**** BEGIN LOGGING AT Fri Mar 06 02:59:57 2009 Mar 06 03:37:16 gm Mar 06 03:38:17 oommmmmmmm Mar 06 05:28:35 03Angus Ainslie  07fso/milestone5.5 * r79518be4f1 10openembedded.git/conf/distro/include/sane-srcrevs.inc: Bump the efl version for elm binding fixes Mar 06 06:45:39 03Koen Kooi  07org.openembedded.dev * rcdd27f17b7 10openembedded.git/packages/linux/linux.inc: linux.inc: add back 'yes '' | oe_runmake oldconfig' that got dropped Mar 06 06:50:53 hmm, I'm trying to build the package neon on two computers.. On the first it fails when checking config.log and on the other it works perfectly. When I diff the config.log files it diffs on, among other things, that the failing one uses "/usr/bin/install" while the working one uses the install from "tmp/staging/i686-linux".. any ideas on what could cause one build environment to use the host binaries? Mar 06 07:15:55 erbo: are the local.conf files same on both computers ? Mar 06 07:19:51 khem: no, something special to look out for? Mar 06 07:23:38 one is for a beagleboard and the other builds for an armv6 cpu. The local.conf for the failing one just sets distro=angstrom-2008.1, machine to a home made machine and enable_binary_locale_generation = 1 Mar 06 07:24:08 (the armv6 which is a home made machine if the failing one) Mar 06 07:27:48 erbo: hmm are you using assume provided Mar 06 07:30:51 no, not that I know of.. Nothing in the machine conf or local.conf Mar 06 07:34:44 erbo: do you have coreutils-native installed on both machines staging area Mar 06 07:39:03 is there a good way of checking whats installed in staging? I coreutils-native in my tmp/work/i686-linux/ but how do I check the staged version? Mar 06 07:39:43 iow do you have install program installed in staging Mar 06 07:42:43 hmm, on the buildenv that works there is no install .. but the failing one has an install Mar 06 07:43:49 can I see which version of coreutils-native that is in staging? to see if they differ Mar 06 07:46:03 morning Mar 06 07:46:31 I think I might have coreutils-native 5.3-r4 on the working one and 6.0-r3 on the failing one Mar 06 07:46:40 morning Mar 06 07:46:41 erbo: just try bitbake -c rebuild coreutils-native Mar 06 07:46:54 on failing machine Mar 06 07:47:02 and also on working one Mar 06 07:47:18 btw. Do you have all the updates Mar 06 07:47:23 for OE Mar 06 07:48:54 I updated them both yesterday so they're at the same status Mar 06 07:49:32 I'm rebuiling coreutils-native on both machines now.. Mar 06 07:50:06 and I really appreciate your help Mar 06 07:59:33 khem: when I rebuilt coreutils-native they both built 6.0-r3 and now they both have install in staging but one still build neon while the other fail on qa_conf.. Bitbake says "This autoconf log indicates errors, it looked at host includes." Mar 06 08:00:12 what does it look for? I'll try to grep -i '\-i/usr/include' on both and see if it differs Mar 06 08:05:29 I get some hits on both machines, lines where is uses the cross-compiler but adds both the staging and host include dir Mar 06 08:05:34 m,orning Mar 06 08:05:39 morning Mar 06 08:23:24 erbo: can you rebuild neon same way Mar 06 08:23:36 bitbake -c rebuild neon Mar 06 08:31:49 I did that, the one that worked before still works and the other fail Mar 06 08:36:21 hmm strange Mar 06 08:37:02 r they same running same os Mar 06 08:38:23 yes, both are running arch linux Mar 06 08:38:44 I've got to run away for a few min, i'll be back soon Mar 06 08:53:30 back Mar 06 08:56:11 erbo: look in config.log file Mar 06 08:56:23 search for Badness Mar 06 09:03:33 morning Mar 06 09:08:28 khem: I get four of these -> "CROSS COMPILE Badness: /usr/include in INCLUDEPATH: /usr/include" on the failing machine Mar 06 09:12:34 the line after each Badness says "cc1: internal compiler error: in add_path, at c-incpath.c:371" Mar 06 09:13:26 erbo: first of all you should post whole build log on pastebin.ca or similar site Mar 06 09:13:37 will do Mar 06 09:14:07 it looks like neon took some headers from your system. so it probably lack some deps Mar 06 09:14:09 hi XorA Mar 06 09:14:48 hey hrw Mar 06 09:15:13 bb in 45m Mar 06 09:19:18 the log is at: http://www.unixshell.se/stuff/config.log Mar 06 09:21:44 Hi, how do we update to midori to 0.1.4? http://cgit.openembedded.net/cgit.cgi?url=openembedded/tree/packages/gtk-webcore Mar 06 09:24:48 erbo: can you post good config.log too Mar 06 09:24:58 -I/usr/include is the problem Mar 06 09:25:17 how does that get added to compiler invocation Mar 06 09:30:59 http://www.unixshell.se/stuff/config_good.log Mar 06 09:36:11 khem: I'm not sure how it's added, I still have a lot to learn about the inner workings of OE Mar 06 09:41:55 TeLLuS: submit a patch Mar 06 09:43:00 erbo: now you can try to run one of those commands manually Mar 06 09:44:53 erbo: in bad case can you see if you have gssapi.h in you target staging area Mar 06 09:49:33 hi all, I refreshed my openembedded directory (via git pull), and now I cannot compile the omap2430sdp. The compilation process stops during kernel compilation. May you tell my any suggestion? Mar 06 09:50:57 can you show us more errors Mar 06 09:54:55 just a moment... Mar 06 09:55:03 khem: in a devshell or regular shell? Mar 06 09:56:26 where the heck is the arm-angstrom-xxxx-toolchain hidden? :) Mar 06 09:56:34 XorA: Still trying to build the first one. There is also a midori_git.bb, could that be used or just bump the version on 0.1.2? cp midori_0.1.2.bb midori_0.1.4.bb What else is needed? Bump version in branches fso/milestone5.5 and fso/milestone5? I have not done this before.. Mar 06 09:58:01 found it Mar 06 09:58:31 strange... now the compilation seems to work... this morning I again updated oe. Mar 06 09:59:42 NOTE: package linux-omap2-2.6.26-r64: task do_configure: completed.NOTE: package linux-omap2-2.6.26: completed Mar 06 09:59:42 Yesterday is was not working Mar 06 09:59:52 hmm, _git.bb have the version hardcoded. 0.1.2 should be ok, I'l try that Mar 06 10:02:03 morning all Mar 06 10:02:18 erbo: in regular shell Mar 06 10:02:18 What is the difference between Poky Linux and Angstrom? Mar 06 10:03:09 khem: is conftest.c some special file? Can't find any in the work/armxx/neon dir Mar 06 10:03:31 erbo make a file say foo.c Mar 06 10:03:40 and add the following Mar 06 10:04:25 #include Mar 06 10:04:32 thats all Mar 06 10:04:49 then compile it with bad case compiler Mar 06 10:05:10 arm-angstrom-linux-gnueabi-gcc -E -isystem/home/erik/wip-svn/xxxx_build/tmp/staging/armv6-angstrom-linux-gnueabi/usr/include -D_LARGEFILE64_SOURCE -DNE_LFS -I/usr/include -I/usr/include Mar 06 10:09:29 khem: http://pastebin.com/d2ac27227 <-- output Mar 06 10:10:02 ok so thats the problem Mar 06 10:10:33 now try again and get rid of the -I/usr/include -I/usr/include Mar 06 10:10:41 compile again Mar 06 10:14:39 http://pastebin.com/m712cf4aa Mar 06 10:14:58 seems like gssapi.h isn't available in staging Mar 06 10:17:07 yea Mar 06 10:17:33 erbo: now compile same with good compiler Mar 06 10:17:48 just did, works when I have -I/usr/include Mar 06 10:18:06 and if I remove those it fails on not finding the .h file Mar 06 10:18:17 and without ? Mar 06 10:19:10 hmm I see you probably are using codesourcery compiler for armv7 Mar 06 10:19:59 That seems strange, I used OE to build my entire cross-comp environment Mar 06 10:20:40 beagleboard profile on the working one (which is the only armv7 of the two I'm using) Mar 06 10:21:01 I mean the sources for gcc Mar 06 10:22:04 my guess is that the patch we have to detect include poison is missing from armv7 compiler Mar 06 10:22:26 and hence it pulls in gssapi.h from host but does not complain Mar 06 10:23:14 what is your gcc version for armv7 case ? Mar 06 10:24:10 ./arm-angstrom-linux-gnueabi-gcc --version Mar 06 10:24:10 arm-angstrom-linux-gnueabi-gcc (GCC) 4.3.1 Mar 06 10:25:07 ok Mar 06 10:25:29 can you check cross-gcc build Mar 06 10:25:45 and see if zecke-no-host-includes.patch was applied Mar 06 10:29:46 it's there in the patches dir Mar 06 10:30:20 what does quilt applied say Mar 06 10:31:54 I checked the source and the patch is applied Mar 06 10:32:21 ah, quilt applied confirms it Mar 06 10:32:27 much easier though :9 Mar 06 10:33:08 interesting what did you get with armv7 compiler when you removed -I/usr/include in above cmdline Mar 06 10:34:36 http://pastebin.com/m12e483e6 Mar 06 10:34:39 TeLLuS: leave fso branches alone, FSO team will mess with them Mar 06 10:35:52 erbo: and no error when you add -I/usr/include ? Mar 06 10:36:20 that would mean that patch does not work as intended on this compiler Mar 06 10:39:32 http://www.unixshell.se/stuff/foo.log Mar 06 10:39:38 a hell of a lot output :) Mar 06 10:40:28 but no warning about Badness Mar 06 10:40:47 yeah the patch is not working with v7 compiler Mar 06 10:41:08 your problem is still there Mar 06 10:41:38 one way to avoid would be get rid of krb5-config stuff from your host Mar 06 10:43:44 someone should fix neon to cross compile properly Mar 06 10:43:49 time to sleep now Mar 06 10:43:54 good night Mar 06 10:44:30 good night, and thanks for helping me! Mar 06 10:44:35 it's lunch time for me :) Mar 06 10:53:02 re Mar 06 11:28:38 hrw: morning Mar 06 11:29:06 hrw: do you have that special defconfigs for Z offhand (700kb)? Mar 06 11:34:34 ant_work: moment Mar 06 11:38:28 http://marcin.juszkiewicz.com.pl/download/zaurus/ Mar 06 11:38:34 ant_work: read README Mar 06 11:41:06 thx Mar 06 12:09:02 morning all Mar 06 12:11:15 <_boto> hi Mar 06 12:25:47 03Koen Kooi  07org.openembedded.dev * r1e8947ddee 10openembedded.git/packages/images/distcc-slave-image.bb: distcc-slave-image: add wirelesstools and iptables Mar 06 12:25:55 03Koen Kooi  07org.openembedded.dev * r3355738d8c 10openembedded.git/ (3 files in 2 dirs): gnash: add 0.8.5 Mar 06 12:26:24 03Koen Kooi  07org.openembedded.dev * r87f1b5d8c0 10openembedded.git/packages/linux/ (linux-omap-pm/beagleboard/defconfig linux-omap-pm_git.bb): linux-omap-pm: enable iptables for beagle Mar 06 12:37:10 hello everybody Mar 06 12:39:51 some OE culture: to backup a situation I need: OE git hash, bitbake svn version, tmp/download . That's all? Mar 06 12:40:26 someone told me that there are images that are no nostamp, but I've not understood what does it mean Mar 06 12:47:26 some OE culture: to backup a situation I need: OE git hash, bitbake svn version, tmp/download . That's all? Mar 06 12:52:40 hi Mar 06 12:53:13 bitbake console-image Mar 06 12:53:23 ............. Mar 06 12:53:27 File "/home/aaron/oe/bitbake/lib/bb/persist_data.py", line 70, in delDomain Mar 06 12:53:27 self.connection.execute("DROP TABLE IF EXISTS %s;" % domain) Mar 06 12:53:27 sqlite3.OperationalError: attempt to write a readonly database Mar 06 12:53:38 what wrong with me Mar 06 12:53:56 recalcati_oe: ? Mar 06 12:54:11 aaron: old python-sqlite? Mar 06 13:17:27 hi mckoan ! Mar 06 13:17:32 ciao Mar 06 13:17:42 back in italy ? Mar 06 13:18:06 yes at last! :-) Mar 06 13:19:05 how was ? Mar 06 13:19:09 linux world Mar 06 13:19:25 gremlin[it]: Embedded World 2009 Mar 06 13:19:44 yes :) Mar 06 13:21:14 seems Italy and Germany are most OE-involved countries :) Mar 06 13:22:11 Jay7: yep ;-) Mar 06 13:22:48 * gremlin[it] is at now the first OE italian member :) Mar 06 13:26:26 i have a break in the process of bitbake console-image last time.but when i do bitbake console-image again . i meets some errors ,how to do now thanks Mar 06 13:27:18 any body would help me ,thank u so much Mar 06 13:33:25 aaron: which distro you use on your computer? Mar 06 13:33:32 ubuntu 7.x? fedora 8? Mar 06 13:38:34 good morning Mar 06 13:38:57 mpoullet: hi Mar 06 13:39:03 mpoullet: you work for AVM? Mar 06 13:39:59 does bitbake -c clean does a little bit more cleaning than bitbake -c rebuild? Mar 06 13:40:06 hrw: yep Mar 06 13:40:52 mpoullet: you use OE for Fritz devices maybe? Mar 06 13:41:19 hi mickeyl Mar 06 13:41:28 morning hrw Mar 06 13:42:02 hrw: maybe ;-) Mar 06 13:42:13 mpoullet: I am asking because I am considering buying Sinus W500V for home voip Mar 06 13:42:53 the cheapest avm device now (used) Mar 06 13:46:42 ok. Mar 06 13:46:47 * hrw -> food Mar 06 13:46:56 ~bon appetit Mar 06 13:46:57 bon appetit is, like, smacznego. Guten Appetit. Eet Smakelijk. God Appetitt. Buon Appetito. Buen apetito Bom Apetite. buen apetito Smaklig måltid!. Hyvää ruokahalua. Bo Proveito Mar 06 13:51:12 hi hrw Mar 06 13:51:15 online Mar 06 13:51:17 ? Mar 06 13:51:31 ubuntu 8.04 Mar 06 13:52:58 hi aaron, what's your problem? Mar 06 13:53:04 i have a break in the process of bitbake console-image last time.but when i do bitbake console-image again . i meets some errors ,how to do now Mar 06 13:53:50 do you have parallel builds? Mar 06 13:57:41 i don't know parallel builds? Mar 06 14:00:05 http://paste.pchowtos.co.uk/?id=1535&key=e4cb9c18142ef Mar 06 14:00:26 look it for me Mar 06 14:00:46 methril|work: look it for me Mar 06 14:01:12 ok, i'll see Mar 06 14:01:48 remove in your download dir the linux-omap dir Mar 06 14:02:08 aaron: try bitbake -c clean linux-omap_2.6.28 Mar 06 14:02:22 aaron: then rebuild your console image. Mar 06 14:06:13 methril|work: are parallel builds often causing problems? Mar 06 14:06:26 sometimes Mar 06 14:07:01 if you break a compilation, with ^C, sometimes some dependencies get in unstable state Mar 06 14:07:34 o.k. thnx Mar 06 14:15:04 ~seen rschuster Mar 06 14:15:38 rschuster was last seen on IRC in channel #oe, 34d 22h 8m 5s ago, saying: 'mckoan: not sure. IMO it would be better if any of the OE developers is hired by such a company for a specific job'. Mar 06 14:31:05 hi pleemans Mar 06 14:31:31 Hi hrw, how are you? Mar 06 14:32:23 remove in your download dir the linux-omap dir Mar 06 14:32:34 where i can find the dir Mar 06 14:32:42 it's in your settings Mar 06 14:33:05 usually {$OEDIR}/downloads Mar 06 14:33:44 pleemans: good, fighting with patch queqes Mar 06 14:34:35 ur means tmp/downloads ? Mar 06 14:34:45 nice, i'm fighting with perl/otrs atm. Mar 06 14:35:53 aaron: if it's your config, yes (usually is in this location) Mar 06 14:42:51 where i can find the bitbake's IRC Mar 06 14:47:17 aaron: nowhere Mar 06 14:48:08 aaron: is your OE the latest? Mar 06 14:53:54 i have never say the so much stupid tool like oe Mar 06 14:54:38 just once break, could not runs next time Mar 06 14:55:23 i have build for 4 hours ,just break once time Mar 06 14:55:47 2nd time run bitbake .meets errors Mar 06 14:55:59 fuck Mar 06 14:56:08 it got me mad Mar 06 14:56:21 aaron: I'm trying to help you and you just ignore me???!!?!?! Mar 06 14:57:04 http://paste.pchowtos.co.uk/?id=1537&key=a9eb250410252 Mar 06 14:57:13 thank u so much Mar 06 14:57:32 but the problem got me mad Mar 06 15:01:11 aaron: koen is touching linux-omap Mar 06 15:01:25 but here is just a fetch error it seems Mar 06 15:04:41 hi chouimat|work ! Mar 06 15:08:04 at the first time i runs bitbake ,when the process fetch the linux-omap_2.6.28,my pc breaks for power adapter ,when i reboot my pc ,i could not runs bitbake Mar 06 15:09:02 the first time .it runs for 4 hours , Mar 06 15:11:23 >fatal: Not a git repository Mar 06 15:11:46 you're trying to fetch from a non git repository Mar 06 15:12:26 this could be beacuse you've a git clone in your downloads, but it didn't has the index properly indexed Mar 06 15:13:04 bitbake -c clean linux-omap Mar 06 15:13:15 but i never modify any .bb files and .git dir Mar 06 15:13:18 this doesn't delete the git downloaded repo Mar 06 15:13:52 show your conf/local.conf in a pastebin Mar 06 15:14:11 i have http://paste.pchowtos.co.uk/?id=1540&key=5c4934671ba93 Mar 06 15:14:20 i have rm -rf tmp Mar 06 15:15:12 i think oe is seemed so stupid Mar 06 15:15:36 when i runs it ,i takes me so much time Mar 06 15:16:00 next again ,occus erros Mar 06 15:16:24 but when i run autotools Mar 06 15:16:32 make clean;make Mar 06 15:16:39 all 's fine Mar 06 15:19:27 aaron: bitbake -c clean linux-omap \ bitbake linux-omap is not soo difficult, isn't? Mar 06 15:20:05 you stumbled in disk error due to power cut...this is not standard Mar 06 15:25:16 ant_work thank u Mar 06 15:25:44 before ur mention bitbake -c clean linux-omap \ bitbake linux-omap Mar 06 15:26:02 i have try it so many many times Mar 06 15:26:11 no use Mar 06 15:29:13 mwester, ping Mar 06 15:29:36 the process is bitbake console-image ,it 'runs ok now . Mar 06 15:29:36 the process is bitbake console-image ,it 'runs ok now . after i rm -rf tmp Mar 06 15:29:36 so fuck oe Mar 06 15:29:36 it got me mad Mar 06 15:29:36 never saw so bad and stupid tools Mar 06 15:30:04 aaron: it is long time since I saw so bad and stupid user Mar 06 15:30:57 aaron, complaining endlessly is no help, people are trying to help and you are alienating them Mar 06 15:31:03 aaron: remember - there is always buildroot, you can also use crosstool(-ng) to generate toolchain and then compile everything by hand Mar 06 15:31:09 aaron, if you are unhappy with OE, try some of the alternatives Mar 06 15:31:30 we are not the solution to everyones problem Mar 06 15:33:07 before the time .i still compile packages hand by hand Mar 06 15:34:08 but becuse of the company use it ,i force to use it Mar 06 15:35:16 aaron: I tried to do compiling by hand. after few hours I gave up because it would take me month to build image. Mar 06 15:35:33 aaron: you know. some people in this channel are actually responsible for what you say is stupid Mar 06 15:35:53 thebohemian: have a time for query? Mar 06 15:36:14 aaron: calling it like this does not solve your problem and does not make people eager to help you as well Mar 06 15:36:24 hrw: sure Mar 06 15:36:45 hrw: took me a month to port es5 to arm ... all by hand Mar 06 15:36:51 natively compiled Mar 06 15:37:09 Crofton: Pong Mar 06 15:37:39 chouimat|work: what is this es5? Mar 06 15:37:51 I am curious why you would build a tarball and build from it Mar 06 15:38:01 redhat entreprise server 5 or is not entreprise linux 5 Mar 06 15:38:16 s/not/now Mar 06 15:39:31 Crofton: Release procedures for SlugOS require that we release a buildable tarball of all the sources along with the binary. The tarballs of all the packages themselves are easy; we simply expose the DL_DIR (basically) on the NSLU2 website, but for the OE snapshot we need to create the tarball. Mar 06 15:39:45 u r right ,but hand by hand .it's never occus those problems Mar 06 15:39:46 ok Mar 06 15:39:50 I kind of thought that Mar 06 15:39:55 :) Mar 06 15:40:13 atmost make clean ;make Mar 06 15:40:40 I'd love to just release a git tag, and point users to OE -- but that would just end up with 1,000 angry and upset users posting angry comments here -- and that's not productive ;-) Mar 06 15:40:45 nerver meets erros Mar 06 15:40:57 when u do it again Mar 06 15:41:20 mwester, thanks for taking us into account :) Mar 06 15:41:30 chouimat|work: rhel5? Mar 06 15:41:35 yup Mar 06 15:41:52 blasted name changes. Mar 06 15:43:08 Is it anything more than a name change (like licensing terms, or support terms)? Mar 06 15:43:24 * mwester supposes he should go to their website and read that for himself... Mar 06 15:46:01 huhu Mar 06 15:46:10 hoho Jin^eLD Mar 06 15:47:09 hey :) Mar 06 15:47:26 how are things in OE land? :> Mar 06 15:48:32 http://www.computerworld.com.au/article/278831/new_online_community_launches_embedded_linux_developers Mar 06 15:49:55 *reading* Mar 06 15:51:48 interesting, I'll have a look at it Mar 06 15:54:01 hello Mar 06 15:54:14 how do I make OE forget it has just made any debs Mar 06 15:54:31 and have bitbake build from the existing working directories again? Mar 06 15:54:53 (this is for chinook-compat...) Mar 06 16:01:36 bitbake target -cpackage -f;bitbake target -cpackage_write -f Mar 06 16:01:45 udovdh: bitbake target -cpackage -f;bitbake target -cpackage_write -f Mar 06 16:02:16 hrw aha. thanks. **** BEGIN LOGGING AT Fri Mar 06 16:41:29 2009 Mar 06 16:48:41 hi Mar 06 16:50:16 hi woglinde Mar 06 17:02:35 gstreamer is failing for me Mar 06 17:02:55 complains that mkdir is not in staging bin dir ... Mar 06 17:07:34 hi jineld Mar 06 17:13:56 When I build several packages fail on do_install because they try to use the host strip program rather than the cross armv4t strip program (when invoking install -s). Mar 06 17:14:08 I can get around this by changing the Makefile, but is there a way to do this globally? Mar 06 17:14:31 its the known automake 1.10.2 bug Mar 06 17:15:02 please look at the mailing list fot the patch form khem Mar 06 17:15:03 Gotcha, so I've been looking in the wrong place. Thanks, I'll look into automake then. Mar 06 17:15:07 oe-dev Mar 06 17:15:16 and we did not downgraded to 1.10.1 to get it 'solved' Mar 06 17:15:28 hrw *g* Mar 06 17:15:29 grr... Mar 06 17:15:39 mickeyl could do it I am sure Mar 06 17:15:44 *g* Mar 06 17:25:42 fyi: i managed to build navit for nokia n810 using OE Mar 06 17:25:48 see the wiki at http://wiki.navit-project.org/index.php/Navit_on_OpenEmbedded_for_n810 Mar 06 17:25:54 feedback welcome Mar 06 17:26:39 udo ah cool Mar 06 17:29:43 hm I dont like the last patch Mar 06 17:30:13 and Mar 06 17:30:16 # Make use of my SMP box PARALLEL_MAKE="-j1" BB_NUMBER_THREADS = "1" Mar 06 17:30:22 isnt really usefull Mar 06 17:31:15 yep Mar 06 17:31:33 I prefer -j4 and 8 bbthreads on quad Mar 06 17:32:21 same on 8core Mar 06 17:33:01 | /bin/sh: /home/balister/oe/tmp/staging/i686-linux/usr/bin/mkdir: No such file or directory Mar 06 17:33:09 any thoughts on this problem? Mar 06 17:33:17 bitbake coreutils-native Mar 06 17:33:45 do we need to get this into depends? Mar 06 17:34:08 helps to spell coreutils-native .... Mar 06 17:35:57 * mwester wonders if it's really necessary to have coreutils-native; can't we assume that's provided on any host capable of building OE? Mar 06 17:36:00 hmmm Mar 06 17:37:28 mwester: same thinking here. but I have no idea why OE from time to time wants rm/mkdir/cp etc from staging Mar 06 17:38:30 coreutils developers do not give enough importance to cross building env Mar 06 17:38:32 ok, than at least I'm not insane. :D Mar 06 17:38:46 you need to patch them to make them work for a cross env Mar 06 17:38:55 I am sure host coreutils wont work in all cases Mar 06 17:39:00 I wonder if the coreutils folks have just assumed that all cross dev work uses busybox instead? Mar 06 17:39:10 yes could be Mar 06 17:39:35 but they have even started removing things that cross dev used Mar 06 17:40:14 and unless someone who considers this it may not be possible to use vanilla coreutils for cross building Mar 06 17:41:36 on the other hand less dependence on host provided tools means more uniformity in OE builds on different machines Mar 06 17:44:36 as long as we get coreutils-native right, it will save future headaches Mar 06 17:54:30 03Philip Balister  07org.openembedded.dev * rc707ef1efb 10openembedded.git/conf/machine/davinci-sffsdr.conf: davinci-sffsdr : Remove dsplink dependency until dsplink works on sffsdr board. Mar 06 17:54:31 03Philip Balister  07org.openembedded.dev * r79e14954d7 10openembedded.git/packages/linux/linux-davinci_2.6.25.bb: linux-davinci-2.6.25 : Fix recipe so it builds reliably. Update git url. Mar 06 17:57:05 woglinde, I know about the smp stuff, I set it to 1 for debugging etc Mar 06 17:57:22 woglinde, the last patch is needed Mar 06 17:57:57 because bitbake for some reason (.pc file?) libgpsbt/libgpsmgr won't be find automagically Mar 06 17:58:02 during the configure phase Mar 06 17:58:06 so yes this is ugly Mar 06 17:58:15 but if I know how to fix this... Mar 06 17:59:29 libgpsmgr/bt build OK with the patches supplied. Mar 06 17:59:34 I list them as depends Mar 06 17:59:44 but they never get linked against during the build Mar 06 17:59:58 so when I install the navit deb on n810 Mar 06 18:00:07 the gps is not autostarted Mar 06 18:02:58 they never get linked against during the build because the PKG_CHECK_MODULES Mar 06 18:03:15 doesn't find the package for some reason Mar 06 18:03:25 maybe because of the .pc file? (dunno) Mar 06 18:03:51 if that issue is solved, we can drop the two patches at the bottom of the wiki page Mar 06 18:04:34 should I file a bug somewhere for this set of issues? Mar 06 18:04:47 a) not building without the patches listed under Issues Mar 06 18:04:51 pb___: ping Mar 06 18:04:53 b) package not being found Mar 06 18:05:00 if so: where? Mar 06 18:05:29 http://bugs.openembedded.net/ Mar 06 18:08:45 pb__ saw your msg on the mailinglist wrt cdrkit, can I just try the debian image on my system ???? Mar 06 18:09:56 hmmm Mar 06 18:09:58 In file included from flv.cpp:24: Mar 06 18:09:59 /home/frans/oe/tmp_angstrom/staging/armv7a-angstrom-linux-gnueabi/usr/include/boost/detail/endian.hpp:69:3: error: #error The file boost/detail/endian.hpp needs to be set up for your CPU type. Mar 06 18:10:08 any idea how to fix this? Mar 06 18:37:44 eFfeM-away: yes, the debian package should "just work" Mar 06 18:48:27 pb___: thanks will give it a try tomorrow, need to leave now Mar 06 18:48:46 was afraid there would be lib incompatibilities Mar 06 18:51:44 re Mar 06 19:23:01 how do you reverse commits made in your local repo ? Mar 06 19:23:23 git-revert? Mar 06 19:23:26 no Mar 06 19:23:28 :) Mar 06 19:23:33 I want to undo them Mar 06 19:23:41 hm.. Mar 06 19:23:44 I am pretty sure you can undo commits until you push them Mar 06 19:24:34 Crofton: git reset --hard HEAD~1 Mar 06 19:24:43 Crofton: or better yet, git rebase -i Mar 06 19:25:19 hmm Mar 06 19:25:21 urg Mar 06 19:25:30 git ate my homeowrk Mar 06 19:25:33 ;D Mar 06 19:25:41 still recoverable Mar 06 19:25:41 next time it will eat your dog Mar 06 19:25:46 heh Mar 06 19:26:11 Crofton: git-rebase -i v2.6.24 (will rebase all changes since v2.6.24, which is how I'm usually working) Mar 06 19:26:33 yeah, I had one commit where I forgot some files Mar 06 19:26:43 I should have commited them then used git reabase Mar 06 19:26:47 too late now Mar 06 19:31:54 Crofton, you have a bad git workflow. "Think then do" is better than "do then think" Mar 06 19:31:57 :) Mar 06 19:32:12 git sux :) Mar 06 19:34:41 That's rather like the carpenterr saying the hammer sux. :D Mar 06 19:45:11 Crofton: git reset Mar 06 19:46:34 Crofton: btw, what's the difference between davinci sffsdr and evm? Mar 06 19:47:41 cbrake: for some reason rebase never worked for me smoothly... :( Mar 06 19:48:41 ~seen kergoth Mar 06 19:48:44 kergoth was last seen on IRC in channel #oe, 2h 12m 44s ago, saying: 'hmmm'. Mar 06 19:49:06 denix: have you tried rebase interactive? That has always worked very well for me, but perhaps I'm using it differently. Mar 06 19:50:55 cbrake: no, I haven't... need to try it next time Mar 06 19:55:05 Hello there. Quick q. How much space does one need to build a lil image with this oe system? Mar 06 19:55:11 denix, radio interface Mar 06 19:55:19 (harddrive space, yeah) Mar 06 19:55:20 the board is made by Lyrtech Mar 06 19:57:03 schmx: depends. from 2 to 24GB Mar 06 19:57:47 Thanks, hrw Mar 06 19:57:52 Guess it's time to clear up some space then (: Mar 06 19:58:42 Disk space is cheap (although I have thought that Seagate and Western Digital should offer OE some donations). Mar 06 19:59:14 Same with Intel -- I do wonder how many processor upgrades they sold because of OE ;-) Mar 06 20:00:16 mwester: nice one! :) Mar 06 20:00:48 mwester: my upgrades were amd mostly Mar 06 20:01:11 mwester: and you forgot about memory upgrades few years ago (when bitbake did not fit in 0.5GB) Mar 06 20:02:18 Crofton: do you have a link to the board? Mar 06 20:02:32 denix: pong Mar 06 20:02:39 * kergoth was in class Mar 06 20:02:47 kergoth: how's class? Mar 06 20:03:02 not bad, have a ton of work to do this weekend on a paper Mar 06 20:03:10 Is there somewhere to look for information on building stuff for qemu? Mar 06 20:03:47 schmx: expand question Mar 06 20:04:06 Crofton: this one? http://www.lyrtech.com/index.php?act=view&pv=SFF%20SDR%20evaluation%20module Mar 06 20:04:15 http://www.lyrtech.com/index.php?act=view&pv=SFF_space_SDR_space_evaluation_space_module Mar 06 20:04:34 yeah that looks right Mar 06 20:04:50 hrw: I want to build an image to use with qemu terrier emulation. Mar 06 20:04:56 Crofton: thanks Mar 06 20:05:23 schmx: MACHINE=spitz DISTRO=angstrom-2008.1 and do build Mar 06 20:05:39 Oh like normal (: Mar 06 20:06:00 Right. Time to see if this baby runs faster than last time I tried it ;) Mar 06 20:06:23 kergoth: if you are too busy, I don't mind commiting the enterprise/BBMASK fix... haven't heard from zecke though - is he back? Mar 06 20:06:23 ;D Mar 06 20:06:39 schmx: x86-64 is better host for qemu then 32bit Mar 06 20:06:45 no idea. feel free to commit it, otherwise i'll add it to my list Mar 06 20:06:51 hrw: I don't quite understand ? Mar 06 20:07:01 schmx: you have 32 ot 64bit machine? Mar 06 20:07:17 hrw: 64 Mar 06 20:07:42 I can't run 32bit binaries, so that is no worry ;) Mar 06 20:08:16 ok Mar 06 20:12:46 Hey this is all coming back to me. I remember this was a very helpful channel too. Big up! Mar 06 20:18:31 03Philip Balister  07org.openembedded.dev * r7ca6daec39 10openembedded.git/packages/linux/linux-davinci/ (5 files in 3 dirs): linux-davinci : Move linux-davinci directory to linux-davinci-2.6.25. Mar 06 20:18:32 03Philip Balister  07org.openembedded.dev * rb0b1ce5846 10openembedded.git/packages/linux/ (15 files in 2 dirs): Mar 06 20:18:32 linux-davinci-git : Update recipe so it builds for sffsdr board. Mar 06 20:18:32 * Rename recipe. Mar 06 20:18:34 03Philip Balister  07org.openembedded.dev * rd3a1be4261 10openembedded.git/conf/machine/davinci-sffsdr.conf: davinci-sffsdr : Use Default preference in kernel recipes to select kernel. Mar 06 20:18:37 03Philip Balister  07org.openembedded.dev * r477cb9e7de 10openembedded.git/packages/linux/linux-davinci_2.6.27.bb: Mar 06 20:18:40 linux-davinci_2.6.27 : DEFAULT_PREFERENCE = "-1" Mar 06 20:18:42 Recipe may not build. Mar 06 20:21:47 Qemu 0.10.0 released Mar 06 20:23:27 http://qemu-forum.ipi.fi/viewtopic.php?f=1&t=4895 Mar 06 20:23:45 wow Mar 06 20:27:36 Jay7: does it build at all? :) Mar 06 20:27:47 denix: I don't know :) Mar 06 20:27:51 just read news :) Mar 06 20:28:16 *reading Mar 06 20:29:29 03Angus Ainslie  07fso/milestone5.5 * rfa7c9e7467 10openembedded.git/packages/openmoko2/openmoko-alsa-scenarios.bb: Add scenario logic for 3d7k and mark scenario files as CONFFILES Mar 06 20:34:52 re Mar 06 20:45:21 any thought? Mar 06 20:45:22 http://rafb.net/p/R4qLWt81.html Mar 06 20:45:26 this is weird ... Mar 06 20:45:49 uh nice Mar 06 20:46:33 ndeed Mar 06 20:46:49 ~curse icedtea6-native Mar 06 20:46:59 May you be reincarnated as a Windows XP administrator, icedtea6-native ! Mar 06 20:47:56 hrw whats wrong with icedtead? Mar 06 20:48:20 * mwester likes a nice glass of iced tea. Mar 06 20:48:24 any suggestions on how to fix? Mar 06 20:48:31 it was written by developers which never heard about crosscompilation Mar 06 20:48:37 hihi long island ice tea Mar 06 20:48:45 hrw hm? Mar 06 20:48:57 robert and worked a lot on openjdk Mar 06 20:49:06 thats why it is in oe Mar 06 20:49:07 woglinde: I had to add -I$(includedir) here, -L$(libdir) there etc Mar 06 20:49:14 and jalimo overlay Mar 06 20:49:23 I am using jalimo overlay Mar 06 20:49:30 hm Mar 06 20:49:31 icedtea6-native_1.3.1.bb Mar 06 20:49:46 | /usr/bin/ld: cannot find -lz Mar 06 20:49:56 hm I tought robert eliminate all such occurs Mar 06 20:50:12 at least for chinook it worked good enough Mar 06 20:50:32 hrw for which machine do you need it? Mar 06 20:50:34 it works ok when you have freetype-dev, zlib-dev installed on host Mar 06 20:50:37 woglinde: BUG Mar 06 20:50:43 hm oh Mar 06 20:50:53 okay Mar 06 20:50:54 and Robert knows that I am working on it Mar 06 20:51:03 please drop us an email on thos Mar 06 20:51:56 woglinde: he got first patchset today Mar 06 20:52:05 with java-native.bbclass added to it Mar 06 20:52:05 thanks :) Mar 06 20:53:18 I decided to not get r/w on jalimo ;) Mar 06 20:53:22 for now atleast Mar 06 20:54:28 hm yes we have to integrate it into oe anyway Mar 06 21:01:15 NOTE: package icedtea6-native-1.3.1-r0: task do_compile: completed Mar 06 21:01:45 hrw :) Mar 06 21:01:58 after 3 or even 4 hours Mar 06 21:02:03 hihi Mar 06 21:02:04 yes Mar 06 21:03:18 * schmx notes that the bitbakeing was faster than last year. Horray for quickier CPU (: Mar 06 21:04:42 schmx: and bitbake is multithreaded Mar 06 21:05:12 Lovely. Mar 06 21:05:21 http://marcin.juszkiewicz.com.pl/download/diffs/0001-icedtea6-native-added-missing-dependencies-made-it.patch Mar 06 21:06:56 hrw wyh automake 10? Mar 06 21:07:42 schmx: there is page about running akita images in qemu Mar 06 21:07:48 woglinde: Poky 'pinky' uses 1.9.6 Mar 06 21:07:52 schmx: you can find it useful Mar 06 21:08:11 schmx: http://www.linuxtogo.org/gowiki/AngstromAndQemu/Akita Mar 06 21:08:53 Jay7: "-serial stdio" is nice too Mar 06 21:10:02 btw.. Mar 06 21:10:29 I should write why qemu-system-arm fails with kexec and reboot Mar 06 21:14:12 * hrw -> off Mar 06 21:14:12 bye Mar 06 21:14:56 chbye hrw Mar 06 21:15:16 well, I'm going sleep too.. Mar 06 21:15:53 nite jay7 Mar 06 21:27:26 03Marcin Juszkiewicz  07org.openembedded.dev * r38402a4968 10openembedded.git/packages/mbw/mbw_1.1.bb: (log message trimmed) Mar 06 21:27:26 mbw: added 1.1 version of simple memory bandwidth benchmark Mar 06 21:27:26 Results from my devices: Mar 06 21:27:26 root@bug:~# mbw 16|grep AVG Mar 06 21:27:26 AVG Method: MEMCPY Elapsed: 0.16403 MiB: 16.00000 Copy: 97.544 MiB/s Mar 06 21:27:27 AVG Method: DUMB Elapsed: 0.12401 MiB: 16.00000 Copy: 129.023 MiB/s Mar 06 21:27:31 AVG Method: MCBLOCK Elapsed: 0.09725 MiB: 16.00000 Copy: 164.519 MiB/s Mar 06 21:29:04 does anyone know if oplinux distro is being maintained? seems to have some problems. like trying to svn some non-existent versions Mar 06 21:29:32 for example, gconf-dbus_svn.bb fails Mar 06 22:24:40 emlinux_: feel free to send patches Mar 06 22:26:44 nytowl: please cherry-pick the next commit Mar 06 22:26:45 g'night Mar 06 22:26:50 03Michael 'Mickey' Lauer  07org.openembedded.dev * reecd3b4269 10openembedded.git/packages/freesmartphone/ (libgsm0710mux/om-3d7k/abyss.conf libgsm0710mux_git.bb): libgsm0710mux: fix modem path on om-3d7k Mar 06 22:26:54 mickey|zzZZzz, gn Mar 06 22:27:12 Anyone see anything like this happen when on do_rootfs? http://pastebin.com/m765934ba Mar 06 22:27:22 nytowl: for fso/ms5.5 you should also bump to the latest versions of frameworkd, libgsm0710mux, and fso-abyss Mar 06 22:27:30 looks like opkg-cl is causing a huge barf. Mar 06 22:39:36 http://pastebin.com/m32ac189e Mar 06 22:40:21 Manual download of this file - helps... Mar 06 22:40:36 link wrong... Links to testing kernels = evil Mar 06 22:41:56 03Thomas Kunze  07org.openembedded.dev * r768fe337b0 10openembedded.git/packages/linux/ (linux-2.6.24/simpad/defconfig linux_2.6.24.bb): linux_2.6.24: enable serial_cs for simpad Mar 06 22:46:46 mewyn: maybe its selinux related. Mar 06 22:48:15 The_Niz, I do not think Mar 06 22:48:31 double free issues are most likely code issues **** ENDING LOGGING AT Sat Mar 07 02:59:57 2009