**** BEGIN LOGGING AT Thu Jun 25 02:59:57 2020 Jun 25 06:13:18 Morning Jun 25 06:13:45 Herrie: the old nodejs code only fails to build on dunfell, whereas my nodejs12 branch builds on both Jun 25 06:13:54 we might have missed a step here Jun 25 06:17:01 Herrie: yes, I agree with your analysis of the traces for halium 9 Jun 25 07:09:12 Morning Jun 25 07:09:48 Tofe: I guess this would be the most proper branch so we keep history as well: https://github.com/webOS-ports/nodejs-module-webos-pmlog/commits/herrie/webOS-OSE Jun 25 07:10:07 Tofe: Any idea why ueventd might take up so much memory ? Jun 25 07:10:40 I didn't update the UDEV rules yet, but I understood that shouldn't make a difference at this point yet Jun 25 07:15:47 ueventd doesn't use udev, it's an equivalent, if I understood correctly Jun 25 07:20:55 Tofe: That was my understanding too Jun 25 07:24:10 I guess logcat might give some clues as to the cause of ueventd using so much memory, but a bootloop is not helping there... Jun 25 07:29:11 JaMa: Would something like this be OK? https://github.com/webOS-ports/nodejs-module-webos-pmlog/commits/herrie/webOS-OSE Jun 25 07:33:33 morning Jun 25 07:33:38 +1 on herrie/webOS-OSE git history Jun 25 07:37:26 JaMa: OK let me fix that now then. Jun 25 07:40:02 JaMa: https://github.com/webOS-ports/nodejs-module-webos-pmlog/commits/webOS-ports/webOS-OSE Jun 25 07:40:26 Tofe: Will try this to see if I can prevent the panic & reboot: https://github.com/Herrie82/android_kernel_xiaomi_mido/commit/49354ac834125197b0ccae97a6274deab2e3d6f1 Jun 25 07:54:30 Herrie: if only we could retrieve the logcat... Jun 25 07:55:13 do you have adb access ? I see in the traces that you have about 20-30s before the memory is full Jun 25 07:56:00 Tofe: No not yet Jun 25 07:56:14 I will try again, so far didn't succeed into getting in with ADB Jun 25 08:01:39 If we're stuck, maybe we can insert a background service during startup that starts logcat and outputs it on a file, like "/system/bin/logcat > /logcat.txt &" Jun 25 08:20:09 Tofe: Yeah that's a good idea but let me first play with a few defconfig options Jun 25 08:20:31 If I can get it to not reboot I should be able to get to logcat already Jun 25 08:43:21 JaMa/Tofe: Bump nyx-modules-hybris a bit (forgot to do that back in December and added a fix for it to work correctly now as well :P) Jun 25 08:43:31 So i guess it will trigger some quite big rebuilds :S Jun 25 08:48:47 nyx-modules-hybris is only used by the packagegroup, shouldn't cause any big rebuilds other than image Jun 25 08:54:16 JaMa: OK :) Jun 25 08:54:22 It didn't at my end, so I guess we're good Jun 25 08:54:44 JaMa: You know if there will be an OSE release this month? Jun 25 09:00:27 Tofe: I didn't flash Android 9 on my mido yet, so that could likely be the cause as well somehow Jun 25 09:12:34 Herrie: there should be, but I haven't seen anything super exciting Jun 25 09:13:04 JaMa: Well I don't expect anything super exciting... Jun 25 09:13:05 I have probably the most of the commits there and I'm not even working on webOS anymore :) Jun 25 09:13:37 Got my WAM rework up to date with latest Chromium and builds now on all targets it seems, just wayland keeps crashing it seems, so I'll leave that for Tofe at some point Jun 25 09:13:49 JaMa: What are you working on then? Jun 25 09:13:54 ROS Jun 25 09:14:01 Ah OK Jun 25 09:25:06 Herrie: I'll switch to WAM when I get a preview on the camera :) Jun 25 09:26:02 should we try dunfell builds again soon or do you have a lot of other things in queue to merge before that? Jun 25 09:26:31 OSE is supposed to be upgraded relatively soon to dunfell, I don't want OSE to beat us there :) Jun 25 09:27:08 JaMa: Well there's nothing really big pending now, so we can test with dunfell Jun 25 09:27:32 ok, will trigger unstable builds once pmlog is fixed Jun 25 09:28:39 JaMa: You're waiting for something from me or at your end for it to be fixed? Jun 25 09:29:37 I can update the branch and bump SRCREV, I didn't know if you're working on it already Jun 25 09:33:05 JaMa: https://github.com/webOS-ports/nodejs-module-webos-pmlog/commits/webOS-ports/webOS-OSE Jun 25 09:33:12 Just bump of SRCREV should be enough Jun 25 09:35:13 Tofe: No way to get into adb shell it seems Jun 25 09:35:26 I can try multiple times a second and it simply doesn't connect Jun 25 09:35:28 ok, doing it now Jun 25 09:35:52 The WAM rework was mainly for newer Chromium, so that could be rebased to dunfell easily if needed Jun 25 09:38:01 Herrie: in this case, try to insert a logcat call in some lxc start script, it might be the easiest way Jun 25 09:42:52 Tofe: In lxc's pre-start.sh or somewher eelse? Jun 25 10:33:34 in pre-start it should be all good Jun 25 10:40:24 Tofe: "/data/luneos/android/system/bin/logcat > /data/luneos/media/internal/logcat.txt &" ? Or ... ? Jun 25 11:40:05 why /data/luneos prefix ? this will be run from within luneos boot Jun 25 11:56:22 Tofe: I'm always confused about the mounts, bind mounts etc Jun 25 11:56:38 Testing something on Hammerhead now and will then push to mido Jun 25 12:10:57 Tofe: Does it matter where I put it inside the .sh? Jun 25 12:11:15 Like this oK? https://bpa.st/DOAA Jun 25 12:11:42 Herrie: yes, that should be ok Jun 25 15:10:46 Tofe: Hmmz doesn't seem to work really :S Jun 25 15:11:05 My hammerhead refuses to finish unpacking the rootfs Jun 25 15:11:13 I think my storage might be a bit flaky on hammerhead Jun 25 15:21:37 Tofe: Seem it might be CWM related somehow, installed TWRP now Jun 25 15:28:25 Tofe: On Hammerhead have a /media/internal/logcat.txt but it's empty :S Jun 25 16:17:45 ah, damn Jun 25 16:17:50 maybe we start logcat too soon Jun 25 16:41:45 Tofe: Also not sure Jun 25 17:13:51 Herrie: I can propose you another line to start, if you want Jun 25 17:17:50 Tofe: sure Jun 25 17:18:26 Something in an init.Rc? Jun 25 17:18:37 ( /bin/sleep 10 ; /android/system/bin/logcat ) > /media/internal/logcat.txt & Jun 25 17:18:40 no, same place Jun 25 17:18:44 Tofe: seems now it doesn't get that far even with same image Jun 25 17:18:58 I mean my pstore has less vs yesterday Jun 25 17:19:05 But I used same image Jun 25 17:19:17 And the output is garbled a bit again too Jun 25 17:19:21 I guess there's one service that eats it all Jun 25 17:19:55 Yeah will play a bit with it later today or tomorrow, dinner soon then kids to bed :P Jun 25 17:20:12 I won't be around much the next 4 days Jun 25 18:01:27 Tofe: seems the same Jun 25 18:13:50 damn Jun 25 18:23:23 Well I did flash LOS 16.0 in between... Maybe I should revert back to 7.0 rom in general to have the same results Jun 25 18:27:01 flashing LOS 16.0 looks like a good idea Jun 25 18:27:24 is it mido ? Jun 25 18:29:10 Yes Mido Jun 25 18:29:14 Mrs still got my Tissot Jun 25 18:31:36 You would say that flashing 16.0 is a good idea seeing device and kernel trees come from LOS 16.0 too **** ENDING LOGGING AT Fri Jun 26 03:01:39 2020