**** BEGIN LOGGING AT Mon Sep 28 02:59:57 2009 Sep 28 03:35:14 03Angus Ainslie  07shr/import * r0c7765d3eb 10openembedded.git/recipes/bt-gps/bt-gps.bb: bt-gps: update the dependencies Sep 28 03:35:24 03Angus Ainslie  07shr/import * rc0349344fd 10openembedded.git/recipes/tasks/task-shr-feed.bb: task-shr-feed: add pyring , bt-gps and bt-configure Sep 28 05:22:37 03Klaus Kurzmann  07shr/import * r7052ad67c5 10openembedded.git/ (2 files in 2 dirs): Sep 28 05:22:37 libgphoto2: bump to 2.4.1 Sep 28 05:22:37 Signed-off-by: Klaus Kurzmann Sep 28 05:22:38 03Klaus Kurzmann  07shr/import * r742d19e0a4 10openembedded.git/: Merge branch 'shr/import' of ssh+git://git@git.openembedded.net/openembedded into shr/import Sep 28 05:22:40 03Martin Jansa  07shr/import * re461787da0 10openembedded.git/recipes/shr/ (3 files): Sep 28 05:22:43 GTK themes: remove link without waiting for enter in postinst. Sep 28 05:22:45 Signed-off-by: Klaus Kurzmann Sep 28 05:22:49 03Klaus Kurzmann  07shr/import * r20f064c791 10openembedded.git/conf/distro/include/preferred-shr-versions.inc: Sep 28 05:22:52 gphoto2: bump preferred version to 2.4.2 Sep 28 05:22:56 Signed-off-by: Klaus Kurzmann Sep 28 05:25:59 hello - I'm having problems with rxvt-unicode's do_install task - it tries to us the local /usr/X11R6/bin/tic and gets a "permission denied" error - anyone know of work around (besides 'bitbake -k'?) Sep 28 07:59:49 morning all Sep 28 08:09:41 gm Sep 28 08:11:36 good morning recalcati ,gremlin[it] Sep 28 08:11:57 hi recalcati, rkirti Sep 28 08:14:16 Mooring.. I got a question. Today my setup looks like this. I got a laptop from which I build my oe images. Then I copy the files onto a external IDE drive. Then I move the drive (fysically) over to a diffrent machine and boot it up. This is quite time consuming and wonder if you guys have any ideas of an more effective way of doing this.. I've been thinking of doing something with vmware but havn't really figured out a good solution. Sep 28 08:44:38 good morning Sep 28 08:45:29 morning all Sep 28 08:45:40 hi thesing Sep 28 08:55:24 good morning Sep 28 08:59:32 hi dth Sep 28 09:03:26 03Florian Boor  07org.openembedded.dev * r514979f141 10openembedded.git/recipes/xserver-common/xserver-common_1.32.bb: xserver-common: Add 1.32 - still with DEFAULT_PREFERENCE -1 but expexted to work. Sep 28 09:07:59 03Howard Chu  07org.openembedded.dev * re5d715b372 10openembedded.git/recipes/python/python-wpactrl_svn.bb: Sep 28 09:07:59 python-wpactrl: add 1.0.1 Sep 28 09:07:59 Signed-off-by: Howard Chu Sep 28 09:07:59 Signed-off-by: Koen Kooi Sep 28 09:11:13 03Graeme Gregory  07org.openembedded.dev * r9c38c20814 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git@git.openembedded.org/openembedded into org.openembedded.dev Sep 28 09:11:14 03Graeme Gregory  07org.openembedded.dev * rcdc32dd34c 10openembedded.git/ (conf/checksums.ini recipes/transmission/transmission_1.75.bb): transmission_1.75.bb : new version of transmission torrent client Sep 28 09:54:38 guten morgen Sep 28 09:55:26 hello Martin-B Sep 28 11:18:59 03Klaus Kurzmann  07shr/import * r47adaadea4 10openembedded.git/recipes/tasks/task-x11.bb: Sep 28 11:18:59 task-x11: bump PR to make it rebuild after XSERVER change to xorg Sep 28 11:18:59 Signed-off-by: Klaus Kurzmann Sep 28 12:24:23 Please HELP - I have this issue: http://pastebin.ca/1582464 Sep 28 12:25:36 Hasse_, have you take a look at the log.do_compile file ? Sep 28 12:31:37 gremlin[it]: Hi :) Yep, it's damn long though... Sep 28 12:38:08 gremlin[it]: The first few lines with "error" in it starts here: http://pastebin.ca/1582480 Sep 28 12:44:09 Hasse_, not sure, but seem there isn't a kernel configured at the time madwifi is compiled ... Sep 28 12:44:37 Hasse_, based on this error "asm/types.h: No such file or directory" ... Sep 28 12:45:52 gremlin[it]: hm hm.. Do you know how to check or fix this? Sep 28 12:47:47 Hasse_, you should find asm/type.h in the staging dir Sep 28 12:48:22 Hasse_, just to check try to force to build the kernel with 'bitbake virtual/kernel' Sep 28 13:29:18 gremlin[it]: I have something called types.h, but not type.h Sep 28 13:31:36 gremlin[it]: I guess this is the directory, we are talking about: http://pastebin.ca/1582541 Sep 28 13:36:26 Can anyone point out some source where I can get info about the initramfs-image? Is there any documentation about it? Sep 28 13:37:15 Hi everyone Sep 28 13:38:16 Hasse_, types.h ... Sep 28 13:38:30 Would you mind helping me? Sep 28 13:38:53 I'm trying to bitbake my gumstix-oe (ever heard of it?) Sep 28 13:39:21 but the gmp-native_4.2.1 fails Sep 28 13:39:38 guest_: some super upto date distro? Sep 28 13:40:01 well, its in the org.openembedded.snapshot... Sep 28 13:40:39 but according to the official tutorial this is supposed to be so... Sep 28 13:41:07 guest_: I mean we used to have gmp-native problems when fedora went to version 10 and ubuntu went to 9.04 Sep 28 13:41:18 guest_: maybe gumstix fork is older than that point Sep 28 13:41:22 im on debian 5 here^ Sep 28 13:41:36 * XorA has no idea what version debian 5 is Sep 28 13:41:56 current stable Sep 28 13:42:04 but that might suffer the same issue Sep 28 13:42:14 give mainline OE a spin Sep 28 13:42:23 guest_, gumstix oe for overo, or older products? Sep 28 13:42:34 it's for verdex pro Sep 28 13:42:35 (ie lenny) Sep 28 13:42:46 * XorA lost track of debian versions after about Mr Potoato Head Sep 28 13:42:59 guest_, unfortunately that is pretty old :( Sep 28 13:43:15 i'm doing this for research at the ETH in zurich, so i'm kinda bound to it^^ Sep 28 13:43:31 but is there a way to maybe update it? Sep 28 13:43:33 ask on the gumstix list Sep 28 13:43:40 or try building from oe.dev Sep 28 13:43:41 they sent me to you guys... Sep 28 13:43:43 luckilly debian still has gcc 4,2 and 4,1 try installing them Sep 28 13:43:46 hmm Sep 28 13:43:57 and making them default Sep 28 13:44:14 "making them default"? Sep 28 13:44:22 * XorA doesnt know how to make then default broonie??? Sep 28 13:45:15 You see, i'm from the windows world, isn't there a version for idiots, like "copy those files over the old ones and run again"? Sep 28 13:45:19 holy crap Mr Potato Head was a long time ago Sep 28 13:45:28 guest_, no :) Sep 28 13:45:36 damn ;( Sep 28 13:46:12 well, any other idea, for example how i could track the bug? Sep 28 13:46:18 so I could at least fix it... Sep 28 13:46:33 gumstic verdex stuff is an old fork Sep 28 13:46:53 guest_: look at gmp-native reciope in stable/2009 and merge it with what you have Sep 28 13:47:00 they have done a much better job with the overo stuff Sep 28 13:47:10 yeah, i'm not happy about it either... 16MB Disk space my ass Sep 28 13:47:32 @xora, would you lend me a hand with a link? Sep 28 13:47:59 guest_: www.openembedded.org follow getting started until you have fetched the git repo Sep 28 13:48:17 * XorA has no gumstix so never looked at the fork Sep 28 13:48:34 i'm getting paid by the hour, so no worries^^ Sep 28 13:49:46 * jovox copies that and sends it to guest__s employer Sep 28 13:50:07 they were trying 2 months and failed Sep 28 13:50:18 and those guys are postdocs^^ Sep 28 13:50:47 gremlin[it]: ...does that sound odd? Or am I looking in the wrong place? Sep 28 13:51:22 hhahaha postdocs are the worst source of technical knowhow Sep 28 13:51:25 Having degrees from universities doesn't necessarily translate into the ability to do practical things, guest_! Sep 28 13:52:36 I already figured that out Sep 28 13:53:10 owfs <--- anyone actively using that package? I'm just trying to get a feel for impact of any changes I might wish to make to the recipe... Sep 28 13:53:12 But as long as I also dont get it to work, well Sep 28 13:54:36 guest_, try using oe.dev Sep 28 13:54:50 hah who summoned sakoman :-D Sep 28 13:54:51 it should work for verdex, and if you have problems, we can answer questions Sep 28 13:55:15 or get gumstix to pay sakoman to get overo-oe working for verdex :) Sep 28 13:57:51 well, I installed GIT, but when I execute in the shell it still says GIT: Command not found? Sep 28 13:58:18 er, linux is case-sensitive... Sep 28 13:59:08 yes, I actually typed git Sep 28 13:59:22 just as said in the "getting started" Sep 28 14:00:22 install git-core on debian Sep 28 14:00:33 git is a filemananger from GNU :-) Sep 28 14:04:04 sorry, git-core installs well, but simply typing "git-core" in the shell after the install returns "command not found" Sep 28 14:07:24 the command is git, the package name is git-core. Sep 28 14:07:26 * kergoth yawns Sep 28 14:08:04 noone's using oe's initramfs image? Sep 28 14:17:52 @kergoth: THANKS that did it. now downloading the 300k objects, so, what to do after i have them? Sep 28 14:18:23 jovox: what kind of information do you need? Sep 28 14:26:18 gremlin[it]: I'll have to go now... Thanks for helping out! :) Sep 28 14:30:44 florian, well, I'm trying to figure out how it's supposed to be used. I took a look in the /init script and found that the ROOT_DEVICE is parsed from /proc/cmdline. So far so good. But what baout BOOT_ROOT? Am I supposed to edit the initscript and set it? Another thing I don't understand is that it seem to boot my "main" rootfs even though the initramfs fails to boot it as I havn't set the BOOT_ROOT which is a bit odd. I thought it would just stop if it couldt bo Sep 28 14:35:50 okay, DLed everything, but how do I get gmp out of there? Sep 28 14:41:21 jovox: I have been using an older version some time ago. There the actual boot scripts (e.g. 85-blockboot.sh) set it. Sep 28 14:43:00 Ah ok, I see, BOOT_ROOT is set to /mnt Sep 28 14:44:06 thanks! Sep 28 14:47:01 Another question, when booting it's it prints out that its uncompressing kernel etc, but after a few seconds the screen goes black and stays black for quite a while.. Is this due to some missing splash image or something? Sep 28 14:47:05 jovox: yw Sep 28 14:48:00 It makes it a bit difficult to see whats going on as dmesg doesnt contain any usful info at all Sep 28 14:50:50 jovox: if you use jffs2: mouting a large jffs2 filesystem can take some time e.g. Sep 28 14:52:00 it's ext3 on a a hard drive Sep 28 14:52:29 morning Sep 28 14:53:14 could it be that it's trying to display a splash and I havn't setup any vga mode in grub? Sep 28 15:19:15 * rkirti looks around for advice on docs.... Sep 28 15:20:01 I am putting an introductory talk on OE and cant seem to locate some of the good presentations I remember seeing earlier.. :-/ Sep 28 15:20:22 aren't some of them in contrib/ in the git repo? Sep 28 15:20:25 * kergoth pokes his memory Sep 28 15:20:50 03Thomas Zimmermann  07shr/import * r262328a73d 10openembedded.git/ (3 files in 3 dirs): Sep 28 15:20:50 xserver-xorg: update to 1.7RC3 Sep 28 15:20:50 Signed-off-by: Klaus Kurzmann Sep 28 15:29:48 I'm trying todo a tar xvf xxx-image.tar.gz from initramfs but get a 'invalid tar magic'. I thought it was caused by a corrupt tar but after copying it from the disk into my ubuntu it worked fine to extract it. Does it have anything todo with the version of tar in the initramfs image? Sep 28 15:31:24 jovox: try "tar xzvf" Sep 28 15:32:10 oki, will do Sep 28 15:33:01 hi rkirti Sep 28 15:33:17 the 'z' tells tar to gunzip the file Sep 28 15:33:32 ok Sep 28 15:34:27 hi florian :) Sep 28 15:36:42 thesing, that worked better. Thank you Sep 28 15:37:36 jo Sep 28 15:38:22 hello woglinde Sep 28 15:38:43 * rkirti never knew zecke_ was speaker at FOSS.IN :) Sep 28 15:39:00 rkirti *g* Sep 28 15:39:24 I always look at his trophy in his berlin flat Sep 28 15:42:09 woglinde: berlin flat ? isnt he from taiwan ? Sep 28 15:43:38 rkirti: you never know _where_ he is, but imho he officially lives in berlin. Sep 28 15:44:10 now he is in .tw again Sep 28 15:45:17 :) Sep 28 15:45:27 woglinde: sure? ;) Sep 28 15:45:43 florian I met him last thursday Sep 28 15:45:54 morning Sep 28 15:46:06 woglinde: in .tw? Sep 28 15:46:09 good morning hrw Sep 28 15:46:09 hi hrw Sep 28 15:46:19 hi hrw Sep 28 15:48:02 hrw: I finally managed to push a new xserver-common bb Sep 28 15:48:12 florian: cool - will look in next days Sep 28 15:58:46 03Koen Kooi  07shr/import * r6d87f76d0a 10openembedded.git/recipes/lua/ (4 files in 2 dirs): lua 5.1: attempt to get in in a buildable state again Sep 28 15:58:47 03Klaus Kurzmann  07shr/import * rbef46e0c5b 10openembedded.git/recipes/efl1/evas/rtl_cursor.patch: Sep 28 15:58:47 evas: update rtl_cursor.patch to apply against current EFL Sep 28 15:58:47 Signed-off-by: Klaus Kurzmann Sep 28 15:58:55 03Klaus Kurzmann  07shr/import * r5b9dcba9e6 10openembedded.git/recipes/efl1/ (edje-native_svn.bb edje_svn.bb): Sep 28 15:58:55 edje(-native): add depend for lua Sep 28 15:58:55 Signed-off-by: Klaus Kurzmann Sep 28 15:59:02 03Klaus Kurzmann  07shr/import * r89c85c5508 10openembedded.git/recipes/efl1/ecore.inc: Sep 28 15:59:02 ecore.inc: make it call autopoint to configure po Sep 28 15:59:02 Signed-off-by: Klaus Kurzmann Sep 28 15:59:10 03Klaus Kurzmann  07shr/import * r1125e7e56f 10openembedded.git/conf/distro/include/shr-autorev-unstable.inc: Sep 28 15:59:10 shr-autorev-unstable: bump EFL_SRCREV to 42764 Sep 28 15:59:10 Signed-off-by: Klaus Kurzmann Sep 28 16:01:05 03Dirk Opfer  07org.openembedded.dev * r01698e5b4b 10openembedded.git/recipes/lua/lua5.1_5.1.4.bb: Sep 28 16:01:05 lua5.1: Don't delete ${STAGING_LIBDIR}/*.so* Sep 28 16:01:05 This breaks other packages. There is no need to delete ${STAGING_LIBDIR}/*.so*. Sep 28 16:01:05 lua5.1 doesn't generate any *.so* Sep 28 16:01:06 03Dirk Opfer  07org.openembedded.dev * r8e8e320564 10openembedded.git/recipes/fastcgi/ (fastcgi_2.4.0.bb files/cstdio.patch): fastcgi: Add patch to compile with gcc4.4 Sep 28 16:21:58 hey guys Sep 28 16:22:04 heres an annoying problem... Sep 28 16:22:33 anyone know how to get networkmanager to ignore usb0 but still have it automatically setup with a static IP when I plug in my PDA ? Sep 28 16:23:03 networkmanager suckz Sep 28 16:23:09 deinstall it Sep 28 16:23:27 I like it - it manbaes all my wlans and v-lans nicely Sep 28 16:23:27 and use ifplugd and if/updown directly Sep 28 16:23:38 *manages Sep 28 16:23:51 for this you can use wpasupplicant guessnet and ifupdown Sep 28 16:23:54 does ifplugd have a gui and handle both wlan and vlans ? Sep 28 16:23:56 which is more powerfull Sep 28 16:24:08 vi Sep 28 16:24:18 gah. Sep 28 16:24:36 nm is doing everything else fine - I just want it to ignore usb0 Sep 28 16:24:39 iirc networkmanager ignores all interfaces configured in /etc/network/interfaces Sep 28 16:24:43 gui is for people which dont understands the underlying problems and is then pissed by networkmanager Sep 28 16:24:48 Spyro1: NM has option to ignore interfaces Sep 28 16:24:55 florian that must be new Sep 28 16:24:58 woglinde: I think you have no idea who you're talking to :) Sep 28 16:25:15 woglinde: or very old Sep 28 16:25:26 spyro I got pissed from networkmanager often enough Sep 28 16:25:40 hrw: I dont want it to totally ignore it, thats the problem. I want it to automatically give it a static IP when usb0 appears Sep 28 16:25:58 but I dont want it to disconnect me from my wifi and try to use usb0 for internets Sep 28 16:26:10 *g* Sep 28 16:28:49 Spyro1: so you have to tell NM to ignore usb0 Sep 28 16:28:59 NM use only one interface for internet Sep 28 16:29:08 hrw: how? Sep 28 16:29:13 this is its main bug Sep 28 16:29:24 Spyro1: check NM arguments Sep 28 16:29:29 .o(...) Sep 28 16:29:36 * hrw moved to connman on devices and nothing on laptop/desktop Sep 28 16:35:59 *failing to find an option to get networkmanager to keep its mitts off a connection Sep 28 16:36:30 so maybee florian suggestion is right Sep 28 16:36:54 networkmanager will igonre usb0 if you configure it in /etc/network/interfaces Sep 28 16:36:57 but I bet not Sep 28 16:39:45 ah Sep 28 16:39:55 its ok if you leave out allow-hotplug Sep 28 16:40:53 still annoying though Sep 28 16:41:23 should have an option for 'non-internet' or something so that it can manage the connection but not try to disconnect you from the internet... Sep 28 16:42:07 or a managed-by: option in the interfaces file so you can choose another program to manage select interfaces Sep 28 16:42:29 write to the networkmanager bugtracker Sep 28 16:42:36 might do. Sep 28 16:42:41 but I guess there is already a featurerightquest there Sep 28 16:43:04 nmms biggest flaw is that it assumes _eveything_ is a net connection Sep 28 16:44:39 *sigh* time to see why mmc stopped working on _one_ device of my 6 only... Sep 28 16:45:08 rkirti: wb Sep 28 16:45:53 htop showing load average : 18-19 ... :'( Sep 28 16:51:10 bbl Sep 28 17:11:10 btw, is there any unused collie in Europe? Sep 28 17:18:43 jay7 ask better on the dev ml Sep 28 17:24:33 03Koen Kooi  07shr/import * rfb743b5c3b 10openembedded.git/ (5 files in 4 dirs): Sep 28 17:24:33 enlightenment: bump SRCREV Sep 28 17:24:33 * runtime tested on angstrom/beagleboard Sep 28 17:24:33 * upgrade paths intact and working Sep 28 17:24:44 03Koen Kooi  07org.openembedded.dev * rd1a1aa3c3f 10openembedded.git/ (conf/checksums.ini recipes/ekiga/ptlib_2.6.5.bb): ptlib: update to 2.6.5 Sep 28 17:24:45 03Koen Kooi  07org.openembedded.dev * r3908084687 10openembedded.git/ (conf/checksums.ini recipes/ekiga/opal_3.6.6.bb): opal: add 3.6.6 Sep 28 17:24:46 03Koen Kooi  07org.openembedded.dev * rdffaf5aa89 10openembedded.git/conf/machine/omap4430-sdp.conf: omap4430-sdp: add machine file for omap4430-sdp, based on beagleboard machine file Sep 28 17:24:47 03Koen Kooi  07org.openembedded.dev * ra487ae0fa8 10openembedded.git/ (6 files in 5 dirs): Sep 28 17:24:48 glibmm: add 2.22.1, fix packaging Sep 28 17:24:50 cairomm: add 1.8.2 Sep 28 17:24:52 pangomm: add 2.26.0 Sep 28 17:24:54 gtkmm: add 2.18.1 Sep 28 17:24:56 03Koen Kooi  07org.openembedded.dev * r7772e33fa5 10openembedded.git/recipes/linux/ (5 files in 3 dirs): linux-omap git: add support for omap4430-sdp Sep 28 17:25:39 mmmm...omap4 Sep 28 17:38:24 yes it has corect a9 Sep 28 17:38:28 cortex Sep 28 17:38:39 and mp Sep 28 17:38:40 yo khem Sep 28 17:38:59 hey, i meant to get a diff to you a week ago or so Sep 28 17:39:05 03Dirk Opfer  07shr/import * r7ac0489828 10openembedded.git/recipes/lua/lua5.1_5.1.4.bb: Sep 28 17:39:05 lua5.1: Don't delete ${STAGING_LIBDIR}/*.so* Sep 28 17:39:05 This breaks other packages. There is no need to delete ${STAGING_LIBDIR}/*.so*. Sep 28 17:39:05 lua5.1 doesn't generate any *.so* Sep 28 17:39:29 m4t: hi Sep 28 17:39:50 03Thomas Zimmermann  07shr/import * rba6fa01132 10openembedded.git/ (3 files in 3 dirs): Sep 28 17:39:50 ffalarms: Update to 0.2.4 Sep 28 17:39:50 Signed-off-by: Klaus Kurzmann Sep 28 17:39:53 everything has been running stable for a week or so Sep 28 17:39:58 I have been in the middle of move this last week. hopefully now I will get rewired Sep 28 17:40:02 m4t: cool Sep 28 17:40:10 oh alright Sep 28 17:40:16 Did you get uboot 1.41 fixed Sep 28 17:40:23 i ended up using dcron instead of vixie though Sep 28 17:40:32 vixie would crap out if the clock changed suddenly Sep 28 17:40:35 eg. epoch -> current Sep 28 17:40:39 i see Sep 28 17:40:57 and i think it would run a huge amount of iterations of something trying to catch up to the new time Sep 28 17:40:57 not sure, but 100% cpu Sep 28 17:41:14 i also added an rdate init script to try to fix that, to run after networking and before cron Sep 28 17:41:28 it'll all be in the diff though Sep 28 17:46:30 m4t: Do you have time set correctly Sep 28 17:46:44 yes, there is no rtc though Sep 28 17:46:48 on the board Sep 28 17:46:55 ok Sep 28 17:47:04 so now it brings networking up, does rdate, then does openntpd/dcron Sep 28 17:47:16 also this mdadm init script is working properly Sep 28 17:47:25 ok nice Sep 28 17:47:32 it starts monitoring the drives listed in /etc/mdadm.conf at boot, and brings the array down at shutdown Sep 28 17:47:47 mdadm 3.0 w/a uclibc patch Sep 28 17:47:49 neat Sep 28 17:48:16 s/canonicalize_file_name/realpath Sep 28 17:48:49 its set to -Werror so i had to bring a char[] array in for realpath() Sep 28 17:48:52 should be safe though Sep 28 17:50:24 hmmm cant say without looking at the patch Sep 28 17:51:03 'git diff' is all i should need to run, in my 'openembedded' dir? Sep 28 17:51:39 erm hmm Sep 28 17:51:51 i need to diff vs the git tree Sep 28 17:52:55 m4t: yeah Sep 28 17:53:15 m4t: actually you can stage the files you changed with git add Sep 28 17:53:48 and then run git diff --cached Sep 28 17:54:07 ok Sep 28 17:54:37 or you can commit to your tree and then use git show Sep 28 18:00:17 hrm i already added everything via git commit -a Sep 28 18:00:36 the changes show up in 'git log' Sep 28 18:01:41 ok Sep 28 18:02:11 he khem Sep 28 18:15:04 i am 'git add' / 'git commit' ing everything Sep 28 18:15:10 w/comments Sep 28 18:15:18 i will try to get that done tonight Sep 28 18:19:11 re florian Sep 28 18:19:58 re Sep 28 18:20:32 wb florian Sep 28 18:28:29 woglinde: hello Sep 28 18:28:58 m4t: then you can check the commit ids in git log Sep 28 18:29:28 and then do git format-patch Sep 28 18:29:52 this will spew out patches for each commit you did on top of origin Sep 28 18:31:10 khem yea i figured that out, thanks Sep 28 18:31:26 my first ci was mid august i think Sep 28 18:32:01 Date: Fri Aug 28 00:45:20 2009 -0400 Sep 28 18:32:01 not too far behind Sep 28 18:37:01 or you can do git show Sep 28 18:37:11 if you just want one patch Sep 28 18:39:22 gotta run though, i will try to get these diffs tonight Sep 28 18:59:18 hi, what happend if you install abi3 packages on abi2 rootfs? Sep 28 19:00:43 re Sep 28 19:04:25 hrw: wb Sep 28 19:07:54 03Klaus Kurzmann  07shr/import * r5895722827 10openembedded.git/recipes/efl1/ (3 files in 2 dirs): Sep 28 19:07:54 evas_svn.bb: split the rtl patch into one probably appearing upstream soon Sep 28 19:07:54 and one probably not, which makes handling the patches easier on EFL bumps. Sep 28 19:07:54 Signed-off-by: Klaus Kurzmann Sep 28 19:08:39 GNUtoo: the rootfs may not work anymore :) Sep 28 19:08:51 ok thanks Sep 28 19:08:56 so I won't do it Sep 28 19:09:08 basically it was the recent tripplet change Sep 28 19:10:01 GNUtoo: iirc some of my systems were abi2 and now are abi2/3 Sep 28 19:10:23 ok Sep 28 19:10:31 maybe I should backup and try Sep 28 19:10:33 then Sep 28 19:18:59 next days will be tough.. 2.6.31 for boards on desk Sep 28 19:29:59 * mwester slides an NSLU2 onto hrw's desk, and waits... Sep 28 19:31:56 * chouimat goes 2.6.32-rc1 running on his EDmini Sep 28 19:31:59 s Sep 28 20:01:26 mwester: I had one under desk Sep 28 20:14:22 gm Sep 28 20:20:18 hi ant Sep 28 20:22:25 what's the toolchain problem with recent ubuntu? Sep 28 20:30:21 hi woglinde Sep 28 20:32:10 crofton hm Sep 28 20:32:18 I could test wendsday at work Sep 28 20:42:36 03Andrea Adami  07org.openembedded.dev * r5d56edd5ad 10openembedded.git/recipes/kexec/files/fix-arm-arch-detection.patch: kexec-tools_2.0.1: fix compiler warnings (and literal comparison too...) Sep 28 21:15:04 Crofton|work: what sort of toolchain problem.? Sep 28 21:15:38 dunno, I think people have been downgrading gcc on ubuntu to get OE builds to work Sep 28 21:15:48 I was talking to someone who mentioned this Sep 28 21:16:09 Crofton|work: hmmm it could be the fortify sources thing I suspect Sep 28 21:17:16 what is that? Sep 28 21:21:16 oh they turn on defaults FORTIFY_LEVEL=2 for gcc which yields more warnings Sep 28 21:21:32 and if warnings are treated as errors then your compile fails Sep 28 21:23:38 if I get the exact problem I will know more Sep 28 21:23:43 is there an easy workaround Sep 28 21:23:45 cool Sep 28 21:24:31 yeah add -U_FORTIFY_SOURCE to the CFLAGS for the -native recipe which fail Sep 28 21:25:11 but no way to do this on the ubuntu machine? Sep 28 21:25:40 oh changing the defaults hmm Sep 28 21:25:58 Well you could have your own spec file fed to gcc Sep 28 21:28:07 $ gcc -E -dD -x c /dev/null|grep FORTIFY Sep 28 21:28:07 #define _FORTIFY_SOURCE 2 Sep 28 21:28:27 thats my karmic box and I build OE without any issues on it Sep 28 21:29:06 so it could be some other issue which I am not aware Sep 28 21:51:34 Hmm, I'm also currently fixing a bare metal cross toolchain script to deal with recent Ubuntu 9.04 gcc's. Sep 28 21:58:20 likewise: which gcc they use? Sep 28 21:58:40 chouimat: 4.3.3, I think I just fixed it. Sep 28 21:58:47 ok Sep 28 21:58:53 chouimat: toolchain for ARM Cortex-M3 Sep 28 21:59:08 ok Sep 28 22:25:04 likewise: what problem is it posing ? Sep 28 22:25:43 khem: do you have an idea about klcc ldflags? http://fr.pastebin.ca/1583344 Sep 28 22:26:02 ant__: sorry did not have time to look into it so far Sep 28 22:26:09 np Sep 28 22:28:46 ant__: --no-undefined is ld option and my guess is that klcc is calling gcc to do the linking here Sep 28 22:28:58 if you do -Wl,--no-undefined it should work Sep 28 22:29:54 khem: simple formatting errors in binutils end up as warnings, propagating to errors. No biggie. Sep 28 22:30:08 ah Sep 28 22:30:13 khem: I could fix it. Sep 28 22:30:15 using --disable-werror Sep 28 22:30:35 yeah this is the FORTIFY_SOURCES thing Sep 28 22:31:46 khem: upstream binutils had a fix for the formatting warning, so I'm now without --disable-werror again. Sep 28 22:32:01 khem: any idea when that got into Ubuntu? Sep 28 22:32:16 must be somewhere in the last few weeks Sep 28 22:36:47 likewise: oh the fortify sources have been there for years now Sep 28 22:37:00 starting 8.04 I think Sep 28 22:38:34 03Graeme Gregory  07org.openembedded.dev * r725fdd557f 10openembedded.git/recipes/strace/ (3 files in 2 dirs): Sep 28 22:38:34 strace_4.5.18.bb : newer strace for avr32, 4.5.15 wasnt building here Sep 28 22:38:34 Forward ported the neccessary patches and think I made the correct set Sep 28 22:38:34 of changes to get them to merge. Sep 28 22:40:21 khem: where is FORTIFY_SOURCE set to 2 now as a default then? Sep 28 22:42:36 yes it is Sep 28 22:43:14 ant__: I think using -Wl,--no-undefined is probably right thing to do Sep 28 22:43:21 klcc is a perl script yuck Sep 28 22:43:23 khem: I figured, but _where_ is it set? Sep 28 22:43:34 khem: klcc seems to ignore the options I pass Sep 28 22:43:44 likewise: oh its a builtin define :) you need ubuntu gcc srcs Sep 28 22:44:11 khem: ah, now I get it :-) Thanks! Sep 28 22:44:20 likewise: you can undefine it if you like on commandline Sep 28 22:44:33 yup, I understood the rest of what you said. :-) Sep 28 22:44:57 it's too late here for me to be bright enough - I'm off to get some zzz Sep 28 22:45:14 likewise: ok sleep well Sep 28 22:46:26 ant__: what do you mean Sep 28 22:46:34 ant__: which package is it Sep 28 22:46:55 khem: hack lightly, and enjoy your evening Sep 28 22:46:57 kexec-tools-klibc-static_2.0.1 Sep 28 22:47:01 we start with: Sep 28 22:47:02 # standart oe cflags don't work with klcc Sep 28 22:47:02 export CFLAGS="" Sep 28 22:47:02 export CPPFLAGS="" Sep 28 22:47:02 export LDFLAGS="" Sep 28 22:47:07 in the recipe :/ Sep 28 22:47:29 I'm trying to adapt the new 2.0.1 Sep 28 22:49:45 ok export LD too Sep 28 22:49:54 because klcc uses LD and CC Sep 28 22:51:05 and it assumes that ld is invoked directly Sep 28 22:51:12 in your case LD=gcc Sep 28 22:51:17 which is the problem Sep 28 22:51:36 that's not all... Sep 28 22:51:38 and klcc chops the -Wl, Sep 28 22:51:44 I have to add --nostdinc Sep 28 22:52:10 so I patch directly the Makefile Sep 28 22:52:10 well nostdinc is not needed for linking Sep 28 22:52:53 but keep in mind the fact that klcc should use the ld directly and the LDFLAGS should have linker options only Sep 28 22:53:22 assumption that gcc will be used for linking thereby using gcc options will end up in errors like what you are seeing Sep 28 22:53:32 I start to see Sep 28 22:55:31 http://www.google.it/url?sa=t&source=web&ct=res&cd=1&url=http%3A%2F%2Fbooks.google.it%2Fbooks%3Fid%3DwQ6r3UTivJgC%26pg%3DPA288%26lpg%3DPA288%26dq%3Dklibc%2Bstatic%2Bcompile%26source%3Dbl%26ots%3DEJXnNrcBEs%26sig%3DIF_-NrqwSDn_gAFSPMH-QlBmMss%26hl%3Dit%26ei%3Dzj7BSqydBMbdsgbRnd0z%26sa%3DX%26oi%3Dbook_result%26ct%3Dresult%26resnum%3D1&ei=zj7BSqydBMbdsgbRnd0z&rct=j&q=klibc+static+compile&usg=AFQjCNE-sCXc7_6SWwxuhtUwgWXEpYXuKQ Sep 28 22:55:46 ^^ this is using plain gcc...as workaround Sep 28 23:02:27 yeah this will work Sep 28 23:02:28 too Sep 28 23:02:32 avoid klcc Sep 28 23:04:10 but I don't give up so easy... Sep 28 23:04:16 ;) Sep 28 23:18:57 03Andrea Adami  07org.openembedded.dev * r11e60ddab0 10openembedded.git/recipes/kexec/files/kexec-tools-2-klibc.patch: kexec-tools-2-klibc.patch: header was commented out during test, fix Sep 28 23:29:30 khem: thx Sep 28 23:29:35 good night Sep 29 00:41:13 has someone worked in lzma recipes? **** ENDING LOGGING AT Tue Sep 29 02:59:58 2009