**** BEGIN LOGGING AT Thu Mar 27 02:59:58 2014 Mar 27 07:02:45 morning Mar 27 07:26:50 morning Mar 27 07:50:28 good morning Mar 27 08:14:14 howdy Mar 27 08:14:21 anyone here know what nfs-client-utils is? Mar 27 08:55:49 JaMa: Thank you for the feedback on my patches. In the LICENSE of python-twisted only list of contributors changes; but I'm wondering if the contributors should be considered part of the license at all? https://twistedmatrix.com/trac/browser/trunk/LICENSE Mar 27 08:56:14 *only list of contributors changed Mar 27 09:24:19 Anyone here know what the "Running intercept scripts:" part of the image.do_rootfs part is? Mar 27 09:24:30 I am getting weird issues with Yocto 1.5 angstrom Mar 27 09:24:44 in the last strages of the image build process Mar 27 09:24:50 (do_rootfs) Mar 27 09:25:00 it is looking for packages that have not been built Mar 27 09:25:07 because they are not in the image Mar 27 09:25:09 why? Mar 27 09:25:21 this is new Mar 27 09:25:26 yocto 1.4 no problems Mar 27 09:25:32 yocto 1.5 big problems Mar 27 09:26:03 for example Mar 27 09:26:07 this is what I get Mar 27 09:26:08 ls: cannot access /home/griftw/w/danny/setup-scripts/build/tmp-angstrom_v2013_06-eglibc/sysroots/impedo-duo/pkgdata/runtime-reverse/sysvinit-pidof: No such file or directory Mar 27 09:26:08 | readlink: missing operand Mar 27 09:26:43 this happned with nfs-client-utils too Mar 27 09:27:03 onlyu after I did a manual "bitbake nfs-utils" did it fix the nfs-utils-client missing issue Mar 27 09:27:10 but now its on to the next one Mar 27 09:27:19 where are these things coming from? Mar 27 09:31:14 oh yea this is happening on a fresh build after nuking the build tmp folder Mar 27 09:31:22 so I might have broken it soehow that way Mar 27 10:12:25 diego_r: I don't mind if you keep them included Mar 27 10:15:02 morning all Mar 27 10:18:52 JaMa: wrt word-build failures for qemux86 linux-yocto-tiny is now under repair so I expect the -kexecboot version to build soon Mar 27 10:19:19 *world Mar 27 10:44:20 http://patchwork.openembedded.org/patch/57647/ Mar 27 10:44:39 do we know why gcc ptest adding patches were not applied to oe ? Mar 27 10:45:38 how can I automatically build u-boot when I bitbake some-image ? Mar 27 10:49:06 JaMa: ok, thank you. Mar 27 11:33:51 ant_work: thanks Mar 27 11:34:25 bluelightning: hi, do you remember why we have negative D_P in libav-9.10 (and libav-8 is prefered) and if we can change the default now? Mar 27 11:34:46 bluelightning: I'm upgrading mplayer2 because it's again broken by some other upgrades and newer mplayer2 needs newer libav Mar 27 11:35:15 JaMa: gstreamer-0.10 I think, plus being a bit conservative Mar 27 11:36:07 we could add some PROVIDES / DEPENDS to make the link a bit stronger Mar 27 11:38:18 at some point maybe we'll be able to drop gstreamer 0.10 altogether, I wonder if that is something we could do next release Mar 27 11:40:16 ok, I'm fine with saying in mplayer recipe that it needs P_V for libav for current release Mar 27 11:44:03 JaMa: is that obvious enough for the user though? i.e. where/how does it fail with 0.8.x ? Mar 27 12:08:27 bluelightning: configure checks for that Mar 27 12:08:49 bluelightning: unfortunately if you reexecute configure it will pass (hacky configure script in mplayer2.. ) Mar 27 12:09:32 bluelightning: well now it fails a bit later in compile because of incompatible libs, so even when it's not obvious for user it's improvement from current state Mar 27 12:12:25 when the prserver starts, should it show something on the console? should it start automatically? Mar 27 12:13:22 [NOTE: Started PRServer with DBfile: /OE/build/shr-core/cache/prserv.sqlite3, IP: 127.0.0.1, PORT: 44573, PID: 23644 Mar 27 12:40:32 doesnt show up here Mar 27 12:46:00 ah now Mar 27 12:48:42 diego_r: thanks for v3, looks good to me Mar 27 12:49:51 JaMa: you're welcome. Mar 27 13:03:56 [asking same as in #yocto ;-] anyone here can help out a quick question? Attempting to build a Yocto I.MX6 image, works nicely - but libSDL2 is missing from my target image (it is in the SDK) Mar 27 13:04:02 Do I need to do more than 'DEPENDS += 'virtual/libsdl2' and 'IMAGE_INSTALL_append += libsdl2 ? Mar 27 14:05:05 khem: re "17:44 < khem> all meta-toolchain recipes should either disappear soon" Mar 27 14:05:43 khem: one of the issues with populate_sdk task is it did not populate qmake and other things in the SDK -- is there a plan for how to fix that? Mar 27 14:14:27 cbrake: there is a bug open for that, one moment Mar 27 14:15:04 cbrake: https://bugzilla.yoctoproject.org/show_bug.cgi?id=5429 Mar 27 14:15:30 meta-toolchain-qt / meta-toolchain-qte won't be dropped until that is resolved Mar 27 15:07:21 Hi guys, is there a task to generate .dtb from bitbake ? Mar 27 15:08:59 or maybe I missed to set a variable somewher in my kernel recipe Mar 27 15:27:56 so, I need to set UBOOT_CONFIG or UBOOT_MACHINE in my BSP... just where do I find what I can sensibly set it to? Mar 27 15:36:31 Hodapp: the u-boot documentation Mar 27 15:37:07 Hodapp: i.e. u-boot's own manual; we just pass on the make command line for u-boot Mar 27 15:37:20 ohhhh... Mar 27 17:03:32 I'm using B_pn-mypackage = some/place/build-${PN} to have oe build in a directory I want, how can I have the arch or target name in that path Mar 27 18:10:44 Question for the recipe gurus: if I have a package with a git SRC_URI that uses ;tag= semantics, can I remove the SRCREV=${AUTOREV} ? Mar 27 18:10:59 don't use ;tag=. Mar 27 18:11:07 why Mar 27 18:11:36 forces it to contact the remote server on every parse even if you lock down SRCREV, iirc Mar 27 18:11:42 interferes with BB_NON_NETWORK capability Mar 27 18:11:48 s/NON/NO/ Mar 27 18:11:54 its simpler than a srcrv to match the tag for readability, and if i use a tag-string in srcrev, it requires asking the repo for the conversion Mar 27 18:11:57 use SRCREV. you can specify a tag there Mar 27 18:12:14 it will ask the repo when using ;tag= too Mar 27 18:12:19 I can specify a tag there, but the manual states that if i do that it will ask the repo every time Mar 27 18:12:32 again, it will do the same exact thing with ;tag=, it has to Mar 27 18:12:42 it needs the commit hash at parse time, period Mar 27 18:13:11 boo. Okay. Mar 27 18:15:15 SRCREV at least lets you exert more control over it. BB_CACHE_POLICY lets you tell it to only contact once and use that cached value rather than every parse, for example. iirc that doesn't work properly with ;tag= Mar 27 18:15:32 thats my understanding, anyway, I could be missing an aspect, but that's the behavior we've hit when trying to support BB_NO_NETWORK for our customers Mar 27 18:16:00 right, which is possibly something I'd rather not have to fight, but yeah Mar 27 18:16:06 instance of not wanting to fight that. Mar 27 18:16:21 I guess I will have to just deal with that, plus commentation Mar 27 18:16:25 I don't think using it in the url buys you much, if anything, at this point, though Mar 27 18:16:32 other than being slightly prettier Mar 27 18:16:33 :) Mar 27 18:16:47 I honestly hate SRCREV across the board, but we're stuck with it at this point :) Mar 27 18:17:03 * kergoth shrugs Mar 27 18:17:15 its prettier for what I care, and if not for it needing to ask, I'd have been happy with that. Mar 27 18:18:55 yeah, that would be nice. it's possible i'm missing something, but we had to bbappend some recipes to remove ;tag= to get it to stop contacting the remote server Mar 27 18:19:03 maybe we can improve bb.fetch2 in that regard Mar 27 18:19:45 I got tired of re-re-resubmitting my patches for "not patching cleanly to master" when I kept rebasing to master right before submitting... Mar 27 18:22:14 Because it would seem to me, honestly, that once you have a tag made it should not be changing Mar 27 18:22:27 I mean if you made that tag, deleted the tag, re-made the tag .. I suppose. Mar 27 18:22:35 technically its possible ot force push a changed tag with git Mar 27 18:22:41 just as its possible to push a non-fast-forward branch Mar 27 18:23:03 not advised, but.. :) Mar 27 18:35:42 blech, but meh, such is unavoidable at this point, I suppose. Mar 27 18:48:27 well, commit dropped up my tree. Boo. Mar 27 18:56:18 I am using do_split_packages to package the gnuradio libraries and their -dev packages Mar 27 18:56:40 I'd like to add all the -dev packages to a gnuradio-dev package as RDEPENDS Mar 27 18:56:47 any ideas on how to do this? Mar 27 19:36:35 Crofton: do_spit_packages provides built in support for that. see how kernel-modules is handled in the kernel classes Mar 27 20:11:20 thanks, but I am thinking I am tryin gto do something backwards, butI'll check there Mar 27 20:45:57 * Hodapp realizes that he's going to have a really hard time fitting an 8 MB kernel, plus root filesystem, from core-image-minimal... into 4 MB of flash on a WR703N. Mar 27 20:46:01 hmmm. Mar 27 21:01:59 oh. poky-tiny. Mar 27 21:02:49 does poky support squashfs? Mar 27 21:02:51 cant remember. Mar 27 21:03:55 I just built a poky image with squashfs, so I guess so Mar 27 21:04:07 I didn't boot it, so this is in theory Mar 27 21:05:10 ah Mar 27 21:05:24 but, I'm trying to target a WR703N Mar 27 21:05:41 and this is my first attempt at using OE/Bitbake/Yocto/Poky/Angstrom/whatever the heck I'm using Mar 27 21:05:41 does this make sense: d.appendVar('RDEPENDS_gnuradio-dev ', ' '+' '.join(pkgs)) Mar 27 21:06:03 where pkgs comes from do_split_pages and appears to have the right stuff Mar 27 21:06:08 my python is poor Mar 27 21:07:49 that looks sensible to me but should your first string have a trailing space, and your second one a leading space? Mar 27 21:08:24 I'd just like somehing to show up the in the gnuradio-dev package to start Mar 27 21:08:31 spaces I can work out later :) Mar 27 21:09:15 well, if the space is part of the name, that could be problematic if something looks it up without said space Mar 27 21:10:22 good point Mar 27 21:11:44 yep Mar 27 21:11:46 thanks Mar 27 21:11:52 :) Mar 27 21:25:19 can I remove "x11" from the distro features in local.conf ? DISTRO_FEATURES := "${@oe_filter_out('x11', '${DISTRO_FEATURES}', d)}" does not work in local.conf Mar 27 21:31:04 if you're on master, DISTRO_FEATURES_remove = "x11" Mar 27 21:38:03 kergoth: thanks, precisely what I need Mar 27 21:38:05 nice new feature Mar 27 21:38:55 hrmph. still an 8 MB kernel with poky-tiny Mar 27 21:57:36 Hodapp: have you seen Darren's presentation about how he trimmed the configuration? you can go further but you start to have to remove pretty important stuff, like networking... Mar 28 00:23:55 shell is tab indent, right Mar 28 00:26:31 personally i say screw that weird hybrid convention and just use 4 spaces everywhere, but i think its supposed to be tabs in shell at the moment Mar 28 00:26:42 huh, http://rr-project.org/ looks interesting Mar 28 00:32:58 :) Mar 28 00:33:46 I propose a cage match Mar 28 00:55:31 I wonder if you can do cross debugging with it Mar 28 00:55:46 kergoth, any idea if you will make ELC/OEDAM? Mar 28 01:06:52 https://www.youtube.com/watch?v=BKorP55Aqvg Mar 28 01:07:03 I'm pretty sure we have all lived that video Mar 28 01:14:06 dang, bluelightning left before I could ask him for what link he was talking about Mar 28 01:14:14 I need to figure out how to configure the kernel, since I can probably leave out large parts of it **** ENDING LOGGING AT Fri Mar 28 02:59:59 2014