**** BEGIN LOGGING AT Mon Apr 10 03:00:01 2017 Apr 10 04:40:49 Marex: meta-altera supports nios2 and yesh nios2eb would be fine for BE Apr 10 04:41:14 Marex: for r2 I think a tune file change is enough Apr 10 04:42:02 ant_home: oh no kidding ... I added that support 3-/ Apr 10 04:42:21 khem: ah thanks for letting me know that meta-altera supports nios2, I added that 3-/ Apr 10 04:42:56 khem: well r2 has different instruction encoding, so how can a tune be fine ? that's basically a different arch, they are NOT binary compatible Apr 10 04:52:55 hey guys Apr 10 04:54:14 khem: while I have you here , https://patchwork.openembedded.org/patch/134456/ which patch did fix the gstreamer wayland problem ? Apr 10 04:54:30 khem: I see you mentioned you fixed it, but I don't see how ... sigh Apr 10 04:55:24 nevermind,I see it now Apr 10 06:18:43 hi, I have question regarding dnsmasq. Is there any process that waits for the dbus event DhcpLeaseAdded ? Apr 10 10:28:19 good morning Apr 10 11:33:22 JaMa, so you expect a patch for disabling parallel build for klibc. Ok Apr 10 11:35:07 ant_work: better than random failures :) Apr 10 11:35:38 it did not happen on your last builds but yes, I agree Apr 10 11:36:04 I've asked for a fix upstream Apr 10 11:36:27 will take time :) Apr 10 11:36:53 JaMa, btw, when do you think you'll tag meta-openembedded? Apr 10 11:40:28 ok, I see, 3 weeks Apr 10 11:43:29 you mean branch for pyro? Apr 10 11:43:44 yes Apr 10 11:44:27 I have a couple of patches to send before that Apr 10 15:00:28 some people really need to learn how to configure jenkins better :) Apr 10 17:01:05 hi there, i am trying to learn about yocto-compat-layer. does it matter which value MACHINE is when running the tool? should we be running the script against 'oe-core' only, or should i pull all my dependencies (if any)? what does the test_signatures really do? Apr 10 17:03:03 is there any write up / email that explains a bit more about the tool? (i have read a few recent commits in yocto-docs, but it's not alot) Apr 10 17:13:58 ndec: see yocto-compat-layer-wrapper Apr 10 17:35:50 ndec, I'll be a few days behind you figuring this out :) Apr 10 17:43:35 kergoth: i've seen the wrapper, but that doesn't tell me what to do with MACHINE. Apr 10 17:44:37 I wonder if you build for one machine, then change to the MACHINE under test? Apr 10 17:45:48 gah must get act together Apr 10 17:46:07 it runs the tests, not you. just set machine to something in oe-core and it won't matter Apr 10 17:46:52 with your bsp in the layer stack? Apr 10 17:49:16 ok, i was trying to use it with our internal setup, i am trying with poky git now.. the wrapper kind of implies you use poky (or at least you have layers at same places as poky tree) Apr 10 17:50:40 and kergoth , what does 'test_signature' do? if i understand correctly , it computes all signatures for 'bitbake world' without my layer, then with my layer, and without setting MACHINE to a machine from my BSP layer. Apr 10 17:50:55 ouch, that would be good feedback Apr 10 17:57:00 the concept of most of the tests are simple, and based on yocto compliance Apr 10 17:57:09 bsp layers shouldn't change anything without MACHINE set to a machine provided by the bsp layer Apr 10 17:57:14 same for a distro alyer and DISTRO Apr 10 17:57:19 Crofton|work: so to me, the easiest to run on my layer was 1) git clone 'poky', 2) git clone my layer, run setup oe-init-build-env in poky folder, 3) run yocto-compat-layer-wrapper Apr 10 17:58:38 only found 2 issues.. 1 being missing README (arguably I have README.md.. Apr 10 18:13:15 ndc all this is great feedback Apr 10 18:13:25 It really bugs me when simple taks require you use poky Apr 10 18:26:43 Crofton|work: nothing in there which shouldn't work with a bitbake+oe clone afaik Apr 10 19:12:20 khem: Hi, I submitted the python-systemd last week. You suggested inherit systemd, but whats the benefit of that? I still had to add DEPEND="systemd" to make it work Apr 10 20:18:24 noc0lour: if you look through bbclass you will see it does lot of common things that is required for systemd unit files Apr 10 20:21:13 khem: the point is python-systemd doesn't provide any unit files, it is just binding to libsystemd so people can use it in python-scripts, I'll check again in the bbclass if I missed something, but it seems to me the inherit systemd is not necessary in this recipe Apr 10 20:27:04 so I guess we learnt today that these guys are using OE: https://www.savant.com/ Apr 10 20:30:28 bluelightning: indeed, could have done without the spam though :) Apr 10 20:30:51 kergoth: yeah :) Apr 10 20:31:28 I've not used jenkins myself but I'm assuming there's some option that's easy to turn on by accident that emails *everyone* that committed to a repo on failure Apr 10 20:31:53 seeing as this isn't the first time Apr 10 20:32:22 i'm sure it's great for work that doesn't involve open source repositories with thousands of contributors Apr 10 20:32:24 heh Apr 10 20:33:36 right :D Apr 10 20:40:47 Ugh, I really need to either fix 'bb' or figure out how to integrate it into bitbake, because i keep really missing its functonality :| Apr 10 20:42:22 I probably broke it badly with tinfoil2 I should imagine Apr 10 20:42:33 sorry about that... Apr 10 20:51:28 bluelightning: yeah, no worries, it's to be expected, will just take time to fix each command one by one, haven't gotten around to it Apr 10 22:06:05 noc0lour: yes it seems it is not needed there, however we still need to check if systemd is enabled in distro_features or not Apr 10 22:08:54 when I tried to build it without systemd in distro_features it errored out, I thought that is expected and the user has to manually append systemd to distro_features? Apr 11 00:03:52 could uart not show bootlog ? **** ENDING LOGGING AT Tue Apr 11 03:00:02 2017