**** BEGIN LOGGING AT Sat Sep 09 03:00:01 2017
Sep 09 05:15:12 <Herrie|2> Morning!
Sep 09 07:42:42 <Tofe> Morning!
Sep 09 07:43:09 <Tofe> Herrie|Laptop: my qemu build is from yesterday, all updated
Sep 09 07:44:33 <Tofe> so once we sort out this udev issue, we'll be all fine. And we even have a workaround for that (setting VIRTUALBOX_TOUCH_EVDEV_MOUSE=/dev/input/event4 in environment.conf).
Sep 09 07:45:46 <Herrie|2> :D
Sep 09 07:45:57 <Herrie|2> Just Mesa would need to be updated at some point right?
Sep 09 07:50:58 <Herrie|2> But that's not urgent for now
Sep 09 07:58:08 <Tofe> right, we'll need to find a way to use directly the upstream Mesa
Sep 09 08:15:01 <Tofe> Herrie|2: looking at how easily qemu helped us on this issue, I begin to think we should just switch to that emulator. However if we do that we lose the mouse integration and it doesn't really solve the egl issue...
Sep 09 08:28:20 <Herrie|2> Tofe: I can confirm I get html5test.com crashing on Hammerhead
Sep 09 08:28:23 <Herrie|2> It did work for me before
Sep 09 08:28:28 <Herrie|2> But that site updates all the time
Sep 09 08:28:46 <Tofe> Herrie|2: this is with the flags change, right?
Sep 09 08:29:00 <Herrie|2> Tofe: Yeah with flags change
Sep 09 08:29:04 <Herrie|2> It could be the netflix stuff
Sep 09 08:29:09 <Herrie|2> Ehm WideVine I mean
Sep 09 08:29:37 <Tofe> could be, but I doubt it though, we don't have any binary blob yet
Sep 09 08:30:12 <Herrie|2> Tofe: Well me neither, but maybe it somehow detects it
Sep 09 08:30:15 <Herrie|2> And then crashes
Sep 09 08:30:21 <Herrie|2> Or they updated html5test.com in meanwhile
Sep 09 08:30:25 <Herrie|2> And that causes crash
Sep 09 08:30:31 <Herrie|2> Site is updated frequently
Sep 09 08:32:14 <Tofe> we could test an image from august 5th, which was Qt 5.8
Sep 09 08:35:13 <Herrie> Tofe: It worked before with 5.9
Sep 09 08:35:17 <Herrie> I tested it
Sep 09 08:35:26 <Tofe> ah ok
Sep 09 08:35:30 <Herrie> Tofe: I noticed N5 doesn't keep my wifi password
Sep 09 08:35:32 <Herrie> YOu have the same
Sep 09 08:35:40 <Tofe> yes
Sep 09 08:35:44 <Tofe> pretty annoying
Sep 09 08:35:47 <Herrie> Yeah
Sep 09 08:35:53 <Herrie> And that also causes the date issue probably
Sep 09 08:35:55 <Tofe> didn't we stumble on that issue before?
Sep 09 08:35:59 <Herrie> It's some mount not working properly
Sep 09 08:36:02 <Herrie> Pretty sure
Sep 09 08:54:21 <Herrie> Tofe: Sep 08 23:07:27 hammerhead kernel[341]: [    1.924765] bcm_wifi_get_mac_addr: Failed to get information from file /persist/wifi/.macaddr (-2)
Sep 09 08:54:58 <Tofe> .macaddr file missing ?
Sep 09 08:55:11 <Tofe> ah, I've definitely seen that before
Sep 09 08:57:02 <Herrie> Tofe: https://github.com/shr-distribution/meta-smartphone/tree/pyro/meta-lg/recipes-core/systemd/systemd-machine-units/mako
Sep 09 08:57:11 <Herrie> And https://github.com/shr-distribution/meta-smartphone/tree/pyro/meta-lg/recipes-core/systemd/systemd-machine-units/hammerhead
Sep 09 08:57:17 <Herrie> Seems we simply miss some files ;)
Sep 09 08:57:50 <Tofe> yup, good find!
Sep 09 08:57:54 <Herrie> Or at least that's my best guess ;)
Sep 09 09:02:00 <Tofe> I'm now testing the fix
Sep 09 09:02:34 <Herrie> I might not be around that much today, leaving for 2 weeks of hopefully better weather tomorrow ;) So need to do some last minute shopping, packing etc :P
Sep 09 09:02:42 <Herrie> And I'm alone with 2 children whole day for first time LOL
Sep 09 09:03:15 <Tofe> Ah, vacation, always a great idea :)
Sep 09 09:03:35 <Herrie> Didn't have a proper one since at least last summer :P
Sep 09 09:05:37 <Herrie> Tofe: I'm getting this for the camera stuff: https://bpaste.net/show/e4dadffd432d
Sep 09 09:05:51 <Herrie> I did a sudo apt-get install bzr and it installed bzr but still doesn't seem to work?
Sep 09 09:09:10 <Tofe> (09:42:22) JaMa: [19:03:25] and the bzr installation isn't enough, we will need http://lists.openembedded.org/pipermail/openembedded-core/2017-September/141976.html applied in oe-core master and pyro
Sep 09 09:09:39 <Tofe> seems like I have it locally in my cache somehow
Sep 09 09:15:05 <Herrie> Ah ok
Sep 09 09:18:02 <Herrie> Let me try that
Sep 09 09:20:29 <Tofe> Herrie: we do the stable release when you'll be back then?
Sep 09 09:20:39 <Tofe> (if I don't break anything in the meantime :p )
Sep 09 09:23:04 <Herrie> Tofe: JaMa's patch seems to work there
Sep 09 09:23:18 <Herrie> Tofe: Well I should have access to internet :P
Sep 09 09:23:24 <Herrie> And will take a TP anyway and my N5
Sep 09 09:23:31 <Herrie> So could do things :P
Sep 09 09:24:05 <Tofe> Don't forget to do vacations too ;)
Sep 09 09:25:00 <Tofe> ok, so, I did have a .macaddr, containing binary stuff, but the access rights were incorrect
Sep 09 09:25:21 <Tofe> --> did "chmod 755 /persist/wifi", it helped
Sep 09 09:25:55 <Tofe> we can keep the .sh, but it's actually a bit incorrect, as it should write it in binary form (0x4c 0x5b 0x10 etc...)
Sep 09 09:29:13 <Tofe> mmh that didn't fix the issue
Sep 09 09:51:08 <Tofe> ok, I got it, we should not have the wifi driver builtin the kernel, but as a delayed-loaded module
Sep 09 09:51:15 <Tofe> ... like N4 :)
Sep 09 09:58:36 <Herrie> Tofe: OK
Sep 09 10:06:15 <Herrie> I guess it was there for a reason after all :P
Sep 09 10:23:01 <Tofe> exactly :)
Sep 09 10:23:38 <Tofe> I've got the module now, it's delayed, but it's loaded as non-powered and connman thinks it is powered
Sep 09 10:23:50 <Tofe> so it works if we do off->on, but that's also annoying :)
Sep 09 10:24:34 <Tofe> I'm playing a bit with the systemd service to load it a bit before android starts, instead of after. It shouldn't need android anyway.
Sep 09 10:35:05 <Tofe> got it, it works now
Sep 09 10:38:14 <Herrie> :D
Sep 09 10:40:18 <Tofe> Herrie: do you know if that kind of use of "COMPATIBLE_MACHINE" will work ? https://bpaste.net/show/aa83ed5ebb05
Sep 09 10:40:37 <Tofe> I wanted to simplify a bit the recipe, as mako and hammerhead are now very alike
Sep 09 11:04:45 <Herrie> Tofe: I've so far only seen examples with architecture... ie x86, mips, armv7 etc
Sep 09 11:05:01 <Herrie> Ah wait
Sep 09 11:05:46 <Herrie> Seems you can do: COMPATIBLE_MACHINE = "tuna|grouper|mako|tenderloin"
Sep 09 11:06:00 <Herrie> That was used by token-generator before
Sep 09 11:54:38 <Herrie> Ok 2 children asleep :P
Sep 09 11:54:47 <Herrie> For as long as it'll last LOL :P
Sep 09 12:05:57 <MartinHov_> all: great to read about the good progress here in the recent days !  Thanks - and looking forward to test on emulator and N4
Sep 09 12:06:20 <MartinHov_> Herrie: have great vacations ..!
Sep 09 12:09:49 <Herrie> MartinHov_: Yeah seems lots of bugs are being squashed :P
Sep 09 12:10:00 <Herrie> Thanks, sure I will, some Adriatic see will do us good :P
Sep 09 12:21:29 <Herrie> s/see/sea
Sep 09 14:59:29 <Tofe> Herrie: the crash comes from the activation of the "disable-es3-gl-context" from within qtwebenginecore
Sep 09 15:13:45 <Herrie> Tofe: OK
Sep 09 15:13:47 <Herrie> Hmmz weird
Sep 09 15:14:00 <Herrie> We had that switch before I think when I tested it
Sep 09 15:14:33 <Tofe> no, not in the browser
Sep 09 15:14:51 <Herrie> Tofe: Ah OK
Sep 09 15:15:03 <Tofe> that's also a reason why I wanted to activate the switches from within: it is now common for all WebEngineView instances
Sep 09 15:15:20 <Herrie> Yeah that makes sense ;)
Sep 09 15:15:24 <Tofe> otherwise we'd have to also add switches to the luna-qml-launcher call
Sep 09 15:15:27 <Herrie> Just silly it crashes
Sep 09 15:15:42 <Tofe> yes, I guess that's a chromium bug somehow
Sep 09 15:32:44 <Tofe> I think it's crashing the on WebGL2 tests
Sep 09 15:33:48 <Tofe> probably because of this kind of code snippet: https://github.com/qt/qtwebengine-chromium/blob/56-based/chromium/gpu/command_buffer/service/texture_manager.cc#L1913
Sep 09 15:51:09 <Tofe> ok I've got a little patch :p
Sep 09 15:59:30 <Tofe> https://github.com/webOS-ports/meta-webos-ports/pull/243 should handle it. I don't reproduce the crash, at least.
Sep 09 16:18:57 <Herrie> Tofe :D
Sep 09 16:31:10 <Herrie> Seems weird how they deal with it in the original code :S
Sep 09 16:38:08 <Tofe> yes, I think it's a bit borderline...
Sep 09 20:43:22 <Herrie> Tofe: Is it correct I didn't see anything for qemu PR-ed yet? Just wondering :P
Sep 09 20:43:26 <Herrie> Or maybe I missed something
Sep 09 20:49:08 <Tofe> Herrie: it's correct
Sep 09 20:50:18 <Herrie> OK :D
Sep 09 20:50:45 <Tofe> I tried to find a nice fix for the mouse, but didn't find anything
Sep 09 20:50:59 <Tofe> I think I'll just fallback to the workaround
Sep 09 21:00:18 <Herrie> Tofe: OK
Sep 09 21:00:28 <Herrie> As long as we know what to fix when it breaks ;)
Sep 09 21:00:58 <Herrie> What was the issue with sshd?
Sep 09 21:01:17 <Tofe> the ethernet virtual hardware
Sep 09 21:01:26 <Herrie> Ah
Sep 09 21:01:29 <Tofe> setting virtio-net works better
Sep 09 21:01:34 <Herrie> :)
Sep 09 21:03:27 <Tofe> https://github.com/webOS-ports/meta-webos-ports/pull/244 should fix all this
Sep 09 21:03:33 <Tofe> I didn't test the ovf though
Sep 09 21:05:22 <Herrie> OK I'll run it here overnight
Sep 09 21:06:27 <Tofe> Ok!
Sep 09 21:06:32 <Tofe> well, time for bed :)
Sep 09 21:06:38 <Tofe> 'Night!
Sep 09 21:19:10 <Herrie> Night!
**** ENDING LOGGING AT Sun Sep 10 03:00:00 2017