**** BEGIN LOGGING AT Fri Feb 27 02:59:57 2009 Feb 27 04:49:53 I'm getting an error in what seems to be the final stages of the build: Feb 27 04:50:03 | Collected errors: Feb 27 04:50:03 | ERROR: Cannot satisfy the following dependencies for task-base-gumstix: Feb 27 04:50:03 | tslib Feb 27 04:50:18 Is this something missing on my host? Feb 27 04:51:20 when I run "bitbake tslib" there are some QA errors but it completes Feb 27 06:28:27 someone alive? Feb 27 07:12:05 otavio: Guess not :) Feb 27 07:19:57 03Koen Kooi  07org.openembedded.dev * r322be639f1 10openembedded.git/conf/checksums.ini: checksums: add more sums Feb 27 07:19:58 03Koen Kooi  07org.openembedded.dev * rae29da85f9 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 27 07:45:08 good morning Feb 27 07:57:26 morning mckoan ! Feb 27 07:57:43 mckoan, i saw the news on LinuxDevices !!! Feb 27 08:00:46 gremlin[it]: hi, thank you. I hope this could be a stimulus for OE spreading :-D Feb 27 08:05:22 we hope ! Feb 27 08:05:50 gremlin[it]: I'll do my best to achieve it ;-) Feb 27 08:31:03 hi i'm having trouble baking the latest releases of gstreamer and its plugins due to their libtool upgrade Feb 27 09:18:26 mhhh there some technical trouble about gcc-2.95 or isn'tin OE just cause no one need it ? Feb 27 09:28:34 ping flameman Feb 27 09:29:39 :) Feb 27 09:29:43 hi Feb 27 09:30:22 i'm here usually but when I'm at work maybe I'm busy on other tasks ... Feb 27 09:31:02 as here, i am busy for a kernel module i am writing Feb 27 09:31:08 and it has issue Feb 27 09:36:08 gremlin[it]: nobody needs it badly enough to have written a recipe for it. it wouldn't be impossible to add, but it wouldn't be trivial either. Feb 27 09:36:37 pb__, why not trivial ? Feb 27 09:36:56 moin Feb 27 09:36:56 gremlin[it]: the gcc build system is somewhat complicated and I think it changed significantly between 2.95 and 3.x. Feb 27 09:37:27 hi woglinde Feb 27 09:37:31 * pb__ go to the office now Feb 27 09:37:32 bbiab Feb 27 09:37:37 till later pb Feb 27 09:37:40 mhh ok ... good work ! Feb 27 09:38:27 btw: i need it for a i386 machine so maybe i can use it as an external gross compiler Feb 27 09:44:28 03Jan Lübbe  07fso/milestone5.5 * r30cf44b6ce 10openembedded.git/packages/freesmartphone/frameworkd_git.bb: frameworkd: depend on pyalsaaudio for ADC/DAC configuration Feb 27 10:03:38 03Koen Kooi  07org.openembedded.dev * raf7ec0ac75 10openembedded.git/packages/dsplink/gstreamer-ti_svn.bb: gstreamer-ti: catch up with site rename, pass in username and passwd Feb 27 10:10:06 * * OE Bug 5045 has been created by di(AT)fh-wedel.de Feb 27 10:10:08 * * Config_heavy.pl contains options specific to the host compiler rather than the cross compiler Feb 27 10:10:10 * * http://bugs.openembedded.net/show_bug.cgi?id=5045 Feb 27 10:12:57 re Feb 27 10:13:05 gremlin[it]: yah, setting it up externally would be the easiest thing Feb 27 10:13:29 thanks pb_ Feb 27 10:25:53 flameman: are you writing for SA1100? Look at thesing's work here: http://git.linuxtogo.org/?p=thesing/collie.git;a=summary Feb 27 10:39:58 is it possible to do a kernel menuconfig using OE ? Feb 27 10:40:34 yes, bitbake -c menuconfig virtual/kernel Feb 27 10:41:03 ant_work: I get a weird error about X11 doing such command Feb 27 10:41:03 then copy the config out of workdir ! Feb 27 10:41:08 he he Feb 27 10:41:08 thanks Feb 27 10:41:18 try with XAuthority Feb 27 10:41:43 I think this error sorted out after env-mask patches to bb Feb 27 10:42:09 ant_work: ERROR: function do_menuconfig failed + | cannot open display: Feb 27 10:42:18 yes, just check the XAuth settings of your current user Feb 27 10:42:22 but I'm in ssh console ! Feb 27 10:42:37 well, it needs to open a new xterminal Feb 27 10:42:41 window Feb 27 10:42:51 (on Gnome) Feb 27 10:43:16 why is doesn't work as a normal make menuconfig? Feb 27 10:44:05 i'm having trouble baking the latest releases of gstreamer and its plugins due to their libtool upgrade. has anybody tried that yet? Feb 27 10:44:30 ant_work: BTW if I call xterm a window opens Feb 27 10:46:02 mckoan: I googled for 'xauth cannot open display' Feb 27 10:47:31 mckoan: iirc X11 forwarding won't work without xauth Feb 27 10:48:12 ant_work: do you mean I neet to set ForwardX11 yes ? Feb 27 10:48:24 just xhost yourname Feb 27 10:48:53 do you mean in /etc/ssh/ssh_config ? Feb 27 10:49:07 no, as command Feb 27 10:49:36 light$ xhost +this.local.host Feb 27 10:49:57 I did xhost koan Feb 27 10:50:08 koan being added to access control list Feb 27 10:50:15 try now Feb 27 10:50:21 bitbake virtual/kernel -cmenuconfig Feb 27 10:51:29 X11 connection rejected because of wrong authentication. Feb 27 10:51:30 ERROR: function do_menuconfig failed Feb 27 10:51:43 (BTW i'm into screen session) Feb 27 10:52:03 hmm Feb 27 10:53:07 xhost uses an host list, xauth magic cookies Feb 27 10:53:48 is it a command? Feb 27 10:54:13 just for fun, try $ xhost + Feb 27 10:54:22 should diable access-control Feb 27 10:54:35 lumdidum Feb 27 10:54:41 hey wog Feb 27 10:54:44 he ant Feb 27 10:55:09 xhost + Feb 27 10:55:10 access control disabled, clients can connect from any host Feb 27 10:55:16 :/ Feb 27 10:55:22 change this, afterward! Feb 27 10:55:31 I'm puzzled, how do yo usually deal with this? Feb 27 10:55:57 you added your user, not your host Feb 27 10:56:43 iirc the env-var 'xhost' is now filtered...have to cjeck...long ago the issue was not present Feb 27 10:59:10 no way Feb 27 10:59:28 anybody tried 11:53 < ant_work> xhost uses an host list, xauth magic cookies Feb 27 10:59:44 sorry Feb 27 11:00:14 anybody successfully used: bitbake virtual/kernel -cmenuconfig ? Feb 27 11:01:25 ant_work, just cloning the git repositories :P Feb 27 11:02:26 ant_work, you could be happy to know i successfully built all the stage4, now i am entering the stage5 Feb 27 11:02:52 flameman: cpu temp? :-) Feb 27 11:03:01 cool :P Feb 27 11:03:15 sa1100@180Mhz is cool Feb 27 11:03:42 this is the good news Feb 27 11:04:03 the bad news is ... the other project is in stale Feb 27 11:04:23 cause i can't figure out what's wrong with the sandisk mmc card v1 Feb 27 11:04:34 v2 works, v1 doesn't Feb 27 11:04:41 well .... i will fix it Feb 27 11:04:52 flameman: on insanelymac forum I have been flamed by a brillant developer, a bit rude though ;-) Feb 27 11:05:34 as koen from #angstrom ? Feb 27 11:05:43 :-) Feb 27 11:05:49 he banned me just cause he doesn't like my project Feb 27 11:05:52 you see..it just happens Feb 27 11:06:07 yes Feb 27 11:06:17 stupids are just around Feb 27 11:06:32 i mean; to cut of a conversation is stupid Feb 27 11:06:57 (to cut off = to ban) Feb 27 11:07:44 ant_work, what was you talking about on the insanelymac forum ? Feb 27 11:07:54 we're OT Feb 27 11:07:58 go private Feb 27 11:12:27 essen Feb 27 11:12:30 ups Feb 27 11:12:34 lunch Feb 27 11:27:01 mckoan: have you solved it? Feb 27 11:33:36 hey florian, kc = kitchen Feb 27 11:33:38 ? Feb 27 11:37:56 good morning Feb 27 11:38:45 ant_work: no "kernel concepts"... but I have to admit that I have a dedicated computer for the kitchen ;) Feb 27 11:39:24 he ha Feb 27 11:39:46 embedded in the oven? Feb 27 11:40:47 I've already seen DVD for TV and cookbooks hanging around in the new fashion kitchens Feb 27 11:41:31 ant_work: its an old SGI o2... that would be a little bit too big for embedding :) Feb 27 11:41:46 I'll have to buy a smaller one for my wife: hopefully the refrigerator will send alarm before being empty! Feb 27 11:42:20 florian: in the future will be fun with RFID's Feb 27 11:42:25 for foods too Feb 27 11:42:41 03Jeremy Lainé  07org.openembedded.dev * r261a757544 10openembedded.git/packages/udev/ (files/boc01/mount.sh udev_124.bb): udev-124: remove boc01-specific mount.sh Feb 27 11:49:08 hmm, truly amazing - I have built meta-toolchain and there is no gcc inside. There should be, right? Feb 27 11:51:56 how to force bitbake every time fetch updated svn version on building a package? Feb 27 11:52:21 looks like opkg's gcc-cross-sdk.list have paths cut at 99 character o.O Feb 27 11:52:22 It seems that it uses a version it fetched one time in the past Feb 27 11:53:09 booxter: Maybe CVS_TARBALL_STASH = "" in local.conf Feb 27 11:53:37 or similar Feb 27 11:56:31 ant_work: :) Feb 27 11:56:45 pocek: this sounds like there is something really broken Feb 27 11:58:07 I got it. Just specified SRCREV=${AUTOREV} Feb 27 11:58:11 seems to work Feb 27 11:58:57 devil inside Feb 27 11:59:10 booxter, I was about to say that :) Feb 27 12:01:20 http://lists.openmoko.org/pipermail/openmoko-devel/2007-September/001507.html Feb 27 12:01:43 (I somehow kept the link...) Feb 27 12:04:10 03Jeremy Lainé  07org.openembedded.dev * r4930180982 10openembedded.git/classes/qtopia4core.bbclass: qtopia4core.bbclass: depend qt-embedded, qtopia-core is obsolete Feb 27 12:04:27 03Jeremy Lainé  07org.openembedded.dev * r5353963f92 10openembedded.git/packages/qt4/ (5 files in 2 dirs): qt4: get rid of bogus qtopia-core-4.4.3 package Feb 27 12:22:33 hi effem Feb 27 12:23:45 hi woglinde Feb 27 12:40:54 hi gnutoo Feb 27 12:41:17 woglinde, hi I didn't post the mail yet...i'll do it today Feb 27 12:44:35 ni Gnutoo Feb 27 12:44:39 hi Feb 27 12:44:41 hi Feb 27 12:51:00 03Jeremy Lainé  07org.openembedded.dev * r8aaaf61339 10openembedded.git/packages/qt4/qt_packaging.inc: qt_packaging.inc: make phonon4-dev honour QT_LIBINFIX Feb 27 12:51:48 03Jeremy Lainé  07org.openembedded.dev * r97547278b0 10openembedded.git/packages/qt4/ (qt-embedded_4.4.3.bb qtopia-core.inc qtopia-core_4.3.3.bb): qt4: split qt-embedded and qtopia-core Feb 27 13:02:50 03Jeremy Lainé  07org.openembedded.dev * r8f50054be5 10openembedded.git/ (3 files in 2 dirs): qt-embedded: restore qtopia4core.bbclass, add qt4e.bbclass Feb 27 13:11:51 03Leon Woestenberg  07mickey/org.openembedded.dev * rcf89155284 10openembedded.git/packages/rt-tests/rt-tests_0.21.bb: rt-tests: Moved from 0.21 to 0.28. Feb 27 13:11:52 03Cliff Brake  07mickey/org.openembedded.dev * rd42294dad3 10openembedded.git/conf/checksums.ini: checksums.ini: add checksum for pengutronix canutils Feb 27 13:11:53 03Theodore A. Roth  07mickey/org.openembedded.dev * raf0e9d42be 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: Feb 27 13:11:53 getting_oe.xml: Update documentation. Feb 27 13:11:55 * New section about directory structure. Feb 27 13:11:57 * Update 'Getting BitBake' section. Feb 27 13:11:59 Resubmitted against .dev instead of .documentation. Feb 27 13:12:01 Signed-off-by: Theodore A. Roth Feb 27 13:12:03 03Theodore A. Roth  07mickey/org.openembedded.dev * r83ef846a36 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: Feb 27 13:12:06 getting_oe.xml: Update getting openembedded section. Feb 27 13:12:08 Resubmitted against .dev branch instead of .documentation. Feb 27 13:12:10 getting_oe.xml: Update getting openembedded section. Feb 27 13:12:12 Signed-off-by: Theodore A. Roth Feb 27 13:12:14 03Theodore A. Roth  07mickey/org.openembedded.dev * rf339e1b9b8 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: Feb 27 13:12:17 getting_oe.xml: Add Configuring OpenEmbedded section. Feb 27 13:12:19 This patch fills in the stub for the "Configuring OpenEmbedded" section. Feb 27 13:12:29 Signed-off-by: Theodore A. Roth Feb 27 13:12:31 03Koen Kooi  07mickey/org.openembedded.dev * r587bd013e9 10openembedded.git/conf/checksums.ini: checksums: add fakeroot 1.12.1 Feb 27 13:12:34 03Koen Kooi  07mickey/org.openembedded.dev * rfb87b496ca 10openembedded.git/packages/tasks/task-gstreamer-ti.bb: task-gstreamer-ti: add gst-plugins-bad Feb 27 13:12:39 03Mike Westerhof  07mickey/org.openembedded.dev * rb295242c18 10openembedded.git/ (5 files in 3 dirs): SlugOS: reinstate fakeroot 1.9.6 and set as the PREFERRED_VERSION Feb 27 13:12:42 03Philip Balister  07mickey/org.openembedded.dev * rdea80abd5c 10openembedded.git/conf/checksums.ini: Feb 27 13:12:45 checksums.ini : Add checksum for libnetfitler_queue-0.0.16.tar.bz2. Feb 27 13:12:47 * I know I am slow adding this. Feb 27 13:12:49 03Philip Balister  07mickey/org.openembedded.dev * rc73b66a463 10openembedded.git/conf/checksums.ini: checksums.ini : Add checksum for base-passwd-3.5.20. Feb 27 13:12:58 03woglinde  07mickey/org.openembedded.dev * r216894213e 10openembedded.git/packages/libpng/ (libpng-native_1.2.35.bb libpng_1.2.35.bb): Feb 27 13:13:01 libpng: add version 1.2.35 Feb 27 13:13:03 * fixes security issues Feb 27 13:13:05 03Khem Raj  07mickey/org.openembedded.dev * r16fb36ea23 10openembedded.git/conf/checksums.ini: checksums.ini: Add checksums for fakeroot_1.12.1.tar.gz and netbase-4.21.tar.gz Feb 27 13:13:10 03Frans Meulenbroeks  07mickey/org.openembedded.dev * rbaa37f3a42 10openembedded.git/ (4 files in 3 dirs): pmount: updated to 0.9.18 Feb 27 13:13:13 03Steffen Sledz  07mickey/org.openembedded.dev * r3c3d070084 10openembedded.git/conf/checksums.ini: checksums.ini: checksums for another e2fsprogs-1.41.4 url added Feb 27 13:13:16 03Jeremy Lainé  07mickey/org.openembedded.dev * r8a7418580d 10openembedded.git/packages/qt4/ (5 files in 2 dirs): qt4: get rid of bogus qtopia-core-4.4.3 package Feb 27 13:13:23 03Jeremy Lainé  07mickey/org.openembedded.dev * rfbd582b8de 10openembedded.git/packages/udev/ (files/boc01/mount.sh udev_124.bb): udev-124: remove boc01-specific mount.sh Feb 27 13:13:26 03Khem Raj  07mickey/org.openembedded.dev * r80a378e8a7 10openembedded.git/conf/checksums.ini: checksums: Add base-passwd checksum Feb 27 13:13:31 03Jeremy Lainé  07mickey/org.openembedded.dev * rf5cf1567d7 10openembedded.git/packages/qt4/qt_packaging.inc: qt_packaging.inc: make phonon4-dev honour QT_LIBINFIX Feb 27 13:13:36 03Koen Kooi  07mickey/org.openembedded.dev * rb6ccea256e 10openembedded.git/packages/dsplink/gstreamer-ti_svn.bb: gstreamer-ti: bump to r127 Feb 27 13:13:39 03Philip Balister  07mickey/org.openembedded.dev * rd3aaa3b121 10openembedded.git/conf/checksums.ini: tzcode-native : Add checksums for txcode and tzdata. Feb 27 13:13:44 03Theodore A. Roth  07mickey/org.openembedded.dev * rbe49495243 10openembedded.git/docs/usermanual/chapters/getting_oe.xml: getting_oe.xml: Update "Building Software" section. Feb 27 13:13:47 (26 lines omitted) Feb 27 13:14:24 what's up with CIA? Feb 27 13:15:17 note that it are commits to mickey/org.openembedded.dev not to org.openembedded.dev ! Feb 27 13:16:35 http://wiki.openembedded.net/snapshots/ - we need to kill that directory Feb 27 13:17:14 ah monotone Feb 27 13:17:27 guess it is only of historical interest Feb 27 13:17:43 yes, horror museum Feb 27 13:18:53 morning btw Feb 27 13:18:59 hrw: please remove the embedix cruft when/if possible Feb 27 13:19:20 he...somebody asked for gcc-2.95 today.... Feb 27 13:21:00 guess he can use the mtn snapshots as well :-) Feb 27 13:28:27 ant_work: I haven't solved it yet, seems that anybody tried bitbake virtual/kernel -cmenuconfig Feb 27 13:28:39 and it does not work Feb 27 13:29:08 hm...last time I did it was 1st feb Feb 27 13:29:19 then I reinstalled all on host Feb 27 13:29:43 I'll try later, but 'xhost +local' should fix the things iirc Feb 27 13:32:53 mckoan: http://www.gumstix.net/wiki/index.php?title=Kernel_Reconfiguration Feb 27 13:33:04 if all above fails I dunno Feb 27 13:33:42 what's on your host? Gentoo ? Feb 27 13:44:10 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * rb697819f7e 10openembedded.git/packages/linux/linux.inc: linux.inc: fix the infamous 'install ./-m0644: file not found' error Feb 27 13:46:47 ant_work: this is the best howto I've read about Kernel Reconfiguration, but this operation is defintely a pain with OE Feb 27 13:47:52 ~curse OE Kernel Reconfiguration Feb 27 13:49:17 where's the problem? Feb 27 13:49:24 bitbake -c devshell virtual/kernel Feb 27 13:49:38 make ARCH=whatever CROSS_COMPILE=whatever menuconfig Feb 27 13:49:57 May you be reincarnated as a Windows XP administrator, OE Kernel Reconfiguration ! Feb 27 13:50:04 then save the .config somewhere Feb 27 13:50:09 mickey|zzZZzz: devshell and menuconfig wants to run gnome-terminal (or konsole or xterm depends on setup) Feb 27 13:50:10 and build it agin Feb 27 13:50:24 yes? Feb 27 13:50:44 mickey|zzZZzz: same problem -> X11 connection rejected because of wrong authentication. Feb 27 13:50:47 ERROR: function do_devshell failed Feb 27 13:50:52 ah well Feb 27 13:51:00 where's the OE problem in that :) Feb 27 13:51:01 my way: bitbake virtual/kernel -cconfigure;sh; cd tmp/work/MACHINE*/*kernel*/$S/;ARCH=arm make menuconfig Feb 27 13:51:05 I'm in ssh console with screen session Feb 27 13:52:14 mickey|zzZZzz: maybe is not an OE problem but involves it strongly Feb 27 13:53:17 also gunstix Kernel Reconfiguration page describe it as a complex activity Feb 27 13:53:32 well Feb 27 13:53:39 at the end of the day it's a cd Feb 27 13:53:44 followed by a make Feb 27 13:53:53 and people used to configure and build a kernel standalone (not in OE tree) are used to have less pain Feb 27 13:53:55 so i fail to see the complexity Feb 27 13:54:15 you just need to remember to have OE build it first Feb 27 13:54:17 I'm not the only person saying that Feb 27 13:54:20 then you can move into the directory Feb 27 13:54:26 and edit the config in every possible way Feb 27 13:54:41 on the other hand, if all you want is to compile a kernel Feb 27 13:54:43 then do it standalone Feb 27 13:54:49 OE is certainly overkill for that Feb 27 13:55:24 my doubt are: 1. who most of OE skilled users do that? Feb 27 13:55:28 OE as all other "full" packages is an offer. if you chose to accept that offer, you lose a bit of control. that's the deal. Feb 27 13:55:39 2.could be possible to imagine an easier way? Feb 27 13:55:57 distinction of tasks is an important thing in your life Feb 27 13:56:06 if you're in 'i build a distro' mode, use OE Feb 27 13:56:18 if you're in 'i hack kernel' mode, use OE's toolchain and do it standalone Feb 27 13:56:46 that's the way I have been used to work since now Feb 27 13:56:55 good strategy Feb 27 13:58:06 so people who ansered me to use bitbake virtual/kernel -cmenuconfig are wrong or never tried such command before? Feb 27 13:58:12 mckoan: this looks like good idea for blog post ;D Feb 27 13:58:20 mckoan: I never used that before Feb 27 13:59:36 * mckoan doesn't want to engage in controversy with anybody, he wants only to understand and learn :-) Feb 27 14:01:22 I'd like to create a command which do a make menuconfig with ncurses, who could point me into the right direction? hwre should I start analyzing the code? in bitbake? Feb 27 14:01:53 I meant a command which do a make menuconfig _in OE_ with ncurses Feb 27 14:02:08 mckoan: no idea. bitbake is multithreaded :( Feb 27 14:03:00 hrw: I supposed that, so my approach would be to create a parallel configuration system Feb 27 14:03:22 I can use to do different configuration activities Feb 27 14:03:45 mhhh have sense to split in 2 part, 1) be able to pass a .config file to bitbake function to compile kernel 2) prepare a custom .config in a native recipe ? Feb 27 14:04:33 test Feb 27 14:04:42 okay bye Feb 27 14:05:36 this configuration system would be useful for other tools like busybox options Feb 27 14:06:32 mckoan, have sense to use a native package to do only the interactive part of the configuration phase ? Feb 27 14:07:14 gremlin[it]: why not, but it must be built in a minute not after 4 hours Feb 27 14:08:22 mhhh true ... Feb 27 14:09:03 a little package into OE tree which builds and run as quickly as Kconfig Feb 27 14:10:18 and considering that most of us works on remote powerful multicore servers a curses interface qould be the best choice Feb 27 14:10:54 although I REALLY would like to see it into Eclipse environment ;-) Feb 27 14:13:50 03Koen Kooi  07org.openembedded.dev * r433e8f22bb 10openembedded.git/docs/usermanual/ (Makefile chunk-utf8.xsl): usermanual: fix 'nochunk' types Feb 27 14:17:14 * mckoan wonders how to involve someone into the development of this configuration system Feb 27 14:23:56 google summer of code ? Feb 27 14:25:25 gremlin[it]: I don't know how does it work, I have to read it Feb 27 14:26:09 don't sure but maybe OE will try to propose something to GSOC ... Feb 27 14:27:38 someone know tool like gnome-device-manager but for ncurses? Feb 27 14:31:41 gremlin[it]: GSOC woul be a good idea, but I can't wait until August 26, 2009 Feb 27 14:32:55 GSOC requires one person only for Administrivia stuff :-P Feb 27 14:33:33 I think I must do a Koan Spring Of Code ;-) Feb 27 14:34:00 hahahahha Feb 27 14:34:19 * mckoan is not joking Feb 27 14:34:38 anybody interested? Feb 27 14:36:09 03Jeremy Lainé  07org.openembedded.dev * r8c06fe62fe 10openembedded.git/conf/checksums.ini: checksums.ini: add entry for libsidplay-1.36.59.tgz Feb 27 15:07:06 hi :) Feb 27 15:09:36 hi Genesis Feb 27 15:14:13 who do you gotta blow to bitbake a kernel that boots? Feb 27 15:45:50 still looking at this error: http://pastebin.com/d34adad44 Feb 27 15:45:56 is the error the cause of a host-side autoconf problem? Feb 27 15:46:00 (Fedora 10 x86_64) Feb 27 15:54:19 03Frans Meulenbroeks  07org.openembedded.dev * r6da0d74d81 10openembedded.git/contrib/angstrom/build-feeds.sh: added ivman and pmount to angstrom build-feeds.sh Feb 27 15:54:50 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r753b13026a 10openembedded.git/packages/busybox/ (7 files in 4 dirs): busybox-static 1.11.3: make it more convenient to maintain the static version Feb 27 16:05:26 So, uh, has anyone heard of icecc being slower than not using it? Feb 27 16:06:19 I'm getting a build failure on pkgconfig-native trying to build for nokia800 + chinook-compat http://pastebin.com/m2a831cf9 Feb 27 16:06:39 any ideas? Feb 27 16:10:48 hmm... my overo build uses version 0.23 (and builds fine) where this was using 0.22... I'll try changing the PREFERRED_VERSION and see what I get Feb 27 16:21:22 that did the trick Feb 27 16:25:45 ScriptRipper, I already had the same error message...may be related to the fact that you may have multiple version of a toolchain and/or not the last version of it if I remember well Feb 27 16:25:49 oops Feb 27 16:25:51 sorry Feb 27 16:26:03 scruggs, I already had the same error message...may be related to the fact that you may have multiple version of a toolchain and/or not the last version of it if I remember well Feb 27 16:26:16 ScriptRipper, sorry tab complettion... Feb 27 16:30:10 Gnutoo: overo uses Angstrom and it sets its preferred version to 0.23. I just set it manually for my chinook-compat build. Feb 27 16:30:37 My builds for the 2 are in different directories so I don't think they could be getting confused Feb 27 16:31:18 scruggs, no that's not what I meant by multiple toolchains: I meant things like 2 autotool in staging Feb 27 16:31:32 of the same temp directory Feb 27 16:31:34 ahh... misunderstood Feb 27 16:33:59 seems that nfs-utils are not compilable... Feb 27 16:34:31 all of the in-tree versions Feb 27 16:41:32 after bitbake does its thing. shouldnt i be able to run qemu with or without an initramfs to see if it works? Feb 27 16:41:48 just the kernel that is Feb 27 16:44:18 I have news about bitbake virtual/kernel -cmenuconfig Feb 27 16:44:37 If I do such command from the local machine it works Feb 27 16:44:49 If I do such command from a ssh -X session it fails Feb 27 16:45:19 If I launch xterm (or whatever grapchic program) from ssh -X session it works Feb 27 16:45:30 so I argue is a bitbake problem Feb 27 16:45:40 mckoan: not quite Feb 27 16:46:00 bitbake clears environment before start build. so DISPLAY variable is also cleared Feb 27 16:46:34 mckoan: export BB_ENV_EXTRAWHITE="DISPLAY" before build and test Feb 27 16:46:44 hrw: clearing DISPLAY is a bad idea Feb 27 16:46:56 I tried also setting this way: GNOME_TERMCMD = 'gnome-terminal --disable-factory -t "$TERMWINDOWTITLE" --display localhost:10.0' Feb 27 16:47:18 florian: provide patch? Feb 27 16:47:33 or adding DISPLAY = localhost:10.0 before GNOME_TERMCMD in conf/bitbake.conf -> same problem Feb 27 16:48:19 with export BB_ENV_EXTRAWHITE="DISPLAY" fails too Feb 27 16:51:40 Gnutoo: I just built (but didn't tested the app) emacs for qemuarm without any problems Feb 27 16:52:06 mr_nice, ok...what's your platform and distro? Feb 27 16:52:12 thanks by the way Feb 27 16:53:34 mr_nice, mine is openmoko distro and gta02 machine Feb 27 16:53:35 Gnutoo: distro is angstrom Feb 27 16:54:06 Gnutoo: machine is qemuarm (armv5te) Feb 27 16:55:02 mr_nice, ah ok i've an armv6 with angstrom dev tmpdir i'll try Feb 27 16:56:49 mr_nice, and what's the cpu of the build machine? 64bit? Feb 27 16:57:29 Gnutoo: 32 bit Feb 27 16:57:46 ok me too(pentium M 2GHz) Feb 27 17:02:22 hrw: It should be sufficient to set BB_ENV_WHITELIST="DISPLAY" in bitbake.conf Feb 27 17:07:03 re Feb 27 17:07:21 woglinde hi Feb 27 17:07:37 he mr_nice Feb 27 17:07:45 florian: I added BB_ENV_WHITELIST="DISPLAY" in bitbake.conf -> same problem Feb 27 17:08:25 florian: I have this var in my own env Feb 27 17:08:55 is there a simple tool like 'procinfo' that monitors interrupt rates that people like to use? I notice there is no 'procinfo' package in current oe so I figure its not that commonly used Feb 27 17:09:09 perhaps people use 'top' but I dont' believe top will watch interrupt rates Feb 27 17:09:19 hrw: this or BB_ENV_EXTRAWHITE? Feb 27 17:09:28 procinfo just monitors /proc/interrupts (among other things similar to top) and gives you rate info Feb 27 17:09:28 ops... extra one Feb 27 17:09:52 tharvey: add it into OE then? Feb 27 17:10:15 it needs only ncurses ;) Feb 27 17:10:31 hrw, well i will if its useful, but I wonder if there is something else that takes its place Feb 27 17:10:52 yes, its a very simple recipe - but its not a maintained project possibly obsolete Feb 27 17:11:32 * hrw -> afk Feb 27 17:16:37 03Koen Kooi  07org.openembedded.dev * rd5ba616147 10openembedded.git/ (conf/checksums.ini packages/hal/gnome-device-manager_0.2.bb): gnome-device-manager: add 0.2 Feb 27 17:16:38 03Koen Kooi  07org.openembedded.dev * r6b77b81ca4 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Feb 27 17:16:38 03Koen Kooi  07org.openembedded.dev * r47a6f40a74 10openembedded.git/packages/linux/ (3 files in 3 dirs): linux-omap-pm: add cpufreq support for omap3evm, update defconfig Feb 27 17:16:49 03Koen Kooi  07org.openembedded.dev * rc5cfeb7646 10openembedded.git/conf/machine/omap3evm.conf: omap3evm: prefer -pm kernel, add ubifs stuff Feb 27 17:17:09 mckoan: hum nasty... did it parse again? Feb 27 17:18:59 03Koen Kooi  07org.openembedded.dev * r2505a09c7d 10openembedded.git/packages/dsplink/ (4 files): gstreamer-ti + depends: bump PR for omap3evm kernel change Feb 27 17:29:23 florian: what do you mean? Feb 27 17:31:07 * mckoan will be back later Feb 27 17:31:41 mckoan|away: did bitbake parse the bb files agin after that change? Feb 27 17:46:28 mckoan|away: I guess I see what might be the reason: SSH_AUTH_SOCK needs to stay there too Feb 27 17:49:18 * florian adds some more ssh environment variables, invalidates cace and tries again Feb 27 17:49:41 03Angus Ainslie  07fso/milestone5.5 * re072256943 10openembedded.git/packages/linux/linux-openmoko-stable_git.bb: Add the stable git version of the openmoko kernel Feb 27 18:16:24 has anyone an idea how to automate "The working tree was created by an older version of quilt. Please run 'quilt upgrade'" ? Feb 27 18:32:35 | {standard input}:30250: Error: selected processor does not support `swpb r2,r1,[r4]' Feb 27 18:32:54 the respective command line was: Feb 27 18:33:07 | ccache arm-oe-linux-gnueabi-g++ -march=armv4t -mtune=arm920t -mthumb-interwork -mthumb -c -pipe -isystem/local/pkg/oe/om-gta02/tmp/staging/armv4t-oe-linux-gnueabi/usr/include -fexpensive-optimizations -frename-registers -fomit-frame-pointer -O2 -isystem/local/pkg/oe/om-gta02/tmp/staging/armv4t-oe-linux-gnueabi/usr/include -fexpensive-optimizations -frename-registers -fomit-frame-pointer -O2 -fpermissive -fvisibility-inlines-hidden -Wall -W -D_REE Feb 27 18:33:07 NTRANT -fPIC -DQT_NO_DEBUG -DQT_GUI_LIB -DQT_CORE_LIB -DQT_SHARED -I/local/pkg/oe/om-gta02/tmp/cross/armv4t/share/qt4/mkspecs/linux-gnueabi-oe-g++ -I. -I/local/pkg/oe/om-gta02/tmp/staging/armv4t-oe-linux-gnueabi/usr/include/qt4/QtCore -I/local/pkg/oe/om-gta02/tmp/staging/armv4t-oe-linux-gnueabi/usr/include/qt4/QtCore -I/local/pkg/oe/om-gta02/tmp/staging/armv4t-oe-linux-gnueabi/usr/include/qt4/QtGui -I/local/pkg/oe/om-gta02/tmp/staging/armv4t-oe-lin Feb 27 18:33:08 ux-gnueabi/usr/include/qt4/QtGui -I/local/pkg/oe/om-gta02/tmp/staging/armv4t-oe-linux-gnueabi/usr/include/qt4 -I/local/pkg/oe/om-gta02/tmp/staging/armv4t-oe-linux-gnueabi/usr/include/qt4/Qt -I/local/pkg/oe/om-gta02/tmp/staging/armv4t-oe-linux-gnueabi/usr/include/python2.6 -I. -I. -o sipQtCorecmodule.o sipQtCorecmodule.cpp Feb 27 18:33:13 any idea what could be the problem? Feb 27 18:38:13 hmm, i saw something like that the other day... Feb 27 18:38:43 it might have something to do with thumb instructions Feb 27 18:38:54 pb_: ? Feb 27 18:39:08 03Angus Ainslie  07fso/milestone5.5 * re29e8f5074 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Bump up qi rev and add linux-openmoko-stable rev Feb 27 18:39:30 mickeyl, if I remember well thumb doesn't support swp but here it's swpb Feb 27 18:39:40 i'll check Feb 27 18:40:02 s/thumb/thumb or armv4 don't remember/ Feb 27 18:41:50 swp is not in thumb Feb 27 18:43:41 ok I don't remember well...it was long time ago Feb 27 18:43:43 hi booxter Feb 27 18:43:50 and swpb must be swap byte Feb 27 18:44:11 gnutoo whats the problem? Feb 27 18:44:20 atomic ops? Feb 27 18:44:24 woglinde, hi... Feb 27 18:44:45 woglinde, was mplayer-svn if I remember well Feb 27 18:45:03 hm sounds like atomic op problem Feb 27 18:45:16 hi ph5 Feb 27 18:45:37 hej woglinde Feb 27 18:46:10 I had also problems with boost_36 Feb 27 18:46:17 jepp Feb 27 18:46:20 atomic ops Feb 27 18:46:33 boost is using libatomic from boehm Feb 27 18:47:14 ah now I remember Feb 27 18:47:17 mplayer => swp Feb 27 18:47:28 s/swp/pld Feb 27 18:47:33 mplayer => pld Feb 27 18:47:37 hihi Feb 27 18:47:42 boost => swp Feb 27 18:48:10 so swp isn't avaliable for thumb if I remember well...I'll check Feb 27 18:48:32 i'll look at my boost bugreport and uncommitable fix Feb 27 18:52:29 hmm, bummer Feb 27 18:52:53 * mickeyl tries with ARM_INSTRUCTION_SET=arm Feb 27 18:53:01 hm Feb 27 18:53:11 whe should better spot which file breaks Feb 27 18:53:23 and only compile this without thumb support Feb 27 18:53:28 which package is it? Feb 27 18:53:54 ah a qtopia package... Feb 27 18:54:05 s/qtopia/qt-extended Feb 27 18:54:39 it's pyqt4 Feb 27 18:54:55 ok thanks Feb 27 18:59:26 if I remember rigth libgc also has this problem Feb 27 19:04:23 ok, it seems to help Feb 27 19:04:35 i'll put this in qt4x11.bbclass and qte4.bbclass for now Feb 27 19:04:49 we can improve granularity later Feb 27 19:05:04 back Feb 27 19:05:57 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r19d7f35751 10openembedded.git/classes/ (qt4e.bbclass qt4x11.bbclass): qt4[e|x11].bbclass: set ARM_INSTRUCTION_SET to arm Feb 27 19:06:09 Gnutoo: swp is depricated even for thumb instruction set Feb 27 19:06:15 sorry arm instruction set Feb 27 19:06:22 ok Feb 27 19:06:29 from armv6 onwards Feb 27 19:06:33 so avoid it Feb 27 19:06:54 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r9d8e810b59 10openembedded.git/packages/qmake/qmake2-native_2.10a.bb: qmake2-native: revert patch that broke qmake2 finding its spec files Feb 27 19:06:57 boost used fixes for beeing able to be compiled as thumb: Feb 27 19:08:03 for boost 1.36: http://rafb.net/p/axDxfR48.html Feb 27 19:08:43 so maybe changing the include could fix it...maybe... Feb 27 19:15:45 s/ok/yes/ Feb 27 19:16:55 (for the record: it is in the manual page B1-12 in the 14128.pdf of the arm architecture reference manual) Feb 27 19:17:08 (for the deprecation) Feb 27 19:20:32 NOTE: package python-pyqt-4.4.3: completed Feb 27 19:20:44 ok, after a long period of breakage, pyqt builds again Feb 27 19:21:09 RP: why did you 'inherit cross' in qmake2-native? this breaks building everything based onq t Feb 27 19:26:43 hi khem Feb 27 19:32:25 gnutoo sorry I dont understand the patch Feb 27 19:33:12 is __thumb__ for arm && thumb? Feb 27 19:34:06 i'm back Feb 27 19:36:01 woglinde, mabe you're right... Feb 27 19:36:10 s/mabe/maybe/g Feb 27 19:36:25 hm which boostlib is it in? Feb 27 19:36:40 36 Feb 27 19:36:52 but if arm is defined and not thumb it matches... Feb 27 19:36:54 detail Feb 27 19:38:20 whaha Feb 27 19:38:22 this suckz Feb 27 19:38:24 http://svn.boost.org/svn/boost/trunk/boost/detail/spinlock_gcc_arm.hpp Feb 27 19:38:26 swp used Feb 27 19:38:33 hello woglinde Feb 27 19:38:36 the whole bugreport is at http://bugs.openembedded.net/show_bug.cgi?id=4977 but I didn't give much details Feb 27 19:38:37 hi ant Feb 27 19:38:52 and I took some patches from debian or mailing-list Feb 27 19:38:54 mckoan: bitbake - c menuconfig working out-of-the-box here... Feb 27 19:39:51 woglinde, yes...but with some patch you can avoid using the swp Feb 27 19:39:58 sure Feb 27 19:40:01 they should fix it Feb 27 19:40:04 for atomic operation Feb 27 19:40:12 s/operations/operations/ Feb 27 19:41:06 maybe after fixing the emacs problem I should look again in boost Feb 27 19:42:49 hm Feb 27 19:42:51 ah Feb 27 19:42:55 morons Feb 27 19:43:11 mckoan: http://imagebin.ca/view/a9kp5c.html Feb 27 19:43:14 why the hack they dont use the gcc function __sync_lock_test_and_set Feb 27 19:43:17 even on arm Feb 27 19:44:18 he ifdef should go for armv4 Feb 27 19:44:22 not arm general Feb 27 19:48:43 ok Feb 27 19:49:04 re Feb 27 19:49:46 woglinde: howdy Feb 27 19:50:20 but I will check gcc source code now Feb 27 19:51:13 ok Feb 27 19:52:15 Gnutoo: did you got cable? Feb 27 19:52:38 hrw, not yet...the intalian post office are slow Feb 27 19:54:23 woglinde: usually people avoid compiler dependencies Feb 27 19:55:54 khem haha yes Feb 27 19:56:00 but look at the boost code Feb 27 19:56:05 its already in Feb 27 19:56:35 elif defined(__GNUC__) && ( __GNUC__ * 100 + __GNUC_MINOR__ >= 401 ) Feb 27 19:56:56 hm Feb 27 19:57:07 where the heck it is implemented for arm Feb 27 19:57:22 ant_home: bb in 15 min Feb 27 20:02:03 woglinde: moreover I dont know if __sync builtins work that well for gcc arm Feb 27 20:02:11 khem Feb 27 20:02:15 I am just looking Feb 27 20:02:20 where to find it Feb 27 20:02:24 atleast for pre armv6 Feb 27 20:02:29 they were not there Feb 27 20:02:39 khem right Feb 27 20:02:46 thats why the boostcode is wrong Feb 27 20:02:53 in my opinion Feb 27 20:03:42 gcc/config/arm/linux-atomic.c Feb 27 20:04:25 but the name says linux so boost will run out of luck for no linux Feb 27 20:05:25 khem huh? which gcc is this? Feb 27 20:05:38 stock gcc-4.3.2 dont have this file Feb 27 20:06:48 haha Feb 27 20:06:52 its from jan 2009 Feb 27 20:07:59 hm Feb 27 20:08:05 they are using the kernel op Feb 27 20:08:13 failure = __kernel_cmpxchg (oldval, val, ptr); Feb 27 20:08:15 woglinde: its trunk Feb 27 20:08:25 okay Feb 27 20:08:33 I hope its in gcc-4.4 Feb 27 20:08:43 yes thats what it will become Feb 27 20:08:44 that would be make a lot thinks easier Feb 27 20:09:09 but you still need to ensure that boost builds with older compilers Feb 27 20:11:05 i dunno if it is the right place to ask but .... anybody working with mmc/sd (spi) over gpio ? Feb 27 20:11:27 ant_home: ye you're right it works out-of-the box also here but only if used on local machine Feb 27 20:12:05 khem? Feb 27 20:12:14 khem did you see the line I posted? Feb 27 20:12:23 they already have it Feb 27 20:12:35 gcc > 4.0.1 for boos 36 Feb 27 20:12:54 ant_home: as I said at 17:44 if I do bitbake virtual/kernel -cmenuconfig Feb 27 20:13:02 If I do such command from the local machine it works Feb 27 20:13:04 oh thats the minimal gcc version they support ? Feb 27 20:13:08 If I do such command from a ssh -X session it fails Feb 27 20:13:14 If I launch xterm (or whatever grapchic program) from ssh -X session it works Feb 27 20:13:19 so I argue is a bitbake problem Feb 27 20:14:02 I tried all the afternoon without success, I have no idea Feb 27 20:14:33 khem for boost 36 yes Feb 27 20:15:08 args Feb 27 20:15:09 sorr Feb 27 20:15:10 y Feb 27 20:15:20 lol Feb 27 20:15:27 what morons Feb 27 20:15:33 for all other archs Feb 27 20:15:49 khem take a look yourself Feb 27 20:15:51 http://svn.boost.org/svn/boost/trunk/boost/detail/spinlock.hpp Feb 27 20:17:15 hm intressting the kernel cmp_exchange dont uses swp at all Feb 27 20:18:33 khem is robtow some of your co-workers? Feb 27 20:19:12 "some"? Feb 27 20:19:20 *g* Feb 27 20:19:23 <---is a singular being Feb 27 20:28:54 "call me legion" Feb 27 20:29:18 mwester for what? Feb 27 20:29:54 reference to robtow being a singular being. :) (Title of Zalazny novel) Feb 27 20:30:27 Hmm. I am *not* a Prince of Amber. Feb 27 20:30:38 lol Feb 27 20:30:42 never heard of it Feb 27 20:30:53 "Nine Princes in Amber" Feb 27 20:33:50 good book series Feb 27 20:43:18 hrw polish author? Feb 27 20:43:28 no Feb 27 20:43:40 Czech iirc Feb 27 20:43:47 roger zalazny Feb 27 20:44:19 Zelazny was born in Euclid, Ohio, the only child of Polish immigrant Feb 27 20:44:26 from Wikipedia :) Feb 27 20:44:27 hehe Feb 27 20:44:31 wiki lies Feb 27 20:44:34 as every time Feb 27 20:44:39 *g* Feb 27 20:45:07 okay will take a bath now Feb 27 20:45:10 bye till later Feb 27 20:47:24 can i ask about how to clone a git rep ? here i am a bit confused Feb 27 20:47:47 flameman whats the problem? Feb 27 20:47:58 did you read our git phrasbook? Feb 27 20:48:20 http://wiki.openembedded.net/index.php/GitPhraseBook Feb 27 20:48:27 yes, i want to clone the colie rep about sa1100 so git clone git://git.linuxtogo.org/home/thesing/collie.git Feb 27 20:48:37 but i get an empty .git folder Feb 27 20:48:56 so it is empty Feb 27 20:48:58 hm Feb 27 20:49:08 ~seen thesing Feb 27 20:49:11 thesing was last seen on IRC in channel #oe, 4d 5h 59m 13s ago, saying: 'I would do a git-bisect and write a mail do arm-ml if you can't fix it yourself.'. Feb 27 20:49:28 see http://git.linuxtogo.org/?p=thesing/collie.git;a=shortlog;h=refs/heads/pcmcia-rework Feb 27 20:49:36 how to get it ? Feb 27 20:50:59 hm good question Feb 27 20:51:03 :P Feb 27 20:51:13 did you try projects instead if home? Feb 27 20:51:22 git.linuxtogo.org/projects/thesing/collie.git Feb 27 20:51:28 let me try now Feb 27 20:52:40 could not switch to git.linuxtogo.org/projects/thesing Feb 27 20:52:56 #To clone one of these trees Feb 27 20:52:56 # Feb 27 20:52:56 # Feb 27 20:52:56 # git clone git://git.linuxtogo.org/home/ + project path. Feb 27 20:52:56 # Feb 27 20:52:57 # Feb 27 20:53:56 home/projects/thesing/collie.git maybee? Feb 27 20:54:54 not working :P Feb 27 20:55:06 ~seen mr_nice Feb 27 20:55:08 mr_nice is currently on #opensimpad.org #oe. Has said a total of 2 messages. Is idling for 1h 55m 40s, last said: 'if I remember rigth libgc also has this problem'. Feb 27 20:55:44 why didnt mrnice didn told me that he has a simpad git repose on linuxtogo Feb 27 20:55:48 *mumbles* Feb 27 20:57:33 flameman: I think you get that error because git clone expects there to be a master branch Feb 27 20:57:51 "git checkout -b mybranch origin/pcmcia-rework" in that empty dir should do the trick Feb 27 20:58:04 hm right Feb 27 20:58:10 ok i try :) Feb 27 20:58:29 "git branch -r" shows all the remote branches that git repo knows about Feb 27 21:00:02 OK working: thank you a lot :) Feb 27 21:15:02 woglinde: sorry *g - but it is not up to date anyway Feb 27 21:18:14 see you Feb 27 21:22:35 woglinde: I only know Jan Jelezny (javelin athelete ) Feb 27 21:22:38 :) Feb 27 21:23:43 it's Zelezny :p Feb 27 21:24:39 hmm syntax error Feb 27 21:53:51 03Jan Lübbe  07fso/milestone5.5 * r5a10e54bb4 10openembedded.git/packages/cellhunter/cellhunter_0.4.1.bb: cellhunter: fix RDEPENDS Feb 27 22:02:56 mr_nice would be good if you could update it Feb 27 22:03:39 how i can disable the checksum.ini ? Feb 27 22:04:02 gremlin its on the mailing list Feb 27 22:04:16 gremlin[it]: OE_STRICT_CHECKSUMS = "" Feb 27 22:04:22 in your local.conf Feb 27 22:04:30 thanks khem ! Feb 27 22:04:48 gremlin[it]: but you may get botched downloads Feb 27 22:04:54 you were warned Feb 27 22:06:28 mhhh i don't think is a really useful feature ... instead is annoying till checksum.ini isn't full populated :( ... Feb 27 22:07:07 if an entry exist, ok checkit, if don't exist warn at most ... but don't fail ! DAMN! Feb 27 22:07:31 gremlin[it]: thats what you will get with the above settings Feb 27 22:08:39 mhh ok ... but why a big file instead of a separate file or an assign in the corresponding .bb file ? Feb 27 22:08:53 its a choice you can make if you leave it to OE then you get what it wants Feb 27 22:09:12 gremlin[it]: there is a long discussion on it on ml Feb 27 22:09:17 you may want to read it it Feb 27 22:09:25 and post your inputs if you like Feb 27 22:10:21 got it ... i read just a couple messages ... Feb 27 22:10:39 gremlin[it], what is missing a checksum? Feb 27 22:11:05 automake-native-1.10.2-r0 Feb 27 22:11:11 btw. I am running a automated download for recipes, I might add some of them in coming days Feb 27 22:11:15 but they will not be all Feb 27 22:12:03 mickeyl added automaket 10.2 Feb 27 22:12:05 what i dislike is the behaviour of .dev banch: start to modify, push, break, start to resolve .... Feb 27 22:12:05 hehe Feb 27 22:12:13 gremlin[it], agreed Feb 27 22:12:36 gremlin[it], let me see if I can fix quickly Feb 27 22:12:43 usually we should stabilize on a separate branch before pushing to master Feb 27 22:12:58 woglinde, i'm using mickey branch to test a non angstrom (minimal) image Feb 27 22:13:03 people who push should not diable checksums ..... Feb 27 22:13:09 ah Feb 27 22:13:21 in his branch disable checksums Feb 27 22:13:28 khem, exactly you should ... Feb 27 22:15:39 khem, i'll read the thread on mailinglist in the weekend ... but ... how checksum.ini behave with cvs (and other) snapshot ? Feb 27 22:16:22 i do not think they are checked Feb 27 22:16:33 that would be more of a can of worms :) Feb 27 22:17:16 i really prefear a total different behaviour ... Feb 27 22:17:54 The current system does proide useful benefits Feb 27 22:17:58 it is not perfect Feb 27 22:18:01 different in what way Feb 27 22:18:06 the big problem is handling overlays Feb 27 22:18:15 2 things Feb 27 22:18:42 1) if a checksum exist use it, if don't exist do nothing ... Feb 27 22:19:06 the env var selects that behavior ... Feb 27 22:19:14 gremlin[it]: unless we have a source repository where we suck in the source tars etc and make sure they are ones we need and sign them everything falls short. The current solution is not evrything but it sure it something Feb 27 22:19:40 2) checksum info should be part of the recipe (where uri is defined) not a big file for all ... Feb 27 22:19:59 there was a suggestion about the second point Feb 27 22:20:03 which looks good Feb 27 22:20:18 ok ... i'll prefear the default behaviour is what a said ... if someone is paranoid can enforce it with env variable ... Feb 27 22:22:08 Yeah, #2 has some merit Feb 27 22:22:10 sincerelly i use OE more time with new of selfmade sources than just rebuild a stable distro ... so a strict check isn't what i expect ... Feb 27 22:22:17 being paranoid is good :) Feb 27 22:22:39 some us want to make sure sources do not change Feb 27 22:22:49 * gremlin[it] use Fedora with SElinus disable :P Feb 27 22:22:53 SElinux Feb 27 22:23:33 last question ... why 2 checksum md5 and sha256 ? Feb 27 22:26:20 not sure, I am wondering myself Feb 27 22:28:13 re Feb 27 22:30:30 jo ant Feb 27 22:31:05 sha is preferred over md5 in general Feb 27 22:41:23 about checksums, the whole issue has to be redesigned, like th eGentoo guys did (I posted links in ML) Feb 27 22:41:33 my 2 cents Feb 27 22:41:53 one file per recipe is sick Feb 27 22:42:15 in my opinion Feb 27 22:45:21 http://www.gentoo.org/news/20080204-digest-files-removed.xml Feb 27 22:49:26 Crofton|work: the link proposes max 3 hashes each, at least SHA1 Feb 27 22:49:48 I find all above very reasonable Feb 27 22:50:20 all above = http://www.gentoo.org/proj/en/glep/glep-0044.html Feb 27 23:51:05 * * OE Bug 5046 has been created by daniel.faulkner(AT)boltblue.com Feb 27 23:51:07 * * rx1950 machine conf typo Feb 27 23:51:09 * * http://bugs.openembedded.net/show_bug.cgi?id=5046 Feb 27 23:56:36 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * rcd9af345ea 10openembedded.git/packages/ (3 files in 2 dirs): python-dweba: new recipe; successor to python-spyro. adjust task-python-everything Feb 28 00:18:23 03Khem Raj  07org.openembedded.dev * r6641422f8e 10openembedded.git/conf/machine/rx1950.conf: rx1950.conf: Add missing quote in MACHINE_FEATURES Feb 28 00:21:06 * * OE Bug 5046 has been RESOLVED (FIXED) by raj.khem(AT)gmail.com Feb 28 00:21:08 * * rx1950 machine conf typo Feb 28 00:21:10 * * http://bugs.openembedded.net/show_bug.cgi?id=5046 Feb 28 00:21:35 03Khem Raj  07org.openembedded.dev * r6eb5682d1b 10openembedded.git/conf/checksums.ini: checksums.ini: Add checksum entry for docbook-dsssl-1.79.tar.bz2 Feb 28 00:29:25 Hello! http://pastebin.com/m7f460dc5 - old problem... still actual... x86 -> errors... Feb 28 00:39:58 Are you building in a virtualized environment? Feb 28 00:44:42 ?.. Feb 28 00:44:59 what is virtualized? Feb 28 00:45:10 it's just a pyhthon error... Feb 28 00:45:26 ImportError: No module named traceback Feb 28 00:45:39 ERROR: Task 537 (/home/s/my/oe/openembedded/packages/linux/linux_2.6.25.20.bb, do_package) failed Feb 28 00:46:14 One question: I prefer to build kernel myself - how to throw out Linux kernel from building by OE? Feb 28 00:46:30 virtualized, like vmware and stuff Feb 28 00:46:41 xen Feb 28 00:46:56 the actual error seems to be: Feb 28 00:46:59 no, I just building Angstrom for x86, minimal image... Feb 28 00:47:13 on your host pc that is not virtualized? Feb 28 00:47:42 ERROR: Exception: Message:[Errno 24] Too many open files: '/home/gilligan/dev/x86_test/tmp/work/x86-angstrom-linux/linux-2.6.28-r3/tmp9MAiIG' Feb 28 00:47:42 no, just plain Linux, no virtualizaed Feb 28 00:47:48 this seems to be your problem Feb 28 00:47:52 too many open files Feb 28 00:48:19 if that comes, all can happen Feb 28 00:48:37 sorry, it's not my errors, now I paste my log: Feb 28 00:48:43 and as it's an error message given by your operating system, I'm pretty much out of ideas Feb 28 00:49:11 http://pastebin.com/m4a8b9fca Feb 28 00:49:27 this is my own errors: http://pastebin.com/m4a8b9fca Feb 28 00:49:39 I have latest bitbake from SVN Feb 28 00:50:05 yes Feb 28 00:50:06 again Feb 28 00:50:07 ERROR: Exception: Message:[Errno 24] Too many open files: '/home/s/my/oe/build/work/x86-angstrom-linux/linux-2.6.25.20-r2/tmpm7vLC5' Feb 28 00:50:20 yeah, really :) Feb 28 00:50:26 how to fix that? Feb 28 00:50:33 fix your operating system Feb 28 00:50:36 no idea, sorry Feb 28 00:50:39 i have never seen this Feb 28 00:51:00 but i wouldn't search in bitbake or OE, the answers lies somewhere else Feb 28 00:51:00 I using Ubuntu 8.10 i386... Feb 28 00:54:00 did you try google yet? Feb 28 00:54:13 ~google ubuntu too many files Feb 28 00:54:23 first hit may apply Feb 28 00:54:43 yes, I searched... Feb 28 00:55:10 and increasing file-max didn't help? Feb 28 00:56:33 can't find how to increase this number, but I think I can solve this problem... :) Feb 28 00:56:47 i hope se Feb 28 00:56:52 first google link Feb 28 00:57:10 less than 10 lines to read :D Feb 28 00:58:56 https://bugs.launchpad.net/ubuntu/+source/bittorrent/+bug/13371 ? :) Feb 28 01:07:21 ulimit -n 4096 Feb 28 01:07:21 ulimit -n unlimited Feb 28 01:07:40 http://confluence.atlassian.com/display/DOC/Fix+%27Too+many+open+files%27+error+on+Linux+by+increasing+filehandles Feb 28 01:09:23 thanks ;) Feb 28 01:09:35 but ulimit -n - easier :) Feb 28 01:13:02 do not help... Already set to 300000 opened files... Feb 28 01:43:13 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * rb09627d8bf 10openembedded.git/packages/python/python-pylinda_0.6.bb: python-pylinda: change SRC_URI to mirror Feb 28 01:43:20 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * rbdca7e76cc 10openembedded.git/packages/moin/moin_1.8.1.bb: moin: no longer needs install hack Feb 28 01:49:22 ls Feb 28 01:49:28 oops Feb 28 01:50:43 hehe Feb 28 01:56:20 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r8a013c3930 10openembedded.git/packages/dtnrg/dtn_2.5.0.bb: dtn: fix packaging and constrain arm instruction set Feb 28 02:01:49 03Michael 'Mickey' Lauer  07mickey/org.openembedded.dev * r9685c48499 10openembedded.git/packages/python/python-pygtk-1.2_0.6.12.bb: python-pygtk 1.2: fix -dev packaging **** ENDING LOGGING AT Sat Feb 28 02:59:58 2009