Magisk Versions Post 23.0 Cause Bootloop on Fresh Install

[akano12]

Device: Galaxy J6+ (SM-J610f) Android version: 9.0 Magisk version name: 5d6d241 Magisk version code: 23015

I was running 23015 without any problems before flashing stock firmware yesterday, I attempted to root my phone again by flashing 23015 but my phone bootlooped (it was working fine with 23015 on same firmware), same symptoms as in this report https://github.com/topjohnwu/Magisk/issues/4810#issue-1034511424 , i tried flashing 2-3 other versions with no success, however i flashed 23.0 and the phone booted, reflashed 23015 and phone also booted without issues.

Grabbed those files from /sys/fs/pstore because adb was inaccessible

TL;DR flashing 23015 =>bootloop flashing 23.0 then 23015 =>no bootloop dmesg.zip

[yujincheng08]

dm-verity corrupted, printing data in hex

[canyie]

What is the latest version that works with fresh installing on your device? 23001?

[akano12]

What is the latest version that works with fresh installing on your device? 23001?

Not sure about the code but it's the one released in may, v23.0

[canyie]

Any luck with Magisk v24? Some bugs has been fixed.

[akano12]

Any luck with Magisk v24? Some bugs has been fixed.

Issue discussed in this thread is only present on fresh install and sadly i can't provide/help with any tests for now as i'm extremely busy during the next 2 weeks.

Will gladly provide test results and logs when i'm free...

[singhh9596]

Agreed, some privious as well as the latest build v24.2 also caused me a blootloop.

[akano12]

Agreed, some privious as well as the latest build v24.2 also caused me a blootloop.

on fresh install or after updating?

[singhh9596]

Agreed, some privious as well as the latest build v24.2 also caused me a blootloop.

on fresh install or after updating?

While Updating.

[akano12]

Agreed, some privious as well as the latest build v24.2 also caused me a blootloop.

on fresh install or after updating?

While Updating.

then i suggest you open a new issue as the problem described here is different from yours.

[canyie]

@singhh9596 I suggest you try the latest CI build first, the bug may already be fixed. https://github.com/topjohnwu/Magisk/actions/runs/1932106012

[osm0sis]

@akano12 try with stable 24.3, and if that doesn't work, latest Canary 24304 please.

[canyie]

@akano12 Any updates with Magisk 24305? Devs may wanna close stale issues...

[akano12]

@osm0sis

I know it's been a while but can you reopen? Or should i open new issue?

[osm0sis]

I'll reopen if you test and provide some new logs from the latest canary.

[akano12]

@osm0sis (sorry for the pings)

Tried flashing latest canary and flashing failed, also note that flashing 23.0 works well, after phone boots with v23 i can upgrade to lastest canary without issues, initial flashing with anything post 23 fails.

if other logs are needed i can provide

Below you can find dmesg and recovery logs dmesg.log recovery.log

- Device is system-as-root
- Legacy SAR, force kernel to load rootfs
- No vbmeta partition, patch vbmeta in boot image
- Target image: /dev/block/mmcblk0p23
- Device platform: armeabi-v7a
- Constructing environment
- Unpacking boot image
Parsing boot image: [/dev/block/mmcblk0p23]
HEADER_VER      [0]
KERNEL_SZ       [12466999]
RAMDISK_SZ      [871500]
SECOND_SZ       [0]
EXTRA_SZ        [0]
OS_VERSION      [10.0.0]
OS_PATCH_LEVEL  [2021-04]
PAGESIZE        [2048]
NAME            [SRPRF05A008KU]
CMDLINE         [console=null androidboot.hardware=qcom msm_rtb.filter=0x237 ehci-hcd.park=3 androidboot.bootdevice=7824900.sdhci lpm_levels.sleep_disabled=1]
CHECKSUM        [82788b6ef3b8f1668d515d1834585789cc67a1c2000000000000000000000000]
KERNEL_DTB_SZ   [1710255]
ZIMAGE_KERNEL
KERNEL_FMT      [gzip]
RAMDISK_FMT     [gzip]
SAMSUNG_SEANDROID
unexpected ASN.1 DER tag: expected SEQUENCE, got APPLICATION [19] (primitive)
- Checking ramdisk status
- Stock boot image detected
- Patching ramdisk
Value too large for defined data type (os error 75)
Failed to process cpio
! Unable to patch ramdisk

[canyie]

This looks like a new regression compared to the original one, maybe it's better to open a new issue rather than reopening this one 🤔 Btw, I remember I received similar issues before, maybe already fixed on master?

[osm0sis]

@akano12 hit Reference in new issue on your last reply

[akano12]

This looks like a new regression compared to the original one, maybe it's better to open a new issue rather than reopening this one 🤔 Btw, I remember I received similar issues before, maybe already fixed on master?

Does magisk have github actions releases? Or do i have to build myself?

And sure i can open new issue (but will test what you suggested before, maybe it's resolved)

[osm0sis]

Yep, right where it always is: https://github.com/topjohnwu/Magisk/actions 👍