**** BEGIN LOGGING AT Fri Jun 14 02:59:58 2013 Jun 14 05:47:46 hi when i was running the clearstate command Jun 14 05:47:49 i am getting Jun 14 05:47:50 autoconf_2.69.bb, do_compile) failed with exit code '1' Jun 14 05:47:54 as error Jun 14 05:48:09 any help would be really helpful Jun 14 05:48:17 :P Jun 14 07:18:25 hi. is it possible to register a commercial project as Yocto Project Compatible, even if it includes non-free software? assuming that all patches to oe-core, meta-oe, .. are contributed back, and all non free 'stuff' is located into a separate (non public) layer. Jun 14 07:30:57 good morning Jun 14 07:45:52 I have a package with a .tar.gz that doesn't have a toplevel directory, say, foobar-1.0. Is it possible to twek the unpack stage so that it creates one? Jun 14 08:08:52 morning all Jun 14 08:15:33 good morning Jun 14 10:53:02 hello everyone, I am trying to install ADT plugin in Eclipse Indigo Jun 14 10:53:29 followed the steps from documentation (installed Autotools as well) Jun 14 10:53:55 but eclipse update manage is whining about missing dependency Jun 14 10:54:52 Missing requirement: Yocto Project ADT Plug-in 1.4.0.201306121752 (org.yocto.sdk.feature.group 1.4.0.201306121752) requires 'org.eclipse.cdt.autotools.core 1.1.0' but it could not be found Jun 14 10:55:16 autotools is installed though... any clues? any success stories? Jun 14 11:53:02 synthnassizer: join an restate question -> #eclipse Jun 14 11:53:42 synthnassizer: maybe alsow try #eclipse-dev Jun 14 13:18:32 khem: ping Jun 14 13:24:10 JaMa: ping Jun 14 13:27:55 JaMa: http://cgit.openembedded.org/meta-openembedded-contrib/commit/?h=jansa/in-test&id=8122a8f57690c1861c1fe5cb8cf7c1d0bb816de3 should be completely replaced with [PATCH v2] llvm: move cross script install to separate sysroot preprocess function Jun 14 13:29:13 synthnassizer: unfortunately the folks who might be able to help are either away or asleep atm... if you can wait until a bit later I might be able to point someone your way Jun 14 13:29:32 synthnassizer: alternatively you could email the mailing list Jun 14 13:29:55 thank you very much. I will wait, if you mean the meta-freescale, I already have Jun 14 13:30:39 @bluelightning: thanks Jun 14 13:31:02 synthnassizer: I mean the ADT question Jun 14 13:31:51 JaMa: it seems you squashed "llvm: install llvm-config in crossscripts" into the commit which is to be disregarded - see http://patchwork.openembedded.org/patch/50959/ Jun 14 13:32:42 bluelightning: I haven't got any response back about getting an account registered for patchwork yet. Jun 14 13:33:09 Net147: khem has been away, he's back now, hopefully he can sort you out with an account Jun 14 13:33:32 I think I might raise this at the next TSC meeting, try to get the issue of patchwork account registration solved finally Jun 14 13:33:51 OE TSC that is Jun 14 13:34:05 bluelightning: I find a reCAPTCHA helps with spam Jun 14 13:34:23 Net147: right, we could probably integrate that without too much trouble Jun 14 13:35:42 #bluelightning yes, the date of my post/email is 2013-06-13 Jun 14 13:36:15 bluelightning: spam registrations dropped to zero when I implemented it on a forum Jun 14 13:36:21 bluelightning, solve the world spam problem also Jun 14 13:37:01 Crofton|work: it's already on my todo list, right next to "achieve world peace" Jun 14 13:37:04 ;) Jun 14 13:37:29 it is so frustrating to deal with Jun 14 13:37:55 synthnassizer: oh, you already sent something? to the yocto@yoctoproject.org mailing list? Jun 14 13:37:56 bluelightning: I have also posted this in https://community.freescale.com/message/334427#334427 Jun 14 13:38:27 to meta-freescale@yoctoproject.com, you believe I show try yocto@ ? Jun 14 13:38:52 synthnassizer: if it's about just being able to get started with the ADT I would think so yes Jun 14 13:39:06 alright I will try. thanks Jun 14 13:40:40 bluelightning: argh IRC question. how do you notify me of a msg? just "synthnassizer:" in the beginning of every post towards me? or is there some other operator too? Jun 14 13:41:06 synthnassizer: yes, it's IRC convention Jun 14 13:41:34 synthnassizer: most IRC clients should highlight messages containing your nick and possibly even notify you some other way at the same time Jun 14 13:41:51 so ":" is not necessary Jun 14 13:42:06 synthnassizer: I don't think most clients care about that, no Jun 14 13:42:23 bluelightning: cool, thank you! Jun 14 13:42:32 my client puts it in automatically if I tab-complete the nick at the start of the message Jun 14 13:43:09 whatever you have you definitely want a client with tab completion of nicks :) Jun 14 13:44:24 bluelightning: I don't have tab completion for nicks, so I get typing practice =) Jun 14 13:45:00 bluelightning: but if I am lazy I can highlight nick in window to copy to clipboard Jun 14 13:45:13 Net147: I guess that works too :) Jun 14 13:46:20 bluelightning: which OS are you on? Jun 14 13:46:30 Net147: Kubuntu Jun 14 13:49:14 bluelightning: sounds good Jun 14 13:52:48 bluelightning, i have XchatIRC, I just hadn't used IRC for many years. Jun 14 13:53:49 synthnassizer: do you have the TAB key on your kb? Xchat is just fine Jun 14 13:54:32 bluelightning, yes yes, i tried it already. tab completion rocks :) Jun 14 13:54:39 you can have automatic completion even, w/out tab Jun 14 13:54:57 bluelightning: apparently I have tab completion too =P Jun 14 13:55:42 bluelightning: learn something new every day Jun 14 14:05:13 With Poky and systemd, what is responsible for mounting /var/volatile? I am having problems because it is mounted too late, so that when e.g. run-postinsts.service runs, /var/log (which is a link to /var/volatile/log) does not exist yet, and all the postinst scripts fail because they cannot write to /var/log/postinstall.log. Jun 14 14:08:57 Saur: what version of Poky you using? Jun 14 14:09:28 Net147: master, somewhere after 1.4... Jun 14 14:09:51 Saur: I though I fixed this in http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=0116e82631a0267be3e5b7c04007980cb4e10171 Jun 14 14:10:02 Saur: do you have that change? Jun 14 14:10:38 Net147: Nope, which could explain the problem :) Jun 14 14:10:54 ah, I thought I'd seen that fix as well Jun 14 14:11:11 Time to move my master forward some I guess... :) Jun 14 14:16:58 http://git.yoctoproject.org/cgit/cgit.cgi/poky/commit/?id=07ba7936d431f5d2d56ad7d2b6ac29d65edf4a0f - it was committed 336 seconds before it was authored? =P Jun 14 14:18:32 RP: are you time travelling? Jun 14 14:19:00 Net147: interesting :) Jun 14 14:19:18 Net147: I can only assume the mail headers were confused Jun 14 14:20:17 RP: you must have signed the patch off pretty quickly Jun 14 14:23:17 umm... Package version for package kernel-module-xfrm-ipcomp went backwards which would break package feeds from (0:3.8.13+gitedd6461602_0+1f973c0fc8-r4.1.1.3 to 0:3.8.13+git1+edd6461602_1f973c0fc8-r4.1.1.3) Jun 14 14:27:12 Net147: right, that's expected if you built over the change to shortened git revisions in versions Jun 14 14:27:29 Net147: unfortunately that change may mean you need to clean TMPDIR Jun 14 14:27:44 bluelightning: will bitbake -c cleansstate linux-yocto do? Jun 14 14:28:37 bluelightning: I don't see any shortening of revision ids in the error message Jun 14 14:29:10 Net147: ah, I misread... it looks like the format changed there Jun 14 16:31:16 Net147: probably fallout from http://git.yoctoproject.org/cgit.cgi/poky/commit/?id=cd7b7de91a98e712e796a8d6a3a8e3741950396e :( Jun 14 16:31:34 yo Jun 14 16:44:27 RP: do we have way to manually add a dependency on a file's contents, basically the equivalent of what bitbake is doing with the local file:// files? I'm thinking probably not, but it could be useful from time to time, for example when calling an external script from a layer Jun 14 16:44:56 RP: maybe a 'filedeps' flag is in order Jun 14 16:45:25 kergoth: you could add it to SRC_URI? :) Jun 14 16:45:41 kergoth: seriously, I don't think we have exposed the API but as you say, we have the infrastructure there Jun 14 16:45:53 true. don't need it there, but that'd work for now Jun 14 16:47:02 I kind of like the idea of introducing it as a flag, and setting one on SRC_URI with the resolved file:// paths, would probably help reduce the tight binding between bb.fetch2 and the rest of the code Jun 14 16:47:06 * kergoth ponders Jun 14 17:03:03 RP: if i call a def'd function from a ${@} block inside a vardeps flag, and use 'd' in that function, it works fine. if i do a getVar() from that def'd function, with expand=True, and *taht* variable has a ${@} block that uses 'd', it will get a NameError saying d doesn't exist. I'm guessing there's something with our context setup that isn't 100% at that point Jun 14 17:03:11 * kergoth 'll nail down a reproducer and open a bug Jun 14 17:03:35 * kergoth scratches head Jun 14 17:21:52 seebs, ping? Jun 14 17:48:22 Net147: still around? Jun 14 17:48:39 Net147: can you send me patchwork IDs for last versions of llvm patches you want to merge? Jun 14 17:52:51 are 4 changes in http://patchwork.openembedded.org/user/bundle/79/?q=llvm&archive=both right set? Jun 14 17:55:08 Net147: and it still looks like llvm-config will be missing in runtime packages Jun 14 18:25:29 JaMa: Net147 is Australia based, he's probably asleep by now! Jun 14 18:39:08 i'm not finding a recipe for atop Jun 14 18:39:30 am i missing something or should i just make one? Jun 14 18:47:20 found an acct recipe, but still no atop Jun 14 19:29:27 sgw_: OK, it's in my queue now, so it will take 3 days to build it again :) Jun 14 21:31:05 I'm thinking about how to make my continuous integration builds faster...right now each build is clean except for the download and sstate caches being preserved. However, the re-packaging that happens after an sstate cache hit seems to soak up all of the time saved by sstates.... Jun 14 21:31:18 Thoughts, anyone? Jun 14 21:58:06 Just pulled from poky master and did a build of core-image-sato for qemux86. It failed on task 5601 of 5603 do_rootfs for recipe core-image-sato. Jun 14 21:58:22 What would a somewhat more experienced person do in this situation. Jun 14 21:58:26 ? Jun 14 22:09:34 mulhern: well...according to the log, src/set_ld.c:234 of mpfr has an unrecognizable instruction Jun 14 22:09:53 mulhern: someone integrate a new version of some toolchain component? Jun 14 22:11:11 Which log? Jun 14 22:11:24 http://autobuilder.yoctoproject.org:8011/builders/nightly-fsl-ppc/builds/143/steps/Building%20Images/logs/stdio Jun 14 22:12:25 mulhern: oh, crap. that wasn't you that said that Jun 14 22:12:56 mulhern: sorry, YoctoAutoBuilder spat some stuff out right in between two lines by you Jun 14 22:13:47 mulhern: but anyway, first thing I'd do is look at the log for the failing do_rootfs ;-) Jun 14 22:23:01 I'm a little bit confused about the different builds. The log you mentioned shows some sort of a syntax error set_ld.c as you point out in a do_compile step. But I've got an error having to do with dependencies in a do_rootfs step which is way later. Is there a way to match up the current state of my build with what the builder is building somehow? Jun 14 22:24:33 mulhern: ...probably.... that particular build claims to be for PowerPC ("nightly-fsl-ppc"), you probably just saw YoctoAutoBuilder announce one for ARM ("nightly-fsl-arm") Jun 14 22:25:04 mulhern: I assume there's one for qemux86 too, or at least there's one in the queue Jun 14 22:25:42 mulhern: for 100% precision you'd want to look at which commit SHA was built, of course Jun 14 22:26:33 mulhern: my project's on 1.2 currently so I don't pay a lot of attention to master, sorry Jun 14 22:26:58 Yes, what does the lsb suffix stand for? Jun 14 22:29:44 mulhern: maybe Linux Standard Base Jun 14 22:35:07 mulhern: I suspect the *-lsb builds are with DISTRO == poky-lsb Jun 14 22:38:26 AFAIK that's correct Jun 14 22:38:51 evanp: What is 1.2? Jun 14 22:38:53 sometimes the do_rootfs failures have a useful log message and sometimes not... Jun 14 22:39:19 i would assume yocto-1.2 ? Jun 14 22:39:27 yes, correct Jun 14 22:39:30 mulhern: i'm building off master Jun 14 22:39:43 for raspberrypi anyway... Jun 14 22:40:32 you might want to update, remove tmp and your cache dirs and do a full clean build Jun 14 22:40:57 did you try just "bitbake image-name" again? Jun 14 22:41:25 mr-science: Yup, it's repeatable. Jun 14 22:41:47 when was your last update on master? Jun 14 22:42:15 hi, anyone here know if systemd should work for the beagleboard BSP (latest release) ? Jun 14 22:42:18 mr-science: I've just backed up to the last commit on the most recent succeeding build. I hope that will be a success. Jun 14 22:42:44 Kernel boot hangs with systemd[1]: Failed to mount /run: No such file or directory Jun 14 22:43:23 if you can spare the time, i'd do the remove tmp and cache dirs first Jun 14 22:43:41 mr-science: Why? Jun 14 22:44:10 i've had some odd failures after git pull if i just leave everything intact Jun 14 22:44:27 cleaning and rebuilding seems to take care of it Jun 14 22:46:12 winfried: my pi build is a simple xorg-nodm build that doesn't use systemd, but IIRC the /run thing and related paths/volatile config was just updated recently Jun 14 22:46:45 which branch are you building from? Jun 14 22:47:31 if you're building off current master then any image with systemd in distro features should work i would think Jun 14 22:47:31 Thanks, at the moment just the 1.4 ti BSP downloaded from the yocto home page. Jun 14 22:47:49 I'll try to use latest git. Jun 14 22:48:07 mr-science + evanp: But what does yocto-1.2 stand for? A future release? Jun 14 22:48:17 older Jun 14 22:48:27 1.4 is current i think Jun 14 22:56:32 anyone seen this? Exception: AttributeError: 'module' object has no attribute 'pool' Jun 14 22:56:42 latest bitbake and oe-core Jun 14 22:56:57 JaMa: [14:01:42] roric_: Fix pushed Jun 14 22:57:00 JaMa: your patches in-stage look correct Jun 14 22:57:01 refers to this, i think Jun 14 22:57:08 not sure though, haven't tested master in a while Jun 14 22:57:55 kergoth: thanks, I'll try to update, but I did like 30 minutes ago.. complete log http://pastebin.com/mrJaQ3DX Jun 14 22:58:13 JaMa: http://cgit.openembedded.org/meta-openembedded-contrib/log/?h=jansa/in-test has the correct patches Jun 14 22:58:33 ok missed one bitbake commit which looks like fix Jun 14 22:58:40 Net147: ok, thanks for confirmation Jun 14 22:59:07 Net147: what you think about missing config script in runtime package (I must admit I haven't checked this yet) Jun 14 22:59:17 JaMa: you want llvm-config to be able to run on target? Jun 14 23:00:17 I don't need it (at least not yet), but it's a possibility for someone to expect it e.g. in PN-dev Jun 14 23:01:13 but if it hardcodes all sysroot paths and cannot be just installed with do_install then we can postpone this until someone will really need it on target Jun 14 23:07:45 JaMa: would probably need to compile twice with different hardcoded paths or change to use environment variables like we do for qmake Jun 14 23:08:07 JaMa: at least with LLVM 3.x which has the llvm-config script written in C++ instead of Perl Jun 14 23:12:19 Net147: fair enough for me Jun 14 23:12:40 bluelightning: JaMa: do we know what qconfig.pri is supposed to be used for qt4-embedded? currently it's always pointing to the qt4 qconfig.pri... Jun 14 23:13:24 Net147: not off the top of my head Jun 14 23:14:04 neither do I Jun 14 23:14:31 I never used qt4-embedded except building it for tests Jun 14 23:15:05 sgw_: do you have /usr/share/qtopia/mkspecs/qconfig.pri or similar in target sysroot? Jun 14 23:22:30 LOL bitbake-worker decafbad Jun 15 00:18:14 Net147: yes, I have such a file in that location Jun 15 00:19:05 sgw_: I submitted a patch to fix the environment variable to point to the right one Jun 15 00:19:28 Net147: yeah, I saw that I am about to start a build Jun 15 00:49:20 Net147: hmm, I am still getting the qgl.h not found. I know it's in my qt4 include dir, but not in the qtopia side. Jun 15 00:49:41 Net147: do I need to clean something? This is a "world" build BTW Jun 15 00:58:46 sgw_: I had an error building qt4-embedded. it went away after I fixed the environment variable, did cleansstate qt4-embedded Jun 15 00:59:30 sgw_: I am building at the moment. it's currently at qt4-embedded do_package stage. after, it's going to build qt-mobility-embedded. Jun 15 00:59:57 Ok cleaning qt4-embedded Jun 15 01:00:20 maybe it's worth cleaning qt-mobility-embedded too just to be sure there is nothing left over? Jun 15 01:01:13 done and done! **** ENDING LOGGING AT Sat Jun 15 02:59:58 2013