mirror of
https://github.com/kuhyx/testsAndMisc.git
synced 2026-11-11 19:54:01 +01:00
Krzysztof kuhy Rudnicki
1b43ba093c
Prep for a Ulefone Armor X12 Pro (MediaTek MT6765) that has not arrived, validated read-only against the Pixel 6a that has. Partition discovery never hardcodes boot vs init_boot. It reads /dev/block/by-name/, derives the slot scheme from observed _a suffixes, and picks the ramdisk carrier by preference order init_boot then boot over what is actually present. The AOSP launch-version rule explains why that ordering is right; the code consults the filesystem, not the API level. One path therefore serves both a Pixel (no init_boot, falls to boot_a) and a 13-launch MediaTek device (init_boot_a taken). Patching the wrong one yields a phone that will not boot. Classification refuses rather than guesses: ULEFONE needs a positive MediaTek signal AND a vendor match, with no "not a Pixel therefore Ulefone" fallback. Carrier lock is always reported UNDETERMINED, because sys.oem_unlock_allowed, ro.oem_unlock_supported and settings global oem_unlock_allowed read identically whether a device is carrier-locked or merely has the toggle off; asserting a negative would wrongly rule out a path. No flashing script, deliberately. Automating irreversible partition writes defeats the confirmation steps that catch a wrong partition or missing slot suffix. Three bugs found by running the code rather than only linting it: - SIGPIPE: `adb version | head -1` under pipefail exited 141 and killed 00-preflight.sh mid-run - 20-dump-stock.sh exited silently when no recon had run, because find on a nonexistent cache dir fails under set -e - the read-only static scan had a hole: stripping all quoted strings to avoid false positives on printf text also hid `adb shell "reboot bootloader"`. Now only printf/echo/cat arguments are stripped, and a negative self-test asserts the scan fails on injected commands Also narrowed a world-writable udev catch-all (idVendor=="*", MODE="0666") that granted every local process read/write access to every USB device; commented with a backup, adb connectivity re-verified afterwards. Validated: 58/58 offline fixture tests; 00-preflight.sh HOST READY (12/12); 10-recon.sh read-only against the Pixel reports PIXEL, boot_a, A/B, no init_boot, path C-or-D indeterminate, with the device unchanged afterwards. Untested and labelled as such in READY.md: the init_boot branch on real hardware, mtkclient over BROM, and the whole flash/verify sequence. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018ReVsw7qovYkkGbqcU9tn7
97 lines
4.6 KiB
Markdown
97 lines
4.6 KiB
Markdown
# mtk_root
|
|
|
|
Host tooling for rooting a MediaTek Android device: preflight, read-only
|
|
recon, stock partition dump, and post-flash verification.
|
|
|
|
Built ahead of a Ulefone Armor X12 Pro (MediaTek Helio G36 / MT6765) arriving,
|
|
and validated read-only against a Pixel 6a.
|
|
|
|
**There is no flashing script, deliberately.** Automating irreversible
|
|
partition writes would defeat the confirmation steps that catch a wrong
|
|
partition name or a missing slot suffix. Flashing stays manual.
|
|
|
|
## Scripts
|
|
|
|
| Script | Device needed | What it does |
|
|
| ------------------- | ------------- | ------------------------------------------------------------------------------- |
|
|
| `install.sh` | no | Installs deps, clones mtkclient + builds its venv, installs narrowed udev rules |
|
|
| `00-preflight.sh` | no | Verifies the host is intact. Run this first when the phone arrives |
|
|
| `10-recon.sh` | yes | **Read-only.** Reads every fact needed to pick a path, writes a facts file |
|
|
| `20-dump-stock.sh` | yes | MediaTek only. Dumps the stock ramdisk + vbmeta via mtkclient |
|
|
| `90-verify-root.sh` | yes | Post-flash checks, reported one by one |
|
|
|
|
## Usage
|
|
|
|
```bash
|
|
./install.sh # once, or after a Python upgrade
|
|
./00-preflight.sh # confirm the host is ready
|
|
./10-recon.sh # read the device, decide the path
|
|
./20-dump-stock.sh # Ulefone only, after recon
|
|
# ... patch with Magisk and flash MANUALLY ...
|
|
./90-verify-root.sh # check what actually worked
|
|
```
|
|
|
|
Add `--serial <serial>` to any device-facing script when more than one phone
|
|
is attached; they refuse to guess rather than picking one.
|
|
|
|
## Design notes
|
|
|
|
**Partition discovery never hardcodes `boot` vs `init_boot`.** It reads
|
|
`/dev/block/by-name/`, derives the slot scheme from observed `_a` suffixes, and
|
|
picks the ramdisk carrier by preference order `init_boot` → `boot` over what is
|
|
actually present. Per AOSP, devices launching on Android 13+ carry the ramdisk
|
|
in `init_boot`; earlier devices keep it in `boot` even after upgrading. That
|
|
rule explains why the preference order is right, but the code consults the
|
|
filesystem, not the API level — so one code path serves both a Pixel (no
|
|
`init_boot`, falls through to `boot_a`) and a 13-launch MediaTek device
|
|
(`init_boot_a` taken). Patching the wrong one produces a phone that will not
|
|
boot.
|
|
|
|
**Classification refuses rather than guesses.** `ULEFONE` requires a positive
|
|
MediaTek platform signal _and_ a vendor/model match. There is no "not a Pixel,
|
|
so presumably the Ulefone" fallback: a third phone attached by accident lands
|
|
in `UNKNOWN` and every caller stops.
|
|
|
|
**Carrier lock is reported as `UNDETERMINED`, always.** No readable property
|
|
distinguishes carrier-locked from "the OEM unlocking toggle is simply off":
|
|
`sys.oem_unlock_allowed` mirrors the toggle, `ro.oem_unlock_supported` is not
|
|
populated on Pixels, and `settings get global oem_unlock_allowed` is null until
|
|
the toggle is used. Asserting a negative here would wrongly rule out a path, so
|
|
the script reports the raw signals and refuses to conclude. Settle it out of
|
|
band via an IMEI check.
|
|
|
|
**`10-recon.sh` is read-only and the tests enforce it.** `tests/run_tests.sh`
|
|
statically scans it for `fastboot`, `magiskboot`, `adb reboot/push/install`,
|
|
`settings put`, `dd`, `mkfs` and `erase`, and separately asserts the scan
|
|
_fails_ when such a command is injected into a copy — a guard never observed
|
|
failing is not known to work.
|
|
|
|
## Artifacts
|
|
|
|
Images, APKs and facts files go to `~/.cache/mtk-root/`, never into the repo:
|
|
the pre-commit hook rejects all binaries, and a factory image runs to several
|
|
GB. Only text manifests belong in git.
|
|
|
|
Ulefone firmware is deliberately **not** pre-fetched — the stock ramdisk is
|
|
dumped off the device itself, so it matches that exact unit and build.
|
|
|
|
## Tests
|
|
|
|
```bash
|
|
./tests/run_tests.sh
|
|
```
|
|
|
|
58 assertions, fully offline. Fixtures are `props.txt` + `by-name.txt`
|
|
directories fed in via `MTK_ROOT_FIXTURE`; `MTK_DUMP_CMD` stubs the mtkclient
|
|
invocation so the dump/manifest logic is exercised without hardware.
|
|
|
|
`tests/fixtures/pixel-boot-ab/` is harvested from a real Pixel 6a. The
|
|
`ulefone-*` fixtures are synthetic and encode _assumed_ property values — see
|
|
`NOTES.md`.
|
|
|
|
## Related
|
|
|
|
`../single_use/utils/root_bl9000.sh` is an older, single-device rooting script
|
|
for a Blackview MT6893. It is untouched by this toolkit and still works; the
|
|
two share no code by design, so changes here cannot break it.
|