**** BEGIN LOGGING AT Sat Sep 12 02:59:56 2009 Sep 12 03:22:10 <_whitefox> Crofton: are you there? Sep 12 05:07:33 03Martin Jansa  07shr/import * r7494c539a1 10openembedded.git/recipes/intone/intone_svn.bb: Intone: Partially revert changes as they were applied upstream for rev > 64, make regexp for /usr/doc/intone -> /usr/share/doc/intone a bit safer Sep 12 06:35:49 03Khem Raj  07org.openembedded.dev * r53c0673ba0 10openembedded.git/ (6 files in 3 dirs): (log message trimmed) Sep 12 06:35:49 powertop: Make it compile on uclibc. Sep 12 06:35:49 * Use INC_PR Sep 12 06:35:49 * Move common stuff to powertop.inc Sep 12 06:35:49 * Add -lint for uclibc bases systems Sep 12 06:35:51 * Add a patch to tinker with order of CFLAGS Sep 12 06:35:53 in commandline Sep 12 08:10:25 03Steve Sakoman  07org.openembedded.dev * r23c18144b1 10openembedded.git/recipes/ebtables/ebtables_2.0.6.bb: ebtables: fix GNU_HASH error Sep 12 09:36:43 Hi, i'm trying to compile blueman (blueman-project.org) for the OpenMoko FreeRunner on a x86-64 machine, but while configuring i'm getting this error: shr-unstable/tmp/staging/x86_64-linux/usr/include/python2.6/pyport.h:685:2: error: #error "LONG_BIT definition appears wrong for platform (bad gcc/glibc config?). Sep 12 09:39:07 I found a bugreport for RedHat where is mentioned that one should split pyconfig.h in a 32 and 64 bit version: https://bugzilla.redhat.com/show_bug.cgi?id=139911 Sep 12 09:39:27 for libxml2 this is solved in OE with the option --without-python Sep 12 09:39:37 has someone a hint, what i can do? Sep 12 10:33:07 hmmm what happened to the beagleboard channel? Sep 12 10:33:46 julemore: #beagle. Sep 12 10:33:53 oooh true Sep 12 11:19:05 omg... I wanted to use oe for cross-compiling.. i am not making any image... but when i write bitbake gcc-cross i get that error Sep 12 11:19:21 FATAL: kernel too old Sep 12 11:19:33 and I dot have in the HOST 2.6.29 Sep 12 11:19:37 whats up? Sep 12 11:25:36 good day Sep 12 11:34:21 morning jest Sep 12 11:34:34 ~seen florian Sep 12 11:34:37 florian was last seen on IRC in channel #oe, 11h 38m 25s ago, saying: 'good night'. Sep 12 11:38:10 morning rsalveti Sep 12 11:38:12 <_whitefox> Hi, I am trying to make openbts, but I am getting No rule to make target 'install' Sep 12 11:38:21 <_whitefox> how can I fix this kind of make problem? Sep 12 11:38:50 _whitefox: probably the src directory is different from where bitbake expects it to be Sep 12 11:39:19 _whitefox: is this your own recipe ? Sep 12 12:17:15 has anyone used boost-asio? Sep 12 12:51:40 <_whitefox> rkirti: I want to config the {S} Sep 12 12:51:44 <_whitefox> but I am not sure how to Sep 12 13:22:00 hi Sep 12 13:59:27 03Koen Kooi  07org.openembedded.dev * rc49513ce80 10openembedded.git/recipes/cairo/cairo_git.bb: cairo git: update SRCREV Sep 12 13:59:28 03Koen Kooi  07org.openembedded.dev * r6efa4799a8 10openembedded.git/ (6 files in 3 dirs): gtk+: add 2.16.6 Sep 12 13:59:29 03Koen Kooi  07org.openembedded.dev * r18158e3377 10openembedded.git/recipes/glibc/glibc-package.bbclass: Sep 12 13:59:29 glibc-package bbclass: add hack to avoid "kernel too old" error when generating locales in angstrom Sep 12 13:59:30 The proper fix would be to patch linux-libc-headers with Sep 12 13:59:32 0001-implement-TIF_RESTORE_SIGMASK-support-and-enable-the.patch and all arm Sep 12 13:59:34 kernels as well. Since that's quite an undertaking angstrom needs to short term Sep 12 13:59:36 hack like this commit to keep things building while the proper fix is being Sep 12 13:59:38 worked on. Sep 12 13:59:40 03Koen Kooi  07org.openembedded.dev * rbd4726dac9 10openembedded.git/conf/distro/include/angstrom-2008-preferred-versions.inc: angstrom: prefer gtk+ 2.16.6 Sep 12 14:01:52 03Koen Kooi  07org.openembedded.dev * reabc33a1f1 10openembedded.git/recipes/clutter/clutter.inc: clutter.inc: rework armv7a machines a bit and add omap3517-evm to compatible machines Sep 12 15:29:44 03Michael 'Mickey' Lauer  07org.openembedded.dev * rca6953db99 10openembedded.git/ (3 files in 3 dirs): libgee: move to git version while it's so much under development and depending on vala-native Sep 12 15:29:45 03Antonio Ospite  07org.openembedded.dev * rd34d1145f5 10openembedded.git/files/device_table-ezx.txt: Sep 12 15:29:45 device_table-ezx: Add static video device node. Sep 12 15:29:46 Useful if we compile the video driver statically in the kernel Sep 12 15:29:48 Signed-off-by: Antonio Ospite Sep 12 15:29:50 03Antonio Ospite  07org.openembedded.dev * r09d4daf8af 10openembedded.git/files/device_table-ezx.txt: Sep 12 15:29:53 device_table-ezx: Static mux nodes from mux1 to mux8. Sep 12 15:29:55 Userspace is not supposed to open mux0 anymore; now that kernel mapping Sep 12 15:29:57 has been removed, each muxX corresponds to dlci X (with X in [1,16]). Sep 12 15:29:59 Signed-off-by: Antonio Ospite Sep 12 15:30:01 03Antonio Ospite  07org.openembedded.dev * r90a1a0ca36 10openembedded.git/files/device_table-ezx.txt: Sep 12 15:30:04 device_table-ezx: Add header Sep 12 15:30:06 Signed-off-by: Antonio Ospite Sep 12 15:30:08 03Antonio Ospite  07org.openembedded.dev * r4e8143441f 10openembedded.git/files/device_table-ezx.txt: Sep 12 15:30:29 device_table-ezx: Add static ttyIPC device node Sep 12 15:30:29 This device node is used by ts0710_mux_usb driver to expose the muxed Sep 12 15:30:29 communication with the BP and configure the line discipiline Sep 12 15:30:29 Signed-off-by: Antonio Ospite Sep 12 15:38:24 03Michael 'Mickey' Lauer  07org.openembedded.dev * r0c4cf4dcde 10openembedded.git/recipes/freesmartphone/fso-alsa-data/default/dummy: fso-alsa-data: create dummy file so that git doesn't remove the directory Sep 12 15:45:40 03Koen Kooi  07stable/2009 * r583297ac29 10openembedded.git/recipes/quake/quake2_svn.bb: Sep 12 15:45:40 quake2: fix QA errors Sep 12 15:45:40 Signed-off-by: Koen Kooi Sep 12 15:45:40 Acked-by: Philip Balister Sep 12 15:45:51 03Koen Kooi  07stable/2009 * rbb74c9fb7a 10openembedded.git/recipes/ogre/freeimage_3.11.0.bb: Sep 12 15:45:51 freeimage: fix QA errors Sep 12 15:45:51 Signed-off-by: Koen Kooi Sep 12 15:45:51 Acked-by: Philip Balister Sep 12 16:36:27 re Sep 12 16:36:34 wb Sep 12 16:37:06 he rkirti whats with your write access? Sep 12 16:37:43 havent asked as yet. I cleaned up bits this weekend...will mail Florian today Sep 12 16:39:32 :) Sep 12 16:39:36 hi Sep 12 16:39:37 hire mrmoku Sep 12 16:39:41 hi ao2 Sep 12 16:39:46 hi all Sep 12 16:40:01 has anybody used boost-asio? Sep 12 16:40:26 woglinde: I have a query .... after the last git pull the ABI version changed, and I couldnt get gcc-cross to build. ...some compiler options seem unrecognised Sep 12 16:43:18 jest hm no, but we lack some versions behind Sep 12 16:43:27 rkirti hm Sep 12 16:43:38 rkirti did you clean the oetmp dir? Sep 12 16:43:44 I think makedevs is not interpreting right the "count" field in device_table, now it thinks the count value meand "count from 0" even if we have a !=0 value for the start field. Shouln't it be doing like this: http://pastebin.ca/1563293 ? It should take the count field as "count from 'start' minor number" Sep 12 16:43:52 mickeyl, ^ Sep 12 16:43:52 yep. its a fresh build from scratch Sep 12 16:43:53 I guess... I had problems with using it with boost::bind Sep 12 16:44:07 rkirti hm best write the error to ml Sep 12 16:44:15 woglinde: is there a reason to keep two packages boost-asio and asio? Sep 12 16:44:22 jest dont know Sep 12 16:45:00 OK, I'll try upgrade the recipe for the newer version Sep 12 16:45:09 jest boost? Sep 12 16:45:12 good lick Sep 12 16:45:15 aeh luck Sep 12 16:45:23 boost/jam building suckz a lot Sep 12 16:45:28 for cross compiling Sep 12 16:45:49 yeah, I my eyes opened wide when I save the recipe... Sep 12 16:54:10 woglinde: just fyi : http://pastebin.com/d581bccf6 Sep 12 16:54:55 I tried decreasing PR...r1 is what used to be built for 4.3.3 earlier. with r6.1/6.2 it broke. bit decrementing pr doesnt help (besides the fact that its wrong) Sep 12 16:57:43 rkirti hm Sep 12 16:57:50 the settings comes from gcc inside Sep 12 17:00:02 rkirti hm could please search configure or compile log Sep 12 17:00:10 for checking whether gcc supports -pedantic -Wno-long-long -Wno-variadic- macros -Wno-overlength-strings. Sep 12 17:00:19 ok Sep 12 17:02:14 woglinde: it does Sep 12 17:02:53 (its the ccache gcc thats mentioned in the logs. How are they different ?) Sep 12 17:05:44 rkirti hm could you try without ccache for now? Sep 12 17:06:13 woglinde: how do I do that ? Sep 12 17:06:30 *g* Sep 12 17:06:33 deinstall it Sep 12 17:06:45 or move the exec to ccache.bak Sep 12 17:08:30 ah..that way. Sep 12 17:08:42 I thought it was some bitbake option to disable it Sep 12 17:08:59 hm there is one but cant remeber Sep 12 17:09:40 ok, do I need to rebuild from scratch now that I have moved the binary or just continue ? Sep 12 17:09:57 nope Sep 12 17:09:59 hi Sep 12 17:10:07 clean gcc-cross-initial Sep 12 17:10:14 and gcc-cross-intermediate Sep 12 17:10:20 should be enough Sep 12 17:10:42 ok Sep 12 17:11:16 hi jay7 Sep 12 17:11:25 hey woglinde Sep 12 17:15:16 question: I need to force kdrive to compile the sis300 server. In xserver-kdrive-common.inc teh packages are defined in PACKAGES =+ . Can I add some package in this variable but without touch the inc file (maybe in local.conf)? Sep 12 17:15:37 sushisan hm no Sep 12 17:16:10 mmm :-( Sep 12 17:16:13 txs Sep 12 17:23:07 03Klaus Kurzmann  07shr/import * r35befdd560 10openembedded.git/conf/distro/include/ (shr-autorev-unstable.inc shr-autorev.inc): Sep 12 17:23:07 shr-autorev(-unstable): set pisi, webkit-efl and eve to AUTOREV Sep 12 17:23:07 Signed-off-by: Klaus Kurzmann Sep 12 17:23:08 03Klaus Kurzmann  07shr/import * r10929d330b 10openembedded.git/recipes/webkit/webkit-efl_git.bb: Sep 12 17:23:10 webkit-efl: use SRCPV and switch to current master branch Sep 12 17:23:12 Signed-off-by: Klaus Kurzmann Sep 12 18:49:54 woglinde: the patch yesterday worked ok for cross gcc but failed for native gcc on micro :( I am now testing this one http://sakrah.homelinux.org/Public/diff Sep 12 18:50:28 woglinde: this patch avoids the muck if we are not building a cross compiler Sep 12 19:34:00 re Sep 12 19:34:24 khem hm Sep 12 19:35:55 woglinde: this works fine though for target gcc as well as cross gcc now Sep 12 19:36:12 I think I will commit this one Sep 12 19:36:27 jepp Sep 12 19:36:33 looks good Sep 12 19:36:56 woglinde: this could be done better but our gcc build is complex and we have so many restrictions in place Sep 12 19:37:14 by nature of the usage it can not be simplified Sep 12 19:37:16 khem did you see the error from rkirti? Sep 12 19:37:25 woglinde: no Sep 12 19:37:35 http://pastebin.com/d581bccf6 Sep 12 19:37:38 * khem looks into irc logs Sep 12 19:37:50 thx Sep 12 19:38:19 what host is it on? Sep 12 19:38:45 the failure is because the gcc on the host system does not understand the option Sep 12 19:39:09 I would not recommend using gcc 3.x to compile gcc 4.3+ Sep 12 19:40:38 hmmm there is no information on what the host gcc version is so I would assume its old enough to not understand the option which means its not good enough to compile gcc 4.3 Sep 12 19:42:54 this option was added somewhere around gcc 4.x release time frame look here http://gcc.gnu.org/ml/gcc-patches/2006-01/msg01920.html Sep 12 19:44:55 I would say gcc 4.1 should be minimal to compile gcc 4.3.x Sep 12 19:49:26 khem he said it compiles before your updates Sep 12 19:49:46 hm its intermediate Sep 12 19:49:52 so inital was build Sep 12 19:49:57 and is now used Sep 12 19:50:46 woglinde: it does not do fixinclude stuff with initial Sep 12 19:52:06 woglinde: and it uses the native gcc to build a tool which does fixinclude stuff and the build is failing during building this tool Sep 12 20:02:16 hm Sep 12 20:03:49 it could be that previous compile was of gcc 4.2 Sep 12 20:04:08 angstrom changed the default compiler after that compile Sep 12 20:04:11 its possible Sep 12 20:06:03 please excuse me a lame question, but what is the best practice to modify other's recipe by changing its variable from the outside? More specifically, I want to append some text to it. Sep 12 20:07:54 like what Sep 12 20:08:04 recipes consist of tasks Sep 12 20:08:09 do you want to modify a task ? Sep 12 20:10:25 no, I want to append text to recipes variable. But I'd prefer not to fork my own version of recipe. Sep 12 20:12:07 jest: I dont know of a way of doing such a thing Sep 12 20:12:18 What exactly I want is to modify BOOST_LIBS and append "asio" to it in boost recipe. Sep 12 20:13:16 you need to modify recipe Sep 12 20:13:43 thanks Sep 12 20:14:07 you could just do BOOST_LIBS += "asio" Sep 12 20:15:12 sure I can, but the question is _where_ to do it? :) If I do it in e.g. local.conf, it will be later overwritten by the recipe (I guess...) Sep 12 20:21:40 in recipe itself Sep 12 20:21:50 why dont you want to modify the recip Sep 12 20:21:53 recipe Sep 12 20:26:02 In this aspect (selection of libraries of Boost being compiled) I believe it should be a part of OE configuration, which should be done outside of recipe. Sep 12 20:28:18 recipe defines the rules and environment to build packages out of it and there are some common stuff that is abstacted into classes that a recipe can inherit and also bitbake provides environment for some common variables Sep 12 20:28:39 BOOST_LIBS seems to be specific to boost which means the recipe is the right place to define it Sep 12 20:31:06 I think the choice of Boost libs is the compromise between functionality and size of the library. But it might be building problems with some parts and disablement of them. I'm checking. Sep 12 20:31:48 thats done be recipes Sep 12 20:44:33 good morning Sep 12 20:44:40 hi florian Sep 12 23:19:27 03Khem Raj  07org.openembedded.dev * r56ca890cdd 10openembedded.git/recipes/gcc/gcc-configure-common.inc: Sep 12 23:19:27 gcc-configure-common.inc: Fiddle with NATIVE_SYSTEM_HEADER_DIR only form cross gcc build Sep 12 23:19:27 * target gcc does not like the NATIVE_SYSTEM_HEADER_DIR being tweaked. Sep 12 23:19:27 for now conditionalize it so it gets included only for cross builds Sep 12 23:19:27 it could be that target gcc for micro wont work because it needs headers Sep 12 23:19:30 in /usr/include and the actual headers are in /include Sep 12 23:19:32 Signed-off-by: Khem Raj Sep 12 23:23:14 03Khem Raj  07org.openembedded.dev * r462d0b6fb1 10openembedded.git/recipes/uclibc/ (6 files): Sep 12 23:23:14 uclibc-initial: Create ${UCLIBC_STAGE_PREFIX}/lib Sep 12 23:23:14 Signed-off-by: Khem Raj Sep 12 23:59:25 whew, think i'm finally happy with the conversion to logging / deprecation of bb.msg... now to revamp the semantics of -l/-D/-v Sep 13 00:01:55 good nite Sep 13 00:36:40 hrm Sep 13 01:21:31 * kergoth tries to figure out exactly what the usability issues are with bitbake & OE Sep 13 01:42:10 hmm, okay, have a few ideas to improve usability Sep 13 02:43:24 hmm **** ENDING LOGGING AT Sun Sep 13 02:59:56 2009