**** BEGIN LOGGING AT Mon Jul 23 03:00:01 2018
Jul 23 04:07:58 <yocti> New news from stackoverflow: Yocto mounting initramfs / initrd image to Raspberry Pi <https://stackoverflow.com/questions/50466075/yocto-mounting-initramfs-initrd-image-to-raspberry-pi>
Jul 23 06:08:20 <yocti> New news from stackoverflow: How to make a recipe .bb file in Yocto 2.0 <https://stackoverflow.com/questions/51472331/how-to-make-a-recipe-bb-file-in-yocto-2-0>
Jul 23 07:18:20 <mckoan> good morning
Jul 23 10:30:37 <rburton> khem: so how is wayland building for me/on the ab without your patches?
Jul 23 11:09:14 <yocti> New news from stackoverflow: how to add optee in barebox for imx6ul <https://stackoverflow.com/questions/51477022/how-to-add-optee-in-barebox-for-imx6ul>
Jul 23 13:09:37 <yocti> New news from stackoverflow: How to make a recipe .bb file in Yocto 2.0 [on hold] <https://stackoverflow.com/questions/51472331/how-to-make-a-recipe-bb-file-in-yocto-2-0>
Jul 23 13:39:42 <yocti> New news from stackoverflow: Is Yocto SDK's character encoding affected by anything other than locale settings? <https://stackoverflow.com/questions/51479976/is-yocto-sdks-character-encoding-affected-by-anything-other-than-locale-setting>
Jul 23 14:09:48 <yocti> New news from stackoverflow: How to port grpc helloworld(CPP) example on yocto <https://stackoverflow.com/questions/49975541/how-to-port-grpc-helloworldcpp-example-on-yocto>
Jul 23 14:47:51 <eduardas_m> hello, I am using Yocto Sumo.. for me some executables in /lib/systemd seem to be missing the execution flag: https://pastebin.com/7NWYtXjR
Jul 23 14:48:37 <eduardas_m> something like a missing execution flag on systemd-sysctl causes a systemd unit failure
Jul 23 14:49:28 <eduardas_m> on Kubuntu 16.04 for example, every file in /lib/systemd is actually marked as executable, so I assume something wrong is going on
Jul 23 15:38:10 <khem> rburton: are you removing x11 from DISTRO_FEATURES ?
Jul 23 15:40:29 <kergoth> hrmph, packagegroup-qt5-toolchain-target & the others in meta-qt5 reference packages that don't even exist anymore
Jul 23 15:43:11 <rburton> khem: oh, is that the missing piece
Jul 23 15:44:30 <rburton> we should have that combo ye
Jul 23 15:48:46 <rburton> khem: http://git.yoctoproject.org/cgit/cgit.cgi/yocto-autobuilder/tree/buildset-config.controller/nightly-no-x11.conf
Jul 23 17:36:20 <khem> rburton: yes so you wont see the problem since x11 is removed there
Jul 23 17:37:35 <khem> kergoth: I have sent a patch and pull both see https://github.com/meta-qt5/meta-qt5/pull/83
Jul 23 17:38:25 <khem> rburton: that reminds me, that for whatever testing we do with systemd, we need to insert a musl variant there too
Jul 23 17:38:40 <khem> infact we should build more "musl"
Jul 23 17:51:30 <rburton> khem: ok but our normal world builds have x11+wayland
Jul 23 17:51:48 <rburton> but yes we only test musl and systemd in isolation
Jul 23 17:51:58 <rburton> maybe a musl+security+systemd combo would exercise a lot more paths
Jul 23 18:01:59 <khem> rburton: yeah that will be killer app
Jul 23 18:02:21 <khem> for next releases probably security should become default
Jul 23 18:02:31 <khem> there is no need to have separate profile
Jul 23 18:03:05 <khem> all distros have switched their defaults to PIE/sec-flags
Jul 23 18:03:33 <khem> unlike in 2012 when it made sense to keep it as a opt-in
Jul 23 18:05:12 <rburton> khem: send a patch to make it in default-distro
Jul 23 18:19:01 <JPEW> rburton: Did you see my comment on pylibfdt in u-boot (http://lists.openembedded.org/pipermail/openembedded-core/2018-July/153128.html)? I'm not inclined to think that pythonnative.bbclass is wrong, but I can't see how it would work for u-boot.
Jul 23 18:19:22 <JPEW> Argh, bad URL, sorry
Jul 23 18:38:23 <yates> which packages provides the connection manager (connman) for the rootfs?
Jul 23 18:39:27 <yates> it must be squirreled under a group of of stuff since grep PROVIDES.*connman was not in any .bb
Jul 23 18:55:04 <rburton> yates: remember that grepping for PROVIDES is mostly useless as the default PROVIDES is ${PN}
Jul 23 18:55:11 <rburton> the connman recipe is, cunning, called connman
Jul 23 18:55:53 <rburton> if you want to be all web 2.0, http://layers.openembedded.org/layerindex/branch/master/recipes/?q=connman
Jul 23 18:59:59 <khem> rburton: I will change poky.conf and send a patch
Jul 23 19:00:27 <khem> or may be include it in default distro config itself
Jul 23 19:19:01 <rburton> khem: so how do i replicate your wayland failures with poky then?  standard poky is wayland+x11, and we test wayland-without-x11 too
Jul 23 19:19:16 <rburton> not saying they're wrong, just curious why it works now
Jul 23 19:21:54 <zeddii_home> aehs29:  I don’t kow if you noticed, but I pushed a lot of my pending -stable updates to the linux-yocto trees. They just need SRCREV bumps on the various -stable branches of oe-core/yocto.
Jul 23 19:38:04 <ente> hi, I'm compiling Qt with PACKAGECONFIG_append = " freetype fontconfig" but it doesn't link the resulting libraries against fontconfig. Any ideas?
Jul 23 19:39:55 <ente> I tried PACKAGECONFIG_append = " freetype fontconfig doesnotexist" just to be sure, and sure enough the other two didn't print a warning
Jul 23 19:46:41 <khem> rburton: my disto is poky + security flags + pkgconf instead of pkg-config + (opengl+x11+wayland+ptest)
Jul 23 19:47:51 <khem> rburton: see https://github.com/cbrake/meta-bec/blob/master/conf/distro/bec.conf
Jul 23 19:49:38 <khem> rburton: does it work with plain poky + pkgconf ?
Jul 23 20:02:24 <rburton> aah i bet its pkgconf
Jul 23 20:02:32 <rburton> its always been a bit special as to how it works
Jul 23 20:02:45 <rburton> fedora started it iirc, did they switch yet?
Jul 23 20:03:28 <rburton> okay tin hat alert, pkgconf moved away from github because MS bought them
Jul 23 20:03:55 <rburton> moved to a host which isn't responding, great work
Jul 23 20:04:12 <rburton> conclusion:pkgconf is still dumb
Jul 23 20:05:48 <rburton> ok fedora switched but they also have patches
Jul 23 20:06:05 <rburton> khem: i'd be interested if anything in https://src.fedoraproject.org/rpms/pkgconf/tree/master was relevant
Jul 23 20:10:35 <khem> rburton: archlinux switch to it too
Jul 23 20:14:05 <khem> rburton: I think these options            --with-system-includedir=%{_includedir} \
Jul 23 20:14:07 <khem>            --with-system-libdir=%{_libdir}
Jul 23 20:14:18 <khem> along with the patch https://src.fedoraproject.org/rpms/pkgconf/blob/master/f/0001-environment-variables-for-system-paths.patch
Jul 23 20:14:25 <khem> are interesting.
Jul 23 20:15:04 <khem> although I like the fact that it exposes some cross compile badness, that otherwise pkg-config is adding internal goop to fix
Jul 23 20:15:51 <khem> I saw similar problem with chromium, which I fixed, and later realised that the patch would have been needed for musl support anyway
Jul 23 20:17:49 <khem> and yes fedora switch to it as well since f26 see https://fedoraproject.org/wiki/Releases/26/ChangeSet?rd=Releases/26#pkgconf_as_system_pkg-config_implementation
Jul 23 20:30:05 <emg> It seems that building dbus-native relies on the machine's libdbus instead of building a libdbus. Here is the error I get: http://ix.io/1icP
Jul 23 20:31:06 <emg> my libdbus is 1.12.8 so it fails
Jul 23 20:31:44 <rburton> dbus-native *is* libdbus
Jul 23 20:32:05 <emg> ok, then I'm confused, the error there shows that it's looking in my /usr/lib instead of in the build directory
Jul 23 20:32:09 <rburton> its being a pain and linking to the host anyway
Jul 23 20:32:15 <rburton> file a bug and all the logs please
Jul 23 20:32:23 <emg> ok, will do
Jul 23 20:32:43 <emg> in the meantime, any tips for debugging/fixing this so I can fix the bug?
Jul 23 20:35:36 <rburton> just figure out why its linking to the host
Jul 23 20:35:52 <emg> ok, time to dig
Jul 23 21:11:05 <yocti> New news from stackoverflow: bitbake qemu_2.7.0 failed do_compile <https://stackoverflow.com/questions/51487160/bitbake-qemu-2-7-0-failed-do-compile> || Replace busybox ntpd with ntp from meta-openembedded <https://stackoverflow.com/questions/51486985/replace-busybox-ntpd-with-ntp-from-meta-openembedded>
Jul 23 22:29:00 <aehs29> zeddii_home: alright, thanks bruce
**** ENDING LOGGING AT Tue Jul 24 03:00:00 2018