97 Commits

Author SHA1 Message Date
topjohnwu
33aebb5976 Stop embedding executables 2024-07-24 22:49:48 -07:00
topjohnwu
fb5ee86615 Install one single ABI in ramdisk 2024-03-31 22:01:22 -07:00
LoveSy
b6c24a3a8a No more sony init.real tricks
Co-authored-by: canyie <a1364259@163.com>
Co-authored-by: vvb2060 <vvb2060@gmail.com>
2024-02-29 23:40:00 -08:00
topjohnwu
75d905a56d Fix device detection scripts and logic 2023-08-28 22:13:36 -07:00
vvb2060
d12d9e82f1 Force kernel to load rootfs only for legacy SAR devices 2023-08-18 17:18:34 -07:00
LoveSy
33a9abc946 Fix backup fails when ramdisk does not exist 2023-06-30 15:57:09 -07:00
topjohnwu
d7d0a44693 Remove randomness from Magisk 2023-06-14 17:05:49 -07:00
topjohnwu
feb44f875e Migrate PREINITDEVICE in recovery mode
Close #6917
2023-06-02 16:49:04 -07:00
osm0sis
18fe0e6442 Fix scripts
manager.sh + boot_patch.sh:
- all listed files from boot_patch.sh header are required for boot patching, but stub.apk was being removed so install_magisk via addon.d.sh would fail without it; leave it in place

addon.d.sh:
- remove old redundant recovery_actions call (it's also performed by setup_flashable in initialize)
- print ABI to match flash_script.sh output

boot_patch.sh:
- catch and abort on any errors from ramdisk.cpio patching in the future

util_functions.sh:
- fix hiding of mount_partitions /system_root umount stderr
- quote mount_apex .pb DEST name parsing charset for safety even though both work

Fixes #6828
2023-04-08 21:13:40 -07:00
vvb2060
46ce765860 Use stable random seed 2023-04-03 18:32:11 -07:00
topjohnwu
4e2b88b3d0 Rename rules to preinit
It is possible that we will allow more preinit files for modules.
Rename the partition and folders from rules to preinit.
2023-03-21 00:40:11 -07:00
LoveSy
1c2fcd14b5 Mount sepolicy.rules for migration 2023-03-21 00:40:11 -07:00
vvb2060
362eea741f Refactor sepolicy.rules resolve native
Co-authored-by: LoveSy <shana@zju.edu.cn>
2023-03-21 00:40:11 -07:00
vvb2060
2c53356bfd Remove unexpected files 2023-02-26 03:47:38 -08:00
残页
ae34659b26
No kernel repack if it isn't patched at all
It turns out that decompressing and recompressing the kernel is enough to break booting on many devices that use MT6763.
Fix #5124, fix #6204, fix #6566 

Co-authored-by: LoveSy <shana@zju.edu.cn>
Co-authored-by: 南宫雪珊 <vvb2060@gmail.com>
2023-02-17 15:00:14 -08:00
LoveSy
9e8c68af12
Refactor sepolicy.rules resolve
We resolve available partitions for sepolicy.rules when patching
boot and bind mount the partition by magiskinit.

For older devices, the previous logic won't work because the part name
is never readable.

Co-authored-by: topjohnwu <topjohnwu@gmail.com>
2023-02-12 00:36:38 -08:00
topjohnwu
76fbf4634a Update scripts 2022-12-26 16:07:04 -08:00
vvb2060
ad0e6511e1 Stop embedding stub.apk in magiskinit 2022-12-26 16:07:04 -08:00
残页
d155da52ce
More friendly and clear error message 2022-07-01 15:15:54 -07:00
topjohnwu
6e5df1f06b Abort when unsupported dtb is detected 2022-06-16 01:47:23 -07:00
topjohnwu
985249c3d0 Support GKIs without ramdisk
Fix #5819
2022-05-12 03:04:55 -07:00
topjohnwu
490e51c1d7 Don't set RECOVERYMODE if recovery_dtbo exists
Apparently some boot images with ramdisk still have recovery_dtbo,
so this assumption is no longer safe to do. Expect the user to
set this option properly themselves in the app.

Fix #4976, close #5070, close #5184
2022-01-11 00:17:47 -08:00
topjohnwu
f4ed6274a4 Invert vbmeta header patching config
vbmeta header should not be patched in most cases
2021-12-14 04:52:25 -08:00
topjohnwu
228570640e Introduce KEEPVBMETAFLAG env variable
Close #4447, close #4906, close #4901, close #4964
2021-11-23 22:14:12 -08:00
Chris Renshaw
258e89c964 Fix script typo for Sony init.real support
- though may still be broken on magiskinit side, see #4885
2021-11-06 23:44:43 -07:00
topjohnwu
ae23ae2d37 Remove code in scripts that should be removed 2021-10-31 11:30:48 -07:00
Chris Renshaw
6663fd3526
Support custom legacy Sony devices with init.real setup
Custom ROM bring-ups of legacy Sony devices contain the following:
/init (symlink to /bin/init_sony)
/init.real (the "real" Android init)
/bin/init_sony (this was /sbin/init_sony on Android <11)

Kernel loads the ramdisk and starts /init -> /bin/init_sony
/bin/init_sony does low-level device setup (see: https://github.com/LineageOS/android_device_sony_common/blob/lineage-18.1/init/init_main.cpp)
/bin/init_sony unlinks /init and renames /init.real to /init
/bin/init_sony starts /init

Since init_sony needs to run first magiskinit needs to replace init.real instead, so add workarounds based on detection of init.real to boot patcher and uninstaller

Thanks @115ek and @bleckdeth

Fixes #3636

Co-authored-by: topjohnwu <topjohnwu@gmail.com>
2021-10-30 18:59:20 -07:00
topjohnwu
5e1fba3603 Build a single APK for all ABIs 2021-05-13 00:21:04 -07:00
osm0sis
7b9a71c9af scripts: improve boot_patch 64bit detection
- use existing api_level_arch_detect function

Fixes #3961
2021-03-24 03:06:57 -07:00
osm0sis
901d22cdfa scripts: add boot_patch unpack error catching
- failure to unpack wasn't being caught so boot_patch would continue to build a new cpio out of nothing and attempt to repack it (identified in #4049)
2021-03-24 03:06:57 -07:00
topjohnwu
f37e8f4ca8 Fix boot image patching 2021-02-07 01:54:08 -08:00
topjohnwu
3f0136362b Move nand flash handling into boot_patch.sh 2021-01-25 03:37:41 -08:00
topjohnwu
ec8fffe61c Merge Magisk install zip into Magisk Manager
Distribute Magisk directly with Magisk Manager APK. The APK will
contain all required binaries and scripts for installation and
uninstallation. App versions will now align with Magisk releases.

Extra effort is spent to make the APK itself also a flashable zip that
can be used in custom recoveries, so those still prefer to install
Magisk with recoveries will not be affected with this change.

As a bonus, this makes the whole installation and uninstallation
process 100% offline. The existing Magisk Manager was not really
functional without an Internet connection, as the installation process
was highly tied to zips hosted on the server.

An additional bonus: since all binaries are now shipped as "native
libraries" of the APK, we can finally bump the target SDK version
higher than 28. The target SDK version was stuck at 28 for a long time
because newer SELinux restricts running executables from internal
storage. More details can be found here: https://github.com/termux/termux-app/issues/1072
The target SDK bump will be addressed in a future commit.

Co-authored with @vvb2060
2021-01-22 02:29:54 -08:00
topjohnwu
d2bc2cfcf8 Install both 32 and 64 bit binaries 2021-01-18 12:37:08 -08:00
topjohnwu
5a71998b4e Stop embedding magisk in magiskinit 2021-01-18 04:25:26 -08:00
topjohnwu
bb303d2da1 Remove old unused code 2021-01-14 05:59:53 -08:00
vvb2060
984f32f994 Move copy_sepolicy_rules to manager
We don’t need it when patch boot
2020-12-19 13:58:53 -08:00
osm0sis
ce84f1762c scripts: fix sloppy/unpredictable cmd && this || that statements
- be a bit more POSIX to avoid any potential issues when full shell stdout/err are redirected
- actual logic chains remain unchanged
2020-12-06 23:48:43 -08:00
topjohnwu
16e4c67992 Significantly broaden sepolicy.rule compatibility
Previously, Magisk uses persist or cache for storing modules' custom
sepolicy rules. In this commit, we significantly broaden its
compatibility and also prevent mounting errors.

The persist partition is non-standard and also critical for Snapdragon
devices, so we prefer not to use it by default.

We will go through the following logic to find the best suitable
non-volatile, writable location to store and load sepolicy.rule files:

Unencrypted data -> FBE data unencrypted dir -> cache -> metadata -> persist

This should cover almost all possible cases: very old devices have
cache partitions; newer devices will use FBE; latest devices will use
metadata FBE (which guarantees a metadata parition); and finally,
all Snapdragon devices have the persist partition (as a last resort).

Fix #3179
2020-11-02 23:20:38 -08:00
topjohnwu
85755e3022 Tone down our DTB patching
- Do not attempt to patch DTB anywhere outside of boot images as they
are no longer essential. This makes Magisk installation to only modify
strictly boot/recovery partitions again.
- The only required patch for DTB is to strip verity out of partitions
2020-05-05 03:29:36 -07:00
osm0sis
0b70bd2b60 scripts: make remaining header/section dividers uniform
- match other recent formatting updates from topjohnwu
2020-03-21 04:32:20 -07:00
topjohnwu
3049a81c3b Update several scripts
- Update backup format as we might be patching multiple partitions
- Update uninstaller to remove files in persist (sepolicy.rule)
- Better handling for dtb/dtbo partition patching
2020-01-01 14:02:44 +08:00
topjohnwu
16b400964b Update vars for 2SI 2019-09-22 06:45:23 -04:00
topjohnwu
cf2d02c0dd Don't wipe ramdisk when A-only SAR 2019-09-22 06:17:54 -04:00
topjohnwu
0e5a113a0c Support patching mnt_point in fstab in dtb 2019-09-22 04:17:15 -04:00
topjohnwu
a1ccd44013 Change MagiskBoot patch behavior
Use environment variables to toggle configurations for patching ramdisk
2019-09-21 05:55:23 -04:00
topjohnwu
4d91e50d6d Update dtb patch to not use in-place modification 2019-09-21 05:30:04 -04:00
topjohnwu
86481c74ff Allow user to select recovery mode
Close #1674
2019-09-08 00:44:26 -04:00
topjohnwu
1cd45b53b1 Support recovery based Magisk
Some devices (mainly new Samsung phones we're talking here...) using
A only system-as-root refuse to load ramdisk when booted with boot
no matter what we do. With many A only system-as-root devices, even
though their boot image is kernel only, we can still be able to add
a ramdisk section into the image and force the kernel to use it as
rootfs. However the bootloader on devices like the S10 simply does
not load anything within boot image into memory other than the kernel.
This gives as the only option is to install Magisk on the recovery
partition. This commits adds proper support for these kind of scenarios.
2019-03-30 00:49:48 -04:00
topjohnwu
b7e2e972c7 Support boot_img_hdr_v2 2019-03-13 16:51:22 -04:00