**** BEGIN LOGGING AT Sun Jun 07 02:59:56 2020 Jun 07 08:19:12 Morning! Jun 07 08:31:44 Herrie, JaMa: here is what I'm trying to achieve with PmLogDaemon: https://www.dropbox.com/s/u9ym3984ltzkky7/Syslog_with_journald.svg?dl=0 Do it look like the right thing to you ? I'm especially not sure about binding /run/systemd/journal/syslog to PmLogDaemon, but journald's documentation says both to use that and to not bind to it, so... Jun 07 08:32:20 ^corresponding PR : https://github.com/webOS-ports/pmlogdaemon/pull/1 Jun 07 08:32:40 I'll try to also remove busybox-syslogd from our images Jun 07 09:52:29 Tofe: I'm not too familiar with either to say much useful about it. Jun 07 09:52:42 JaMa or nizovn might have some thougts Jun 07 12:23:37 Herrie: looks like we broke phone app a bit with ACG Jun 07 12:23:43 I'm on it Jun 07 14:24:07 Tofe: Ah I didn't test it that extensively on qemux86 as you can imagine so there might be a few things here and there Jun 07 14:26:17 They should be easy enough to fix though Jun 07 14:26:48 While you're on it you might want to test these implicit declaration in cardshell as well Jun 07 14:27:00 Seeing it spams the logs quite a bit Jun 07 14:30:33 Herrie: I thought you fixed those already? Jun 07 14:30:50 And that it was just my image which was outdated Jun 07 14:30:54 Tofe: Well I proposed some Jun 07 14:31:05 You had comments, then you weren't sure Jun 07 14:31:19 So it's still sitting in a branch somewhere Jun 07 14:31:26 :) ah, yes, it was unfinished business, let me have a second look Jun 07 14:33:14 ah, yes, I see some fixes to do. I'll do that. Jun 07 14:37:40 My WIP should be sitting there somewhere Jun 07 14:42:22 yes, in the herrie/qt5-15 Jun 07 14:59:36 Tofe: A bit under the weather today will have a look at your PR's probably tomorrow Jun 07 15:08:14 https://github.com/webOS-ports/luna-next-cardshell/pull/326 here we go, I didn't bother create another branch... Jun 07 15:12:26 that's better Jun 07 15:12:32 I'll merge it Jun 07 15:13:18 I have some comments on your phone app PR Jun 07 15:13:25 sure Jun 07 15:15:05 Not sure why you still included public & private in appinfo.json? Shouldn't need that anymore? also we probably need database and/or database.internal for database queries Jun 07 15:15:57 Herrie: that was back then, when I didn't really understand all I was doing, to be on the safe side Jun 07 15:17:27 audio-service[2445]: [] [pmlog] INVALID_JSON {"MSGID":"LS_INVALID_HANDLE", "CAUSE":"The json string is wrong.","JSON":"{\"HANDLER\":0x21393580,\"CREATE_DLI_FNAME\":\"(unknown)\",\"CREATE_DLI_SNAME\":\"(unknown)\",\"DESTR_DLI_FNAME\":\"(unknown)\",\"DESTR_DLI_SN ..."} <-- when trying to initiate a call... might be an old issue, too Jun 07 15:18:50 It's a bit easier now to track issues with everything in the journal Jun 07 15:23:06 Tofe: so /var/log/messages is gone? Or I Misunderstood? Jun 07 15:23:26 And all is in journalctl or somewhere else? Jun 07 15:26:39 not gone, no Jun 07 15:26:59 I redirected the flows of syslog messages Jun 07 15:27:16 Ah OK, so all is in 1 place now? Jun 07 15:27:24 it was a real mess: journald, busybox-syslog and pmlogdaemon all fighting to get the syslog :) Jun 07 15:27:51 well, busybox is out of the way -- it was useless anyway Jun 07 15:28:16 and now pmlogdaemon get the syslog messages that are output by journald Jun 07 15:28:42 journald is the initial syslog daemon (for /dev/log) and also aggregates the other errors for the various services Jun 07 15:29:08 then it's able to output it all as new syslog calls, to another stream Jun 07 15:29:27 So I binded pmlogdaemon to that other stream Jun 07 15:30:05 So we still have the message management from pmlogdaemon, which means differnt log files for different contexts Jun 07 15:30:17 and we have all the messages also in journald Jun 07 15:30:36 (c.f. https://www.dropbox.com/s/u9ym3984ltzkky7/Syslog_with_journald.svg?dl=0 which... is maybe quite unclear) Jun 07 15:32:43 in my svg the missing information is that journald also aggregate messages from other sources (if some services output to stdout/stderr instead of syslog, it would be lost in OSE with only pmlogdaemon) Jun 07 15:49:47 OK still not very clear to me but could be my head today Jun 07 15:50:41 So we can see everything in 1 place now? Or we still need to check both /var/log/messages and journalctl? Jun 07 15:51:01 Legacy-log was mainly for nodejs services it seems Jun 07 15:51:41 I didn't know there was a ls-hubd.log Jun 07 15:52:07 That's where ls2 writes all info to? Jun 07 15:52:19 And what ls-monitor displays or... ? Jun 07 15:54:47 journalctl is enough, now Jun 07 15:55:09 Actually I could have dropped pmlogdaemon, it's not *that* useful Jun 07 15:55:47 I'm not sure what goes exactly in that ls-hubd.log file, I just took some examples from pmlogdaemon current config Jun 07 16:03:17 OK will have a look probably tomorrow. Head is a bit funny today Jun 07 16:03:28 Have some rest, yes Jun 07 16:03:58 Your phone PR contains public/private but no database/database.internal in appinfo.json Jun 07 16:04:28 For call history etc you'd need the database and or database.internal pretty sure Jun 07 16:05:46 Public/private we shouldn't need really since we migrated all Jun 07 16:10:14 For requiredPermissions I usually grep around for luna:// or palm:// and see what we need to add Jun 07 16:12:57 ok, I'll check that Jun 07 18:22:02 Herrie: it seems including miniafservice in Halium improves the situation for tissot; we should schedule a new build to prepare a srcrev bump Jun 07 18:22:42 Tofe: improves which situation? For calls? Jun 07 18:22:55 well, it complains a bit less Jun 07 18:23:13 but I'm still not at the same level as before acg, though I don't quite see why Jun 07 18:27:53 Did we ever have calls working on Tissot? Jun 07 18:28:01 I don't remember Jun 07 18:28:17 well, it looked like I had the call working but no sound Jun 07 19:02:12 Herrie: it was the QML Connections migrations, there was a fatal mistake :) Jun 07 19:03:40 Now I'm at same level as before ACG regarding telephony: the "outgoing call" screen on the phone app appears, but no sound. Jun 07 19:03:52 Even a bit better, as the profile switching is now fixed. Jun 07 19:04:45 I think I'll stop there for today, I "just" wanted to reach same status as before acg for the phone app Jun 07 19:07:35 I've pushed and PR some of it; for the rest it's awaiting some merges so we'll see tomorrow Jun 07 19:07:46 I'll also start a rebuild of tissot's Halium Jun 07 19:13:49 I haven't had time to review it today, but LGTM in principle, getting rid of another logger is always good Jun 07 19:16:53 JaMa: ok good :) Jun 07 19:17:26 I'll try to find a place for this little SVG diagram of mine on the wiki Jun 07 19:17:39 maybe near the architecture ones Jun 07 19:53:08 https://www.webos-ports.org/wiki/Logging_Infrastructure here we go, I didn't spend time for nothing :) Jun 07 21:26:55 Hello. I'm trying to get AsteroidOS boot correctly for qemux86 (virgil , qt5 and eglfs), but its not quite working. The #asteroid people suggested I ask over here. I was able to build it, adding "virglrenderer" and "kmscube" to the yocto bitbake bbclass script, but it doesn't run correctly still. Either it gives a seemingly endless stream of errors Jun 07 21:26:56 saying `Could not queue DRM page flip on screen Virtual1 (Device or resource busy)`, or when it does run sometimes the mouse is stuck in the top-left and jumps rapidly between pos 0 and 1/3 of the screen in x and at whatever y the mouse was (but also bound to the first 1/3 of the screen.) Any ideas? Any help would be appreciated, thanks! Jun 07 21:27:19 They confirm the same errors. Jun 07 21:27:34 Their 1.0 version from about ~2018 does not have this issue. Jun 07 22:03:59 wyattwork: we made some changes at our end to those bits when we upgraded yocto and mesa if I recall correctly Jun 07 22:04:23 It's midnight where I am now so a bit tired to find those commits Jun 07 22:04:40 They should be in meta-webos-ports layer though Jun 07 22:05:52 Ok, thanks! I'll relay that too. Have a good night Jun 07 22:07:31 This is part of it I recall: https://github.com/webOS-ports/meta-webos-ports/commit/c1ab0637f3cbd3cbec3d327d634d2b7c6cf799f7#diff-bd287a1bac08c467e83cf706eb490b47 Jun 07 22:07:45 nice! thanks Jun 07 22:07:55 Tofe/JaMa did those changes they might recall what they did Jun 07 22:08:50 thanks Herrie I'll relay that over too **** ENDING LOGGING AT Mon Jun 08 02:59:58 2020