**** BEGIN LOGGING AT Fri Dec 11 02:59:56 2020 Dec 11 11:17:05 @Shakendo I tried to use the asus_qcom_common device tree and the EDL reboots are happening on the reference rootfs too. I'll now try to disable selinux and tell you if it worked... Dec 11 12:01:04 However, I don't see any pulseaudio errors there. There are just some rmt_storage errors repeating so that's probably why the device stops. Dec 11 12:19:52 Is this something that's started in android init? Does it happen when you disable lxc? Dec 11 12:24:38 @dohniks [Is this something that's started in android init? Does it happen when you disabl …], Here's how the repeating error looks like: … [ 350.999038] [ 350.999038] (CPU:2-pid:3010:rmt_storage) [12:00:01.935307659] IPC_RTR: msm_ipc_router_bind: rmt_storage Do not have permissions … It looks like something "sees" this and forces the Dec 11 12:24:38 device to go to EDL. I'll try if it happens with stopped LXC... Dec 11 12:26:33 Another weird thing is that it only happens when modem is mounted Dec 11 12:30:29 @dohniks [Is this something that's started in android init? Does it happen when you disabl …], hmm it's taking to long to stop the container... the device goes to EDL before LXC stops Dec 11 12:35:07 you can disable it somehow so it never starts in the first place Dec 11 12:35:14 I think it's in the docs Dec 11 12:38:34 @dohniks [you can disable it somehow so it never starts in the first place], What's rmt_storage btw? So it is not required for Halium? What about Ubuntu Touch? Dec 11 12:39:32 dunno 🙈 Dec 11 12:53:34 I tried to run dmesg -w before the stop happened. It's something with modem: https://pastebin.com/7EuASBGi Dec 11 13:38:31 @Korko123 [What's rmt_storage btw? So it is not required for Halium? What about Ubuntu Touc …], iirc its used for parts of the network stack, I imagine you will have to try disabling a couple things one by one (and re-enabling if it doesnt work) before you figure it out Dec 11 14:00:43 @Shakendo [iirc its used for parts of the network stack, I imagine you will have to try dis …], Ok. rmt_storage is one of the vendor blobs. Is it safe to remove the file? I know the system image build will fail without it, but I can still remove it from proprietary-files.txt Dec 11 14:03:04 for testing purposes, yes its ok to remove it, however you will likely need it at some point for something to work, so I would backup and files you delete/modify before making said changes Dec 11 14:03:48 (Photo, 410x74) https://irc.ubports.com/pOV2Ak8W.png Ok, I'll try to remove these files while the system is booted... Dec 11 14:05:03 you can likely remove them from the source side before compile and have it not look for the file Dec 11 14:06:10 @Shakendo [you can likely remove them from the source side before compile and have it not l …], I'll just remove it from proprietary-files.txt Dec 11 14:06:24 (Photo, 540x70) https://irc.ubports.com/nRdybSQI.png Btw, how can I remount /system rw? Dec 11 14:06:53 sudo mount -o remount,rw / is how we do that in UT Dec 11 14:07:07 I know but I want the system image Dec 11 14:14:50 Hmm another MCLK error? https://pastebin.com/cX2BpQsP Dec 11 14:27:51 I'm trying to get halium9 gsi running on my galaxy a20 … 1. I patched the kernel and it boots to SSH. … 2. LXC container is running. … But I'm still on my bootloader screen. hwcomposer test returns terminated. … Could someone have a look at the strace and can tell me what is wrong? … https://pastebin.com/EVdyAqrY Dec 11 14:28:54 Here is the kernel source so far … https://gitlab.com/arpio/kernel_samsung_a20_halium Dec 11 14:31:04 Anubhab Sarkar was added by: Anubhab Sarkar Dec 11 14:43:30 @Korko123 [Hmm another MCLK error? https://pastebin.com/cX2BpQsP], qmuxd is partnof your cellular stack Dec 11 14:44:24 @Shakendo [qmuxd is partnof your cellular stack], Hmm... am I the only one who had these issues on MSM8916 or is there someone who had them? Dec 11 14:44:43 @Korko123 [Hmm another MCLK error? https://pastebin.com/cX2BpQsP], is this a 32 or 64-bit device? Dec 11 14:44:57 @Shakendo [is this a 32 or 64-bit device?], ARM64 Dec 11 14:46:40 ok, dumb question, did this device have android on it and work fine? Dec 11 14:46:49 I'm using this rootfs: http://bshah.in/halium/halium-rootfs-20170630-151006.tar.gz Dec 11 14:47:05 @Shakendo [ok, dumb question, did this device have android on it and work fine?], Yep, it has CM and LOS port from Android 6 up to 10 Dec 11 14:47:13 @Shakendo [ok, dumb question, did this device have android on it and work fine?], [Edit] Yep, it has CM and LOS ports from Android 6 up to 10 Dec 11 14:47:22 I just used CM-14.1 sources Dec 11 14:47:38 ok, Im starting to wonder if your vendor blobs are possibly corrupted Dec 11 14:48:54 @Shakendo [ok, Im starting to wonder if your vendor blobs are possibly corrupted], I think the device tree is a bit wrong... proprietary-files.txt contains blobs that even don't exist on LineageOS 14.1 port... when I execute extract-files.sh script, it'll show lots of missing files errors and system image will fail to build because of mi Dec 11 14:48:54 ssing files. Dec 11 14:52:09 @Shakendo I used sources from this ROM: https://forum.xda-developers.com/t/rom-z00ed-z00rd-unofficial-unified-lineageos-14-1-aokp.3520986/ … https://github.com/audahadi Dec 11 14:53:20 hmm... now, I think I used the sources incorrectly... this ROM is unified for both Z00ED and Z00RD, and I used only Z00ED sources along with asus_qcom_common Dec 11 14:57:05 that could be part of your issue Dec 11 14:58:22 @Shakendo [that could be part of your issue], Ok. How are those 'unified' builts made? If I execute breakfast Z00xD, it'll just show that device doesn't exist. Dec 11 14:58:31 @Shakendo [that could be part of your issue], [Edit] Ok. How are those 'unified' builds made? If I execute breakfast Z00xD, it'll just show that device doesn't exist. Dec 11 15:00:09 the unified builds vary between devices, on my device the unified build is just a script that copies the blobs for the right device to the right location on first boot Dec 11 15:00:45 @Shakendo [the unified builds vary between devices, on my device the unified build is just …], Oh ok... so the devices use the same device tree? Dec 11 15:01:19 I'll try to build Halium for Z00RD and boot it on Z00ED Dec 11 15:01:57 in my devices case yes, I had to find non-unified sources cause the unified script wont work in UT like it does in android Dec 11 15:02:47 @Shakendo [in my devices case yes, I had to find non-unified sources cause the unified scri …], Oh... there are 2 device trees available: Z00ED and Z00RD. But, the vendor blobs contain a directory called "Z00xD", which is unified. Dec 11 15:07:38 does theMuppets have your device? Dec 11 15:08:07 @Shakendo [does theMuppets have your device?], It only has proprietary_vendor_asus, which contains only msm8916-common directory, not "Z00xD" Dec 11 15:08:33 ok, sonyou would still need the Z00? Dec 11 15:09:35 @Shakendo [ok, sonyou would still need the Z00?], I can build without "Z00xD", but lots of features won't work... e. g. Ubuntu Touch didn't get past the splash screen even though I added udev rules. Dec 11 15:11:20 ok, if you build without Z00 does all the hardware get detected (can check through logs/dmesg/logcat) Dec 11 15:11:53 @Shakendo [ok, if you build without Z00 does all the hardware get detected (can check throu …], I'll try it after I build for Z00RD... I'll let you know. Dec 11 15:12:07 ok Dec 11 15:37:35 Butw there is ubports halium 9 armhf rootfs? Dec 11 15:37:42 or arm64 only? Dec 11 15:48:53 @Lukapanio [Butw there is ubports halium 9 armhf rootfs?], afaik its arm64 since everything that shipped with android8 or newer was arm64 Dec 11 15:49:53 @Shakendo [afaik its arm64 since everything that shipped with android8 or newer was arm64], Hm sad Dec 11 15:50:24 hm Dec 11 15:50:26 here https://github.com/ubports-on-fxtec-pro1/rootfs-builder-debos-android9/releases/ Dec 11 15:50:30 it was armhf Dec 11 16:03:22 Hochberg hiltunen was added by: Hochberg hiltunen Dec 11 16:03:27 (Photo, 449x745) https://irc.ubports.com/djl6EwF0.png Dec 11 16:08:33 @Shakendo I tried the Z00RD build and it's same... Dec 11 16:25:56 maitreya29 was added by: maitreya29 Dec 11 16:32:40 @Korko123 [@Shakendo I tried the Z00RD build and it's same...], had a feeling itnwas going to be Dec 11 16:33:58 @Shakendo [had a feeling itnwas going to be], I'm now building for Z00ED, but without the device specific vendor blobs (Z00xD) Dec 11 16:34:58 @Lukapanio [Butw there is ubports halium 9 armhf rootfs?], there are no builds as of now, but possible to enable it since packages are still compiled Dec 11 16:35:04 do you have a device to test it? Dec 11 16:46:16 @NotKit [do you have a device to test it?], yep, karnak Dec 11 16:46:31 after update it become real android 9 Dec 11 17:20:31 SHB_nik was added by: SHB_nik Dec 11 17:47:38 @arpio1 [I'm trying to get halium9 gsi running on my galaxy a20 … 1. I patched the kernel a …], Is this related? Syslog at start … https://pastebin.com/h7WCiRSd Dec 11 18:35:56 hı. how install halium-9? Dec 11 18:36:09 ı use —system-as-root. but ı dont see ssh Dec 11 18:36:24 why ı fix this problem? Dec 11 21:09:15 @Shakendo The device doesn't stop and go to EDL when I build without the "Z00xD" vendor blobs... the same errors in dmesg are showing though. Dec 11 21:09:54 Also I was wrong. Ubuntu Touch gets past the splash screen even when I don't use these blobs. But, the device is a bit "laggy" without them. **** ENDING LOGGING AT Sat Dec 12 02:59:56 2020