**** BEGIN LOGGING AT Mon Apr 05 02:59:56 2010 Apr 05 10:28:33 hi all Apr 05 10:30:24 anybody can guide me to build openjdk for angstrom Apr 05 10:45:48 Hello! Apr 05 10:45:59 Can anyone help? I'm using openembedded toolchain to bitbake angstrom's console-image and I get an error when compiling u-boot. Anyone has a clue on how to fix? Apr 05 10:46:17 bitbake fails on do_compile for u-boot Apr 05 10:46:53 And this is his log file in the log file I have this first error: (...)/OE/angstrom-dev/work/beagleboard-angstrom-linux-gnueabi/u-boot-2009.05+r30+gitrd363f9cb0918a1b6b92e2e20d01543d0cf53274-r30/git/lib_generic/sha1.c:64: error: expected ')' before '*' token Apr 05 10:46:55 :) Apr 05 10:47:14 siji: thanks! :) Apr 05 10:47:21 :) Apr 05 10:47:39 thats new Apr 05 10:48:17 XorA, Is it cose of any bug in the Uboot Source Apr 05 10:48:44 no idea, totally new error to me Apr 05 10:48:54 ok Apr 05 10:51:40 any other details I can give that would help figuring this out? Apr 05 10:52:55 Am not sure but my wild guess is smthing wrong in that sha1.c Apr 05 10:53:58 if you are having the same version of uboot soruce , just cross check the files Apr 05 10:54:59 thats not org.openembedded.dev anyway Apr 05 10:59:26 Or may be a prblm with gcc Apr 05 10:59:38 I mean gcc version Apr 05 11:12:34 siji: " if you are having the same version of uboot soruce , just cross check the files" Can you explain? Apr 05 11:18:58 ,bfre that can u tell me how are trying to build this Uboot in OE Apr 05 11:23:33 siji: I'm just doing: MACHINE=beagleboard bitbake console-image Apr 05 13:33:37 GNUtoo, https://www.codeaurora.org/gitweb/quic/xwin/?p=xf86-video-msm.git;a=shortlog;h=refs/heads/chromium Apr 05 13:33:50 GNUtoo, xf86-video-msm updates Apr 05 13:34:02 dcordes, nice,I'll look later Apr 05 13:35:17 GNUtoo, we should check if it brings any changes on dream, update in OE, and most importantly fix the build in OE finally Apr 05 13:35:28 GNUtoo, the current version is still broken Apr 05 13:35:29 ok Apr 05 13:35:35 I noticed Apr 05 13:35:59 dcordes, it's easy to fix,just cp the headers to the src dir Apr 05 13:36:01 leviathan wanted to give me the patches but we never really managed Apr 05 13:36:08 ok Apr 05 13:36:18 I'll try to fix it when I'll have the time Apr 05 13:36:25 I'm currently struggeling with wifi Apr 05 13:36:34 that would be good if somebody jsut does it Apr 05 13:36:39 the problem is, I don't have dream to test! Apr 05 13:39:20 I'm looking forward to test it on the leo. hopefully it will work without the neon removal Apr 05 15:10:11 i just uploaded a custom image from narcissus but after uncompressing linux the device seems to turn off Apr 05 15:10:55 nusse_, you might want to ask in the #angstrom IRC channel Apr 05 15:11:09 ill try there Apr 05 15:28:01 does the firmwareloader needs userland helpers? Apr 05 15:31:54 does OE have a target to build a nfs rootfs and kernel image? Apr 05 15:32:18 like can I tell OE I have a nfs server at x.x.x.x and my rootfs be mounted from /myrootfs and have it build everything? Apr 05 15:33:31 just set your IMAGE_FSTYPE to tar, bake the image you want, adn unpack the tarball as root where you want it Apr 05 15:33:32 bfrog, build a tar.gz and uncompress it where ever you want to NFS boot from. Apr 05 15:40:09 makes sense Apr 05 15:40:17 what if I wanted to do development Apr 05 15:40:26 like can OE build me a chroot like scratchbox sort of? Apr 05 15:40:59 or would I just use qemu and chroot in to the rootfs Apr 05 15:41:26 I'm just trying to figure out this whole arm9 development setup deal Apr 05 15:41:52 bfrog, I tend to boot the actual device. Apr 05 15:42:18 with gcc and whatever else you need on the device? Apr 05 15:42:23 isn't that slow? Apr 05 15:42:46 I guess it wouldn't be that slow Apr 05 15:47:23 bfrog, see what you mean - you can always use the cross compiler generated by OE. Apr 05 16:00:53 bfrog, if you need gcc etc... bitbake devshell and you'll be able to use gcc etc... Apr 05 16:01:04 GNUtoo: thanks Apr 05 16:01:11 will that build a new image, or just a set of ipkgs? Apr 05 16:01:26 bfrog, else if you don't want to build on the builder you need to bitbake an sdk,a nativesdk or a canadian sdk Apr 05 16:01:47 bfrog, it will build a shell script Apr 05 16:01:57 oh! Apr 05 16:01:58 you will source that script Apr 05 16:02:05 and then I can develop as usual? Apr 05 16:02:11 the you'll have gcc in your path etc... Apr 05 16:02:17 oe_runmake etc... Apr 05 16:02:19 what about the includes for the arm though Apr 05 16:02:28 ? Apr 05 16:02:28 those are likely very different Apr 05 16:02:35 you mean staging Apr 05 16:02:36 ? Apr 05 16:02:42 it can deal with it I think Apr 05 16:02:44 no, like I want to write a C program Apr 05 16:02:50 yes Apr 05 16:02:52 and have it run on the beagle board Apr 05 16:03:01 it will uses arm includes if you call the right gcc Apr 05 16:03:02 but I need all the includes to be correct for the beagleboard Apr 05 16:03:09 gotcha Apr 05 16:03:13 what about kernel includes? Apr 05 16:03:16 same thing? Apr 05 16:03:30 same thing Apr 05 16:03:33 nice Apr 05 16:03:40 I'm liking OE more now :-) Apr 05 16:03:45 source ~/oetmp/deploy/addons/arm-oe-linux-gnueabi-shr-htcdream-devshell Apr 05 16:04:17 nice Apr 05 16:04:20 thanks Apr 05 16:04:26 arm-oe-linux-gnueabi-gcc playwav2.c -o playwav Apr 05 16:04:40 file playwav => playwav: ELF 32-bit LSB executable, ARM, version 1 (SYSV), dynamically linked (uses shared libs), for GNU/Linux 2.6.16, not stripped Apr 05 16:05:32 and this playwav had standard c libraries includes Apr 05 16:05:47 like stdio.h stdlib.h etc... Apr 05 16:06:16 then all is good Apr 05 16:36:11 FATAL: Unable to start to 'None' UI due to exception: 31338. .. interesting Apr 05 16:38:01 RP: ping Apr 05 17:54:12 mickeyl: pong Apr 05 17:54:19 * RP won't be here for long though :/ Apr 05 17:55:55 RP: Good news, I don't think anything broke on my torture builds, w/ SCAN_CMD as Joshua did it Apr 05 17:56:08 Tartarus: Cool :) Apr 05 17:56:09 Haven't had time to dig the logs, but it took about as long as a working build should Apr 05 17:56:16 RP: There's still the binconfig problem tho Apr 05 17:56:32 * Tartarus needs to tell Koen about that since it's a bit premature to put target pstaging up Apr 05 17:56:47 Tartarus: I've not got to email yet, I assume that will make sense when I do Apr 05 17:56:59 RP: Nope, just part of the pastebin I did before Apr 05 17:57:08 staging//usr/bin//foo-config Apr 05 17:57:19 pstage'd into target stuff Apr 05 17:58:01 Tartarus: ah, right Apr 05 17:58:17 Tartarus: That shouldn't be too dfficult to solve, good point though Apr 05 17:58:26 RP: it's not, no, sec.. Apr 05 17:59:59 RP: http://pastebin.com/fwiACsPH Apr 05 18:00:11 afk, work phone time Apr 05 18:01:43 Tartarus: As part of the cross changes, I want to move the cross binaries directory. That should make these moves unrequired which will be nice Apr 05 18:07:42 PING 1270490863 548576 Apr 05 18:11:51 * RP afk Apr 05 18:32:18 RP: argh Apr 05 18:32:21 too late :) Apr 05 18:32:59 RP: i'm pondering whether you would accept a LED trigger that implements brightness changes in a soft dimming way Apr 05 18:33:11 * mickeyl wants breathing LEDs Apr 05 18:34:08 mickeyl: We've already got things like the various blinking triggers. Apr 05 18:34:31 yes, but no smooth dimming Apr 05 18:34:35 only hard Apr 05 18:34:40 RP: Depends totally on what the dir structure ends up as Apr 05 18:34:47 Yes, but that's precedent for that style of operation. Apr 05 18:35:00 Main difference is that triggers don't (by and by large) read the brightness. Apr 05 18:35:01 RP: These are run-on-host to tell build-time where stuff is, target-specific stupid scripts Apr 05 18:35:03 broonie: right. so i guess such a trigger would be acceptable to mainline Apr 05 18:35:17 RP: If we can place them someplace that's not encoded, yes, it goes away Apr 05 18:37:08 mickeyl: Incidentally, with questions like that you're generally better off just asking them :) Apr 05 18:39:30 broonie: well... want to check the waters before diving into coding such a thing Apr 05 18:39:47 mickeyl: I meant rather than pinging. Apr 05 18:39:57 heh Apr 05 18:39:59 true Apr 05 18:40:18 Folks keep doing that on ASoC for questions I could answer in one line if they were actually still on line when I reappear. Apr 05 18:42:06 ;D Apr 05 18:46:55 mickeyl: good evening Apr 05 18:47:27 Tartarus: the binconfig scripts? they should just go away, period. Apr 05 18:47:45 pb_: Yes, them and .la files need to alt.diediedie Apr 05 18:48:27 pb_: good evening Apr 05 18:50:04 re Apr 05 19:10:50 greetings! is there any documentation on how the PR variable is set? I'm seeing other packaged get 0.1, 0.2 etc if their .bb file is edited, I would like my .bb recipes to behave the same Apr 05 19:19:58 ? Apr 05 19:20:04 the user bumps PR when they edit a recipe Apr 05 19:20:09 there's no magic Apr 05 19:20:51 kergoth_: ok I think I just answered my own question, apparently INC_PR in included recipes was throwing me off Apr 05 19:21:06 INC_PR is just a variable you can use when you set PR Apr 05 19:21:23 I thought there was something behind the scenes doing checksums/watching timestamps of .bb files Apr 05 19:21:25 to ensure that changes to the .inc automatically affect the recipes that include it, so you don't have to bump the PR of every single recipe manually when you change the .inc Apr 05 19:21:28 nope Apr 05 19:21:40 right, that makes sense Apr 05 19:21:47 thanks for the help Apr 05 19:22:07 the thing is, there are lots of changes you can make to recipes that don't actually affect the output Apr 05 19:22:15 so there'd be no point in rebuilding due to a timestamp change in those cases Apr 05 19:22:18 yes, that would be bad Apr 05 19:22:27 we're moving in that direction, though Apr 05 19:22:35 just need to keep better track of what variables are used where, etc Apr 05 19:22:37 really? Apr 05 19:22:39 ah Apr 05 19:23:38 we're looking to drop the whole stamp concept in favor of checksums of the input/output of the tasks / metadata, long term Apr 05 19:23:54 ah interesting Apr 05 19:24:47 kergoth_: isn't it like e2factory works? Apr 05 19:26:05 yep Apr 05 19:33:07 mario-goulart: kergoth_: :) Apr 05 19:53:19 can I build mamona using OE? Apr 05 19:54:33 bfrog: probably. Maybe you'll get better answers from #mamona Apr 05 20:03:06 it seems like angstrom is sort of the semi-standard oe distro? Apr 05 20:27:59 hi kergoth Apr 05 20:42:52 hey pb_ Apr 05 20:55:30 bump. just hit the wall while building latest x11-image on devel-branch. libpng-native do_configure is failing Apr 05 20:55:51 /home/nazgee/workspace/oe/openembedded/build/tmp-igep0020/work/x86_64-linux/libpng-native-1.2.42-r1/temp/run.do_configure.19884: /home/nazgee/workspace/oe/openembedded/build/tmp-igep0020/staging/x86_64-linux/usr/bin/automake: /home/nazgee/workspace/oe/openembedded/build/tmp-igep0020/staging/x86_64-linux: bad interpreter: No such file or directory Apr 05 20:56:07 that's what logs do tell me Apr 05 20:56:26 whcich is not enoeugh for a dummy as me to get forward Apr 05 20:56:49 that's rather funky. what does the first line of /home/nazgee/workspace/oe/openembedded/build/tmp-igep0020/staging/x86_64-linux/usr/bin/automake look like? Apr 05 20:56:52 all directories listed above do exist. anyone got a clue? Apr 05 20:58:03 #!/home/nazgee/workspace/oe/openembedded/build/tmp-igep0020/staging/x86_64-linux/usr/bin/perl -w Apr 05 20:58:07 # -*- perl -*- Apr 05 20:58:09 # Generated from automake.in; do not edit by hand. Apr 05 20:58:11 eval 'case $# in 0) exec /home/nazgee/workspace/oe/openembedded/build/tmp-igep0020/staging/x86_64-linux/usr/bin/perl -S "$0";; *) exec /home/nazgee/workspace/oe/openembedded/build/tmp-igep0020/staging/x86_64-linux/usr/bin/perl -S "$0" "$@";; esac' if 0; Apr 05 20:58:38 does that perl binary exist? Apr 05 20:58:49 gimme a sec Apr 05 20:59:20 hah. it does not ;] Apr 05 20:59:33 rebuilding perl shall do it? Apr 05 21:00:34 perl-native, I guess, though I'm not quite sure how automake came to depend on perl from staging. Apr 05 21:01:14 i had this image built some time ago Apr 05 21:01:29 today for some reason i decided to pull some news Apr 05 21:01:33 and i end up here ;] Apr 05 21:01:37 on the first error Apr 05 21:01:46 bitbake -f perl does fail either Apr 05 21:01:52 i'll try with native Apr 05 21:04:23 perl-native is being baked now. it seems that it will do it. TIA. devel-branch is not for me ;] i should stick with a stable one Apr 05 21:09:53 pb_: it worked. thank you. Apr 05 21:12:50 very good Apr 05 21:41:07 Hi everybody, I have a question. I want to work on BeagleBoard and I need a buildsystem. I have decided to use OpenWrt, because I am a bit familiar with. Apr 05 21:41:21 But one of my friend suggested openembedded Apr 05 21:41:53 If anybody works on BeagleBoard, I want to ask if openembedded will be nice with it Apr 05 21:50:18 any advice will be really appreciated :) Apr 05 21:58:50 mgundes: there is a #beagle which is full of people specialising in OE on beagle Apr 05 21:59:06 mgundes: OE is what I use on my two, Angstrom distro to be precise Apr 05 22:03:08 thanks, I will ask in beagle room, you are right Apr 05 22:08:46 mgundes: www.angstrom-distribution.org if you want to build a quick demo image Apr 05 22:27:48 there is a problem in tzcode recipe again - in latest devel branch it points to tzdata-2010g located in unreachable locations. previously there was simillar issue, but with 2010f Apr 05 22:30:07 Try changing g to h Apr 05 22:30:44 then submit patches Apr 05 22:30:56 mickeyl: I see no reason that wouldn't be accepted Apr 05 22:31:11 RP: ok, thanks. i might take a stab at it. Apr 05 22:31:21 these days many leds use pwm Apr 05 22:31:30 Tartarus: Yes, I have a path layout in mind that would make that directory not host dependent Apr 05 22:32:04 Tartarus: there is noe need for that. I've grabed a copy from ftp://munnari.oz.au/pub/oldtz/tzdata2010g.tar.gz and genereated md5 Apr 05 22:32:05 RP: Good :) Apr 05 22:44:46 ok. antoher blocker. i got: "/usr/bin/ld: cannot find -lldap" from the error log Apr 05 22:45:41 damn. i should have stayed with my a bit delayed version of oe. there was really no need to be updated. Apr 05 22:47:32 here goes log output: Apr 05 22:47:35 NOTE: make -j2 Apr 05 22:47:35 Making all in libbb Apr 05 22:47:35 Making all in libopkg LINK libopkg.la Apr 05 22:47:35 /usr/bin/ld: cannot find -lldap Apr 05 22:47:35 collect2: ld returned 1 exit status Apr 05 22:47:38 make[2]: *** [libopkg.la] Error 1 Apr 05 22:47:40 make[1]: *** [all] Error 2 Apr 05 22:47:43 make: *** [all-recursive] Error 1 Apr 05 22:59:02 installing libldap2-dev helped. it's strange it did not come out before Apr 05 22:59:20 as I've build angtrom for few times now **** ENDING LOGGING AT Tue Apr 06 02:59:56 2010