**** BEGIN LOGGING AT Sun Jan 10 02:59:56 2010 Jan 10 03:03:53 I'm trying to build an image with the angstrom-2008.1 distro from the stable/2009 branch, and it's trying to build madwifi even though wifi isn't one of the target machine's features. Jan 10 05:18:04 * * OE Bug 4197 has been RESOLVED (FIXED) by raj.khem(AT)gmail.com Jan 10 05:18:06 * * Add a new image minimal-native-sdk-image.bb Jan 10 05:18:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4197 Jan 10 08:21:39 so I just filled x11-image Jan 10 08:21:46 err Jan 10 08:21:54 finished* Jan 10 08:22:17 Is there any way to test my webcam?> Jan 10 08:22:52 the debug output is here: http://pastebin.com/m491d72de Jan 10 09:11:19 morning all ! Jan 10 11:13:59 mickeyl: good morning Jan 10 11:14:25 good morning phil Jan 10 11:15:04 03Michael 'Mickey' Lauer  07org.openembedded.dev * r657fea552c 10openembedded.git/recipes/linux/linux-on-qualcomm-s-msm/htcraphael/defconfig: Jan 10 11:15:04 linux-on-qualcomm-s-msm: htcraphael: first step sanitizing defconfig for GNU/Linux Jan 10 11:15:04 disable virtual keyboard, enable sysv ipc, enable nfs, enable cifs, enable rtc Jan 10 11:15:15 03Michael 'Mickey' Lauer  07org.openembedded.dev * r2610f65bd7 10openembedded.git/ (5 files in 2 dirs): sysstat: refactor into .inc file, add 9.0.6 Jan 10 11:19:22 firefox won't build for me, asserts (struct sizes) in firefox-3.5.5-r2/mozilla-1.9.1/security/nss/lib/jar/jarfile.c around line 57 are false Jan 10 11:26:17 03Koen Kooi  07org.openembedded.dev * r8274bc3cdd 10openembedded.git/recipes/mplayer/mplayer_svn.bb: mplayer: bump SRCREV Jan 10 11:26:18 03Koen Kooi  07org.openembedded.dev * rb086d73777 10openembedded.git/recipes/ (ffmpeg/ffmpeg_svn.bb gstreamer/gst-ffmpeg_0.10.8.bb): ffmpeg: bump SRCREV Jan 10 11:26:18 03Koen Kooi  07org.openembedded.dev * r31113502d6 10openembedded.git/recipes/navit/ (navit-fpu.inc navit.inc): navit: fix build with qt4 in staging, only enable floats when you have an fpu Jan 10 12:27:05 * * OE Bug 3768 has been REOPENED by Jan 10 12:27:07 * * SharpROM-compilation emits long (but harmless?) errors when compiling Jan 10 12:27:09 * * http://bugs.openembedded.net/show_bug.cgi?id=3768 Jan 10 12:54:42 JaMa: Good morning. Are you aware who is behind the freesmartphone/milestone5.5 builder on tinderbox? Jan 10 12:55:17 From the looks of it it's another buildbot that learns nothing but keeps recompiling the same stuff and failing at the same place Jan 10 12:56:21 Laibsch: probably mickeyl knows Jan 10 12:57:12 Laibsch: btw openmoko builder stoped sending builds to tinderbox (but its still running based on changing donwloads on openmoko) Jan 10 12:57:25 OK Jan 10 12:57:26 good Jan 10 12:57:34 I will unblock them, then Jan 10 12:57:43 Laibsch: ah wait Jan 10 12:57:54 Laibsch: if you blocked it than maybe its not disabled Jan 10 12:58:13 Laibsch: I just sent 1 e-mail and then seen that more reports on tinderbox Jan 10 12:58:35 Laibsch: so I hoped that right people have seen my e-mail.. but probably not Jan 10 13:06:19 khem: I am currently reorganizing the bug tracker a bit. I think you are our toolchain guru, aren't you? I have created a toolchain component and would like to set you as default assignee. Is that OK with you? This can always be overridden for individual reports. Another option would be to have the default assignee be the mailing list as always but set you to cc as default. That as well can be overridden per ticket. Let me know what you thi Jan 10 13:06:58 * Laibsch thinks having some kind of experts for certain areas in OE would be quite helpful Jan 10 13:07:10 03Stanislav Brabec  07org.openembedded.dev * rfaf1793d47 10openembedded.git/recipes/xorg-xserver/ (8 files): Jan 10 13:07:10 xserver-kdrive: Build against xextproto-70-includes: Jan 10 13:07:10 * Ancient xservers are incompatible with xextproto >= 7.1. Jan 10 13:07:10 * Nobody would want to port working ancient code. Jan 10 13:07:10 * For more see http://lists.linuxtogo.org/pipermail/angstrom-distro-users/2010-January/003011.html Jan 10 13:07:14 * Introduced INC_PR. Jan 10 13:07:22 03Stanislav Brabec  07org.openembedded.dev * r371ee0f79e 10openembedded.git/recipes/alsa/alsa-lib_1.0.20.bb: alsa-lib: Bump PR, r1.5 appeared somehow in angstrom unstable feed and moves current r0.5 to morgue. Jan 10 13:07:23 03Stanislav Brabec  07org.openembedded.dev * rf3aefd497f 10openembedded.git/recipes/xorg-proto/xextproto-70-includes_7.0.5.bb: xextproto-70-includes: New old package, that provides headers required to compile ancient X servers (e. g. kdrive). Jan 10 13:14:04 * * OE Bug 4748 has been RESOLVED (FIXED) by Jan 10 13:14:06 * * qtopia-core-gui-image fails with qtopia-core Jan 10 13:14:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4748 Jan 10 13:15:04 * * OE Bug 4382 has been RESOLVED (FIXED) by Jan 10 13:15:06 * * qtopia-core-gui-image.bb' failed Jan 10 13:15:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4382 Jan 10 14:17:05 * * OE Bug 4389 has been RESOLVED (FIXED) by Jan 10 14:17:06 * * libxcomposite-1_0.4.0 needs libxext Jan 10 14:17:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4389 Jan 10 14:17:38 03Stanislav Brabec  07org.openembedded.dev * r85908133b0 10openembedded.git/recipes/xorg-proto/xextproto-70-includes_7.0.5.bb: xextproto-70-includes: Fixed comments. Jan 10 14:17:49 03Stanislav Brabec  07org.openembedded.dev * r474c06d9c5 10openembedded.git/recipes/balsa/ (balsa-2.4.1/gnome-icon-theme.patch balsa_2.4.1.bb): balsa: Update to version 2.4.2. Jan 10 14:28:44 03Andrea Adami  07org.openembedded.dev * rc9157c4a6a 10openembedded.git/recipes/kexecboot/linux-kexecboot-2.6.32+2.6.33-rc3/ (akita/defconfig spitz/defconfig): linux-kexecboot_2.6.32+2.6.33-rc3: update defconfigs for akita/spitz Jan 10 14:36:44 Laibsch: hi Jan 10 14:37:01 Hi Ant Jan 10 14:37:03 How are you? Jan 10 14:37:25 fine, just very busy Jan 10 14:37:42 I have time for my hobbies overnight ;) Jan 10 14:37:49 pls, would you mind to check the xserver/xorg situation in zaurus-2.6.inc? Jan 10 14:38:37 what exactly? Jan 10 14:39:12 seems strange that akita and spitz differs Jan 10 14:39:24 I just regrouped the settings there Jan 10 14:39:53 and the whole XSERVER = is a bit unclear to me Jan 10 14:40:09 recently Angstrom switched to xserver-xorg Jan 10 14:46:02 OK, I will take a look Jan 10 14:46:21 Maybe you in turn can have a glimpse at the mtd-utils and mtd-utils-native situation? Jan 10 14:46:55 It seems to unreproducably fail (unfortunately, for me it always fails, though) Jan 10 14:51:58 It seems mtd is not a problem on my setup. I last built a couple of weeks ago, though Jan 10 14:52:03 I'll retry Jan 10 15:06:06 * * OE Bug 1189 has been RESOLVED (LATER) by Jan 10 15:06:08 * * 'bitbake unionfs-modules' compile fails Jan 10 15:06:10 * * http://bugs.openembedded.net/show_bug.cgi?id=1189 Jan 10 15:09:03 03Rolf Leggewie  07org.openembedded.dev * r848d298214 10openembedded.git/recipes/unionfs/ (3 files): Jan 10 15:09:03 unionfs-modules: move to nonworking. Closes #1189 Jan 10 15:09:03 * These packages must have been non-compiling for about 3,5 years at least Jan 10 15:09:11 03Rolf Leggewie  07org.openembedded.dev * r7d5928037a 10openembedded.git/conf/distro/sharprom-compatible.conf: Jan 10 15:09:11 sharprom-compatible.conf: define PREFERRED_VERSION-linux-libc-headers Jan 10 15:09:11 * define PREFERRED_VERSION-linux-libc-headers to keep bitbake happy Jan 10 15:09:11 about recipes/prismstumbler/prismstumbler_0.7.3+0.7.4pre1.bb which Jan 10 15:09:13 assumes this variable to be defined Jan 10 15:17:53 ant__: I would consider moving the collie to xorg after some testing. You should talk to somebody who owns an akita if a different version of xserver works on the device Jan 10 15:18:18 But to compile images, I first need to get past the mtd-utils barrier Jan 10 15:47:26 Laibsch: I rebuilt mtd-utils and mtd-utils-native for angstrom-dev. no issues Jan 10 15:47:31 bbl Jan 10 15:48:50 could be that Angstrom is using an older gcc that is more forgiving Jan 10 15:49:17 But I'm not sure that guess is in sync with the failure I see Jan 10 16:31:08 RP: ping Jan 10 16:34:59 03Stanislav Brabec  07org.openembedded.dev * rd6279520f7 10openembedded.git/recipes/libgmime/libgmime-2.4_2.4.7.bb: Jan 10 16:34:59 gmime: Deleted libgmime-2.4 in favor of gmime. Jan 10 16:34:59 * gmime provides the same 2.4 API. Jan 10 16:34:59 * balsa package migrated to gmime by 474c06d9. Jan 10 16:39:03 i'm puzzled by the architecture number in arch.conf. my board say that the number for the board is 36, but I see other boards using that same number (and for yet another board I see that 36 is for arm6 Jan 10 16:39:08 how does that work? Jan 10 16:42:18 XorA, do you know anywhere the devfs stuff is explained? Jan 10 16:42:26 eFfeM: there is difference between arch number and ARM machine number Jan 10 16:42:56 Crofton|work: http://www.kernel.org/pub/linux/kernel/v2.6/ChangeLog-2.6.32 commit ab86e5765d41a5eb4239a1c04d613db87bea5ed8 Jan 10 16:43:25 likewise: ah ok Jan 10 16:43:59 arch number is defined by ELF. ARM machine number is from the registry at the arm linux web site Jan 10 16:47:34 03Stanislav Brabec  07org.openembedded.dev * rfd9e3c2b07 10openembedded.git/recipes/bzip2/ (5 files in 3 dirs): Jan 10 16:47:34 bzip2: Security update to version 1.0.5 (CVE-2008-1372, CERT-FI 20469) and a migration to autoconf. Jan 10 16:47:34 * More about security issue: Jan 10 16:47:34 * https://www.cert.fi/haavoittuvuudet/joint-advisory-archive-formats.html Jan 10 16:47:34 * http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2008-1372 Jan 10 16:47:38 * More about autoconf: See README.autotools in the patched source. Jan 10 16:48:32 XorA: pong Jan 10 16:49:39 RP: your email sounded like you had sent out TSC summary, I havent received it Jan 10 16:49:48 gmmm Jan 10 16:49:53 XorA: I did :/ Jan 10 16:49:57 I thought devfs went away Jan 10 16:50:00 now it is back Jan 10 16:50:12 XorA: I got one reply saying it was fine, nothing else :/ Jan 10 16:50:37 somewhere in the great mail system it has been eaten, I even checked my spam folders Jan 10 16:50:57 http://lkml.indiana.edu/hypermail/linux/kernel/0601.0/0283.html Jan 10 16:51:00 XorA: dp@xora.org.uk ? Jan 10 16:51:10 Crofton|work: I guess someone noticed udev sucks ass Jan 10 16:51:15 kergoth, mickey|sofa ping? Jan 10 16:51:16 yeah Jan 10 16:51:18 RP: yes thats the right address Jan 10 16:51:28 trying to find the emails about it coming back Jan 10 16:51:29 XorA: hmm. Let me resend now Jan 10 16:52:48 Crofton|work: as far as I can see it just puts the nodes both in sysfs and /dev/ but really useful for embedded devices Jan 10 16:53:03 RP: pong Jan 10 16:53:56 http://kerneltrap.org/node/6744 Jan 10 16:54:07 trying to find the logic behind the return Jan 10 16:56:09 kergoth: Did you get the TSC summary email? Jan 10 16:56:20 don't think so, no Jan 10 16:56:32 XorA: anything turned up now? Jan 10 16:56:40 saw a couple new replies on the tsc thread, but nothing else tsc related Jan 10 16:57:04 RP: has arrived now Jan 10 16:57:34 kergoth: I'd sent again. I don't know what my mail program is playing at :/ Jan 10 16:58:00 okay, yeah, got it now, thanks Jan 10 16:58:53 http://www.flickr.com/photos/44849007@N00/sets/72157623050010225/ was what I was doing yesterday :) Jan 10 17:01:17 RP: that looks pretty cool Jan 10 17:01:31 * eFfeM1 didn't take the challenge to go outside on his ATB Jan 10 17:04:11 eFfeM1: hard work but good fun Jan 10 17:30:37 ant__: mtd-utils fails for me for angstrom as well. I'm starting to suspect it may be a 64bit issue. Are you compiling on a 32bit host? Jan 10 17:30:41 has anyone compiled dsplink with oe-dev? Jan 10 17:30:50 Aditya1: check tinderbox first Jan 10 17:32:07 tinderbox? Jan 10 17:32:19 tinderbox.openembedded.org Jan 10 17:32:21 cigars? Jan 10 17:32:22 :P Jan 10 17:34:51 hmm Jan 10 17:35:00 maybe I should try building dsplink inside oe then Jan 10 17:35:48 03Koen Kooi  07org.openembedded.dev * re4876cdfb3 10openembedded.git/contrib/angstrom/build-feeds.sh: angstrom-feed-builder: add netbook-launcher-efl and sysstat Jan 10 17:35:59 03Koen Kooi  07org.openembedded.dev * r685431c719 10openembedded.git/recipes/linux/linux-kirkwood_2.6.33-rc1.bb: linux-kirkwood 2.6.33rc1: fix parsing and make PV adhere to our versioning policy (yes, it went backwards, but the recipe has never built properly in the first place) Jan 10 17:36:48 actual Laibsch: since you are a new face to me, was wondering if you'd used IMGLIB at all Jan 10 17:38:43 not knowingly Jan 10 17:41:04 gah Jan 10 17:41:11 internet freaked out Jan 10 17:41:31 Laibsch: sorry did you say anything? Jan 10 17:41:54 as an answer to your earlier question Jan 10 17:41:59 (18:38:42) Laibsch: not knowingly Jan 10 17:42:05 This channel is logged Jan 10 17:42:14 !logs Jan 10 17:42:15 Channel logs for #oe are archived at: Jan 10 17:42:16 http://hentges.net/tmp/logs/irc/%23oe Jan 10 17:42:17 Live-logs are available at Jan 10 17:42:19 http://hentges.net/tmp/logs/irc/livelogs/%23oe.livelog Jan 10 17:42:21 See ?? help-logs for usage instructions Jan 10 17:42:34 ooh nice Jan 10 18:31:55 03Rolf Leggewie  07org.openembedded.dev * r7695c100b0 10openembedded.git/conf/distro/sharprom-compatible.conf: Jan 10 18:31:55 sharprom-compatible.conf: fix spelling error, albeit a significant one Jan 10 18:31:55 underscore, not dash Jan 10 18:31:55 -PREFERRED_VERSION-linux-libc-headers = "2.4.18" Jan 10 18:31:55 +PREFERRED_VERSION_linux-libc-headers = "2.4.18" Jan 10 18:36:59 mtd-utils and mtd-utils-native just won't compile for me. Others seem completely unaffected or the builds are flakey and succeed at times while not at others: http://tinderbox.openembedded.org/packages/mtd-utils/ Jan 10 18:37:21 | /data/oe/laibsch/build/minimal/mkfs.ubifs/crc16.o: file not recognized: File format not recognized Jan 10 18:37:21 is one of the errors I get Jan 10 18:37:56 but the file does not seem to be anything out of the ordinary Jan 10 18:38:01 /data/oe/laibsch/build/minimal/mkfs.ubifs/crc16.o: ELF 64-bit LSB relocatable, x86-64, version 1 (SYSV), not stripped Jan 10 18:38:38 I'm kind of stuck at the moment Jan 10 18:49:58 What's the command that produces that error? Jan 10 18:53:39 pb_: please take a look at http://paste.debian.net/56266/ and http://paste.debian.net/56267/ Jan 10 18:54:48 The errors aren't always the same, but at least for me they always seem to be ubifs or ubi related Jan 10 18:57:43 Laibsch: based on the "file" output you provided, it sounds as if you have a problem with ${CC} not being respected. Jan 10 18:58:04 it looks like it is accidentally compiling one of the object files for your host but then trying to link it into a target binary. Jan 10 18:59:14 Hi, I tried to reproduce what Henning wrote on http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-January/016082.html on my local system. Jan 10 18:59:17 if you delete crc{16,32}.o and then rerun the build, what does the compile command for those two files look like? Jan 10 18:59:41 g++ --version g++ (Debian 4.4.2-8) 4.4.2 Jan 10 18:59:51 But I do not get any warnings. Jan 10 19:01:08 Now I want to try this in the OE build environment and want to set the tools according to Henning’s ones. Jan 10 19:01:40 Do I need to change those by adding `PREFERRED_PROVIDER` in my `local.conf`? Jan 10 19:03:08 others may answer with more authority, but I think, yes, that would be one way Jan 10 19:03:32 set them, and then build the targets individually to make sure you are actually getting the ones you want. Jan 10 19:08:04 just for the record, I am hating the udev caching stuff atm Jan 10 19:08:44 I meant PREFERRED_VERSION… of course. Jan 10 19:10:43 where does udev cache the dev nodes? Jan 10 19:13:27 Crofton: /dev/.udev/ is what I have on desktop Jan 10 19:13:38 thanks Jan 10 19:15:17 Crofton: can you please upgrade my patchwork account "jama" to maintainer? Jan 10 19:15:35 yeah Jan 10 19:15:39 Crofton: I would close few patches from fellow SHR devs where I know the state.. Jan 10 19:15:39 give me a sec Jan 10 19:16:12 pb_: now it's a different file: | /data/oe/laibsch/build/minimal/mkfs.ubifs/lpt.o: file not recognized: File format not recognized Jan 10 19:16:34 /data/oe/laibsch/build/minimal/mkfs.ubifs/lpt.o: ELF 64-bit LSB relocatable, x86-64, version 1 (SYSV), not stripped Jan 10 19:17:29 seems like the number of people being bitten by the udev stuff increases steadily Jan 10 19:17:34 JaMa, try now Jan 10 19:17:51 Crofton: works, thanks Jan 10 19:18:52 pb_: next one was /data/oe/laibsch/build/minimal/mkfs.ubifs/devtable.o Jan 10 19:20:49 and /data/oe/laibsch/build/minimal/mkfs.ubifs/hashtable/hashtable.o and /data/oe/laibsch/build/minimal/mkfs.ubifs/hashtable/hashtable_itr.o and then it looks like http://paste.debian.net/56273/ Jan 10 19:21:35 should I add myself to MAINTAINERS file, or do I have to be selected as ie maintainer for SHR, gta02, spitz? Jan 10 19:22:11 I would say you are doing a good job Jan 10 19:23:38 i just wanted to put contact somewhere in case someone have problem with some SHR recipe or some machine config I've changed.. Jan 10 19:23:55 but maybe commit author from git is enough :) Jan 10 19:29:30 Personally, I consult git log frequently, but I haven't looked at Maintainers file in years, I think Jan 10 19:36:41 so Jan 10 19:36:52 ti-dsplink-module definitely fails in oe-dev Jan 10 19:37:01 tried building out of and inside OE Jan 10 19:37:08 fails in both places with same error Jan 10 19:37:22 "error: implicit declaration of function 'IO_ADDRESS'" Jan 10 19:42:47 Laibsch: did you answer my previous question? I couldn't see anything obvious in the new output you pasted. Jan 10 19:49:39 pb_: do you want me to rerun the build via bitbake (as I did) or by calling make directly? I thought my posts contained the information you requested, my bad. Jan 10 19:50:05 should I call -c devshell? Jan 10 20:03:52 Laibsch: via bitbake is fine Jan 10 20:04:37 good, did I not provide enough backlog in the console logs I pasted, pb? Jan 10 20:04:50 yeah, I think that was the problem Jan 10 20:04:54 OK Jan 10 20:05:06 let me rerun things and paste again Jan 10 20:05:07 if you could delete those two files again, rerun the build, and then paste the whole log, that would be perfect Jan 10 20:05:13 ok Jan 10 20:05:16 just a minute Jan 10 20:05:59 pb_: does http://tinderbox.openembedded.net/public/logs/task/4318744.txt contain the information you need? Jan 10 20:06:54 that's from the previous run Jan 10 20:10:05 Laibsch: sadly no, it doesn't seem to Jan 10 20:13:36 pb_: http://paste.debian.net/56284/ is the log of a complete run, but I'm not sure it contains what you are looking for Jan 10 20:14:14 BTW, I have disabled PARALLEL_MAKE in the recipe locally. I was thinking maybe that is the reason for some of the erratic behaviour. Jan 10 20:14:18 not sure it helped much Jan 10 20:15:03 lines 45 and 46 separate the two runs Jan 10 20:16:33 It looks like a different file may have been affected this time Jan 10 20:16:46 03Martin Jansa  07org.openembedded.dev * rbe89272e36 10openembedded.git/MAINTAINERS: Jan 10 20:16:46 maintainers: add Martin Jansa entry Jan 10 20:16:46 * In case someone wants to blame/praise without checking git log :) Jan 10 20:16:46 Signed-off-by: Martin Jansa Jan 10 20:16:47 /data/oe/laibsch/build/minimal/flash_eraseall.o Jan 10 20:16:57 03Frederik Sdun  07org.openembedded.dev * r5f1354772a 10openembedded.git/recipes/libsdl/libsdl-ttf_2.0.9.bb: Jan 10 20:16:57 libsdl-ttf: Fix path to sdl-config Jan 10 20:16:57 Signed-off-by: Frederik 'playya' Sdun Jan 10 20:16:57 Signed-off-by: Martin Jansa Jan 10 20:17:24 yeah Jan 10 20:17:34 clearly something is a bit wrong with those pathnames as well. Jan 10 20:17:53 I guess it might be worth trying to debug why the files are landing in these weird locations before worrying about why they have the wrong contents. Jan 10 20:20:22 03Martin Jansa  07org.openembedded.dev * r957f6b2865 10openembedded.git/conf/distro/include/sane-srcrevs-fso.inc: Jan 10 20:20:22 libfsoframework: bump srcrev with fixed libnl-2.0.vapi Jan 10 20:20:22 Signed-off-by: Martin Jansa Jan 10 20:20:47 pb_: http://paste.debian.net/56287/ has the run including removal of flash_eraseall.o Jan 10 20:25:11 03Martin Jansa  07org.openembedded.dev * r8a75c3794c 10openembedded.git/conf/distro/include/preferred-shr-versions.inc: Jan 10 20:25:11 Revert "shr: temporary disable libsdl-ttf_2.0.9" Jan 10 20:25:11 * This reverts commit 814f07a935d8be208fa219c5f04a43a72bbd4ea6. Jan 10 20:25:11 * The the issue is fixed. Jan 10 20:31:06 pb_: I think it is mtd-utils-native that puts them there Jan 10 20:31:15 Maybe a broken staging function Jan 10 20:31:30 ah, that would make sense Jan 10 20:31:42 mtd-utils-native fails to build as well Jan 10 20:31:45 at least for me Jan 10 20:31:46 I doubt it is staging though, I think both mtd-utils and the -native one probably just have broken makefiles Jan 10 20:32:00 so they are, effectively, sharing a build directory Jan 10 20:32:08 yes and the wrong one Jan 10 20:32:26 you'll probably need to inspect the makefiles yourself to figure out why that is happening Jan 10 20:34:16 drat Jan 10 20:34:23 that's frightening Jan 10 20:34:40 but it looks like a somewhat known problem Jan 10 20:34:51 there are comments to that effect in the bb files themselves Jan 10 20:36:42 It looks like the makefiles use $PWD as the workdir Jan 10 20:36:51 that's where the stuff gets put Jan 10 20:41:44 03Martin Jansa  07org.openembedded.dev * r21499373b6 10openembedded.git/recipes/navit/files/shr/ (navit.desktop navitD.desktop): navit: use --force for fsoraw in SHR launchers (Thanks Zoff) Jan 10 20:41:55 03Christian Rüb  07org.openembedded.dev * rf293421563 10openembedded.git/recipes/navit/ (files/navit.xml files/plugins.xml navit_svn.bb): Jan 10 20:41:55 navit: enable dbus plugin in navit by default Jan 10 20:41:55 Signed-off-by: Martin Jansa Jan 10 20:42:00 yes, it's the mtd-utils-native package that does that Jan 10 20:42:25 mtd-utils-native fails in do_package_write Jan 10 20:42:40 no Jan 10 20:42:46 in do_populate_staging Jan 10 20:55:40 is it possible to set the continue flag in the bb shell? Jan 10 21:10:31 03Graeme Gregory  07org.openembedded.dev * r8a3b0c0dd6 10openembedded.git/ (2 files in 2 dirs): transmission_1.76+1.80b4.bb : add new version Jan 10 21:10:31 03Graeme Gregory  07org.openembedded.dev * r09be6bb208 10openembedded.git/: Merge branch 'org.openembedded.dev' of git+ssh://git.openembedded.org/openembedded into org.openembedded.dev Jan 10 21:10:32 03Graeme Gregory  07org.openembedded.dev * rd85d9ce302 10openembedded.git/MAINTAINERS: MAINTAINERS : fix my entry to remove openmoko which I never touch these days Jan 10 21:28:47 pb_: can you take a look at lines 922 and 923 of http://paste.debian.net/56298/? sysroot-destdir looks kind of fishy Jan 10 21:30:53 seems like at least a $ is missing, I think Jan 10 21:32:08 03Michael  07org.openembedded.dev * rb87ef25ed7 10openembedded.git/ (conf/checksums.ini recipes/python/pyrtm_0.2.bb): Jan 10 21:32:08 pyrtm: Added Python library for Remember the Time Jan 10 21:32:08 Signed-off-by: Martin Jansa Jan 10 21:34:51 ah crap Jan 10 21:35:25 patchwork should highlight Author in patch not name from "From: " Jan 10 21:35:52 whats pyrtm? Jan 10 21:35:55 can I git amend it as long its in HEAD? Jan 10 21:36:06 better not amend head Jan 10 21:36:13 I mean last commit Jan 10 21:36:18 aeh the remote Jan 10 21:36:27 no Jan 10 21:36:33 its not good practice Jan 10 21:36:38 localy okay Jan 10 21:36:41 but when you pushed Jan 10 21:39:11 I know that rebase/amend or any other history rewritting command shouldn't be used on remote.. I was hoping that there is way to make exception if its last commit but you're right of course.. Jan 10 21:39:51 JaMa you cant be sure someother commited stuff Jan 10 21:40:31 hi bluelightning Jan 10 21:40:47 why? git push checks that no? Jan 10 21:41:30 ok Jan 10 21:41:35 so I have a fix for oe-dev Jan 10 21:41:38 for dsplink Jan 10 21:41:49 I am not entirely sure what to do about it Jan 10 21:41:49 Aditya11 hm Jan 10 21:41:58 for some reason dsplink wasn't building with oe-dev Jan 10 21:41:59 write to the dev mailinglist Jan 10 21:42:08 or show me your patch Jan 10 21:42:21 and it was because IO_ADDRESS wasn't defined as IO_ADDRESS, it has a different name Jan 10 21:42:29 but dsplink still uses IO_ADDRESS Jan 10 21:42:59 ... I need to go figure out how to make a patch file heh Jan 10 21:44:04 git commit Jan 10 21:44:10 git format-patch Jan 10 21:46:22 woglinde: If you have a minute, would you be so kind to take another look at mtd-utils-native? Jan 10 21:46:38 woglinde: would that change the main repo? Jan 10 21:46:45 cause I dont want to break anything for everyone else Jan 10 21:46:45 Aditya11: no Jan 10 21:46:51 only your local copy of it Jan 10 21:46:55 Aditya11 with out push you modify all local Jan 10 21:47:07 ah cool Jan 10 21:47:30 laibsch hms I am really tired Jan 10 21:47:57 ok Jan 10 21:48:23 where do these patches go? Jan 10 21:49:40 http://upload.pixelfaction.com/files/4b4a4a1b_io.h.patch -- thats all the changes I made its just one line, not sure what a recipe could do to do that. Maybe after the kernel src is downloaded the recipe could add that line before doing aything else Jan 10 21:51:43 Aditya11 ~pastebin Jan 10 21:51:50 ~pastebin Jan 10 21:51:51 [~pastebin] A "pastebin" is a web-based service where you can paste anything over 3 lines without flooding the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://www.rafb.net/paste Jan 10 21:53:49 I know about pastebin Jan 10 21:53:56 just used to using that site for sharing files Jan 10 21:53:59 sorry Jan 10 21:54:51 http://pastebin.com/m1be93303 there Jan 10 22:06:24 how can I provide my own do_fetch so that it will be compatible with the rest? Jan 10 22:06:50 make do_fetch() { Jan 10 22:06:51 \ Jan 10 22:06:52 args Jan 10 22:06:54 } Jan 10 22:06:58 in your recipe Jan 10 22:08:54 woglinde: I mean... where should I put the sources, how should I handle updating sources etc Jan 10 22:09:05 aeh??? Jan 10 22:09:09 woglinde: I can't seem to find any docs about it Jan 10 22:09:26 woglinde: chromium uses its own weird program for syncing with svn (gclient) Jan 10 22:09:26 seems I dont understand your problem Jan 10 22:09:30 please explain it Jan 10 22:09:37 ah Jan 10 22:09:43 that has to go into bitbake Jan 10 22:09:51 woglinde: it fetches and configures sources at the same time Jan 10 22:09:57 woglinde: what do you mean? Jan 10 22:10:28 bitbake is the tool which builds the stuff oe has as meta-data for the various software Jan 10 22:11:14 yeah I know Jan 10 22:11:19 I am developing a bb file for chromium Jan 10 22:11:25 the google browser Jan 10 22:11:42 lol, seems like code.google.com is down Jan 10 22:13:53 woglinde: it's just that I can't use the svn fetch method, I need to use gclient (http://209.85.135.132/search?q=cache:Fv8gKQIHyngJ:code.google.com/p/gclient/+gclient&cd=1&hl=pl&ct=clnk) Jan 10 22:22:06 woglinde_: do you have a sec? Jan 10 22:23:04 woglinde_: classpath-native_0.98 picks antlr directly from build machine from tools/Makefile: Jan 10 22:23:07 > ANTLR = runantlr Jan 10 22:23:10 > ANTLR_JAR = /usr/share/antlr/lib/antlr.jar Jan 10 22:23:23 and fails on other machine where is no antlr installed, bot have antlr-native, any idea? Jan 10 22:23:32 hm Jan 10 22:23:36 noz yet Jan 10 22:23:37 sorry Jan 10 22:23:42 maybee its need a fix Jan 10 22:23:51 the 2nd machine has < ANTLR = /home/shr/shr-unstable/tmp/staging/x86_64-linux/usr/bin/java-initial -classpath antlr.Tool Jan 10 22:23:54 < ANTLR_JAR = Jan 10 22:25:33 hm Jan 10 22:25:44 sorry I am not sure at the moment if antlr is needed Jan 10 22:27:22 I tried even with antlr-native removed on 2nd.. but still fails :/ Jan 10 22:27:49 who is kaeilos-user on tinderbos? Jan 10 22:28:43 they have built it fine http://tinderbox.openembedded.net/packages/classpath-native/ but it failed the same before for them.. so I'm curious if they just installed antlr on buildhost or fixed it somehow Jan 10 22:30:40 woglinde: do you maybe have a hint for me? Jan 10 22:30:50 [23:13] < filip> woglinde: it's just that I can't use the svn fetch method, I need to use gclient (http://209.85.135.132/search?q=cache:Fv8gKQIHyngJ:code.google.com/p/gclient/+gclient&cd=1&hl=pl&ct=clnk) Jan 10 22:31:38 filip hm look into the bitbake code Jan 10 22:31:49 woglinde_: would you mind if I add --disable-gjdoc to configure params for -native? Jan 10 22:31:54 no Jan 10 22:34:53 woglinde: what bitbake version are you using? Jan 10 22:35:08 mom Jan 10 22:35:16 woglinde, JaMa: do you guys have an interest in Java in OE? Jan 10 22:35:46 aeh??? Jan 10 22:36:14 BitBake Build Tool Core version 1.8.999, bitbake version 1.8.999 Jan 10 22:36:24 Laibsch: not so much I would say.. Jan 10 22:36:40 woglinde_: Is the question about Java unclear? Jan 10 22:36:50 Laibsch: just we had cacao in feeds.. (based on some users request) Jan 10 22:37:01 woglinde_: thank you for letting me know your bitbake version Jan 10 22:37:04 Laibsch: but it always failed to build.. Jan 10 22:37:04 JaMa: thanks Jan 10 22:37:18 Laibsch: and worked on my host.. so I wanted to fix/workaround it Jan 10 22:37:25 Laibsch all recipes besides openjdk were transfered from jalimo to oe.dev Jan 10 22:37:26 woglinde_: do you mean that I should add a 'gclient' fetch method to the bitbake sources? Jan 10 22:37:36 filip jupp Jan 10 22:37:52 woglinde_: will it be accepted upstream? Jan 10 22:37:59 hm Jan 10 22:38:08 filip if it works why not Jan 10 22:38:32 woglinde_: the reason I am asking is that I have created theme-based categories in OE. python, java, cjkv, etc. People who are interested that the respective area in OE works nicely could be added to the cc list of that category. Jan 10 22:38:35 Laibsch: but as java programmer, I would be interested in having good vm also on my freerunner sometime.. Jan 10 22:38:40 woglinde_: makes sense? Jan 10 22:38:41 woglinde_: well it's specific for one application Jan 10 22:38:56 JaMa: let me know if I should add you to cc, then Jan 10 22:38:58 filip hm for now Jan 10 22:39:03 maybee there will other come Jan 10 22:39:18 Laibsch: add me, I'm interesed enough for cc.. Jan 10 22:39:39 you can always unsubscibe on an individual, per-ticket basis Jan 10 22:40:32 Laibsch: I'm looking forward to see how it will work.. nice idea! thanks Jan 10 22:45:18 JaMa: added Jan 10 22:50:02 http://pastebin.com/f19b9efc is the log.do_compile from my failing mtd-utils-native build. The second line is already fishy creating a directory where it shouldn't. Is there a script somewhere that I can inspect? It doesn't seem to be run.do_compile but I'm not sure where else to look. Jan 10 22:59:08 hrw|gone: I see you've touched mtd-utils in the past and deduct a certain interest from that ;-) The recipe puts stuff in strange places even for people where it doesn't complete blow up Jan 10 22:59:35 The binaries are put into /data/oe/laibsch/build/minimal, my build directory and the build blows up Jan 10 23:00:11 for bluelightning, stuff is put into /home/paul/oe/tmp/oe-spitz-dev2/work/i686-linux/mtd-utils-native-1.2.0+git-r4/image/home/paul/oe/tmp/oe-spitz-dev2/staging/i686-linux/usr/ which completes but doesn't look right Jan 10 23:00:23 Maybe you can have a look when you find the time Jan 10 23:11:04 * * OE Bug 5371 has been created by  Jan 10 23:11:06 * * kaffeh-native fails to compile Jan 10 23:11:08 * * http://bugs.openembedded.org/show_bug.cgi?id=5371 Jan 10 23:15:10 woglinde_: and as you asked about pyrtm.. its for task scheduling.. someone sent gui for that.. but I asked few changes.. so I pushed only library now http://patchwork.dev.bearstech.com/patch/611/ Jan 10 23:16:06 jama hm ah okay Jan 10 23:20:47 03Martin Jansa  07org.openembedded.dev * r73a66c2746 10openembedded.git/recipes/mplayer/mplayer_git.bb: Jan 10 23:20:47 mplayer: update SRCREV for om-gta02 Jan 10 23:20:47 Signed-off-by: Martin Jansa Jan 10 23:20:58 03Martin Jansa  07org.openembedded.dev * r762fa9e865 10openembedded.git/recipes/classpath/ (3 files): Jan 10 23:20:58 classpath-native: disable-gjdoc, it calls antlr and fails without antlr installed on buildhost (runantlr from antlr-native is found by configure, but antlr.jar from staging doesn't work) Jan 10 23:20:58 Signed-off-by: Martin Jansa Jan 10 23:20:59 03Martin Jansa  07org.openembedded.dev * rfa39ec90d3 10openembedded.git/recipes/navit/files/shr/ (navit.desktop navitD.desktop): Jan 10 23:21:02 navit: use fsoraw params in right order Jan 10 23:21:04 Signed-off-by: Martin Jansa Jan 10 23:21:19 PaulePanter: I think you want PREFERRED_VERSION, not PREFERRED_PROVIDER, though I can't think offhand why the version of gcc should be especially significant for your issue. I suspect your problem has more to do with missing feature test macros than any particular compiler. Jan 10 23:23:52 pb__: Yes, thanks. See my message 20:08. Jan 10 23:25:55 pb__: Henning wrote that gcc and g++ since version 4 deal differently with headers. Jan 10 23:26:28 pb__: Although I could not reproduce that on Debian [1]. Jan 10 23:26:30 http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-January/016095.html Jan 10 23:26:35 paule only g++ Jan 10 23:26:43 *g* Jan 10 23:26:57 woglinde: Ok, thanks for clarifying. Jan 10 23:27:05 I don't know what he was thinking of there, but I can't think of any significant way in which headers have different semantics in gcc 4. As I said, i think your problem is with the feature test macros. Jan 10 23:27:19 pb__ c-header inclusion Jan 10 23:27:30 is stricter with newer gcc versions Jan 10 23:27:35 aeh g++ Jan 10 23:27:36 args Jan 10 23:27:40 *g* Jan 10 23:27:58 stricter in what way? Jan 10 23:28:13 I try to reproduce this in OE with newer g++ and binutil versions. But in Debian there was no problem. Jan 10 23:28:20 you sometimes needs to include more or the right headers Jan 10 23:28:25 one example is vdr Jan 10 23:29:13 oh, I guess you mean that the library is less prone to include headers for you implicitly. Jan 10 23:29:24 I don't think the compiler itself is behaving any differently. Jan 10 23:29:34 he yeah Jan 10 23:29:43 technical its the preprocessor Jan 10 23:30:54 no, I don't think the preprocessor is doing anything different either, I think it's just the library headers that are different. Jan 10 23:31:09 although the g++ 4 preprocessor is different to the g++ 3 version, the semantics are not that different. Jan 10 23:32:02 but yeah, I could imagine the net effect being that you sometimes could avoid including all the necessary headers with older toolchains. Jan 10 23:32:31 *g* cstring vs string.h Jan 10 23:33:01 To test it in OE I now set PREFERRED_VERSION_binutils{,cross,cross-sdk} and P…V…_gcc{,cross,cross-sdk,cross-initial,cross-intermediate} to 4.4.2 Jan 10 23:33:15 binutil* to 2.20 Jan 10 23:33:19 binutils should be irrelevant, it has nothing to do with library processing Jan 10 23:33:27 … in my `local.conf`. Jan 10 23:33:33 the c++ headers are resolved long before your code gets anywhere near binutils. Jan 10 23:34:03 pb__: Thanks for all the information. I do not know anything about this. Jan 10 23:34:06 not that it will hurt to set your version of that, of course, but I don't think it will make any difference in this case. Jan 10 23:34:17 yeah sorry Jan 10 23:34:18 Anyway I hope I can reproduce Henning’s findings. Jan 10 23:34:25 binutils has nothing to do with it Jan 10 23:34:32 dont know why I wrote it Jan 10 23:34:58 woglinde: Oh now I know who Henning is. ;-) Thanks for helping me out with VDR. Jan 10 23:35:16 *g* Jan 10 23:35:48 How do all you guys manage your different configurations (local.conf)? Jan 10 23:36:22 Bitbake is looking for it in ${OE}/build/conf/ by default, is not it. Jan 10 23:38:04 * * OE Bug 4236 has been RESOLVED (WONTFIX) by Jan 10 23:38:06 * * ERROR: Patches_Linux_2.4.24_1.00.tar.gz: not in gzip format Jan 10 23:38:08 * * http://bugs.openembedded.net/show_bug.cgi?id=4236 Jan 10 23:46:23 hi rkirti Jan 10 23:48:04 * * OE Bug 2059 has been RESOLVED (INVALID) by Jan 10 23:48:06 * * glibc-utils wants to install file /usr/bin/iconv Jan 10 23:48:08 * * http://bugs.openembedded.net/show_bug.cgi?id=2059 Jan 10 23:57:11 Thanks for all your help. See you on the list. Jan 11 00:25:27 IS v4l2 already included in the x11-image in OE-DEV? Jan 11 02:45:28 aaaanyone use a webcam with OE? **** ENDING LOGGING AT Mon Jan 11 02:59:57 2010