**** BEGIN LOGGING AT Mon Aug 28 03:00:01 2017
Aug 28 03:38:49 <bshah> Tofe: no, I don't even know what dsme is.. well I know it's something-mer but, what it does. nah
Aug 28 05:41:40 <saidinesh5> iirc dsme was something power related
Aug 28 05:41:51 <saidinesh5> similar to mce but for power iirc
Aug 28 05:43:00 <saidinesh5> Description: Device State Management Entity
Aug 28 05:43:00 <saidinesh5>  This package contains the Device State Management Entity which provides state management features such as service monitoring, process watchdog and inactivity tracking.
Aug 28 05:43:03 <saidinesh5> Ahh..
Aug 28 06:50:44 <Tofe> Morning!
Aug 28 06:51:11 <Tofe> Herrie: could it be related to the various options you activated in your last commit ?
Aug 28 07:11:59 <Herrie|Pre3> Tofe: testing that now :) Seems my cleaning operation helped BTW :P
Aug 28 07:12:58 <Tofe> Ah, good!
Aug 28 07:15:47 <Tofe> bshah: in that case migrating to Qt5.9 should be pretty smooth for you, apparently dsme is the main issue locusf has got
Aug 28 07:33:18 <bshah> Tofe: I am not hopeful as you :P ... given the list of the bugs we have with QtWayland 5.9 and kwin
Aug 28 07:33:21 <bshah> mmhe
Aug 28 07:40:13 <Tofe> ah yes, your compositor is quite a bit more complex than ours, it's true
Aug 28 07:51:00 <Herrie|Pre3> Tofe: The issue with callback that's there still will cause stuff like relaunch, launchparams, app menu not to work?
Aug 28 08:01:38 <Tofe> Herrie|Pre3: I think it's more about the additionalFeature stuff that is causing this
Aug 28 08:01:51 <Tofe> when an Enyo window open with "height=xxx" for instance
Aug 28 08:04:08 <Herrie|Pre3> Tofe: OK just to understand where I can expect issues when testing and what to look forr.
Aug 28 09:57:31 <JaMa> Herrie: FYI: I've removed BuildFlow plugin from our jenkins because of security issues, it's replaced with Pipeline-Workflow
Aug 28 09:57:55 <JaMa> Herrie: I've updated my jobs which were using BuildFlow, but it looks like the LuneOS BuildFlows were disabled anyway
Aug 28 09:58:09 <JaMa> novaldex|away: ^^
Aug 28 10:14:13 <Herrie|Pre3> JaMa: Doesn't ring a bell for me, what does it do in short?
Aug 28 10:14:53 <Herrie|Pre3> Tofe: WebEngine build finished OK with proprietary codecs & pepper-plugins. So now testing by adding other 3 one by one.
Aug 28 10:23:49 <Herrie|Pre3> JaMa: Any suggestions on how to debug qemu when ssh is not working?
Aug 28 10:30:06 <JaMa> Herrie|Pre3: e.g. luneos-testing_build-supported-machines job, which just lists which other jobs to start and in which order
Aug 28 10:30:54 <Herrie|Pre3> JaMa: OK.
Aug 28 10:31:02 <JaMa> ssh stopped working with Pyro upgrade, right?
Aug 28 10:31:09 <Herrie|Pre3> JaMa: Yeah seems so
Aug 28 10:32:03 <Herrie|Pre3> Long time ago it worked fine without me fiddling at all. Then around Krogoth I needed to explicitly set Bash for it to work. Now it seems broken completely.
Aug 28 10:32:07 <JaMa> I'll check if there are some packages missing since the upgrade
Aug 28 10:32:18 <JaMa> /bin/sh to point to bash?
Aug 28 10:32:56 <JaMa> fwiw with Rocko we should really try to find a way to use newer mesa
Aug 28 10:33:23 <JaMa> we're still using x years old mesa, because of fbdev and how it's used in our QPA
Aug 28 10:33:49 <JaMa> but I don't want to keep this old mesa buildable with updated oe-core forever, so I've dropped it in master branch
Aug 28 10:34:19 <JaMa> https://github.com/webOS-ports/meta-webos-ports/commit/7719acd8b4a08c2e94e1e5cf2ed3cc6f65a79315
Aug 28 10:35:51 <Herrie|Pre3> JaMa: Yeah in my WinSCP needed to set Shell to /bin/bash while before "Default" worked
Aug 28 10:36:01 <Herrie|Pre3> But with pyro that broke as well :P
Aug 28 10:36:58 <JaMa> ok, I've kicked the build without webappmanager, will check if I can get to console first
Aug 28 10:37:42 <Herrie|Pre3> JaMa: Tofe fixed webappmgr in general
Aug 28 10:37:56 <Herrie|Pre3> But for me it's stuck on "Booting kernel"
Aug 28 10:38:00 <Herrie|Pre3> Could be some systemd issue
Aug 28 10:38:48 <JaMa> I've seen some commits in herrie/qt5 and toffee/herrie/qt5, but 1) i don't want to rebuild qtwebengine now 2) I would prefer to use some cleaned up version of your many patches :)
Aug 28 10:39:28 <Herrie|Pre3> JaMa: We'll do that once we're happy with things
Aug 28 10:39:31 <Herrie|Pre3> This is WIP
Aug 28 10:39:48 <JaMa> and FWIW there is still issue with qtwebengine and qemux86 build, so I've merged the 5.9.2 to master-next but I'm not going to merge it before fixing it
Aug 28 10:40:03 <Herrie|Pre3> JaMa: We fixed ours
Aug 28 10:40:08 <Herrie|Pre3> qemux86 builds for me
Aug 28 10:40:10 <JaMa> http://lists.openembedded.org/pipermail/openembedded-devel/2017-August/114413.html http://lists.openembedded.org/pipermail/openembedded-devel/2017-August/114414.html
Aug 28 10:40:15 <Herrie|Pre3> Turned out it was patches path
Aug 28 10:40:37 <JaMa> this is without meta-webos-ports layers
Aug 28 10:41:00 <JaMa> and it's the same issue I've reported to Samuli on the first 5.9.0 upgrade
Aug 28 10:48:37 <Herrie|Pre3> JaMa: The bits/c++config.h missing error?
Aug 28 10:49:51 <JaMa> http://lists.openembedded.org/pipermail/openembedded-devel/2017-August/114417.html
Aug 28 10:49:58 <JaMa> yes, that's the missing multilib issue
Aug 28 10:50:22 <JaMa> or incorrectly passing -m32 in host build depending on which explanation is actually correct
Aug 28 10:51:15 <JaMa> maybe we need 32bit build of native mksnaphost to create correct v8 snapshot for 32bit arm
Aug 28 10:52:12 <Herrie|Pre3> Not sure why we don't get this on our builders?
Aug 28 10:52:54 <Herrie|Pre3> I mean locally
Aug 28 10:55:14 <JaMa> I've multilib installed in my gentoo
Aug 28 10:55:20 <JaMa> you probably have it as well
Aug 28 10:55:29 <JaMa> and the jenkins slaves don't
Aug 28 10:56:56 <Herrie|Pre3> JaMa: Well I'm running Ubuntu 16.04 x64 on my local machine and it seems to work fine
Aug 28 10:58:28 <JaMa> do you have g++-multilib installed?
Aug 28 10:58:56 <JaMa> on the slaves we have gcc-multilib, but g++-multilib adds few more packages:
Aug 28 10:58:59 <JaMa> The following NEW packages will be installed:
Aug 28 10:59:02 <JaMa>   g++-5-multilib g++-multilib lib32stdc++-5-dev libx32stdc++-5-dev
Aug 28 10:59:11 <JaMa> and one of them probably provides this bits/c++config.h
Aug 28 11:00:28 <JaMa> https://codereview.qt-project.org/#/c/188558/
Aug 28 11:08:46 <Herrie|Pre3> JaMa: Not sure never installed it myself, but maybe it was included when I setup the environment using the instructions on Wiki.
Aug 28 11:10:16 <JaMa> yes, it's quite common
Aug 28 11:39:45 <Tofe> Herrie|Pre3: your UI doesn't boot on N5 ?
Aug 28 11:41:22 <Herrie|Pre3> Tofe: No I didn't try yet :P
Aug 28 11:41:27 <Herrie|Pre3> UI is on qemu
Aug 28 11:41:37 <Herrie|Pre3> That doesn't boot at all
Aug 28 11:41:46 <Herrie|Pre3> But not sure that's qt 5.9
Aug 28 11:41:52 <Tofe> ok, and that's worse than before... :/
Aug 28 11:42:00 <Herrie|Pre3> Since we couldn't ssh in before with pyro either
Aug 28 11:42:08 <Herrie|Pre3> So cannot check log
Aug 28 11:42:20 <Herrie|Pre3> Tofe: Likely something minor but hard to say without log
Aug 28 11:42:29 <Herrie|Pre3> Could be my flaky build as well
Aug 28 11:43:28 <Tofe> can we mount the virtual disk image ?
Aug 28 11:44:23 <Tofe> that way we could create the /var/log/journal folder, which would make the journal persistent, and display it on our host machine
Aug 28 11:56:01 <Herrie|Pre3> Tofe: I guess we could also just put it in the recipe permanently? So we can always get it from here?
Aug 28 11:56:09 <Herrie|Pre3> Whenever we have an issue
Aug 28 11:56:15 <Herrie|Pre3> At least on dev images
Aug 28 12:13:50 <Tofe> Herrie|Pre3: ok for dev images, but then make sure it doesn't grow over a reasonable size :)
Aug 28 12:26:17 <Herrie|Pre3> Tofe: That will be trickier :P
Aug 28 12:43:28 <Tofe> Herrie|Pre3: it should already be the case, in journald.conf the default is 50Mb I think
Aug 28 12:45:11 <Herrie|Pre3> Tofe: OK. Would this make sense for all targets or just qemux86?
Aug 28 12:48:48 <JaMa> Herrie|Pre3: I've squashed all your commits in herrie/qt5 and added it to my jansa/master branch now I'm rebuilding qemux86 with that
Aug 28 12:49:09 <JaMa> Herrie|Pre3: there were some changes from upstream, so it was rebuilding big chunk of image anyway
Aug 28 12:51:01 <Herrie|Pre3> JaMa: OK
Aug 28 12:55:22 <Herrie|Pre3> JaMa: I was having some compile error probably caused due to some of my EXTRA_QMAKEVARS_CONFIGURE. I have confirmed that -proprietary-codecs -pepper-plugins and -printing-and-pdf work fine. Now testing -spellchecker and still need to do -webrtc afterwards
Aug 28 12:55:22 <Herrie|Pre3> .
Aug 28 13:02:11 <Tofe> Herrie|Pre3: just for qemux86 for now
Aug 28 14:18:05 <Herrie|Pre3> Tofe: What kind of permissions should be there?
Aug 28 14:18:43 <Herrie|Pre3> In base-files we have a recipe that creates some webOS folders with 700 and 755 rights
Aug 28 14:35:42 <Herrie|Pre3> Ah we have some 777 as well it seems :P
Aug 28 14:42:50 <Tofe> journald should be 755 I guess
Aug 28 15:06:53 <Herrie|Pre3> Tofe: OK made an untested commit in herrie/qt59 for now that might work :P
Aug 28 17:21:08 <Tofe> Might work :)
Aug 28 17:21:21 <Tofe> On my side I'm trying to debug this crash on additionalFeature
Aug 28 17:24:25 <Herrie|Pre3> Tofe: OK
Aug 28 17:25:08 <Herrie|Pre3> Tofe: Just arriving home
Aug 28 17:52:55 <Herrie> Seems last one also build fine
Aug 28 17:52:58 <Herrie> Must have been local glitch
Aug 28 23:25:50 <JaMa> Herrie: I've disabled -webrtc as well in last build (which was successful), but spellchecker seems to be enabled by default (even without -spellchecker in EXTRA_QMAKEVARS_CONFIGURE)
Aug 28 23:26:07 <JaMa> at least log.do_configure shows:
Aug 28 23:26:08 <JaMa>   Spellchecker ........................... yes
Aug 28 23:26:08 <JaMa>   WebRTC ................................. no
**** ENDING LOGGING AT Tue Aug 29 03:00:00 2017