**** BEGIN LOGGING AT Mon Jan 20 02:59:59 2014 Jan 20 06:34:26 03214021369 Jan 20 06:34:48 sorry, wrong window Jan 20 09:12:00 morning all Jan 20 09:14:17 hi bluelightning Jan 20 09:31:38 hi woglinde Jan 20 10:03:24 good morning Jan 20 10:49:33 is there an EFI feature in OE-core? Jan 20 10:49:46 MACHINE or IMAGE or both? Jan 20 10:52:15 koen: not that I'm aware of Jan 20 10:53:26 hi can someone confirm, when building native-sdk under dora, that the enviroment-setupfiles are not inculded? Jan 20 10:54:02 woglinde: building what exactly? "native-sdk" isn't a target Jan 20 10:54:41 bl is it a task? Jan 20 10:54:56 woglinde: what are you building? Jan 20 10:55:05 custom image Jan 20 10:55:23 with -c populate_sdk ? Jan 20 10:55:25 and I found serval places where the enviroment files are touched Jan 20 10:55:29 no Jan 20 10:55:30 mom Jan 20 10:58:07 ah strange this time it was included Jan 20 10:59:01 okay maybee it comes from running -c populate_sdk before Jan 20 10:59:20 it had some strange dirs like /opt/..../sysroots/opt dirs too Jan 20 10:59:34 but after clean rebuilds its gone Jan 20 10:59:35 good Jan 20 11:02:09 woglinde: I don't know what to make of that... Jan 20 11:12:59 bluelightning just forget it ;) Jan 20 14:18:44 so, we're on the last day of the bug sprint Jan 20 14:18:54 anyone need any help with working on bugs? Jan 20 14:20:27 I've tried looking through th ejanitor bugs Jan 20 14:20:33 but nothing super easy there Jan 20 14:20:45 except maybe clsiong the boost fail on Arch Jan 20 14:21:09 anything in state "new" is fair game as well Jan 20 14:21:10 based on jackmitchell claiming it worked on his Arch system Jan 20 14:27:26 http://tinyurl.com/ptvu6b8 Jan 20 14:28:01 some of those might be being worked on, but it's likely most aren't Jan 20 14:41:20 morning all Jan 20 14:42:19 hi pb_ Jan 20 14:45:43 hi bluelightning Jan 20 14:58:10 RP, JaMa: do you see any correlation between the bitbake hangs and the number of bitbake threads? Jan 20 15:02:27 koen: no Jan 20 15:02:41 koen: I run with it set to 48 most of the time FWIW Jan 20 15:02:55 8 is the max here Jan 20 15:04:11 koen: I'd expect if it was related to the number I'd have seen the issue more by now. Jan 20 15:04:45 My hardware isn't 48 core in case anyone wonders, I just like to stress out the parallel races :) Jan 20 15:05:27 above 8 IO becomes the bottleneck here Jan 20 15:23:31 koen: I don't see hangs anymore (except when aborting job on jenkins), I'm also using 8 max, usually even less Jan 20 15:26:23 what I see now is AUTOINC numbers not getting incremented (or correctly replaced) on one of my builders :/ Jan 20 15:26:51 -rw-r--r-- 1 bitbake bitbake 101 Jan 20 11:04 tmp-eglibc/deploy/images/om-gta04/uImage-5-3.12.0+gitr0+a4aa4cd999-r16.1-om-gta04-20140119235254.bin.md5 Jan 20 15:26:54 -rw-r--r-- 2 bitbake bitbake 3.2M Jan 20 15:03 tmp-eglibc/deploy/images/om-gta04/uImage-5-3.12.0+gitr0+a4aa4cd999-r16.1-om-gta04-20140120132947.bin Jan 20 15:26:57 -rw-r--r-- 1 bitbake bitbake 101 Dec 20 00:43 tmp-eglibc/deploy/images/om-gta04/uImage-5-3.12.0+gitr0+c44404cbeb-r16.1-om-gta04-20131201054336.bin.md5 Jan 20 15:27:47 sqlite3 prserv.sqlite3 "select * from PRMAIN_nohist where version like 'A%linux-gta04%';" Jan 20 15:27:50 AUTOINC-linux-gta04-5_3.4.9+gitr|om_gta04|AUTOINC+5c4603590af40cccce1d75396d4cb3fdffe0fdef|9 Jan 20 15:27:53 AUTOINC-linux-gta04-5_3.12.0+gitr|om_gta04|AUTOINC+a4aa4cd999|0 Jan 20 15:28:13 not sure where "AUTOINC-linux-gta04-5_3.12.0+gitr" for c44404cbeb got lost :/ Jan 20 15:31:19 hmm maybe it's some mismatch between PRserv server using bitbake with SRCREV shortening and builder using something older with full SRCREV Jan 20 15:36:11 JaMa: I'm doing most of my work on 'dora' where linux-yocto will always rebuild Jan 20 15:36:35 need to hunt down the fix for that and shepherd it into the releasebranch Jan 20 15:40:23 koen: I've sent pull-request for dora with that a while ago Jan 20 15:40:46 http://lists.openembedded.org/pipermail/openembedded-core/2014-January/088074.html Jan 20 15:50:02 JaMa: I guess I need to reply to rhat :) Jan 20 17:22:34 koen: do you know if systemd.bbclass properly works with systemduserunitdir (instead of systemdsystemunitdir)? Jan 20 17:22:49 koen: e18 is using systemduserunitdir from pkg-config for e18.service Jan 20 17:23:26 I'm tempted to add sed for systemdsystemunitdir as e-wm installed in image is more like "system" :) Jan 20 17:27:48 JaMa: haven't tried Jan 20 17:27:59 ok, sed added :) Jan 20 17:28:24 makes it easier to use ${systemd_unitdir} in FILES variable Jan 20 17:28:35 right Jan 20 17:28:51 is anyone else still using meta-efl? Jan 20 17:29:02 it did surprise me that ${systemd_unitdir} is /lib/systemd and not lib/systemd/system Jan 20 17:29:24 I'm still building e17 from it Jan 20 17:29:55 have you seen RFC about efl-1.8? Jan 20 17:30:14 I'll send few more tweaks today, but it's big change so more testing would be nice Jan 20 17:30:41 I must have missed that rfc Jan 20 19:58:01 extra breakage on master lately? Jan 20 19:58:26 several odd/trivial failures after updating yesterday Jan 20 20:00:19 what failed update-alternatives has got some changes Jan 20 20:00:24 that can break stuff Jan 20 20:02:07 have to wait until i get home and do a git status to see my changes Jan 20 20:02:36 was trying to debug why my rootfs is all jacked up now Jan 20 20:03:17 missing proc/run/sbin/lib and it looks like all the stuff that should be under /lib is dumped in / Jan 20 20:03:29 weird... Jan 20 20:03:48 the only rootfs that worked was the rpi-basic-image Jan 20 20:41:18 oh, apmd was one problem Jan 20 20:41:57 and one of the recipes (openssh maybe?) had some error output stuffed in the middle of the recipe Jan 20 20:42:28 i'd suspect a botched merge/commit somewhere... Jan 20 23:35:19 okay, i have a classic build that seems to be confused between arm-none and armv7a-none Jan 20 23:36:48 it's TI/arago/code sourcery toolchain, the latter of which is prefix "arm-none-linux-gnueabi" Jan 20 23:37:20 the only place i see armv7a-none-linux-gnueabi is in the TI paths in openmax, etc Jan 20 23:37:46 is there a way to set this to some value and make it stick? **** ENDING LOGGING AT Tue Jan 21 02:59:59 2014