**** BEGIN LOGGING AT Thu Oct 27 03:00:00 2016 Oct 27 08:02:21 I've modified a shell script (a part of openssl recipe) used at the time of rootfs creation only. Now bitbake rebuilds everything including kernel. Is there a way to skip full rebuild and to do image only? Oct 27 11:50:39 after giving command of "bitbake rpd-console-image" Oct 27 11:51:06 i am getting following error Oct 27 11:51:08 WatchManagerError: add_watch: cannot watch /home/kalyan/OE_yocto/new/build-rpb-wayland/conf/../../layers/openembedded-core/meta/recipes-graphics/matchbox-session WD=-1, Errno=No space left on device (ENOSPC) Oct 27 16:27:33 hello, I am the maintainer of the documentation of a project in the openembedded-core layer. I don't know OE very well; my question is: are there official OE releases or is it a rolling distribution? Oct 27 16:28:20 are the 2011-1, 2015-4, and 2016-4 tags releases? Oct 27 16:30:42 The official releases are in sync with Yocto Project: https://wiki.yoctoproject.org/wiki/Releases Oct 27 16:31:35 oh wow, thanks for this Oct 27 16:34:32 and what's this "poky" thing? is it a version? Oct 27 16:35:10 I see it's a "reference distribution" Oct 27 16:35:28 so, if this is a distribution, then what is Yocto? is it just a naked set of OE layers? Oct 27 16:36:39 yocto is an umbrella project which includes various projects, one of which is a build system consisting of bitbake, oe-core, and the poky distribution. Oct 27 16:37:00 the yocto project consists of a number of companies collaborating on a common base system for their work Oct 27 16:37:44 ok I see Oct 27 16:37:50 thank you Oct 27 18:35:45 kergoth: http://lists.openembedded.org/pipermail/bitbake-devel/2016-June/007590.html Oct 27 18:37:01 kergoth: this is an issue where bitbake figures that there is no usable sstate for the given component so it rightly goes and rebuilds the component. This is fine, problem is that the return code from bitbake is still -1 Oct 27 18:37:16 so that tricks the autobuilders to think that build has failed Oct 27 19:04:15 khem: i haven't seen an sstate fetch failure in some time, it shouldn't occur, as far as i can tell, unless the remote removed it from the mirror between the check and the download Oct 27 19:04:54 hmm Oct 27 19:13:44 bitbake-layers add-layer let's me add a layer with unresolved deps, but after that I can't add more layers because of unresolved deps... Oct 27 19:14:09 yes, the add-layer doesn't actually process dependencies Oct 27 19:15:08 mattsm: i have a local patch series to let bitbake-layers work without parsing bitbake.conf/etc. it doesn't make add-layer handle deps, but it does mean you can't get yourself into a situation you can't get back out of with bitbake-layers Oct 27 19:15:12 which helps scripting Oct 27 19:15:20 haven't submitted it yet, was waiting until more patches are hitting master Oct 27 19:16:05 kergoth: good deal, just wanted to see if anyone else knew about the issue Oct 27 19:16:31 mattsm, others know about it for sure -- but I don't know if there is a bug on it.. this is something worthy of a bug IMHO Oct 27 19:17:00 agreed Oct 27 19:17:34 ideally need to rework a bit of bitbake to make it easier to just parse bblayers+layer.confs Oct 27 19:19:53 mattsm: https://github.com/openembedded/bitbake/compare/master...kergoth:bitbake-layers-only-parse-bblayers Oct 27 19:21:01 kergoth: thanks Oct 27 23:13:53 kergoth: its when the sstate cache management script it run on server Oct 27 23:14:13 and server is exporting the cache via nfs Oct 27 23:14:34 kergoth: then some artifact may be gone Oct 27 23:15:02 so it ends with Exception: IOError: [Errno 116] Stale file handle Oct 27 23:15:56 kergoth: here is the stack trace https://gist.github.com/kraj/b013026bd3d13d34e43c957125d7a5fb Oct 27 23:15:56 ah, indeed. might have to add the error suppression hack i used to use Oct 27 23:16:51 pointer ? Oct 27 23:21:34 mattsm: I meant this PROVIDES is only defined in fsl layer its not a global virtual Oct 27 23:39:55 khem: i saw that, I was wondering why if never made it farther up Oct 27 23:43:05 mattsm: wayland portions are not as widely used yet Oct 27 23:43:22 khem: is moving that up to oe-core overall reasonable then? Oct 27 23:43:51 khem: i'm personally coming up to speed on all this domain as well Oct 27 23:52:05 it is, if we have vendor provided drivers shadowing some OSS portions Oct 27 23:52:12 and it should be more than 1 Oct 27 23:54:16 well fsl already does it, and possibly more ;) Oct 27 23:54:40 khem: i suspect some of the others will be hard to push up... maybe we can pressure some companies there ;) **** ENDING LOGGING AT Fri Oct 28 03:00:00 2016