step procedure Mar 09 10:31:10 same here, I also want to contribute to OE Mar 09 10:33:40 I hope that what I said was clear Mar 09 10:34:17 yes Mar 09 10:34:34 I just express my interest that's all :) Mar 09 10:34:45 ok Mar 09 10:35:04 hrw: I have already sent my new machine releated patches to ML, but it is missed. Mar 09 10:35:18 hrw: Correct me if I am wrong, OEML reviews the submitted patches and forward the reviewed patches to commit request mailing list. Mar 09 10:36:53 bala_holems: patches lands on oeml and in patchwork.openembedded.org then devs look and push or comment Mar 09 10:37:11 bala_holems: and you prefer to be unnamed developer or have some name? Mar 09 10:37:28 I am asking because /whois bala_holems does not give that info Mar 09 10:37:52 florian: good morning Mar 09 10:38:29 hey pb_ Mar 09 10:38:40 hi cbrake Mar 09 10:38:49 bala_holems: there is no "commit request mailing list". if you send your patches to openembedded-devel then someone ought to check them in. if that doesn't happen, you might need to send a reminder after a while. Mar 09 10:39:21 hrw : I am balakrishnan, next time I will register properly Mar 09 10:40:10 regulus pxa270 machine? Mar 09 10:40:23 ~curse @ifgnumake@ macro Mar 09 10:40:25 May the fleas of a thousand camels infest your most sensitive regions, @ifgnumake@ macro ! Mar 09 10:41:13 pb_ : what about "openembedded-commits@lists.openembedded.org"? Mar 09 10:41:46 hrw: yeah. correct Mar 09 10:41:54 bala_holems: that is for notifications, not requests Mar 09 10:42:50 hrw : last time what I have sent is missed, So I am going to resent to ML Mar 09 10:43:00 pb_ : ok, thanks Mar 09 10:52:29 i would have assumed to find novfp, thumb and interwork in MACHINE_FEATURES Mar 09 10:53:34 those are more distro features than machine. I think we did have a conversation at one point about allowing MACHINEs to veto the presence of certain distro features, and I thought that this actually got implemented, but maybe I am imagining that last bit. Mar 09 11:05:51 hi. are there any plans for a new stable release? Mar 09 11:06:06 or is it possible to use "ovelay classes" like "overlay recipes"? Mar 09 11:06:37 I think we will have a new stable branch fairly soon. Mar 09 11:48:00 9 files changed, 123 insertions(+), 503 deletions(-) Mar 09 11:48:11 thats just e2fsprogs change Mar 09 12:00:13 gm Mar 09 12:04:44 lib/bb/utils.py: "The lines resulting into this error were:" is that english? "The lines leading to this error" or "causing that error" Mar 09 12:05:43 probably "to result in" was meant Mar 09 12:06:00 "The lines resulting in this error ..." Mar 09 12:07:29 ah, that's better. Thanks Mar 09 12:07:30 blindvt`: your ideas are indeed better, the lines were probably written by developers not having English as their native tongue. Mar 09 12:08:16 blindvt`: you from .at, so non-English as well? :-) Mar 09 12:08:30 likewise, yep, i'm not a native speaker either Mar 09 12:09:45 likewise, but my parser is as picky as bitbakes ;) Mar 09 12:15:27 :-) Mar 09 12:26:27 def nope(): return ("foo","bar") Mar 09 12:26:28 XorA, Y = ${@nope()} Mar 09 12:26:40 is not supposed to work, is it? Mar 09 12:27:08 s/orA//;# darn completion, sorry Mar 09 12:35:27 Hi All Mar 09 12:35:52 I have build Clutter 1.0 now Mar 09 12:36:03 but still getting errors for 0.8 Mar 09 12:36:19 Collected errors: Mar 09 12:36:20 * Package libclutter-eglx-0.8-dev md5sum mismatch. Either the opkg or the package index are corrupt. Try 'opkg update' Mar 09 12:42:20 any idea why this error cming? Mar 09 12:51:30 re Mar 09 12:52:52 hrw, for me ?? Mar 09 12:58:02 :-) Mar 09 13:12:16 blindvt: hehe Mar 09 14:08:38 hi All Mar 09 14:08:53 how to build cluttter with eglnative backend Mar 09 14:09:29 I have modified clutter.inc (beagleboard) Mar 09 14:09:40 but not generated Mar 09 14:28:49 hi, anyone have compiled mono-2.4.2.3 for amrv5te? I need libmono-corlib but at the end of the compilation I don't have libraries Mar 09 15:04:09 03Sebastian Spaeth  07org.openembedded.dev * rdd28281a55 10openembedded.git/conf/distro/include/sane-srcrevs-fso.inc: Mar 09 15:04:09 sane-srcrevs-fso: bump frameworkd, previous rev turned out faulty. Mar 09 15:04:09 Signed-off-by: Sebastian Spaeth Mar 09 15:08:46 03Denys Dmytriyenko  07org.openembedded.dev * r64a9355721 10openembedded.git/ (2 files in 2 dirs): linux-davinci: update to the latest staging snapshot Mar 09 15:21:05 mhm, can someone give me a hint with a git src uri... I have a repo on the intranet, I setup the src uri with the url that I usually use to clone something like "git://path/repo;protocol=git" Mar 09 15:21:13 it gets the sources, creates a tar Mar 09 15:21:37 and then in the end tries to do a "git read-tree 1" for some reason (whatever that is) Mar 09 15:21:39 and this is failing Mar 09 15:21:42 any ideas? Mar 09 15:22:16 Jin^eLD: you forgot to set the SRCREV Mar 09 15:22:44 03Holger Hans Peter Freyther  07org.openembedded.dev * rc495ba7d82 10openembedded.git/ (9 files in 4 dirs): Mar 09 15:22:44 opensll: Upgrade to version 0.9.8m Mar 09 15:22:44 * Upgrade both openssl and openssl-native to 0.9.8m Mar 09 15:22:44 * Upgrade the Makefile and Config changes Mar 09 15:22:44 * Switch to checksums inside the recipe Mar 09 15:23:20 oh..thanks Mar 09 15:33:17 * Tartarus tries, <3s git stash Mar 09 15:36:36 i like that you can name the stash when you save it.. i often forget to do it, but can be handy trying to make sense of the stash pile later on :) Mar 09 15:36:38 * kergoth_ yawns Mar 09 15:40:09 oh bite me, cisco Mar 09 15:40:52 ~curse cgit/gitosis/git Mar 09 15:40:53 kergoth_, could you please rephrase the sentence with "resulting into" in lib/bb/utils.py ? "leading to" or "resulting in" come to my mind, but i'm not a native speaker Mar 09 15:40:53 May the fleas of a thousand camels infest your most sensitive regions, cgit/gitosis/git ! Mar 09 15:41:08 kergoth_: also, messages for the stash Mar 09 15:41:12 blindvt`: okay, will do Mar 09 15:41:27 I have local copies of several git trees and would like to use cgit for them. but I failed Mar 09 16:18:07 urg Mar 09 16:18:12 tired of vacationing Mar 09 16:19:47 heh Mar 09 16:20:50 hi Crofton Mar 09 16:20:58 hi Mar 09 16:22:04 4 days of skiing is kicking my rear Mar 09 16:25:29 hi gtk+ seem not to build for some time now...(maybe a week) http://pastebin.com/tvta2HMg Mar 09 16:25:52 failed to load "./apple-red.png": Couldn't recognize the image file format for file './apple-red.png' Mar 09 16:26:04 I'll look file './apple-red.png' Mar 09 16:26:14 GNUtoo|oeee: XorA sent patch for it on ML Mar 09 16:26:33 ah ok,thanks a lot Mar 09 16:26:37 * JaMaWrk confirmed it, but nobody pushed it Mar 09 16:27:23 ok nice,my psu had bad caps,and I couldn't find the same replacement caps so I have only my eeepc under angstrom which has only the following mail clients: Mar 09 16:27:32 *claws mails Mar 09 16:27:40 *mutt Mar 09 16:28:00 so following mails with claw mails is harder as delete is too slow Mar 09 16:30:57 http://home.haerwu.biz/openembedded/log/?h=org.openembedded.dev - my queue for pushing Mar 09 16:34:08 morning kergoth Mar 09 16:45:10 JaMaWrk: you may push it :-D Mar 09 16:46:54 XorA: but than someone can blame for that gdk*-native removal.. :) Mar 09 16:47:41 I don't use gtk much so I know nothing about it.. so I was hoping that someone who "broke" it, will be willing to fix it Mar 09 16:48:37 hey pb_ Mar 09 16:56:21 JaMaWrk: its not a git-am patch so you can edit it :-) Mar 09 17:05:09 uh frist patches for tegra Mar 09 17:32:40 03Roman I Khimov  07org.openembedded.dev * r3d18f7dff9 10openembedded.git/recipes/qemu/ (14 files in 2 dirs): (log message trimmed) Mar 09 17:32:40 qemu: add version 0.12.3 Mar 09 17:32:40 qemu-native tested to compile on openSUSE 11.2 and Debian Lenny. Binary Mar 09 17:32:40 locale generation tested for qemuarm/glibc, simpad/glibc, Mar 09 17:32:40 x86-prescott/eglibc, efika/glibc. Mar 09 17:32:41 Signed-off-by: Roman I Khimov Mar 09 17:32:42 Acked-by: Tom Rini Mar 09 17:32:42 03Roman I Khimov  07org.openembedded.dev * r44ecf67474 10openembedded.git/classes/sanity.bbclass: (log message trimmed) Mar 09 17:32:43 sanity: remove /proc/sys/vm/mmap_min_addr check Mar 09 17:32:43 Binary locale generation fails on openSUSE 11.2 and probably would fail Mar 09 17:32:44 on some other hosts with 2.6.31+ kernels since mmap-ing page zero may be Mar 09 17:32:44 forbidden for security reasons even with /proc/sys/vm/mmap_min_addr set Mar 09 17:32:45 to 0 (this also affects Ubuntu 9.10, although it's been fixed there: Mar 09 17:32:45 https://bugs.launchpad.net/ubuntu/+source/linux/+bug/423513 Mar 09 17:32:46 03Roman I Khimov  07org.openembedded.dev * r677630b31b 10openembedded.git/ (110 files in 6 dirs): (log message trimmed) Mar 09 17:32:47 qemu: remove 0.9.x Mar 09 17:32:53 yes Mar 09 17:33:52 pb_: ping Mar 09 17:34:35 ynezz: you here? Mar 09 17:35:06 hrw: yes Mar 09 17:35:38 ynezz: can you next time move ep93xx generic patches to ep93xx/ instead of ts72xx/ dir? Mar 09 17:36:21 03Marcin Juszkiewicz  07org.openembedded.dev * r5b2b133db7 10openembedded.git/recipes/lftp/lftp_3.5.1.bb: Mar 09 17:36:21 lftp: upgrade to 4.0.5 Mar 09 17:36:21 Signed-off-by: Marcin Juszkiewicz Mar 09 17:36:22 03Marcin Juszkiewicz  07org.openembedded.dev * r6b128e9173 10openembedded.git/recipes/iksemel/ (iksemel-1.4/r25.diff iksemel_1.4.bb): Mar 09 17:36:22 iksemel: make it build with GnuTLS 2.8.5 - from upstream SVN Mar 09 17:36:22 Signed-off-by: Marcin Juszkiewicz Mar 09 17:36:24 03Marcin Juszkiewicz  07org.openembedded.dev * rfda4472f28 10openembedded.git/recipes/gnome/ (gnome-vfs/03_gnutls27.patch gnome-vfs_2.24.1.bb): Mar 09 17:36:24 gnome-vfs 2.24.1: added patch from Debian to use new GnuTLS Mar 09 17:36:24 Signed-off-by: Marcin Juszkiewicz Mar 09 17:36:25 03Marcin Juszkiewicz  07org.openembedded.dev * r11beaab126 10openembedded.git/recipes/iksemel/iksemel_1.2.bb: iksemel 1.2: drop glibc dependency Mar 09 17:36:25 03Marcin Juszkiewicz  07org.openembedded.dev * r52d301bd10 10openembedded.git/recipes/loudmouth/ (2 files in 2 dirs): Mar 09 17:36:26 loudmouth: added patch from Debian to use new GnuTLS Mar 09 17:36:26 Signed-off-by: Marcin Juszkiewicz Mar 09 17:36:27 03Marcin Juszkiewicz  07org.openembedded.dev * ref73bcd899 10openembedded.git/recipes/e2fsprogs/ (9 files): (log message trimmed) Mar 09 17:36:28 e2fsprogs: move libblkid/libuuid/fsck to util-linux-ng + new staging/native Mar 09 17:36:29 Dropped libblkid, libuuid and fsck binary and use those which are in Mar 09 17:36:29 util-linux-ng. Mar 09 17:36:42 03Marcin Juszkiewicz  07org.openembedded.dev * rd1c51b5cd6 10openembedded.git/recipes/tasks/task-mamona-base.bb: task-mamona-base: adapt to e2fsprogs<>util-linux-ng changes Mar 09 17:36:42 03Marcin Juszkiewicz  07org.openembedded.dev * r67b7fffc17 10openembedded.git/recipes/tasks/task-openprotium.bb: task-openprotium: adapt to e2fsprogs/util-linux-ng changes Mar 09 17:36:42 03Marcin Juszkiewicz  07org.openembedded.dev * ra8ead97cbb 10openembedded.git/recipes/libgpg-error/libgpg-error.inc: Mar 09 17:36:42 libgpg-error: add native Mar 09 17:36:42 Signed-off-by: Marcin Juszkiewicz Mar 09 17:36:42 03Marcin Juszkiewicz  07org.openembedded.dev * r098f31a905 10openembedded.git/recipes/libgcrypt/libgcrypt.inc: Mar 09 17:36:43 libgcrypt: add native Mar 09 17:36:43 Signed-off-by: Marcin Juszkiewicz Mar 09 17:36:57 hrw: I would do that, but I don't have any other hardware with ep93xx to test it on, just ts72xx Mar 09 17:37:04 03Marcin Juszkiewicz  07org.openembedded.dev * r0d41b93b52 10openembedded.git/recipes/gnutls/ (12 files in 3 dirs): Mar 09 17:37:04 gnutls: drop old versions Mar 09 17:37:04 GnuTLS 2.7 moved from M4 macros + libgnutls-config to pkg-config. Mar 09 17:37:04 All software which required gnutls got fixed for this and is not Mar 09 17:37:05 buildable with older versions. Mar 09 17:37:05 Signed-off-by: Marcin Juszkiewicz Mar 09 17:37:06 03Petr Štetiar  07org.openembedded.dev * r601eb6c5ae 10openembedded.git/recipes/linux/ (18 files in 2 dirs): Mar 09 17:37:06 linux 2.6.33: add support for ts72xx Mar 09 17:37:07 Signed-off-by: Petr Štetiar Mar 09 17:37:07 03Marcin Juszkiewicz  07org.openembedded.dev * rf5bea767a3 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: Mar 09 17:37:08 Revert "angstrom: prefer gnutls 2.4.2, 2.8.5 induces too much breakage" Mar 09 17:37:23 ynezz: thats does not matter - I have sim.one and can test Mar 09 17:37:34 ynezz: sim.one has few other ep93xx patches Mar 09 17:37:40 hrw: ah, good then Mar 09 17:38:04 This reverts commit de69350beb7739b944c728980ca724e171492595. Mar 09 17:38:04 All recipes which depend on gnutls was fixed to build with 2.8.5 Mar 09 17:38:04 version. Mar 09 17:38:04 Signed-off-by: Marcin Juszkiewicz Mar 09 17:38:05 Acked-by: Koen Kooi Mar 09 17:38:53 hrw: and thanks for pushing :) Mar 09 17:39:05 np Mar 09 17:39:14 what's the status on gtk+ breakage? Mar 09 17:39:16 hrw: I dropped the kernel image from sim.one images yesterday. now a simple console image is close to fit into internal flash :) Mar 09 17:39:32 I wonder how many people will complain about stuff which I just pushed Mar 09 17:39:52 florian: I am booting from nfs iirc (long time not booted) Mar 09 17:40:06 I always test with nfs only Mar 09 17:40:12 florian: and can you move that kernel stuff to machine config? I want .33 on sim.one Mar 09 17:40:35 ynezz: any plans on merging ts72xx to mainline? Mar 09 17:41:13 hrw: yes this might be a good idea. This migh make maintennace a little bit easier. Mar 09 17:42:15 * hrw -> off Mar 09 17:43:01 hrw|gone: well I don't have time nor the skills yet to do that :) Mar 09 17:43:09 bye hrw Mar 09 17:43:38 hrw|gone: and thanks to others it's slowly getting into the mainline Mar 09 17:43:51 hi jama Mar 09 17:43:51 hi florian Mar 09 17:45:05 hi woglinde Mar 09 17:47:16 hi woglinde Mar 09 17:58:55 hi, I am trying to add alsamixer to my image, I added IMAGE_INSTALL += "alsa-utils" but it is not in the resulting filesystem, where am I wrong? Mar 09 18:07:22 hi laibsch Mar 09 18:26:25 there is no install phase in alsa-utils, is it normal? Mar 09 18:27:40 gm all Mar 09 18:27:57 woglinde Laibsch hello Mar 09 18:33:13 mckoan: you mean in recipe ? Mar 09 18:33:39 mckoan: then it should use the default do_intall task Mar 09 18:33:44 mckoan: alsa-utils generates few pacakges. Mar 09 18:33:58 mckoan: you probably want alsa-utils-alsamixer o sth like that Mar 09 18:35:02 mckoan: IMAGE_INSTALL lists packages, not recipes Mar 09 18:35:19 alsa-utils package is pretty much empty Mar 09 18:35:36 greetings! does anyone know how I can tell OE not to cache/archive the svn checkout for my recipe? Mar 09 18:37:03 oneshel: no such option Mar 09 18:37:19 or you can point SRC_URI to your sources Mar 09 18:38:18 hrw: ah ok, you're saying point SRC_URI to my separate local checkout? Mar 09 18:38:31 thats option Mar 09 18:38:55 hrw: I have two recipes that checkout subdirectories of each other that break each other Mar 09 18:43:05 03Koen Kooi  07org.openembedded.dev * r9fa5bd8ff0 10openembedded.git/recipes/powervr-drivers/libgles-omap3/rc.pvr: libgles-omap3: prepare initscript for ES5.x (DM37xx and OMAP36xx)3 Mar 09 19:02:27 afternoon :-) Mar 09 19:06:34 oneshel: break each other how? do you mean that the two tarballs collide in the DL_DIR? Mar 09 19:06:43 hrw|gone, denix, khem` : thx Mar 09 19:08:37 although alsa-uils says PACKAGES += "alsa-utils-alsamixer \ Mar 09 19:10:31 denix: alsa-utils package is pretty much empty, so what would be its purpose? Mar 09 19:12:35 hi,I'm looking into base files like /etc/passwd...I bet they are provided by a task in task-base or something like that Mar 09 19:13:43 maybe it's in a postinstall script Mar 09 19:13:44 mckoan: none. all the utils are packaged into own packages Mar 09 19:14:16 indeed Mar 09 19:14:18 base-passwd Mar 09 19:14:57 at least for micro Mar 09 19:15:23 ok got it Mar 09 19:17:42 denix: yep, I added alsa-utils-alsamixer and now is into the fs, thx Mar 09 19:17:47 mmm Mar 09 19:17:52 /etc/init.d/dbus-1 start => Starting system message bus: start-stop-daemon: unknown user messagebus Mar 09 19:17:58 messagebus is a group Mar 09 19:18:12 I'll look Mar 09 19:18:42 maybe chown Mar 09 19:19:03 /bin/chown -> busybox Mar 09 19:19:14 mmm Mar 09 19:19:33 and note that I do not use shadow anymore Mar 09 19:20:22 re Mar 09 19:21:08 ahhh Mar 09 19:21:09 ok Mar 09 19:21:23 not in /etc/passwd or /etc/groups after reinstall Mar 09 19:27:43 nice,hald now re-works Mar 09 19:29:42 basically I think I messed up my system with shadow Mar 09 19:46:16 hi woglinde, florian, mickeyl Mar 09 19:47:23 jo pb Mar 09 19:49:32 hey pb__ Mar 09 19:49:53 mickey|office: that's a good time to show up at the office :) Mar 09 19:50:59 hmm Mar 09 19:51:04 stupid iMac Mar 09 19:51:07 spurious wakeups Mar 09 19:51:08 he mickeyl Mar 09 19:51:10 i'm actually at home Mar 09 19:51:16 snowing Mar 09 19:52:48 mickeyl: heh... I se the same effect with my brother's mac and Skype :) Mar 09 20:01:47 snowing here too Mar 09 20:09:28 hi what good rss reader are there in oe? Mar 09 20:13:31 http://google.com/reader/ ? :) Mar 09 20:13:39 GNUtoo|oeee: since nobody answered, I have a suggestion marginally relevant: sage for firefox. I don't know if you use firefox and there's not a recipe for sage, but it is a rss reader. :-) Mar 09 20:14:05 * Tartarus wonders if any of the other pstage folks are about Mar 09 20:14:09 mario-goulart, I use firefox Mar 09 20:14:25 but I'd like a standalone app Mar 09 20:14:34 GNUtoo|oeee: ah, ok. Mar 09 20:14:42 like liferea Mar 09 20:14:46 but it's not in oe Mar 09 20:14:57 GNUtoo|oeee: really? I hated standalone readers Mar 09 20:15:08 But I'm weird about missing stuff because I was offline Mar 09 20:15:16 ok Mar 09 20:15:23 And some feeds are real spammy :) Mar 09 20:16:14 lol Mar 09 20:16:25 hi woglinde Mar 09 20:17:26 /ME/ME / cdfcdsf Mar 09 20:19:20 ~curse Ubuntu on Acer laptop Mar 09 20:19:21 May you be reincarnated as a Windows XP administrator, Ubuntu on Acer laptop ! Mar 09 20:19:23 mario-goulart, at the end did you commit some images for routers in oe? Mar 09 20:20:08 GNUtoo|oeee: no... I ended up switching to some higher priority projects. Mar 09 20:20:21 ~curse Telecom connection while it's snowing Mar 09 20:20:22 May the fleas of a thousand camels infest your most sensitive regions, Telecom connection while it's snowing ! Mar 09 20:20:27 lol Mar 09 20:20:32 (for snow) Mar 09 20:20:41 mario-goulart, ok Mar 09 20:20:49 GNUtoo|oeee: I lost connection Mar 09 20:20:52 03Martin Jansa  07org.openembedded.dev * r6be933980e 10openembedded.git/recipes/tasks/task-shr-feed.bb: Mar 09 20:20:52 task-shr-feed: add bootchart-lite Mar 09 20:20:52 Signed-off-by: Martin Jansa Mar 09 20:20:53 03Martin Jansa  07org.openembedded.dev * r09e89d01e7 10openembedded.git/recipes/tasks/task-shr-minimal.bb: Mar 09 20:20:53 task-shr-minimal: readd libx11-locale Mar 09 20:20:53 Signed-off-by: Martin Jansa Mar 09 20:20:54 ok Mar 09 20:21:21 ~hail Ubuntu on Acer laptop Mar 09 20:21:22 * ibot bows down to Ubuntu on Acer laptop and chants, "I'M NOT WORTHY!!" Mar 09 20:21:31 mckoan, I've connection problems recently,restarting the alice box fixes it(i use it only as a modem) Mar 09 20:22:04 maybe I should look if we own the alice box,and if so, install GNU/Linux on it Mar 09 20:22:09 like openwrt for instance Mar 09 20:22:17 GNUtoo|oeee: Telecom have too many bailout problems to focus on a better service Mar 09 20:22:18 because it could be the box Mar 09 20:22:24 ok Mar 09 20:23:03 I heard that some telecom focus on better service and assistance,they are not aimed at the general public but at companies Mar 09 20:23:04 GNUtoo|oeee: my Alice router + Linksys OpenWRT are an excellent couple Mar 09 20:23:34 I've alice router->gentoo on x86 box Mar 09 20:23:56 linksys is not powerfull enough for running all what I run Mar 09 20:24:00 GNUtoo|oeee: actually I faced to a problem with a broken router (business) they replaced afer 7 days Mar 09 20:24:09 ouch Mar 09 20:24:09 ok Mar 09 20:24:40 GNUtoo|oeee: in the meantime I bought a new Linksys router as spare, unfortunately wotked as dynamic IP Mar 09 20:24:53 ok Mar 09 20:25:01 GNUtoo|oeee: but it worked :-D Mar 09 20:25:47 ok Mar 09 20:26:07 I wanted to install openwrt *on* the alice pirelli box Mar 09 20:26:42 soldering serial etc... Mar 09 20:26:58 but I should look if the box is our proprety or alice's Mar 09 20:27:09 GNUtoo|oeee: it's not worty Mar 09 20:27:12 ok Mar 09 20:27:18 it would have been fun Mar 09 20:27:20 what's the best way to read an environment variable from within my .bb file? bb.data.getVar doesn't seem to be working for me Mar 09 20:27:51 but the problem would have been that if it failed=>no more internet for a while Mar 09 20:27:56 that's not acceptable Mar 09 20:27:58 Linksys stuff is the best choice for OpenWRT Mar 09 20:28:41 * mckoan wonders why OE wouldn't be the same as OpenWRT Mar 09 20:29:40 I've done (in collaboration with mario-goulart and some irc help)s tuff with wrt54gs and oe: Mar 09 20:30:05 init segfaulted,and also all dynamically linked programs Mar 09 20:30:09 GNUtoo|oeee: hmmm interesting Mar 09 20:30:34 we used uclibc, because of compiler version changes,some stuff in ld.so weren't inlined Mar 09 20:30:48 so everything segfaulted Mar 09 20:31:17 fortunately openwrt worked,that made debugging easier Mar 09 20:31:28 it was only a mather of chroot + gdb Mar 09 20:31:39 and NFS of course Mar 09 20:32:52 I wonder why OE couldn't support such devices, maybe there is not interest from developers Mar 09 20:33:20 we have the micro distribution Mar 09 20:33:24 so yes it could Mar 09 20:33:29 but I didn't leave NFS Mar 09 20:33:37 I hate to work with OpenWRT every since and then Mar 09 20:33:42 ok Mar 09 20:33:57 do you have serial cable for your router? Mar 09 20:34:14 I've soldered it,made serial voltage adapter on breadboard Mar 09 20:34:16 no Mar 09 20:34:18 ok Mar 09 20:34:30 I did everything from eth Mar 09 20:34:40 too bad Mar 09 20:34:51 because you've got access to pmon Mar 09 20:35:16 basically the electronic thing is easy and non intrusive Mar 09 20:35:25 as I soldered only some pins Mar 09 20:35:31 and used some computer cable Mar 09 20:35:39 like for instance the cdrom audio cable Mar 09 20:36:11 you used netconsole for the kernel messages? Mar 09 20:36:26 or you just installed openwrt and used it? Mar 09 20:36:37 the latter Mar 09 20:36:44 ok Mar 09 20:39:40 03Phil Blundell  07org.openembedded.dev * r0135ae31d1 10openembedded.git/recipes/gtk+/ (gtk+.inc gtk+_2.18.6.bb): Mar 09 20:39:40 gtk+: remove special-casing for native installation, declare provider for gdk-pixbuf-csource-native when building natively Mar 09 20:39:40 (based on original patch from Graeme Gregory with minor edits) Mar 09 20:41:02 wow gtk will build again Mar 09 20:41:03 pulling Mar 09 20:44:15 * mckoan is rebuilding from scratch to verify the latest pushes Mar 09 20:44:23 i need some real linux crack helping me a s2ram issue on a intel unit .. anyone know such guy ? Mar 09 20:44:35 yay! Mar 09 20:44:50 pb_: thanks! Mar 09 20:47:01 rob_w I fear you have to ask on the kernel-ml Mar 09 20:48:13 kk i feared that Mar 09 20:54:31 we need to port that toolchain into openembedded Mar 09 20:54:42 oops wrong channel Mar 09 21:00:34 gn8 Mar 09 21:05:57 oops, what's this? /usr/include/sys/types.h:100: error: two or more data types in declaration specifiers Mar 09 21:06:38 that's while building gcc-cross-initial-4.3.3 Mar 09 21:10:45 denix: Oh boy, you can make it fail too maybe? Yay :) Mar 09 21:11:30 what's that supposed to mean? :-P Mar 09 21:11:31 denix: After talking with kergoth more we've decided, I think, that something like that should be correct for the default BB_STAMP_POLICY (perfile) but that whitelist / full would be OK Mar 09 21:11:57 denix: I mean I haven't tested our big boxes with more than 4 threads but make -j6 Mar 09 21:12:15 Still need to find out if there's a bigger set of optimial values, for from scratch at leastr Mar 09 21:12:21 *least Mar 09 21:14:10 ok, let's try w/o -j6 :) Mar 09 21:34:56 Tartarus: hmm, disabling parallel build actually worked... Mar 09 21:36:22 Well, it's a race right? Mar 09 21:37:23 I'm a bit surprised that disabling PARALLEL_MAKE (as opposed to BB_N_T) would make a difference. It seems like it must be an inter-package race, not a race within a single build. Mar 09 21:37:55 well, unless gcc's makefiles are just busted. I guess that's possible, but it doesn't seem very likely. Mar 09 21:38:15 Well, denix's was outside of gcc i thought Mar 09 21:38:27 The gcc one is a thread race and stamp policy un-fun Mar 09 21:38:32 * Tartarus wonders if kergoth replied yet :) Mar 09 21:38:45 ah, maybe I misunderstood what denix was talking about then. Mar 09 21:39:26 which version of gcc is it ? Mar 09 21:39:37 Tartarus: kergoth seems very quiet today. I guess he must be thinking or something. Mar 09 21:39:52 I know with 4.1 series there were issues with make -j with install Mar 09 21:39:56 pb__ :) Mar 09 21:40:46 pb__: I think those are 2 different issues Mar 09 21:41:00 the one with gcc-4.3.3 and -j6 is one Mar 09 21:41:20 the other with BB_N_T and pstage inter-package race Mar 09 21:41:40 Tartarus: replied to what? Mar 09 21:42:02 kergoth_: What you figured out about stamp policy Mar 09 21:42:03 oh, damnit, gmail archived the thread even though it wasnt muted again Mar 09 21:42:12 * kergoth_ didn't see any of the followup mails on teh thread Mar 09 21:42:15 and that yes, this change should be right Mar 09 21:42:41 again, if i'd seen something to reply to, i would have replied toi t Mar 09 21:42:51 heh Mar 09 21:42:54 Anyhow Mar 09 21:43:30 * Tartarus goes chasing after another gcc issue that would just be a no-op if OE used sysroot stuff by default rather than needing -isystem this and --sysroot= that Mar 09 21:44:48 03Kristoffer Ericson  07org.openembedded.dev * r1fb6cc37b0 10openembedded.git/recipes/linux/ (2 files in 2 dirs): Mar 09 21:44:48 linux/linux/linux-jlime-jornada7xx_2.6.32+git.bb: Mar 09 21:44:48 * Add newest HP Jornada 700-series kernel recipe. Mar 09 21:44:48 Signed-off-by: Filip Zyzniewski Mar 09 21:44:48 Signed-off-by: Kristoffer Ericson Mar 09 22:09:52 Tartarus, kergoth_: were there any more discussions on fixing pstage with DEPLOY_DIR outside of TMPDIR? :) Mar 09 22:11:10 Not about that issue in particular, no. There were discussions on general pstage changes, but those stalled due to disagreements over the implementation. Mar 09 22:11:36 * kergoth_ grumbles, OE really doesn't want to cooperate today Mar 09 22:12:58 yeah, I looked through those disucssions on pstage changes - I understand they are not directly related to my issue, but do you think that would help in my case? :) Mar 09 22:19:49 denix: yes. in my proposal, all do_deploys would go based on the installed files, regardless of whether those came from the pstage package or do_install directly. I think changing how do_deploy is handled is a good step regardless of the rest of the changes we may want to implement, though. Mar 09 22:20:04 * kergoth_ gets food Mar 09 22:38:23 * kergoth_ is beginning to regret trying to go cold turkey on caffeine Mar 09 22:39:51 hehe Mar 09 22:40:04 * XorA avoid kergoth_s state for a while Mar 09 22:40:23 probably best.. either i have a horrible headache, or i'm half asleep, or both :) Mar 09 22:43:49 * kergoth_ ponders Mar 09 22:49:23 ah, that is no fun Mar 09 22:49:36 I think it is easier to give it up gradually, you don't tend to get the headaches that way. Mar 09 22:50:31 * XorA drinks almost zero caffeine :-) Mar 09 22:50:48 and of course, if you take painkillers for your headache, half of them have caffeine in anyway so you might as well not have bothered :-) Mar 09 22:51:34 XorA: still waiting for someone to invent caffeinated rum, I suppose Mar 09 22:51:46 caffeince makes me ill Mar 09 22:51:56 Id rather just have more rum Mar 09 22:52:46 It seems like I've been trying to cut sodas and caffinated beverages both from my diet for years, off and on.. Mar 09 22:52:54 clearly I need a better approach :) Mar 09 22:53:17 I was glad to learn on last USA trip there is now caffeine free root beer Mar 09 22:54:21 * kergoth_ thinks they need to open stores like http://popsoda.com/noname.html in every major city Mar 09 22:54:36 the good stuff :) Mar 09 22:55:09 OMG kergoth_ take me there :-) Mar 09 22:55:29 * XorA wonders if anyone ever made properly fermented root beer Mar 09 22:56:08 * kergoth_ is about 10 minutes from the store for that site, and its probably not very good for his waistline or pocketbook, but man he'd miss it if he moves Mar 09 22:57:18 * kergoth_ thinks about sitting down and splitting up base.bbclass some Mar 09 22:57:37 So Mar 09 22:57:43 I think its past time for oe_libinstall & friends to just die. Mar 09 22:58:06 If I was to bring in zlib 1.2.3.3 from debian, should I link to the uri for the patch, or just copy, strip out debian/ dir stuff and use that? Mar 09 22:59:02 Tartarus: if it wasn't debian, I would say link to the url. but, since debian urls tend to be somewhat transient, probably best in this case to copy. Mar 09 22:59:03 in the past we've mostly done the former, but I'd suggest the latter for ease of maintenance and avoiding unnecessary deps on debian, personally Mar 09 22:59:07 heh, yeah Mar 09 22:59:17 My thoughts too Mar 09 23:03:26 kergoth_: yah, splitting up base.bbclass would be cool. there is a whole lot of weird stuff in there. Mar 09 23:04:14 shit's been piling up in there since the project was started, I think.. far worse than the recipes, due to the lack of exposure, code wise.. most are scared away from what's under the hood, I think :) Mar 09 23:04:16 I don't harbour any particular malice towards oe_libinstall, though obviously it does tend to be an indication that old school staging may be close at hand Mar 09 23:05:02 but, yeah, base.bbclass does seem to have become just some kind of grab-bag of random utility functions. it's a bit sad to force every recipe to parse that stuff. Mar 09 23:06:06 check_gcc3(), for example. in what sense could that possibly be "base"? Mar 09 23:06:17 I moved that ;) But yeah Mar 09 23:06:28 heh Mar 09 23:06:29 It's the "more than one file needed this so, BASE!" Mar 09 23:06:31 file Mar 09 23:06:38 cbrake, ping? Mar 09 23:07:36 yeah. or, I suspect in some cases, "some other file might want this in the future, maybe... so, BASE!" Mar 09 23:07:44 the melo server will be offline for a little bit for some hardware upgrades...bbi1hr or so. Mar 09 23:08:00 mmmm new hardwares Mar 09 23:08:54 I vote these "unexport"s in base.bbclass die too Mar 09 23:09:06 if a var breaks a recipe, it should get unexported in that recipe.. Mar 09 23:10:38 well, I dunno, I think recipes probably do have a right to expect that their environment will be relatively hygenic. **** ENDING LOGGING AT Tue Mar 09 23:10:55 2010 **** BEGIN LOGGING AT Wed Mar 10 21:48:19 2010 Mar 10 21:48:22 pb__: in theory yes Mar 10 21:49:47 khem: well, if it doesn't work today, it can be fixed. there is no fundamental reason why the recipe needs to be different. Mar 10 21:50:08 you might be right that there are perhaps some linux-isms in the existing gcc.bb but that is arguably just a bug. Mar 10 21:50:23 pb__: right build sequence would be a bit differnet Mar 10 21:50:29 you dont need kernel headers Mar 10 21:50:43 and shared lib support to build gcc Mar 10 21:51:29 but as a design it should be doable to stick in same recipes Mar 10 21:51:58 just put 'freestanding' into your distro feature and fix it up with a chainsaw Mar 10 21:52:05 heh Mar 10 21:53:12 We have to provide machine specific libgloss port Mar 10 21:53:12 khem, btw.. i'm doing the very same now for nls lfs rpc ipv4 ipv6 Mar 10 21:53:17 well, a baremetal Xorg, with integrated fiflesystem support, all static, would rock Mar 10 21:53:48 fiflesystem....lol typos Mar 10 21:53:51 filesystem Mar 10 21:54:00 good luck with xorg on that setup Mar 10 21:54:06 Why? Mar 10 21:54:12 Brokie: xorg might be a big ask Mar 10 21:54:38 I have Xorg, ext3, static running on NXP CX8785 Mar 10 21:54:47 as baremetal Mar 10 21:55:05 broonie: what do you use for a libc Mar 10 21:55:08 and pthread Mar 10 21:55:23 xorg doesn't need pthread, does it? it never used to Mar 10 21:55:26 static Mar 10 21:55:26 Brokie, ext3 != bare metal, is it? Who provides an ext3 accessor ? Mar 10 21:55:42 EXT3 only for USB drives Mar 10 21:55:50 not for bootup Mar 10 21:55:59 Boot direct from ROM Mar 10 21:56:20 I had an xfree86 server running on a bare metal system at one point back in the day, it wasn't all that hard. the xserver doesn't (didn't) use as many os services as you might expect. Mar 10 21:56:29 nope Mar 10 21:56:37 it provides alot of the services itself Mar 10 21:56:51 pb__, last time i tried it did not. Not sure if that was xfree86 or already xorg, though Mar 10 21:57:51 We have an STi7109, Flash based system, treats flash as rom Mar 10 21:58:05 it might have gotten slightly worse more recently, in that I think it tends to lean on the OS for PCI support and suchlike, and I guess the input system is probably a bit more complex nowadays, but I don't think it is fundamentally very different to how it used to be. Mar 10 21:58:10 runs Xorg, has EXT3, NTFS, and runs QT4/e Mar 10 21:58:12 pb__, i was under the implression that xorg nowadays used libpciaccess (and thus /sys, /proc and whatnot) and other fancyness but i may be wrong Mar 10 21:58:32 s/impl/imp/ Mar 10 21:58:59 I like implressions, they keep my pants in one spot Mar 10 21:59:02 blindvt: yeah, that's probably true; you'd need some way to finesse that. obviously if your video is not interfaced by pci then it is more or less a non-issue anyway. Mar 10 21:59:11 pb__: for mac someone wrote xorg-libpthread-stubs so I thought it did Mar 10 21:59:25 and, even if it is interfaced via pci, you can provide your own implementation of libpciaccess that does the right thing Mar 10 21:59:40 Xorg can run threaded... but not necessary for compile, or function Mar 10 22:00:03 khem: I dunno, I've never heard of pthreads being mandatory. if so, that is a new thing. Mar 10 22:00:15 pb__, yes, for !pci it should still be relatively painless to get running i'd guess. Blind guess though Mar 10 22:00:42 ST40 has no PCI Mar 10 22:00:47 Brokie: ok, and what does your system use for libc Mar 10 22:00:54 and runtime Mar 10 22:01:03 libgcc.a Mar 10 22:01:05 is it newlib Mar 10 22:01:11 no Mar 10 22:01:33 the C library I meant Mar 10 22:01:34 We use a library from ST, called STOS40 Mar 10 22:01:54 I see Mar 10 22:02:47 for an opensource solution one has to see if newlib/libgloss can be used or not Mar 10 22:02:49 lookie here... Mar 10 22:02:52 http://www.stlinux.com/ Mar 10 22:02:59 whatever that may be Mar 10 22:04:02 Brokie: stlinux ? that does not seem to be baremetal Mar 10 22:04:17 it can be Mar 10 22:04:24 on our system it was Mar 10 22:04:40 we only had 32MB flash, so we went baremetal route Mar 10 22:05:45 i'm still not convinced that your notion of bare-metal matches mine, but anyway Mar 10 22:05:53 http://www.youtube.com/watch?v=MKq15o6kJo0 Mar 10 22:05:59 Pay attention to the update Mar 10 22:06:01 no, thanks Mar 10 22:06:28 its sent as PGM Mar 10 22:06:29 that stuff usually doesn't render for me Mar 10 22:06:35 its a compressed binary Mar 10 22:06:44 No filesystem in flash Mar 10 22:06:48 it runs baremetal Mar 10 22:07:06 I was engineer for ViewsatUSA Mar 10 22:07:44 \"Hello, you either have JavaScript turned off or an old version of Adobe's Flash Player\". Not bare enough Mar 10 22:07:47 plays fine in google chrome... Mar 10 22:08:03 grab google chrome, compile it, and watch it with google chrome Mar 10 22:08:56 Brokie, thanks for the hint but lynx renders everything i'm interrested in Mar 10 22:09:11 blindvt: ugh lynx Mar 10 22:09:27 or links[0-9] for that matter ;) Mar 10 22:10:01 ok Mar 10 22:10:06 blindvt: firefox is also decent except sometimes its a memory hog Mar 10 22:10:14 smash firefox... Mar 10 22:12:08 khem: eglibc (why do you ask)? Mar 10 22:12:55 broonie: Q was for Brokie Mar 10 22:13:04 wrong tab completion Mar 10 22:25:44 cbrake ping? Mar 10 22:26:27 can somebody tell me if the git server is working again? **** BEGIN LOGGING AT Thu Mar 11 09:00:15 2010 Mar 11 09:03:59 ~logs Mar 11 09:04:00 All conversations are logged to http://ibot.rikers.org/channel, where "channel" is replaced by the URL-encoded channel name, such as %23freenode for #freenode. Lines starting with spaces are not logged. logs are updated daily Mar 11 10:11:10 hm, tried to build vala, but as the oe servers are down it fetches from the source, but http://www.freesmartphone.org/sources does not even exist Mar 11 10:11:37 and the recipe is at 0.7.10.3 but the vala home page only goes to 0.7.10 (which is from feb) Mar 11 10:12:38 eFfeM: it exists.. downloaded yesterday 0.7.10.4 ok Mar 11 10:14:05 eFfeM: in .inc I see http://downloads.freesmartphone.org/sources/ not www Mar 11 10:14:29 git should be coming up soon Mar 11 10:15:09 JaMa: hm, my inc says: SRC_URI = "http://www.freesmartphone.org/sources/vala-${PV}.tar.bz2" Mar 11 10:15:22 rwhitby: great & thanks for the info Mar 11 10:16:14 JaMa: was also a little bit surprised to find newer versions on freesmartphone than on the official site of the project Mar 11 10:16:31 rwhitby: woowoo Mar 11 10:16:39 XorA: all good? Mar 11 10:16:41 rwhitby: I take it you guys been having a bad day :-) Mar 11 10:16:44 eFfeM: hmm I really have it in recipes/vala/vala.inc Mar 11 10:17:18 XorA: there was a problem with the second NIC on the new MB, so a new card had to be sourced for that. then the PV migration to the new RAID array took much longer than expected. Mar 11 10:17:25 i can't recall having touched it but maybe it got changed after my last pull Mar 11 10:17:32 So an 8 hour job turned into 36 hours. Mar 11 10:17:35 rwhitby: sounds just like my other work :-) Mar 11 10:17:41 mmm Mar 11 10:17:43 [ 160.942111] Terminal[993]: segfault at 4 ip b6d0c618 sp bfe3ad7c error 6 in libglib-2.0.so.0.2200.4[b6cfb000+c0000] Mar 11 10:17:49 ka6sox-work had to extend his stay at OSUOSL for an extra day Mar 11 10:18:25 XorA: can you confirm everything for OE is working? Mar 11 10:18:32 rwhitby: git isnt up yet Mar 11 10:18:39 eFfeM: ah sorry, you're right, mickey updated it in his last commit (not yet in oe.dev) Mar 11 10:18:50 rwhitby: but the port is no longer a dead zone, its replying with connection refused Mar 11 10:19:09 eFfeM: http://git.freesmartphone.org/?p=openembedded.git;a=commit;h=640462b0b5bbc3abfaaf17eba6d19ce00b532b2d Mar 11 10:19:19 XorA: still not responding? Mar 11 10:19:27 rwhitby: there is goes Mar 11 10:19:34 rwhitby: git+ssh is working Mar 11 10:19:59 03Steffen Sledz  07org.openembedded.dev * rabfc008565 10openembedded.git/ (classes/image.bbclass conf/bitbake.conf): Mar 11 10:19:59 bitbake.conf: introducing new image fstype .cpio.gz.u-boot Mar 11 10:19:59 * initrd images need to be prepared with mkimage to be usable from u-boot Mar 11 10:19:59 Signed-off-by: Steffen Sledz Mar 11 10:19:59 Acked-by: Martyn Welch Mar 11 10:20:00 Acked-by: Tom Rini Mar 11 10:20:10 and CIA obviousky :-) Mar 11 10:20:13 ka6sox: dyoung brought up melo and XorA confirms working Mar 11 10:20:29 rwhitby, thanks...I just remembered that... Mar 11 10:20:51 long day today. Mar 11 10:20:58 glad its done Mar 11 10:21:00 make sure those guys have a beer :-) Mar 11 10:21:12 ka6sox, thanks alot! Mar 11 10:21:17 if I did that I Mar 11 10:21:22 JaMa: that commit also changed the dir in .inc Mar 11 10:21:27 would probably fade to black! Mar 11 10:21:37 http://git.freesmartphone.org/?p=openembedded.git;a=blobdiff;f=recipes/vala/vala.inc;h=7de0cf9f91fc43fb9ba7277a046e540945cf0465;hp=620e2a6454d865dbdca8b575379828d0c3169698;hb=640462b0b5bbc3abfaaf17eba6d19ce00b532b2d;hpb=9fa5bd8ff0fceac54273850140b71319bc2c3b7e Mar 11 10:21:49 as I had still .3 i also had the old inc Mar 11 10:23:06 03Koen Kooi  07org.openembedded.dev * r72ae9db372 10openembedded.git/recipes/xorg-lib/ (3 files in 2 dirs): pixman git: add some ARMv6 optimizations Mar 11 10:23:07 ka6sox: thanks for your work dude! Mar 11 10:23:54 mmm [ 638.779526] xfce4-panel[957]: segfault at 4 ip b6e13618 sp bf90001c error 6 in libglib-2.0.so.0.2200.4[b6e02000+c0000] Mar 11 10:23:55 again Mar 11 10:25:50 XorA, thats what we do...this server should be a lot faster than the old one. Mar 11 10:26:10 ka6sox: as an admin I like to thank people for their work! Mar 11 10:29:35 much appreciated. Mar 11 10:29:55 eFfeM: yes.. that's why I said 11:19:52 < JaMa> eFfeM: ah sorry, you're right, mickey updated it in his last commit (not yet in oe.dev) Mar 11 10:31:47 ah JaMa didn't read well enough Mar 11 10:35:47 03Michael 'Mickey' Lauer  07org.openembedded.dev * rff7af94a4f 10openembedded.git/recipes/freesmartphone/fsogsmd_git.bb: fsogsmd: depend on ppp and msmcommd; build the qualcomm_palm modem plugin Mar 11 10:35:58 03Michael 'Mickey' Lauer  07org.openembedded.dev * r7cb4c48134 10openembedded.git/ (4 files in 2 dirs): vala: 0.7.10.3 -> 0.7.10.4 Mar 11 10:35:58 03Michael 'Mickey' Lauer  07org.openembedded.dev * r355d43e7e3 10openembedded.git/recipes/freesmartphone/ (fso-abyss_git.bb libgsm0710mux_git.bb): libgsm0710mux|fso-abyss: catch up with SRC_URI; both have moved into cornucopia Mar 11 10:36:05 *yawn* Mar 11 10:41:30 Hi Guys. I pushed a patch to the mailing list yesterday to add support for the Freescale p2020ds demo board. Doing an update this morning it seems that an addition to the linux 2.6.33 bitbake file has caused the patch to not apply cleanly. Should I resend an updated patch? Mar 11 10:41:46 03Martin Jansa  07org.openembedded.dev * r7291f24141 10openembedded.git/recipes/linux/ (3 files in 2 dirs): Mar 11 10:41:46 linux-openmoko-2.6.32: two more patches Mar 11 10:41:46 Signed-off-by: Martin Jansa Mar 11 10:41:51 03Martin Jansa  07org.openembedded.dev * rdcd2eb8f94 10openembedded.git/recipes/navit/navit_svn.bb: Mar 11 10:41:51 navit: remove D_P -1 as older release 0.1.0 is not compatible with current gpsd Mar 11 10:41:51 Signed-off-by: Martin Jansa Mar 11 10:41:51 03Martin Jansa  07org.openembedded.dev * r1a037cc05d 10openembedded.git/recipes/efl1/ (3 files in 2 dirs): Mar 11 10:41:52 ecore: split exit_uclibc patch, as one part needs maxrev=47076 Mar 11 10:41:52 * ecore_con_dns.c was removed in r47077 Mar 11 10:41:53 Signed-off-by: Martin Jansa Mar 11 10:41:53 03Martin Jansa  07org.openembedded.dev * r24a32f4da4 10openembedded.git/ (5 files in 3 dirs): Mar 11 10:41:54 qi-ubi: add ubifs rootfs params also for gta01, move SRCREV to recipe Mar 11 10:41:54 Signed-off-by: Martin Jansa Mar 11 10:41:55 03Martin Jansa  07org.openembedded.dev * r9bd7e087b7 10openembedded.git/conf/compatibility-providers.conf: Mar 11 10:41:55 compatibility-providers: add gtk+-native as provider for gdk-pixbuf-csource-native Mar 11 10:41:56 Signed-off-by: Martin Jansa Mar 11 10:41:57 03Martin Jansa  07org.openembedded.dev * re6b545ce0c 10openembedded.git/recipes/xorg-driver/xf86-video-glamo_git.bb: Mar 11 10:41:57 xf86-video-glamo: bump SRCREV for xrandr fix Mar 11 10:41:58 Signed-off-by: Martin Jansa Mar 11 10:42:27 flood gates are open :D Mar 11 10:42:42 ~prase ka6sox for all his and the other's work Mar 11 10:42:45 err Mar 11 10:42:48 ~praise ka6sox for all his and the other's work Mar 11 10:42:49 All hail ka6sox for all his and the other's work! Mar 11 10:44:28 MWelchUK_work_: yes Mar 11 10:45:04 OK, shall I keep the ack I've already received (since it is rather trivial)? Mar 11 10:46:18 hm, no idea Mar 11 10:46:53 eFfeM, I'll leave it off then and CC the ack'er Mar 11 10:47:52 ok Mar 11 10:56:15 mickey|zzZZzz: good morning Mar 11 10:57:36 morning pb_ Mar 11 10:57:42 * mickey|zzZZzz not quite awaken yet Mar 11 10:58:18 03Frans Meulenbroeks  07org.openembedded.dev * r6d39f386da 10openembedded.git/recipes/lsof/lsof_4.78.bb: Mar 11 10:58:18 lsof: updated to 4.83 Mar 11 10:58:18 Signed-off-by: Frans Meulenbroeks Mar 11 10:58:21 03Frans Meulenbroeks  07org.openembedded.dev * r21cc37782c 10openembedded.git/ (recipes/libsndfile/libsndfile1_1.0.20.bb site/arm-common): Mar 11 10:58:21 libsndfile1: updated to 1.0.21 Mar 11 10:58:21 also fixed download location and moved two config vars to arm-common Mar 11 10:58:21 (not sure about ac_cv_sizeof_off_t=8, this seems odd or recipe specific, so left that one) Mar 11 10:58:21 Signed-off-by: Frans Meulenbroeks Mar 11 11:00:11 eFfeM: are you sure about that libsndfile commit? Mar 11 11:00:40 what part of it ? Mar 11 11:01:12 eFfeM: moving those values to site Mar 11 11:01:25 eFfeM: I generated them to work around a massive bug in the configure script Mar 11 11:01:48 eFfeM: they were never checked to see if they were the correct values for arm, just the correct values to pass the buggy tests Mar 11 11:01:59 eeeh, ok, didn' t know that, that is wy I left the off_t=8 as that seemed quite weird) Mar 11 11:02:11 blush, didn't verify them eihter Mar 11 11:02:28 was a head scratching day of reverse engineering some really bad logic :-) Mar 11 11:02:48 eFfeM: find the debian build logs, thats always a good method of verifying :-) Mar 11 11:02:51 i just noticed some vars on a place I did not expected them, and moved them to what seemed the proper location Mar 11 11:02:55 will do Mar 11 11:03:16 * XorA did put a comment in the commit message but you would have to git-log --follow to find it these days Mar 11 11:04:22 morning Mar 11 11:04:43 03Marco Cavallini  07org.openembedded.dev * rdc3ecdbab2 10openembedded.git/recipes/linux/linux_2.6.28.bb: Mar 11 11:04:43 linux_2.6.28.bb: updated to new paches for Ronetix PM9263 Mar 11 11:04:43 * used new 003_linux-2.6.28-at91-ronetix-20112009.patch Mar 11 11:04:43 * bump DEFAULT_PREFERENCE_ronetix-pm9263 Mar 11 11:04:44 03Marco Cavallini  07org.openembedded.dev * rb6eabf5ce1 10openembedded.git/conf/distro/include/kaeilos.inc: kaeilos.inc: added some new options Mar 11 11:04:44 03Marco Cavallini  07org.openembedded.dev * r3ff1dc5295 10openembedded.git/recipes/base-files/base-files/kaeilos/ (issue issue.net): kaeilos/issue: updated welcome message (also in kaeilos/issue.net) Mar 11 11:07:33 XorA: this does not seem to help me much: https://buildd.debian.org/fetch.cgi?pkg=libsndfile;ver=1.0.21-2;arch=armel;stamp=1260883504 Mar 11 11:08:36 eFfeM: hrmm, I saw in same logs it chucked out the config.log results somewhere Mar 11 11:09:22 this place: http://gnu.ethz.ch/debian/cinelerra/cinelerra-4.1/thirdparty/libsndfile-1.0.11/config.log ? Mar 11 11:09:29 that has the answers you give Mar 11 11:10:04 guess simplest solution would be to test on a target, but i don't have one handy Mar 11 11:11:28 always the issue Mar 11 11:13:46 03Koen Kooi  07org.openembedded.dev * rcd332df361 10openembedded.git/recipes/udev/ (udev-151/init udev_151.bb): udev 151: create modules.dep if it doesn't exist, fixes scary warnings at first boot Mar 11 11:21:57 XorA: what do you suggest to do with it? undo ? Mar 11 11:24:00 eFfeM: leave it, see who complains :-) Mar 11 11:24:50 xora actually i think it is ok: see the excerpt from the debian log: http://www.pastebin.ca/1833517 Mar 11 11:25:21 still find it strange that off_t is not always the same size Mar 11 11:25:59 eFfeM: from memory the configure script actually runs the same test twice, the 8 is effectively discarded, but if you dont force it to 8 it sets it wrong in the second running of the test Mar 11 11:26:43 eFfeM: it was reason why openmoko built on 64bit machines was very quiet, on 32bit it made noises Mar 11 11:26:54 yuk Mar 11 11:27:49 thats what I thought Mar 11 11:28:39 XorA: peeked a little bit further in the log see http://pastebin.ca/1833522 Mar 11 11:28:52 it seems to first test 4 then 8 Mar 11 11:29:00 first and last line of paste Mar 11 11:29:24 yeah, that brings back memories Mar 11 11:29:55 hmmm, I think it was what is size with -DNO_LARGE_FILE then what is size with -DLARGE_FILE or something like that Mar 11 11:30:10 but it never seperates the sizes into different vars Mar 11 11:30:41 seems a good explanation Mar 11 11:30:45 anyway.... Mar 11 11:30:50 lunchtime ! Mar 11 11:33:32 03Phil Blundell  07org.openembedded.dev * rf6774ecd46 10openembedded.git/recipes/gtk+/gtk+.inc: gtk+: generate gdk-pixbuf.loaders during native install Mar 11 11:33:33 03Phil Blundell  07org.openembedded.dev * ra9089155a6 10openembedded.git/classes/base.bbclass: base: also stage ${sysconfdir} Mar 11 11:37:14 03Koen Kooi  07org.openembedded.dev * re2fb432ca9 10openembedded.git/recipes/linux-libc-headers/ (linux-libc-headers.inc linux-libc-headers_2.6.31.bb): Mar 11 11:37:14 linux-libc-headers: reinstate scsi headers in staging that got lost with the new-style staging conversion Mar 11 11:37:14 * this makes udev build again Mar 11 11:37:17 03Michael 'Mickey' Lauer  07org.openembedded.dev * r5310bdc1b3 10openembedded.git/recipes/readline/readline-5.2/fix-includes.patch: Mar 11 11:37:17 readline 5.2: add missing include to stdio.h Mar 11 11:37:17 Include headers are usually required to be self-contained; only Mar 11 11:37:17 the readline team sees that quite differently. Mar 11 11:37:32 03Michael 'Mickey' Lauer  07org.openembedded.dev * r4891edc5e0 10openembedded.git/recipes/readline/readline-5.2/fix-includes.patch: Mar 11 11:37:33 readline 5.2: add missing include to stdio.h Mar 11 11:37:33 Include headers are usually required to be self-contained; only Mar 11 11:37:33 the readline team sees that quite differently. Mar 11 11:37:57 hmm Mar 11 11:37:59 double? Mar 11 11:38:14 morning Mar 11 11:38:16 indeed, seems like double trouble Mar 11 11:38:52 morning hrw Mar 11 11:38:53 strange Mar 11 11:39:03 and neither is the main file patched Mar 11 11:39:09 pb_: can I push libxext-native added to gtk+-native DEPENDS? for me it fails without (libx11-native failed before, but is pulled by libxext-native) Mar 11 11:39:20 JaMa: yeah, go ahead Mar 11 11:40:12 03t.fromm  07org.openembedded.dev * r20f3db21f3 10openembedded.git/recipes/linux/linux-2.6.24/hipox/hipox.patch: Mar 11 11:40:12 linux-2.6.24: fix kernel oops upon driver unload for hipox machine Mar 11 11:40:12 Signed-off-by: Thilo Fromm Mar 11 11:40:12 Acked-by: Steffen Sledz Mar 11 11:40:13 03Steffen Sledz  07org.openembedded.dev * rb35a5e4e57 10openembedded.git/: Merge remote branch 'upstream/org.openembedded.dev' into org.openembedded.dev Mar 11 11:40:14 03sledz  07org.openembedded.dev * r6064bd2778 10openembedded.git/recipes/busybox/files/hipox/syslog.conf: Mar 11 11:40:14 busybox: syslog to file instead of buffer for hipox machine by default Mar 11 11:40:14 Signed-off-by: Steffen Sledz Mar 11 11:40:26 03sledz  07org.openembedded.dev * rc7b148b0c8 10openembedded.git/recipes/udev/udev-141/hipox/default: Mar 11 11:40:26 udev: we don't like a DEVCACHE at hipox machine Mar 11 11:40:26 Signed-off-by: Steffen Sledz Mar 11 11:41:24 03Michael 'Mickey' Lauer  07org.openembedded.dev * r82311ec396 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Mar 11 11:41:24 03Michael 'Mickey' Lauer  07org.openembedded.dev * rcad405defe 10openembedded.git/recipes/readline/readline_5.2.bb: readline 5.2: and apply the patch to fix the missing include directive Mar 11 11:41:35 JaMa: this one will fix the msmcommd compilation Mar 11 11:43:25 mickey|zzZZzz: thanks Mar 11 12:06:07 can someone change the channel msg (git up) Mar 11 12:06:24 13:05 -!- #oe You're not channel operator Mar 11 12:06:42 me neihter, who is? Mar 11 12:07:09 are poky related questions off-topic herE? Mar 11 12:07:22 * XorA doesnt even have the /topic command :-) Mar 11 12:07:55 c0le: no, we like poky Mar 11 12:07:56 eFfeM: mickey|zzZZzz is an operator, I tihnk Mar 11 12:08:19 probably kergoth too Mar 11 12:08:54 ~curse smc95xx driver Mar 11 12:08:55 XorA: :) I am trying to evaluate poky for an industrial computer application. But am confused about where its heading to.. as in poky/moblin/meego .. Mar 11 12:08:55 May you be reincarnated as a Windows XP administrator, smc95xx driver ! Mar 11 12:09:18 c0le: for that info you would have to ask RP Mar 11 12:09:52 bah git-native requires python development packages on my host :-( Mar 11 12:10:20 what sick twisted mind made git use both perl and python Mar 11 12:10:20 c0le: feel free to try Poky Mar 11 12:10:59 hrw: yeah, I already started working on it. and lately saw this thing about intel acquiring o-hand.. and looked like poky was under o-hand. Mar 11 12:11:01 c0le: OH team got merged into Moblin(+ few other Intel teams) 1.5 year ago and Poky still exists Mar 11 12:11:26 * hrw was on "OH is now Intel" conference Mar 11 12:11:56 hrw: and was unsure whether poky is going to be poky or will be merged into moblin. coz, moblin seems to be targetted for intel chips.. Mar 11 12:12:33 c0le: there was attempt to make Moblin Poky based but it failed. Poky will not be merged into Moblin Mar 11 12:13:08 hrw: so, I can safely assume that poky is a stable release cycle of the oe infrastructure? Mar 11 12:13:15 not quite Mar 11 12:13:33 c0le: Poky is OE based but it is not 'stable release of oe' Mar 11 12:14:09 hrw: ok.. Mar 11 12:15:18 anyone an idea on how to merge m4-native and m4? when I tried I got a cycle, with autoconf, guess INHIBIT_AUTOTOOLS_DEPS = "1" in the native recipe breaks the dep, but no idea how to resolve that iwth BBCLASSEXTEND Mar 11 12:15:22 hi GNUtoo|oeee Mar 11 12:15:35 bye GNUtoo|oeee :-) Mar 11 12:26:46 hrw: so, which is recommended. as in, poky or oe? Mar 11 12:28:35 I try to use a prebuild toolchian for oe to build something. I do it follow this doc: http://docs.openembedded.org/usermanual/html/commonuse_prebuilt_toolchain.html . I run 'bitbake zlib" ,and it get many errors(http://ubuntu.pastebin.com/p5jy0xHL) And the local.conf(http://ubuntu.pastebin.com/Q0N2NuQd) ... Mar 11 12:29:59 c0le: Poky has less changes during development then OE and it's stable branch is more fresh then stable/2009 branch of OE Mar 11 12:30:12 03Phil Blundell  07org.openembedded.dev * rcf770ec168 10openembedded.git/recipes/gtk+/gtk+.inc: gtk+: work around problem with loaders not yet being installed at the point where we need to query them Mar 11 12:30:43 c0le: try Poky, try OE and decide which one fits you better Mar 11 12:31:59 hrw: ok.. I'll evaluate both then. :) Mar 11 12:32:04 hrw: thanks a ton :) Mar 11 12:33:04 uf... bugbase2.0 ethernet is now eth0 Mar 11 12:35:17 03Klaus Kurzmann  07org.openembedded.dev * ra94764b15b 10openembedded.git/conf/distro/include/sane-srcrevs-fso.inc: Mar 11 12:35:17 sane-srcrevs-fso.inc: bump PR of libframeworkd-glib Mar 11 12:35:17 Signed-off-by: Klaus Kurzmann Mar 11 12:41:36 03Martin Jansa  07org.openembedded.dev * r21c2ae56fb 10openembedded.git/recipes/sdlperl/sdl-perl_1.20.3.bb: Mar 11 12:41:36 sdl-perl: remove replace for smpeg.h in SDL_perl.xs as it's not needed anymore for newer smpeg Mar 11 12:41:36 Signed-off-by: Martin Jansa Mar 11 12:41:37 03Martin Jansa  07org.openembedded.dev * r8c203a8fe6 10openembedded.git/conf/compatibility-providers.conf: Mar 11 12:41:37 fsodeviced: use it as default alsa scenarii provider Mar 11 12:41:37 Signed-off-by: Martin Jansa Mar 11 12:41:38 03Martin Jansa  07org.openembedded.dev * r13c2e6cede 10openembedded.git/recipes/smpeg/ (smpeg/as-fix.patch smpeg_svn.bb): Mar 11 12:41:38 smpeg: convert to new staging and bump SRCREV a bit Mar 11 12:41:38 * please note that headers are now as upstream puts them in /usr/include/smpeg/* Mar 11 12:41:38 instead /usr/include/SDL Mar 11 12:41:39 * smpeg-config returns right paths, so it should be OK for sane packages Mar 11 12:41:39 Signed-off-by: Martin Jansa Mar 11 12:41:40 03Martin Jansa  07org.openembedded.dev * rb6cef8b521 10openembedded.git/recipes/python/ (python-pygame-1.8.1/Setup python-pygame_1.8.1.bb): Mar 11 12:41:40 python-pygame: add version 1.9.1, with new staging and smpeg compatibility Mar 11 12:41:41 Signed-off-by: Martin Jansa Mar 11 12:42:25 03Martin Jansa  07org.openembedded.dev * rca622ee8a9 10openembedded.git/recipes/gtk+/gtk+_2.18.6.bb: Mar 11 12:42:25 gtk+_2.18.6: add libxext-native to DEPENDS Mar 11 12:42:25 Signed-off-by: Martin Jansa Mar 11 12:57:30 Hi. I'm trying to use xinput-calibrator to calibrate an eGalax touchscreen and I noticed that the output for xorg.conf has maxx and miny inverted, compared to udev and hal rules. Is that expected? Here's the output: http://paste.debian.net/63685/ Mar 11 12:58:45 mario-goulart: I always used only settings for xinput, if you don't get answer here, try to contact author (if it's bug) Mar 11 13:00:19 JaMa: Thanks. I have Xorg 1.5.3 and Xinput 1.3.0 here, so I guess xinput is not an option. I'm gonna try to contact the author. Mar 11 13:08:21 mario-goulart: xinput-calibrator works with xorg 1.5.3 Mar 11 13:09:55 hrw: I mean the 'xinput set' commands. xinput-calibrator doesn't print them if Xorg < 1.6.0, as far as I understand. Mar 11 13:12:29 hrw: see the DEBUG messages at http://paste.debian.net/63500/ Mar 11 13:13:38 ok then Mar 11 13:13:51 I added xinput-calibrator output to xorg.conf Mar 11 13:15:17 I'm trying to use hal, but maxx and miny look inverted compared to the output for xorg. I don't know if it is a bug or if xorg.conf uses different references. Mar 11 13:15:39 It's very suspicious. Mar 11 13:23:07 It really looks like a bug: http://paste.debian.net/63697/ Mar 11 13:41:48 Hi all, Mar 11 13:42:00 does anyone can help me to solve this problem: Mar 11 13:42:03 ERROR: tiff-3.8.2+4.0.0beta2: http://download.osgeo.org/libtiff/old/tiff-4.0.0beta2.tar.gz has no checksum defined, cannot check archive integrity Mar 11 13:42:19 in the case of building x86 machine, Mar 11 13:43:48 mmm I've a trace Mar 11 13:44:16 NOTE: Running task 830 of 1929 (ID: 1710, /opt/my_OE/openembedded/recipes/libtiff/tiff_3.8.2.bb, do_fetch) NOTE: Missing checksum ERROR: tiff-3.8.2+4.0.0beta2: http://download.osgeo.org/libtiff/old/tiff-4.0.0beta2.tar.gz has no checksum defined, cannot check archive integrity ERROR: Error in executing: /opt/my_OE/openembedded/recipes/libtiff/tiff_3.8.2.bb ERROR: Exception: Message:1 ERROR: Printing the env Mar 11 13:45:05 guilig, there is a python script for that in a dir....let me find it Mar 11 13:46:39 contrib/source-checker/oe-checksums-sorter.py Mar 11 13:47:31 guilig, basically do the following adapting the paths Mar 11 13:47:36 cat ${HOME}/oetmp/checksums.ini >> ${HOME}/oe/org.openembedded.dev/conf/checksums.ini Mar 11 13:47:48 cd ${HOME}/oe/org.openembedded.dev/conf/ Mar 11 13:47:51 ok i will try Mar 11 13:47:53 python ../contrib/source-checker/oe-checksums-sorter.py checksums.ini >tmp Mar 11 13:48:00 mv tmp checksums.ini Mar 11 13:48:09 or simply add the checksums inside the recipe Mar 11 13:48:16 like in the newer wesnoth recipe Mar 11 13:48:27 and in both cases send a patch Mar 11 13:48:38 example of checksums inside the recipe: Mar 11 13:49:12 SRC_URI = "${SOURCEFORGE_MIRROR}/wesnoth/${PN}-${PV}.tar.bz2" -> SRC_URI = "${SOURCEFORGE_MIRROR}/wesnoth/${PN}-${PV}.tar.bz2;name=tarball" Mar 11 13:49:21 SRC_URI[tarball.md5sum] = "493826bbd9ba355930765a7e8fe3749a" Mar 11 13:49:22 SRC_URI[tarball.sha256sum] = "7ef047ae364278a5bf9bdc69228f77d825f793f1c4d9adae8b47f0882e7f30d7" Mar 11 13:51:29 http://pastebin.com/zngZyvD5 for my problem Mar 11 13:51:36 (not related to checksums) Mar 11 13:51:39 GNUtoo|oeee: python ../contrib/source-checker/oe-checksums-sorter.py -i checksums.ini Mar 11 13:51:48 ok thanks Mar 11 13:54:15 what's the problem on pastebin ? Mar 11 13:55:36 segmentation fault of glib on x86 target when there is hald running Mar 11 13:58:03 ok understand Mar 11 13:58:28 actually peeked at the tiff recipe earlier today guess it should be named 3.9.2+4.0.0beta2 as 3.9.2 is the last version before 4.0 Mar 11 13:58:47 tried to bring it to beta5, but that seemed to require a newer autoconf Mar 11 14:00:14 hi, after booting a fresh baked image, i get no login prompt in my serial terminal. how can i fix this? Mar 11 14:03:16 hexor: Are you sure you have the correct baudrate? Mar 11 14:04:04 hexor: kernel output is present? Mar 11 14:05:32 eFfeM, hi, I've some strange glib segfault: terminal,xfce4-pannel,xchat segfault under the eee701,but only when hald is running,if I kill hald it doesn't segfault anymore,I've a trace: http://pastebin.com/zngZyvD5 but unfortunately my terminal segfaulted so I couldn't go into the frame etc... Mar 11 14:05:45 I heard that hald will die Mar 11 14:05:58 and that it may be replaced by some udev thing Mar 11 14:06:09 is it true,and if so could I bitbake it and install that? Mar 11 14:08:20 ah ok udev-extras according to wikipedia Mar 11 14:10:48 GNUtoo|oeee: sounds like a bug, but don't have any glib knowledge Mar 11 14:11:07 I've udev 151 Mar 11 14:11:10 ok Mar 11 14:11:36 would it be a good idea if a checksum is missing and OE_STRICT_CHECKSUMS is true, still to emit the checksums (so dev's can even with that setting easily get the right setting and clip it in the recipe) Mar 11 14:11:49 in udev_151 there is: # Needed for udev-extras Mar 11 14:12:18 MWelchUK_work__ hrw kernel output is present Mar 11 14:12:35 ok Mar 11 14:12:38 im passing the correct baudrate to the kernel to boot it Mar 11 14:12:49 but I didn't find udev-extras Mar 11 14:13:07 hexor: did you got 'init respawning too fast' like message? Mar 11 14:13:24 GNUtoo|oeee: I'm using udev for xorg input devices and you need to prepare udev rules, but then it works ok Mar 11 14:13:30 hexor: check does you system has /dev/ttyS0 (or which ever you used for console login) Mar 11 14:13:33 hrw no Mar 11 14:13:42 JaMa, is there a howto somewhere? Mar 11 14:13:43 hrw let me check Mar 11 14:13:52 GNUtoo|oeee: If you check latest xserver-xorg, then there is support to choose input manager in distro config Mar 11 14:13:52 what do you mean by prepare? Mar 11 14:13:59 ok Mar 11 14:14:25 hrw it has ttySAC0 (its the serial for my arch/processor) i cannot Mar 11 14:14:26 GNUtoo|oeee: GNUtoo|oeee http://who-t.blogspot.com/2010/01/new-configuration-world-order.html Mar 11 14:14:27 I used angstrom Mar 11 14:14:36 ok thanks Mar 11 14:16:09 hexor: make initscript which will do "ls /dev/" and add it to rootfs. this will tell you how your ttySAC0 is named Mar 11 14:16:25 GNUtoo|oeee: http://bugs.gentoo.org/show_bug.cgi?id=305371 Mar 11 14:16:25 hexor: I had such problem with BeagleBoard and custom rootfs Mar 11 14:16:46 thanks a lot Mar 11 14:16:50 what about automount Mar 11 14:16:54 is it done automatically? Mar 11 14:17:06 if not blacklisted in udev rules Mar 11 14:17:43 hrw i will try, thanks :) Mar 11 14:18:27 ok thanks a lot Mar 11 14:23:08 I'll reboot and try Mar 11 14:30:00 03Martin Jansa  07org.openembedded.dev * r44ccc03cf9 10openembedded.git/recipes/linux-libc-headers/linux-libc-headers_2.6.32.bb: Mar 11 14:30:00 linux-libc-headers_2.6.32: reinstate scsi headers, PR bump after e2fb432ca9f375defad1491a3ae83c473f6036fb Mar 11 14:30:00 Signed-off-by: Martin Jansa Mar 11 14:30:38 is udev-151 broken? Mar 11 14:31:47 tasslehoff: It's broken for me today. Mar 11 14:32:38 koen fixed it for linux-libc-headers_2.6.31 Mar 11 14:32:53 for 2.6.32 it needs a bit more than this PR bump :/ ^ Mar 11 14:32:59 MWelchUK_work__: not just me then. thanks. :) Mar 11 14:33:24 ok Mar 11 14:33:34 tasslehoff, Is this what you are seeing? http://pastebin.ca/1833791 Mar 11 14:34:36 MWelchUK_work__: yep, that's the one Mar 11 14:34:38 MWelchUK_work__: I have the same for 2.6.32 Mar 11 14:35:06 JaMa, Ah, just saw your post above. Mar 11 14:35:20 reverting scsi.h chunk from http://gitweb.icadyptes.org/?p=icadyptes-core/.git;a=commitdiff_plain;h=1fd88711e329776387bb985e573e3318ab1fc115 worked for me, but probably not best solution Mar 11 14:35:28 MWelchUK_work__: so you have 2.6.32 to? Mar 11 14:35:29 too? Mar 11 14:35:39 or is it the same error for 2.6.31? Mar 11 14:36:27 I've been trying to build 2.6.33. Funny thing was that it build for me yesterday - think I may have had a slightly odd machine.conf when I built udev Mar 11 14:36:56 MWelchUK_work__: linux-libc-headers_2.6.33? Mar 11 14:38:24 I think I may have been forcing it to a very old version of the header :-s Mar 11 14:44:11 eFfeM: i tried to clean up m4 and hit the same issue you did, INHIBIT_*_DEPS_virtclass-native doesn't take effect. it seems that the deps are queried before the overrides are processed, or so. Mar 11 14:49:14 hrw i think my no prompt problem is because tinylogin is not being compiled and baked into my rootfs Mar 11 14:49:38 strange Mar 11 14:49:50 hrw inittab looks for it u know (actually looks for getty, but getty is a symlink to tiniylogin and tinylogin is a broken symlink) Mar 11 14:51:51 kergoth: tried the same thing as you did with the same amount of success Mar 11 14:52:15 eFfeM: i noticed m4-native recipes still exist in poky, i guess rp hit it too and didn't feel like fixing it just then :) Mar 11 14:52:19 * kergoth doesn't either, really.. Mar 11 14:53:22 kergoth: the only things I could think of (but did not test) is, is doing the auto stuff only for target or a dedicated configure or so Mar 11 14:53:23 JaMa, Ah! No, I think I'm using linux-libc-headers_2.6.31, but it probably pulled the tree before Koen's fix. Mar 11 14:54:39 I added a print of the value of it from the autotools_dep_prepend, and for a single recipe, it shows multiple executions, the first couple show None, the second couple show the correct value Mar 11 14:55:07 so it seems something is pulling those vars before the overrides, but who/what/why? :) Mar 11 14:55:21 thought about printing a traceback there Mar 11 14:55:28 but i doubt itd be of much help Mar 11 14:58:37 DEPENDS_virtclass-native_prepend = "" should do the same thing I think Mar 11 15:01:47 no. Mar 11 15:02:04 the autotools & default deps are prepended/appended already Mar 11 15:02:08 you can't override one prepend with another Mar 11 15:02:10 MWelchUK_work__: then you would see error like this extras/ata_id/ata_id.c:33:23: error: scsi/scsi.h: No such file or directory http://tinderbox.openembedded.net/packages/513458/ Mar 11 15:04:30 kergoth: oh yeah, forgot prepend was special Mar 11 15:05:09 JaMa, Yes - I saw that earier this morning. Mar 11 15:09:26 hrw: not sure if it would help opkg resolve conflict automaticaly, but shouldn't util-linux-ng RCONFLICTS with e2fsprogs-uuidgen? Mar 11 15:10:40 feel free to add it if needed Mar 11 15:11:29 ok I'll try if it will remove old e2fsprogs-uuidgen before upgrading to newer util-linux-ng Mar 11 15:20:21 kergoth: i guess the only way out would be not inheriting autohell and do things explicit. that way with virtclass-native one can get two versions Mar 11 15:20:37 (btw, still think it is a good idea to also have a _target override) Mar 11 15:20:44 autohell? :) Mar 11 15:20:45 I think the better route would be to find the root cause of the problem, rather than trying to work around it Mar 11 15:20:51 this is clearly a bug somewhere. Mar 11 15:21:28 kergoth: true, but at least my pythonese and knowledge of the build system is not good enough for that Mar 11 15:22:09 I'd say leave it alone then. It'll get fixed eventually, and having those recipes around until then isn't a life or death proposition, and will mean not having to remove the cruft you add later on Mar 11 15:22:42 mwester: somehow every once in a while i get lost in autohell, debugging strange autoconf/automake/configure etc issues Mar 11 15:23:03 kergoth: true; maybe RP has an idea for it Mar 11 15:23:40 I love the term; it completely describes my experiences with the auto-* tools. :) Mar 11 15:29:17 sigh...without HAL(maybe with it it's the same story) It doesn't detect my sdcards on eee701 Mar 11 15:29:35 doesn't detect means cat /proc/partition has only sda{1} Mar 11 15:29:50 and nothing in dmesg Mar 11 15:30:13 I've realy too much bugs Mar 11 15:30:23 it worked quite ok the first install time Mar 11 15:34:49 pb_: http://pastebin.ca/1833886 or rather add STAGING_SYSCONFDIR to bitbake.conf? Mar 11 15:36:37 native has STAGING_ETCDIR_NATIVE Mar 11 15:41:06 btw talking about autohell: anyone interested in having a stab at autoconf to bring it from 2.63 to 2.65? Latest tiff beta requires at least 2.64, but I don't feel too confident with autoconf to patch it (and the 2.63 patches did not apply) Mar 11 15:41:26 03Martin Jansa  07org.openembedded.dev * rd25468cd60 10openembedded.git/ (classes/base.bbclass conf/bitbake.conf): Mar 11 15:41:26 base: don't write to system /etc dir, use new STAGING_ETCDIR instead Mar 11 15:41:26 * ETCDIR instead SYSCONFDIR, because there already was STAGING_ETCDIR_NATIVE Mar 11 15:41:26 Signed-off-by: Martin Jansa Mar 11 15:41:26 03Martin Jansa  07org.openembedded.dev * r4f9e9c309b 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Mar 11 15:41:27 pyphonelog: bump SRCREV Mar 11 15:41:27 Signed-off-by: Martin Jansa Mar 11 15:42:38 pb_: feel free to correct me, if I overlooked something.. without this I got http://pastebin.ca/1833869 Mar 11 15:58:37 03Martyn Welch  07org.openembedded.dev * r964958b0b1 10openembedded.git/recipes/util-linux-ng/util-linux-ng.inc: (log message trimmed) Mar 11 15:58:37 util-linux-ng: Correct vi error Mar 11 15:58:37 Commit bbd436a98df5d4a439e519651c38fb48d81e5a78 appears to have added a via Mar 11 15:58:37 error to "pkg_prerm_${PN}": Mar 11 15:58:37 - update-alternatives --remove pivot_root pivot_root.${PN} Mar 11 15:58:37 + update-alternatives --remove pivot_root pivot_root.${PN}i Mar 11 15:58:38 Signed-off-by: Martyn Welch Mar 11 16:07:00 Yay! That ^^^^ would be my first patch to go back. Mar 11 16:07:34 congrats Mar 11 16:08:20 Got a few queued :-) Mar 11 16:36:48 JaMa: ah, yeah, that looks reasonable. Mar 11 16:37:03 thanks for fixing it. Mar 11 16:37:28 hi kergoth Mar 11 16:37:32 hey pb_ Mar 11 16:38:35 eFfeM: there was patch for 2.65 on ML and there is also 2.65 in Poky Mar 11 16:52:11 I keep hitting weird race issues with new staging and multiple BB_N_T... Mar 11 16:52:55 here's the latest one: install: cannot create directory `/OE/arago-tmp/staging/i686-linux/usr/bin/armv5te-none-linux-gnueabi': File exists Mar 11 16:53:41 http://pastebin.org/109932 Mar 11 16:54:13 With that patch I posted still installed? Mar 11 16:57:29 hmm, should "install -d" fail if the directory already exists? this code: install -d ${SYSROOT_DESTDIR}${STAGING_BINDIR_CROSS} Mar 11 16:58:02 dont think so, i thought it was more like mkdir -p Mar 11 16:58:07 * kergoth could be wrong, though Mar 11 16:58:36 Tartarus: ah, I think I have your patch applied in another tree. let me port it over... Mar 11 17:00:18 denix: that error message sounds like it could be someone tried to copy into that dir before it existed, and inadvertantly created a *file* with the name of the directory, but that's just a guess :) Mar 11 17:00:23 * kergoth shrugs Mar 11 17:00:32 kergoth: normally it doesn't fail Mar 11 17:00:52 install -d a/b/c/d works Mar 11 17:00:58 So yes, it's like mkdir -p Mar 11 17:01:01 And, what kergoth said :) Mar 11 17:01:14 kergoth: spot on! Mar 11 17:01:15 denix: and? that doesn't make what i described an impossibility, it makes it dependent upon task execution order.. Mar 11 17:01:22 * kergoth gets caffeine Mar 11 17:02:06 ha, xml2-config shell script got written into that name... Mar 11 17:03:27 hehe, oops Mar 11 17:03:54 when you do a cp somefile /somepath, often best to add a trailing / to the second argument, so it fails if its not an existing directory Mar 11 17:04:07 hehe, http://img.ffffound.com/static-data/assets/6/e4911ee39b3f5228c10ccfd39344b76eda9cfb0c_m.jpg Mar 11 17:05:16 kergoth: good one! :) Mar 11 17:05:45 yeah, there was recent change around binconfig in libxml... Mar 11 17:06:32 meh, binconfig Mar 11 17:07:37 hm, that is odd Mar 11 17:07:48 my other xchat seems to have vanished Mar 11 17:09:44 pb____________________: how many do you have? :) Mar 11 17:10:26 two Mar 11 17:10:37 * denix sometimes also runs up to 3 xchats... :rolleyes: Mar 11 17:10:52 but the one for pb_, although apparently still running (and logged in!) is nowhere to be seen on my desktop Mar 11 17:10:57 just tried restarting metacity but that didn't help Mar 11 17:11:31 * kergoth used to just leave a bip irc proxy going and connected to that from everywhere, but hasn't bothered nowadays Mar 11 17:11:56 oh well, I guess I'll kill him off by hand Mar 11 17:12:31 pb_: sayonara Mar 11 17:13:19 bletch, my tea mug is covered in salty residue Mar 11 17:13:25 dishwasher must be malfunctioning again Mar 11 17:18:44 denix: I run one irssi but several windows in it and 3 irc servers. also few channels of same name on different networks Mar 11 17:22:43 kergoth: found the problem with libxml... Mar 11 17:23:06 Tartarus: sorry, false alarm, this one wasn't staging related :) Mar 11 17:24:53 heh Mar 11 17:24:58 Apply that patch still please :) Mar 11 17:25:09 I'll repost it and bug RP sometime soon Mar 11 17:25:15 denix: what's wrong with libxml? Mar 11 17:34:43 Tartarus: I'll do Mar 11 17:36:08 JaMa: with binconfig enabled, it unmasked that xml2-config was not installed correctly Mar 11 17:36:37 bye Mar 11 17:36:54 JaMa: I have the fix, I'll push it in a minute - I'm verifying it builds :) Mar 11 17:47:32 denix: thanks Mar 11 17:52:20 03Denys Dmytriyenko  07org.openembedded.dev * r33fc60bdc0 10openembedded.git/recipes/libxml/libxml2.inc: Mar 11 17:52:20 libxml: actually install xml2-config in STAGING_BINDIR_CROSS Mar 11 17:52:20 Make sure STAGING_BINDIR_CROSS is a directory and not the target filename Mar 11 17:52:20 for xml2-config. Mar 11 17:52:20 Signed-off-by: Denys Dmytriyenko Mar 11 17:59:33 gm Mar 11 18:02:29 Glad to see git's back - time to remove the admin message from the channel status? Mar 11 18:09:11 I think do_split_packages() shouldn't be in -native packages, right? Mar 11 18:11:54 otherwise there are multiple providers of the same runtime entity from pkg and pkg-native... Mar 11 18:23:20 hrw,thanks for the 0.65 reference (was afk) Mar 11 18:26:18 is a do_install which only contains oe_runmake DESTDIR="${D}" install useful? (no autotools) or can i just remove do_install? Mar 11 18:27:21 its useful if you plan to install anything.. Mar 11 18:27:35 base.bbclass doesn't give you a functional install, autotools.bbclass does Mar 11 18:31:35 denix: hm, I guess the native packages should be getting split with "-native" in their names. probably a few hardcoded strings need changing for ${PN} or something. Mar 11 18:32:05 just not running splitting for -native would be a solution, I guess, but it would be nice to keep the native packaging as a mirror of the target packaging in so far as that's feasible Mar 11 18:32:50 pb__: AFAIK -native packages are not generated, so why bother? Mar 11 18:33:04 they are in toolchain-desuck Mar 11 18:33:20 nice name :) Mar 11 18:33:36 in the long term, I think it would be desirable to package everything consistently. Mar 11 18:33:46 sdk generation is much easier that way Mar 11 18:33:57 ditto staging management Mar 11 18:38:53 pb__: is there a var or function like base_prune_suffix() which actually extracts the suffix? something like PN minus BPN? Mar 11 18:43:34 kergoth: ah ok, i assumed base.bbclass would also give an install, but when I looked I could not find it, thought I was overlooking things (not too familiar with the whole bbclass structure) Mar 11 18:44:06 non-autotools packages can't be relied upon even having an install, and if they do, they can't be relied upon using DESTDIR Mar 11 18:44:19 many use a different name for that variable, or don't have one at all Mar 11 18:44:49 ah ok, understood Mar 11 19:08:05 re Mar 11 19:11:26 denix: if all you want is the suffix, why not just use bb.data.inherits_class()? Mar 11 19:13:04 morning :) Mar 11 19:13:26 kergoth: with little bit of extra code that should work :) Mar 11 19:13:40 i think we should enhance inherits_class to behave like isinstance Mar 11 19:13:51 with isinstance, the second argument can be a string -or- an iterable of strings Mar 11 19:13:58 erm, s/string/class/ Mar 11 19:14:20 would be nice to do if bb.data.inherits_class(["native", "cross", "sdk"], d) Mar 11 19:14:22 or something Mar 11 19:22:13 could a hyperviser be run on the dsp, to monitor execution on the arm core? Mar 11 19:24:05 it would be interesting to see a "trusted" platform running on the BB... Mar 11 19:25:46 Brokie: if you ask TI nicely they will give you omap3 with TPM enabled Mar 11 19:26:03 nono Mar 11 19:26:08 not TPM module Mar 11 19:26:20 Trusted Execution Platform Mar 11 19:26:38 XorA: hey.. Timur has answered but I fear too late :) Mar 11 19:27:06 Jay7: I got it, and I shall send next week Mar 11 19:27:13 XorA: ah, good :) Mar 11 19:27:16 http://software.intel.com/en-us/articles/intel-trusted-execution-technology-a-primer/ <--- intel's crap version Mar 11 19:27:22 Jay7: working from home next week so courier can come, that has been the issue lately Mar 11 19:27:47 * XorA needs to find some bubbe wrap :-) Mar 11 19:27:54 Like this... Mar 11 19:28:07 A Memory ACL is written to a table Mar 11 19:30:08 Areas designated "1" are DATA areas, Areas designated "2" are CODE areas, Areas designated "4" are Protected Stack Frame, Areas designated "0" are Free areas (cannot write to them, till they are allocated by the visor) Mar 11 19:31:32 So, each step, is checked by the execution visor, to make sure no OOB data is executed, no SIB code is written to till declared free, and reallocated.... Mar 11 19:32:36 and noone can write to the stack without passing data via jmp, bra, jre, or jra, POP, PUSH, or CLMP Mar 11 19:40:28 of course the execution visor runs DSP side... Mar 11 20:14:38 eagle Mar 11 20:31:43 03Thomas Zimmermann  07org.openembedded.dev * r907c4dbbcc 10openembedded.git/recipes/pingus/ (files/pingus-gta012.sh pingus_0.7.2.bb): Mar 11 20:31:43 pingus: update wrapper script for gta{01|02} Mar 11 20:31:43 * old script does not work with newer xrandr anymore Mar 11 20:31:43 Signed-off-by: Thomas Zimmermann Mar 11 20:57:22 ok guys, I goofed, and need some help to fix it: I've just pushed my branch again but did it wrong so instead of pushing to eFfeM/sanity I made a new branch sanity. how can I get rid of that new branch and how can I push to eFfeM/sanity ? Mar 11 21:04:08 eFfeM: you can't delete branches in oe, ask admins Mar 11 21:04:59 eFfeM: git push : Mar 11 21:05:58 that's the syntax when localbranch name differs from remotebranch Mar 11 21:06:01 see also git push --help :-} Mar 11 21:06:34 denix ok, thanks, mistyped things as when pushing it i was told to use my id as prefix and forgot when pushing again Mar 11 21:06:44 yeah, RTFM is always suggested :) Mar 11 21:06:47 who can delete the faulty branch ? Mar 11 21:07:33 the git doc is not always too userfriendly :-( or correct, it tells I can delete a branch with git pull origin :branch (or something like taht) Mar 11 21:07:58 anybody with admin privs on the server. I think those überhackers include zecke, cbrake and rp, maybe mickey too Mar 11 21:10:04 as for deleting branches, the git docs are correct, it's just that this feature is disabled on the oe servers. Mar 11 21:10:18 ah ok Mar 11 21:10:38 to keep tards from destroying thousands of man hours from being destroyed by a punk Mar 11 21:10:43 eFfeM1: as I found out today, remote delete is really powerful and a pain in the ass to fix when you fuck it up Mar 11 21:14:48 XorA: yeah Mar 11 21:15:09 i would expect to be able to kill the branch I created Mar 11 21:15:22 read the help for git push --mirror more carefully before using Mar 11 21:16:02 eFfeM1: that was discussed, but gitosis doesnt export $KEYNAME to git so hooks dont know which user it is Mar 11 21:16:06 eFfeM, I think the idea was to protect your work from account theft Mar 11 21:16:16 or that Mar 11 21:16:30 ah ok Mar 11 21:17:22 XorA: anything that alters the raw objects or history is dangerous, including rebase... Mar 11 21:34:54 cbrake: rp if you see this message: can you remove the sanity branch for me (NOT: eFfeM/sanity !) thanks alot, will take more care next time Mar 11 21:35:03 and calling it a day now Mar 11 21:35:16 at least almost Mar 11 21:40:11 eFfeM1: send email to mickey|zzZZzz Mar 11 21:40:19 he will delete it for you Mar 11 21:42:19 ok, thanks Mar 11 21:43:20 What's the best way to verify packages after converting a recipe to using BBCLASS? Mar 11 21:43:36 I'm doing my first conversion with the zlib recipe. Mar 11 21:43:57 er, that's BBCLASSEXTEND, of course Mar 11 21:46:42 anyone using angstrom on mistral's omap3evm? Mar 11 21:50:26 bitbake master is choking on me Mar 11 21:50:29 hmm Mar 11 22:02:01 khem, where is it choking? Mar 11 22:02:24 oneshal, yes, the mistral evm works... Mar 11 22:06:50 Brokie: cool, just got one in today, need to build angstrom for it now Mar 11 22:20:37 Brokie: it sits there doing nothing Mar 11 22:21:45 hm.. udev_1.5.1 fails on do_compile here Mar 11 22:25:38 ant__: what error message do you get Mar 11 22:25:54 is on linking Mar 11 22:26:02 tyle=gnu -o extras/firmware/firmware extras/firmware/firmware.o -L/oe/build/tmp/staging/armv5te-angstrom-linux-gnueabi/usr/lib libudev/.libs/libudev-private.a Mar 11 22:26:02 | make[1]: *** [all-recursive] Error 1 Mar 11 22:26:26 starts with: arm-angstrom-linux-gnueabi-libtool: link: arm-angstrom-linux-gnueabi-gcc Mar 11 22:26:27 the real message somewhere else Mar 11 22:26:37 search for error: Mar 11 22:26:59 oops..I don't have the log onscreen, mom Mar 11 22:27:41 before you ask, I just pulled Mar 11 22:27:52 it's udev_151-r4 Mar 11 22:28:17 khem: Mar 11 22:28:18 included from extras/ata_id/ata_id.c:33: Mar 11 22:28:18 d/tmp/staging/armv5te-angstrom-linux-gnueabi/usr/include/scsi/scsi.h:145: error: expected specifier-qualifier-list before 'u8' Mar 11 22:28:57 ant__: which libc? Mar 11 22:29:06 glibc Mar 11 22:29:19 iirc koen recently readded scsi headers Mar 11 22:29:38 to glibc ? why Mar 11 22:30:08 [oe-commits] Koen Kooi : linux-libc-headers: reinstate scsi headers in staging that got lost with the new-style staging conversion Mar 11 22:30:34 then there are 3-4 udev commits... Mar 11 22:30:41 scary ... Mar 11 22:32:33 I'm cleaning and logging the build on tinderbox now Mar 11 22:33:14 ah Mar 11 22:33:28 so its otherway around Mar 11 22:33:44 the headers came from libc but now they come from kernel Mar 11 22:36:31 ant__: can you paste your d/tmp/staging/armv5te-angstrom-linux-gnueabi/usr/include/scsi/scsi.h somewhere Mar 11 22:36:44 see..now I did -c clean udev Mar 11 22:36:55 on rebuild the new libc headers are built.. Mar 11 22:36:58 these headers were coming from libc in the past Mar 11 22:37:05 and now kernel is exporting them too Mar 11 22:37:09 http://tinderbox.openembedded.org/builds/64216/ Mar 11 22:37:14 so I think we have to fix libc Mar 11 22:37:39 and for newer kernels which export them we should have wrapper headers in glibc which call the kernel headers Mar 11 22:37:56 I'm checking headers in staging Mar 11 22:38:27 ant__: I guess you need http://patchwork.kernel.org/patch/38102/ Mar 11 22:38:49 what is your kernel version for linux-libc-headers Mar 11 22:39:22 NOTE: Running task 708 of 934 (ID: 441, /oe/org.openembedded.dev/recipes/linux-libc-headers/linux-libc-headers_2.6.31.bb, do_build) Mar 11 22:39:31 scsi.h: Mar 11 22:39:32 http://fr.pastebin.ca/1834632 Mar 11 22:40:45 ok so your header is coming from kernel-headers Mar 11 22:41:01 actually I think it just overwrote what glibc has installed for you Mar 11 22:41:28 may be you can bitbake -c populate_staging glibc Mar 11 22:41:37 and see if it reinstalls the one from glibc Mar 11 22:42:07 scsi headers from kernel are not so user friendly yet Mar 11 22:42:17 khem: btw my kernel is still 2.6.26 Mar 11 22:42:36 ant__: that does not matter here Mar 11 22:42:55 problem is if header should come from glibc or from linux-libc-headers Mar 11 22:44:10 actually I will ask you to apply the patch I suggested to linux-libc-headers and rebuild them Mar 11 22:44:17 my bitbake is constipated Mar 11 22:44:23 so I can do no testing atm Mar 11 22:44:29 er Mar 11 22:44:38 if the kernel installs them for userland, they should be :) Mar 11 22:44:55 headers_install only after 2.6.27 afaik Mar 11 22:45:05 Tartarus: userspace has for long used the one glibc provided Mar 11 22:45:25 so we have to fix glibc first before starting to use kernel ones now Mar 11 22:45:31 if 2.6.33 only would be a bit better... Mar 11 22:46:36 ant__: if you apply the above patch to linux-libc-headers_2.6.31.bb it should fix your issue Mar 11 22:47:01 ah..now glibc fails too :/ Mar 11 22:47:40 hmm where ? Mar 11 22:48:07 Er Mar 11 22:48:11 Are they conflicting now? Mar 11 22:48:22 Or always conflicted in names? Mar 11 22:48:22 thats my guess Mar 11 22:48:38 they are different contents and same location and names Mar 11 22:48:40 hmm Mar 11 22:48:51 maybe the rm needs to be put back in, but compat w/ the nativesdk stuff Mar 11 22:49:00 he http://tinderbox.openembedded.org/public/logs/task/5314230.txt Mar 11 22:49:23 khem: I'm applying the patch Mar 11 22:49:38 ant__: glibc needs kernel-headers Mar 11 22:49:53 and when you tried to build them the build did not go thru Mar 11 22:49:58 so it deinstalled them Mar 11 22:50:06 but did not install back into staging Mar 11 22:50:20 so now when you try to build glibc gcc finds no kernel headers Mar 11 22:50:22 ah, ok, as klibc (still needing kernel headers for a while) Mar 11 22:50:48 ant__: apply the patch rebuild the kernel-headers and stage them Mar 11 22:50:57 yep Mar 11 22:51:01 then I think udev will build Mar 11 22:51:15 I think glibc might complain Mar 11 22:51:30 but you have it built already so ..you may care less Mar 11 22:51:44 but I wonder what would happen when building from scratch Mar 11 22:51:50 gj on the git server Mar 11 22:51:52 khem: I can rebuild from scratch and see...perhaps is best option Mar 11 22:52:04 ant__: to verify Mar 11 22:52:15 just do linux-libc-headers patched first Mar 11 22:52:20 and try udev Mar 11 22:52:31 if it builds then you know the fix Mar 11 22:52:40 and then you can try to build from scratch Mar 11 22:52:43 et.c etc Mar 11 22:53:38 can someone change status? GIT SERVER IS NO LONGER DOWN... Mar 11 23:02:05 khem: as I expected udev build is now stopped by glibc failure...but it builds Mar 11 23:04:32 khem: pls have a look at the latest patch in oe right now... Mar 11 23:30:18 ant__: Brokie admins may not be here atm Mar 11 23:30:33 ant__: glibc failure ? Mar 11 23:30:46 ant__: what sort of failure are you getting Mar 11 23:31:18 I think once you have linux-libc-headers staged and also staged glibc-initial you should be good Mar 11 23:32:12 khem: glibc as failing with configure: error: C preprocessor "arm-angstrom-linux-gnueabi-gcc -E" fails sanity check Mar 11 23:32:20 so I wiped tmp Mar 11 23:33:44 ant__: ok so you are doing build from scratch now ? Mar 11 23:34:09 yes, with patched linux-libc-headers Mar 11 23:34:40 but have you seen the specular patch in OE? Mar 11 23:35:26 http://article.gmane.org/gmane.comp.handhelds.openembedded.scm/32816 Mar 11 23:36:47 thats ugly Mar 11 23:36:52 dont apply that Mar 11 23:39:35 no no Mar 11 23:39:52 at first I thought was in dev branch, isn't Mar 11 23:39:57 push the one I suggested it should build for you Mar 11 23:40:04 its a private branch Mar 11 23:40:16 but I am scared that its gonna come in as a huge merge Mar 11 23:40:23 so it must be flagged Mar 11 23:40:55 khem: obviously same applies to 2.6.32 headers Mar 12 00:39:16 anyone know the proper way for a filesystem to mount debugfs? I would think /etc/fstab but my ubuntu sys has it mounted yet its not in /etc/fstab Mar 12 00:46:49 Um Mar 12 00:46:54 depends? :) Mar 12 00:53:55 tharvey, manually Mar 12 00:54:22 anything that does it automatically is not really in the spirit of debugfs Mar 12 00:55:03 if you have a fstab entry it should definitely be 'noauto' Mar 12 00:57:31 fair enough - just curious how my ubuntu system was doing it - didn't see it in any init script Mar 12 01:03:20 Just checking to see that the git server and any other services on melo are working okay... Mar 12 02:40:12 ka6sox-work: git is ok Mar 12 02:59:26 http://imggmi.info/DSC-1268362376.jpg/ do my tits look big? **** ENDING LOGGING AT Fri Mar 12 02:59:57 2010