**** BEGIN LOGGING AT Thu Aug 23 02:59:57 2007 Aug 23 03:14:15 03daniel 07org.oe.dev * rb10ad1f1... 10/ (3 files in 3 dirs): linux-gta01: Add version 2.6.21.6 from OpenMoko svn overlay. Aug 23 04:34:15 03daniel 07org.oe.dev * rc4778a74... 10/ (4 files in 2 dirs): linux-gta01: Reflect svn path changes between 2.6.21 and 2.6.22 patchset Aug 23 04:34:24 03daniel 07org.oe.dev * rda18e14b... 10/ (4 files in 2 dirs): disapproval of revision 'c4778a747ca17f52ef9afa2e2c5254081401273a' Aug 23 04:34:35 03daniel 07org.oe.dev * r62e56b40... 10/ (4 files in 2 dirs): linux-gta01: Reflect svn path changes between 2.6.21 and 2.6.22 patchset Aug 23 06:59:35 RP: http://pastebin.ca/667676 Aug 23 06:59:40 morning Aug 23 07:04:28 hi Aug 23 07:05:13 someone built qt4-x11-free in OE recently? Aug 23 07:05:15 hi Aug 23 07:15:20 Crofton: using iwmmxt on a non iwmmxt cpu wil give you SIGILL Aug 23 07:15:35 (or on a kernel without iwmmxt support) Aug 23 07:19:58 * * OE Bug 2817 has been RESOLVED (FIXED) by Aug 23 07:20:00 * *  abiword-2.5.2-r0-do_configure Aug 23 07:20:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2817 Aug 23 07:25:57 * * OE Bug 2814 has been RESOLVED (WORKSFORME) by Aug 23 07:25:59 * *  webkit-gtk-0.0+svnr25169-r25169-r1-do_install Aug 23 07:26:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2814 Aug 23 07:29:09 someone built qt4-x11-free in OE recently? Aug 23 07:36:34 03koen 07org.oe.dev * ree306d87... 10/ (4 files in 2 dirs): gstreamer: add depends on directfb (should fix #2802) Aug 23 07:37:22 hi Stelios Aug 23 07:37:43 good morning all Aug 23 07:37:49 hrw : hi ! Aug 23 07:38:07 good morning all Aug 23 07:38:26 koen hi ! Aug 23 07:38:27 hey ppl Aug 23 07:39:09 * * OE Bug 2832 has been created by carbonfreeze(AT)gmail.com Aug 23 07:39:11 * * gpe-irc segfaults Aug 23 07:39:12 just cracked the digitizer on my zaurus... any ideas where i can get the parts ? Aug 23 07:39:13 * * http://bugs.openembedded.org/show_bug.cgi?id=2832 Aug 23 07:39:48 rshakin: which model? Aug 23 07:39:54 sl5500 Aug 23 07:40:16 Hi there. Many many packages have disappeared during my last update of org.openembedded.dev. Especially packages in xorg-driver, where I'm right now... Do anybody know why, and where I find the drivers? Aug 23 07:40:18 rshakin: put it into drawer Aug 23 07:40:36 rshakin: replacing screen will be more expensive then buying new pda Aug 23 07:40:58 * * OE Bug 2833 has been created by autobuild(AT)openembedded.org Aug 23 07:41:00 * * gst-plugins-bad-0.10.5-r1-do_compile Aug 23 07:41:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2833 Aug 23 07:41:10 * * OE Bug 2833 has been RESOLVED (INVALID) by Aug 23 07:41:12 * *  gst-plugins-bad-0.10.5-r1-do_compile Aug 23 07:41:14 * * http://bugs.openembedded.org/show_bug.cgi?id=2833 Aug 23 07:41:37 damn does anyone know the part number of the screen, or any other option Aug 23 07:41:55 rshakin: moment Aug 23 07:45:33 rshakin: DUNTL2262YCZZ probably Aug 23 07:46:01 rshakin: but as I said - replacing screen will be more expensive then buying new pda Aug 23 07:46:41 rshakin : look at ebay. maybe you can find any zaurus that is sold for parts Aug 23 07:47:48 yeah going to do that Aug 23 07:54:57 * * OE Bug 2834 has been created by carbonfreeze(AT)gmail.com Aug 23 07:54:59 * * horizon fails to start due to bad hardcoded path Aug 23 07:55:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2834 Aug 23 07:58:25 Do anybody know where I can find the xorg-driver/xf86-video-chips package that existed before? Aug 23 08:02:11 Hi, how can I tell Bitbake to apply a patch only if a certain kernel (2.6) is used? Aug 23 08:02:26 03koen 07org.oe.dev * r6d7ffa86... 10/ (1 packages/gsm/libgsmd_svn.bb): gsmd: fix packaging Aug 23 08:02:58 03hrw * r946 10bitbake/ (ChangeLog lib/bb/cooker.py): bitbake: use Psyco only on x86 (no support for other archs) - taken from Poky Aug 23 08:03:34 03hrw 07bitbake-1.8 * r947 10/ (ChangeLog lib/bb/cooker.py): bitbake: use Psyco only on x86 (no support for other archs) - taken from Poky Aug 23 08:04:57 * * OE Bug 2835 has been created by carbonfreeze(AT)gmail.com Aug 23 08:04:59 * * Kazehakase segfaults when attempting to open a webpage Aug 23 08:05:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2835 Aug 23 08:09:58 * * OE Bug 2802 has been marked as DUPLICATE of bug 2833 by Aug 23 08:10:00 * *  gst-plugins-bad-0.10.5-r0-do_compile Aug 23 08:10:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2802 Aug 23 08:15:10 raphman : to apply a patch where ? Aug 23 08:16:25 ~seen zecke Aug 23 08:16:28 zecke was last seen on IRC in channel #oe, 10h 54m 43s ago, saying: 'bye'. Aug 23 08:20:58 * * OE Bug 2833 has been REOPENED by Aug 23 08:21:00 * *  gst-plugins-bad-0.10.5-r1-do_compile Aug 23 08:21:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2833 Aug 23 08:21:58 * * OE Bug 2836 has been created by carbonfreeze(AT)gmail.com Aug 23 08:22:00 * * gpe-gallery segfaults during image manipulation Aug 23 08:22:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2836 Aug 23 08:22:59 the weekly changelog: http://lists.linuxtogo.org/pipermail/openembedded-devel/2007-August/002857.html Aug 23 08:32:21 pidgin 2.1.1 was released Aug 23 08:39:46 03hrw 07org.oe.dev * r38e2cd40... 10/ (5 files in 3 dirs): python-dbus: make it build on amd64 Aug 23 08:40:11 03koen 07org.oe.dev * r19b3553f... 10/ (1 packages/gtk-webcore/midori_0.0.5.bb): midori: update to 0.0.5 Aug 23 08:40:21 03hrw 07org.oe.dev * r3a4584bc... 10/ (4 files in 3 dirs): python-openmoko: make it build on amd64 Aug 23 08:40:34 03pH5 07org.oe.dev * r52b1e168... 10/ (10 files in 6 dirs): Aug 23 08:40:45 libgsmd: combine machine specific configuration files, package plugin Aug 23 08:40:58 debugging info - no machine specific (lib)gsmd packages anymore Aug 23 08:41:10 03koen 07org.oe.dev * r873dfa47... 10/ (1 classes/seppuku.bbclass): seppuku.bbclass: add more debugging and keep attaching buildlogs for known bugs Aug 23 08:42:22 hey zecke Aug 23 08:42:28 hi zecke Aug 23 08:42:45 zecke: can you adapt patch from r3a4584bc... to python-openmoko? Aug 23 08:44:10 03koen 07org.oe.dev * r1549b0af... 10/ (4 files in 2 dirs): pidgin: update to 2.1.1, enable avahi Aug 23 08:45:39 moin Aug 23 08:45:55 hrw: I would prefer to wait for mickeyl :) Aug 23 08:46:21 zecke: could you commit ikada to contrib/ in .dev? Aug 23 08:47:12 zecke: to get 'yes, patch is ok'? Aug 23 08:47:58 * * OE Bug 2837 has been created by autobuild(AT)openembedded.org Aug 23 08:48:00 * * uboot-openmoko-1.2. 0+git8993e54b6f397973794f3d6f47d3b3c0c98dd4f6+svn20070820-r1-do_configure Aug 23 08:48:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2837 Aug 23 08:49:53 Thanks for fixing rm_work Aug 23 08:50:27 koen: http://www.openembedded.org/~zecke/tinderbox-server-snapshot-20070823.tar.gz Aug 23 08:51:21 I don't know if this snapsot actually works :) Aug 23 08:55:02 yakov: as I said, use your brain. If binutils configure fails. Take a look at the bb file for binutils, take a look at config.log Aug 23 08:55:49 hahahah what's classes/seppuku.bbclass ... is to suicide bitbake ?!?! Aug 23 08:56:21 zecke, yes, I've thought about it late yesterday, if it fails I'll try to patch bb files.. thanks! Aug 23 08:56:48 zecke : is this the source to the tinderbox server ? Aug 23 08:57:44 steliosk: no, to bonsai, ikada and my incomplete version of tinderbox Aug 23 08:58:33 zecke : the one running now which one is it ? Aug 23 08:59:09 zecke : we are thinking of getting one setup locally as the OE one delays the builds considerably Aug 23 09:00:17 steliosk: I can send you a copy Aug 23 09:00:40 good morning Aug 23 09:00:43 zecke : thanks ! that woule be great Aug 23 09:02:38 hrw: why is this patch for amd64? Aug 23 09:02:55 hi all! Aug 23 09:03:59 I have a little question regarding a build problem .. Aug 23 09:04:05 zecke: it is for any combo of misc size between host<>target. so 32host->64target, 64host->32target etc Aug 23 09:04:23 zecke: python stuff search in wrong place for headers Aug 23 09:04:49 when bitbake says: makedevs: ... file can not be created! Aug 23 09:05:20 jan_js: update your metadata Aug 23 09:05:25 Is there something missing on my system, or is it missing in oe ? Aug 23 09:05:55 koen: o.k, will try that first. Thanks. Aug 23 09:06:16 jan_js : There was an issue and was fixed yesterday, update your metadata Aug 23 09:07:01 steliosk: All right, thanks for the info! Aug 23 09:07:10 * * OE Bug 2838 has been created by autobuild(AT)openembedded.org Aug 23 09:07:12 * * linux-gta01-2.6.21.6-moko11-r2-do_populate_staging Aug 23 09:07:14 * * http://bugs.openembedded.org/show_bug.cgi?id=2838 Aug 23 09:07:45 steliosk: http://page.mi.fu-berlin.de/freyther/crappy-tinderbox.tar.gz Aug 23 09:08:25 zecke : Thanks ! Aug 23 09:08:58 zecke : btw i see you follow the ubuntu naming scheme. aka "crappy" :) Aug 23 09:09:01 hrw: you want to set a maxsrcdate for the patch Aug 23 09:09:17 that is horrible mozilla perl code :} Aug 23 09:09:35 someone else is tesint oabi ? Aug 23 09:09:42 instead of giving it to you, it should just die Aug 23 09:09:45 tesint --> testing Aug 23 09:09:57 * * OE Bug 2839 has been created by carbonfreeze(AT)gmail.com Aug 23 09:10:00 * * pidgin/ finch packages currently in feed lack network plugins Aug 23 09:10:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2839 Aug 23 09:10:46 zecke: I do not want. recipe is marked as 0.1 so I do not care does it works for someone or not. Aug 23 09:11:09 zecke: if upstream do not cooperate with marking releases in repo then why I should? Aug 23 09:11:23 hrw: why did you patch it then? Aug 23 09:11:24 zecke: and nothing changed just because you are the upstream ;D Aug 23 09:11:43 zecke: because it broke task-python-everything which I tried to build Aug 23 09:11:56 and it will break again in a second :) Aug 23 09:11:57 * * OE Bug 2840 has been created by  Aug 23 09:11:59 * * gsmd startup script broken Aug 23 09:12:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2840 Aug 23 09:12:10 well, actually only after my work day... Aug 23 09:12:26 Was task-bootstrap obsoleted by task-boot? Aug 23 09:12:38 B_Lizzard: by task-base Aug 23 09:12:53 Ok, thanks Aug 23 09:13:04 zecke: then update recipe to have _svn.bb instead of floating _0.1.bb Aug 23 09:13:33 did I create that recipe ;) Aug 23 09:14:34 ;) Aug 23 09:20:46 zecke : does apache needs any special config to run the perl files outside cgi-bin ? Aug 23 09:21:48 zecke pasted "cgi-bin config" at http://paste.lisp.org/display/46591 Aug 23 09:22:03 florian: I saw in the log of org.openembedded.dev that you deleted a lot of files in the xorg-driver package directory, 2007-08-09. Aug 23 09:22:16 zecke annotated #46591 with "complete config" at http://paste.lisp.org/display/46591#1 Aug 23 09:22:49 majje: uh? quite unlikely... Aug 23 09:23:31 florian: hmmm. wierd florian@openembedded.org, isn't you? Aug 23 09:23:54 majje: which revision it was? Aug 23 09:24:02 hrw: Revision: 7a67ee2faf59e6803887617211f99fcc2f9cf4e7 Aug 23 09:25:03 majje: that's me... but I can't imagine touching xorg and I can't find that revision Aug 23 09:25:11 someone forgot to exclude merges from log? Aug 23 09:25:39 XorA: aha, probably ;) Aug 23 09:26:39 majje: it was koen Aug 23 09:26:50 778aeb6e918d6ec53d4abbfa67d0d3832c8d717e revision Aug 23 09:26:58 ~curse koen twice Aug 23 09:26:59 May you be reincarnated as a Windows XP administrator, koen twice ! Aug 23 09:27:16 ~blame koen this game is fun Aug 23 09:27:17 * ibot blames koen this game is fun (and Canada) for all the evil in the world Aug 23 09:27:23 hehe Aug 23 09:27:26 koen: ping Aug 23 09:27:30 hrw, majje: read my mail to the mailinglist Aug 23 09:27:54 koen: ok, I will :) Aug 23 09:27:58 * * OE Bug 2839 has been RESOLVED (FIXED) by Aug 23 09:28:00 * *  pidgin/finch packages currently in feed lack network plugins Aug 23 09:28:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2839 Aug 23 09:28:06 "Hi, Aug 23 09:28:06 I just finished merging the Xorg cleanup from Poky. All my test cases build and run, but Aug 23 09:28:07 there might be small bugs left, please report those to bugzilla." Aug 23 09:28:10 3I just finished merging the Xorg cleanup from Poky. All my test cases build and run, but Aug 23 09:28:30 koen: but Poky use only subset of OE X.org recipes Aug 23 09:28:40 Good morning, all! Aug 23 09:28:51 morning russf Aug 23 09:29:02 There is a broken link (for me anyway) from DevelopmentBranches on the wiki Aug 23 09:29:10 hrw: right, that's why I added recipes needed for angstrom-x11-image as well Aug 23 09:29:21 should link to viewmtn Aug 23 09:29:22 (e.g. "my test cases") Aug 23 09:29:57 can I get permission to view 'viewmtn' or can it be opened up? Aug 23 09:30:09 here it is: http://www.openembedded.org/viewmtn/ Aug 23 09:30:20 koen: and as it looks you also managed to break builds of other people Aug 23 09:30:36 hrw: such is life Aug 23 09:32:14 hrw, koen: How do I revive my packages? Aug 23 09:33:47 majje: checkout a revision before the deletion? change something, merge? Aug 23 09:33:50 majje: I would update to before 778aeb6e918d6ec53d4abbfa67d0d3832c8d717e was created and copy xorg* dirs. then update to current metadata Aug 23 09:34:28 update needed recipes, send to bugtracker Aug 23 09:37:13 hrw: OK. Do I attach all the recipes to the issue? Aug 23 09:38:16 all which you need? yes Aug 23 09:38:32 ok, good. thanx for the help! Aug 23 09:38:37 I have a problem with spectrum, a big wait while configuration Aug 23 09:40:26 I guess it's grow to infinity Aug 23 09:53:00 good morning Aug 23 09:56:16 03florian 07org.oe.dev * rfd86559d... 10/ (1 packages/librsvg/librsvg_2.11.1.bb): librsvg: Add 2.11.1 again Aug 23 09:56:24 03florian 07org.oe.dev * r0c3cbec7... 10/ (1 packages/fbreader/fbreader_0.8.2a.bb): fbreader: Add libiconv depencency Aug 23 09:56:35 03florian 07org.oe.dev * r8712c7fb... 10/ (3 files in 3 dirs): gpe-bootsplash: Update GPE version number for Angström Aug 23 09:56:50 03florian 07org.oe.dev * r6cbb4963... 10/ (1 packages/gpephone/libsettings_0.3.bb): libsettings: Update download location Aug 23 09:57:10 03mickeyl 07org.oe.dev * rf469b433... 10/ (3 files in 3 dirs): add openmoko-tasks2 and ship it in the default image Aug 23 09:57:22 03mickeyl 07org.oe.dev * r289b2122... 10/ (4 files in 3 dirs): python 2.5.1: catch up with some build system changes from 2.4 - 2.5 causing only half of the packages being installed into the prop places Aug 23 10:01:05 what's the vasiable to enable parallele package build on local.conf ? Aug 23 10:01:14 vasiable -> variable Aug 23 10:03:17 03woglinde2 07org.oe.dev * r488594bd... 10/ (1 MAINTAINERS): Aug 23 10:03:19 MAINTAINERS: add myself to MAINTAINERS Aug 23 10:03:21 * now I am at MAINTINERS Aug 23 10:03:23 * added my recipes cacao-cldc, classpath-minimal, midpath Aug 23 10:03:33 03woglinde2 07org.oe.dev * rcbc9095c... 10/ (1 MAINTAINERS): MAINTAINERS: added my irc-nick Aug 23 10:03:49 03woglinde2 07org.oe.dev * r16fca46d... 10/ (1 packages/classpath/classpath-minimal_0.95.bb): classpath-minimal: next time I should not forget to bump PR Aug 23 10:05:30 OMG Aug 23 10:05:40 how stupid can people get?!?!? Aug 23 10:05:48 ~lart florian Aug 23 10:05:48 * ibot whips out a shotgun, trudges over to florian, and goes postal Aug 23 10:05:50 woglinde: could you lart florian? Aug 23 10:06:17 koen oehm Aug 23 10:06:34 did I something wron again? Aug 23 10:07:02 woglinde: not you a certain Mr Florian just broke all glibc builds Aug 23 10:08:13 CIA-1: florian org.oe.dev * r0c3cbec7... / (1 packages/fbreader/fbreader_0.8.2a.bb): fbreader: Add libiconv depencency Aug 23 10:08:50 ~lart florian Aug 23 10:08:50 * ibot whips out a shotgun, trudges over to florian, and goes postal Aug 23 10:09:32 koen, you guys have it easy, yesterday I had to sit through my research groups kick off meeting for the new year yesterday Aug 23 10:09:37 uh?! Aug 23 10:10:10 koen: hmm... It didn't build without... Aug 23 10:10:13 * koen attempts to write a lua postcommit check for libiconv Aug 23 10:10:19 gremlin[it]: PARALLEL_MAKE="-j4" ? and then e.g. BB_NUMBER_THREADS = "2" Aug 23 10:10:37 florian: time for seppuku? Aug 23 10:11:04 florian: s|libiconv|virtual/libiconv| Aug 23 10:11:34 XorA: ah oops... Aug 23 10:11:53 you know need to remove tmp and rebuild Aug 23 10:12:06 and everybody that built libiconv due to that change as well Aug 23 10:12:08 florian: are you using uclibc? Aug 23 10:12:11 03koen 07org.oe.dev * r28930b93... 10/ (1 packages/fbreader/fbreader_0.8.2a.bb): disapproval of revision '0c3cbec705ed9bb1d768da4c43b21a210b3a4609' Aug 23 10:12:14 Hey, can anyone tell me what I need to do to get rid of this message on mtn up: mtn: misuse: path _MTN/tmp/1 already exists Aug 23 10:12:44 rm _MTN/tmp/ -rf Aug 23 10:12:53 (no, that's not a joke) Aug 23 10:13:20 koen: hmm.. the resulting build was fine, why? Aug 23 10:13:27 XorA: no Aug 23 10:13:46 florian: libiconv and glibc iconv are incompatible, that will start breaking all iconv using apps Aug 23 10:13:52 florian: you CANNOT mix glibc-iconv and libiconv Aug 23 10:14:33 * florian has no idea why - but fbreader uses it and fbreader in the image works Aug 23 10:15:20 florian: try rebuilding glib now Aug 23 10:15:43 bitbake -crebuild glib-2.0 (feeling is believing) Aug 23 10:15:56 Did udev stop depending on libvolume-id? Aug 23 10:16:11 A general question, about config files: What is the correct way of adding your own config file to an already existing recipe? (A config file which isn't useful for anyone but me.) Do I create a new local recipe with the same name, and let it require the old recipe, and add URI_SRC_append with my config file? Aug 23 10:16:32 I made a new userland after a long hiatus, and it doesn't have any static device entries in /dev, nor libvolume_id in /lib Aug 23 10:16:37 Which makes it unbootable Aug 23 10:16:59 I was searching maybe something changed in the task-* files, but nothing Aug 23 10:17:10 hi all Aug 23 10:17:29 Everything seems the same between the new (non-working) userlands and the old (working) ones, except for the above facts Aug 23 10:17:36 mickey|bbiab: have you looked at http://www.gnome.org/~alexl/glick/ for openmoko? Aug 23 10:17:40 B_Lizzard: a known problem. the fix is to put these lines in devices*.txt Aug 23 10:17:46 zecke: urgs... what kind of effect is this?! Aug 23 10:17:51 /dev d Aug 23 10:17:55 /bin d Aug 23 10:17:57 et.c Aug 23 10:18:09 florian: the drepper effect Aug 23 10:18:12 makedev is being used to make up the -filesystem- erroneously .. Aug 23 10:18:12 koen: Thanks, I also needed to mtn list unknown packages/ |xargs rm -r Aug 23 10:18:29 otherwise the problem would reappear Aug 23 10:18:31 Oh, OK Aug 23 10:18:34 Lemme check... Aug 23 10:18:38 koen: hehe Aug 23 10:19:06 koen: is that similar to the rmk effect? ;) Aug 23 10:19:17 sort of Aug 23 10:19:21 make sure makedevs.c has been updated recenlty, also, there is add_dir_item() // somethingorother added.. Aug 23 10:19:29 rmk usually doesn't break userspace :) Aug 23 10:19:52 Also, a questions, why is udev still at version 100? Aug 23 10:20:20 you mean 092 Aug 23 10:20:30 Well, I use 100 Aug 23 10:20:33 everything newer is either broken or more broken Aug 23 10:20:51 :) Aug 23 10:20:53 Aug 23 10:21:04 Well, 100 served me well until, um, now Aug 23 10:21:05 the broken do_root codefix on makedevs is half-done (*txt files not updated with 'd' entries) .. i have a feeling because its sorta not makedevs any more, its makedirs too .. Aug 23 10:21:27 koen: kernel people? userspacers? what is this, some sort of game of footie? Aug 23 10:22:58 * * OE Bug 2841 has been created by magnus.ericmats(AT)sbg.se Aug 23 10:23:00 * * Missing xf86-video-chips and xf86-input-elographics recipes in org. openempedded.dev Aug 23 10:23:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2841 Aug 23 10:25:35 Ehm, device_table.txt has a (commented) "/dev/ d" thing, is that it? Aug 23 10:25:54 Plus a /dev/msys and a /dev/pts thing Aug 23 10:25:56 :) Aug 23 10:25:58 * * OE Bug 2842 has been created by simon.vogl(AT)researchstudio.at Aug 23 10:26:00 * * Linux 2.6.21 support for gumstix boards Aug 23 10:26:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2842 Aug 23 10:28:10 interesting Aug 23 10:28:16 where is the patch .... Aug 23 10:28:41 B_Lizzard: yup Aug 23 10:28:57 Ok... Aug 23 10:29:03 Thanks a lot, Aug 23 10:29:25 Crofton|home: which one? i posted the gumstix stuff, if you mean that Aug 23 10:29:46 koobla, you just created the bug? Aug 23 10:30:00 Should I wait till this is fixed in upstream, though? Aug 23 10:30:05 Crofton|home: yop Aug 23 10:30:07 Or will that take a long time Aug 23 10:30:20 koobla, I do not see the patch in bugzilla Aug 23 10:30:21 I don't like deviation or duplication is such things Aug 23 10:30:33 a number of people have been working on the gumstix already Aug 23 10:30:47 Crofton|home: I see - mom Aug 23 10:30:50 partail support in the generic linux kernel build already Aug 23 10:32:23 ok, I see - the full patch set worked oob, at least, had not played with the generic stuff yet, as I came from another custom board... Aug 23 10:32:41 please post the patcj Aug 23 10:32:46 into the bug Aug 23 10:32:57 how much testing have you done? Aug 23 10:33:12 there are a couple of guys here working on gumstix Aug 23 10:35:40 I have a question about bb-packages system. can I ask here? Aug 23 10:36:24 step2back: dont ask to ask but ask Aug 23 10:37:06 Crofton|home: do you know who's involved? Aug 23 10:37:31 asmola and a guy who is mainly on the oe-users list Aug 23 10:37:54 please do add the patch to bugzilla so I can look at it later today Aug 23 10:38:23 from opie-taskbar-images.inc PACKAGES = "${PN}-480x640 ${PN}-640x480 ..." How to know what is current package name called? Aug 23 10:38:52 step2back from the bla.bb name Aug 23 10:38:55 ok, i saw asmola's bug report Aug 23 10:39:13 step2back: recipes are named: PN_PV.bb Aug 23 10:39:16 he has a verdex, the other guy has a low end board Aug 23 10:39:30 03florian 07org.oe.dev * rd9028008... 10/ (1 packages/fbreader/fbreader_0.8.2a.bb): fbreader: fix dependency Aug 23 10:39:53 ok, I have some connex'es here and 2 verdex (but need network cards first for those) Aug 23 10:41:13 no no... I want to know what vars contained current name while building. Aug 23 10:41:22 step2back: PN Aug 23 10:42:33 ok, found the stuff in the .21 bb... will try that later. Aug 23 10:44:03 I need to make sure I clear on how the file:// interacts witht he machine name linux-2.6.21 directory to try and move all the gumstix patches out off the root Aug 23 10:44:27 and work out how to select betweeen the two defconfigs Aug 23 10:44:44 but I can't look at it right now Aug 23 10:44:59 Crofton|home: two defconfigs? Aug 23 10:45:06 Crofton|home: patches to packages/linux/linux-2.6.21/, configs into packages/linux/linux-2.6.21/gumstix1/defconfig packages/linux/linux-2.6.21/gumstix2/defconfig Aug 23 10:45:10 one for verdex one for the others Aug 23 10:45:14 03florian 07org.oe.dev * rd9028008... 10/ (1 packages/fbreader/fbreader_0.8.2a.bb): fbreader: fix dependency Aug 23 10:45:14 koen: vertex<>connex Aug 23 10:45:16 Oh damn, it doesn't work Aug 23 10:45:26 Never mind, I'll just wait for the fix in upstream Aug 23 10:45:35 why all the patches in "root" Aug 23 10:45:37 Crofton: put the verdex one in package/linux/linux-2.6.21/gumstix-verdex/ Aug 23 10:46:08 and the connex one in package/linux/linux-2.6.21/gumstix-connex Aug 23 10:46:16 Crofton|home: you can put them in 'root' or copy into each gumstix dir Aug 23 10:46:25 and add conf/machine/gumstix-verdex and -connex Aug 23 10:46:37 yeah, this sounds like what I was thinking Aug 23 10:47:03 my one main concern is the number of patches "polluting" the 2.6.21 directory Aug 23 10:47:14 but we would have to have two sets in teh machine directories Aug 23 10:47:40 Crofton|home: put them as gumstix-*.patch so it will be clear Aug 23 10:48:00 * Crofton|home hates renaming from updatream .... Aug 23 10:48:05 er upstream Aug 23 10:48:22 Crofton|home: gumstix-ORIGINAL-GUMSTIX-NAME Aug 23 10:48:30 hrw :) Aug 23 10:49:56 I also need to do a review of the gumstix list to see if they have anything going on that we chould think about Aug 23 10:52:05 B_Lizzard: the 'd' means "create this directory" .. its only really valid for /dev (since this is makedevs), but makedevs is being used to construct a filesystem here (since /bin has nothing to do with devs.. or shouldn't) .. the problem is, this whole issue might get fixed better in the do_root construction of a filesystem, or it could stay as a text-file trigger to a highly dubiously named makedevs.c too .. i think this requires someone responsible for the rootf Aug 23 10:52:34 Crofton|home: how about putting patches into an arch-specific subdirectory? ....-2.6.21/verdex/foobar.patch ? Aug 23 10:52:38 Hmmm Aug 23 10:52:53 in my case, adding /dev d and /bin d to the .txt files locally fixed the rootfs so that it could complete, and i've moved on to other problems too .. Aug 23 10:52:56 So, how did this whole thing break? Aug 23 10:52:57 koobla, we can, but I would have to dupe between the two machines Aug 23 10:53:16 I suspect hrw's suggestion will win :) Aug 23 10:53:22 i guess it broke because someone considered makedevs.c a good way to set up a protected/linked /bin/tinylogin for *sure* .. Aug 23 10:53:25 hm, you could keep common ones in gumstix/ :) Aug 23 10:53:44 koobla: you cant Aug 23 10:53:49 this does not work withthe automagic file selection mechanism :) Aug 23 10:53:59 oh, i see Aug 23 10:54:09 koobla: there is no such thing as 'gumstix' - gumstix-verdex and gumstix-connex exists Aug 23 10:55:09 koobla, atm I need to talk with a guy about some other stuff Aug 23 10:55:10 it coiuld really more complicated ... you'd have additional sub-versions depending on the types of i/o boards you connect Aug 23 10:55:14 so I am doing real work until then Aug 23 10:55:25 Well, uncommenting /dev d and rebuilding initscripts and the userland did nothing :/ Aug 23 10:55:37 koobla, I think the kernel just has modules for the additional boards Aug 23 10:55:44 Thank you for you help, I'll check it out myself, I guess Aug 23 10:55:49 koobla: extra boards are only addons. Aug 23 10:55:59 koobla: system works on base gumstix Aug 23 10:56:14 first we need to ave good support for the basic stuff :) Aug 23 10:56:25 the kernel, yes - but there are some incompatible combinations for user space (pcmcia cs mmc) Aug 23 10:56:36 03florian 07org.oe.dev * rd9028008... 10/ (1 packages/fbreader/fbreader_0.8.2a.bb): fbreader: fix dependency Aug 23 10:56:37 but that's a different set of prbs, agreed Aug 23 10:56:51 koobla: userspace are other thing ;D Aug 23 10:57:20 I am going to hide Aug 23 10:57:30 btw, must head off to route my zigbee board Aug 23 10:57:36 neet Aug 23 10:58:57 * * OE Bug 2772 has been RESOLVED (FIXED) by Aug 23 10:58:59 * *  recent change in fbreader breaks glibc build Aug 23 10:59:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2772 Aug 23 11:00:41 hmmm... is it normal: RRECOMMENDS_task-opie-base = -> ERROR: No providers of runtime build target opie-taskbar-images-480x320 Aug 23 11:03:35 B_Lizzard: do a bitbake -mrproper on the rootfs .. make *sure* you're getting the right devs*.txt file, and also make sure makedevs.c has a 'd' switch on its readint of the .txt file .. Aug 23 11:03:46 if not all above, then make update, &etc. Aug 23 11:04:52 make update? Aug 23 11:04:58 you mean mtn up, right? Aug 23 11:05:11 i am using mokomakefile, so i just make update to get changes .. Aug 23 11:05:41 ah right Aug 23 11:05:55 you're one of the 'too stupid to type mtn up' people Aug 23 11:06:08 excuse me, if i got questions about tinderbox, is this the right place to ask? Aug 23 11:06:17 no, I'm a "hoard it all in the Makefile" people Aug 23 11:06:28 that's what I said Aug 23 11:06:32 i don't particularly think mtn is all that special, either. Aug 23 11:07:07 OMG OE IS BROKENE!!!!!!ONE!!! Aug 23 11:07:08 no point adding yet another grand lexicon to the adventure when Makefile can do the job grandly, anyway .. Aug 23 11:07:18 i mean tinderclient.bbclass Aug 23 11:07:49 johncylee: sure .. just ask. Aug 23 11:08:15 OE is always broken Aug 23 11:08:37 hah Aug 23 11:08:37 it seems bitbake will fire multiple "NoProvider" event even if the build is successful Aug 23 11:08:41 i agree. Aug 23 11:09:00 :) Aug 23 11:09:02 bitbake -cbuild tinderclient Aug 23 11:09:54 johncylee: right, bitbake-1.8 started to do that Aug 23 11:09:56 * Bernardo is away: Ausente por agora. Aug 23 11:10:21 yeah... so the build will always seems failed on webpage Aug 23 11:12:13 what is the proper way to fix this? I have commented out the related lines in tinderclient.bbclass but it's just a workaround Aug 23 11:12:34 johncylee: .dev has the same thing commented out as well Aug 23 11:13:16 johncylee: You could change bitbake to fire a new event FataNoProvider or add a parameter to the NoProvider to be able to see if this is a fatal error or just misconfiguration Aug 23 11:15:04 zecke: i see. thanks Aug 23 11:21:58 hm Aug 23 11:22:32 why dont sdl RPROVIDES virtual/libsdl Aug 23 11:22:50 You shouldn't Rprovide virtuals Aug 23 11:23:00 rp hm okay Aug 23 11:23:06 rp reason? Aug 23 11:23:30 you link against it anyway Aug 23 11:23:41 woglinde: ipkg has no idea about PREFERRED_PROVIDERS so how does it understand multiple virtual/ entries ? Aug 23 11:23:59 *sigh* java Aug 23 11:24:02 *g* Aug 23 11:25:07 hm Aug 23 11:25:12 crofton, swapping the generic kernel for the gumstix one with iwmmx support worked, angstrom now boots with the tune-iwmmx set from the gumstix.conf file Aug 23 11:26:28 how should I do than this: jar needs a runtime libsdl, but its regardless if it is fb or x11 Aug 23 11:27:44 hm wie dont have the | as in .deb dependencies? Aug 23 11:28:57 * * OE Bug 2772 has been REOPENED by Aug 23 11:28:59 * *  recent change in fbreader breaks glibc build Aug 23 11:29:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2772 Aug 23 11:29:59 asmola, I am confused Aug 23 11:30:21 if you use the kernel that comes from linux_2.6.21 everythign is ok? Aug 23 11:31:06 if i use the kernel from the gumstix buildroot (which is a linux 2.6.21) itworks fine, that kernel has mmx support I belive Aug 23 11:31:18 if I use the 2.6.21 kernel that OE is creating it doesn't work Aug 23 11:32:04 in looking at when this is putting I am seeing that "Ive" is running before init Aug 23 11:32:19 asmola, if you look at the oe defconfig for gumstix you find: # CONFIG_IWMMXT is not set Aug 23 11:33:20 asmola, koobla has some stuff in the bug tracker Aug 23 11:33:44 i will take a look, add the #CONFIG_IWMMXT line and check and see what I get Aug 23 11:33:45 later today, I need to compare it with the stuff in linux_2.6.21 Aug 23 11:34:26 guys: maybe let we start adding version of gumstix when we talk about testing on hardware - ok? Aug 23 11:34:47 definitely, i am on verdex Aug 23 11:34:49 current OE has only old gumstix (pxa255 'connex') support Aug 23 11:35:09 and angstrom provides connex kernel/rootfs only Aug 23 11:35:14 hrw, we are moving that into linux.inc Aug 23 11:35:16 atm Aug 23 11:36:44 I moved that into linux 2.6.21 yesterday Aug 23 11:36:50 koobla, the defconfig from the gumstix website has the "right stuff" in it? Aug 23 11:37:11 it would be great if we could get a seperate machine conf file for verdex class that way we can start pushing some stuff back in without breaking connex/basix Aug 23 11:37:18 yeah Aug 23 11:37:20 ok Aug 23 11:37:26 I am going to shower Aug 23 11:37:32 then I will do guimstix work Aug 23 11:37:46 I am so damn pissed at work, I need a break Aug 23 11:37:59 can't focus on this stupid writing Aug 23 11:38:26 gumstix people: mtn pull ; mtn up Aug 23 11:39:19 koen, did you make some change in the past 12 hrs? Aug 23 11:39:31 he did Aug 23 11:39:37 ok, cool Aug 23 11:39:37 mtn: adding conf/machine/gumstix-connex.conf Aug 23 11:39:37 mtn: adding conf/machine/gumstix-verdex.conf Aug 23 11:39:37 mtn: renaming conf/machine/gumstix.conf to conf/machine/include/gumstix.inc Aug 23 11:39:43 awesome Aug 23 11:39:58 nice :) Aug 23 11:40:28 cia is broken, so no irc notifications Aug 23 11:43:39 03hrw 07org.oe.dev * r3fc90cd6... 10/ (5 files in 3 dirs): disapproval of revision '38e2cd40a04ac88008e433a43002415c562145e4' Aug 23 11:43:46 :/ Aug 23 11:43:56 03koen 07org.oe.dev * rc6e16b4a... 10/ (4 files in 3 dirs): gnuplot: update to 4.2.0 Aug 23 11:43:56 That was...ironic Aug 23 11:44:06 03koen 07org.oe.dev * rdc3cf7ca... 10/ (5 files in 3 dirs): gumstix: split into connex and verdex Aug 23 11:44:16 03hrw 07org.oe.dev * r9ffaf3c8... 10/ (3 files in 3 dirs): python-dbus: make it build on amd64 Aug 23 11:44:17 http://pastebin.com/d11ad2aca Aug 23 11:44:19 hi mickeyl Aug 23 11:44:24 What's "Ive" on line 78? Aug 23 11:44:26 03hrw 07org.oe.dev * r1479b877... 10/ (1 MAINTAINERS): MAINTAINERS: updated my entry Aug 23 11:44:34 03koen 07org.oe.dev * r0d0e74d6... 10/ (5 files in 3 dirs): Aug 23 11:44:39 hi woglinde Aug 23 11:44:44 * mickeyl fixing uicmoc dependency on X Aug 23 11:44:46 linux 2.6.21: split gumstix into connex and verdex Aug 23 11:44:56 * defconfigs need updating for both Aug 23 11:45:13 * DO NOT SET LOCALVERSION IN THE DEFCONFIG!!!!!!! Aug 23 11:45:21 03koen 07org.oe.dev * r616723bf... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom: add feed archs for new gumstix machines Aug 23 11:45:27 iwmmxt needs to get turned on in the verdex defconfig Aug 23 11:45:28 03koen 07org.oe.dev * re0613a17... 10/ (3 files in 2 dirs): gumstix: fix name Aug 23 11:46:53 udevd[162]: init_udevd_socket: error getting socket: Address family not supporte Aug 23 11:46:54 d by protocol Aug 23 11:47:01 yay for broken kernel configs Aug 23 11:49:46 asmola: I always hack in the configs, but propably - no Aug 23 11:49:56 mom, will check Aug 23 11:50:51 asmola: the verdex config does include mmx support Aug 23 11:51:13 in the machine.conf? Aug 23 11:51:41 i looked at the linux kernel config Aug 23 11:51:43 i'm *very* new to this so my apologies in advance for any comments/questions that don't make sense Aug 23 11:54:02 ok, I am clean Aug 23 11:54:29 koen, I am going to review the patches koobla posted, versus what we have Aug 23 11:57:45 Crofton|home: I assume the patches are identical, as I copied them over from the official gumstix repo Aug 23 11:58:05 koobla, I need to check versus what we have in linux_2.6.21 Aug 23 11:58:09 the verdex defconfig will need a look-at though Aug 23 11:58:29 where did the verdex defconfig in your patch come from? Aug 23 12:00:36 this is the current svn defconfig Aug 23 12:00:42 ok Aug 23 12:00:45 buildroot/target/device/Gumstix/verdex/linux.config Aug 23 12:00:48 so it should be good Aug 23 12:01:15 will need some tweaking for glibc support, i suppose (I think futex is disabled) Aug 23 12:01:39 no, its in Aug 23 12:01:48 asmola, tune-xscale is correct for verdex? Aug 23 12:01:56  Aug 23 12:02:04 what is correct tune for basix? Aug 23 12:02:07 yes, but you can also do tune-iwmmx Aug 23 12:02:07 tune-iwmmxt is the one we want Aug 23 12:02:15 ok Aug 23 12:02:35 tune-xscale is for the basix and connex i beleive Aug 23 12:02:35 so tune is different for basix and verdex? Aug 23 12:02:42 yes Aug 23 12:02:50 pxa255 -> pxa270 Aug 23 12:02:54 basix/connex are pxa255 so xscale Aug 23 12:03:02 verdex as pxa27x is iwmmxt Aug 23 12:03:17 03crofton 07org.oe.dev * r87df0f67... 10/ (1 packages/ltp/ltp-20070228/runltp-path.patch): ltp : Commit patch intended for earlier commit. from bug #2078. Aug 23 12:05:21 no, tune-iwwmxt is now what we want Aug 23 12:05:42 koen, they differ by machine Aug 23 12:06:02 regardless Aug 23 12:06:21 I have a feeling that OE is completely borked Aug 23 12:06:32 I don't even get ipkg in my userlands anymore :/ Aug 23 12:06:36 is not completely borked just a little bit. Aug 23 12:06:40 koen, I think I should move the tune into the specific machien file from the .inc Aug 23 12:06:45 B_Lizzard: works fine for all my machines and images Aug 23 12:06:46 Hello Aug 23 12:06:52 i've been working on it for 5 days straight and i'm now at the point where i can at least keep a build, and keep up with the daily updates. Aug 23 12:06:54 hey vlo Aug 23 12:07:06 but i am yet to get my images tested in both QEMU and on my neo .. Aug 23 12:07:46 i won't feel properly safe about settling into coding/porting for the neo until i can run a current image of my own builds under both qemu and neo, as well as flip to the mainline builds too. Aug 23 12:08:01 03koen 07org.oe.dev * r79480b22... 10/ (3 files in 3 dirs): linux 2.6.21: fix gumstix patches and verdex defconfig (works with iwmmxt now) Aug 23 12:08:22 Still no /dev/* entries in the userland :/ Aug 23 12:08:32 yeah, i think that require needs to go in the machine specifc file because when theres no way the connex is going to run with that require tune-iwmmx Aug 23 12:08:35 One more thing to do... Aug 23 12:08:47 B_Lizzard: rebuild makedevs/do_rootfs. Aug 23 12:08:53 bitbake -cmrproper rootfs Aug 23 12:09:14 for me, all i had to do was turn on /dev d in the .txt and it worked. Aug 23 12:09:15 replace rootfs with my image name? Aug 23 12:09:21 B_Lizzard: bitbake -crebuild makedevs-native IMAGE-NAME Aug 23 12:09:25 whatever the build step name is, i forget it. Aug 23 12:09:25 Cause I did that Aug 23 12:09:48 Ok, thanks Aug 23 12:09:54 I have defined a new machine in my OE tree (very similar to mx31ads, but still different on some points). However, when I try to build the bootstrap-image (with angstrom), the image class fails at fakeroot do_rootfs because in device_table-minimal.txt there is a /bin file to be added and there is no ${IMAGE_ROOTFS}/bin directory. Is this a known bug ? Aug 23 12:09:59 right, maybe makedevs hasn't been updated .. Aug 23 12:10:03 The image just went up by 600KB for no reason :/ Aug 23 12:10:09 I'll check that out Aug 23 12:10:29 vlo: mtn pull ; mtn up Aug 23 12:10:48 mtn to; mtn the bummer; mtn baby Aug 23 12:11:30 Hi! I try to build the "helloworld-image" but it can't find the "helloworld-*.ipk" package. A "find -name he*.ipk" at the build directory didn't show any results. Isn't this helloworld-image uptodate? Aug 23 12:11:45 koen: sorry, I had done it yesterday, I thought I was up to date Aug 23 12:12:02 yesterday is never up to day :) Aug 23 12:12:19 you might even pull hourly Aug 23 12:12:22 always update before trying to debug things Aug 23 12:13:09 koen, you done fondling the gumstix files? Aug 23 12:14:08 yeah, the verdex should be 'done' now Aug 23 12:14:48 ok Aug 23 12:14:53 yay it works :D Aug 23 12:15:03 I am going to move the tuning into the specifc machine files Aug 23 12:15:07 It probably was makedevs-native after all Aug 23 12:15:16 Thanks a lot torpor and hrw :D Aug 23 12:16:00 Crofton|home: NO! Aug 23 12:16:25 koen, right now you call for tune-xscale in gumstix.inc Aug 23 12:16:33 right Aug 23 12:16:47 since that's appropriate for pxa2xx cpus Aug 23 12:16:52 verdex needs tune-xwmmx Aug 23 12:16:59 no it doesn't Aug 23 12:17:00 I know it is spelled wrong Aug 23 12:17:32 it can *optionally* use tune-iwmmxt for people that like to shoot themselves in the foot Aug 23 12:18:03 how do we select the tune-iwmmxt Aug 23 12:18:10 we don't Aug 23 12:18:16 since it breaks a few apps Aug 23 12:18:28 as a compromise - add a commented-out tune-iwmmxt statement :) Aug 23 12:18:59 so on a verdex, we would get broken apps? Aug 23 12:19:59 if you'd use tune-iwmmxt , yes Aug 23 12:20:21 we only want tune-iwmmxt on the verdex part Aug 23 12:20:22 iwmmxt needs to get switched on at a case-by-case basis (e.g. like we do with mplayer) Aug 23 12:20:35 a couple notes... everything built OK with the tune-iwmmxt set with the gumstix last night Aug 23 12:20:41 so I would think the tune shoudl be included in the gumstix-*.conf Aug 23 12:20:47 Crofton|home: you only want tune-iwmmxt if you like random segfaults Aug 23 12:20:49 asmola, maybe runtime breage Aug 23 12:20:53 ok Aug 23 12:21:22 (or if you use g3nt00, but that's the same as random segfaults) Aug 23 12:21:55 anyway, can I move the tune-xsale into the specific machine files Aug 23 12:22:06 so it easier for people to shoot themselves Aug 23 12:22:14 and test when toolchains are updated? Aug 23 12:26:23 Newest pidgin depends on gstreamer even though --disable-audio is on Aug 23 12:26:28 Add --disable-gstreamer Aug 23 12:28:45 03pH5 07org.oe.dev * r632a3f74... 10/ (10 files in 2 dirs): Aug 23 12:28:57 angstrom task packages: use task.bbclass, automatically builds -dev and Aug 23 12:29:05 -dbg versions Aug 23 12:29:17 03pH5 07org.oe.dev * r02e047a8... 10/ (1 packages/tasks/task-base.bb packages/tasks/task-boot.bb): task-{base,boot}: use task.bbclass Aug 23 12:29:34 03pH5 07org.oe.dev * r4491a5c2... 10/ (12 files in 2 dirs): task-gpe*: use task.bbclass Aug 23 12:29:37 koen: bluez 3.15 (with A2DP alsa plugin) is out. Will you do the update our should I take a look later today? Aug 23 12:29:41 03pH5 07org.oe.dev * r0d292da9... 10/ (7 files in 2 dirs): task-{opie,qpe}*: use task.bbclass Aug 23 12:29:51 03pH5 07org.oe.dev * r3780347a... 10/ (1 packages/tasks/task-mokogateway.bb): task-mokogateway: use task.bbclass Aug 23 12:29:56 03pH5 07org.oe.dev * r65fc07a2... 10/ (1 packages/tasks/task-mythfront.bb): task-mythfront: use task.bbclass, only the epia version is machine specific Aug 23 12:30:09 03pH5 07org.oe.dev * rc8b45a18... 10/ (1 packages/tasks/task-xterminal.bb): task-xterminal: use task.bbclass Aug 23 12:30:20 03pH5 07org.oe.dev * r513114f1... 10/ (1 packages/tasks/task-xfce-base.bb): task-xfce-base: use task.bbclass Aug 23 12:30:27 03pH5 07org.oe.dev * rbe98539a... 10/ (1 packages/tasks/task-dvb.bb): task-dvb: use task.bbclass Aug 23 12:30:38 03pH5 07org.oe.dev * rd7dd304a... 10/ (1 packages/tasks/task-e-x11.bb): task-e-x11: use task.bbclass Aug 23 12:30:51 03pH5 07org.oe.dev * r3c3bca35... 10/ (1 packages/tasks/task-gmae.bb): task-gmae: use task.bbclass Aug 23 12:31:01 03pH5 07org.oe.dev * r01f0affd... 10/ (1 packages/tasks/task-proper-tools.bb): task-proper-tools: use task.bbclass Aug 23 12:31:10 03pH5 07org.oe.dev * r082dc9bd... 10/ (1 packages/tasks/task-sdl.bb): task-sdl: use task.bbclass Aug 23 12:31:16 03pH5 07org.oe.dev * r93a2f8dd... 10/ (7 files in 3 dirs): avahi: drop old versions Aug 23 12:31:26 03pH5 07org.oe.dev * r518fd97b... 10/ (5 files in 2 dirs): Aug 23 12:31:31 crofton, i am looking in packages/linux and i see a bb and folder for linux-gumstix-2.6.15 but not for the new kernel ver, is this correct? Aug 23 12:31:38 avahi: prepare for building gtk and python code, let only -daemon Aug 23 12:31:44 recommend libnss-mdns, have only -daemon and -autoipd depend on Aug 23 12:31:55 update-rc.d Aug 23 12:31:58 asmola, we are using the linux_2.6.21.bb for gumstix now Aug 23 12:32:03 03pH5 07org.oe.dev * r4d744e0a... 10/ (3 files in 3 dirs): avahi-python: provides avahi, includes python bindings Aug 23 12:32:10 there is one patch with changes Aug 23 12:32:13 ahh, i see Aug 23 12:32:20 03pH5 07org.oe.dev * r9beb0f7d... 10/ (1 packages/xorg-lib/libice_1.0.4.bb): libice: add 1.0.4 Aug 23 12:32:28 03pH5 07org.oe.dev * r0c3f4348... 10/ (3 files in 2 dirs): libx11, diet-x11: add 1.1.3 Aug 23 12:32:29 pxafb-definition.patch Aug 23 12:32:37 03pH5 07org.oe.dev * r63683be1... 10/ (1 packages/xorg-doc/xorg-sgml-doctools_1.2.bb): xorg-sgml-doctools: add 1.2 Aug 23 12:32:45 03pH5 07org.oe.dev * rf0f80b10... 10/ (1 packages/xorg-driver/xf86-input-mouse_1.2.2.bb): xf86-input-mouse: add 1.2.2 Aug 23 12:32:54 03pH5 07org.oe.dev * r75c1d0c8... 10/ (1 packages/xorg-driver/xf86-video-ati_6.6.193.bb): xf86-video-ati: add 6.6.193 Aug 23 12:32:58 http://bugs.openembedded.org/show_bug.cgi?id=2829 Aug 23 12:33:06 03pH5 07org.oe.dev * rf79db263... 10/ (1 packages/xorg-driver/xf86-video-nv_2.1.3.bb): xf86-video-nv: add 2.1.3 Aug 23 12:33:06 I am just wondering how that fits in Aug 23 12:33:07 oh those spammer s ;-) Aug 23 12:33:15 03pH5 07org.oe.dev * raa20f879... 10/ (1 packages/xorg-driver/xf86-video-savage_2.1.3.bb): xf86-video-savage: add 2.1.3 Aug 23 12:33:25 03pH5 07org.oe.dev * r3cbaa873... 10/ (1 packages/xorg-app/setxkbmap_1.0.4.bb): setxkbmap: add 1.0.4 Aug 23 12:33:33 03pH5 07org.oe.dev * r98d41dc7... 10/ (1 packages/xorg-lib/libxaw_1.0.4.bb): libxaw: add 1.0.4 Aug 23 12:33:39 03pH5 07org.oe.dev * r033f15d2... 10/ (1 packages/xorg-app/xrdb_1.0.4.bb): xrdb: add 1.0.4 Aug 23 12:33:41 03pH5 07org.oe.dev * rd2bd63e1... 10/ (1 packages/xorg-app/sessreg_1.0.3.bb): sessreg: add 1.0.3 Aug 23 12:33:47 03pH5 07org.oe.dev * r46d29205... 10/ (1 packages/xorg-lib/libxpm_3.5.7.bb): libxpm: add 3.5.7 Aug 23 12:33:47 I am going to look at his configs next Aug 23 12:33:49 03pH5 07org.oe.dev * rb47e1ac8... 10/ (1 packages/xorg-lib/libxrender_0.9.3.bb): libxrender: add 0.9.3 Aug 23 12:33:51 03pH5 07org.oe.dev * r60e3e2e4... 10/ (1 packages/xorg-app/xdm_1.1.6.bb): xdm: add 1.1.6 Aug 23 12:33:51 ok Aug 23 12:33:59 I think they come from the gumstix site Aug 23 12:33:59 03pH5 07org.oe.dev * rf5f444fe... 10/ (1 packages/xorg-app/xmodmap_1.0.3.bb): xmodmap: add 1.0.3 Aug 23 12:34:01 (13 lines omitted) Aug 23 12:34:10 koen: OK, problem fixed, the updated device_table-minimal.txt works better Aug 23 12:34:25 03mickeyl 07org.oe.dev * r30eacf6a... 10/ (4 files in 3 dirs): gtk+ 2.10.14: add updated versions of cell-renderer and treeview-checkbox patches Aug 23 12:34:36 03mickeyl 07org.oe.dev * rdbdf31b9... 10/ (3 files in 3 dirs): remove old versions of libxslt and twisted Aug 23 12:34:44 bascically, we should be as close to what gumstix cretes through buildroot at first Aug 23 12:34:53 03mickeyl 07org.oe.dev * r9da4b755... 10/ (1 packages/e17/exhibit_cvs.bb packages/e17/expedite_cvs.bb): exhibit|expedite: use E_CVS variable Aug 23 12:35:05 03mickeyl 07org.oe.dev * re30e8794... 10/ (5 files in 2 dirs): qmake2: update to 2.10a from Qt/Embedded 4.3.1, thus removing X as dependency Aug 23 12:35:14 03pH5 07org.oe.dev * r9064298a... 10/ (1 packages/dbus/dbus-glib-native_0.72.bb): Aug 23 12:35:24 dbus-glib-native: stage dbus-binding-tool for eds manually Aug 23 12:35:33 * autotools_stage_all doesn't install binaries Aug 23 12:37:13 now I have a very "general" question: I have rebuilt my whole system. I would like to download it on my boad. With the dev board, I used to do it with the bootloader and TFTP. Now I would like to be able to do it directly on the Linux I am loged in (because I have no Ethernet, no TFTP and 6 MB on the serial port is very long !). How can I do this easily ? Aug 23 12:37:43 Crofton|home: the buildroot defconfigs break udev Aug 23 12:37:51 urg Aug 23 12:38:13 koen, where did teh verdex defconfig come from? Aug 23 12:38:24 that we have atm Aug 23 12:38:33 from the original defconfig, then fixed by me Aug 23 12:38:56 ok Aug 23 12:39:04 so I can diff Aug 23 12:39:07 and check Aug 23 12:39:28 did you update patched from gumstix site? Aug 23 12:39:47 no Aug 23 12:39:49 I found one small difference in koobla tarball he posted on bug tracker Aug 23 12:40:00 I will push the diff then Aug 23 12:40:04 vlo: you have any storage? Aug 23 12:42:03 koen, RP: apologies for the incorrect posting on oe-devel. Aug 23 12:42:26 I need to check the mailing list header more closely before replying on auto-pilot. Aug 23 12:43:00 hrw: the flash the filesystem is on and I can use the SD card too Aug 23 12:43:34 03koen 07org.oe.dev * r555ca250... 10/ (1 packages/bluez/bluez-libs_3.15.bb): bluez-libs: update to 3.15 Aug 23 12:43:46 * rwhitby says 50 hail-bitbakes and 20 blessed-angstroms as penance Aug 23 12:45:36 vlo: which bootloader you have? Aug 23 12:45:45 vlo: and what machine it is? Aug 23 12:46:13 rwhitby: heh Aug 23 12:46:43 hrw: bootloader is redboot and the machine is a custom mx31 board Aug 23 12:48:07 hmm.. never used redboot Aug 23 12:48:55 someone built qt4-x11-free recently? Aug 23 12:49:44 hrw: last week or so Aug 23 12:49:57 ok ... let's download it by serial port. Aug 23 12:52:03 vlo: you can do small rootfs partition in flash to be able to boot into it and then rewrite other flash parts Aug 23 12:52:34 hm when I remeber correctly redboot has acces to sd cards Aug 23 12:52:48 woglinde: u-boot for sure can have it Aug 23 12:52:57 u-boot has Aug 23 12:53:05 vlo: in SlugOS, we pivot-root to a ramdisk to reflash the jffs2 partition. Aug 23 12:53:06 the bootloader isn't the problem Aug 23 12:53:24 rob_w|laptop: ping Aug 23 12:53:30 pong Aug 23 12:53:45 rob_w|laptop: Are you up to anything on the 15th Sept? Aug 23 12:53:57 rob_w|laptop: (Software Freedom Day) Aug 23 12:54:06 what for ? Aug 23 12:54:18 rwhitby: yes that sounds a very interesting idea. Do you have a pointer to where I can find more info on this ? Aug 23 12:55:05 because with USB networking under Linux, the download is going to be much quicker Aug 23 12:55:07 pschulz01, do we know each other ? Aug 23 12:55:31 Are 'rod_w' and 'rwhitby' the same person? Aug 23 12:55:38 vlo look at initrd stuff in oe Aug 23 12:55:42 rob_w|laptop: Sorry.. Aug 23 12:55:47 rwhitby: ping Aug 23 12:56:00 ah Aug 23 12:56:15 rob_w|laptop: Jumped in too quickly. Aug 23 12:56:41 no problem Aug 23 12:57:02 vlo: http://www.openembedded.org/repo/org.openembedded.dev/packages/slugos-init/files/boot/ram - but that's a very slugos-specific implementation which should be made more generic before being used elsewhere. Aug 23 12:57:07 pschulz01: gu Aug 23 12:57:43 rwhitby: Right,.. now that I have the right person.. Aug 23 12:57:46 pschulz01: sorry, can't assist on that date. Aug 23 12:58:13 rwhitby: Rats .. any change of borrowing a phone? Aug 23 12:58:27 rwhitby: chance Aug 23 12:58:45 pschulz01: sure Aug 23 12:59:16 you can have both for the weekend if you like Aug 23 12:59:29 rwhitby: :-) Ta. Aug 23 12:59:56 drop by my office on friday to pick them up, and drop them back on monday. Aug 23 12:59:58 koen, what is teh gumstix defconfig that breaks udev, I am looking at diffs atm Aug 23 13:00:01 rwhitby: See - http://sfs07.mawsonlakes.org for details Aug 23 13:00:13 rwhitby: More than one? Aug 23 13:00:45 Crofton: networking -> socket Aug 23 13:01:19 rwhitby: Excellent.. thank you. Aug 23 13:01:33 pschulz01: Bv03 and Bv04 Aug 23 13:01:52 rwhitby: Can calls be made from one to the other? Aug 23 13:01:59 nope Aug 23 13:02:09 neither of mine make calls at the moment Aug 23 13:02:48 rwhitby: Np.. it will still good to display. Aug 23 13:03:12 you can show off the kinetic scrolling Aug 23 13:03:18 rwhitby: the libgsmd part of the problem got fixed today Aug 23 13:03:28 the dialer still segfaults Aug 23 13:03:29 rwhitby: Cool. Aug 23 13:03:47 pschulz01: things should be fixed by 15 september however Aug 23 13:04:15 rwhitby: I'm happy to show whatever I can get :-) Aug 23 13:04:49 rwhitby: I'll have the OLPC there. Aug 23 13:05:01 send me an email closer to the date to get the office address and to remind me to make sure I have both flashed and at work that day. Aug 23 13:06:40 rwhitby: Cool.. I'll do that. I'm in though with mithro as well.. so he may be able to collect them for me :-) Aug 23 13:06:58 rwhitby: ..In contact with.. Aug 23 13:07:43 03koen 07org.oe.dev * r0a1bbba5... 10/ (1 packages/bluez/bluez-utils_3.15.bb): bluez-utils: update to 3.15, get out your a2dp headsets :) Aug 23 13:07:58 morning Aug 23 13:08:03 hey chouimat Aug 23 13:08:30 pschulz01: depends when he starts ;-) Aug 23 13:13:20 anyone else build an image last night and have it miserably fail on the device? I have akita and my x11 image drops me to console.... seems to be missing pcmcia and /dev/null somehow? Aug 23 13:15:09 why doesn't the gumstix conf enable a FPE (the onf from buildroot) Aug 23 13:15:43 and what does koen mean by networking -> socket Aug 23 13:19:48 mallum: is there a pimlico ml archive ? Aug 23 13:20:07 and the final question, is why does teh gumstix build have a defconfig, and then patch it? Aug 23 13:20:49 ade|desk: lists.o-hand.com/ Aug 23 13:21:03 Crofton|home: because of idiots ;) Aug 23 13:21:10 koen, I know that :) Aug 23 13:21:25 actually, I was muttering about the lack of lfow upstream Aug 23 13:21:36 and the problem may be one overworked guy Aug 23 13:21:38 -CONFIG_PACKET=m Aug 23 13:21:39 +CONFIG_PACKET=y Aug 23 13:21:39 CONFIG_PACKET_MMAP=y Aug 23 13:21:40 -CONFIG_UNIX=m Aug 23 13:21:41 +CONFIG_UNIX=y Aug 23 13:21:42 ~lart the Aug 23 13:21:42 * ibot keeps mailing the free America Online CDs until he drowns Aug 23 13:21:45 ade|desk: yeah a 'beta' one - lists.o-hand.com Aug 23 13:21:50 ah Aug 23 13:21:57 you really want those as 'y, not as 'm' Aug 23 13:22:00 in kernel, not as modules Aug 23 13:22:03 got it Aug 23 13:22:17 Crofton|home: you can remove FPE as long as you use soft-float Aug 23 13:22:38 I think defconfig.patch turns on an emulator Aug 23 13:22:40 which not all platforms in OE have Aug 23 13:22:46 whomever posted the update on bugzilla indicated that they were able to get ethernet workign merely by switching the smc driver from a module to a y Aug 23 13:23:28 asmola, that was me - i boot via nfs Aug 23 13:23:49 ahhh, ok, i thought it was you but i wasn't sure, i'm new here :) Aug 23 13:23:57 no prb Aug 23 13:24:08 i drop by only now and then myself Aug 23 13:24:12 I am going to update teh defconfig with teh one from buildroot Aug 23 13:24:19 and fix the bits koen describes Aug 23 13:24:59 I suppose nfs boot is hopeless unless teh drier is built in? Aug 23 13:25:23 Crofton|home: could you update the connex defconfig as well please - teh config_packet is module there as well Aug 23 13:25:30 If I do this, you verdex guys can give me feedback, right? Aug 23 13:25:37 will do Aug 23 13:25:53 basically, I will do this and you guys can post boot logs Aug 23 13:26:01 yeah, not a problem Aug 23 13:26:09 no prb Aug 23 13:26:14 DISTRO_EXTRA_RRECOMMENDS += " \ Aug 23 13:26:14 kernel-module-vfat \ Aug 23 13:26:15 kernel-module-ext2 \ Aug 23 13:26:16 kernel-module-af-packet \ Aug 23 13:26:16 " Aug 23 13:26:29 I added that especially for messed up defconfigs Aug 23 13:26:56 that installs the modules henthe images are built? Aug 23 13:27:31 in images that use task-base, yes Aug 23 13:27:34 mallum , koen: thanks Aug 23 13:28:26 does this change the requirement on packet and unix being built in? Aug 23 13:28:49 not really Aug 23 13:28:53 ok Aug 23 13:29:00 the RRECOMMENDS is a last resort Aug 23 13:29:14 * Crofton|home needs to get this right so he can move linux-omap stuff over at some point Aug 23 13:29:38 also, the OABI/EABI selector works no matter what is in teh defconfig? Aug 23 13:33:07 alright, let me make sure these kernels build, and I will push Aug 23 13:44:25 Crofton: as long as the defconfig is called 'defconfig' in SRC_URI, the selector will work Aug 23 13:46:07 ok, my concern is fixes it up no matter what the exisitng ABI settings are Aug 23 13:46:13 my sed is really poor Aug 23 13:46:41 it deletes the line Aug 23 13:46:50 excellent Aug 23 13:46:52 and adds a new one in Aug 23 13:47:31 * koen threatens Crofton|home with clam chowder for not working on his thesis Aug 23 13:47:39 soon Aug 23 13:47:42 I need a break Aug 23 13:52:32 * koen remembers the neighbouts have moved out and turns up the volume Aug 23 14:00:38 koen: stop being awkward :P thats the 3d time i've said its possible to have differnt idle plugins... Aug 23 14:01:18 qemu builds in OE on AMD64 machine :-) Aug 23 14:01:37 i can try :) Aug 23 14:03:29 great Aug 23 14:03:40 the uni killed parts of the internet again Aug 23 14:03:53 which makes bitbake cry Aug 23 14:04:17 läs mailen! Aug 23 14:05:46 Crofton, all the changes are in right? if so, I am going to pull, build and test it out Aug 23 14:06:05 asmola, a couple of miutes Aug 23 14:06:09 the kernels build Aug 23 14:06:21 pushing .... Aug 23 14:07:14 .... done Aug 23 14:07:17 btw, Crofton|home, kernel bbclass could be extended with FILESPATH += "${FILE_DIRNAME}/${P}/${MACHINE}" so you can keep machine-specific patches in subdirs Aug 23 14:07:27 ok, try the sumstic and let us know what happens Aug 23 14:08:20 i will let you know Aug 23 14:08:28 thanks guys Aug 23 14:16:06 03crofton 07org.oe.dev * r76cb562b... 10/ (1 packages/linux/linux-2.6.21/pxafb-definition.patch): linux : Update pxafb-definition.patch from gumstix svn. Aug 23 14:23:58 Crofton|home: http://lists.linuxtogo.org/pipermail/openembedded-commits/2007-August/008133.html Aug 23 14:24:05 -CONFIG_LOCALVERSION="" Aug 23 14:24:06 +CONFIG_LOCALVERSION="gum" Aug 23 14:24:15 koen: Thanks for the bluez update. Time to buy myself a nice A2DP headset. :) Aug 23 14:24:22 ok :) Aug 23 14:24:28 I'll fix in a few minutes Aug 23 14:24:32 stefan_schmidt: you get one donated by FIC :) Aug 23 14:24:46 s/you/you mean/ Aug 23 14:25:35 koen: I fear not. But I should be able to buy one myself with their money. ;) Aug 23 14:25:43 :) Aug 23 14:26:05 hmm, perhaps I should ask. Somebody needs to test this... ;) Aug 23 14:26:24 Anyone tips for a good A2DP headset? Aug 23 14:26:25 why ween need python.tk Aug 23 14:26:34 that sneaks in libx11 Aug 23 14:26:43 even for a image which dont uses x11 Aug 23 14:27:28 no it doesn't Aug 23 14:27:43 x11 for python is a *buildtime* dep Aug 23 14:27:49 not a *runtime* dep Aug 23 14:28:19 *sigh* lasting build one hour more Aug 23 14:31:19 i'm looking right now on the embeddedarm TS-7800 sound nice (but not exacly cheap) ... did anyone already try or know the marvel processor ?? (500MHz arm9) Aug 23 14:31:29 why don't we set LOCALVERSION? Aug 23 14:33:15 .clear Aug 23 14:41:21 03crofton 07org.oe.dev * r76cb562b... 10/ (1 packages/linux/linux-2.6.21/pxafb-definition.patch): linux : Update pxafb-definition.patch from gumstix svn. Aug 23 14:45:35 03crofton 07org.oe.dev * r76cb562b... 10/ (1 packages/linux/linux-2.6.21/pxafb-definition.patch): linux : Update pxafb-definition.patch from gumstix svn. Aug 23 14:46:38 Crofton|home: testing kernel will take some time, as i need to update uboot - machine is set to gumstix-connex now, which is not recognized by the boot loader Aug 23 14:48:22 I wish I had some gumstix to test on, it is tough supporting stuff I do not have direct access to Aug 23 14:48:36 thanks for your help Aug 23 14:49:10 will see what i can do Aug 23 14:50:02 UBOOT_MACHINE = "gumstix" in gumstix.inc? Aug 23 14:51:38 koen, that solves problem? Aug 23 14:54:20 hmm. digging. I get 'Bad Magic Number' so the problem is somewhere else... Aug 23 14:56:14 there is an old and new boot method for conne boards Aug 23 14:56:22 or so I hear Aug 23 14:58:54 ok, can't mount my nfsroot... boots into local fine Aug 23 14:59:34 03crofton 07org.oe.dev * r76cb562b... 10/ (1 packages/linux/linux-2.6.21/pxafb-definition.patch): linux : Update pxafb-definition.patch from gumstix svn. Aug 23 15:01:02 Crofton|home: the kernel boots my original gumstix environment, will recompile with nfsroot support tomorrow and see if angstrom works as well Aug 23 15:01:23 ok Aug 23 15:01:30 how do I add a package that should be built with my machine ? I guess I have to add something in the machine-name.conf file, but what is the variable name ? is it MACHINE_EXTRA_RDEPENDS ? Aug 23 15:01:46 a guy is working on gumstix to send me some boards so I can test, hopefully this happens Aug 23 15:02:59 03crofton 07org.oe.dev * r76cb562b... 10/ (1 packages/linux/linux-2.6.21/pxafb-definition.patch): linux : Update pxafb-definition.patch from gumstix svn. Aug 23 15:03:34 Can someone tell me where the file is with the IPKG_INSTALL? I see that the final rootfs has a kernel and gdb and some ipk stuff that I won't need in my build and I want to remove those... Aug 23 15:06:15 03crofton 07org.oe.dev * r76cb562b... 10/ (1 packages/linux/linux-2.6.21/pxafb-definition.patch): linux : Update pxafb-definition.patch from gumstix svn. Aug 23 15:06:26 03crofton 07org.oe.dev * rd7266d01... 10/ (1 packages/linux/linux-2.6.21/gumstix-verdex/defconfig): linux : Update defconfig for gumstix verdex from gumstix svn. Aug 23 15:06:39 03crofton 07org.oe.dev * rb7def8e0... 10/ (1 packages/linux/linux-2.6.21/gumstix-connex/defconfig): linux : Update defconfig for gumstix connex/basix from gumstix svn. Aug 23 15:06:52 03crofton 07org.oe.dev * r0632006d... 10/ (3 files in 3 dirs): linux : Do not set LOCALVERSION in kernel defconfig. Aug 23 15:06:57 * * OE Bug 2843 has been created by jstoefer(AT)schuett-elektronik.de Aug 23 15:06:59 * * unmet dependency on bluez-dtl1-workaround Aug 23 15:07:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2843 Aug 23 15:07:03 03xora 07org.oe.dev * r045e5642... 10/ (1 packages/gstreamer/gst-plugins-bad_0.10.5.bb): gst-plugins-bad_0.10.5.bb : disable directfb as its broken Aug 23 15:07:28 christopher: you can exclude the kernel image from your build by placing the following in the kernel recipe: FILES_kernel-image = "" Aug 23 15:08:29 cbrake: ok... Aug 23 15:11:05 vlo: MACHINE_EXTRA_RDEPENDS would be my guess based on exsiting conf files Aug 23 15:11:55 cbrake: so, if it is building 2.6.21, can I just have it not build it? or do I just place that in the 2.6.21.bb file? Aug 23 15:12:47 cbrake: I already have linux and u-boot up and going. I was just looking for something to build the rootfs nicely other than buildroot Aug 23 15:12:51 christopher: it is best to add a kernel recipe for your build, so that you can put kernel modules into the rootfs Aug 23 15:13:04 cbrake: ok. guess that makes sense then. Aug 23 15:13:14 christopher: otherwise, you will need to track down the dependency and figure out how to remove it. Aug 23 15:13:21 christopher: bitbake -g might help with that Aug 23 15:13:43 christopher: but you are fighting the normal flow of things :-) Aug 23 15:13:48 cbrake: so, just put that line in the bb file for .21 so it doesn't go in the rootfs? I was reading in the manual that there is something that specifies which packages to install. Aug 23 15:14:03 cbrake: yeah, I don't want to fight it...just want to go along with it and tailor it to my board Aug 23 15:14:19 christopher: what section of the manual? Aug 23 15:14:31 cbrake: rootfs_ipk Aug 23 15:14:49 cbrake: p106 Aug 23 15:17:11 christopher: I'm not sure IPKG_INSTALL is meant to be used directly Aug 23 15:19:58 * * OE Bug 2844 has been created by skoroneos(AT)digital-opsis.com Aug 23 15:20:00 * * lograte fails to package Aug 23 15:20:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2844 Aug 23 15:20:33 christopher: looks like it is now name PACKAGE_INSTALL Aug 23 15:21:23 christopher: and then you would set it in your [image].bb Aug 23 15:22:28 christopher: grep the OE tree for PACKAGE_INSTALL and you will see examples of how it is used. Aug 23 15:24:45 christopher: to remove things from PACKAGE_INSTALL, you will likely need to create you own image bb Aug 23 15:24:57 * * OE Bug 2432 has been RESOLVED (FIXED) by skoroneos(AT)digital-opsis.com Aug 23 15:24:59 * *  Mesa fails on Efika Aug 23 15:25:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2432 Aug 23 15:31:37 cbrake: ok, I'll take a look for that and see what I can find. thanks! Aug 23 15:34:57 * * OE Bug 2845 has been created by gremlin(AT)gremlin.it Aug 23 15:34:59 * * gtk+-2.10.14-r5 fail to compile Aug 23 15:35:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2845 Aug 23 15:36:00 who will take care of this bug have just to take a look at one of the patch applied that seem to me wrong ... Aug 23 15:42:58 ~listppaste7 Aug 23 15:43:06 ~lisppaste7 Aug 23 15:43:12 lisppaste7: help Aug 23 15:43:12 To use the lisppaste bot, visit http://paste.lisp.org/new/oe and enter your paste. Aug 23 15:43:16 sorry Aug 23 15:43:36 asmola_work: no need to apologize Aug 23 15:44:53 Crofton|work__, build from scratch failed Aug 23 15:44:59 asmola_work pasted "build_fail" at http://paste.lisp.org/display/46605 Aug 23 15:46:05 bugger all Aug 23 15:46:26 ~lart uboot-mkimage Aug 23 15:46:27 * ibot raises middle finger to uboot-mkimage Aug 23 15:46:34 rofl Aug 23 15:46:49 standby Aug 23 15:47:19 why didn't the dependency fire? Aug 23 15:50:17 because some idiot renamed the machines Aug 23 15:50:41 which one of us idiots? Aug 23 15:51:24 * koen <- Aug 23 15:51:31 asmola_work, I have fired off a build at home Aug 23 15:52:02 asmola_work, I have comprehended the idiot remark :) Aug 23 15:52:10 roflmao Aug 23 15:52:16 of course, I should have seen that Aug 23 15:52:30 koen, you fix, or do you want me to fix/ Aug 23 15:53:03 mtn pull Aug 23 15:53:12 asmola_work, pull Aug 23 15:53:23 pulling Aug 23 15:53:33 bitbake -c rebuild linux, then build the image :) Aug 23 15:53:48 thanks koen Aug 23 15:53:51 again Aug 23 15:53:54 thanks guys Aug 23 16:02:01 What does the bitbake message "NOTE: the following files were installed but not shipped in any package" imply? Why are installed files not shipped? Aug 23 16:03:03 03ifaistos 07org.oe.dev * rdf87aa96... 10/ (7 files in 3 dirs): packages/asterisk/asterisk_1.2.18.bb : Remove old asterisk version Aug 23 16:03:14 03ifaistos 07org.oe.dev * rf5d5db62... 10/ (7 files in 3 dirs): dropped packages/asterisk/asterisk_1.2.23.bb : remove old asterisk version Aug 23 16:03:21 03ifaistos 07org.oe.dev * r308da270... 10/ (8 files in 3 dirs): packages/asterisk/asterisk_1.4.11.bb : Add asterisk 1.4.11 Aug 23 16:03:32 cyberdeck: that depends. It can be a packaging error, new files from a new release where the bb file was just moved, or left out on purpose (e.g. with busybox) Aug 23 16:03:32 03koen 07org.oe.dev * rca7f351c... 10/ (1 packages/linux/linux.inc): linux.inc: catch up with depends after gumstix changes Aug 23 16:03:39 cyberdeck: files are installed to a temp directory, and then packaged into ipk files. "not shipped" probably means they did not make it into any ipk files Aug 23 16:04:22 zecke: I tried to package the helloworld sample. but no ipk is created. Aug 23 16:04:54 cyberdeck: sample from where? interested in fixing it with us? Aug 23 16:05:36 zecke: the "helloworld-1.0.0-r0" from "org.openembedded.dev". Aug 23 16:06:03 zecke: and yeah. it would be a pleasure for me to fix it with you! Aug 23 16:07:26 03koen 07org.oe.dev * r68489b71... 10/ (1 packages/linux/linux_2.6.21.bb): linux 2.6.21: bump PR after all those changes Aug 23 16:07:54 zecke: hmm. it is no .bb error. it must be my filesystem. if i do a "bitbake -f -c package helloworld" afterwards it runs without an error. Aug 23 16:08:38 cyberdeck: package has already been run, so unless you clean the package, delete the timestamp, etc, bitbake considers it done Aug 23 16:09:01 cbrake: therefore i "-f" forced it to rerun. Aug 23 16:09:43 cyberdeck: ahh Aug 23 16:10:22 cyberdeck: I'm just done with parsing Aug 23 16:10:47 cyberdeck: I have two ipk's Aug 23 16:13:16 zecke: if i do "bitbake -f -c rebuild helloworld" I have zero ipk's. Aug 23 16:13:39 zecke: if i follow with a ""bitbake -f -c package helloworld" afterwards i have also two ipks. Aug 23 16:13:57 zecke: btw I use bitbake 1.8.8 Aug 23 16:16:04 * zecke too, from the dir that was used to create the 1.8.8 release ;) Aug 23 16:16:38 but I admit to have a older version of OE :) Aug 23 16:17:12 older version means "days older" or "hours ago"? Aug 23 16:17:40 cbrake: ok, I'm lost looking around at these bb files. I am just doing the angstrom distro with a sarge-at91 machine. I see where the install is, but then that points onward. for example, where does gdb get put into the rootfs from an ipk? Aug 23 16:18:06 cyberdeck: days :) Aug 23 16:18:23 maybe something is broken now? Aug 23 16:21:42 * cyberdeck waves to say goodbye. I'll leave work soon and will be in 30min @home! Aug 23 16:29:03 hmm, adding a md5sum=e34320c97d768191217424af64de380b check to a SRC_URI does not trigger a failure if the packages has already been downloaded. Aug 23 16:29:19 and checksum.ini? Aug 23 16:29:38 koen: let me check ... Aug 23 16:29:51 koen: is checksum.ini the preferred location for this Aug 23 16:30:34 cbrake: right, it will only check on do_fetch Aug 23 16:30:47 cbrake: you could move the test to do_unpack, or do it in both tasks Aug 23 16:31:05 cbrake: moving would be bad as the do_fetch task wouldn't fail Aug 23 16:31:55 zecke: both seems right in this case, as I want to force a check on unversioned package download. Novell is not versioning their 1.2.5preX mono downloads. Aug 23 16:35:08 cbrake: feel to copy the invocation to do_unpack Aug 23 16:35:19 zecke: ok Aug 23 16:35:38 cbrake: and our fetcher should add a version to the tarball :} Aug 23 16:37:08 zecke: so should I copy the md5 check to do_unpack just for mono, or globally? Aug 23 16:40:00 cbrake: globally in base.bbclass, hopefully it is just a call to the check method + for loop for the SRC_URI's Aug 23 16:41:05 zecke: is checksums.ini updated automatically, or should I commit changes to it Aug 23 16:42:43 cbrake: commit... Aug 23 16:43:00 cbrake: any hints on how to commit it automatically? Aug 23 16:45:13 zecke: no, I just thought maybe someone had a golden downloads directory, and the checksums were genereted from the that, but I agree the checksums should probably be entered by the package maintainer. Aug 23 16:46:32 cbrake: we generated most of them from a golden downloads directory (CELF) Aug 23 16:46:44 zecke: nod Aug 23 16:47:12 but I would love to see people adding checksums... Aug 23 16:47:38 zecke: looks like if the checksums are entered in checksums.ini, do_fetch does flag an error on rebuild, even if the package has already been downloaded. Aug 23 16:48:17 zecke: so this should be adequate for the mono case. But the behaviour is different than md5sum in the SRC_URI Aug 23 16:49:01 cbrake: md5sum, was from bitbake, checksums is handled by OE/base.bbclass. But in both cases no do_fetch stamp should be created Aug 23 16:49:17 cbrake: option b) is to just add the correct tarball into OE Aug 23 16:51:02 hi! Aug 23 16:51:16 koen: how do I add a tarball to OE? Aug 23 16:51:40 cp tarball package/mono ; mtn add tarball ; mtn commit tarball Aug 23 16:52:05 pH5: hey pH5. somebody reported about wizard with OE yet? Aug 23 16:52:25 03koen 07org.oe.dev * rd839f215... 10/ (1 packages/linux/linux_2.6.22+2.6.23-rc3.bb): linux: add 2.6.22+2.6.23-rc3 for at91 machines Aug 23 16:52:48 koen: I hate to bloat the mtn repo with a 17MB mono tarball Aug 23 17:02:50 dcordes: don't know Aug 23 17:21:44 hello Aug 23 17:22:38 hi, yakov Aug 23 17:23:09 crofton, just did an mtn pull & update and i am seeing an error on the bitbake -c rebuild linux Aug 23 17:25:04 asmola_work pasted "failed build (gumstix)" at http://paste.lisp.org/display/46609 Aug 23 17:26:01 koen, it looks like there is still an issue with the renamed gumstix stuff Aug 23 17:27:14 Hey I am getting this error http://pastebin.ca/668180 Aug 23 17:27:33 on 5912osk build from yesterdays mtn Aug 23 17:37:34 bye Aug 23 17:38:30 asmola_work: could you try 'bitbake u-boot-mkimage-openmoko-native' Aug 23 17:38:57 sure, that's what i did a few days ago when i had a problem that lokoed like this and it worked (crofton told me to try it) Aug 23 17:39:33 linux.inc seems to do the right thing: DEPENDS_gumstix-verdex = "u-boot-mkimage-openmoko-native" Aug 23 17:41:05 koen, uboot-tools noew builds mkimage Aug 23 17:41:49 but not uboot-mkimage Aug 23 17:44:27 it worked as of last night prior to the gumstix split Aug 23 17:45:11 well, that is just a name Aug 23 17:45:28 and if you use the kernel uIamge target you need it to be mkimage :) Aug 23 17:45:53 anyone knows what does ``"build crosses midnight GMT" bitbake bug'' mean?! Aug 23 17:46:34 re ph5 Aug 23 17:46:45 hej woglinde Aug 23 17:47:07 03stefan 07org.oe.dev * r64b4d4f1... 10/ (1 packages/openmoko2/openmoko-dialer2_svn.bb): openmoko-dialer2: Switch dependency. The dialer uses pulseaudio instead of gstreamer now. Aug 23 17:48:09 drat Aug 23 17:48:16 wot? Aug 23 17:48:26 asmola_work, I am running a console image, so still building .... Aug 23 17:48:28 * koen discovered 'bridges' on oh-puzzles Aug 23 17:49:08 koen: timesink? Aug 23 17:49:33 w00t! Aug 23 17:49:39 finally have an image to burn of my own! ;) Aug 23 17:57:04 * khem ftp.rpm.org seems to be down Aug 23 17:57:42 my build is stuck at NOTE: fetch ftp://ftp.rpm.org/pub/rpm/dist/rpm-4.1.x/popt-1.7.tar.gz Aug 23 17:59:23 the angstrom source mirror should have that Aug 23 18:00:23 crofton, minimal image completed with new kernel, im going to burn them right now, brb Aug 23 18:01:16 koen: I dont think its enquiring that Aug 23 18:01:28 press ctrl-c Aug 23 18:04:45 asmola_work, great! Aug 23 18:04:57 I could stand some success :) Aug 23 18:05:23 :-) Aug 23 18:06:36 koen: is there any timeout on fetch () Aug 23 18:06:50 yes, when wget times out Aug 23 18:06:53 (or svn or git) Aug 23 18:07:20 hmmm it seems that it was stuck there for circa 30 mins Aug 23 18:07:47 it took 2 Ctrl Cs to make it chose from local repo Aug 23 18:12:13 cbrake: what sort of command line do you guys use for the kernel with the angstrom console image. I got everything to build and it loaded up the filesystem, but it just sits there and doesn't run a login or a console although I do see a /dev/ttyS0 in the rootfs Aug 23 18:16:07 crofton, it booted up Aug 23 18:18:05 great Aug 23 18:18:17 how many error/warnings? Aug 23 18:18:26 Did it find the coprocessor? Aug 23 18:18:35 yes Aug 23 18:18:40 ill post a bootlog Aug 23 18:18:49 thanks Aug 23 18:18:57 now if we get some testing on teh connex Aug 23 18:19:55 asmola_work pasted "gumstix-verdex-boot" at http://paste.lisp.org/display/46612 Aug 23 18:20:20 i've bene going back with sakoman via email, he is working on it Aug 23 18:20:27 cool Aug 23 18:20:31 we owe you guys Aug 23 18:20:38 for testing this stuff Aug 23 18:20:55 not a problem, i think we owe you guys! lol Aug 23 18:21:23 i have a connex.... Aug 23 18:21:31 it is now detecting the clock speed, it wasn't doing that before Aug 23 18:21:32 i have no idea where it is though Aug 23 18:21:43 buried in a sea of gadgets Aug 23 18:21:45 :) Aug 23 18:21:45 timtimred, :) Aug 23 18:22:00 i will test it out if i find it Aug 23 18:22:13 :) Aug 23 18:22:15 ok, we need to turn on tmpfs in kernel config Aug 23 18:23:08 i'm not seeing an eth0 still when i do an ifconfig -a Aug 23 18:23:27 or do i need a console image for ethernet support? Aug 23 18:23:32 lsmod Aug 23 18:23:44 try modprobe (the name of the module) Aug 23 18:23:47 i guess gumstix are still using buildroot!!! aghhh :) Aug 23 18:23:57 lsmod sees no modules Aug 23 18:24:16 see if they are installed on the image Aug 23 18:25:33 timtimred, we are updating OE support Aug 23 18:25:50 asmola_work, the biggest problem I see in the log is lack of tmpfs Aug 23 18:25:51 christopher: it depends a lot. This will give you some examples: find -name defconfig | xargs grep CMDLINE Aug 23 18:25:55 crofton, sorry, how do i do that? Aug 23 18:26:07 I will add that to defconfig when I get a kernel built again Aug 23 18:26:28 see if there are modules? Aug 23 18:26:47 yeah there are Aug 23 18:26:49 look in /lib/modules/...... Aug 23 18:27:01 modprobe the ethernet one Aug 23 18:27:11 hmmm I am getting gtk+ build failure Aug 23 18:27:13 http://pastebin.ca/668247 Aug 23 18:27:27 now when I updates today Aug 23 18:27:27 khem when did you pull? Aug 23 18:27:38 just an hour back may be less Aug 23 18:27:46 you are building an x11 image? Aug 23 18:28:00 I really like the ascii logo steliosk made for us Aug 23 18:28:13 i see some modules here but not the eternet one Aug 23 18:28:17 nope Aug 23 18:28:17 on Angstrom Aug 23 18:28:27 I am building angstrom-console-image Aug 23 18:28:31 hmm Aug 23 18:28:34 for omap5912osk Aug 23 18:28:36 me too, except for gumstix Aug 23 18:28:44 that is running at work Aug 23 18:28:57 ascii logo is pretty impressive Aug 23 18:30:29 asmola, what modules are there Aug 23 18:30:58 alias, ccwmap, dep, ieee1394map, inputmap, isapnpmap, ofmap, pcimap, seriomap, symbols, usbmap Aug 23 18:32:25 hmm, how do I force it to install the network module on the image no matter what Aug 23 18:32:33 do you know what the module is called? Aug 23 18:32:45 i have it written down Aug 23 18:32:47 hold on Aug 23 18:33:07 maybe look in the ipks directory for modules .... Aug 23 18:33:33 its called networking AFAIR Aug 23 18:36:17 I hate the idea of building in the kernel, since I think you can run a gumstix without the networking interface attached? Aug 23 18:36:25 i don'tcorrect Aug 23 18:36:32 sorry, two responses in one Aug 23 18:36:47 yes you can run the gumstix without the network interface Aug 23 18:37:54 i guess i don't know what the module is called, sorry, although i am looking in koobla's reply to the bug and he changed the config_smc91x line, but the verdex network card doesn't use the smc91x it uses the smc911x, which appears to be commented out and a few lines down in the patch file Aug 23 18:38:37 hmm Aug 23 18:39:13 i take that back Aug 23 18:39:24 do you minf switching gears a little and working through setting up ossie and BBCOLLECTION while I wait for my buidl to finish? Aug 23 18:39:26 they have the same chipset Aug 23 18:39:54 its technically an SMC LAN91C111 on both Aug 23 18:40:13 i don't know why they need that many ones in the model number (well i suppose to confuse people) Aug 23 18:40:21 sure, i have been wanting to work on that actually Aug 23 18:40:57 is it possible to search mail archive? Aug 23 18:41:03 oe-devel.. Aug 23 18:41:46 asmola_work, http://pastebin.ca/668271 Aug 23 18:42:03 apply common sense to the paths Aug 23 18:42:23 excellent Aug 23 18:42:28 will do Aug 23 18:42:41 I was hoping to test this b4 you got here :) Aug 23 18:42:56 haha Aug 23 18:42:57 I have changed the paths, and haven't tested since our symposium :( Aug 23 18:43:27 next symposium you guys have i will have to come down, i know that jeff came down for that one Aug 23 18:43:44 they are good fun Aug 23 18:43:51 cbrake: am I wrong when I look at the inittab that it doesn't use busybox for it's init/console? the issue I had was that the jffs2 patches weren't applying so I just let it build the rootfs and then make the rootfs using mkfs.jffs2...but it may have only had the utils for ext2 instead of jffs2. the ext2 image was 60 MB! while the rootfs directory was only 15MB Aug 23 18:44:20 brb Aug 23 18:44:25 np Aug 23 18:45:27 christopher: I'm not sure I follow everything. The ext2 image is formatted to a specific size, so it will be as large as the ext2 filesystems specified. Aug 23 18:45:37 christopher: there is a variable that defines this size Aug 23 18:45:56 cbrake: ok. I guess i need to look for that block size. I just figured it would be set to auto size it Aug 23 18:46:18 but with the directory at 15MB, that ext2 file will probably be bigger than 16 MB (size of my flash) Aug 23 18:46:45 christopher: yes, use jffs2 for flash Aug 23 18:46:53 cbrake: yeah, sorry for the bad explanation. basically, I went to build it for jffs2, and it failed on some of the patches. Aug 23 18:47:13 cbrake: so, I took that out although jffs2 was what I wanted. Aug 23 18:47:15 christopher: if you want to pastebin the jffs2 problem, I'll take a look. Aug 23 18:47:27 christopher: I've not had any problems with jffs2 for a long time. Aug 23 18:48:38 cbrake: ok. I am starting another build and I'll see what it was. I'll see if I can find the bb file with the patches specified...I had to comment out two of them Aug 23 18:49:18 christopher: ok. I'm not sure about the init/console issue either without some digging as it is just working for me Aug 23 18:49:58 cbrake: what are you running it on? is it an arm? Aug 23 18:50:31 hm, guys, there's no way in real to search MLs? Aug 23 18:50:46 i couldnt find one one oe.org.. :/ Aug 23 18:51:36 christopher: yes, compulab-pxa270 machine Aug 23 18:51:44 yakov: google it with "+site:ml-archive-adress ". at least it works for me (TM). Aug 23 18:52:09 thanks! Aug 23 18:53:01 cbrake: ok. cool. Can I ask what sorts of graphics type things you use? that is, if it is an arm9, can you play video/mp3s on it ? We have 16MB of ram currently and this is what I am trying to get running. I am using the sarge-at91 machine file Aug 23 18:55:01 shit. how this could happend. directory build.i386-darwin.arm-angstrom-linux-gnueabi is empty under binutils temp building dir... Aug 23 18:55:18 of-course there's no do_configure which is needed by bitbake then.. :( Aug 23 18:57:48 christopher: you can't do much with 16MB of ram these days Aug 23 18:58:16 cbrake: yeah, figured that. I didn't design the hardware and have requested that we go to a two chip solution for at least 32MB of ram. how much is on yours? Aug 23 18:58:53 christopher: you might be able to get directfb or something light running, but most people these days run gtk and then you at least 32 MB ram -- I would not go less than 64MB Aug 23 19:00:05 cbrake: ok. so, basically, with 16MB of ram, what could I run on top of tiny-x ? anything? not even matchbox? Aug 23 19:01:11 khem, I will commit linux-omap1 stuff in a bit Aug 23 19:01:48 christopher: anything you do with 16MB will be a lot of work. You might get matchbox to run, but then you still need a graphical toolkit. Aug 23 19:01:53 Crofton|home: thanks ! Aug 23 19:02:01 Crofton|home: How is theris going on ? Aug 23 19:02:06 slow Aug 23 19:02:14 been editing chapter 3 for days Aug 23 19:02:20 cbrake: ok. that's good info then. it's hard to know before you start getting into it. Aug 23 19:02:24 what is your topic is it for MS or PhD Aug 23 19:02:34 MS Aug 23 19:02:49 christopher: nod Aug 23 19:02:53 basically system reseource usage measurements of a software defined radio Aug 23 19:03:21 wow EE project seems like Aug 23 19:03:26 yeah Aug 23 19:03:29 sort of Aug 23 19:03:42 work for EE's that like software :) Aug 23 19:03:55 christopher: the tighter your contraints are, your work will grow exponetially. So its not so much a question of can it be done, but do I have 2 man-years to try to make it work. Aug 23 19:04:24 how could anyone not like the sca? :) Aug 23 19:04:42 christopher: with things as complex as Linux in embedded systems, its best to stay on the beaten path unless you have a lot of dev resources :-) Aug 23 19:05:48 does uclibc support internationalization Aug 23 19:05:56 broken Aug 23 19:05:58 * * OE Bug 2846 has been created by carbonfreeze(AT)gmail.com Aug 23 19:06:00 * * Request: add package htop to Angstrom feeds Aug 23 19:06:01 locales Aug 23 19:06:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2846 Aug 23 19:06:02 hi Aug 23 19:06:06 cbrake: yeah, no doubt. the guys I work with aren't working on the software but want it to be the least amount hw wise. Aug 23 19:06:20 so it has limited support for internationalization Aug 23 19:06:24 cbrake: I had to push to get them to up it to 16MB (it was 8 till I got the kernel and fs stable and up) Aug 23 19:06:49 crofton, about how big is this ossie-image? Aug 23 19:06:51 cbrake: oe seems like it maintains things well and that things generally build and is why I would like to learn how it is structured so we can use it for rootfs work Aug 23 19:07:18 cbrake: lots of programs out there and I'd rather not reinvent the wheel and struggle with compiling this or that or knowing what version of what works with what. Aug 23 19:07:34 cbrake: so many people have worked really hard on it, it would be foolish not to use it Aug 23 19:07:39 christopher: you might want to note down OE revisions that work for you if work against .dev. It's moving forward quite rapidly, and stuff is sometimes broken inbetween. Aug 23 19:07:48 hmmm Aug 23 19:07:50 I forget Aug 23 19:07:56 it fits in 32M of flash Aug 23 19:08:00 :) Aug 23 19:08:07 that's what i was looking for Aug 23 19:08:07 I run it on the OSK Aug 23 19:08:21 it will build gnuradio also :) Aug 23 19:08:22 i know that i read the specs on the OSK but it was escaping me how big the flash was Aug 23 19:08:34 I need to mess around with how we get the libusrp Aug 23 19:08:45 likewise: yeah, that is what I am working with. I haven't gotten the fs up and going but did finally get the console-image to build which I was happy to see. but that is why buildroot is frustrating because they change things so much that nothing ever seems to build Aug 23 19:09:07 christopher: yup, OE is the way to build and maintain a rootfs Aug 23 19:09:52 did you have the gnu radio plugged into the osk, does the osk have usb 2.0? Aug 23 19:10:26 cbrake: yep. I am a little over my head at the moment trying to follow one bb file to the next and figure out where things get added in. I still don't see where the kernel gets added to the fs or some other programs like gdb that are 2MB+ Aug 23 19:12:30 christopher: RDEPENDS drives everything that gets added to a rootfs. If you trace through RDEPENDS, you can figure things out. Aug 23 19:14:22 christopher: if you grep the conf directory, you will see that gdb is included in distro/angstrom-2007.1.conf Aug 23 19:14:42 cbrake: yeah, I saw that in the console image...so, it calls out task base extended, extra install which is empty, and that the base contains MACHINE_FEATURES, and others which are defined elsewhere I suppose. so, then you just fine the task-base-extended and follow it Aug 23 19:15:00 cbrake:oh...I guess I wasn't looking in the conf directory. I was looking in the packages Aug 23 19:17:09 i see several *.bb files unde packages/binutils, how is one picked when i issue bitbake binutuils with latest distro? Aug 23 19:17:21 can i manualy build *latest* *.bb file? Aug 23 19:17:36 i mean, could this manual build break distro sanity later? Aug 23 19:17:56 christopher: yes, packages/tasks/... Aug 23 19:19:26 christopher: so task-base is where it all starts. Any package that task-base includes can have RDEPENDS to additional packages Aug 23 19:19:30 Crofton|home: till last year or so, this was the biggest tent in the world: http://www.flickr.com/photos/koenkooi/1215986270/in/set-72157601612773118/ Aug 23 19:21:31 cbrake: so, those determine which are built...and which are installed is determined by something dealing specifically with the rootfs, right? Aug 23 19:22:11 christopher: actually, you start at the [image].bb file and work backwards from there. Aug 23 19:22:34 cool, where is the debauchery? Aug 23 19:22:40 christopher: so it is generally image.bb->task.bb->RDEPENDS Aug 23 19:26:33 asmola_work, looks like tmpfs is in the kernel ... Aug 23 19:27:56 ahhh Aug 23 19:28:28 cbrake: just found the bitbake.conf file with the ext command in it. searching for the IMAGE_ROOTFS_SIZE now to see where that is. perhaps it is set in this distro which causes that file to be so large Aug 23 19:28:40 03cbrake 07org.oe.dev * r0626c248... 10/ (8 files in 4 dirs): Aug 23 19:28:44 mono-1.2.5pre5: add mono 1.2.5pre5 Aug 23 19:28:46 first version that runs on ARM EABI without Aug 23 19:28:48 floating point problems. Aug 23 19:28:50 Still needs a lot of packaging work. Aug 23 19:28:57 03cbrake 07org.oe.dev * r5be94893... 10/ (1 MAINTAINERS): MAINTAINERS: update my entry Aug 23 19:30:28 damn ... with filebrowser (oe web site) is impossoble to div einto gtk+ dir cause of '+' sign ... uff uff Aug 23 19:39:54 um, I am having a realy crappy time pulling and updating ...... Aug 23 19:40:47 Crofton|home: tried ewi546.ewi.utwente.nl? Aug 23 19:41:08 http://rafb.net/p/f9hRX923.html Aug 23 19:41:22 weird stuff going on Aug 23 19:42:06 Crofton|home: hmm, my local checkout is that exact revision Aug 23 19:42:17 Crofton|home: had no probs here Aug 23 19:43:01 bother Aug 23 19:44:12 do you guys use a gui for the monotone stuff? Aug 23 19:44:19 no Aug 23 19:44:27 no Aug 23 19:44:31 gui's are for the weak Aug 23 19:44:33 :) Aug 23 19:44:57 * * OE Bug 2183 has been RESOLVED (FIXED) by cliff.brake(AT)gmail.com Aug 23 19:44:59 * *  Mono and mcs throw a SIGILL exception after executing Aug 23 19:45:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2183 Aug 23 19:45:05 perhaps but sometimes I perform a pull/update and really regret it Aug 23 19:45:09 * * OE Bug 2847 has been created by carbonfreeze(AT)gmail.com Aug 23 19:45:11 * * gpe-lock-display on screen keyboard is cut-off when using rotated screen Aug 23 19:45:13 * * http://bugs.openembedded.org/show_bug.cgi?id=2847 Aug 23 19:45:39 keesj: how does a gui help? you want to go back easily? Aug 23 19:46:25 I don't know perhaps you can "tag" or select a more advanced method Aug 23 19:46:53 or even preview the files that are changed Aug 23 19:49:12 I do use revert to remove my own changes Aug 23 19:56:34 preview, you can, hold on. Aug 23 19:57:14 I have a shell alias called "mtn-in" to preview incoming, I think I submitted it to the OE wiki Aug 23 19:57:44 keesj: ^^^^ mtn diff -r `mtn automate get_base_revision_id` -r `mtn automate heads` Aug 23 19:58:26 keesj: run this after a pull, before the update Aug 23 20:00:57 03likewise 07org.oe.dev * rb7d8ca31... 10/ (3 files in 3 dirs): x86-prescott: Added new machine. Aug 23 20:01:58 * * OE Bug 2848 has been created by autobuild(AT)openembedded.org Aug 23 20:01:58 03koen 07org.oe.dev * r9b04f62c... 10/ (1 packages/mono/mono_1.2.5pre5.inc): mono: fix PV for 1.2.5pre so we don't need an epoch for the 1.2.5 release (see our versioning policy at http://www.openembedded.org/versioningpolicy ) Aug 23 20:02:00 * * gnuradio-3.0.4+svnr6169-r2-do_compile Aug 23 20:02:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2848 Aug 23 20:02:09 * * OE Bug 2849 has been created by autobuild(AT)openembedded.org Aug 23 20:02:12 * * pulseaudio-0.9.6-r5-do_compile Aug 23 20:02:14 * * http://bugs.openembedded.org/show_bug.cgi?id=2849 Aug 23 20:02:46 likewise: thanks for the tip Aug 23 20:04:36 mtn: misuse: 29 workspace conflicts Aug 23 20:04:47 anyone have any suggestions? Something bad is happening .... Aug 23 20:05:36 Crofton|home: Merge conflicts I think Aug 23 20:07:15 Crofton|home: maybe ask on the monotone channel, dunno Aug 23 20:07:27 grrr Aug 23 20:07:35 * likewise hides Aug 23 20:07:40 time to nuke the work dir and start over .... Aug 23 20:07:52 days like this I heart svn :) Aug 23 20:07:53 Crofton|home: copy the work dir first? Aug 23 20:08:06 likewise, good idea Aug 23 20:08:10 Crofton|home: then create a clean checkout and diff them? Aug 23 20:08:29 That's why I did when I had something similar months ago Aug 23 20:08:35 likewise, thanks for being the voise of reason Aug 23 20:08:35 re Aug 23 20:09:12 Crofton|home: glad it made sense to someone :-) Aug 23 20:10:57 * * OE Bug 2850 has been created by autobuild(AT)openembedded.org Aug 23 20:11:00 * * webkit-gtk-0.0+svnr25203-r25195-r1-do_compile Aug 23 20:11:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2850 Aug 23 20:11:12 Anyone here building on NFS? Aug 23 20:12:16 flo_lap: sometimes yes Aug 23 20:12:46 likewise: Did you ever notice strange effects? Aug 23 20:13:29 flo_lap: using NFS? what do you mean? I use my target as a NFS client to my development (and NFS) host Aug 23 20:13:33 ok Aug 23 20:13:38 now more editing Aug 23 20:13:58 then reapply the patch from khem and fix teh def_pref lines and push Aug 23 20:14:08 likewise: using NFS on your build host. Aug 23 20:14:10 what does this 'psyco' JIT compiler help in terms of speed (now that my build server is 32-bit) Aug 23 20:14:48 flo_lap: pls explain. "Using NFS" is reasonably ambiguous. Aug 23 20:16:07 likewise: hmmm true... having $TMPDIR on NFS would be more precise. Aug 23 20:17:15 flo_lap: no, didn't do that ever. Aug 23 20:20:03 flo_lap, do you build uclibc system? Aug 23 20:20:16 (that is regarding fbreader thing) Aug 23 20:21:06 polyonymous_: no Aug 23 20:21:16 flo_lap, hmm.. and you needed -liconv ? Aug 23 20:21:19 polyonymous_: fixed this already Aug 23 20:21:44 flo_lap, you fixed dependency - that is it won't screw up people's staging, but it won't build with -liconv either. Aug 23 20:22:33 ph5: ping Aug 23 20:22:34 polyonymous_: yes, right - otherwise it stopped with some unresolved symbols Aug 23 20:23:06 flo_lap, weird. but glibc-based build should have no libiconv at all (and it doesn't for me) Aug 23 20:23:35 polyonymous_: were you able to build fbreader? Aug 23 20:23:41 ph5: if you read this, pls check my comment on the OE ML on your patch. Aug 23 20:23:50 flo_lap, after removing -liconv from patch and .bb :) Aug 23 20:26:50 polyonymous_: hrm... strange, that happened when I built Angstrom for my TX Aug 23 20:27:14 flo_lap, I don't know much about TX builds... Perhaps I should try building for TX... Aug 23 20:27:28 I'm now rebuilding just in case it will build as is for me... Aug 23 20:27:43 Actually, I don't mind broken fbreader, but only as long as it doesn't get in the way ;-) Aug 23 20:28:12 likewise: that's why I added a second block that still includes the wildcard (after the dot), hoping though that there will always only be one libgcc_s.so.x.y in the cross-dir ... Aug 23 20:29:31 polyonymous_: well... the TX should be like any PXA based device Aug 23 20:29:44 pH5: http://bugs.openembedded.org/show_bug.cgi?id=2840 Aug 23 20:29:45 ph5: ok, check! Aug 23 20:29:47 flo_lap, I'm build for spitz... Aug 23 20:29:54 ph5: my bad, looks fine! Aug 23 20:31:58 * * OE Bug 2849 has been RESOLVED (FIXED) by Aug 23 20:32:00 * *  pulseaudio-0.9.6-r5-do_compile Aug 23 20:32:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2849 Aug 23 20:32:57 * * OE Bug 2848 has been RESOLVED (FIXED) by Aug 23 20:32:59 * *  gnuradio-3.0.4+svnr6169-r2-do_compile Aug 23 20:33:01 * * http://bugs.openembedded.org/show_bug.cgi?id=2848 Aug 23 20:33:09 * * OE Bug 2850 has been RESOLVED (FIXED) by Aug 23 20:33:11 * *  webkit-gtk-0.0+svnr25203-r25195-r1-do_compile Aug 23 20:33:13 * * http://bugs.openembedded.org/show_bug.cgi?id=2850 Aug 23 20:33:27 koen, what was wrong with gnu radio? Aug 23 20:34:03 likewise: the mtn diff -r `mtn automate get_bas... is in on the wiki on the monotonephrasebook . I guess that's a good place Aug 23 20:37:00 keesj: yes, that's where I put it Aug 23 20:37:15 gtk+ is not building atm .... Aug 23 20:38:41 here it is libxcalibrate Aug 23 20:44:58 * * OE Bug 2851 has been created by autobuild(AT)openembedded.org Aug 23 20:44:59 * * mono-1.2.4+1.2.5pre5-r3-do_patch Aug 23 20:45:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2851 Aug 23 20:45:59 flo_lap, fails fine, unable to find -liconv, of course. I think you have your staging screwed up. Aug 23 20:46:53 polyonymous_: maybe... I wonder how that happened, that was a clean rebuild Aug 23 20:47:37 flo_lap, well, if that was a clean rebuild with libiconv non-virtual dependency - that might explain it. if not, then maybe there's another package like that... Aug 23 20:47:49 I think bitbake lacks blocker dependencies.... Aug 23 20:48:14 but that would take packaged-staging first... Aug 23 20:48:50 anyway, flo_lap, if you have libiconv in your staging libdir - it's broken... Aug 23 20:49:16 polyonymous_: ah no, i added the dependency after the link failure of course Aug 23 20:49:49 flo_lap, do you by chance still have original failure? Maybe in temp/log.do_compile.* ? Aug 23 20:49:56 Is PACKAGE_ARCH supposed to end up in the GNU machine tuple? Aug 23 20:50:45 polyonymous_: no, I cleaned it again but I'll try to reproduce Aug 23 20:51:39 flo_lap, well, I hope you're at it, because for what we have no my fix would be simply undoing your changes, but I guess that wouldn't exactly fix it for you. Aug 23 20:51:55 s/no/now/ Aug 23 20:52:48 Crofton|home: nothing, I control-c'ed and forgot sepukuu was running Aug 23 20:53:18 heh Aug 23 20:53:24 gtk+ is not building Aug 23 20:54:12 gtk does build for me, otherwise sepukku would bitch about it :) Aug 23 20:54:36 this is annoying Aug 23 20:55:12 'night all Aug 23 20:55:16 gn Aug 23 21:01:11 03koen 07org.oe.dev * rf3f91ac3... 10/ (3 files in 2 dirs): mono: move patchdir, should fix #2851 Aug 23 21:05:32 koen: ouch.. Aug 23 21:06:09 lumidum, Aug 23 21:06:10 re Aug 23 21:06:16 hi Aug 23 21:08:49 I m looking for an initramfs-image.bb Aug 23 21:10:13 03pH5 07org.oe.dev * r733d4e48... 10/ (1 packages/gsm/files/default packages/gsm/libgsmd_svn.bb): libgsmd: fix /etc/default/gsmd script Aug 23 21:10:31 babel look initrd it makes initramfs too Aug 23 21:10:34 +at Aug 23 21:14:01 thanks Aug 23 21:27:57 * * OE Bug 2852 has been created by autobuild(AT)openembedded.org Aug 23 21:27:59 * * mono-1.2.4+1.2.5pre5-r3-do_compile Aug 23 21:28:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2852 Aug 23 21:29:20 koen: for some reason gsmd still doesn't depend on initscripts, although RDEPENDS_gsmd is set. I guess it is overwritten because gsmd is in PACKAGES_DYNAMIC? Aug 23 21:31:36 ph5 -> < koen> 'night all Aug 23 21:31:56 22:55 Aug 23 21:32:16 woglinde: oh, thanks Aug 23 21:39:15 so good night it is. Aug 23 21:39:21 nite ph5 Aug 23 21:43:07 have a agood night all! bye. Aug 23 21:52:49 good nite Aug 23 22:03:38 * cbrake learns the right way to do pre-release versioning ... Aug 23 22:03:47 is any of you guys going to the embedded linux conference in Linz, .at? Aug 23 22:04:18 yes me Aug 23 22:04:24 CELF, you mean .. Aug 23 22:04:33 yes Aug 23 22:04:40 i submitted a proposal for a lecture, based on what i'll be doing at shop.subotron.com .. Aug 23 22:05:06 I'll go there too I guess Aug 23 22:05:17 only 2 hours from Vienna or something Aug 23 22:05:44 do you konw if anyone else is going? actually, would be a good place to present OE too Aug 23 22:06:18 i'm in vienna too .. Aug 23 22:06:35 yeah, i was thiinkig aboutu including a bit of details about openmoko .. would make a lot of sense. Aug 23 22:07:44 oh, you live in .at? Aug 23 22:08:29 yes i live in vienna Aug 23 22:09:12 me 2 heh, sers ;) Aug 23 22:09:33 03rpurdie * r948 10bitbake/lib/bb/providers.py: providers.py: Adjust logic so providers with a PREFERRED_VERSION set take priority over providers without a PREFERRED_VERSION available Aug 23 22:09:52 03rpurdie * r949 10bitbake/lib/bb/runqueue.py: runqueue.py: Fix debugging function Aug 23 22:11:29 03rpurdie * r950 10bitbake/lib/bb/ (. fetch/ parse/ parse/parse_py/ ui/): ignore pyo files Aug 23 22:13:45 03rpurdie * r951 10bitbake/lib/bb/providers.py: ptoviders.py: Remove debugging Aug 23 22:26:44 rp still awake? Aug 23 22:50:44 nite Aug 23 22:52:53 03rpurdie * r952 10bitbake/lib/bb/providers.py: providers.py: Sort providers by default preference as well as priority before processing Aug 24 00:07:49 Crofton: gtk+ is failing for me to, on CentOS 5 (RHEL5) Aug 24 00:19:09 fc6 Aug 24 00:19:12 grrr Aug 24 00:19:25 it is running on koen's autobuilds .... Aug 24 00:21:54 hopefully it gets fixed in the EU AM :) Aug 24 00:22:34 ad hopefully I do not wake up at 0430 again ... Aug 24 00:23:42 Crofton|home, my ossie build failed earlier in the day and then I called it quits, it's trying to get some ossie tarball? Aug 24 00:23:43 http://pastebin.com/m6ac7386 Aug 24 00:28:12 Crofton: it builds fine for me in Debian Etch. Aug 24 00:28:46 And I assume Koen's running some Debian derivative rather than some RedHat derivative, so I doubt it will be magically fixed for us. Aug 24 00:31:16 Crofton|home, I just tried getting the files manually but the links here aren't working on http://ossie.wireless.vt.edu/trac/wiki/Downloads Aug 24 00:31:28 although I am not 100% sure that's what I need anyway Aug 24 00:39:15 my angstrom-console-image build is dying in gtk+ Aug 24 00:39:19 anyone seeing this issue Aug 24 01:00:58 * * OE Bug 2853 has been created by autobuild(AT)openembedded.org Aug 24 01:01:00 * * linux-2.6.21-r7-do_unpack Aug 24 01:01:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2853 Aug 24 01:06:22 asmola, it is trying to grab from svn, have you set up an acount on the server? Aug 24 01:07:00 that's what it eventually does after it fails trying to get a tarball from angstrom-distribution.org Aug 24 01:07:12 no i haven't setup an acct on the svn Aug 24 01:07:23 it does that automagically, the source only exists in svn' Aug 24 01:07:34 ossie.wireless.vt.edu/trac Aug 24 01:07:38 and create an account Aug 24 01:07:40 ahhhh, i c Aug 24 01:07:50 then swear you get the GPL/LGPL Aug 24 01:07:57 yes teh uni lawyer is an idiot Aug 24 01:11:03 ok, im registered Aug 24 01:11:41 ntsux123 Aug 24 01:11:47 doh Aug 24 01:12:14 rofl Aug 24 01:12:41 ok Aug 24 01:12:44 try again Aug 24 01:12:52 asmola what w you used Aug 24 01:12:56 yeah Aug 24 01:13:19 ok, im able to browse the source from the web Aug 24 01:14:23 should i just manually pull down the files and dump them somewhere or is there a place i can enter my username/passwd so that the script will enter that in when it does the svn request? Aug 24 01:17:45 bitbake ossie-image, you'll need to enter the username pw once Ithink Aug 24 01:17:47 or Aug 24 01:18:00 check out something and it will save teh info Aug 24 01:18:05 http://www.balister.org/~balister/otr_a_while_ago.png Aug 24 01:18:13 ahhh, i see Aug 24 01:18:14 and for something completely different :) Aug 24 01:32:58 what is that pic of? Aug 24 01:37:41 finally got it to work Aug 24 01:37:42 whew Aug 24 01:45:34 yay Aug 24 01:45:37 hwh Aug 24 01:45:46 I will be there over labourt day Aug 24 01:46:02 do not expect any qulaity support Thursday through Monday Aug 24 01:46:21 Although they claim tehy will have wireless, so there may be low quality discussion Aug 24 01:46:58 I need to get ossie-image on my OSK so I can help you do something Aug 24 01:49:23 asmola, the current ossie-image does not include the stand alone demo waveform Aug 24 01:49:29 I will need to add that tomorrow Aug 24 01:49:48 I need to do some of this for my sdr forum paper :) Aug 24 02:02:17 seems the busybox patch for gcc4 was lost during upgrading qemu 0.8.2 to qemu 0.9.0 Aug 24 02:03:45 are you presenting in denver? Aug 24 02:03:51 yeah Aug 24 02:03:58 I should be Aug 24 02:04:04 we all want to go ski :) Aug 24 02:04:19 :) Aug 24 02:05:39 i am thinking of writing up a procedure on how to get ossie on gumstix, i don't know where i will post it, but i will find somewhere Aug 24 02:05:52 we joke about the SDR forum accepting 130% of the papers submitted Aug 24 02:06:05 lol Aug 24 02:06:09 we have a couple of wikis Aug 24 02:06:21 the audience pays attention though Aug 24 02:06:33 this is why I like the SDR Forum tech conf Aug 24 02:07:16 I need to clean up the image some Aug 24 02:07:27 so it does something useful without access to a USRP Aug 24 02:07:47 We need a better demo that runs without hardware Aug 24 02:08:27 that would be slick Aug 24 02:08:35 yeah Aug 24 02:08:46 we have a sample now, but it is verging on silly :) Aug 24 02:09:04 I also need to work out the gumstix/angstrom gory details Aug 24 02:09:09 I think fstab needs work Aug 24 02:09:21 Hopefully, Steve can get me some gumstix Aug 24 02:09:47 yeah, i menteiond it to him today in one of my emails, he's working on it, he met with one of the gumstix guys today i guess Aug 24 02:10:07 if it doesn't pan out, i'll see what i can do from here Aug 24 02:10:29 yeah, he is on a boondoggle Aug 24 02:10:56 wearing my OE hat, we are always glad for hw so we can imporve support for vendro hw Aug 24 02:11:12 I am trying to Linux going on the Lyrtech SFFSDR board Aug 24 02:11:19 that will rock Aug 24 02:11:51 that would be pretty neat Aug 24 02:11:57 * * OE Bug 2157 has been REOPENED by xjqian(AT)gmail.com Aug 24 02:12:00 * *  qemu fails to compile with gcc 4.x Aug 24 02:12:02 * * http://bugs.openembedded.org/show_bug.cgi?id=2157 Aug 24 02:12:10 i'm completely new to sdr, just started at sca mid june Aug 24 02:12:10 ~lart qemu Aug 24 02:12:11 * ibot whacks qemu upside the head Aug 24 02:12:31 sdr is cool for EE's that like software Aug 24 02:12:42 i'm a comp sci that likes hardware :( Aug 24 02:12:45 :)* Aug 24 02:12:52 the really hard part is getting people that have some comms and some software dev knowledge Aug 24 02:13:16 comms guys that think they can write sw are the worst :) Aug 24 02:13:30 rofl Aug 24 02:13:51 koen is in a comms group Aug 24 02:14:05 I had been talking to him for like a year b4 I found that out :) Aug 24 02:14:12 haha Aug 24 02:14:56 craig (the gumstix guy) also mentioned moving over to oe on the gumstix list, and steve said he was seriousa bout it (as per their meeting) so that's good Aug 24 02:15:30 yeah Aug 24 02:15:56 buildroot is more traigth forward, but the OE guys have alot of packages Aug 24 02:16:02 very complete systems Aug 24 02:17:53 the OSK has 11Mb USB right? how did that work with the USRP? Aug 24 02:18:06 not well Aug 24 02:18:53 there is another guy interested in 1.1 atm though Aug 24 02:19:18 I got the USRP to connect, but the FX2/FPGA interface did not handle the small packets well Aug 24 02:20:05 I cheated and put a usb2.0 card in my Efika :) Aug 24 02:20:09 haha Aug 24 02:21:37 as far as the gumstix, i had a few ideas (although i have no idea how viable they are) one would be trying to leverage the 100Mb ethernet and the other is making the FPGA do more work (although its counter to the GNU radio GPP philosophy) Aug 24 02:23:07 there are lots of neat things to try :) Aug 24 02:23:43 the gnu radio guys have a pretty narrow focus Aug 24 02:23:53 they assume everyone loves PC's Aug 24 02:23:58 and hdtv Aug 24 02:24:06 and they perform well within that space Aug 24 02:24:28 http://www.tapr.org/dcc.html Aug 24 02:24:36 This might be a good intro to SDR Aug 24 02:25:12 Bob McGwier is really interesting Aug 24 02:25:28 that looks interesting and it's close Aug 24 02:26:07 yeah Aug 24 02:26:09 and cheap Aug 24 02:26:26 and I they aren't guys that will draw equations on the board :) Aug 24 02:27:04 :) Aug 24 02:27:18 ok, I need to go to bed Aug 24 02:27:24 been waking up ay to early Aug 24 02:27:46 goodnight crofton, thanks again for the help Aug 24 02:28:35 no problem, it is really nice to have people to work with! Aug 24 02:28:38 l8r **** ENDING LOGGING AT Fri Aug 24 02:59:56 2007