| 30 Oct 2024 |
 atemu12 | IT WAS MAGISK | 10:46:08 |
| atemu12 | WHAT | 10:46:11 |
| atemu12 | Flashed the OS without flashing Magisk again and it works now | 10:46:24 |
| atemu12 | Perhaps you need to boot without magisk once?! | 10:46:56 |
| atemu12 | I'll try re-flashing magisk | 10:47:02 |
| atemu12 | Still works | 10:49:23 |
| atemu12 | Maybe there's some sort of migration that needs to happen on boot but it doesn't happen when Magisk is there? Perhaps it detects that the boot image was modified? | 10:53:32 |
| atemu12 | Man it's so great to not have to wipe data | 10:53:53 |
|  mraethel changed their display name from mraethel [new] to mraethel. | 14:28:54 |
 samueldr | that's interesting, I wonder if it's some sort of "protection" built into magisk for the usual use case (hostile vendor software) that is causing surprising behaviour here | 16:38:14 |
| samueldr | did your tests on other devices use magisk too? | 16:38:25 |
| atemu12 | Now that you mention that, I think I skipped that step | 17:16:56 |
| atemu12 | I'll experiment with this when I get the chance to | 17:17:13 |
| samueldr | yeah, no worries, I was wondering if that thought could help you :) | 17:21:18 |
| 1 Nov 2024 |
| atemu12 | So I've been trying to get LOS21 and the build starts successfully but at some point javac's JVM segfaults O.o | 10:57:50 |
| atemu12 | --------------- S U M M A R Y ------------
Command Line: -XX:OnError=cat hs_err_pid%p.log -XX:CICompilerCount=6 -XX:+UseDynamicNumberOfGCThreads -DremoveAndroidCompatAnnotations=true out/host/linux-x86/framework/jarjar.jar process packages/modules/OnDevicePersonalization/federatedcompute/apk/jarjar_rules.txt out/soong/.intermediates/packages/modules/OnD
ompute/apk/FederatedCompute/android_common_apex33/737f09fcc0b17bc4650c68e8a26b3dd4/combined/FederatedCompute.jar out/soong/.intermediates/packages/modules/OnDevicePersonalization/federatedcompute/apk/FederatedCompute/android_common_apex33/737f09fcc0b17bc4650c68e8a26b3dd4/jarjar/FederatedCompute.jar
Host: AMD Ryzen 9 7950X3D 16-Core Processor, 32 cores, 62G, Linux
Time: Thu Oct 31 00:24:26 2024 UTC elapsed time: 3.084017 seconds (0d 0h 0m 3s)
--------------- T H R E A D ---------------
Current thread (0x00007fff2c0021e0): JavaThread "C2 CompilerThread3" daemon [_thread_in_native, id=958115, stack(0x00007fffa7b00000,0x00007fffa7c00000)]
Current CompileTask:
C2: 3084 1248 4 org.objectweb.asm.ClassReader::readCode (5113 bytes)
Stack: [0x00007fffa7b00000,0x00007fffa7c00000], sp=0x00007fffa7bfb830, free space=1006k
Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code)
V [libjvm.so+0x522a80] PhaseChaitin::Register_Allocate()+0x1d90
V [libjvm.so+0x6acb72] Compile::Code_Gen()+0x4c2
V [libjvm.so+0x540b74] C2Compiler::compile_method(ciEnv*, ciMethod*, int, bool, DirectiveSet*)+0x1cc4
V [libjvm.so+0x6b649e] CompileBroker::compiler_thread_loop()+0x11be
V [libjvm.so+0x766b2f] JavaThread::run()+0x26f
V [libjvm.so+0x71d835] thread_native_entry(Thread*)+0x395
siginfo: si_signo: 11 (SIGSEGV), si_code: 1 (SEGV_MAPERR), si_addr: 0x0000000000000000
| 10:58:56 |
| atemu12 | Oh wait no it's actually jarjar | 10:59:30 |
| atemu12 | Still I don't think I've ever seen a JVM segfault before | 10:59:41 |
| atemu12 | # Problematic frame:
# V [libjvm.so+0x522a80] PhaseChaitin::Register_Allocate()+0x1d90
| 11:02:09 |
| atemu12 | Huh, apparently it was a one-off. | 12:22:41 |
| atemu12 | 
Download Screenshot_20241101-144637_Settings.png | 13:49:14 |
| atemu12 | 🥳 | 13:49:24 |
| atemu12 | https://github.com/nix-community/robotnix/pull/250 | 14:13:13 |
| atemu12 | Okay so apparently signature spoofing patches aren't required anymore but µG doesn't appear to have the permission or ability | 14:53:45 |
| atemu12 | Anyone have a clue why that might be? | 14:54:00 |
| atemu12 | Ugh they check that the µG signature matches upstream's key | 15:12:04 |
| atemu12 | Wtf | 15:12:05 |
| atemu12 | And we override it with a snakeoil cert for signing purposes | 15:12:28 |
 magic_rb | Wait so normally compiled uG will just work with lineageos? They finally figured it out? | 16:10:55 |
| atemu12 | Yeah that happened a while ago | 16:11:25 |
