**** BEGIN LOGGING AT Thu Sep 30 02:59:57 2010 Sep 30 03:01:54 Is there any way to set a verbosity option that would associate bitbake actions with the task that they're helping execute? i.e. 'running task 654...' and then something in the task breaks but if you've got multiple bitbake threads you kinda gotta dig around to associate the problem with the task Sep 30 03:02:54 you can run bitbake -v anything-that-you-will-do and it will show verbose of all tasks Sep 30 03:04:06 Yeah.. What I'm shooting for is less log data to sift through though. And it still doesn't associate the action with the task directly. Sep 30 03:04:10 Altho good suggestion Sep 30 03:41:34 Bye all Sep 30 03:52:36 loadammo: are the log files that every task generates in some way insufficient? Sep 30 03:53:23 I have to go dig around to find those. Sep 30 03:56:56 When the pertinant information is available in the output of the bitbake I just ran, it seems like there's just a matter of formatting the output so it includes the task. I'm playing around with a couple ideas, like overriding oenote/oewarning/oefatal so that it's echo "${PN}" "$*", which basically gets me what I need. Sep 30 03:57:38 The problem is, I've got a co-worker who's not too DIY, so when the build breaks he just comes and bugs me without looking at pretty much any logs. So trying to find a cut/paste solution to make my life easier. Sep 30 03:58:58 loadammo, your co-worker is broken and needs to be patched Sep 30 03:59:07 I suggested a catapult Sep 30 03:59:32 and he's senior to me, so. :/ Sep 30 03:59:43 problem definitely exists between keyboard and chair. Sep 30 04:15:35 loadammo: oenote/oewarning/oefatal are useless shell scripts that emit messages that only go to the log, not to the console Sep 30 04:15:44 you'd have to monkeypatch bb.msg for the python messages Sep 30 04:16:02 I likely do not. I shall google. Sep 30 04:16:09 in bitbake master, we're using the python logging module, so it'd be much easier to change formatting or add extra handlers to send output to alternative locations in alternative formats Sep 30 04:16:30 ahh Sep 30 04:19:04 that'd do it. Sep 30 04:20:44 thank you Sep 30 05:57:55 Aarti, here ? Sep 30 05:57:59 ya Sep 30 05:58:14 u are here already ! Sep 30 06:29:31 03Frans Meulenbroeks  07org.openembedded.dev * r285f58b3a4 10openembedded.git/ (2 files in 2 dirs): (log message trimmed) Sep 30 06:29:31 Revert "wpa-supplicant : moved unused files to obsolete dir" Sep 30 06:29:31 This reverts commit 9a5382aad4e5a3e52fa52722cf5587fdba958e9f. Sep 30 06:29:31 This one accidentally also dragged in classes/kernel.bbclass Sep 30 06:29:31 I will redo the commit for wpa-supplicant later. Sep 30 06:29:32 classes/kernel.bbclass will be se a separate commit Sep 30 06:29:32 Thanks for Paul Menzel for spotting this Sep 30 07:04:59 03Martin Jansa  07master * rfe0535e783 10openembedded.git/recipes/kbd/ (kbd_1.12.bb kbd_1.15.2.bb): Sep 30 07:05:00 kbd: upgrade from 1.12 to 1.15.2, add native version Sep 30 07:05:00 * native version is usefull in some kernel builds where drivers/char Sep 30 07:05:00 calls loadkeys (which are not available on some buildhosts) Sep 30 07:05:00 Signed-off-by: Martin Jansa Sep 30 07:12:20 hi all, m getting error tht do_compile failed Sep 30 07:13:23 because it's not getting one .h file which is specified in my Makefile Sep 30 07:13:51 Aarti, pastebin ur error .. Sep 30 07:14:03 also provide the compilation command here .. Sep 30 07:14:45 harsh, for compilation m doing bitbake Sep 30 07:18:08 http://pastebin.com/xY6WUhbe Sep 30 07:19:32 * harsh is busy Sep 30 07:23:36 Aarti: the makefile is busted, you need to fix it and add the local include paths to compiler flags using -I Sep 30 07:24:16 good morning Sep 30 07:24:17 you could also add those paths using CFLAGS += in recipe but I would rather suggest to fix makefile Sep 30 07:24:29 recalcati: hello Sep 30 07:25:30 03Martin Jansa  07master * re770146203 10openembedded.git/recipes/freesmartphone/ (cornucopia.inc fsodeviced_git.bb): Sep 30 07:25:31 cornucopia: bump SRCREV for new alsa state files for 2.6.34 Sep 30 07:25:31 Signed-off-by: Martin Jansa Sep 30 07:25:33 03Martin Jansa  07master * r88fc0e2153 10openembedded.git/recipes/freesmartphone/ (2 files in 2 dirs): Sep 30 07:25:33 frameworkd: limit oeventsd workaround for buggy kernel only to 2.6.32 Sep 30 07:25:33 Signed-off-by: Martin Jansa Sep 30 07:25:34 Are anybody going to OEDEM. I'll be in the previous two conferences, and I need a confirm that OEDEM is in the same hotel Sep 30 07:25:57 the first was a question... Sep 30 07:26:00 recalcati: I am not there Sep 30 07:26:39 ok, thx anyway Sep 30 07:30:15 good morning Sep 30 07:34:41 thnx khem Sep 30 07:36:56 i was bitbake gcc cross i got this error message Sep 30 07:37:05 Please install following missing utilities: C++ Compiler (${BUILD_PREFIX}g++ Sep 30 07:37:11 whats this error? Sep 30 07:40:00 anyone alive?? Sep 30 07:49:57 khem, i have done in my recipe CFLAGS += "-Wall -fPIC -I. -I$(INCPATH)" Sep 30 07:50:17 but my problem is still not solved Sep 30 07:53:55 recalcati: i'm going to elc/oedem, afaik there has not been any communication on the oedem location Sep 30 07:57:06 nor on the agenda Sep 30 08:14:37 03Holger Hans Peter Freyther  07master * r843f051cf3 10openembedded.git/recipes/meta/meta-toolchain-neuros.bb: Sep 30 08:14:37 meta-toolchain-neuros: Remove recipe, meta-toolchain-qte is a better replacement Sep 30 08:14:37 meta-toolchain-qte is a good replacement for this target. Koen has Sep 30 08:14:37 acked the removal and it opens the door to some other cleanup. Sep 30 08:14:45 03Holger Hans Peter Freyther  07master * rf454442543 10openembedded.git/recipes/uicmoc/ (3 files): Sep 30 08:14:45 uicmoc4: qt4-tools is the replacement. Sep 30 08:14:45 qt4-tools exists for 4.5, 4.6, 4.7. There is nothing in the OE Sep 30 08:14:45 tree that is still using the uicmoc4* target. Remove it so people Sep 30 08:14:45 do not think that they need to use it. Sep 30 08:14:46 03Holger Hans Peter Freyther  07master * r5703890dbe 10openembedded.git/recipes/qt4/ (3 files in 2 dirs): Sep 30 08:14:46 qt4-tools-native: Build the tools as -embedded to avoid X11 deps Sep 30 08:14:47 We only need command line utilities and can avoid requiring X11. Sep 30 08:18:13 morning Sep 30 08:53:48 zecke: what about that build error I've sent you with embed qt4-tools-native? Sep 30 08:54:04 JaMa: aeh Sep 30 08:54:15 JaMa: I thought -embedded worked for you? Sep 30 08:54:16 oh zhecke is awake :3 Sep 30 08:55:27 ./oe-2010-09-28:20:36 < JaMa> zecke: http://tinderbox.openembedded.net/packages/793224/ Sep 30 08:56:38 JaMa: okay, funny part is... a lot of people have issues with freetype headers lately Sep 30 08:57:25 JaMa: will add a -no-freetype to the list Sep 30 09:10:37 03Holger Hans Peter Freyther  07master * rfaec5db53b 10openembedded.git/recipes/qt4/qt4-tools-native.inc: Sep 30 09:10:37 qt4-tools-native: Add -no-freetype to the build Sep 30 09:10:37 Avoid any dependencies on freetype, just disable it. We are Sep 30 09:10:37 not interested in the GUI module right now, we build it to Sep 30 09:10:37 have the tools/qdbus build, which also builds the qdbusviewer. Sep 30 09:10:49 03Holger Hans Peter Freyther  07master * r2e1c5469e2 10openembedded.git/recipes/tasks/task-qte-toolchain-target.bb: Sep 30 09:10:49 task-qte-toolchain-target: The QtAssistantClient is deprecated and removed in 4.7 Sep 30 09:10:49 Do not install the header files as they are deprecated and removed Sep 30 09:10:49 from Qt 4.7 and onwards. Sep 30 09:12:10 good morning Sep 30 09:12:29 03Martin Jansa  07master * reefd79ec83 10openembedded.git/recipes/freesmartphone/libfsotransport_git.bb: Sep 30 09:12:29 libfsotransport: enable palmpre-hsuart for all Sep 30 09:12:29 * machine override without MACHINE_ARCH is bad idea Sep 30 09:12:29 * 11:05:19 < mickey|office> palmpre hsuart support doesn't hurt Sep 30 09:12:29 * 11:05:59 < mickey|office> palmpre support can be compiled in unconditionally for all platforms Sep 30 09:22:55 JaMa: I added -no-freetype but couldn't test. i have started a build but it will take a while Sep 30 09:23:57 zecke: I've started build too, I'll ping you with result Sep 30 09:29:50 JaMa: thanks Sep 30 09:37:43 florian: linuxtogo is not responsive Sep 30 09:38:01 I'd say slow as hell Sep 30 09:41:43 * florian checks Sep 30 09:43:47 ant_work: What did you try to access? The machine looks pretty much idle... Sep 30 09:44:03 git.linuxtogo Sep 30 09:44:14 and lists. too Sep 30 09:45:05 now git.linuxtogo is better Sep 30 09:45:09 strange... gitweb is comparably fast for me Sep 30 09:45:14 heh Sep 30 09:45:36 hm.. Sep 30 09:45:41 XML Parsing Error: no element found Sep 30 09:45:43 Location: http://git.linuxtogo.org/ Sep 30 09:48:37 florian: now is slow again :/ Sep 30 09:48:58 is someone remote-building on linuxtogo? Sep 30 09:49:07 it did happen :D Sep 30 09:50:36 heh..I see strange xml errors, Sep 30 09:50:39 XML Parsing Error: unclosed token Sep 30 09:50:41 Location: http://git.linuxtogo.org/?p=groups/kexecboot/kexecboot.git;a=shortlog Sep 30 09:50:43 ant_work: no... doesn't seem so. narcissus is running but with a load of 3 the machine is mostly idle. Sep 30 09:50:43 Line Number 175, Column 246: Sep 30 09:51:28 huh? Sep 30 09:51:35 this url works for me Sep 30 09:51:42 yes, probably the page is not loaded fast enough Sep 30 09:51:53 try to zap between links Sep 30 09:52:01 summary / shortlog / .. Sep 30 09:52:28 you should see a yellow page with red error msg Sep 30 09:52:55 then disappears Sep 30 09:53:11 seems smthg is not in sync Sep 30 09:53:41 ah that's firefox... yes it seems to try to parse the file before it read it completely. Sep 30 09:53:49 managed to get this one with the log Sep 30 09:54:07 yep, firefox Sep 30 09:55:18 fraxinas: well, what is your current state? Sep 30 09:55:50 hmm were we stopped the other night Sep 30 09:56:10 if i build with -fjni then i get 7000+ resource not found errors Sep 30 09:56:23 fraxinas: can you give me the source package you are trying to compile? Sep 30 09:56:41 sure i'll mail you what i got so far Sep 30 09:58:38 fraxinas: you should see the same problem when statically linking on your desktop Sep 30 10:21:50 fraxinas: hmm wehre should openssl/ec.h come from? :) Sep 30 10:22:53 hm mom Sep 30 10:23:31 ./mipsel-oe-linux/usr/include/openssl/ec.h Sep 30 10:26:19 comes from openssl-0.9.8j-r12.0 Sep 30 10:28:26 fraxinas: it is gone from 1.0.0a :( Sep 30 10:28:42 really??? annoying!! Sep 30 10:28:50 good thing i still use the outdated junk :) Sep 30 10:35:54 zecke: qt4-tools-native built fine here for armv7a.. thanks Sep 30 10:42:39 JaMa: native armv7a build? Sep 30 10:46:02 zecke: ofc just native, but building with MACHINE=nokia900 Sep 30 10:55:07 fraxinas: where is the java code? Sep 30 10:57:22 zecke it's in dumphd-0.61-mipsel/source/dumphd/ Sep 30 10:59:49 03Martin Jansa  07master * r6045791ded 10openembedded.git/recipes/freesmartphone/ (fso-specs_git.bb libfso-glib_git.bb): fso-specs,libfso-glib: bump SRCREV Sep 30 11:00:50 fraxinas: not in the tarball I downloaded. :) Sep 30 11:01:03 what!? moment Sep 30 11:02:53 zecke the url i gave you was no real tarball Sep 30 11:03:12 it's just my work directories compressed, and it's in there Sep 30 11:10:02 03Martin Jansa  07master * r5918bb5442 10openembedded.git/ (4 files in 2 dirs): moko(suite,wm,wm-inf-ecore): bump SRCREV and remove AUTOREV from shr-autorev Sep 30 11:21:38 03Brijesh Singh  07org.openembedded.dev * r7be610b634 10openembedded.git/recipes/ti/ (7 files in 2 dirs): Sep 30 11:21:39 ti-dvtb: add new recipe to build DVTB (Digital Video Test Bench) Sep 30 11:21:39 Signed-off-by: Brijesh Singh Sep 30 11:21:39 Signed-off-by: Denys Dmytriyenko Sep 30 11:21:39 Signed-off-by: Roger Monk Sep 30 11:21:39 Signed-off-by: Koen Kooi Sep 30 11:21:58 03Brijesh Singh  07org.openembedded.dev * r048e1a1955 10openembedded.git/recipes/ti/ (ti-codecs-dm365_4.00.00.00.bb ti-eula-unpack.inc): (log message trimmed) Sep 30 11:21:58 ti-codecs-dm365_4.00.00.00: add new recipe for dm365 codecs Sep 30 11:21:58 * This recipe downloads the individual codec.s from the web and puts them together to create a codec package. Sep 30 11:21:58 * The output of codec installer is 'tar' file and we needed to extract this Sep 30 11:21:58 file during unpack and since ti-eula-unpack has a logic for extracting Sep 30 11:22:32 tar file, but it assumes gzip version instead of regular tar. And I don't see Sep 30 11:22:32 any recipe currently using TARFILE variable hence updating ti-eula-unpack.inc Sep 30 11:22:32 03Brijesh Singh  07org.openembedded.dev * r936ed14cdb 10openembedded.git/recipes/ti/ti-codecs-dm365_4.00.00.00.bb: Sep 30 11:22:32 ti-codecs-dm365_4.00.00.00: update SRC_URI to use latest dm365 codecs Sep 30 11:22:32 Signed-off-by: Brijesh Singh Sep 30 11:22:33 Signed-off-by: Denys Dmytriyenko Sep 30 11:22:34 Signed-off-by: Roger Monk Sep 30 11:22:34 Signed-off-by: Koen Kooi Sep 30 11:22:35 03Brijesh Singh  07org.openembedded.dev * r4430b2a315 10openembedded.git/recipes/ti/ (ti-c6accel.inc ti-c6accel_1.01.00.01.bb): Sep 30 11:22:35 ti-c6accel_1.01.00.01: add new c6accel version. Sep 30 11:22:36 Signed-off-by: Brijesh Singh Sep 30 11:22:42 Signed-off-by: Denys Dmytriyenko Sep 30 11:22:42 Signed-off-by: Roger Monk Sep 30 11:22:42 Signed-off-by: Koen Kooi Sep 30 11:22:42 03Brijesh Singh  07org.openembedded.dev * rdfdd6a064c 10openembedded.git/recipes/ti/gstreamer-ti/dm365-evm/loadmodules.sh: Sep 30 11:22:42 gstreamer-ti: update loadmodules.sh Sep 30 11:22:42 Signed-off-by: Roger Monk Sep 30 11:22:43 Signed-off-by: Koen Kooi Sep 30 11:22:44 03Roger Monk  07org.openembedded.dev * rac9b39905b 10openembedded.git/recipes/ti/ (10 files): (log message trimmed) Sep 30 11:22:44 ti-recipes: Switch remaining buildable ti recipes to use KERNEL_PR Sep 30 11:22:45 * Switch recipes to use KERNEL_PR + append if they are buildable Sep 30 11:22:45 (91 lines omitted) Sep 30 11:24:49 03Martin Jansa  07master * r7587cd5255 10openembedded.git/recipes/freesmartphone/ (libframeworkd-glib_git.bb libfreesmartphone-glib_git.bb): Sep 30 11:24:49 lib(frameworkd,freesmartphone)-glib: bump SRCREV Sep 30 11:24:49 Signed-off-by: Martin Jansa Sep 30 11:55:57 Hi all Sep 30 11:56:10 do we have expect package recipe in openembedded Sep 30 11:56:30 zecke, does it work now Sep 30 11:56:33 ? Sep 30 11:56:38 I skimmed through the recipe folder but could not find any recipe related to expect packahe Sep 30 12:00:10 the tcl thing? Sep 30 12:01:27 I have a package and I wish to integrate an extra icon theme and a launcher script into the recipe, what's the best way of doing that? Sep 30 12:01:50 tar.gz it up, add it in SRC_URI and add some do_install hooks? Sep 30 12:16:12 hi Noor, does ./recipes/python/python-pexpect_2.3.bb or ./recipes/perl/libexpect-perl_1.20.bb got anything to do with what you are looking for? Sep 30 12:23:27 let me see but I have doubts that expect package is nothing to so with python Sep 30 12:33:08 fahad python-pexpect is a different package than expect Sep 30 12:42:19 Noor, and what about ./recipes/perl/libexpect-perl_1.20.bb Sep 30 12:50:29 fahad it perl module :) Sep 30 14:05:11 I was consider to add lttng support to my kernel, I can see that lttng-control recipe is there. If i build that recipe will I get the lttng packagae working in my Linux distro or I have to patch the kernel as well? Sep 30 14:51:45 Noor you have to enable it in kernerl config too Sep 30 14:53:25 JaMa: how did compilation go for armv7 Sep 30 14:54:44 zecke: around ? Sep 30 14:55:05 khem, I'm rebuilding from scratch with your patches Sep 30 14:55:25 GNUtoo|laptop: cool. How long does it take for you to rebuild Sep 30 14:55:47 a coupple of hours Sep 30 14:55:58 I don't know exactly Sep 30 14:55:58 ok Sep 30 14:56:00 core i7 Sep 30 14:56:03 4G ram Sep 30 14:56:07 slow hdd Sep 30 14:56:12 hmmm Sep 30 14:56:22 at least 3 hours Sep 30 14:56:34 ok cool Sep 30 14:56:43 at least means more than Sep 30 14:56:50 I don't know exactly Sep 30 14:56:56 maybe 5 Sep 30 14:56:59 maybe 4 Sep 30 14:57:07 which image do u build Sep 30 14:57:11 shr-image Sep 30 14:57:22 for nokia900 Sep 30 14:57:31 on a branch that is not hosted in oe Sep 30 14:57:34 ok how many tasks does it say Sep 30 14:57:46 task 2110 of 8095 Sep 30 14:58:01 do u have rm_work enabled Sep 30 14:59:51 no Sep 30 14:59:52 never Sep 30 15:00:12 ok then it could take some time Sep 30 15:00:30 8000 is some amount Sep 30 15:00:31 btw can you identify an assembly kind just by looking at an instruction? Sep 30 15:00:37 but core i7 Sep 30 15:00:38 yes Sep 30 15:00:40 ok Sep 30 15:00:43 let me find it Sep 30 15:01:22 push {r4, r5, r6, r7, r8, lr} Sep 30 15:01:29 it's SMID? Sep 30 15:01:36 no Sep 30 15:01:53 its normal stack operation Sep 30 15:01:55 I've illegal instruction from gcc init stuff in wesnoth for armv4 Sep 30 15:01:57 stmia Sep 30 15:01:59 ok Sep 30 15:02:08 but I didn't find it in armv4t Sep 30 15:02:10 manual Sep 30 15:02:41 its a thumb instruction Sep 30 15:02:51 ahh ok Sep 30 15:03:00 but wesnoth should be ARM? Sep 30 15:03:02 I'll check Sep 30 15:03:08 r8 could be a problem Sep 30 15:03:15 ah? Sep 30 15:03:20 thumb only uses r0 to r7 Sep 30 15:03:24 ok Sep 30 15:03:30 let me boot the device Sep 30 15:06:34 http://pastebin.com/B7pBcZpC Sep 30 15:06:56 maybe it's not at 0 Sep 30 15:06:59 but here: Sep 30 15:07:04 => 0x00416e64 <+204>: blx r7 Sep 30 15:07:09 I'll lookup blx Sep 30 15:08:47 khem: hi Sep 30 15:09:09 GNUtoo|laptop: yes its blx and its not present on armv4t Sep 30 15:09:17 ok Sep 30 15:09:28 and accroding to l it's from gcc right? Sep 30 15:09:42 zecke: I have 450M of webkit library Sep 30 15:09:56 zecke: if I strip it then its 16M Sep 30 15:10:05 zecke: there are two on them Sep 30 15:10:31 khem: where? in the package? staging area? Sep 30 15:10:44 zecke: I have disabled stripping to ease debugging and it wont fit into any .ext2 now Sep 30 15:11:05 zecke: its in staging as well as root file system Sep 30 15:11:17 zecke: why is debug info so bloated Sep 30 15:11:28 khem: webkit is horrible big Sep 30 15:11:29 is it hitting some gcc bug or something Sep 30 15:11:35 khem: lot of template usage Sep 30 15:11:41 zecke: the final binary is 16M Sep 30 15:11:54 zecke: hmm ok Sep 30 15:12:09 give me a bit, i always took it for granted Sep 30 15:12:10 brb Sep 30 15:12:34 zecke: I think 450 meg is still a lot Sep 30 15:12:42 even for templates Sep 30 15:13:07 zecke: and you also had a patch to delete .lib dirs in rm_work Sep 30 15:13:13 it was not applied Sep 30 15:13:22 can you revive that too Sep 30 15:13:41 because these libs are under .lib and .lib escapes rm_work Sep 30 15:15:33 khem, what should I do for blx? gcc-4.5/libgcc/../gcc/config/arm/lib1funcs.asm seem to come from gcc and so I bet it's the loading assembly functions Sep 30 15:17:07 GNUtoo|laptop: its not libgcc where the problem lies Sep 30 15:17:28 these are string templates from libstdc++ Sep 30 15:18:27 whats is it passing in -march or -mtune Sep 30 15:18:46 I'll look Sep 30 15:19:02 GNUtoo|laptop: my guess is that those flags are not passed to gcc Sep 30 15:19:10 I bet it too Sep 30 15:19:18 I'll look Sep 30 15:19:23 khem: do you know if stock gcc has an option to leave the dwarf inside the .o files? Sep 30 15:19:24 therefore gcc configures for default which is armv5te Sep 30 15:19:34 huhh I'm stupid Sep 30 15:19:41 It's comming from SHR repo Sep 30 15:19:52 jo Sep 30 15:19:55 and they use rm_work Sep 30 15:20:02 hm nice next bombfound in berlin Sep 30 15:20:08 zecke: ? thats the default if you use -g Sep 30 15:20:19 woglinde: scary Sep 30 15:20:23 then I'll have to build it Sep 30 15:20:30 khem, indeed scary Sep 30 15:20:34 good that have to leave the train earlier Sep 30 15:20:53 khem: hmm. but when linking the dwarf stuff gets copied into the .so? Sep 30 15:21:19 zecke: it will but I mean its still huge Sep 30 15:21:56 khem: on OSX the default appears that the dwarf stuff is never copied to the .so Sep 30 15:22:26 but first I've to finish nokia900 build Sep 30 15:22:28 I'll wait then Sep 30 15:23:10 zecke: they dont build with -g ? Sep 30 15:24:07 khem: sorry. On OSX it appears that when you build with -g the dwarf stuff is not copied into the .so, debugging works by going through the .o files or such Sep 30 15:24:20 khem: at least linking is very fast on OSX and the debug build is small Sep 30 15:24:55 okay till later Sep 30 15:25:05 zecke: interesting dont they use gcc Sep 30 15:25:37 GNUtoo|laptop: there are two solutions, 1 we could force gcc defaults back to armv4t Sep 30 15:25:39 khem: patched one, and a forked linker.. so the magic might be somewhere else or I am confused Sep 30 15:25:52 GNUtoo|laptop: or we could pass --with-arch option to gcc Sep 30 15:25:57 ok Sep 30 15:25:58 second one is more elegant Sep 30 15:26:18 I don't know second at all Sep 30 15:26:22 zecke: could be Sep 30 15:27:58 khem: which .lib patch are you talking about? Sep 30 15:31:36 khem: I just cleaned my OE build (going Out of Disk space) Sep 30 15:31:44 03Roger Monk  07org.openembedded.dev * r81162f4b07 10openembedded.git/recipes/ti/ (3 files in 2 dirs): (log message trimmed) Sep 30 15:31:44 gstreamer-ti: various fixes Sep 30 15:31:44 * fix sed for initscript Sep 30 15:31:44 * fix typos in recipe Sep 30 15:31:44 * fix typos in initscript Sep 30 15:31:45 * fix omapdmaifbsink missing symbols Sep 30 15:31:45 Signed-off-by: Roger Monk Sep 30 15:34:47 RP: ping Sep 30 16:07:12 zecke: rm_work currently does not delete any folder starting with a dot(.) Sep 30 16:07:29 and IIRC you once proposed a patch to take care of that Sep 30 16:07:37 I cant find the patch link Sep 30 16:07:45 so could be it was on IRC Sep 30 16:07:50 I dont remember Sep 30 16:08:13 .lib in webkit takes 1G Sep 30 16:08:17 :( Sep 30 17:25:56 03Koen Kooi  07org.openembedded.dev * r6fd8d018ae 10openembedded.git/recipes/ti/ti-codecs-omapl138_4.00.00.00.bb: ti-codecs-omapl138 4.00.00.00: update SRCREV to integrate c6accel Sep 30 17:34:35 * kergoth thinks about resurrecting interactive mode Sep 30 17:35:11 khem, 7942 of 8095 Sep 30 17:35:20 GNUtoo|laptop: cool Sep 30 17:37:55 kergoth: lt-cd-drive do u know anything about this process ? Sep 30 17:38:09 seems to be launched when bitbaking OE Sep 30 17:38:17 and takes 100% cpu Sep 30 17:38:57 what repo should I be using for a verdex? ( IE has the Q1 2010 switch to the overo repo that was mentioned in the wiki occurred ) Sep 30 17:40:14 khem: no idea, never heard of it Sep 30 17:41:27 dante, ask on the gumstix list Sep 30 17:44:56 kergoth: lt-cd-drive is coming from /home/kraj/work/oe/tmp/work/i686-oe-linux/libcdio-0.81-r0/libcdio-0.81/src/.libs/lt-cd-drive Sep 30 17:45:05 now I wonder why this package is being pulled in Sep 30 17:45:13 its a native package Sep 30 17:45:48 wait no its target package Sep 30 17:46:03 qemux86 Sep 30 17:47:56 machine does not have any CD drive Sep 30 17:47:57 khem: n900 build: NOTE: Running task 10102 of 12504, sofar good Sep 30 17:48:07 JaMa: cool Sep 30 17:48:19 JaMa:how long before you finish Sep 30 17:48:23 can it happen today Sep 30 17:50:12 recipes/gnome/gvfs_1.6.0.bb:DEPENDS = "libcdio... Sep 30 17:50:20 khem: I think all long packages are done.. so maybe an hour or so Sep 30 17:50:25 recipes/gstreamer/gst-plugins.inc:DEPENDS = "libcdio Sep 30 17:50:40 khem: I've started it today, because 2 full backups were running till then :/ Sep 30 17:50:55 JaMa: ok Sep 30 17:54:42 JaMa: if you are able to boot well then I will commit the gcc change Sep 30 17:55:27 khem: I was able to boot it after rebuild (when only your last commit was missing) Sep 30 17:55:41 khem: in case that's enough for proof Sep 30 17:55:54 JaMa: eh cool Sep 30 17:55:58 I did not know that Sep 30 17:58:04 khem: I had whole image built when you pushed that fix for mplayer build.. and that one worked Sep 30 17:59:13 cool Sep 30 17:59:28 so it seems safe to push those fixes Sep 30 18:39:03 khem: I don't think I proposed such a patch, but it seems like a good idea Sep 30 18:39:18 zecke: ok Sep 30 18:39:29 there are two versions of libsoup Sep 30 18:39:42 most recipes use 2.4 including webkit Sep 30 18:39:54 but some use older versions too Sep 30 18:40:14 and it seems libsoup versions can not not coexist Sep 30 19:09:03 Does anyone remember how matchbox-desktop is themed? Sep 30 19:09:05 The icons Sep 30 19:09:31 Do I just follow these naming guidelines and throw everything in /usr/share/pixmaps? Sep 30 19:09:32 http://matchbox-project.org/documentation/themes/ar01s06.html#id2465637 Sep 30 19:09:56 jo re Sep 30 19:10:50 lets see if I can finish pyqt today Sep 30 19:33:54 03Tom Rini  07org.openembedded.dev * ra9a0e0250a 10openembedded.git/recipes/sysvinit/sysvinit_2.86.bb: Sep 30 19:33:54 sysvinit: Add mountpoint to u-a Sep 30 19:33:54 Signed-off-by: Tom Rini Sep 30 19:34:04 03Tom Rini  07org.openembedded.dev * raf3f52db17 10openembedded.git/recipes/util-linux-ng/util-linux-ng.inc: Sep 30 19:34:04 util-linux-ng: Make cal and script use u-a Sep 30 19:34:04 Signed-off-by: Tom Rini Sep 30 19:35:02 hi khem I'll install the image Sep 30 19:35:14 GNUtoo|laptop: ok fingers crossed :) Sep 30 19:37:07 hehe Sep 30 19:37:13 jo effem Sep 30 19:37:45 hi woglinde, all Sep 30 19:41:00 03Frans Meulenbroeks  07org.openembedded.dev * rb6cc4bc217 10openembedded.git/classes/kernel.bbclass: (log message trimmed) Sep 30 19:41:00 kernel.bbclass: avoid NOTE messages for files that are staged but not packaged Sep 30 19:41:00 copy the files directly from the work area to staging. That way Sep 30 19:41:00 there are no unpackaged files in image/ so no NOTE messages. Sep 30 19:41:01 Additional advantage is that it is faster as we save the copy to the image dir. Sep 30 19:41:01 See also http://thread.gmane.org/gmane.comp.handhelds.openembedded/37487 Sep 30 19:41:02 Signed-off-by: Frans Meulenbroeks Sep 30 19:44:08 khem, it booted Sep 30 19:44:56 which machine? Sep 30 19:45:02 nokia900 Sep 30 19:45:08 hehe Sep 30 19:45:27 ~lart zecke for enabling qt-examples Sep 30 19:45:27 * ibot whips zecke with a wet and grimy noodle just because for enabling qt-examples Sep 30 19:51:40 03Frans Meulenbroeks  07org.openembedded.dev * r8809a602b9 10openembedded.git/recipes/wpa-supplicant/files/use-channel.patch: Sep 30 19:51:40 wpa-supplicant : moved unused files to obsolete dir Sep 30 19:51:40 Signed-off-by: Frans Meulenbroeks Sep 30 19:53:06 03Vladimir Sorokin  07org.openembedded.dev * r77de34d15b 10openembedded.git/recipes/pacemaker/pacemaker_1.0.9.1.bb: Sep 30 19:53:06 pacemaker: packaging fixes, move some files between packages Sep 30 19:53:06 Signed-off-by: Vladimir Sorokin Sep 30 19:53:06 Signed-off-by: Roman I Khimov Sep 30 19:53:06 03Vladimir Sorokin  07org.openembedded.dev * r5b45248c45 10openembedded.git/recipes/groff/groff_1.20.1.bb: Sep 30 19:53:06 groff-1.20: disable parallel make Sep 30 19:53:07 Fails from time to time. Sep 30 19:53:07 Signed-off-by: Vladimir Sorokin Sep 30 19:53:08 Signed-off-by: Roman I Khimov Sep 30 19:53:08 03Vladimir Sorokin  07org.openembedded.dev * r128966821f 10openembedded.git/recipes/corosync/ (corosync_1.2.8.bb files/init): Sep 30 19:53:09 corosync: fix init script Sep 30 19:53:09 Allows to get status from init script when no corosync.cfg exists. Sep 30 19:53:10 Signed-off-by: Vladimir Sorokin Sep 30 19:53:11 Signed-off-by: Roman I Khimov Sep 30 19:53:11 03Vladimir Sorokin  07org.openembedded.dev * r68481f1a67 10openembedded.git/recipes/linux-ha/cluster-glue_1.0.5.bb: Sep 30 19:53:12 cluster-glue: packaging fixes, move some files between packages Sep 30 19:58:14 khem, mplayer seem to work Sep 30 19:59:15 GNUtoo|laptop: cool Sep 30 20:00:07 GNUtoo|laptop: I will make the big move and push it of Sep 30 20:00:16 JaMa how far are you Sep 30 20:00:25 nice Sep 30 20:00:47 khem: mmt just reboot Sep 30 20:01:03 khem linaro patches? Sep 30 20:01:53 woglinde: yes Sep 30 20:02:08 once JaMa is able to boot then its out of the kitty Sep 30 20:04:41 you fixed enricos bug? Sep 30 20:04:42 khem: ok enlightenment started ok, but be aware I'm using kernel image from different build (so testing only user space) Sep 30 20:11:01 JaMa: userspace is where problems would have been Sep 30 20:11:09 the new patches should not effect kernel build Sep 30 20:11:20 woglinde: thats was not a bug at all Sep 30 20:11:31 he was using some broken local patches Sep 30 20:11:35 ah Sep 30 20:11:37 okay Sep 30 20:11:38 hehe Sep 30 20:11:43 ugh, the bitbake modules are so terribly intertwined and interdependent Sep 30 20:13:14 03Frans Meulenbroeks  07org.openembedded.dev * rd843dff07b 10openembedded.git/recipes/mpfr/dont_use_mips_h_constraint.patch: Sep 30 20:13:14 mpfr : moved unused files to obsolete dir Sep 30 20:13:14 Signed-off-by: Frans Meulenbroeks Sep 30 20:13:26 03Frans Meulenbroeks  07org.openembedded.dev * re29ef6f6bc 10openembedded.git/recipes/musicbrainz/files/libmusicbrainz-2.1.5-gcc43.patch: Sep 30 20:13:26 musicbrainz : moved unused files to obsolete dir Sep 30 20:13:26 Signed-off-by: Frans Meulenbroeks Sep 30 20:13:26 03Frans Meulenbroeks  07org.openembedded.dev * rea007eae79 10openembedded.git/recipes/mpfr/mpfr-canadian_3.0.0.bb: Sep 30 20:13:27 mpfr-canadian: fixed patching by adding FILESPATHPKG Sep 30 20:13:27 Signed-off-by: Frans Meulenbroeks Sep 30 20:13:28 03Frans Meulenbroeks  07org.openembedded.dev * r8baedf3a44 10openembedded.git/recipes/monotone/files/ (3 files): Sep 30 20:13:28 monotone : moved unused files to obsolete dir Sep 30 20:13:29 Signed-off-by: Frans Meulenbroeks Sep 30 20:13:29 03Frans Meulenbroeks  07org.openembedded.dev * ra820865e11 10openembedded.git/recipes/musicpd/mpd/fix-mod-support.patch: Sep 30 20:13:30 musicpd : moved unused files to obsolete dir Sep 30 20:13:31 Signed-off-by: Frans Meulenbroeks Sep 30 20:13:31 03Frans Meulenbroeks  07org.openembedded.dev * reef3871ee6 10openembedded.git/recipes/modphp/files/ (autotools.patch endianness.patch pear.patch): Sep 30 20:13:32 modphp : moved unused files to obsolete dir Sep 30 20:13:32 Signed-off-by: Frans Meulenbroeks Sep 30 20:13:33 03Frans Meulenbroeks  07org.openembedded.dev * r6981c76b60 10openembedded.git/recipes/mysql/ (2 files in 2 dirs): Sep 30 20:14:14 Signed-off-by: Frans Meulenbroeks Sep 30 20:15:43 a bunch of these bidirectional links too, server uses commands, commands uses server Sep 30 20:15:47 * kergoth mutters Sep 30 20:17:03 as far as i can tell, its not possible to write a script that just runs commands rather than having to use the server/ui mechanism Sep 30 20:17:08 that just seems wrong to me Sep 30 20:17:21 * kergoth wonders what a good next step is in untangling this Sep 30 20:17:36 * woglinde wonders too Sep 30 20:18:27 kergoth: is ui overstepping on to server Sep 30 20:19:03 alright here comes the big push Sep 30 20:19:10 03Khem Raj  07master * r3880477ea5 10openembedded.git/recipes/gcc/ (gcc-4.5.inc gcc-4.5/gcc-vmovl-PR45805.patch): Sep 30 20:19:10 gcc-4.5: Fix wrong instruction generation for vmovl pattern Sep 30 20:19:10 Signed-off-by: Khem Raj Sep 30 20:19:13 03Khem Raj  07master * r5c65c92e50 10openembedded.git/recipes/eglibc/ (7 files): (log message trimmed) Sep 30 20:19:13 eglibc: Fix eglibc-dev aliases and locales Sep 30 20:19:13 packages which had glibc-dev in their RDEPENDS Sep 30 20:19:13 caused glibc to be pulled along with eglibc. Sep 30 20:19:13 eglibc-dev was not properly providing glibc-dev Sep 30 20:19:13 Refresh the locale generation changes with glibc(poky) Sep 30 20:19:14 Add missing eglibc-static to PACKAGES Sep 30 20:19:21 03Khem Raj  07master * r28c05be128 10openembedded.git/recipes/meta/nylon-feed.bb: Sep 30 20:19:21 nylon-feed.bb: Use RDEPENDS instead of DEPENDS Sep 30 20:19:21 Signed-off-by: Khem Raj Sep 30 20:19:31 03Khem Raj  07master * re43cc49ace 10openembedded.git/recipes/prismstumbler/ (2 files in 2 dirs): Sep 30 20:19:32 prismstumbler_0.7.3+0.7.4pre1.bb: Add patch to replace gps_query call Sep 30 20:19:32 Signed-off-by: Khem Raj Sep 30 20:19:43 03Khem Raj  07master * rc5a2504372 10openembedded.git/recipes/xbmc/libcdio_0.81.bb: Sep 30 20:19:44 libcdio: Upgrade to 0.82 release Sep 30 20:19:44 * 0.81 is not cross-compiling safe. It tries to run generated programs Sep 30 20:19:44 to get version information to put into manpages which hangs these Sep 30 20:19:44 programs on virtual build machines which dont have cd drives. Sep 30 20:19:44 Signed-off-by: Khem Raj Sep 30 20:20:01 03Thilo Fromm  07master * r35686bb67b 10openembedded.git/recipes/qemu/qemu_0.12.5.bb: (log message trimmed) Sep 30 20:20:01 qemu-0.12.5: add ncurses, gnutls dependencies Sep 30 20:20:01 QEmu build breaks occassionally because some of the libraries required Sep 30 20:20:01 have not been built at the time qemu is being built. I ran into problems Sep 30 20:20:01 with the "qemu-native" build which tried to link against host libraries Sep 30 20:20:01 and subsequently failed: Sep 30 20:20:04 .... Sep 30 20:20:09 03Khem Raj  07master * r91aeba1aed 10openembedded.git/recipes/gcc/ (87 files in 2 dirs): Sep 30 20:20:09 gcc-4.5: Import Linaro patches Sep 30 20:20:09 * Tested gcc on efikamx. Sep 30 20:20:09 Signed-off-by: Khem Raj Sep 30 20:20:09 Marcin Juszkiewicz Sep 30 20:20:17 03Khem Raj  07master * r7a92e52370 10openembedded.git/recipes/gcc/ (6 files): Sep 30 20:20:17 gcc: Make use of INC_PR consistently. Sep 30 20:20:17 Signed-off-by: Khem Raj Sep 30 20:20:23 03Khem Raj  07master * r993fa56b77 10openembedded.git/recipes/gcc/ (12 files): Sep 30 20:20:23 gcc: Separate recipes for linaro gcc are redundant, remove Sep 30 20:20:23 * Linaro patches has been ported into gcc 4.5 in OE Sep 30 20:20:23 we dont need to have separate set of recipes. Sep 30 20:20:23 Signed-off-by: Khem Raj Sep 30 20:20:23 Marcin Juszkiewicz Sep 30 20:20:24 03Scott Garman  07master * r6ccae1e558 10openembedded.git/recipes/grep/ (3 files in 3 dirs): Sep 30 20:20:24 grep: update 2.6.3 to 2.7 Sep 30 20:20:25 Signed-off-by: Scott Garman Sep 30 20:20:25 Signed-off-by: Khem Raj Sep 30 20:20:45 * eFfeM could use another ack to for linux-uml removal, or a statement that someone still finds it useful. Sep 30 20:20:48 it appears that commands supports async ones, and in supporting those it interacts directly with the server to handle the callbacks Sep 30 20:20:51 see http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-September/024654.html Sep 30 20:20:54 03Graham Gower  07master * rf4d1cd6975 10openembedded.git/recipes/gdb/ (20 files): Sep 30 20:20:54 gdb: Don't clobber files installed by binutils. Sep 30 20:20:54 Also, move INC_PR into gdb-common.inc and propagate it into all recipes. Sep 30 20:20:54 Signed-off-by: Graham Gower Sep 30 20:20:54 Signed-off-by: Khem Raj Sep 30 20:20:55 03Fahad Usman  07master * rad833a8f47 10openembedded.git/recipes/python/ (python-setuptools-native_0.6c9.bb python-setuptools_0.6c9.bb): Sep 30 20:20:55 python-setuptools: mergerd native and target Sep 30 20:21:09 i could fix that api to let you register the callbacks for async commands instead, perhaps, and avoid the dep back on the server.. Sep 30 20:21:13 * kergoth ponders Sep 30 20:21:13 eFfeM: seems ok to me Sep 30 20:21:25 * merged native and target recipes and converted to new style staging Sep 30 20:21:26 * bumped PR Sep 30 20:21:26 * added BBCLASSEXTEND = "native" and NATIVE_INSTALL_WORKS = "1" Sep 30 20:21:26 Signed-off-by: Fahad Usman Sep 30 20:21:26 Signed-off-by: Khem Raj Sep 30 20:21:27 03Noor Ahsan  07master * r6c21af6a85 10openembedded.git/recipes/tcltk/tcl_8.5.8.bb: Sep 30 20:21:27 tcl: Run oe-stylize.py script and modify the recipe accordingly Sep 30 20:21:28 * Run the oe-stylize.py script on the recipe and modify the recipe according to the output of the script. Sep 30 20:21:28 Signed-off-by: Noor Ahsan Sep 30 20:21:57 Signed-off-by: Khem Raj Sep 30 20:21:57 03Graham Gower  07master * r9730a28e66 10openembedded.git/recipes/gtk+/ (files/xkb-ifdef.patch gtk+_2.20.1.bb): Sep 30 20:22:09 can anybody build mupdf for angstrom? Sep 30 20:22:47 gtk+_2.20.1.bb: Fix build with --disable-xkb. Sep 30 20:22:47 See https://bugzilla.gnome.org/show_bug.cgi?id=619114 Sep 30 20:22:47 Signed-off-by: Graham Gower Sep 30 20:22:48 Signed-off-by: Khem Raj Sep 30 20:22:48 03Graham Gower  07master * rb98c234bf7 10openembedded.git/recipes/ (4 files in 4 dirs): (log message trimmed) Sep 30 20:22:49 imlib_1.9.15, qiv_1.9: move to obsolete. Sep 30 20:22:49 Motivation for removal: Sep 30 20:22:50 ERROR: QA Issue with staging: imlib.pc failed sanity test (tmpdir) in path /mnt/oe/tmp/sysroots/mipsel-oe-linux/usr/lib/pkgconfig Sep 30 20:22:50 This 3 year old qiv recipe is the only in tree user of imlib1. Newer Sep 30 20:23:42 versions of qiv which use imlib2 are available from http://spiegl.de/qiv/ Sep 30 20:23:42 (38 lines omitted) Sep 30 20:24:07 http://tinderbox.openembedded.net/builds/91362/ Sep 30 20:24:19 mupdf is failed on -lfreetype Sep 30 20:28:15 khem, may I add your ack to it and make and push the patch with the two acks I have ? Or can you ack on the list ? Sep 30 20:28:46 add it Sep 30 20:28:52 jay7 I will look tomorrow into it Sep 30 20:29:09 woglinde: 10x Sep 30 20:29:19 khem, thanks Sep 30 20:29:55 it should be easy enough.. looks like lack of -Ldir or wrong staging of freetype Sep 30 20:30:02 I'm just not at home now Sep 30 20:30:09 -L I guess Sep 30 20:33:24 khem, btw did you see this thread? http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-September/024816.html Sep 30 20:33:40 I am not too happy by that solution (see my last reply in the thread) Sep 30 20:34:58 right, of course.. a typical repl won't work with bitbake/oe, not if you want to support async commands and not clutter up the display -- would need the ncurses ui to be solid Sep 30 20:35:02 * kergoth pokes at the ncurses ui Sep 30 20:41:55 03Noor Ahsan  07org.openembedded.dev * r5d7a31de21 10openembedded.git/recipes/libfribidi/ (fribidi_0.19.2.bb libfribidi-native_0.10.4.bb): Sep 30 20:41:56 fribidi: Extended recipe to build native package, add BBCLASSEXTEND Sep 30 20:41:56 * Add BBCLASSEXTEND = "native" to fribidi recipe Sep 30 20:41:56 * Bump PR to "r1" Sep 30 20:41:56 * Remove libfribidi-native_0.10.4.bb Sep 30 20:41:56 Signed-off-by: Noor Ahsan Sep 30 20:41:56 Signed-off-by: Tom Rini Sep 30 20:42:01 03Noor Ahsan  07org.openembedded.dev * r9567a8885b 10openembedded.git/recipes/sqlite/sqlite3_3.6.23.1.bb: Sep 30 20:42:01 sqlite3: Extend target recipe to native recipe using BBCLASSEXTEND Sep 30 20:42:01 * Set BBCLASSEXTEND to "native" to use target recipe as native recipe Sep 30 20:42:01 * In include file tcl-native was present in DEPENDS list, remove it from DEPENDS list by setting DEPENDS_virtclass-native = "readline-native ncurses-native" Sep 30 20:42:03 * Bump PR "${INC_PR}.1" Sep 30 20:42:03 Signed-off-by: Noor Ahsan Sep 30 20:42:03 Signed-off-by: Tom Rini Sep 30 20:42:10 03Noor Ahsan  07org.openembedded.dev * rb673977996 10openembedded.git/recipes/sqlite/ (sqlite-native_2.8.17.bb sqlite_2.8.17.bb): Sep 30 20:42:10 sqlite: Merge target and native recipe using BBCLASSEXTEND, Remove native recipe Sep 30 20:42:10 * Merge target and native recipes using BBCLASSEXTEND, Set BBCLASSEXTEND = "native" Sep 30 20:42:10 * Bump PR to "r7" Sep 30 20:42:10 * Remove sqlite-native_2.8.17 recipe Sep 30 20:42:10 Signed-off-by: Noor Ahsan Sep 30 20:42:11 Signed-off-by: Tom Rini Sep 30 20:42:20 03Noor Ahsan  07org.openembedded.dev * r537d40d95f 10openembedded.git/recipes/uicmoc/uicmoc-native_2.3.10.bb: (log message trimmed) Sep 30 20:42:20 uicmoc-native: Convert to new style staging, Remove do_stage Sep 30 20:42:20 * Convert do_stage to do_install function Sep 30 20:42:20 * Convert ${STAGING_BINDIR} to ${D}${bindir} Sep 30 20:42:21 still no sqlite-3.7? Sep 30 20:42:21 * Set NATIVE_INSTALL_WORKS = "1" Sep 30 20:42:22 * Bump PR to "r5" Sep 30 20:42:22 Signed-off-by: Noor Ahsan Sep 30 20:42:24 *sigh* Sep 30 20:52:14 okay, have a plan of attack to fix up ncurses, now to find the motivation to work on it.. Sep 30 20:52:58 go go go Sep 30 20:53:02 we don't need to implement interactive mode directly, at all. All we need to do is fix up the ncurses UI to actually work properly, and move the "Nothing to do" message from the cooker into the UI Sep 30 20:53:19 richard intended ncurses to be an interactive setup, there's a command input window Sep 30 20:53:21 its just not hooked up Sep 30 20:53:25 woglinde: thanks Sep 30 20:53:34 khem: who is using old libsoup? Sep 30 20:53:38 ~lart zecke again Sep 30 20:53:39 * ibot follow's zecke again with a gauntlet and ... scratch ... HUMILIATION Sep 30 20:54:19 woglinde: well, we can safe the time by not building directfb, gstreamer and this crap. :) Sep 30 20:54:48 woglinde: you should lart GNOME developers... evince is blocking here on a synchronous dbus call to GConf... Sep 30 20:56:50 zecke: geoclue gpe-calendar gpe-scap libsyncml rhythmbox starling Sep 30 20:56:57 gconf should die anyway Sep 30 20:57:02 hm Sep 30 20:57:12 oh... very old... Sep 30 20:57:23 zecke: can they be safely upgraded or removed Sep 30 20:57:24 woglinde: yeah, it does the same thing on GSettings... Sep 30 20:57:27 why the hell pyqt uses libqpycore.a for Sep 30 20:57:41 khem: api/abi is different... i fear that it is a problem Sep 30 20:58:07 zecke: thought so Sep 30 20:58:32 zecke: best bet is to make it co-exist Sep 30 21:00:09 03Frans Meulenbroeks  07org.openembedded.dev * refc6e6b952 10openembedded.git/recipes/linux-uml/ (17 files in 4 dirs): (log message trimmed) Sep 30 21:00:09 linux-uml: removed Sep 30 21:00:09 There are two recipes in this dir, and neither fetches properly. Sep 30 21:00:09 Also the recipes are for very old kernels. Sep 30 21:00:09 (UML is nowadays mainlined) Sep 30 21:00:09 see http://lists.linuxtogo.org/pipermail/openembedded-devel/2010-September/024654.html Sep 30 21:00:10 and replies Sep 30 21:00:11 zecke: I built webkit-gtk and then after that a whole bunch of stuff got built then I went ot rebuild webkit only to find that one of the above has installed older version of libsoup and webkit wont rebuild Sep 30 21:00:18 03Andreas Oberritter  07org.openembedded.dev * r617ac27e70 10openembedded.git/recipes/cdparanoia/ (cdparanoia_10.2.bb cdparanoia_svn.bb): (log message trimmed) Sep 30 21:00:18 cdparanoia: create libcdparanoia packages Sep 30 21:00:18 * Create separate packages for the libraries, which Sep 30 21:00:18 use a different license than the executable: Sep 30 21:00:18 "The cdparanoia command line tool versions 10.1 and later are Sep 30 21:00:19 distributed under the GNU General Public Licence v2.0 or, at your Sep 30 21:00:19 option, any later version of the GPL." Sep 30 21:01:09 khem: we should be fine with different includedirs? pkg-config files are already different Sep 30 21:01:35 khem: hmm... incluedir should be already different Sep 30 21:03:05 zecke: ok checking for LIBSOUP... configure: error: Package requirements (libsoup-2.4 >= 2.28.2) were not met: Sep 30 21:03:19 Consider adjusting the PKG_CONFIG_PATH environment variable Sep 30 21:03:26 Alternatively, you may set the environment variables LIBSOUP_CFLAGS Sep 30 21:03:34 and LIBSOUP_LIBS to avoid the need to call pkg-config Sep 30 21:03:46 even package names are different libsoup vs. libsoup-2 Sep 30 21:04:09 sorry, need to bring my partner to the trainstation, she is going back to Taiwan Sep 30 21:04:12 zecke: names dont matter I think Sep 30 21:04:33 03Fahad Usman  07org.openembedded.dev * ree63893f3c 10openembedded.git/recipes/classpath/classpath-tools-native.bb: Sep 30 21:04:33 classpath-tools-native.bb: Converted to new style staging Sep 30 21:04:33 * Converted do_stage to do_install. Sep 30 21:04:33 * Added NATIVE_INSTALL_WORKS = "1" Sep 30 21:04:33 * Added PR = "r1" Sep 30 21:04:33 Signed-off-by: Fahad Usman Sep 30 21:04:34 Signed-off-by: Tom Rini Sep 30 21:04:34 its the content Sep 30 21:04:34 03Fahad Usman  07org.openembedded.dev * rbf605db758 10openembedded.git/recipes/pngcrush/ (pngcrush-native_1.7.9.bb pngcrush_1.7.9.bb): (log message trimmed) Sep 30 21:04:35 pngcrush_1.7.9: merged native and non-native recipes Sep 30 21:04:36 * removed the pngcrush-native_1.7.9.bb file Sep 30 21:04:36 * added NATIVE_INSTALL_WORKS = "1" Sep 30 21:04:37 * added BBCLASSEXTEND = "native" Sep 30 21:04:37 * added PR = "r1" Sep 30 21:04:38 Signed-off-by: Fahad Usman Sep 30 21:04:45 03Fahad Usman  07org.openembedded.dev * r1cdb3733cf 10openembedded.git/recipes/gnome/gconf-native_2.6.1.bb: (log message trimmed) Sep 30 21:04:45 gconf-native_2.6.1.bb: Converted to new style staging Sep 30 21:04:45 * Converted do_stage to do_install. Sep 30 21:04:45 * Added NATIVE_INSTALL_WORKS = "1" Sep 30 21:04:46 * Added PR = "r1" Sep 30 21:04:46 Note: Not merging it into the non-native recipe as this is a dummy recipe and it does not need all the regular stuff. Sep 30 21:04:47 Signed-off-by: Fahad Usman Sep 30 21:04:47 03Fahad Usman  07org.openembedded.dev * raaa6bfa82a 10openembedded.git/recipes/docbook-sgml-dtd/ (5 files): (log message trimmed) Sep 30 21:04:48 docbook-sgml-dtd-native: converted to new style staging Sep 30 21:04:48 * Converted do_stage to do_install. Sep 30 21:04:49 * Moved the catalog files from ${sysconfdir}/sgml to ${D}${sysconfdir}/sgml so that they get included in the package file resulting with the new style staging. Sep 30 21:05:14 * Added a function docbook_sgml_dtd_native_mangle, to correct the path in .cat file and added it to SYSROOT_PREPROCESS_FUNCS. Sep 30 21:05:14 * Added INC_PR = "r1" to the .inc file and PR = "${INC_PR}.0" to the .bb files Sep 30 21:05:14 Signed-off-by: Fahad Usman Sep 30 21:05:14 03Fahad Usman  07org.openembedded.dev * rfd4d29eed1 10openembedded.git/recipes/opie-i18n/ (3 files): Sep 30 21:05:14 opie-lrelease-native: Converted to new style staging Sep 30 21:05:15 * Converted do_stage to do_install. Sep 30 21:05:15 * Added NATIVE_INSTALL_WORKS = "1" Sep 30 21:05:16 * Added INC_PR = "r1" to the .inc file and PR = "${INC_PR}.0" to the .bb files Sep 30 21:05:16 Signed-off-by: Fahad Usman Sep 30 21:05:17 Signed-off-by: Tom Rini Sep 30 21:05:17 (28 lines omitted) Sep 30 21:05:32 khem: even the pkgconfig file should be libsoup-2.4.pc Sep 30 21:07:03 03Fahad Usman  07org.openembedded.dev * r44af05077f 10openembedded.git/recipes/glib-1.2/ (glib-1.2-native_1.2.10.bb glib-1.2_1.2.10.bb): (log message trimmed) Sep 30 21:07:03 glib-1.2: merged native and non-native recipes Sep 30 21:07:03 * removed the glib-1.2-native_1.2.10.bb file Sep 30 21:07:03 * added NATIVE_INSTALL_WORKS = "1" Sep 30 21:07:03 * added BBCLASSEXTEND = "native" Sep 30 21:07:03 * converted do_stage of native recipe to do_install_virtclass-native in the combined recipe Sep 30 21:07:04 * added DEPENDS_pn-glib-1.2 += "glib-1.2-native", per Vitus' suggestion Sep 30 21:08:14 zecke: yes it is Sep 30 21:08:26 khem: Re http://patchwork.openembedded.org/patch/2920/ Sep 30 21:08:28 zecke: but I think webkit is not reading the right one Sep 30 21:08:41 No, I don't think we can convert to do_install_append since that export modifies how autotools_.. goes Sep 30 21:08:45 I'll add a comment in the bb tho Sep 30 21:10:23 Tartarus: you can export even in do_install_prepend Sep 30 21:10:24 glib1 die die Sep 30 21:10:35 calling it a day, cya all tomorrow & have fun! Sep 30 21:10:42 nite effem Sep 30 21:11:02 khem: Yeah Sep 30 21:11:06 but that's even sillier I think Sep 30 21:11:17 do_install_prepend() 1 line, do_install_append() N lines Sep 30 21:11:26 All to avoid calling out autotools_do_install Sep 30 21:11:29 hm seems I have now emultae pyqt configure.py with qmake stuff and the recipe Sep 30 21:12:55 hm... I see more cases of missing gettext: http://pastebin.ca/1952138 Sep 30 21:12:56 03Fahad Usman  07org.openembedded.dev * rfe6da768cb 10openembedded.git/recipes/omniorb/ (omniorb-native_4.1.4.bb omniorb_4.1.4.bb): (log message trimmed) Sep 30 21:12:56 omniorb_4.1.4: merged native and non-native recipes Sep 30 21:12:56 * removed the omniorb-native_4.1.4.bb file Sep 30 21:12:56 * added DEPENDS_pn-omniorb = "omniorb-native" and DEPENDS_pn-omniorb-native += Sep 30 21:12:56 "python-native" to handle different dependencies for native and target Sep 30 21:12:56 recipes, thanks to Vitus Jensen for guidance Sep 30 21:12:57 * added NATIVE_INSTALL_WORKS = "1" Sep 30 21:13:46 re gnutoo Sep 30 21:14:46 hi Sep 30 21:14:46 khem, was it you who suggested to someone the other day using amend.inc and something else? wanted to read your blog entry again Sep 30 21:15:21 seems to work Sep 30 21:15:25 yeah Sep 30 21:15:31 at least for compiletime Sep 30 21:17:41 tharvey: bbappend Sep 30 21:17:54 tharvey: http://sakrah.dontexist.org Sep 30 21:18:12 thx, that was the blog I wanted to read again - is that your blog? Sep 30 21:18:20 yes Sep 30 21:19:38 khem new benchmark tests with linaro stuff now? Sep 30 21:19:42 and uclibc? Sep 30 21:20:21 we have to benchmark more Sep 30 21:20:30 khem: heh, I just sent a note to the ML about the bbappend stuff :-) Sep 30 21:20:38 so bbappend seems to have the exact same effect as inheriting amend.bbclass and having an amend.inc in FILESBASEPATH correct? - bbappend being an improvement because its automatic? Sep 30 21:20:59 Tartarus: hey Tom, would you pls give impulse to investigations about the missing inherit gettext? http://pastebin.ca/1952138 Sep 30 21:22:12 then there is a strange issue with globs, I think : http://pastebin.ca/1952148 Sep 30 21:23:52 Still seeing that 2nd pastebin? Sep 30 21:24:01 I thought that got fixed rather recently Sep 30 21:24:05 Not sure about the gettext one Sep 30 21:24:50 logs are 25 sept. pull Sep 30 21:24:59 I'll retry tonight Sep 30 21:25:17 ant they will not have changed Sep 30 21:25:53 af5d0dbecc11ada0d3255da195b4674dd6e08782, no? Sep 30 21:25:58 didn't check one by one..I trust you :) Sep 30 21:26:00 Which is odd since that should be in the build :( Sep 30 21:27:36 hm Sep 30 21:29:22 khem, not quite understanding the bblayers and bbappend examples at http://sakrah.homelinux.org/node/2 - it seems to me both these mechanisms already existed in OE: COLLECTIONS have been there a long time - I'm not understanding the diff betwen bblayers and collections. As far as providing your own /etc/network/interfaces - modifying FILESPATHBASE to prepend your own path has always worked to allow you to use your own localfil Sep 30 21:29:22 es instead of whats in a recipe Sep 30 21:43:30 I have a bitbake package selection question. Sep 30 21:47:27 I can't get bitbake to select a u-boot other than root. Sep 30 21:49:17 other than 'root' - what do you mean? Sep 30 21:52:10 err Sep 30 21:52:12 git Sep 30 21:52:14 tharvey: could be, its one use case of it Sep 30 21:52:29 so if I do bitbake u-boot-2009.08 Sep 30 21:52:37 it tries to build BOTH git and 2009.08 Sep 30 21:52:54 tharvey: I could get the complete recipe unmodified from OE and just let it use one file from my overlay Sep 30 21:52:55 ( I don't want to build git because the recipe is failing ) Sep 30 21:53:25 tharvey: with collections I had to copy the whole recipe into my overlay and then modify it Sep 30 21:54:17 tharvay and bblayers and bbappend does same that collections did infact its another mechanism to maintain overlays Sep 30 21:54:47 dante, set PREFERRED_VERSION_uboot Sep 30 21:55:00 tharvey: bbappend is a stub that will get added to original .bb Sep 30 21:55:09 by bitbake Sep 30 21:55:34 ok, for my understanding of bitbake .. is there a reason that it is ignoring the version I have on my request? Sep 30 21:56:15 khem, if the one file you want to modify is in SRC_URI as file:// then FILESBASEPATH takes care of it Sep 30 21:56:54 tharvey: true how would you do it if you did not have bbappend Sep 30 21:56:55 re bbappend I see value in it for sure, just wondering if its different than amend.inc used when inheriting amend.bbclass? Sep 30 21:56:57 dante: it's a bug. Sep 30 21:57:05 ah ok. Sep 30 21:57:28 tharvey: it's supported directly by bitbake, which means a number of things. performance is improved, less work is done (only needs to finalize (apply overrides, etc) once), etc Sep 30 21:58:04 the downside is the fact that the filename has to match directly, which means if upstream prefs change, you get ad ifferent version, and suddenly your overlay's changes aren't applied to the version you're building anymore.. Sep 30 21:58:07 that's my biggest concern with it Sep 30 21:58:09 khem, in my local.conf I have FILESPATHBASE =. "${OEROOT}/myfiles/files:", then I have for example myfiles/files/netbase/interfaces which netbase uses instead of its interfaces Sep 30 21:59:02 kergoth_, your refering to bbappends right? so its simply a bit more efficient than amend.bbclass and preferable correct? Sep 30 21:59:33 you're downplaying the benefit, its not "a bit" more efficient Sep 30 22:00:01 in addition, BBCLASSEXTEND should work with it, which it doesn't with amend.inc today Sep 30 22:00:01 tharvey: if you have more than one thing to override then you will end up cluttering you local.conf Sep 30 22:00:25 khem: hm? no, you wouldn't Sep 30 22:00:28 no worries - I'm just asking. I often find multiple ways of doing things and wonder what the preferred method is these days. Seems like amend.bbclass should perhaps be deprecated then? Sep 30 22:00:37 bbappend is only in master Sep 30 22:00:39 not even 1.10 Sep 30 22:00:44 and bitbake isn't even bumped to require 1.10 yet Sep 30 22:00:48 khem, no, I override local files all the time... no more cluter than what your talking about Sep 30 22:00:54 so at some point far in the future, yes, we'll deprecate it Sep 30 22:01:04 s/bibake isn't/OE isn't/ Sep 30 22:01:38 kergoth_, ahh... bbpapend is a 'bitbake' enhancement? ok that makes sense now Sep 30 22:01:49 perhaps i wasn't clear Sep 30 22:01:51 tharvey: so you put all your files in one single dir ? Sep 30 22:02:04 when i said "it's supported directly by bitbake" i meant *directly* Sep 30 22:02:14 khem, no, its per package, per package-name, etc Sep 30 22:02:27 khem, uses OVERRIDES Sep 30 22:02:28 khem: read FILESPATH in bitbake.conf. Sep 30 22:02:36 OVERRIDES = "local:${MACHINE}:${DISTRO}:${TARGET_OS}:${TARGET_ARCH}:build-${BUILD_OS}:fail-fast:pn-${PN}:${FEED_ARCH}" Sep 30 22:02:41 FILESPATHBASE is only the first component Sep 30 22:02:56 FILESPATHPKG is then built on top of every base, and every override on top of all those combinations Sep 30 22:03:21 the default FILESPATHBASE is only FILE_DIRNAME Sep 30 22:03:29 khem, its very clean - only issue is for things you want to override which are 'not' a localfile... Sep 30 22:03:30 kergoth_: true I confused with myfiles/files: Sep 30 22:04:05 tharvey: not a local file ? Sep 30 22:04:09 yeah, its only useful for file://, doesn't help with anything else Sep 30 22:04:13 what that would be Sep 30 22:04:16 oh urls Sep 30 22:04:20 right Sep 30 22:04:39 anything that isn't file:// files. if it installs a config file that it gets from ${S}, not filespath, that'd require a metadata change to alter it Sep 30 22:04:48 kergoth_: with bbappend I guess we could change SRC_URIs Sep 30 22:05:03 same with amend.inc, yes Sep 30 22:05:07 for example I just ran into trouble because I wanted a custom /etc/sysconfig/gpelogin yet thats not a localfile it comes from the source tarball of gpe-login Sep 30 22:05:47 tharvey: anything that comes from source tar is not metadata I consider Sep 30 22:05:48 I'm using the amend.inc method to apply any patches to handle what I can't with FILESPATHBASE Sep 30 22:05:48 tharvey: amend.inc bumps ram usage by about 14%, and slows parse time as well due to having to check for existance of all the potentially existent files Sep 30 22:05:57 prior to the recent amend.bbclass change, it nearly doubled the up front ram usage Sep 30 22:06:04 for some very strange reason Sep 30 22:06:13 kergoth_: where am I supposed to set this PREFERRED_VERSION_uboot? Sep 30 22:06:23 ya, I'll move to using bbappend for sure Sep 30 22:06:26 dante: uh, the same place you set everything else? Sep 30 22:06:38 dante, you set that in one of your conf files, such as local.conf Sep 30 22:06:39 dante: its a bitbake variable, its local to your build, put it in local.conf. Sep 30 22:07:06 kergoth: is BBPATH the list of dirs under which bb will look for metadata ? Sep 30 22:07:09 tharvey: let me know if you figure out a solution to what i mentioned before, other than a distro wide version lockdown :\ Sep 30 22:07:19 I had a local.inc .. lets see if that changes things. Sep 30 22:07:35 local.inc isn't included by anything, unless you did so yourself in a config file thats already included. Sep 30 22:07:56 So... hunh. font-util-native needs 4200ish tasks to build Sep 30 22:08:02 kergoth_: which problem the file renaming ? Sep 30 22:08:02 I think somewhere there's a wrong dep :) Sep 30 22:08:29 ok .. apparently .. it's not including my local.conf Sep 30 22:09:00 kergoth_, the issue that .bbappend has to match the package name exactly and if package version bumps you need to re-apply? - that may not be a bad downside as you 'should' re-examine your append in the case of a new version Sep 30 22:09:19 tharvey: perhaps you have time to review every commit when updating.. Sep 30 22:09:23 but I would agree its a pain to try to track these changes all the time Sep 30 22:09:28 kergoth_: is there a good place to learn bitbake before it drives me insane? Sep 30 22:09:30 exactly heh Sep 30 22:09:33 the problem is the failure mode Sep 30 22:09:35 rather than erroring Sep 30 22:09:39 it just silently doesn't apply your changes Sep 30 22:10:03 dante: bitbake user manual, oe user manual.. Sep 30 22:10:32 ok, the bitbake user manual seemed .. lacking? Sep 30 22:10:43 not particularly Sep 30 22:10:47 it covers bitbake alone. Sep 30 22:10:50 the way I've been supporting OE for my project is to lock down a specific snapshot of OE, so I don't pull very often (not often enough) - looking for ways to make the pull's less work to sync up Sep 30 22:10:51 bitbake isn't what includes local.conf. Sep 30 22:10:54 OE's bitbake.conf does that Sep 30 22:11:17 currently I just compare package lists from my image before/after pull and check my overrides for anything that bumped versions Sep 30 22:12:00 the big lack i see in bitbake's docs are that its really only reference material and a few examples, almost no conceptual information, and no real beginning to end tutorials on its usage Sep 30 22:12:06 kergoth: may be there could be a bitbake note about form where a file:// is being picked Sep 30 22:12:16 dante, its probably not bitbake driving you insane... I think its more a matter of understanding OE's use of bitbake - where confs get included like kergoth_ metioned Sep 30 22:12:18 so I've put a local.conf where the local.conf.sample is, and put the PREFERRED_VERSION_uboot="2010.06" Sep 30 22:12:29 khem: you mean like http://bitbake.berlios.de/manual/ch03s02.html ? Sep 30 22:12:30 and it still tries to build the git version Sep 30 22:12:38 agreed regarding docs Sep 30 22:13:02 dante, not where local.conf.sample is Sep 30 22:13:14 dante: read the OE user manual. Sep 30 22:13:19 i mentioned both for a reason Sep 30 22:13:29 needs to be in BBPATH Sep 30 22:13:56 kergoth_: sounds good, I'll read that then. Sep 30 22:13:58 where local.conf.sample is will work too, as long as there isn't a conf/local.conf in a high priority location in bbpath Sep 30 22:14:16 dante, typically you will have an environment setup - something that you source in your shell which sets BBPATH... typically you would add a path to a build directory Sep 30 22:14:44 dante, the issue is that we don't know what you have setup for your local configuration so we can't answer you directly... there are many guides out there that setup BBPATH differently Sep 30 22:15:27 and of course, this is one of oe/bitbake's biggest usability problems Sep 30 22:15:29 if you followed http://wiki.openembedded.net/index.php/Getting_Started then you would put your local.conf in "/stuff/build" Sep 30 22:15:32 there are too damn many ways to do things Sep 30 22:15:38 flexibility is nice, but this is just excessive. Sep 30 22:15:47 heh Sep 30 22:15:48 heh... which is why I asked about bbappend vs amend.bbclass Sep 30 22:15:53 indeed Sep 30 22:16:16 khem: last change in eglibc doesn't compile here (in locales) http://tinderbox.openembedded.net/packages/802402/ any hint? Sep 30 22:16:25 i always figured we'd get something like amend into bitbake directly, was just a matter of time, the functionality is too useful, and the implementation via metadata had issues Sep 30 22:16:30 I can't keep up daily... I tend to try to get up to date every couple of months which is very painful as there is no snapshots and release changes to follow Sep 30 22:16:55 03Tom Rini  07org.openembedded.dev * ra1506f34f4 10openembedded.git/recipes/libffi/libffi-native_2.0+gcc4.3.2.bb: Sep 30 22:16:55 libffi-native: Convert from legacy staging Sep 30 22:16:55 Signed-off-by: Tom Rini Sep 30 22:16:56 tharvey: i find it helps to look over the git log just for conf/ and classes/ Sep 30 22:17:00 since those are the highest impact Sep 30 22:17:05 03Tom Rini  07org.openembedded.dev * r18680edeca 10openembedded.git/recipes/mysql/mysql5-native_5.1.40.bb: Sep 30 22:17:05 mysql5-native: Convert from legacy staging, disable pstaging Sep 30 22:17:05 Signed-off-by: Tom Rini Sep 30 22:17:05 03Tom Rini  07org.openembedded.dev * r532afcc89f 10openembedded.git/recipes/apache2/ (apache2-native_2.2.14.bb apache2-native_2.2.3.bb): Sep 30 22:17:06 apache2-native: Convert from legacy staging Sep 30 22:17:06 Signed-off-by: Tom Rini Sep 30 22:17:07 03Tom Rini  07org.openembedded.dev * r53ddbbe512 10openembedded.git/recipes/dcop/dcopidl2cpp-native_3.5.8.bb: Sep 30 22:17:08 dcopidl2cpp-native: Convert from legacy staging Sep 30 22:17:08 Signed-off-by: Tom Rini Sep 30 22:17:08 for the oe side of things anyway Sep 30 22:17:09 03Tom Rini  07org.openembedded.dev * redd73eef19 10openembedded.git/recipes/htmldoc/ (htmldoc-native_1.9.x-r1571.bb htmldoc_1.9.x-r1571.bb): Sep 30 22:17:09 htmldoc-native: Switch to BBCLASSEXTEND, drop legacy staging Sep 30 22:17:10 Signed-off-by: Tom Rini Sep 30 22:17:10 03Tom Rini  07org.openembedded.dev * r4e7a0d99b6 10openembedded.git/recipes/perl/ (2 files): Sep 30 22:17:11 liblocale-gettext-perl: Switch to BBCLASSEXTEND Sep 30 22:17:11 Signed-off-by: Tom Rini Sep 30 22:17:19 dante, so in that wiki guide, the step called 'Create local configuration' where you copy local.conf.sample into build/conf/local.conf and edit - thats the 'conf' we are talking about Sep 30 22:17:43 but if you follow along in that wiki you can see that you can put any number of directories in the path where your local.conf may reside Sep 30 22:18:05 good point on conf/classes Sep 30 22:18:42 course, you have to ignore all the random changes to some distro you don't care about in conf/distro/ Sep 30 22:18:45 but its a start Sep 30 22:18:51 heh Sep 30 22:18:56 tharvey: gotcha. I think I have it now .. I just got very frustrated when I was trying to do the version by hand at the commandline Sep 30 22:19:06 yeah... don't do that, its broken :) Sep 30 22:19:09 and looking though the bb files and being like "WHY IS IT IGNORING MY VERSION?!?!" Sep 30 22:19:15 JaMa: bitbake -c clean eglibc eglibc-initial gcc-cross gcc-cross-initial gcc-cross-intermediate; bitbake eglibc Sep 30 22:19:20 thanks. Sep 30 22:20:40 JaMa: in worst case you might have to rebuild Sep 30 22:22:53 khem: ok, will try Sep 30 22:24:25 khem: I was just surprised, because I've built this dir from scratch 2 days ago (already with linaro patched gcc and only change now was Make use of INC_PR consistently.) Sep 30 22:25:07 JaMa: I ended up in same error but then cleaning gcc and eglibc helped Sep 30 22:25:12 but then I also deleted all Sep 30 22:25:51 oki Sep 30 22:25:58 thanks Sep 30 22:33:42 good night Sep 30 22:43:15 * kergoth_ glares at bitbake Sep 30 23:00:34 03Tom Rini  07org.openembedded.dev * r718dfa9145 10openembedded.git/recipes/inetutils/ (inetutils_1.5.bb inetutils_1.8.bb): Sep 30 23:00:34 inetutils: Add hostname to u-a. Sep 30 23:00:34 Signed-off-by: Tom Rini Sep 30 23:15:11 khem: cleaning whole toolchain and building again helped me too, thanks again Sep 30 23:16:44 hmmmm Sep 30 23:24:24 JaMa: cool np Sep 30 23:35:45 03Chase Maupin  07org.openembedded.dev * ra6d0aa94fe 10openembedded.git/recipes/ti/ (3 files): Sep 30 23:35:45 matrix-gui-common: make versioned recipes Sep 30 23:35:45 * Abstracted common code to .inc file Sep 30 23:35:45 * Made versioned recipe for version 1.1 and 1.2 Sep 30 23:35:45 * Removed the _svn version of the recipe. Sep 30 23:35:45 Signed-off-by: Chase Maupin Sep 30 23:35:45 Signed-off-by: Denys Dmytriyenko Sep 30 23:36:00 03Chase Maupin  07org.openembedded.dev * r7a8f11af65 10openembedded.git/recipes/ti/ (matrix-gui_svn.bb matrix-gui_1.1.bb matrix-gui_1.2.bb): Sep 30 23:36:00 matrix-gui: make versioned recipes Sep 30 23:36:00 * Abstracted common code to .inc file Sep 30 23:36:00 * Made versioned recipe for version 1.1 and 1.2 Sep 30 23:36:00 * Removed _svn version of the recipe Sep 30 23:36:00 Signed-off-by: Chase Maupin Sep 30 23:36:01 Signed-off-by: Denys Dmytriyenko Sep 30 23:36:02 03Chase Maupin  07org.openembedded.dev * rd8d91cad17 10openembedded.git/recipes/ti/am-benchmarks_svn.bb: Sep 30 23:36:02 am-benchmarks: update to latest version Sep 30 23:36:03 * Updated the SRCREV to the latest version of this project Sep 30 23:36:03 which includes support for multiple platforms. Sep 30 23:36:04 * Add COMPATIBLE_MACHINE setting for recipe. Sep 30 23:36:04 Signed-off-by: Chase Maupin Sep 30 23:36:05 Signed-off-by: Denys Dmytriyenko Sep 30 23:36:05 03Chase Maupin  07org.openembedded.dev * r76f664fbe8 10openembedded.git/recipes/ti/ (matrix-tui_svn.bb matrix-tui_1.0.bb): (log message trimmed) Sep 30 23:36:06 matrix-tui: create versioned matrix-tui recipe Oct 01 01:26:37 good evening all Oct 01 01:27:29 I rebuilt glibc, but when issuing "bitbake x11-gpe-image" I still receive "unknown package locale-base-en-us". Oct 01 01:27:41 any idea what is going on? Oct 01 01:40:48 anybody here? Oct 01 01:41:24 i but i can't help you sorry..wait more Oct 01 01:45:06 what machine config should i use for a PXA272 device thats not in the list ? Oct 01 01:45:18 what device? Oct 01 01:45:27 pocket loox 720 Oct 01 01:45:33 i use armv5te for my PX270 Oct 01 01:45:40 ops Oct 01 01:45:43 as target? Oct 01 01:45:43 machine = a1200 Oct 01 01:45:54 ah Oct 01 01:46:25 so that should work for any? Oct 01 01:46:32 PXA27x Oct 01 01:46:48 i think yes.. Oct 01 01:47:55 in the a1200 config i just find require conf/machine/include/motorola-ezx-base.inc Oct 01 01:51:47 em-x270.conf seems to be a generic PXA272 board with the same specs :) Oct 01 01:52:36 motorola devices uses pxa 270 you can try Oct 01 01:52:42 on wait for more help Oct 01 01:52:44 bye all Oct 01 01:53:49 anybody else ;-) Oct 01 01:57:55 IgorK: alright whats your distro/machine Oct 01 02:42:58 khem: Im building angstrom/x11-gpe for mini2440 device (armv4t) with armv4t-angstrom-linux-gnueabi-gcc... **** ENDING LOGGING AT Fri Oct 01 02:59:57 2010