kernel: Add the real UID by parsing the UID from the /data/user_de/0

[ShirkNeko]

Prioritize retrieving the application UID from /data/user_de. If this fails, fall back to retrieving it from packages.list.

Fix unstable application UID acquisition

[Copilot]

Pull Request Overview

This PR adds a new method to retrieve application UIDs by scanning the /data/user_de/0 directory as the primary approach, with a fallback to the existing packages.list parsing method. This aims to improve stability in UID acquisition.

• Implements a new directory scanning approach using filesystem metadata to extract UIDs
• Adds fallback logic to maintain compatibility with the existing packages.list method
• Includes comprehensive error handling and logging for the new scanning functionality

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 3 comments.

File	Description
kernel/throne_tracker.c	Adds new user data directory scanning functions and modifies track_throne() to prioritize the new method with fallback
kernel/kernel_compat.h	Adds list_count_nodes helper function for kernels < 6.6.0

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[ShirkNeko]

revisions have been completed

[ShirkNeko]

提交 e47775a 后不再回退到packages.list，且支持多用户扫描，并优化 track_throne 扫描触发条件

[Copilot]

Pull Request Overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated 4 comments.

Comments suppressed due to low confidence (1)

kernel/core_hook.c:45

• Duplicate include of throne_tracker.h should be removed.

#include "throne_tracker.h"

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[aviraxp]

Do we need additional selinux rules?

[tiann]

Do we need to scan all packages under all userIds?
If userId = 999, the app's uid in the directory is 100000 * 999 + (userId=0's uid).

If we only need to scan the files in the /data/user_de/0 directory, we can directly use iterate_dir without recursion, This is much simpler.

[ShirkNeko]

Do we need to scan all packages under all userIds? If userId = 999, the app's uid in the directory is 100000 * 999 + (userId=0's uid).

If we only need to scan the files in the /data/user_de/0 directory, we can directly use iterate_dir without recursion, This is much simpler.

Great idea!

[ShirkNeko]

This eliminates the overhead caused by unnecessary recursion.

[aviraxp]

adb install --user 10 MyApp.apk will directly install app to user 10, not 0. So i think it is a must to scan all dir.

[ShirkNeko]

However, in most cases, user 0 is always used as the primary user, and support for multiple users is more of a design reserve.

[ShirkNeko]

Filtering and identification using active users

[tiann]

If supporting multiple users is indeed necessary, it can be done as follows:

1. Scan /data/user_de/0 to get all app UIDs.
2. (Optional) Scan /data/user_de to get all users.
3. (Optional) Use the same step 1 to get apps installed separately by other users.

Given that current multi-user environments cannot be completely isolated, I personally believe that installing apps separately in multiple users to evade detection is unnecessary.

[ShirkNeko]

If supporting multiple users is indeed necessary, it can be done as follows:

1. Scan /data/user_de/0 to get all app UIDs.
2. (Optional) Scan /data/user_de to get all users.
3. (Optional) Use the same step 1 to get apps installed separately by other users.

Given that current multi-user environments cannot be completely isolated, I personally believe that installing apps separately in multiple users to evade detection is unnecessary.

好的，也许我还能改进下