**** BEGIN LOGGING AT Sat Sep 16 03:00:01 2017 Sep 16 06:57:19 Morning! Sep 16 07:01:26 Tofe: VirtualBox you tested was still 5.8 based right? Sep 16 07:03:45 Morning! Sep 16 07:03:50 Herrie|Laptop: no, 5.9 Sep 16 07:04:13 Tofe: Hmmz because I don't get graphics on mine Sep 16 07:04:24 And with latest nightly build seems that we have another issue :S Sep 16 07:04:41 NOTE: recipe luneos-dev-image-1.0-r0: task do_bootdirectdisk: Started WARNING: luneos-dev-image-1.0-r0 do_bootdirectdisk: /home/jenkins/workspace/luneos-testing/webos-ports/tmp-glibc/deploy/images/qemux86-64/bzImage doesn't exist Sep 16 07:05:13 Which will result in not having a vmlinuz in the 0.fat file inside the .vmdk Sep 16 07:05:23 Which is probably caused by something JaMa did, just not sure what Sep 16 07:06:19 I don't see anything obvious in our recent changes that would cause this though :S Sep 16 07:06:19 ah, ok, yes we need to look into that Sep 16 07:06:38 Same for qemux86 by the way Sep 16 07:07:28 Just I don't get graphics on 5.9 based.... I get this in log: Sep 15 15:35:43 qemux86 luna-next[485]: FATAL: 15:35:43.008: EGL library doesn't support Emulator extensions Sep 16 07:07:32 Well, we should get qemux86 working first, as qemux86-64 was left non-working for quite a long time iirc Sep 16 07:07:47 Tofe: This is with qemux86 as well ;) Sep 16 07:08:15 This is VirtualBox, right? Sep 16 07:08:35 I'll have to download latest testing too Sep 16 07:08:53 Tofe: Latest testing misses the vmlinuz :P Sep 16 07:09:07 The previous one should be fine, but you'd have to update the .ovf then Sep 16 07:09:24 no pb, I never use it :p Sep 16 07:10:18 Tofe: OK Sep 16 07:10:30 Anyway if you'd do you'd have to update the adapter 0 Sep 16 07:11:28 Didn't I switch the adapter to virtio-net in the ovf ? Sep 16 07:11:42 As per: https://github.com/webOS-ports/meta-webos-ports/commit/70aa1d88a58a1a8661381e70b9ba72885c3d93ad Sep 16 07:11:46 You did half ;) Sep 16 07:11:50 I did the other half ;) Sep 16 07:11:58 See above commit Sep 16 07:12:24 oops :) Sep 16 07:12:47 It's fixed now ;) Sep 16 07:13:06 I don't really understand the ovf values anyway, how they map to what virtualbox is using Sep 16 07:13:12 The problem I get seems to be coming from a file that wasn't there in 5.8 yet, hence I wondered if you tested 5.8 or 5.9 Sep 16 07:13:21 https://github.com/qt/qtbase/blob/e4c39d5e1e7ee8c2bba273e6d613ec519b7fa9c2/src/plugins/platforms/eglfs/deviceintegration/eglfs_emu/qeglfsemulatorintegration.cpp#L86 Sep 16 07:14:54 Tofe: Well I modified mine in UI and checked what changed in .ovf :P Sep 16 07:15:55 Herrie|Laptop: mmh why are we using this platform at all ? it seems to be specific to the Qt emulator Sep 16 07:16:23 Tofe: Not sure Sep 16 07:16:33 Just that's the error it pushes to the log Sep 16 07:16:39 And that's the only reference I can find to it Sep 16 07:16:48 On all of GitHub and most Google searches :P Sep 16 07:17:38 you could try to just remove the corresponding .so in /usr/lib/qt5/plugins/platform (should be eglfs_emu.so or smthg like that) Sep 16 07:18:01 Let me try that Sep 16 07:20:25 Tofe it's not there but I have /usr/lib/qt5/plugins/egldeviceintegrations/libqeglfs-emu-integration.so Sep 16 07:20:58 yep that one Sep 16 07:21:59 When I delete that one I get messed up graphics but at least something Sep 16 07:22:15 ah Sep 16 07:22:31 all this is pretty weird Sep 16 07:22:48 and you're right, maybe my tests were done on a Qt 5.8 image Sep 16 07:23:27 Yeah: http://tinypic.com/r/2qjhma9/9 Sep 16 07:23:33 This is what I get Sep 16 07:23:36 And I've seen this before Sep 16 07:23:39 Not sure what was the fix :P Sep 16 07:23:43 I think JaMa fixed that Sep 16 07:24:54 But I can clearly see that's FirstUse :P Sep 16 07:26:18 mmh how to find that in the logs... :p Sep 16 07:33:16 Well search for virtualbox :P Sep 16 07:33:22 I think it must have been sometime this year Sep 16 07:33:26 Or late last year Sep 16 07:33:32 I'm pretty sure I posted a picture Sep 16 07:37:55 yes, I remember it Sep 16 08:39:40 Tofe: http://logs.nslu2-linux.org/livelogs/webos-ports/webos-ports.20161021.txt Sep 16 10:55:16 Herrie|Laptop: I can boot the build #469 on virtualbox, but removing /etc/modprobe.d/vboxvideo.conf was a mistake, sorry Sep 16 10:55:44 I had to remove /usr/lib/qt5/plugins/egldeviceintegrations/libqeglfs-emu-integration.so too Sep 16 10:57:13 that means we have a little vmlinuz issue between #469 and #470 Sep 16 10:57:45 I'll test #470 right away Sep 16 11:05:00 ok no surprise, same as you Sep 16 11:08:22 Herrie|Laptop: a-piori, it's because of https://github.com/webOS-ports/webos-ports-setup/commit/493fecf4ba443b2c358a9bf04ddb1f7eaabdde21 Sep 16 11:43:14 Herrie|Laptop: I've fixed most issues with mako: for the sound we need to disable the alsa driver, for halium's init the TheMuppets repo's binaries didn't have the "+x" flag Sep 16 11:43:51 We lost orientation sensor, but I think we actually lost that with Qt 5.8 upgrade Sep 16 11:47:53 So, with these two fixes (sound and binaries executable flag), I'll PR the switch to Halium Sep 16 11:58:24 PRs done Sep 16 13:27:15 Tofe: Nice! Sep 16 17:46:12 For the Halium bits, anything I can do to move it to our repos? Sep 16 17:46:21 Tofe: ^ Sep 16 17:46:31 I merged the meta-webos-ports one Sep 16 17:46:43 For meta-smartphone need to wait for JaMa Sep 16 21:08:54 thx Sep 16 21:09:06 For the repos, let me see, there are quite a bunch of them Sep 16 21:11:00 first thing would be to bring the https://github.com/Tofee/android/tree/wop-halium-5.1 branch into our repos Sep 16 21:11:50 then, same for https://github.com/Tofee/android_luneos_hal/tree/halium-5.1 Sep 16 21:12:42 then, same for https://github.com/Tofee/android_device_hp_tenderloin/tree/halium-5.1 Sep 16 21:13:15 ... and https://github.com/Tofee/android_device_hp_tenderloin-common/tree/halium-5.1 Sep 16 21:14:23 .. and https://github.com/Tofee/android_hardware_qcom_display-caf/tree/tenderloin/wop-12.1 Sep 16 21:14:46 for the kernel, we'll simply need to make it point to the current shr repo Sep 16 21:15:14 and I think that's all :p Sep 16 21:15:52 for mako & hammerhead we take directly the code from ubports or from halium, no patch whatsoever Sep 16 21:17:47 After that I'll think about simplifying a bit the way we extract and bind-mount /system; there's one bind-mount too many here **** ENDING LOGGING AT Sun Sep 17 03:00:01 2017