**** BEGIN LOGGING AT Sun Apr 24 02:59:56 2022 Apr 24 08:29:50 Morning Apr 24 08:32:13 Morning! Apr 24 08:32:59 pinephone build started here; I cleaned up sstate and downloads and tmp-work, so it'll take a good part of the day, but I'll be away anyway Apr 24 08:37:57 Tofe: well qemux86-64 still funny for me Apr 24 08:38:13 I get FirstUse but still a non visible card next to it Apr 24 08:38:17 And I cannot center FirstUse Apr 24 08:38:22 home/end don't work Apr 24 08:38:58 PPP and Tenderloin images ready here Apr 24 08:43:29 I'm surprised... Could you share your image, so that I test it here ? it should be the same of course, but who knows Apr 24 08:44:44 "snappy-1.1.9-r0 do_fetch: Fetcher failure: Unable to find revision 2b63814b15a2aaae54b7943f0cd935892fae628f in branch master even from upstream" mmh Apr 24 08:45:03 ah, I forgot to pull latest honister Apr 24 08:59:28 I put it in FromTofe folder Apr 24 09:06:23 damn, I get the same ! Apr 24 09:12:54 I won't have time to debug that now :/ Apr 24 09:13:45 but in /usr/lib/qml/WebOSCompositor/qml/CardShell.qml you can revert https://github.com/webOS-ports/luna-next-cardshell/commit/b3821bd23a245f419f21b6c630716fe3f6f35ecd so that you'll have the gesture area back Apr 24 09:14:06 and at least play with other things Apr 24 09:15:10 "MaliitServer[2251]: [] [pmlog] LS_REQ_NAME {"FILE":"transport.c","LINE":1931} Invalid permissions for com.webos.service.ime.settings" ah, maybe the fix is incomplete for imemanager ? Apr 24 09:16:36 looks like the files in /usr/share/luna-service2 for IME settings have disappeared together with those of /usr/share/ls2 Apr 24 09:16:53 have to go, see you later! Apr 24 09:50:22 Tofe: OK will check that lead Apr 24 10:06:21 Tofe: I have these 4: ./usr/share/luna-service2/client-permissions.d/com.webos.service.ime.perm.json Apr 24 10:06:21 ./usr/share/luna-service2/services.d/com.webos.service.ime.role.json Apr 24 10:06:21 ./usr/share/luna-service2/api-permissions.d/com.webos.service.ime.api.json Apr 24 10:06:21 ./usr/share/luna-service2/roles.d/com.webos.service.ime.role.json Apr 24 15:17:32 back Apr 24 15:17:58 Herrie: shoudn't we have also a manifest file ? Apr 24 15:41:59 Tofe: let me check Apr 24 15:42:33 Yeah we should I guess Apr 24 15:42:37 Not sure why that's not there Apr 24 15:43:17 Let me try something in the recipe Apr 24 15:43:32 the manifest is the one I modified inside qemu to get MaliitServer working Apr 24 15:43:56 (I just removed the sections about prv and pub roles) Apr 24 15:51:49 Tofe: Ah OK Apr 24 15:56:14 OK I will update the recipe to deal with it Apr 24 16:10:38 I still can't explain the black window... it shouldn't appear when using the webos-wayland QPA platform... Apr 24 16:13:00 Tofe: OK recipe fixed it seems Apr 24 16:14:07 great Apr 24 16:19:03 Easy enough by the looks of it: https://github.com/webOS-ports/meta-webos-ports/commit/5cfc98fc98c43e6de0905b0a0d1c7da629199bef Apr 24 16:19:13 Don't have legacy files anymore and a correct manifest Apr 24 16:34:42 I'm just wondering, did we need this WEBOS_SYSTEM_BUS_SKIP_DO_TASKS variable in other recipes ? it's strange that we suddenly stumble on this old pub/prv issue Apr 24 16:35:45 ah, yes, it's actually everywhere in other recipes :) Apr 24 16:41:44 Tofe: Yeah Apr 24 16:41:49 Seems LG oversight somehow Apr 24 16:42:02 Not sure why it works OK on OSE (if it does after all) Apr 24 16:42:16 I also see some LS_REQUIRES_SECURITY errors for listApps and getPreferences in logs Apr 24 16:42:25 Not really sure yet from which calls these come yet Apr 24 16:42:26 yes, the default behavior now should be to not generate old prv/pub unless a variable is set Apr 24 16:42:52 Yes indeed Apr 24 16:43:02 Maybe I need to check the bbclass Apr 24 16:43:06 Could be they updated it Apr 24 16:43:10 And I didn't pick it yet Apr 24 16:43:11 It might be that they set WEBOS_SYSTEM_BUS_SKIP_DO_TASKS at a higher level ? is that even possible ? Apr 24 16:54:50 Herrie: https://github.com/webosose/meta-webosose/blob/master/meta-webos/classes/webos_component.bbclass Apr 24 16:55:29 or, others: https://github.com/search?q=org%3Awebosose+WEBOS_SYSTEM_BUS_SKIP_DO_TASKS&type=code Apr 24 19:00:48 Tofe: Yeah I've seen those Apr 24 19:07:38 Seems inherit webos_component should do the same ? Apr 24 19:11:03 Which we don't have yet ;) Apr 24 19:43:57 Tofe: In my new image seems maliit is failing after a successfull start: https://pastebin.ubuntu.com/p/T59qrVdvfj/ Apr 24 19:53:50 CRITICAL: Failed to create wl_display (No such file or directory) Apr 24 19:54:03 mmh that could be that LSM isn't ready yet Apr 24 19:54:50 or that XDG_RUNTIME_DIR isn't set ? but I've set it in maliit.env... Apr 24 19:59:45 Tofe: The weird thing is, all seems OK until I move my mouse Apr 24 20:00:42 I emailed you a link to a video Apr 24 20:00:46 So you see what I mean Apr 24 20:01:08 As soon as I move the mouse cursor into the VBox screen it seems to create the "extra" invisible/black window and move the card Apr 24 20:02:08 Let me flash PPP soon too and try there Apr 24 20:04:03 yes, it's also what I had when not using the webos QPA platform Apr 24 20:04:29 but that disappeared (for my build) when switching qml-runner and compositor on webos-wayland Apr 24 20:04:47 I also have your issue here, with your image Apr 24 20:06:19 Tofe: OK, well PPP image should be done soon Apr 24 20:06:24 So can test there Apr 24 20:13:28 codepoet: We're not there yet, but we're getting close on our migration for the compositor and webstack! Apr 24 20:13:34 Few minor issues to solve still Apr 24 20:21:04 Seems 1 battery was dead but had another one, booting now Apr 24 20:21:40 Hmmmz on PPP it shows similar as the Phone app before in terms of size Apr 24 20:22:01 Small about 1/9 of the screen on bottom left Apr 24 20:22:04 Rest is white Apr 24 20:23:45 Sent you a picture by email Apr 24 20:23:56 I don't have wifi, so cannot easily pull logs for now yet Apr 24 20:23:59 Can try via USB Apr 24 20:24:02 If needed Apr 24 20:24:38 what is the screen size of the PPP ? Apr 24 20:26:15 1440*720 it seems Apr 24 20:26:21 Same as regular PP Apr 24 20:26:21 ah, yes, found it too Apr 24 20:26:50 And I have this: https://github.com/webOS-ports/configd-data/blob/webOS-ports/webOS-OSE/configs/layers/base/pinephonepro/com.webos.surfacemanager.json Apr 24 20:27:05 I'm looking for a relationship between 1920x1080, 720x1440, and the display issue you see on your screen Apr 24 20:27:17 It seems to be OK in terms of size in app menu and how it displays stuff now after this configd-data fix Apr 24 20:27:33 ^ what do you mean ? Apr 24 20:28:01 ah yes, the compositor itself is fine Apr 24 20:28:02 Well before the configd-data fixes it was showing 1920*1080 wrongly Apr 24 20:28:21 Yes, in terms of gridunits or whatever we call them, it seems pretty OK size too Apr 24 20:28:26 Not too small, not too big Apr 24 20:28:38 mmmh could it be an issue with qml-runner ? or WebOSWindow QML component ? or they qtwayland-webos layer ? Apr 24 20:29:12 Weird thing is you said it was working at your end before Apr 24 20:29:17 So it could be a small patch missing somewhere Apr 24 20:29:28 no no, I never tried it on device Apr 24 20:29:44 it's just the qemu spurious window bug Apr 24 20:30:33 actually, the logs could be interesting in the PPP case Apr 24 20:30:43 Well I'm able to replicate on PPP with tapping around a bit Apr 24 20:30:53 oh Apr 24 20:30:58 I now have a "transparent" card and the FirstUse on the left Apr 24 20:31:04 :( ok Apr 24 20:31:04 So similar to Qemu actually Apr 24 20:31:17 Just the size of FirstUse in qemu is "normal" Apr 24 20:31:35 I hope it's just a PR I forgot or something like that Apr 24 20:31:47 my webruntime isn't built yet for PP Apr 24 20:32:01 I had to restart it, it crashes OOM :p Apr 24 20:32:03 Tofe: It usually is when you touch things in 20 different places Apr 24 20:32:05 crashed* Apr 24 20:32:10 I have the same at times Apr 24 20:33:40 Hmmz funny PPP reboots when I plug in USB C Apr 24 20:33:46 Let's see if it boots to UI at least Apr 24 20:33:52 That would be a kernel bug I guess Apr 24 20:34:23 Hmmz weird Apr 24 20:34:29 Let me find my jack dongle instead Apr 24 20:36:23 meh Apr 24 20:36:32 It doesn't want to be debugged Apr 24 20:37:54 Well dongle found Apr 24 20:40:47 This is at least the log for when I plug USB in: https://pastebin.ubuntu.com/p/jyPZXBx45x/ Apr 24 20:41:02 Let me see if Megi did some magic since my last kernel bump Apr 24 20:41:21 I saw some discussions on battery & USB stuffs Apr 24 20:45:57 looks like an obvious bug in the USB stack, but it's probably actually hard to find Apr 24 20:48:30 Tofe: Well there are some commits in 5.18 that might be related; https://github.com/megous/linux/commits/ppp-drivers-5.18 Apr 24 20:48:32 Recent 3 ones Apr 24 20:49:00 And a bunch more there Apr 24 20:49:17 I'll try to build a 5.18 kernel and see Apr 24 20:49:53 right Apr 24 20:49:56 worth a try Apr 24 20:50:36 Updating should be easy enough usually Apr 24 21:00:13 Seems to build at least Apr 24 21:01:55 Will flash in a minute Apr 24 21:21:45 Still there, but seems I might need latest commit ;) https://github.com/megous/linux/commit/fae7a3ca26232a53ddf4d08a555c267b3094816a Apr 24 21:24:36 So rebuilding as last try for tonight Apr 24 21:45:37 OK no more reboot at least **** ENDING LOGGING AT Mon Apr 25 02:59:56 2022