**** BEGIN LOGGING AT Fri Jul 31 02:59:57 2009 Jul 31 03:45:43 When i create an image the result lacks anything in /dev. So even though the system should use udev, things like /dev/null and /dev/tty are missing so it fails to boot. Jul 31 03:46:17 Any clues as to the recipe responsible for creating basic device files needed for booting? Jul 31 04:40:37 makedevs-native, and you need to either have files/device_table-minimal.txt at the top of one of your BBPATH entries. image.bbclass actually creates the devices though. Jul 31 04:41:37 You can also set IMAGE_DEVICE_TABLES to point to another file other then files/device_table-minimal.txt Jul 31 04:54:20 Ah. Cheers ReaperOfSouls Jul 31 04:56:42 hmm... it is in the manual... bad me. Jul 31 06:57:23 morning Jul 31 06:58:13 how to add a shortcut to an application on desktop of the OE x11-image (angstrom)? Jul 31 07:25:02 how to change the angstrom loading... logo ? Jul 31 08:29:11 good morning Jul 31 09:23:45 morning Jul 31 09:27:07 hi hrw Jul 31 10:23:09 <_dash__> XorA: Thanks for ur help on day b4 yesterday, for the problem of mmap Jul 31 10:23:26 _dash__: no probs, hope you cured it Jul 31 10:24:05 <_dash__> XorA: yup I just moved everything to ext3 and all worked smoothly without error at first time Jul 31 10:24:30 <_dash__> NTFS sucks Jul 31 10:25:02 nothing wrong with NTFS, its just now a linux filesystem Jul 31 10:25:07 s/now/not/ Jul 31 10:25:11 <_dash__> of course it is stupidity to run bitbake in linux env. on NTFS Jul 31 10:25:34 Im actually quite stunned you got that far :-) Jul 31 10:26:29 <_dash__> Got that far then jumped back to big Zero and got that far ahead of it now successfully Jul 31 11:31:08 03Graham Gower  07org.openembedded.dev * ree1df327f9 10openembedded.git/ (classes/kernel.bbclass recipes/linux/linux.inc): Jul 31 11:31:08 kernel.bbclass : moved uImage creation from do_deploy to between do_compile Jul 31 11:31:08 and do_install as discussed on ML. Jul 31 11:31:08 linux.inc : removed uImage creation, its now in kernel.bbclass Jul 31 11:31:08 Signed-off-by: Graham Gower Jul 31 11:31:12 Acked-by: Graeme Gregory Jul 31 11:31:14 Acked-by: Koen Kooi Jul 31 12:19:40 http://www.embedded-computing.com/?p=626 Jul 31 12:53:37 Hi. I bumped into a convoluted linking thing. A bit off-topic, but maybe some guru here knows the answer. ;) Jul 31 12:53:59 I've got a dynamic executable, that uses some dynamic libraries (.so). Jul 31 12:54:17 I'd like to convert all that into a static executable. Jul 31 12:54:46 Problem is: some of the dyn libraries have the same global symbols (some C++ trash). Jul 31 12:55:08 I'm using visibility attribute + default visibility= hidden to make the dyn libraries hade those Jul 31 12:55:16 so, each "has its own"... they don't clash Jul 31 12:55:34 but if I try to link all the sources into one huge static executables, they would clash again. Jul 31 12:56:15 I think I'd need something like (sources that comprise single .so) -> object files -> one obj file that drops the symbols that are to be hidden from other modules Jul 31 12:56:22 and only then link those "hude objects" together Jul 31 12:56:31 does it use libtool? Jul 31 12:56:33 is that possible? aand is it feasible? Jul 31 12:56:42 nope, cmake Jul 31 12:56:46 ok Jul 31 12:56:53 cmake is a build system Jul 31 12:56:56 and I'll be honest with you... it's some proprietary crap Jul 31 12:57:04 libtool is a shared/static library abstraction Jul 31 12:57:15 well, no libtool needed, it's targeted at linux Jul 31 12:57:15 do you have .a files Jul 31 12:57:30 you need static .a files Jul 31 12:57:32 well, I could make them Jul 31 12:57:46 but I'm afraid then the the symbols would clash Jul 31 12:57:56 now, the so linking acts as a barrier that drops some symbols Jul 31 12:58:08 I donn't know then Jul 31 12:58:29 yeah, it sounds rather convoluted... I think it might be possible, but it sounds quite crazy Jul 31 12:58:54 the all issue is that android dyn. linker is... well, totally simplified Jul 31 12:59:08 so I want to work around that using static linking Jul 31 12:59:29 and I figured there might be some binutils/ld gurus around, so... I'm grateful for any comments Jul 31 13:25:11 XorA, btw, ASSUME_PROVIDED=virtual/kernel works in dev Jul 31 14:12:02 ~curse Jul 31 14:12:03 extra, extra, read all about it, swearing is Silence, you sock-clucking mother-trucker Jul 31 14:12:37 'bitbake xserver-xorg' is 2100 tasks ;( toolchain got updates Jul 31 14:14:28 hihi Jul 31 14:14:59 hrw but with words from koen we all have fast 8 core machine so it doesnt matter Jul 31 14:16:53 :) Jul 31 14:18:10 woglinde: I have root on 8core but use it for other builds Jul 31 14:18:21 crofton should I have put sarcastic marks around it? Jul 31 14:18:40 2051/2080 is not so bad Jul 31 14:24:09 heh Jul 31 14:24:28 I understand the headaches of building on not so hot hw Jul 31 14:24:46 occasionally I do stuff on my laptop Jul 31 14:42:20 Crofton: ;) Jul 31 14:42:28 Crofton: I plan to make a build on sheeva Jul 31 14:43:07 awesome :) Jul 31 14:43:20 I like people that run stuiff on machines it is not supposed to work on Jul 31 14:43:51 Crofton: angstrom should be self hosting for ~2 years Jul 31 14:44:05 Upgrading xserver-common on root from 1.30-r1.3 to 1.30-r2.3... Jul 31 14:44:11 no one has it D: Jul 31 14:53:27 morning all Jul 31 14:54:59 hi RP Jul 31 14:56:46 RP: do you know does libdrm_intel is used by non-Intel x86 chipsets? Jul 31 15:08:45 03Klaus Kurzmann  07fso/milestone5.5 * r865cca5af0 10openembedded.git/recipes/freesmartphone/libfsobasics_git.bb: Jul 31 15:08:45 libfsobasics: new recipe Jul 31 15:08:45 Signed-off-by: Klaus Kurzmann Jul 31 15:08:46 03Klaus Kurzmann  07fso/milestone5.5 * r0e27246a8f 10openembedded.git/recipes/libgee/libgee_0.1.6.bb: Jul 31 15:08:47 libgee_0.1.6.bb: add missing inherits for vala and autotools_stage Jul 31 15:08:49 Signed-off-by: Klaus Kurzmann Jul 31 15:08:51 03Klaus Kurzmann  07fso/milestone5.5 * r954b054ab5 10openembedded.git/ (conf/checksums.ini recipes/libgee/libgee_0.1.6.bb): Jul 31 15:08:54 libgee: add 0.1.6 Jul 31 15:08:56 Signed-off-by: Klaus Kurzmann Jul 31 15:08:58 03Klaus Kurzmann  07fso/milestone5.5 * rc67ccdec6c 10openembedded.git/recipes/freesmartphone/cornucopia.inc: Jul 31 15:09:01 cornucopia.inc: do not make libfsobasics depend on libfsoframework Jul 31 15:09:01 hrw: Its not afaik Jul 31 15:09:07 Signed-off-by: Klaus Kurzmann Jul 31 15:09:09 03Klaus Kurzmann  07fso/milestone5.5 * r2a20fc22b1 10openembedded.git/recipes/freesmartphone/libfsoframework_git.bb: Jul 31 15:09:12 libfsoframework: add dependency for libfsobasics Jul 31 15:09:14 Signed-off-by: Klaus Kurzmann Jul 31 15:09:16 03Didier 'Ptitjes  07fso/milestone5.5 * r34fdadfc7b 10openembedded.git/classes/girepository.bbclass: Jul 31 15:09:21 Add a girepository bbclass Jul 31 15:09:23 Signed-off-by: Didier 'Ptitjes Jul 31 15:09:25 03Klaus Kurzmann  07fso/milestone5.5 * rbf63dc9f54 10openembedded.git/recipes/freesmartphone/libfso-glib_git.bb: Jul 31 15:09:28 libfso-glib: switch to use SRCPV instead of SRCREV Jul 31 15:09:30 Signed-off-by: Klaus Kurzmann Jul 31 15:09:32 03Klaus Kurzmann  07fso/milestone5.5 * r531144b0c6 10openembedded.git/recipes/efl1/emotion_svn.bb: Jul 31 15:09:39 emotion: add --with-edje-cc to EXTRA_OECONF Jul 31 15:09:41 Signed-off-by: Klaus Kurzmann Jul 31 15:09:43 03Angus Ainslie  07fso/milestone5.5 * r7297196a42 10openembedded.git/conf/distro/include/preferred-om-2009-versions.inc: preferred-om-2009-versions : update vala versions Jul 31 15:09:46 03Angus Ainslie  07fso/milestone5.5 * r6f902ed2b4 10openembedded.git/conf/distro/include/moko-autorev.inc: moko-autorev : remove deprecated packages Jul 31 15:09:53 03Angus Ainslie  07fso/milestone5.5 * rf593fc552c 10openembedded.git/: Merge branch 'fso/milestone5.5' of git@git.openembedded.org:openembedded into fso/milestone5.5 Jul 31 15:09:56 03Klaus Kurzmann  07fso/milestone5.5 * rf3a61db2d7 10openembedded.git/ (conf/checksums.ini recipes/vala/vala-native_0.7.5+fso1.bb): Jul 31 15:09:59 vala-native: update to 0.7.5+fso2 Jul 31 15:10:03 Signed-off-by: Klaus Kurzmann Jul 31 15:12:02 morning Jul 31 15:12:09 hi kergoth Jul 31 15:24:08 re Jul 31 15:24:16 jo kergoth and rp Jul 31 15:24:55 * kergoth thinks about adding bbclasses to OpenEmbedded for putting recipes in project source trees Jul 31 15:26:21 * RP wonders whether an rsync which can look inside compressed files would be useful Jul 31 15:26:53 kergoth: So parsing would be from remote repos? Jul 31 15:27:23 naw, thats more trouble than its worth, which is why fetching from the parser went away :) Jul 31 15:27:40 kergoth: right, just wondering ;-) Jul 31 15:28:33 thinking more from the perspective of the user checking out their svn tree with a variety of pieces of software, each which has its own recipe, and then adding that checkout to their COLLECTIONS/BBFILES, and having it just work Jul 31 15:28:47 would make it much easier for the active development use case, in my opinion Jul 31 15:31:00 03Holger Hans Peter Freyther  07org.openembedded.dev * rfcc0f281ac 10openembedded.git/ (4 files in 2 dirs): Jul 31 15:31:00 libdbi-drivers: Upgrade to 0.8.3-1 with more db drivers Jul 31 15:31:00 Upgrade to the latest upstream release, coming with MySQL, Jul 31 15:31:00 postgres, sqlite3 and sqlite3 drivers. Jul 31 15:31:04 03Holger Hans Peter Freyther  07org.openembedded.dev * rf9ad231ef4 10openembedded.git/ (conf/checksums.ini recipes/libdbi/libdbi_0.8.2.bb): libdbi: Move from 0.8.2 to 0.8.3 which is the latest version Jul 31 15:31:25 * zecke_ spent some hours today cross compiling glib, expat, libxml2, gstreamer, gst-plugins-base by hand... Jul 31 15:31:54 eep Jul 31 15:32:18 zecke_ is a masochist? :P Jul 31 15:33:37 kergoth: i thought it was easier than figuring out how to get a git checkout in this corporate network Jul 31 15:33:46 * kergoth chuckles Jul 31 15:33:57 kergoth: I can see what you mean but I'd check developers would actually use it... Jul 31 15:34:15 kergoth: developers are funny ;-) Jul 31 15:34:19 RP: i know 6 kernel developers within 50 feet of me that would Jul 31 15:34:43 zecke_: ssh to a remote machine, bitbake -c fetchall, rsync Jul 31 15:35:13 zecke_: Poky does know how to do git through a coporate firewall too btw :) Jul 31 15:35:45 kergoth: You know about the setting S to a local directory hack, right? Jul 31 15:36:01 yes, but it still adds more time to the build/test cycle than they'd prefer Jul 31 15:36:16 RP: well, but for the initial checkout (unless you can use http there too) Jul 31 15:36:40 RP: I will be back on monday, and use their prebuilt toolchain to build some packages Jul 31 15:36:40 zecke_: There are ways of making git go through socks Jul 31 15:36:57 all I'm talking about is a class to make fetch/unpack/patch a noop and set S=${TOPDIR} and such, nothing to it Jul 31 15:37:08 maybe helpers for extracting version info from the scm Jul 31 15:37:19 kergoth: Probably be nice to have... Jul 31 15:39:12 RP: would a meta-qt something pokylinux would ever see? Jul 31 15:39:39 zecke_: If and only if someone stepped up to actively maintain it Jul 31 15:40:04 RP: is your buildbot sending hate mail? Jul 31 15:40:14 zecke_: To where? Jul 31 15:40:19 03Stefan Schmidt  07org.openembedded.dev * rd05d52c497 10openembedded.git/recipes/efl1/emotion_svn.bb: emotion_svn: Give path to edje_cc to fix build. Bump PR. Jul 31 15:40:32 RP: devs, when things break? Jul 31 15:40:47 zecke_: yes Jul 31 15:40:55 RP: e.g., or is actively maintaining == watching the bots? Jul 31 15:41:35 zecke_: I'd settle for someone using it and noticing when it breaks manually. I do have a builder than can send mail though Jul 31 15:42:01 (finally, long story) Jul 31 15:46:00 lumdidum Jul 31 16:33:07 for vmlinux.srec, who generates the srec file? Jul 31 16:33:27 is that a builtin kernel target, or a postprocessing step on the vmlinux image? Jul 31 16:33:36 have a nice rest of day Jul 31 16:45:39 03Holger Hans Peter Freyther  07org.openembedded.dev * r7cdcc9ad25 10openembedded.git/recipes/qt4/ (6 files): Jul 31 16:45:39 qt4: Qt4.5 is released under LGPLv2.1 and GPLv3 Jul 31 16:45:39 Update licensing bits on the Qt classes. Jul 31 16:50:06 re zecke Jul 31 16:51:32 thanks Jul 31 16:52:04 i know it is hard to provide a stable internet connection for 8 euro a night.. Jul 31 17:53:11 03Khem Raj  07org.openembedded.dev * r758a9ce142 10openembedded.git/ (3 files in 3 dirs): Jul 31 17:53:11 tinylogin, unzip: Update URLs. Jul 31 17:53:11 Signed-off-by: Khem Raj Jul 31 18:12:23 03Koen Kooi  07org.openembedded.dev * rc1cf6902c0 10openembedded.git/ (4 files in 2 dirs): enlightenment: bump SRCREV, fix upgrade paths Jul 31 18:12:34 03Koen Kooi  07org.openembedded.dev * r8e6e8ded5a 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.org:openembedded into org.openembedded.dev Jul 31 18:47:13 Hi! i have a problem building libtool-sdk (2.2.6a): the patch does not apply because the libtool-2.2.6a.tar.gz extracts to libtool-2.2.6. I was able to get around this by putting S = "${WORKDIR}/libtool-2.2.6" into libtool-sdk_2.2.6a.bb, but I feel this is just a dirty workaround Jul 31 18:51:55 03Koen Kooi  07org.openembedded.dev * r867c7be51f 10openembedded.git/recipes/tasks/task-beagleboard-demo.bb: task-beagleboard-demo: remove exhibit Jul 31 19:09:40 I'm using the Angstrom distribution on a BeagleBoard. How do I build 2.6.31 kernel with OE? No matter what I do, I get the 2.6.29-r40 kernel. Jul 31 19:20:03 no recipe yet Jul 31 19:20:21 not sure if omap kernel has caught up to .31 Jul 31 19:25:03 Crofton: Thanks Jul 31 19:26:44 Crofton: I don't see a recipe for .30 either Jul 31 19:28:57 yeah Jul 31 19:29:03 you could ask in #beagle Jul 31 19:29:34 Crofton: I'll do that Jul 31 19:34:49 hmm why bitbake would not checkout from git Jul 31 19:35:15 I am trying to fetch a git recipe and it always bails out after looking at the predefined tar locations Jul 31 19:37:01 I've made a g1 machine config and kernel(I say it before someone duplicate my work) Jul 31 19:37:05 I'll try it Jul 31 20:38:07 hmm woglinde this commit b296387c6227c3ab870277f77b2d30110475e00c broke uclibc_nptl and uclibc_nptl recipes Jul 31 20:38:21 should I revert it or I could also override it Jul 31 20:38:46 I would prefer to revert it Jul 31 20:39:25 re Jul 31 20:42:02 khem hm Jul 31 20:42:10 let me see if I can fix it Jul 31 20:45:52 khem I find it better if _svn and _nptl have there own SRC_URI Jul 31 20:46:52 kehm I morer worried about PROVIDES += "${@['virtual/libiconv', ''][bb.data.getVar('USE_NLS', d, 1) != 'yes']}" Jul 31 20:49:33 thats what Ididd Jul 31 20:49:52 I dont change your patch I override it Jul 31 20:50:03 ah okay Jul 31 20:50:20 the push it instead of me Jul 31 20:50:23 tehn Jul 31 20:50:55 khem I also would prefer to drop the older uclibc recipes Jul 31 20:51:07 only keep form .28 Jul 31 20:56:27 but we have a no delete policy dont we Jul 31 20:57:48 khem not really Jul 31 20:58:02 I would be happy to nuke few of them Jul 31 20:58:10 and I am wonder why in the world is using oe with uclibc 0.9.26 Jul 31 20:58:16 args Jul 31 20:58:21 s/why/who Jul 31 20:59:28 khem we should post a rfc and ack it Jul 31 20:59:36 ok Jul 31 20:59:41 and if noboy objects in 2 weeks Jul 31 20:59:45 we are doing it Jul 31 21:00:05 let me send an email then Jul 31 21:08:29 Gnutoo: btw. I found answer to my problem. Basically "ld -r" and "objcopy --localize-hidden" did the trick. :) Jul 31 21:08:57 ok Jul 31 21:09:55 I'll go bye Jul 31 21:10:28 bye gnutoo Jul 31 21:10:38 bye Jul 31 21:10:45 bye Jul 31 21:42:20 hm Jul 31 21:42:33 why always me finding some nifty typos in manuals Jul 31 21:43:27 textual or command examples? Jul 31 21:43:39 hm logical Jul 31 21:43:46 its in beaglboard hw manual Jul 31 21:44:04 in diffs btw the revisions Jul 31 21:44:31 rev 5 to 6 claims now OMAP3530 ES3.0 is used instead of 3.0 Jul 31 21:45:01 then rev 6 to 7 says OMAP3530 Rev ES3.0 silicon versus the ES2.1 Jul 31 21:45:14 Rev B6 board Jul 31 21:46:05 i wonder if it's clear in the C3 manual? Jul 31 21:46:17 or if they don't bother with that kind of thing Jul 31 21:46:18 its the c3 manual Jul 31 21:46:22 ah Jul 31 21:47:12 I am looking up Jul 31 21:47:16 whats needed for hd video Jul 31 21:58:35 03Cliff Brake  07org.openembedded.dev * r3e377a05c6 10openembedded.git/ (9 files in 3 dirs): mono: add 2.4.2.3 recipes Jul 31 21:58:46 03Cliff Brake  07org.openembedded.dev * rabe7a7078f 10openembedded.git/: Merge branch 'mono-dev2-prepare-to-commit' into org.openembedded.dev Jul 31 22:01:45 re kergoth Jul 31 22:30:46 03Cliff Brake  07org.openembedded.dev * re60d8a6b70 10openembedded.git/conf/checksums.ini: Jul 31 22:30:46 checksums.ini: restore entries that got deleted in last commit Jul 31 22:30:46 I should be more careful when I copy files in tarballs **** ENDING LOGGING AT Sat Aug 01 02:59:58 2009