- simplify rdtmp and $ramdisk creation and keep only in `unpack_ramdisk`
- fix `reset_ak` and `setup_ak` behavior if run before $block is set
- only save boot.img and boot-new.img in *-files subdirectories on multi-partition setups since otherwise the tmpfs (in recovery) and ext4 .img (when booted) size usage gets quickly out of hand
- fix slot_select=inactive when used with block=auto or block=boot partition detection
- the expected "cat: write error: No space left on device" gives a non-zero return code so always return true
- devices that flash dtbo abort since the return code would get passed back to the Backend and silently fail as the final command in anykernel.sh (via `write_boot`), so catch seperately with a proper error message for other image write methods
- some device recoveries (e.g. TWRP for wahoo) seem to have issues with the new `cat <image> /dev/zero > <block>` method of flashing, so allow setting customdd="bs=1048576" shell variable in anykernel.sh to work around this by reverting back to dd wipe+flash for dtbo as it does with boot
- abstract out execution directory as $home to fascilitate testing individual scripts (defaults to /tmp/anykernel)
- general script clean ups for uniformity, readability, speed and simplicity
- simplify consecutive `ui_print` commands where possible
Backend only fixes:
- use busybox umount and mount whenever possible to avoid linker warnings in stderr during system-as-root setup
- use `busybox --install -s` for bin/* setup
- consolidate setup to abort earlier if there are issues
- add backwards compatible symlinks for those who might not have updated their anykernel.sh with ak3-core.sh (to be renamed in subsequent commit)
Tools only fixes:
- add magiskboot to replace separate support for AOSP and PXA-AOSP formats, with MTK headers, BLOB, DHTB, SEAndroid and Nook signing
- add magiskpolicy to allow for easy sepolicy patching
- remove xz and lz4 since magiskboot supports these
AK3 Core only fixes:
- use full `ui_print` function implementation like the backend
- add `abort` function to simplify exit on errors
- fix $patch directory being excluded from `reset_ak`
- update $ramdisk_compression handling for magiskboot
- magiskboot also resolves long-standing request for built-in Bump support (#18)
- update Sony ELF and related support for initial repack as AOSP by using mkbootimg as plug-and-play optional add-in binary
- update Sony ELF MTK header support using mkmtkhdr only when ELF is initially unpacked by unpackelf
- add support for MTK RECOVERY ramdisk header type with Sony ELF
- simplify U-Boot dumpimage/mkimage and Sony ELF unpackelf/mkbootimg split_img file handling
- update patch_cmdline to handle both magiskboot as header prop file, and unpackelf or mboot as cmdline.txt
- use `cp -fp` for `backup_file` and `restore_file` to hopefully better preserve attributes if used outside of the ramdisk
- refactor setup into `setup_ak` to simplify `reset_ak` trigger to find slot and block again
- add fallback to busybox for compression formats (i.e. lzop) which magiskboot cannot handle
IMPORTANT NOTES:
1) forcing repack with uncompressed cpio ramdisk is not currently supported by magiskboot
2) forcing repack with lzop ramdisk when previously another format is not currently supported by magiskboot
- extend `reset_ak` with `reset_ak keep` which will keep files to be available to patch the other slot
- add optional slot_select=inactive shell variable to force inactive slot during slot detection
- AOSP has added --dtb, so to avoid conflicts *unpackbootimg and unpackelf have had what was previously dumped as *-dtb renamed to *-dt
- the same corresponds to any replacement version of these included in the zip root, i.e. they are now dt or dt.img for the old style from before hdr_v1, and dtb or dtb.img for the new style for hdr_v2
Note: You MUST update to the newer mkbootimg and unpackbootimg for this commit if you are using architectures other than the included ARM builds!
See AIK-mobile v3.5 for updated unpackelf, and mkbootimg/unpackbootimg binaries:
https://forum.xda-developers.com/showthread.php?t=2073775
- dd uses bs=512 by default, which is far from optimal for sequential reads.
Use bs=1048576(1M) for initial dump.
This improves read performance by a measurable amount.
- Zero-fill only the remaining blocks.
There's no need to zero-fill all blocks,
not to mention the additional/unnecessary wear and tear
to the flash storage.
Linux pipe uses bs=4096 by default. By using cat for this,
we additionally gain much better write performance.
On the OnePlus 6T, final write operation took 7.1 seconds in total with
2 consecutive dd commands. With cat, it now takes 0.08 seconds.
Signed-off-by: Park Ju Hyung <qkrwngud825@gmail.com>
- checks the supported.versions prop in anykernel.sh and matches against it to prevent users flashing incorrect kernels for ROM versions
- supported.versions can be a list, e.g. "7.1.2, 8.0, 8.1" or a range, e.g. "8.1 - 9" of versions in the correct formatting for that Android version in ro.build.version.release in /system/build.prop
split_boot is for fine grained control over what exactly gets flashed
at what point. It makes more sense for flash_dtbo to live in write_boot,
which does everything all at once.
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
Some zips may want to flash just the dtbo so add a function that can be
called in anykernel.sh to make that happen.
Signed-off-by: Nathan Chancellor <natechancellor@gmail.com>
- gzip in busybox 1.29.1 doesn't seem to use compression anymore, with -1 and -9 both coming
out to the same size, generally dramatically larger than previously resulting in
boot-new.img exceeding the partition size on some devices so then failing the check.
e.g. 28mb ramdisk.cpio -> 11.2mb ramdisk.cpio.gz in 1.28.4, is now 20mb in 1.29.1
thanks nathanchance for reporting
- was broken with search strings containing paths (a limitation with busybox sed)
- now matches blank lines and lines containing only white space for next-empty-line matching
- include end-of-file as the final possible end-point for next-empty-line matching
This reverts commit bafe4bd586.
A certain level of functioning recovery is expected by AnyKernel2 (versus this patch's source, the Magisk installer, which is extremely pessimistic).
- search using a list of known ramdisk/kernel or recovery by-name partition names
- block=auto is equivalent to block=boot
- for use with broad, device non-specific AK2-based mod zips
- check for and remove patch for multi-partition/reset_ak (so removal of ramdisk/ and patch/ with the addition of $(basename block)-files is now the full trigger)
- wipe $(basename block)-files/ramdisk to ensure proper copied output, as the original contents are retained in the copied rdtmp
- for ramdisk-only changes that don't require added files or patch snippets, create the $(basename block)-files subfolder for proper output backup with reset_ak
- if no ramdisk is found in the zip, check for and copy files in from "$(basename $block)-files", e.g. boot-files + recovery-files subfolders, each which can contain their own zImage, and ramdisk and patch subfolders, etc.
- as part of the reset_ak command, copy the output files back to the corresponding *-files subfolder for use in debugging
* when ak2-core.sh gets run again by reset_ak after changing block= in anykernel.sh the above process begins again
osm0sis