**** BEGIN LOGGING AT Sat Jun 07 02:59:56 2008 Jun 07 03:33:48 03xjqian 07org.oe.dev * r2be84240... 10/ (1 packages/linux/linux-rp-2.6.24/defconfig-tosa): Jun 07 03:33:48 linux-rp-2.6.24/defconfig-tosa: fix broken config Jun 07 03:33:48 * fix unbootable: +CONFIG_JFFS2_LZO=y, +CONFIG_LZO_COMPRESS=y, +CONFIG_LZO_DECOMPRESS=y Jun 07 03:33:48 * fit size limit: -CONFIG_OABI_COMPAT=y, -CONFIG_BLK_DEV_IDECS=y Jun 07 03:33:49 * unwind builtin crypto API Jun 07 03:33:51 * not bump PR, as tosa patch conflit present in linux-rp_2.6.24.bb. Jun 07 05:04:04 03xjqian 07org.oe.stable * r3598f614... 10/ (7 files in 3 dirs): linux-rp-2.6.24/defconfig: revert built-in crypto api modules Jun 07 07:06:07 * * OE Bug 4108 has been marked as DUPLICATE of bug 4007 by xjqian(AT)gmail.com Jun 07 07:06:09 * * wlan-ng-modules-0.2.7-autobuild Jun 07 07:06:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4108 Jun 07 08:33:02 morning all Jun 07 09:02:53 morning Jun 07 09:03:12 morning woglinde Jun 07 09:13:52 hi slapin Jun 07 09:14:17 woglinde, hi! Jun 07 09:14:25 hi, all! Jun 07 09:16:57 03xjqian 07org.oe.dev * r0d2fc1d2... 10/ (14 files in 5 dirs): Jun 07 09:16:57 wlan-ng: tag 0.2.8 srcrev 1859. use as default for 2.6.24 kernel Jun 07 09:16:57 * update only-the-utils.patch for wlan-ng-utils_svn Jun 07 09:33:35 03woglinde2 07org.oe.dev * rb67982af... 10/ (5 files in 3 dirs): Jun 07 09:33:35 sysfsutils: update to latest version 2.1.0 Jun 07 09:33:35 * add mnt patch from debian Jun 07 09:33:36 * reorganize packages Jun 07 09:42:29 hey zecke! Jun 07 09:42:47 I'd like to help to make qtopia4 working in OM / oe. What is the best I can do to help? Jun 07 09:48:20 cyrilRomain: eeek Jun 07 09:48:56 cyrilRomain: you are insane. With the current Qtopia buildsystem I veto any attempt to put that into OE. Jun 07 09:49:23 cyrilRomain: it has to use system uic/moc/qmake, it has to use the cross built Qt version... and this is a major effort Jun 07 09:49:42 zecke: yes it is what I noticed Jun 07 09:50:03 cyrilRomain: you can file bug reports and complain on the qtopia interest/feedback mailinglist Jun 07 09:50:36 zecke: are you working on spliting qtopia-phone into -native and not native ? Jun 07 09:51:43 cyrilRomain: no, I started splitting that stuff while being at trolltech... Jun 07 09:52:38 cyrilRomain: The thing is it is a major effort to fix their buildsystem, and they will not get it... so you have effectively created a Qtopia fork... but why would you want to do that? Jun 07 09:52:56 cyrilRomain: if they want to create software that no distribution is going to distribute it is their decision :) Jun 07 09:54:42 zecke: well I understand your feeling here. And it seems (reading mailing list) that you don't get too much help/answer from qtopia devs :-/ Jun 07 09:55:24 zecke: for now, I'm trying to run gta02 in qemu, but flash.sh need update regarding NAND partitions Jun 07 09:56:24 zecke: I also tried to built the qtopia-x11 image, but get a bug because it looks for libXtst in host system Jun 07 09:56:59 zecke: but you are already aware of such bugs (as you said, uic/moc, etc. and splitting must be done correctly) Jun 07 09:58:12 hi zecke Jun 07 09:58:33 cyrilRomain: the other part of the story, we have a qtopia-phone-x11_git.bb recipe in the Openmoko OE tree, it is building stuff. Due the above reasons the recipe is not mergable though Jun 07 09:58:33 cyrilRomain: To admit, I didn't bother to look into the libxtst one. Is that coming from E or Qtopia? Jun 07 09:58:54 zecke: qtopia Jun 07 09:59:37 zecke: zecke qtopia-phone-x11_git.bb actually Jun 07 09:59:49 cyrilRomain: they don't get linux... there is only so much one can do... Jun 07 10:00:31 cyrilRomain: they want to live in their parallel universe and don't like to be disturbed... I decided to let darwin sort out the rest Jun 07 10:00:49 zecke: :} Jun 07 10:11:52 cyrilRomain: fixing libxtst is probably easy, it might be my bug :) Jun 07 10:12:47 e.g. I would like to build marble for my neo... this means that I will have to install Qt twice... which is just insane Jun 07 10:13:04 what is marble? Jun 07 10:13:18 woglinde: a opensource 'google earth' like Jun 07 10:13:26 ah okay Jun 07 10:13:28 woglinde: developed in the KDE Jun 07 10:13:56 zecke: due to rm_work, I don't have my log anymore :-/ Jun 07 10:14:10 woglinde: sudo apt-get install marble Jun 07 10:14:24 zecke: but a '-L/usr/lib' was here Jun 07 10:14:26 not yet Jun 07 10:14:45 my wife has to the taxes Jun 07 10:14:47 do Jun 07 10:15:01 cyrilRomain: it has openstreetmap support now, and the drawing code seems to be good (I trust torsten on this) Jun 07 10:15:12 zecke: I added -L${STAGING_LIBDIR} but without success because the -L/usr../ is without doubt comming from qtopia wuild system Jun 07 10:15:35 zecke: cool :) Jun 07 10:16:40 zecke: in fact the build was reported in Openmoko trac, but someone closed it with 'i installed libxtst-dev on my host. Fixed!' :-( Jun 07 10:16:49 s/build/bug/ Jun 07 10:17:34 hehe, regarding qtopia that is probably good enough :( Jun 07 10:18:03 http://git.openmoko.org/?p=qtopia.git;a=commitdiff;h=88e10c03f08b5ff761d61dcbbad8203211d79e15 <- I have to fight stuff like this... Jun 07 10:18:27 8-O Jun 07 10:18:33 cyrilRomain: but patches are welcome, I have other bugs to fix in Qtopia before I can look into these kind of build issues were a workaround exists Jun 07 10:18:52 zecke: I understand that Jun 07 10:19:50 zecke *g* remeber 40 developers fixing bugs in qtopia Jun 07 10:20:03 maybee 39 of them are lorn Jun 07 10:21:04 woglinde: 40 devs on bug fixing ? I think this tells many things ... :-/ Jun 07 10:21:27 cyril that is what lorn told us Jun 07 10:21:42 BTW, I was wondering how split are the Qt and Qtopia team Jun 07 10:21:50 I mean not geographically Jun 07 10:21:53 cyril and I dont believe this Jun 07 10:22:10 seems austrial is makeing qtopia Jun 07 10:22:15 asutralia Jun 07 10:22:17 woglinde: I thougth they were not so many working in Australia Jun 07 10:22:18 args Jun 07 10:24:27 03xjqian 07org.oe.stable * rbd4ba2fd... 10/ (3 files in 3 dirs): Jun 07 10:24:27 linux-rp-2.6.24/defconfig-tosa: fix tosa 2.6.24 kernel build and boot problem Jun 07 10:24:27 approved by: koen and xjqian Jun 07 10:25:05 cyrilRomain|bbl: two different continents Jun 07 10:25:07 * * OE Bug 4007 has been RESOLVED (FIXED) by xjqian(AT)gmail.com Jun 07 10:25:09 * * Angstrom 2.6.24 kernel won't boot on tosa. Jun 07 10:25:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4007 Jun 07 10:27:41 morning all Jun 07 10:28:31 xjqian: ping Jun 07 10:28:34 cyrilRomain|bbl: but good comment. They fix the symptones, I look at the root cause. and it is either "stupid people" or "people asked to be stupid and not revolting" Jun 07 10:29:01 morning: ant Jun 07 10:29:10 you never sleep... Jun 07 10:29:30 I think you're in USA Jun 07 10:29:40 I happen to be in the mood of fixing things Jun 07 10:29:44 I bother you about udev problem: no /dev/shm Jun 07 10:30:00 it only appens to tosa and c7x0 apparently... Jun 07 10:30:08 *happens Jun 07 10:30:14 interesting... Jun 07 10:30:26 the initscript are unchanged Jun 07 10:30:40 I have no deep knowledge about udev. Jun 07 10:30:42 but just downgrading to udev 092 solves Jun 07 10:30:44 ant|work: and you have shared mem compiled into the kernel? Jun 07 10:30:50 oh Jun 07 10:30:51 yes Jun 07 10:31:12 the previous udev works is udev 115 Jun 07 10:31:21 but only in Poky Jun 07 10:31:43 udev 115 works for me on Tosa Jun 07 10:32:53 mickey upgraded udev from 115 to 118, which indroduced the breaking point about one month ago Jun 07 10:34:00 !oebug 4118 Jun 07 10:34:00 * * Bug 4118, Status: NEW, Created: 2008-03-25 08:47 Jun 07 10:34:03 * * xjqian(AT)gmail.com: udev-118 failed to create device nodes on Tosa Jun 07 10:34:05 * * http://bugs.openembedded.net/show_bug.cgi?id=4118 Jun 07 10:34:34 re Jun 07 10:36:10 xjqian: I'm not @home, could you check the MAKEDEVS in your rootfs? Jun 07 10:36:15 03koen 07org.oe.dev * rbb0dc928... 10/ (3 files in 3 dirs): xserver-common: move fallback keymap into case statement Jun 07 10:37:14 *if this closes 4366 then thanks koen Jun 07 10:37:40 BTW did you notice gitweb is out of sync? Jun 07 10:38:04 ant: this may take me a while to get back to you. Seems the keymap has some new problem for 2.6.24 tosa kernel Jun 07 10:38:19 urgh Jun 07 10:38:54 I was talking earlier with XorA on #angstrom about this Jun 07 10:39:13 but wasn't kernel, happily Jun 07 10:39:31 just xserver xmodmap Jun 07 10:43:28 ant: sorry for confusing you. never mind my previous statement Jun 07 10:43:33 ant: I'm checking now Jun 07 10:44:56 thx Jun 07 10:49:13 ant: I don't have MAKEDEV in my /sbin Jun 07 10:50:02 but, I do have makedevs in my /sbin Jun 07 10:50:40 can you inspect /etc/volatile.cache ? Jun 07 10:51:05 XorA suggested to remove it (could be bad) Jun 07 10:51:36 s/bad/b0rked/ Jun 07 10:52:32 I found trace of /dev/shm only in udev's init script... Jun 07 10:52:49 (and in /etc/fstab of course) Jun 07 10:57:20 ant: http://pastebin.ca/1041165 is my  /etc/volatile.cache Jun 07 10:58:54 ant: can you summarize what's XorA's analysis Jun 07 10:59:23 ant: I read your conversation but can't understand where should we dig into Jun 07 10:59:32 Im guessing someone special cased c7x0 somewhere, and that is now breaking Jun 07 11:00:06 somewhere = device creation = udev ? Jun 07 11:00:32 as I said it's the only package I revert. Need to check the dependencies, though Jun 07 11:01:14 the patchsets is very different between 092 and 118 Jun 07 11:01:27 ant: I still don't understand Jun 07 11:01:41 morning all Jun 07 11:01:43 what puzzles me is that akita seems working ok Jun 07 11:02:07 RP: g'mornin' Jun 07 11:02:17 ant: you could try 115 as your starting point Jun 07 11:02:39 RP; morning Jun 07 11:03:23 ant: you mean the culprits are in the patches for udev 118? Jun 07 11:03:37 I can only suspect Jun 07 11:04:24 because lot of files comes sunchanged from 115 Jun 07 11:04:35 *unchanged* Jun 07 11:05:19 http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=commit;h=b6e8d9bf9ab2fd78ed0be815cb3309dee2b7250a Jun 07 11:07:57 ant: don't look like patches. It's not likely kernel problem either. as It breaks for tosa in both 2.6.23 and 2.6.24, with different config options. Jun 07 11:08:16 ant: I think it's in the udev source change from 115 to 118 Jun 07 11:08:49 and it would break only for two machines? Jun 07 11:08:52 ??? Jun 07 11:09:30 ant: indeed puzzling Jun 07 11:09:48 03koen 07org.oe.stable * rf8ad3acb... 10/ (1 conf/distro/angstrom-2007.1.conf): angstrom 2007.12: bump to r16 Jun 07 11:10:19 ant|work: thanks for the gitweb hint Jun 07 11:10:33 you're welcome Jun 07 11:11:18 I had my autoconvert script running twice... this wasn't too clever Jun 07 11:11:29 he Jun 07 11:12:59 ant: mickey probably is the right person to bug. Current I am fine with udev 115, and override the default 118 in dev. I will learn udev once I some more time Jun 07 11:13:18 115 is not in .dev anymore... Jun 07 11:13:32 if you put it as preferred_version, you get 092 Jun 07 11:13:40 (angstrom 2008.1) Jun 07 11:13:48 I'll have to add it Jun 07 11:14:34 * ant|work is scared by koen's statement: PV never goes back Jun 07 11:15:31 xjqian: interestingly Gentoo moved rapidly to 119 Jun 07 11:16:59 xjqian: thanks for your time Jun 07 11:17:05 ant: yes, you have to add it from revision history. But I agree with koen, for oe in general we should move forward. Local hacks are only personal. Jun 07 11:17:44 * ant|work -> food(s) Jun 07 11:19:16 ant: nice to know you are also interested in solving this bug. Jun 07 11:35:50 03xjqian 07org.oe.stable * r8e1304ae... 10/ (56 files in 15 dirs): Jun 07 11:35:50 wlan-ng: use 0.2.8 svnr1859 as default, needed by 2.6.24 kernel Jun 07 11:35:50 * clean up of obsolete recipes Jun 07 11:35:50 approved by Koen and xjqian Jun 07 11:48:43 RP; do you have a minute? Jun 07 12:14:07 * * OE Bug 4366 has been RESOLVED (FIXED) by andrea.adami(AT)gmail.com Jun 07 12:14:09 * * broken xmodmap for c7x0 Jun 07 12:14:11 * * http://bugs.openembedded.net/show_bug.cgi?id=4366 Jun 07 12:26:47 hi florian Jun 07 12:27:22 hi all Jun 07 12:37:04 zecke: hm about qtopia-phone-X11: git/qtopiacore/qt/mkspecs/linux-g++ is used to build host qtopiacore, but this file needs s,/usr/X11R6/, Jun 07 12:37:17 or a linux-oe-g++ spec should be added to the host qtopia build dir and specified when building host qtopiacore Jun 07 12:39:13 hm what about linux-gnueabi-g++ and linux-oe-g++ located in the ficgta01 device spec directory ? Jun 07 12:42:39 cyrilRomain: to admit. I have no idea which spec gets used. I'm confident one inside the devices/ficfta01... gets used Jun 07 12:42:52 cyrilRomain: just by having a gcc42 device they show how flawed their device concept is :} Jun 07 12:43:57 zecke: yeah I just noticed that :-/ Jun 07 12:44:30 zecke: you can see which spec is used in the header of the Makefile generated by qmake Jun 07 12:46:02 so at least I found were the wrong host path come from .. Jun 07 12:46:16 s/were/where/ Jun 07 12:46:54 but I'm wondering if linux-gnueabi-g++ and linux-oe-g++ are used at all Jun 07 12:47:14 cyrilRomain: oh, the host build is failing already? Jun 07 12:47:22 cyrilRomain: ah stupid me :) Jun 07 12:47:31 no qws involved that is x11 Jun 07 12:48:16 -spec ../../git/devices/ficgta01/mkspecs/qws/linux-oe-g++ Jun 07 12:48:18 hmm Jun 07 12:48:24 hmm Jun 07 12:48:40 in your qtopiacore/host/src/gui/Makefile ? Jun 07 12:49:02 host == host, so /usr/lib/libxtst would be fine... Jun 07 12:49:38 it is building two versions of Qt... a full desktop version (for no reason at all...) and the target one Jun 07 12:51:37 zecke: should be fine, except when you build from a chrooted environment with no X :} Jun 07 12:53:35 cyrilRomain: right, the lack of a native x11 is something we tolerate even for the gtk+ stuff. the question is what you can do Jun 07 12:53:45 but qtopia core in oe seem compiling correctly, is there so much difference from qtopia and qtopia phone? Jun 07 12:54:02 cyrilRomain: you mostly just need moc/uic/rcc but also the svg libs. Jun 07 12:54:13 zecke: I can install X in my chroot environment, but I tend to dislike workaround Jun 07 12:54:23 cyrilRomain: oh well, qt embedded linux (qtopia core) is not requiring x11 at all Jun 07 12:54:55 cyrilRomain: your only other option is to build the host version as qtopia core/qt embedded linux. Jun 07 12:55:24 zecke: right Jun 07 12:55:25 and then see how far svgtopicture comes/goes... (think of the -qws parameter) Jun 07 12:57:39 cyrilRomain: but why is only xtst failing? Jun 07 12:58:03 zecke: well I may have some other libs .. Jun 07 12:58:20 zecke: it's a chroot with some applications (text mainly) Jun 07 12:58:34 zecke: but probably due to dependencies I got some other libs with Jun 07 12:59:03 zecke: but I could also fix mkspecs/common/linux.conf, as it is fixed in oe Jun 07 13:01:13 ok, anyway in the end in qtopiacore/host/src/gui/Makefile, you have ficgta01/mkspecs/qws/linux-oe-g+ and not git/qtopiacore/qt/mkspecs/linux-g++ ? Jun 07 13:04:51 zecke: hmm and I guess qtopia core cannot be easily built outside from the qtopia-phone build ? .. Jun 07 13:05:15 zecke: i.e. separating qtopia core, from what qtopia phone bring upon qtopia Jun 07 13:08:12 cyrilRomain: Not easily, or I miss the magic environemnt var... but then it would use a source tree... Jun 07 13:08:48 cyrilRomain: anyway qtopia-phone-x11 will use Qt/X11. The disgusting build script will build Qt for X11 for your host (x86) AND Qt for X11 for your target (neo)... Jun 07 13:08:58 cyrilRomain: so it will use at least two different spec files, as you see Jun 07 13:09:09 cyrilRomain: for one the usage of /usr/lib is just fine Jun 07 13:40:23 *sigh* too few people developing with directfb Jun 07 13:43:28 woglinde: how do you mean? Jun 07 13:49:42 googlebot is spidering my build directory Jun 07 13:50:27 markus .pc files are little bit broken Jun 07 15:37:55 Hmm Jun 07 15:38:01 Has everythig switched to git? Jun 07 15:41:38 re Jun 07 15:41:42 nope Jun 07 15:41:50 and the git is out of sync Jun 07 15:41:55 +demo Jun 07 15:44:26 k, thanks Jun 07 15:44:38 did a sync and didn't see any changes since i thought a few days ago Jun 07 15:47:43 anyone here knows the bootsequence of LAB / kernel on an ipaq? Jun 07 15:47:56 I copied my system from CF to an sd card and edited the labrun file Jun 07 15:48:13 but nwo the kernel (not lab's) panics because it doesn't know the fs Jun 07 15:48:18 (can't find I think) Jun 07 15:54:18 never mind... Jun 07 15:58:16 soccer Jun 07 15:59:00 ? Jun 07 15:59:03 not for me. Jun 07 15:59:20 first gm Jun 07 15:59:33 aeh game of europaen championship Jun 07 16:09:26 hello Jun 07 16:09:40 which package provides libdbus-1-3 (>= 1.0.2) Jun 07 16:09:41 ? Jun 07 16:21:16 dbus 1.2.1 doesn't help here. Jun 07 16:21:30 what o what do I need for libdbus-1-3 (>= 1.0.2) ??? Jun 07 16:26:44 03koen 07org.oe.dev * r4dbd9e3e... 10/ (1 packages/libgpewidget/libgpewidget_0.115.bb): libgpewidget: depend in libxinerama. Parallel builds fail due to debian renaming Jun 07 17:18:40 03bluelightning 07org.oe.dev * r56a62771... 10/ (5 files in 5 dirs): libopie/qte: fix cursor key rotation for iPAQs. Should resolve 3678. Jun 07 17:19:25 woohoo my first official commit Jun 07 17:19:29 :) Jun 07 17:20:12 bluelightning, way to go! Jun 07 17:21:04 HopsNBarley: thanks :) Jun 07 17:46:39 welcome to OE, bluelightning! Jun 07 18:36:03 03mickeyl 07org.oe.dev * r82ea7fb3... 10/ (1 packages/images/fso-image-light.bb): add fso-image-light, removing tools and audio (for now) Jun 07 18:49:05 * * OE Bug 1324 has been RESOLVED (FIXED) by Jun 07 18:49:07 * * Cross compile badness in icewm Jun 07 18:49:09 * * http://bugs.openembedded.net/show_bug.cgi?id=1324 Jun 07 18:49:58 03mickeyl 07org.oe.dev * r62edbc24... 10/ (7 files in 4 dirs): icewm 1.2.30 fix build. close #1324 Jun 07 18:55:06 * * OE Bug 1545 has been RESOLVED (FIXED) by Jun 07 18:55:08 * * setserial does not compile when using a prebuilt toolchain Jun 07 18:55:10 * * http://bugs.openembedded.net/show_bug.cgi?id=1545 Jun 07 18:55:52 03mickeyl 07org.oe.dev * r1ed1a74f... 10/ (4 files in 3 dirs): setserial 2.17 add ldflags. closes #1545 Jun 07 19:36:00 Laibsch: made a little progress with formatting: http://wiki.openembedded.net/index.php/Prototype_Main_Page Jun 07 19:36:10 Laibsch: now, for some content ... Jun 07 19:40:04 cbrake: looks quite nice Jun 07 19:40:20 we should add some colour and then that is what we should use Jun 07 19:40:28 thanks, mickey Jun 07 19:44:10 Laibsch: color would be good -- can probably be tweaked easily with Common.css: http://wiki.openembedded.net/index.php/MediaWiki:Common.css Jun 07 20:29:58 hi ph5 Jun 07 20:30:10 hi woglinde Jun 07 21:06:16 hi Jun 07 22:00:36 xjqian: hi Jun 07 22:00:51 ant: hi Jun 07 22:01:13 I built udev-119 with the same 118 patchset: nothing changes Jun 07 22:01:30 now I'm repatching noasmlinkage in udev-115 Jun 07 22:01:51 image built right now, let see ... Jun 07 22:03:51 flashing ... Jun 07 22:06:30 mounting tmpfs on /dev/shm failed Jun 07 22:06:39 :-/ Jun 07 22:07:41 xjqian: now, with 115, the log appears immediately after the (known) failed mount of /proc (resource busy) Jun 07 22:07:53 hm Jun 07 22:08:02 woglinde: hey Jun 07 22:08:08 eventually means /proc is mounted already Jun 07 22:08:38 woglinde: this is the *minor* issue Jun 07 22:08:51 /dev/shm being the big one Jun 07 22:09:08 (no devices...) Jun 07 22:09:40 hm when you create /dev/shm by hand? Jun 07 22:09:51 maybee your kernel has no ipc Jun 07 22:10:57 woglinde: changing udev version brings it back to work Jun 07 22:11:07 I suppose it's not the kernel Jun 07 22:11:29 ant hm okay Jun 07 22:11:30 BTW linux-rp-2.6.24 Jun 07 22:11:39 you could try udev in debug mode Jun 07 22:11:53 yea, I was hoping 115 would work Jun 07 22:12:14 xjqian: could you please list me the patches you sed for udev 115? Jun 07 22:12:24 *used Jun 07 22:12:58 the solution must be here http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=commit;h=b6e8d9bf9ab2fd78ed0be815cb3309dee2b7250a Jun 07 22:13:08 but mickey|zzZZzz .... Jun 07 22:13:25 ant: I used all the existing patches for udev 115 Jun 07 22:13:39 which one, the deleted one?? Jun 07 22:13:47 noasmlinkage ? Jun 07 22:14:21 118 and 115 share all the resting patches Jun 07 22:14:36 or I could be just renamed 118 patch folder to 115 Jun 07 22:14:53 yea, but the recipe? Jun 07 22:14:55 I'm not sure right now. It's a while ago when I tested Jun 07 22:15:32 I will test with and without tonight noasmlinkage Jun 07 22:15:44 SRC_URI = "\ Jun 07 22:15:44 http://kernel.org/pub/linux/utils/kernel/hotplug/udev-${PV}.tar.gz \ Jun 07 22:15:44 file://noasmlinkage.patch;patch=1 \ Jun 07 22:15:44 file://flags.patch;patch=1 \ Jun 07 22:15:44 file://vol_id_ld.patch;patch=1 \ Jun 07 22:15:45 file://udevtrigger_add_devname_filtering.patch;patch=1 \ Jun 07 22:15:47 file://mount.blacklist \ Jun 07 22:15:49 " Jun 07 22:16:13 without noasmlinkage it fails later, not immediatly after proc Jun 07 22:16:47 (with sed instead of noasmlinkage) Jun 07 22:19:03 woglinde: the "nice" thing is it fails only on two zaurus models (c7x0 and tosa). Others don't complain about udev-118... Jun 07 22:19:50 is the repository down? Jun 07 22:21:27 http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=commitdiff;h=55df9cf2cd11aed3b1b9333f72d518d1763b180f;hp=95c2745634c0164dbd556d068219057f27e8131e Jun 07 22:21:44 ^^^ -udevd_timeout=30 Jun 07 22:21:55 and with sed it works? Jun 07 22:22:07 nope, fails some lines later Jun 07 22:22:25 but this timeout... Jun 07 22:22:57 and this - # wait for the udevd childs to finish Jun 07 22:23:06 - sleep 1 Jun 07 22:23:21 all waitstates removed...doh Jun 07 22:23:41 ant which ine works? Jun 07 22:23:48 udev_092 Jun 07 22:24:03 poky uses 115 Jun 07 22:24:13 (for c7x0 too) Jun 07 22:24:13 all kernels are 2.6? Jun 07 22:24:22 and have the same config? Jun 07 22:24:25 yes, 2.6.2[34] Jun 07 22:25:08 # Pseudo filesystems Jun 07 22:25:08 # Jun 07 22:25:08 CONFIG_PROC_FS=y Jun 07 22:25:08 CONFIG_PROC_SYSCTL=y Jun 07 22:25:08 CONFIG_SYSFS=y Jun 07 22:25:10 CONFIG_TMPFS=y Jun 07 22:25:12 # CONFIG_TMPFS_POSIX_ACL is not set Jun 07 22:25:14 # CONFIG_HUGETLB_PAGE is not set Jun 07 22:25:16 # CONFIG_CONFIGFS_FS is not set Jun 07 22:25:18 Jun 07 22:25:45 hm that are not the options for shared memory Jun 07 22:27:49 * ant_ looking for others? Jun 07 22:28:25 CONFIG_SHMEM=y Jun 07 22:28:37 yes Jun 07 22:28:40 it is Jun 07 22:30:05 * * OE Bug 1609 has been RESOLVED (FIXED) by Jun 07 22:30:07 * * Fixes for text based bluetooth applications Jun 07 22:30:09 * * http://bugs.openembedded.net/show_bug.cgi?id=1609 Jun 07 22:30:16 can you manually mount /dev/shm? Jun 07 22:30:54 03mickeyl 07org.oe.dev * ra62f6316... 10/ (1 packages/bluesnarfer/bluesnarfer.bb): bluesnarfer: fix URI and DEPENDS. partly closes #1609 Jun 07 22:31:00 03mickeyl 07org.oe.dev * rd17dcf16... 10/ (5 files in 4 dirs): add handsfree 0.1 courtesy Joaquim Duran. closes #1609 Jun 07 22:31:18 mount: mounting shm on /dev/shm failed: No such file or directory Jun 07 22:31:48 but /dev/shm is avaibale? Jun 07 22:31:55 woglinde: if I mkdir /dev/shm then I can mount it Jun 07 22:31:55 and what does dmesg says? Jun 07 22:32:05 mkdir? Jun 07 22:32:10 and I see it in mtab Jun 07 22:32:26 hm okay yes Jun 07 22:32:29 yes, mkdir (found on Google tonite) Jun 07 22:32:43 strange, gell Jun 07 22:32:48 sorry I tought it were only a device Jun 07 22:32:50 but okay Jun 07 22:33:13 I would say do it manually before starting udev Jun 07 22:33:24 but wtf is not remounted by /init.d/udev ? Jun 07 22:33:47 dont know Jun 07 22:33:58 as I said try it in debug mode Jun 07 22:34:08 now I will go to bed Jun 07 22:34:27 thx for your help Jun 07 22:34:38 udev I meant Jun 07 22:34:45 k Jun 07 22:34:47 udev should be replaced Jun 07 22:34:53 with a c daemon Jun 07 22:35:04 shell script is sucking for this task Jun 07 23:03:29 03mickeyl 07org.oe.dev * r0f34f029... 10/ (7 files in 3 dirs): dropbear 0.50 update testing version (DEFAULT_PREFERENCE=-1) to 0.51. Consider commenting on bug #2506 Jun 07 23:05:20 xjqian|away: partial success with the "init" of 115 Jun 07 23:12:49 now I got a msg: udevd[270]add_matching_files: unable to open '/dev/.udev/rules.d': No such file or directory Jun 07 23:17:34 ah, I see why.. it's "/etc/udev/rules.d" Jun 07 23:17:51 in rootfs Jun 07 23:50:22 xjqian|away: ok, I mixed too much.. Jun 07 23:50:37 xjqian|away: the trick must be in http://gitweb.openembedded.net/?p=org.openembedded.dev.git;a=commit;h=55df9cf2cd11aed3b1b9333f72d518d1763b180f Jun 07 23:51:17 xjqian|away: the culprit seems to be /udev-118/init Jun 07 23:52:11 this - # if this directory is not present /dev will not be updated by udev Jun 07 23:52:11 - mkdir -p /dev/.udev/ /dev/.udev/db/ /dev/.udev/queue/ /dev/.udevdb/ Jun 07 23:52:29 or this - # wait for the udevd childs to finish Jun 07 23:52:29 - echo "Waiting for /dev to be fully populated" Jun 07 23:52:53 my 2 (euro)cents Jun 07 23:53:10 now is too late Jun 07 23:53:14 good night Jun 08 00:01:07 I need to see if my udev stuff is still in there or if it got lost... Jun 08 01:27:06 Is it possible to figure out why OE is trying to build qemu when building gpe-image? Looking through the distro conf files and image conf files...not seeing where it's pulling the list from Jun 08 01:35:50 df00z: almost certainly because it's building the locales with qemu Jun 08 01:39:57 mwester: any idea if qemu will build with gcc 4 with the org.openembedded.dev branch? Jun 08 01:44:04 I've always install gcc3 on my hosts, so I dunno if it still needs it or not. But won't work on Fedora anyway... Jun 08 01:48:04 ah..hm **** ENDING LOGGING AT Sun Jun 08 02:59:57 2008