**** BEGIN LOGGING AT Sat Oct 18 02:59:57 2008 Oct 18 03:27:57 hi khem! Oct 18 06:02:21 how do i install all deps with ipkg? Oct 18 06:16:14 HopsNBarley_: hey how r u Oct 18 06:49:34 03Mike Westerhof 07org.openembedded.dev * r3607c5935c 10openembedded.git/conf/checksums.ini: checksums.ini: Additions from the SlugOS packages Oct 18 08:14:05 03Phil Blundell 07org.openembedded.dev * rc6aff25fcf 10openembedded.git/packages/gcc/ (gcc-4.1.1.inc gcc-4.1.1/sed-fixinc-backslash.patch): gcc-4.1.1: fix problem with backslash in sed command Oct 18 08:23:50 Morning all Oct 18 08:24:07 * RP has taken his eyes off OE for too long and you've all gone crazy :( Oct 18 08:24:49 Hey RP Oct 18 08:26:15 Can you see whats wrong with my git access :) I am not able to checkout git@git tree Oct 18 08:27:16 khem: Is this an anonymous or ssh checkout? Oct 18 08:27:25 khem: Which url are you using exactly and what is the error? Oct 18 08:27:40 RP: anonymous works problem is with ssh Oct 18 08:27:51 something is wrong with my keys prolly Oct 18 08:29:18 I think I see a problem Oct 18 08:29:22 RP: error is here http://pastebin.ca/1230021 Oct 18 08:30:43 I've just powered on this machine so its going to take it a few minutes to resync things Oct 18 08:31:27 RP: no problems I have waited couple of days few mins is ok :) Oct 18 08:32:38 heh, I don't have access anymore Oct 18 08:34:18 khem: Try now, we should both be fixed ;-) Oct 18 08:35:24 RP: yesss finally thanks a bunch Oct 18 09:04:02 ok, complaint email sent :/ Oct 18 09:05:19 03Koen Kooi 07org.openembedded.dev * r2b4d8829a8 10openembedded.git/contrib/angstrom/build-feeds.sh: angstrom feed builder: add gnome-icon-theme to cleaning list Oct 18 09:21:20 03Jeremy Lainé 07org.openembedded.dev * r77129c8939 10openembedded.git/packages/wt/ (files/cmakelist.patch wt_2.2.0.bb): wt: fix cross-compilation Oct 18 10:55:54 03Koen Kooi 07org.openembedded.dev * rfa469a73af 10openembedded.git/packages/telepathy/libtelepathy_0.3.3.bb: libtelepathy: add telepathy-glib to DEPENDS Oct 18 10:57:36 mickeyl: good morning Oct 18 10:59:21 good morning pb_ Oct 18 11:18:41 morning all Oct 18 11:25:01 03Michael 'Mickey' Lauer 07org.openembedded.dev * r0592d2c50f 10openembedded.git/packages/tasks/task-fso-compliance.bb: taks-fso-compliance: add Oct 18 11:25:28 hi thesing Oct 18 11:25:57 03Michael 'Mickey' Lauer 07org.openembedded.dev * r23c6142305 10openembedded.git/packages/tasks/ (task-x11-illume.bb task-x11.bb): task-x11[-illume]: switch to FILE_PR Oct 18 11:28:59 RP seems to be unhappy about the FILE_PR stuff. I guess he didn't read the ml for some time. Oct 18 11:30:53 thesing: I do read the mailing list ;-) Oct 18 11:31:19 But wasn't the change discussed on ml? Oct 18 11:31:32 thesing: For about 12 hours, yes Oct 18 11:33:10 ok. so not really long. Oct 18 11:33:38 what about the git conversion? is it just "we could have done better?" Oct 18 11:34:12 thesing: The bkcvs part is just broken Oct 18 11:34:25 and we could have done it better so that we could fix it later, yes Oct 18 11:35:40 But its quite complicated to change it now, isn't it? Oct 18 11:36:03 It might not be so bad Oct 18 11:36:16 there would be some pain Oct 18 11:37:15 the one thing that I try to understand is.. how bad is just ignoring it Oct 18 11:37:27 i mean it's just pre-mtn, that's years ago. Oct 18 11:37:36 why do we need to have it perfect? how bad is it really? Oct 18 11:38:18 can't we just rebase the tree when we get a better bk conversion? Oct 18 11:39:29 that's the 2nd question Oct 18 11:39:37 the 1st is... do we really _need_ the better bk conversion Oct 18 11:40:29 bk?? Oct 18 11:40:46 dcordes: yes. Oct 18 11:41:00 I don't think that we _need_ it but if someone wants to invest time in this, we should honor it and use it. Oct 18 11:41:16 dcordes: bk = bitkeeper Oct 18 11:41:37 on the FILE_PR thing, I agree with rp, it does seem like the patch that was landed is a bad one. Oct 18 11:41:50 suddenly changing the format of all .bb files doesn't seem like a great idea Oct 18 11:42:20 hmm Oct 18 11:42:23 i have to agree with that Oct 18 11:42:34 without knowing what it tried fixing Oct 18 11:42:38 (as i must confess) Oct 18 11:42:58 ok, so lets move on to the list, trying not to piss off zecke Oct 18 11:43:46 Right, upsetting zecke isn't my aim, I just think we could do this in a cleaner way Oct 18 11:44:05 he was trying to add a way for distros to globally increment PR for things like the --hash-style change, which is a good aim and one that I supported. Oct 18 11:44:07 mickeyl: AFAIK if you change and .inc file you don't have to bump all PRs using this file now. Oct 18 11:44:43 but, I admit that I didn't pay much attention at the time to how he was doing it, and if I had noticed the details of his patch I would have objected at the time. Oct 18 11:44:49 ok Oct 18 11:44:53 i (kind of) see Oct 18 11:44:55 pb_: I think this could be done at the package_*.bbclass level since we can manipulate the PR values there Oct 18 11:44:56 nothing we can't fix Oct 18 11:44:59 right Oct 18 11:45:37 RP: yeah. what I thought he was going to do was to introduce a new variable, PKG_PR or something, being the concatenation of DISTRO_PR and PR, and then have the packaging use that instead. Oct 18 11:46:18 pb_: right. Can you reply to the thread in support of doing this a better way please? Oct 18 11:46:22 obviously there are a variety of ways to achieve that end, though, and it doesn't really matter how it gets done. Oct 18 11:46:28 mickeyl: Likewise :) Oct 18 11:46:28 yah, okay, will do Oct 18 11:48:19 OT: we really need to get this year's OEDEM planned Oct 18 11:48:49 with git and the changes surrounding various interest groups, we got to talk about some things in person Oct 18 11:49:34 in Berlin would be best ;) Oct 18 11:57:30 not for me... Oct 18 11:57:36 but i'm used to compromise on that Oct 18 11:59:21 mickeyl: presumably berlin would be better than outside Germany? ;-) Oct 18 12:03:42 RP: *nod* Oct 18 12:09:08 mickeyl, don't do it in November .... Oct 18 12:09:50 november too full anyways Oct 18 12:09:55 good Oct 18 12:09:59 early dezember sounds good for me Oct 18 12:10:03 I have an out of town paying gig Oct 18 12:23:19 zecke: yo zecke Oct 18 12:26:57 Hello Oct 18 12:27:17 * otavio has few more patches to push Oct 18 12:27:36 is anyone available to review them? Oct 18 12:28:20 which files do they apply to? Oct 18 12:29:49 03Jeremy Lainé 07org.openembedded.dev * r58774ccd9f 10openembedded.git/packages/boost/ (boost-36.inc boost_1.36.0.bb): boost-1.36.0: disable boost-serialization, it fails on powerpc Oct 18 12:33:01 pb_: I'm pushing them for your review Oct 18 12:33:21 pb_: https://projetos.ossystems.com.br/git/?p=users/otavio/org.openembedded.dev.git;a=summary Oct 18 12:33:39 pb_: cups should be moved to 1.3.9 before commiting Oct 18 12:37:21 okay. I have to go move some firewood around now, but I will look at these later if nobody else has reviewed them in the meantime. Oct 18 12:38:13 pb_: great. I believe they're all safe but is better someone else to look at it Oct 18 12:38:19 RP: and we can probably discuss FILE_PR here... I simply do not agree that we need more python magic to cover up things Oct 18 12:38:47 RP: with PR = FILE_PR.DISTRO_PR it is pretty obvious how the resulting build directory in work will be named Oct 18 12:39:00 zecke: the branch metadata code has a bug; it puts a new line on it Oct 18 12:39:12 otavio: strip it :) Oct 18 12:39:58 zecke: yeah; just wanted to tell you. I'd prefer to leave my hands out of base classes ;-P Oct 18 12:40:21 RP: I think we don't need more magic in OE, in contrast we need less of that. This is good for the runtime speed of OE and good for the users of OE as things are more obvious Oct 18 12:40:44 zecke: what RP complains about? Oct 18 12:40:57 zecke: and yes, OE metadata loading is SLOW Oct 18 12:41:11 otavio: see the mailinglist, but not complains, discusses Oct 18 12:41:57 zecke: oh ok, i'm not native. I didn't meant he was being boring but just a disagreement Oct 18 12:41:59 otavio: pastebin a patch, I rubber stamp it :) Oct 18 12:42:15 zecke: ok, will look at it Oct 18 12:42:50 zecke: I agree that less magic is good. But rp does have a fair point that requiring all the recipes to suddenly define FILE_PR rather than PR is an incompatible format change, of the kind that I am normally a vocal opponent of :-} Oct 18 12:43:33 zecke: a less intrusive change, which would still not require any extra python, would be to let the recipes continue to define PR, make a new variable PKG_PR = "${DISTRO_PR}.${PR}", and patch the various class files to use that instead of the regular PR where appropriate. Oct 18 12:44:08 pb_: yes, but bitbake does handle PR and not PKG_PR Oct 18 12:44:18 pb_: so you don't rebuild when bumping the PR :} Oct 18 12:44:28 the DISTRO_PR Oct 18 12:44:48 zecke: oh, drat, are the stamps all done in bitbake? Oct 18 12:44:54 I thought base.bbclass did that, but I haven't checked. Oct 18 12:45:09 if that's the case then you're right, this plan would lose. Oct 18 12:45:38 * otavio now see the point Oct 18 12:46:29 hm, isn't it enough to change the definition of ${STAMP} in bitbake.conf? Oct 18 12:46:39 zecke: in this case bitbake would need to be change to allow to get the revision field to be set; what breaks all users until they do the upgrade of bitbake .. not good too. Oct 18 12:46:47 pb_: I think there is still some conde in bitbake that treats PR specially Oct 18 12:46:52 zecke: drat Oct 18 12:46:58 bbiab, really firewood duty now Oct 18 12:47:21 wow that was a simple and stupid fix Oct 18 12:47:24 hehe Oct 18 12:47:47 zecke: http://pastebin.com/f14459749 Oct 18 12:48:10 otavio: put it in Oct 18 12:48:52 zecke: i've commited it locally Oct 18 12:49:00 pb_: yes, I agree that there might be a better way of combining less magic and still getting PKG_PR, I think it at least requires one bitbake release until we are there Oct 18 12:49:35 When is bitbake planned to be released? Oct 18 12:49:54 no plans Oct 18 12:50:54 I just say, that any other approach requires work on OE classes and bitbake, and with bitbake it looks RP and me switched role's again (somehow we can not manage to have two active devs at the time) Oct 18 12:52:57 pb_: would you change STAMP, why not PF? Oct 18 12:54:24 zecke: did you mean PR? Oct 18 12:55:23 otavio: instead of? Oct 18 12:55:28 zecke: PF? Oct 18 12:55:37 otavio: open bitbake.conf Oct 18 12:56:04 zecke: got it Oct 18 12:56:19 PF == Package File, right? Oct 18 12:56:43 or Fullname; whatever Oct 18 12:57:40 zecke: this would change the stamp; it looks a possible way to solve those issues; you fear bitbake ignores it and use PR directly? Oct 18 12:58:34 * otavio preparing a new push set (base.bbclass and checksums.ini additions) Oct 18 13:00:06 otavio: git-grep PR in bitbake... at some point it only looked at PV, PR for the dependencies and if i see this correctly it still does Oct 18 13:03:01 03Otavio Salvador 07org.openembedded.dev * r4b1aa6a614 10openembedded.git/classes/base.bbclass: base.bbclass: strip branch name in base_detect_branch Oct 18 13:03:01 03Otavio Salvador 07org.openembedded.dev * rd1a7beac4a 10openembedded.git/conf/checksums.ini: checksums.ini: add xfce-utils 4.4.2 Oct 18 13:03:02 03Otavio Salvador 07org.openembedded.dev * r5886b2b45e 10openembedded.git/conf/checksums.ini: checksums.ini: add xfce-mcs-plugins 4.4.2 Oct 18 13:03:12 03Otavio Salvador 07org.openembedded.dev * r7da643e378 10openembedded.git/conf/checksums.ini: checksums.ini: add xfwm4 4.4.2 Oct 18 13:03:55 hi Oct 18 13:05:34 Any specific reason why OM still uses libpam 0.79? Oct 18 13:08:01 otavio mickeyl is conservative with updates Oct 18 13:09:01 mickey|bbl: https://projetos.ossystems.com.br/git/?p=users/otavio/org.openembedded.dev.git;a=summary <= there're fixed for libpam and with those you might consider using libpam 1.0.2 on OM and then we could drop 0.79 ;-) Oct 18 13:09:16 woglinde: thx Oct 18 13:09:28 woglinde: i hope he takes a look on that later Oct 18 13:11:55 03woglinde 07org.openembedded.dev * r8b78fa6cc5 10openembedded.git/packages/uclibc/ (8 files): Oct 18 13:11:55 uclibc.inc: do not install ldconfig and ldd to staging Oct 18 13:11:55 * do not install this target binaries to install Oct 18 13:11:55 they are not useful and were installed into a Oct 18 13:11:56 false directory Oct 18 13:11:58 * bump PR on all affected recipes Oct 18 13:13:30 woglinde: is X usable at uclibc? Oct 18 13:13:42 woglinde: full Xorg I meant Oct 18 13:15:40 * mwester wanders in, with a very large cup of coffee... Oct 18 13:15:59 Good Morning Oct 18 13:16:39 go the mew Oct 18 13:25:10 gdb fails to build: configure: error: `LDFLAGS' has changed since the previous run Oct 18 13:25:55 The difference is an extra space in the LDFLAGS -- has anyone encountered this? (very recent change must have caused this) Oct 18 13:26:06 zecke: right, yeah, PF would be better Oct 18 13:26:09 than STAMP Oct 18 13:31:42 otavio hm I think yes but I didnt test it yet Oct 18 13:33:27 pb_: I will send two patches to the list, feel free to comment Oct 18 13:36:15 mwester: I just saw exactly the same thing Oct 18 13:37:10 mwester: I suspect it's related to the TARGET_LINK_HASH_STYLE somehow Oct 18 13:37:24 I found that koen sent an email yesterday; apparently it's a casualty of the "enable --hash-style=both" change. Oct 18 13:38:29 oops, I might have introduced a whitespace change... Oct 18 13:38:32 I avoid toolchain stuff as much as possible because I fear black magic... so it will have to wait for one of the Master Mages to fix it... Oct 18 13:38:45 mwester: what is this previous run gdb is talking about? Oct 18 13:39:00 black magic? :) Oct 18 13:39:09 It seems to run configure twice, according to the log... Oct 18 13:39:18 I can pastebin it. Oct 18 13:39:41 ah no, I will have to do university assignments now Oct 18 13:41:12 Ok. Oddly enough, it does the configures in the do_compile task. :( Very dark magic indeed... Oct 18 14:00:37 otavio: yes it is. some time ago i ran x on an arm920t with uclibc Oct 18 14:02:08 cpo_: awesome! Oct 18 14:02:22 but it was a bit tricky Oct 18 14:02:32 hehe Oct 18 14:02:36 i can imagine Oct 18 14:03:16 you have to configure uclibc in a special way to compile the X font stuff Oct 18 14:03:38 cpo_ hm is locales the problem? Oct 18 14:04:31 dont remember. cant find the bug in the bugtracker Oct 18 14:04:37 looks like i cant search for my name Oct 18 14:20:17 zecke: right, I see your patches. I will reply on the list. Oct 18 14:26:23 * RP will wait for that reply before replying Oct 18 14:28:38 03Mike Westerhof 07org.openembedded.dev * r5443f24a52 10openembedded.git/packages/meta/slugos-packages.bb: SlugOS: slugos-packages - demote gdb due to build failure Oct 18 14:32:54 mwester: I can't figure out how one space dissapears in 2nd configure run Oct 18 14:34:35 mwester: the first space is introduced in bitbake.conf where TARGET_LDFLAGS is set in case you do not have TARGET_LINK_HASH_STYLE set Oct 18 14:35:20 mwester: and the second one comes from gdb.inc: LDFLAGS_append = " -s" Oct 18 14:35:57 mwester: but the question is why one of them dissapears in 2nd configure run? Oct 18 14:36:04 RP: sorry, I got distracted by something else for a moment. anyway, I have sent my mail now. Oct 18 14:36:39 mwester: it might be easier to just set TARGET_LINK_HASH_STYLE in distro conf to mask the issue :) Oct 18 14:37:14 kalev: I suspect config.status is stripping out redundant whitespace when it reconstitutes the configure line. Oct 18 14:37:42 that would make sense, yes Oct 18 14:37:52 I wouldn't have thought it should be so hard to stop this duplicated space getting in there. Oct 18 14:38:38 oh, right, I see, it happens when TARGET_LINK_HASH_STYLE expands to nothing. Oct 18 14:38:43 yes Oct 18 14:39:38 I cannot link qpegps on stable branch. ld does not find libstdc++, the standard lib !?!. "warning: libstdc++.so.6, needed by ... not found (try using -rpath or -rpath-link). Has anybody an idea? Oct 18 14:39:51 well, you could rewrite that part of TARGET_LDFLAGS in terms of ${@base_conditional}, which would admittedly make it a bit uglier but would avoid this particular problem. or you could have the particular gdb recipe run the flags through s/ / / before it calls configure for the first time. Oct 18 14:40:32 the latter might be a better answer, given that the problem does seem to be localised to gdb. there's no reason why having repeated whitespace in LDFLAGS is inherently illegitimate, I think the configury in there is just being over-sensitive in this case. Oct 18 14:41:39 ok, I'll do the latter and if more similar issues come up we can always do the global fix later Oct 18 14:41:54 thanks pb__ Oct 18 14:42:17 ok, good luck Oct 18 15:26:23 mwester: I posted the fix to http://bugs.openembedded.net/attachment.cgi?id=11780 Oct 18 15:27:43 mwester: I don't have git push access, so you might want to commit it if it fixes your problem Oct 18 15:29:25 hi, my oe repository doesn't have xserver-kdrive-fbdev Oct 18 15:31:16 that's very sad. which oe repository is that? Oct 18 15:32:06 pb_: http://openzipit.org/Z2Develop/#GetTheFiles Oct 18 15:35:44 03Michael 'Mickey' Lauer 07org.openembedded.dev * re769c7fbe1 10openembedded.git/packages/freesmartphone/frameworkd_git.bb: frameworkd git: ship yaml rules in -config package and tag them as CONFFILE Oct 18 15:42:30 pb_: i built xserver-kdrive, but how do i launch it? Oct 18 15:43:23 just run it. "Xfbdev" or whatever the binary is called. Oct 18 15:48:26 shoragan: ping Oct 18 15:49:09 zecke: I doubt he is online. At least he is not in jabber Oct 18 15:49:50 I wonder how many of us are on jabber these days :) Oct 18 15:50:45 pb_: how do i find out what the binary is? Oct 18 15:50:48 RP: I guess a lot. IMHO the only sane IM :) Oct 18 15:51:08 * RP still has a soft spot for irc though Oct 18 15:51:48 RP: Same here. Different scope. I use IM for private stuff, IRC for devel and public discussions Oct 18 15:52:41 echelon: check the contents of the .ipk file Oct 18 15:54:11 thanks Oct 18 15:54:58 what package does libts belong in? Oct 18 15:55:21 tslib? Oct 18 15:59:30 echelon: yes Oct 18 16:01:12 thanks Oct 18 16:02:02 i just installed matchbox.. how would i run it? just call x-window-system? Oct 18 16:38:13 re Oct 18 17:10:43 03Holger Hans Peter Freyther 07shared/file-pr-revert * r245f7d129f 10openembedded.git/classes/ (5 files): WorkInProgress: Mangle the DISTRO_PR into the package revisions Oct 18 17:10:59 03Phil Blundell 07org.openembedded.dev * rbf1158d4e1 10openembedded.git/packages/gcc/gcc-cross-kernel.inc: gcc-cross-kernel: defeat fixincludes Oct 18 17:11:07 03Phil Blundell 07org.openembedded.dev * refacdf15ac 10openembedded.git/: Merge branch 'org.openembedded.dev' of git@git.openembedded.net:openembedded into org.openembedded.dev Oct 18 17:11:13 03Phil Blundell 07org.openembedded.dev * r2a6217b27d 10openembedded.git/packages/gcc/gcc-configure-common.inc: gcc: respect ${layout_includedir} Oct 18 17:20:48 03Holger Hans Peter Freyther 07shared/file-pr-revert * r8b21a87d8b 10openembedded.git/ (9 files in 2 dirs): WorkInProgress: Mangle the DISTRO_PR into the package revisions Oct 18 17:21:28 RP: oops, I had hoped you do git-commit --amend and remove the WIP note :) Oct 18 17:21:47 ich__: Well, its an RFC now Oct 18 17:21:53 Let me change that Oct 18 17:22:36 *puh* I assumed it was in .dev already Oct 18 17:22:46 okay food first Oct 18 17:22:55 No, I was going to be good and ask for agreement first :) Oct 18 17:23:09 hehe Oct 18 17:25:31 he rp :) Oct 18 17:27:09 ok, better commit description included now Oct 18 17:28:10 * RP -> food too Oct 18 17:55:51 hi sweetlilmre Oct 18 18:12:30 hi Oct 18 18:13:00 woglinde: hi :) how are things? Oct 18 18:13:42 i'm getting issue building shasum really first stage of OE. i follow the wiki OE install, but same. I'm on 64 bits linux. My previous install was running on 32 bit system Oct 18 18:14:40 BabelO: I have heard of 64bit issues from others... not sure what to suggest, but I think that || make may cause issues Oct 18 18:15:49 sweetlilmre: but i disable those two line BB_NUMBER_THREADS and PARALLEL_MAKE Oct 18 18:16:14 BabelO: then that is not the issue, sorry. Oct 18 18:16:33 sweetlilmre: other strange thing is that log is empty Oct 18 18:16:45 sweetlilmre: and i can start the run.compile...... by hand Oct 18 18:16:57 it compile fine the app in the output Oct 18 18:17:10 BabelO: truely weird! Oct 18 18:18:40 :( Oct 18 18:22:53 BabelO: what's the error message you get? Oct 18 18:24:03 kalev: ERROR: log data follows (/home/OE/build/tmp/work/x86_64-linux/shasum-native-1.0-r1/temp/log.do_compile.9204) Oct 18 18:24:16 ERROR: function do_compile failed Oct 18 18:25:18 kalev: full output http://pastebin.com/m688ef008 Oct 18 18:25:19 hi kalev Oct 18 18:36:03 kalev: can i bypass the shasum build by copying the executable ? or do i need to do something else ? Oct 18 18:37:15 you need sgasum anyway Oct 18 18:37:32 otherwise oe will coomplain about missing checksums Oct 18 18:39:02 woglinde: but i mean the application compile fine if i start the script in subdirectory by hand Oct 18 18:39:18 so i don't understand why it say it fail Oct 18 18:40:06 babelo what is in /home/OE/build/tmp/work/x86_64-linux/shasum-native-1.0-r1/temp/log.do_compile.9204 ? Oct 18 18:40:21 it is empty Oct 18 18:40:33 eh Oct 18 18:41:20 hm Oct 18 18:41:25 do_compile is Oct 18 18:41:33 $CC $CFLAGS $CPPFLAGS -o oe_sha256sum main.c sha256.c Oct 18 18:41:42 so either your CC is setting wrong Oct 18 18:41:48 or the CLFAGS Oct 18 18:42:25 woglinde: if i start Oct 18 18:42:27 tmp/work/x86_64-linux/shasum-native-1.0-r1/temp/run.do_compile.9204 Oct 18 18:42:30 by hand Oct 18 18:42:32 it works Oct 18 18:42:41 and compile oe_sha256sum Oct 18 18:43:05 BabelO: no idea but generally if such a simple app does not work... there is something wrong with your setup Oct 18 18:44:02 hm Oct 18 18:44:16 we have to see the output from this compile line Oct 18 18:44:26 but I do not know how to get it Oct 18 18:49:54 hey woglinde Oct 18 18:50:25 maybe something like that: sh -x /home/OE/build/tmp/work/x86_64-linux/shasum-native-1.0-r1/temp/run.do_compile.9224 Oct 18 18:52:05 i ve tried strace bitbake minimal-image Oct 18 18:52:53 kalev hm Oct 18 18:53:03 kalev: with sh -x everything is ok Oct 18 18:53:14 + ccache gcc -isystem/home/OE/build/tmp/staging/x86_64-linux/usr/include -O2 -isystem/home/OE/build/tmp/staging/x86_64-linux/usr/include -o oe_sha256sum main.c sha256.c Oct 18 18:53:17 this is last line Oct 18 18:53:56 maybe i v e to remove ccache Oct 18 18:54:46 hm Oct 18 18:54:53 have you installed ccache? Oct 18 18:55:21 woglinde: yes i ve installed it else it does not call it Oct 18 18:55:24 maybe the problem lies there Oct 18 18:55:30 output of sh -x http://pastebin.com/d297af43f Oct 18 18:56:35 yeah but thats not called inside from bitbake Oct 18 18:56:54 so it doesnot help to find th error Oct 18 18:57:18 anyway if i call the lines by and they works Oct 18 18:57:47 is there a debug command in bitbake Oct 18 18:59:32 you can try setting CCACHE="" in local.conf and see if it changes anything, but I doubt it has much impact Oct 18 18:59:39 hm do not know if it will by showed theire Oct 18 18:59:58 try bitbake -DDD Oct 18 19:01:07 hmm yes good debug, but it show fail after ccache command line Oct 18 19:01:18 i try to remove it and restart from scratch Oct 18 19:04:11 BabelO: if it makes you feel any better, I have just had to do basically that :) Oct 18 19:07:25 ok, woglinde, kalev, sweetlilmre that was the bug, i desintall ccache and all works fine now Oct 18 19:07:34 thanks Oct 18 19:07:52 BabelO: glad it works :) Oct 18 19:08:42 babelO hm the question is why your ccache is broken Oct 18 19:08:43 yess, this problem with ccache is very strange, i install it because wiki say it can be faster with that Oct 18 19:08:50 jepp Oct 18 19:08:53 for recompiling Oct 18 19:09:04 good :) Oct 18 19:09:09 it only needs to copies the .o files Oct 18 19:09:22 instead of making them new Oct 18 19:09:41 ok maybe opensuse 11.0 version of ccache is wrong for your need Oct 18 19:09:51 but the command line was working Oct 18 19:11:40 BabelO: I heard mru or someone say something about ccache being borked on 64bit. look in the #beagle logs Oct 18 19:12:01 sweetlilmre: ok thanks i ll look Oct 18 19:12:29 BabelO: it was a while ago and late so I may be mistaken, i hope not Oct 18 19:13:21 sweetlilmre: ok, now i started with parallel build and make to see Oct 18 19:14:00 w00t! after cleaning everything || make is blasting! :) Oct 18 19:16:02 sweetlilmre: i ve a Q6600 cpu with 8Gb of ram so i hope it is fast to build wioth that Oct 18 19:16:13 BabelO hehe Oct 18 19:16:19 wish I had this machine Oct 18 19:16:27 BabelO: 2.4 quad here, but building under VMWare so only 2 cpus Oct 18 19:16:46 why vmware ? Oct 18 19:17:10 BabelO: haven't made the transition to the One True OS (tm) yet ... :) Oct 18 19:18:32 BabelO: xp host, I am still a windoze boy Oct 18 19:19:13 sweetlilmre: my actual main problem is to get a angstrom toolchian x64 to continue doing devel for htc-linux mobile devices Oct 18 19:19:39 sweetlilmre: linux since 1998 here :) Oct 18 19:19:47 only that at home Oct 18 19:21:22 BabelO are you htc kernel hacker? Oct 18 19:21:45 woglinde: yes for htc artemis and blueangel device Oct 18 19:22:02 cool Oct 18 19:22:30 when I grow up I want to be a kernel hacker :) Oct 18 19:22:45 problem is, I am already 34! :) Oct 18 19:22:45 sweetlilmre hege Oct 18 19:23:05 sweetlilmre hehe sometimes I am on the way too it Oct 18 19:43:51 gn all. Oct 18 19:44:35 nite sweetlilmre Oct 18 20:02:15 03Otavio Salvador 07org.openembedded.dev * r34f093e365 10openembedded.git/conf/checksums.ini: checksums.ini: add terminal 0.2.8 Oct 18 20:07:26 03Otavio Salvador 07org.openembedded.dev * r3a44c41ea3 10openembedded.git/packages/tasks/task-xfce-base.bb: task-xfce-base: use splited xfce4-plugins Oct 18 20:16:44 gregoiregentil: thanks by reporting the xfce task issue; please ping me if you find anything else wrong with it Oct 18 20:19:37 kalev: woglinde: minimal-image just finishing, thanks for help Oct 18 20:20:10 BabelO hehe Oct 18 20:20:11 fine Oct 18 20:24:33 otavio are you some git expert? Oct 18 20:25:01 hi like2wise Oct 18 20:26:15 woglinde: no but i'm used to it; i can try to help Oct 18 20:26:24 woglinde: what is your problem with git? Oct 18 20:28:58 otavio ah sorry zecke is now on again Oct 18 20:29:02 and I am asking him Oct 18 20:32:54 woglinde: ok, np Oct 18 20:33:47 git-stash I now learned Oct 18 20:34:05 03woglinde 07org.openembedded.dev * r8b5e3f5fce 10openembedded.git/packages/libsdl/ (files/fixmfour.patch libsdl-x11_1.2.11.bb): Oct 18 20:34:05 libsdl-x11: fix sdl.m4 for sdl-config Oct 18 20:34:05 * there is a stupid PATH setting which is using the host Oct 18 20:34:05 stuff, just removed it and our staging sdl-config is found Oct 18 20:34:05 * bump PR Oct 18 20:34:15 03woglinde 07org.openembedded.dev * rf5908addf3 10openembedded.git/packages/libsdl/libsdl-image_1.2.6.bb: Oct 18 20:34:15 libsdl-image: remove the integrated stupid sdl.m4 packages Oct 18 20:34:15 * remove the old sdl.m4 likewise the libtool.m4 file Oct 18 20:34:16 * bump PR Oct 18 20:36:32 git stash works Oct 18 20:36:35 and rocks Oct 18 20:36:56 so sdl is now fixed Oct 18 20:37:05 :-) Oct 18 20:37:05 *g* lazy developers again Oct 18 20:37:34 dont think about that the user maybee sets a custom PATH Oct 18 21:41:29 woglinde: do you have time to review a few patches to oe? Oct 18 21:41:39 kalev yes Oct 18 21:42:39 http://git.smartlink.ee/?p=openembedded.git;a=shortlog;h=org.openembedded.dev Oct 18 21:43:25 you can pull from the repository too, cd to your repo and do "git pull git://git.smartlink.ee/openembedded.git org.openembedded.dev" Oct 18 21:48:46 kalev aeh I will create a local branch for it Oct 18 21:48:50 org.openembedded.dev" Oct 18 21:48:54 is my main Oct 18 21:49:25 sure Oct 18 21:51:01 kalev: woglinde the patches look good imo Oct 18 21:57:34 thanks. Can you push them please? Oct 18 21:58:20 hm Oct 18 21:58:28