**** BEGIN LOGGING AT Mon Nov 08 02:59:57 2010 Nov 08 05:25:00 if in a recipe i want to make a shared library and also have it available in the staging directory for other recipes Nov 08 05:25:47 should i copy the library to my target directory and the staging directory and run ln on them seperately Nov 08 05:26:27 or is there a way to copy the shared library from the target directory to the staging directory while maintaing the link Nov 08 06:58:33 Good morning! Nov 08 07:00:17 I am trying to cross compile Free Pascal programs to be installed on an x11-gpe-image. Nov 08 07:01:51 So, which is the best recipe to bake the libraries required on the build host? Nov 08 07:04:08 The cross compile requires things like crti.o and stuff... Nov 08 07:07:46 I tried manually copying files in sysroot/armxxx/usr/lib to the build host, but this seemed not a very elegant way to do things. Nov 08 07:54:31 good morning Nov 08 08:22:07 mornin Nov 08 08:26:41 Jin^eLD: hi Nov 08 08:31:51 good morning yall Nov 08 08:32:49 i, yet again, have a weird problem Nov 08 08:33:28 turning on DEBUG causes the board not to boot Nov 08 08:34:40 even adding one extra printf breaks things Nov 08 08:34:50 Is there a page/man/file explaining what FILESPATHPKG does ? Nov 08 08:48:50 morgen Nov 08 08:50:04 hrw, hi Nov 08 08:54:15 hrw: gm Nov 08 12:54:54 hmm Nov 08 12:54:58 this question might be a bit stupid Nov 08 12:55:15 heh Nov 08 12:55:18 jsut ask Nov 08 12:55:19 but do i need a special ARM gdb? or can i just use the gdb that comes by default with my distro? Nov 08 12:55:54 not sure, but I suspect you need one that knows abut arm Nov 08 12:56:33 Im trying to do remote debugging Nov 08 12:56:58 so start gdb server on one end, and do "target remote..." etc on the client Nov 08 13:00:19 and your client's gdb is cross gdb with arm support? Nov 08 13:01:03 nah i dont think so Nov 08 13:13:23 Hi, what options do I have for building a graphical application (single app) on an embedded device? I have looked at Qt embedded, Nano-X and EFL. Would it be EFL and DirectFb ?? Nov 08 13:15:17 I had EFL working on directfb Nov 08 13:16:13 but the evas/ecore backend is in bad shape Nov 08 13:16:33 when was that? Nov 08 13:16:38 probably not well optimized Nov 08 13:16:44 I see a lot of EFL is now in beta status Nov 08 13:16:50 guyvdb_: ~6/8 months ago Nov 08 13:17:05 what other options do i have? Nov 08 13:17:39 it depends on your device, the kind of hardware accelerator you've access to Nov 08 13:17:54 I am looking at the Mini2440 Nov 08 13:18:21 what's that ? Nov 08 13:18:49 It is with the Samsung s3c2440 chip -- an arm 9 core Nov 08 13:20:08 guyvdb_: and what display size are you targeting ? Nov 08 13:20:22 a 7" 800x480 Nov 08 13:20:42 it is from InnoLux Nov 08 13:21:00 ok Nov 08 13:21:15 no graphical hardware accelerato Nov 08 13:21:16 r Nov 08 13:21:31 so you can basicly use whatever you want Nov 08 13:21:46 I do not think so.. I am looking for 2D, I think... that is why EFL looked appealing Nov 08 13:21:57 Xfbdev is good option Nov 08 13:22:09 after Qt/EFL/Gtk+ Nov 08 13:22:13 The application is a simple data entry app for selling tickets Nov 08 13:22:18 everything is going to work Nov 08 13:22:42 ok will read up on xfbdev Nov 08 13:22:47 thx Nov 08 13:22:50 gtk+/Qt is probably what you're looking for Nov 08 13:23:11 Hmm i only have 64M Ram / Nand flash Nov 08 13:23:37 yep Nov 08 13:23:38 will gtk+ not be too big. I was trying to stay away from x Nov 08 13:24:26 guyvdb_: gtk is fine... but the mini2440 is not that fast with these (comparably large displays) Nov 08 13:24:31 if you go for directfb, Qt is probably the only good choice Nov 08 13:24:40 03Koen Kooi  07org.openembedded.dev * r058ad87c02 10openembedded.git/recipes/gtk-engines/murrine_0.98.1.bb: murrine: update to 0.98.1 Nov 08 13:24:46 ahh Xfbdeb is an xserver Nov 08 13:24:55 yes Nov 08 13:25:22 and on the qt side should i target 4.7? Nov 08 13:25:43 doesnt X have a pretty large footprint compared to qt for example? Nov 08 13:26:06 I would think so... that is my assumption Nov 08 13:26:12 atiti: xfbdev is design to be small Nov 08 13:26:31 florian Nov 08 13:26:37 and directfb backends are not always well maintained (at least on gtk+/EFL) Nov 08 13:26:42 florian: do you have experiance on mini2440 Nov 08 13:27:00 guyvdb_: yes a little bit. Nov 08 13:27:15 guyvdb_: did you check the mini6410? Nov 08 13:27:17 what screen size / resolution did you target? Nov 08 13:27:50 Yea i looked at 6410 but application is very cost sensitive... .wanted to try with 2440 first Nov 08 13:27:51 i have two displays for it, the small qvga and a 7" 800x480 Nov 08 13:28:48 and what did you use for graphical display? Xfbdev? Nov 08 13:29:11 yes mostly... Nov 08 13:29:33 and is it more responsive at qvga than 800x480? Nov 08 13:29:33 i have to admit i didn't try efl Nov 08 13:29:43 yes Nov 08 13:30:15 Well i should have hardware soon i hope... still waiting on shipment... but also got 2 displays 4.3 & 7 " Nov 08 13:30:37 it should be easy to test it, iirc narcissus supports it so you can build angstrom images easily Nov 08 13:31:01 yea.. i have been looking at angstrom for it Nov 08 13:31:28 once i get hardware I will come here and pick your brain :) Nov 08 13:31:58 Has anyone tried nano-x Nov 08 13:32:40 guyvdb,you may also consider free pascal Nov 08 13:33:09 mfhk does it write to framebuffer? Nov 08 13:33:23 guyvdb_: isn't nano-x kind of out-dated ? Nov 08 13:33:34 it is RAD dev tool, not X windows system Nov 08 13:33:35 Is it? I don't know Nov 08 13:34:03 guyvdb_: I though xfbdev was the replacement for frambuffer based systems Nov 08 13:34:34 ahh ok ... Nov 08 14:37:44 Perhaps someone can point me in the right direction: I apparently screwed up something on a .conf or .bb, but I don't know which one. Nov 08 14:37:48 Bitbake just says: Nov 08 14:38:04 NOTE: Handling BitBake files: - (7092/7093) [99 %] Nov 08 14:38:15 ERROR: Command execution failed: Traceback (most recent call last): Nov 08 14:38:42 and just shows a trace of runAsyncCommand, updateCache, parse_next, raising ParsingErrorsFound Nov 08 14:38:51 How do I find out the offending file causing the problems? Nov 08 14:41:11 ~pastebin Nov 08 14:41:11 [~pastebin] A "pastebin" is a web-based service where you should paste anything over 3 lines so you don't flood the channel. Here are links to a few : http://www.pastebin.com , http://pastebin.ca , http://channels.debian.net/paste , http://paste.lisp.org , http://bin.cakephp.org/ , http://asterisk.pastey.net/ , or install pastebinit with yum or aptitude. Nov 08 14:41:22 Put the whole log up there, give us the url and we'll see if anything sticks out Nov 08 14:41:26 ah, ok Nov 08 14:43:59 Odd. Found the problem - my internal git server was down so the recipe was using AUTOREV so it was running git ls-remote to get the SRC_REV Nov 08 14:44:31 since it couldn't contact the server, it puked in a non-intuitive way. Nov 08 14:45:52 Would be nice if bitbake would say something about the git failure - but it does not. Just says ParsingErrorsFound at the end. Oh well. Nov 08 14:50:53 foerster: in the future, you could try using a different bitbake version, i'm pretty sure that bug only exists in master -- needs further investigation Nov 08 14:52:17 ok, thanks. I am on master right now. Would 1.10 be better to use? Nov 08 14:55:34 03Peter Chubb  07master * r01b85608d8 10bitbake.git/lib/bb/fetch/hg.py: (log message trimmed) Nov 08 14:55:34 Allow mercurial fetcher to follow tip Nov 08 14:55:34 There are occasions when developing when I want a package always to Nov 08 14:55:34 grab the latest copy of a package. Witht eh CVS fetcher you can do Nov 08 14:55:34 this by setting the `date' tag to `now'. This patch adds similar Nov 08 14:55:35 functionality to the mercurial fetcher: if the revision to fetch is Nov 08 14:55:36 `tip' then always grab from the server, and don't use the cached Nov 08 14:56:54 foerster: i wouldn't say "better", no, just there's always a slightly higher likelihood of bugs when using trunk/master of any project, this is no different Nov 08 14:59:44 kergoth: I switched to 1.10 branch for now, thanks. Nov 08 15:07:25 03Chris Larson  07master * rb1b0006059 10openembedded.git/recipes/autoconf/autoconf.inc: Nov 08 15:07:26 autoconf: don't build the documentation Nov 08 15:07:26 This kills one of the dependencies upon help2man. In the future, will switch Nov 08 15:07:26 this to be driven by a documentation feature in DISTRO_FEATURES. Nov 08 15:07:26 Signed-off-by: Chris Larson Nov 08 15:07:30 03Chris Larson  07master * r9018df571a 10openembedded.git/lib/oe/patch.py: Nov 08 15:07:30 oe.patch: for 'patch', always use -f Nov 08 15:07:30 Apparently quilt always ignores hunks that seem already applied, so let's do Nov 08 15:07:30 the same via -f. -t would probably be better, but isn't safe to use from a Nov 08 15:07:31 portability standpoint. Nov 08 15:07:31 Signed-off-by: Chris Larson Nov 08 15:07:32 03Chris Larson  07master * r4e8c767f3f 10openembedded.git/recipes/jack/jack_0.118.0.bb: Nov 08 15:07:32 jack: use the svn repository to fetch Nov 08 15:07:33 Signed-off-by: Chris Larson Nov 08 15:12:11 do we really need AUTOREV in jack? Nov 08 15:45:18 JaMa|Off: fair point, will fix Nov 08 16:09:27 does anyone knows why something like PREFERRED_PROVIDER_qt4-x11-free = "qt4-x11-custom" wouldn't work (I've got a strange bitbake dependency loop problem) Nov 08 16:10:39 any advice for generating read-only rootfs's with OE - there are a lot of postinst that want to run.... Nov 08 16:11:58 HopsNBarley, if it's possible then do your first boot with a RW partition then reboot in ro Nov 08 16:12:38 interesting idea. thanks. Nov 08 16:26:57 03Chris Larson  07master * r1f2731798a 10openembedded.git/recipes/jack/jack_0.118.0.bb: Nov 08 16:26:57 jack-0.118.0: fix SRCREV Nov 08 16:26:57 Signed-off-by: Chris Larson Nov 08 16:43:30 For the record it's because the recipe should have been named "qt4-x11-free-custom", not "qt4-x11-custom" Nov 08 16:53:18 does AUTOREV work for git too? Nov 08 16:53:23 * jconnolly could rtfm but thought I'd ask Nov 08 17:05:22 Hi, I don't quite get the toolchain generation process in OE Nov 08 17:05:56 what's the difference between gcc-native, gcc-cross, gcc-cross-intermediate, and gcc-cross-initial Nov 08 17:06:37 they all seem to serve the same purpose Nov 08 17:07:39 can someone explain me why I am wrong ? Nov 08 17:09:48 julian_: gcc-native is gcc built for your host machine Nov 08 17:10:01 gcc-cross is gcc that procuces code for target machine Nov 08 17:10:13 right Nov 08 17:10:30 initial and intermediate are stages in building gcc-cross I think Nov 08 17:10:32 both gcc-native and gcc-cross are running inside the host though Nov 08 17:10:37 yes Nov 08 17:11:01 but gcc-native can't produce code for target machine Nov 08 17:11:13 (if target isn't the same as host) Nov 08 17:11:18 thx for the explanation Nov 08 17:12:03 thing is, I would like to be able to build a multilib gcc-cross Nov 08 17:12:41 but I don't really know which bb file I'm supposed to modify for that purpose Nov 08 17:13:52 that's question for toolchain gurus Nov 08 17:15:38 I'm pretty sure they're gonna tell me I'm crazy Nov 08 17:15:42 :) Nov 08 17:16:21 It's very possible ;) Nov 08 17:20:46 binutils -> gcc-cross-initial -> libc-initial -> gcc-cross-intermediate -> libc -> gcc-cross Nov 08 17:20:59 I just found that on the mailing list Nov 08 17:21:47 so if gcc-cross-intermediate is only used to build glibc, then I'd only need to modify gcc-cross Nov 08 17:22:08 except if I also want to build a multilib glibc as well Nov 08 17:22:14 which is not the case Nov 08 17:24:47 julian_: iirc for multilib stuff you need to modify the stuff under gcc/tc-arch.c Nov 08 17:26:03 Is this what people are refering to when they mention some obscure "multilib debian patches" ? Nov 08 17:26:23 obscure? Nov 08 17:27:06 I found some places on the internet where they say that --enable-multilib is not enough when configuring gcc Nov 08 17:27:17 julian_: the issue is upstream, they do not unify multilibs for all arches, so Debian, that has multiple arches has to move arround things Nov 08 17:30:25 julian_: you need to patch gcc/config/*/t-* Nov 08 17:30:47 allrighty Nov 08 17:30:53 I'm gonna have a look to it Nov 08 17:31:20 julian_: cross lfs i think also has some patches that might help Nov 08 17:32:18 thx for the tip Nov 08 17:32:36 no prob, glad you improve the build system :) Nov 08 17:53:53 angstrom-version failed on testing-next with: Exception RuntimeError: 'maximum recursion depth exceeded while calling a Python object' in ignored - any ideas? Nov 08 17:54:35 bitbake 1.10.1 Nov 08 18:06:48 tharvey: see fe0b4102c33f7300aaf8b139ee2d73820e12d635 Nov 08 18:10:36 kergoth, so does anyone merge later fixes like that back into testing-next or we wait until the next week's cycle? Nov 08 18:16:47 Not sure Nov 08 18:23:14 * kergoth also doesn't know Nov 08 18:35:53 tharvey: we fix the testing-next Nov 08 18:36:08 tharvey: we dont fix the branch per say Nov 08 18:36:19 fixes go into master then testing-next moves to it next week Nov 08 18:39:26 provided enough platform builds went ok Nov 08 18:40:09 the issue you are talking about would fail all angstrom builds which means test-next will not be tagged but moved to master next week and then it will get this fix Nov 08 18:40:54 so someone looking for good start point always uses the latest tag and that tag is kept buildable and not moved until testing-next has enough builds to be tagged Nov 08 18:44:14 damn, i think srctree and packaged-staging step on one anothers toes Nov 08 18:44:23 hmm Nov 08 18:44:54 yes, I assume everyone building angstrom tripped over this issue Nov 08 18:52:51 bug labs are hiring Nov 08 18:53:16 http://jobs.linux.com/job/embedded-linux-developer-new-york-ny-soho-bug-labs-0133009cf3/ Nov 08 18:59:59 It works, yes Nov 08 19:00:00 ga Nov 08 19:19:13 03Chris Larson  07master * r728e072e5d 10openembedded.git/classes/srctree.bbclass: Nov 08 19:19:13 srctree: ensure setscene is still run Nov 08 19:19:13 Signed-off-by: Chris Larson Nov 08 19:27:26 hello, i have a problem compiling webkit-efl_svn.bb: JSSVGFontElement.cpp:(.text+0x174): relocation truncated to fit: R_ARM_THM_CALL Nov 08 19:28:15 on my build server ubuntu 10.04 amd64 it comiles Nov 08 19:28:38 on my desktop ubuntu 10.04 x86 it does not Nov 08 19:29:08 on my desktopn in a ubuntu 10.04 x86 with minimal software installed it compiles not Nov 08 19:29:18 (chroot) Nov 08 19:30:18 so there anybody who can compile webkit-efl on x86 for shr distribution ? Nov 08 19:35:24 Is it possible to override a variable set in the distro config file from a machine config file ? Nov 08 20:19:02 hrm Nov 08 20:19:11 * kergoth ponders srctree issues Nov 08 21:04:32 hrm Nov 08 21:35:14 fidencio, hi I didn't have the time to make the video, because my webcam seem broken, there is some garbage on it + jpeg errors Nov 08 21:35:25 I'll try with my normal camera which runs chdk Nov 08 21:35:43 fidencio, btw someone else is also working to get enjoy in oe Nov 08 21:36:11 it's pespin Nov 08 21:38:45 GNUtoo|laptop, yeah, but I'm having problems with the "test" binary. Nov 08 21:39:01 Won't have time till tomorrow or at the end of the week Nov 08 21:39:07 julian_: thats a good question Nov 08 21:39:08 dinner time now Nov 08 21:39:31 pespin|away, the test binary works Nov 08 21:39:40 pespin|away, you just need the right command line arguments Nov 08 21:40:10 GNUtoo|laptop, it doesn't ge compìled. maybe because of something I modified in the recipe. I'll digg into it when I have time Nov 08 21:40:25 bitbake is geenrating a bash script instead of the binary Nov 08 21:40:35 look in .libs Nov 08 21:40:38 or something like that Nov 08 21:40:40 ok Nov 08 21:40:47 look in hidden dirs Nov 08 21:40:54 then you need the right command line Nov 08 21:40:57 .libs in which dir? Nov 08 21:41:06 in the dir where there is the script Nov 08 21:41:24 look for hidden dirs Nov 08 21:41:38 GNUtoo|laptop: ok. :-) Nov 08 21:42:04 ./src/bin/.libs/test Nov 08 21:56:56 GNUtoo|laptop, yeah, it's there. I will try it later, thanks :) Nov 08 21:57:09 pespin|away, ok also try it this way: Nov 08 21:58:27 pespin|away, http://pastebin.com/7jGGddCB Nov 08 21:58:34 the -m mono is important Nov 08 21:58:38 else it doesn't work Nov 08 22:24:35 GNUtoo|laptop, uhm do I need some gestreamer-whatever package to support mp3? Nov 08 22:24:42 I think I only have mp3 files in my neo Nov 08 22:24:52 pespin|away, yes Nov 08 22:25:00 I have enjoy in my PC from some weeks ago, and it supports mp3 Nov 08 22:25:09 you should RDEPENDS or RSUGGEST gstreamer-foo Nov 08 22:25:38 GNUtoo|laptop, with foo you suggest me looking for the package, or is it really named like that? :P Nov 08 22:26:29 I don't know the exact name Nov 08 22:26:41 it should be gst-plugin-lame Nov 08 22:26:46 or something like that Nov 08 22:26:59 you also need gst-plugin-decodebin2 Nov 08 22:27:04 and gst-plugin-alsa Nov 08 22:27:05 etc... Nov 08 22:28:18 gst-plugin-lame - 0.10.16-r11.1.6 - GStreamer plugin for lame Nov 08 22:30:34 khem, kergoth, the only change between my last good bitbake world and my failing run is this: http://cgit.openembedded.org/cgit.cgi/bitbake/commit/?id=83cdb23f8b89453a3527a276bd0b4deb85d63deb Nov 08 22:30:47 hmm, strange Nov 08 22:31:15 explode_deps is basically just a smarter string split Nov 08 22:31:19 don't see how that would break it Nov 08 22:31:21 hmm Nov 08 22:32:49 grg: yep thats the one Nov 08 22:33:03 seems like it works ok if I remove it Nov 08 22:33:07 what's the error, again? Nov 08 22:33:34 kergoth, http://pastebin.com/sdNDTE3W Nov 08 22:33:56 I see the same now :) Nov 08 22:34:07 http://pastebin.com/hpGF8kCn Nov 08 22:34:15 what is the trigger? Nov 08 22:34:20 or is it random? Nov 08 22:34:46 don't know there's a build which triggered it http://tinderbox.openembedded.net/builds/101757/ Nov 08 22:36:12 ynezz, what did you bitbake? Nov 08 22:36:12 I could not see the trigger point but bitbake -k world produced it always at some point Nov 08 22:36:16 lately Nov 08 22:36:19 that function is only used by the depexp ui and bitbake -g, as far as i can see Nov 08 22:37:42 grg: angstrom-2010.x, machine ts72xx, console-image Nov 08 22:37:53 grg: testing-next branch Nov 08 22:38:12 fidencio, yay! enjoy running in neo freerunner: http://scap.linuxtogo.org/files/581a7187636123c08661719c80f774cf.png Nov 08 22:39:44 kergoth: could there be a bug in the function itself ? Nov 08 22:39:57 don't know what you mean Nov 08 22:40:14 whether there's a bug in the function isn't relevent if its never called Nov 08 22:40:27 bb.event.fire_from_worker(self.queue[:index+8], self.d) Nov 08 22:40:41 yes Nov 08 22:41:37 hmmm I wonder if runqueue preparation is affected by this somehow Nov 08 22:41:52 I mean the commit Nov 08 22:41:56 pespin|away: wow! that's cool! Nov 08 22:42:04 kergoth: when you have 5 mins, I have some questions about http://cgit.openembedded.org/cgit.cgi/openembedded/commit/?h=master&id=4271024e26551da70e710cef6d2dcbfa8740b007 Nov 08 22:42:06 again, that commit is in a function that's only called by bitbake -g and the depexp ui Nov 08 22:42:18 afaict, anyway Nov 08 22:42:23 fidencio, can't play music still though :P Nov 08 22:42:26 clearly something is happening, but i'm not seeing how at the moment Nov 08 22:42:36 kergoth: ok Nov 08 22:42:44 ant___: what's the question? Nov 08 22:42:46 fidencio, it segfaults when I press the play button Nov 08 22:42:56 grg: can you try to revert that commit and see if it ends up with that error again Nov 08 22:43:03 I don't want any i18n in my initramfs image. How? Nov 08 22:43:07 for me it seems to go well. Nov 08 22:43:21 khem, sure. Will know in 5-10 hours Nov 08 22:44:14 kergoth: I sent a msg to the ML about that: I can override after including but perhaps a weak assignment could be better Nov 08 22:45:08 pespin|away: which rev? Nov 08 22:45:28 pespin|away: I think that in HEAD svn is not segfaulting Nov 08 22:46:10 fidencio, its due to gst dependencies Nov 08 22:46:22 fidencio, I'm overcoming some problems, just have to find which do I need Nov 08 22:46:42 segfault problem was missing gst-plugin-playbin Nov 08 22:46:46 grg: cool Nov 08 22:46:49 pespin|away: is segfaulting in desktop version too? Nov 08 22:46:53 pespin|away: hmmm. right Nov 08 22:47:02 fidencio, now it doesn't segfault, but I have toher problems Nov 08 22:47:19 ant___: it looks like that commit unintentionally changed the logic. glibc-localedata-i18n should only be installed if IMAGE_LINGUAS is non-empty. that conditional was lost in the switch to overrides based, as far as i can see. should be easy enough to resurrect. try IMAGE_LINGUAS_linux = "${@base_ifelse(d.getVar('IMAGE_LINGUAS', True), 'glibc-localedata-i18n', '')}" Nov 08 22:47:30 ant___: rather than the current IMAGE_LINGUAS_linux = "glibc-localedata-i18n" Nov 08 22:47:34 ant___: see if that takes care of it Nov 08 22:47:49 ok, I'll try now, thx Nov 08 22:48:20 fidencio, http://paste.pocoo.org/show/288091/ Nov 08 22:48:42 I get infinite number of those last messages when I press play, till I kill it Nov 08 22:48:49 and no sound Nov 08 22:50:28 fidencio, I'm trying to play mp3 files Nov 08 22:52:59 pespin|away: I never used enjoy in embedded systems. Is working in PC Nov 08 22:53:37 fidencio, last time I tried it here in PC didn't work neither Nov 08 22:53:41 I'll try to recompile in pC Nov 08 22:54:25 kergoth: now is like before, no i18n (obviously you meant LINGUAS_INSTALL_linux = blah) Nov 08 22:54:34 pls fix it, thx Nov 08 22:55:31 he.. Angstrom~cpio.gz│ 177410│Nov 7 00:10 vs. Angstrom~cpio.gz│ 84605│Nov 8 23:47 Nov 08 22:55:32 kergoth: there are two generateDepTreeEvent() definitions Nov 08 22:56:11 pespin|away: I'll try too Nov 08 22:56:13 one is in lib/bb/command.py and another one in lib/bb/cooker.py Nov 08 22:56:22 yep Nov 08 22:56:42 the former is a command you can send to the server which runs the cooker method Nov 08 22:57:08 03Chris Larson  07master * r0ad93ec8b5 10openembedded.git/classes/image.bbclass: Nov 08 22:57:08 image: don't include glibc-localedata-i18n when IMAGE_LINGUAS is empty Nov 08 22:57:08 Signed-off-by: Chris Larson Nov 08 22:58:08 ant___: yikes Nov 08 22:58:13 ant___: that should fix it Nov 08 22:58:19 50% in this case ;) Nov 08 23:00:11 kergoth: I am looking into lib/bb/ui/depexp.py:init() Nov 08 23:00:22 that function wont be called when not using -g ? Nov 08 23:00:26 neither Nov 08 23:00:44 depexp.py is a different ui, only used with bitbake -u depexp Nov 08 23:00:54 dependency explorer Nov 08 23:00:56 X11 Nov 08 23:02:44 kergoth: it could be that something we did to metadata thats unearthing this issue in bb Nov 08 23:02:44 Hi all Nov 08 23:04:07 fidencio, I get lots of "Spank spank naughty programmer" eina iterator errors when using enjoy in my PC Nov 08 23:05:14 could it be because an older lightmediascanner/database? Nov 08 23:05:35 pespin|away: let me 10 minutes. I'll try here Nov 08 23:05:48 ok Nov 08 23:09:48 fidencio, same problem with ogg files in openmoko Nov 08 23:10:44 pespin|away: Can you report this erros in #edevelop? Nov 08 23:11:33 pespin|away: you can ping acidx or otaviobp Nov 08 23:14:20 fidencio, I'm asking in #gstreamer. it seems I get same error with gst-launch Nov 08 23:18:33 kergoth: I've tested again about base_version_less_or_equal issue with particular $PV / see http://pastebin.com/DkBGgm0F Nov 08 23:20:07 kergoth, that one is insane but even with 2.6.36+2.6.37-rc1 it would fail Nov 08 23:20:48 pespin|away: ok Nov 08 23:23:13 fidencio, it seems I need gst-plugins-base, but the generated package has an empty data.tar.gz :S Nov 08 23:33:42 fidencio, gst-launch is working now :) in enjoy... I listen a glimpse of music sometimes, but it looks like it's not being capable of playing it quite reliably Nov 08 23:34:08 fidencio, 757 root 20 0 71520 20m 9808 R 98.1 16.8 1:58.97 enjoy Nov 08 23:34:16 wow! Nov 08 23:34:36 pespin|away: are you running enjoy in which device? Nov 08 23:35:16 fidencio, om-gta02 Nov 08 23:35:25 yeah, hardware sucks Nov 08 23:36:17 btw, this is what I needed to have it working: gst-plugin-typefindfunctions gst-plugin-lame gst-plugin-ogg gst-plugin-vorbis gst-plugin-playbin Nov 08 23:36:25 (have added it to enjoy bb recipe) Nov 08 23:40:27 fidencio, uhm, gst-launch is having problems too decoding it reliabily Nov 08 23:40:58 so it's not only an enjoy problem, but a gst one Nov 08 23:41:17 too bad mplayer can play it without problems >.< Nov 08 23:49:36 fidencio, it looks mp3 playback works great though! Nov 08 23:49:39 :D Nov 08 23:50:18 great difference: 850 root 20 0 62900 20m 9764 S 47.3 17.0 0:42.44 enjoy Nov 09 00:23:11 khem: seems like rejecting that offending commit in bitbake fixed that issue, now at task 3980 of 4819 Nov 09 00:50:49 has someone got libxml2 building with uclibc? Nov 09 01:13:48 khem: false alarm, rejecting that commit didn't helped Nov 09 01:14:10 khem: and it's testing-next so it's rm -fr tmpdir before Nov 09 01:25:16 khem, you speak on Wednesday? Nov 09 01:30:10 he'll let you know in a couple days :P Nov 09 02:19:14 anyone tried the new version of valgrind on ARM? Nov 09 02:19:50 was going to create a patch to the recipe; the recipe is for 3.6.0, but still has COMPATIBLE_HOST set to x86 or x86_64, so i was wondering if people had found bugs or just hadnt updated the recipe to reflect ARMv7 support Nov 09 02:21:50 it only supports armv7 and up; doesn't look like you can trivially say that in COMPATIBLE_HOST, since all arm-based .confs set TARGET_ARCH to "arm" and not their specific ISA version Nov 09 02:24:29 mrj10, I suspect I will try it eventually .... Nov 09 02:25:04 i'll try to get it to work, not sure how to integrate it into the recipe in a sufficiently robust way Nov 09 02:30:56 configure pukes because it expects the architecture to be "armv7*" instead of "arm" Nov 09 02:31:59 not sure whether it's more kosher to patch configure.in to assume that arm==armv7*, or change OE to pass in the ISA revision instead. could hack it into just the valgrind recipe, but seems like other packages might like to know as well Nov 09 02:34:57 Don't we have to play a similar game for mplayer? Nov 09 02:35:01 or ffmpeg or something Nov 09 02:35:29 couldn't tell you offhand, i can look though Nov 09 02:35:38 aping that seems like a good idea for the time being Nov 09 02:35:44 yeah, poke around, should be something to rip off :) Nov 09 02:35:48 when is the next patch day? Nov 09 02:46:07 ka6sox, when we collect some patches in the new patchwork :) Nov 09 02:46:38 * Crofton is headed for the hotel, been up since 1AM local time Nov 09 02:46:41 gn Nov 09 02:56:00 valgrind requires the debug version of libc. can we build the debug version of glibc in OE out of the box? Nov 09 02:56:11 i'm seeing it for eglibc but not glibc **** ENDING LOGGING AT Tue Nov 09 02:59:57 2010