**** BEGIN LOGGING AT Mon Feb 02 02:59:58 2015 Feb 02 08:49:38 shoragan: hey Feb 02 08:49:48 shoragan: survived FOSDEM alrighte? Feb 02 09:12:29 mickeyl, mostly :) Feb 02 09:13:18 shoragan: hehe. I know it can be quite exhausting. Anyways, if you could upload the FSO vm to somewhere for me to grab it I'd be grateful. Feb 02 09:13:46 mickeyl, yes. i'll give you a shell account later this we, if thats ok Feb 02 09:13:51 we have to make a new release anyways, so it's not a problem if the code moves to github Feb 02 09:13:57 shoragan: yeah, that'd be good Feb 02 09:40:30 hi mickeyl Feb 02 09:40:49 hey woglinde! how are things? Feb 02 09:40:58 morning florian Feb 02 09:41:21 good morning Feb 02 09:41:21 mickeyl fine so far Feb 02 09:41:43 mickeyl: ui hi, you here? :-) Feb 02 09:42:47 * florian is back from FOSDEM Feb 02 09:42:53 florian: yeah. changed some things with regards to professional work. will work more from home this year and hopefully also more on non-ios things as well, at least that's the idea Feb 02 09:43:18 florian: ya,unfortunately i had appointments. would have loved to go this year as the program had some interesting things Feb 02 09:43:20 mickeyl: sounds like a good plan! Feb 02 09:43:29 :) Feb 02 09:44:17 i want to use OE again – whenever the beagle board x15 ships Feb 02 09:44:26 (or should i say yocto?) Feb 02 09:47:00 that's pretty much up to you how far you want to go :) Feb 02 09:47:41 I have no idea. I've been pretty much out of the loop since the birth of Lara Marie. I can imagine many things have changed since 2011… Feb 02 09:47:58 i would have no idea where to begin to contribute again, but before that, I have to actually try to use all that stuff Feb 02 09:48:02 * mickeyl forgot everything Feb 02 09:54:58 I guess the getting started wiki is more or less up to date :) Feb 02 09:55:07 excellent :) Feb 02 09:56:26 what's the general tendency wtr. init systems in oe these days? i remember we have talked about alternatives some years ago. has there been any work towards that? or is everyone waiting for debian to move to system? Feb 02 09:56:41 d Feb 02 09:57:38 yes yocto documentation covers a lot of topics Feb 02 09:58:03 mickeyl most user are using systemd now I guess Feb 02 09:58:39 I would not switch back to shell stuff Feb 02 09:58:43 *nod* Feb 02 09:58:59 besides the little helper scripts Feb 02 10:19:52 morning all Feb 02 10:20:16 hi florian, mickeyl, woglinde Feb 02 10:20:26 morning bluelightning Feb 02 10:21:43 yeah, we support both sysvinit and systemd; sysvinit is still the default but a lot of folks are now using systemd and it works pretty well within OE now Feb 02 10:21:56 hi bluelightning Feb 02 10:21:59 cool Feb 02 10:29:01 hi bluelightning Feb 02 15:08:05 morning Feb 02 15:12:19 hi kergoth Feb 02 15:37:24 morning kergoth Feb 02 16:16:36 hi Feb 02 16:16:47 is it possible to rebuild the sysroot gcc from bitbake ? Feb 02 16:17:54 afournier: what are you attempting to do? Feb 02 16:18:57 i am attempting to rebuild gcc, i moved my workdir some days ago without the cache, but gcc remembers /usr/include from the old sysroot directories Feb 02 16:19:56 i could put a symlink... but if i dont want to rely on a symlink i can rebuild everything from scratch or rebuild gcc only maybe Feb 02 16:20:39 what's the best option ? Feb 02 16:22:19 you can't move a tmp directory from an oe/yocto build. it will break, guaranteed. Feb 02 16:23:01 right, just delete your TMPDIR (only) and then the build will restore what it can from the sstate cache Feb 02 16:23:22 okay Feb 02 16:23:28 fair enough Feb 02 16:23:40 honestly I thought we had an error pop up when TMPDIR is moved... Feb 02 16:24:27 honestly i thought i rebuilded everything from scratch already, but i may i forgot to delete some parts... Feb 02 16:24:35 pretty sure we used to, the path was stored in a file in conf or something Feb 02 16:25:01 afournier: there's no reason to avoid wiping tmp, pulling unchanged bits from the sstate cache really doesn't take all that long Feb 02 16:25:04 in general Feb 02 16:25:42 so i can keep build/{downloads,sstate-cache} and cache/* also ? Feb 02 16:26:14 yes, you can leave those alone, just wipe tmp and bitbake again, and anything unchagned will be pulled from sstate-cache Feb 02 16:26:31 bitbake too ? Feb 02 16:26:41 ha no Feb 02 16:27:27 ok thanks Feb 02 17:04:58 bluelightning, kergoth, i still have the same problem, i think the cache should be removed also because gcc hardcode /usr/include path, and this gets cached Feb 02 17:05:35 ok, that would definitely be a bug Feb 02 17:05:47 i am not on current btw Feb 02 17:05:58 which release are you using? Feb 02 17:06:18 f8125a1e9b6893a12355d55d4df584a8d97f0bff Feb 02 17:06:33 (oe-core) Feb 02 17:07:16 so that would be dizzy aka 1.7? Feb 02 17:08:02 if you say so... Feb 02 17:08:14 i was on current one day :) Feb 02 17:08:20 i mean master Feb 02 17:09:42 in any case, bitbake -c cleansstate gcc-cross gcc-cross-initial *may* fix it, I don't know for sure Feb 02 17:10:14 i cleaned everything too late to test :) Feb 02 17:10:49 ok, well it was a yucky workaround in any case Feb 02 17:12:50 afournier: I just tested moving TMPDIR, I immediately got an error... you didn't see any errors telling you you needed to rebuild? Feb 02 17:14:46 yes i got that error, so i deleted tmp-workdir, but even so the problem was in the cache of gcc's own compilation Feb 02 17:16:41 sysroot-prefix-cpp -Wp,-v was telling me /usr/include was in the old sysroot Feb 02 17:17:18 so i was getting errors like stdio.h file not found :) Feb 02 17:17:56 wasn't there a relatively recent fix that went in to remove gcc's fall back to the build time sysroot path when —sysroot= wasn't specified? Feb 02 17:18:03 i think i saw something like that from RP on the lists at some point Feb 02 17:20:06 kergoth: yes, in master we default to an invalid path now **** ENDING LOGGING AT Tue Feb 03 02:59:58 2015