# Stealthy-Kernelmode-Injector **Repository Path**: baohongyu/Stealthy-Kernelmode-Injector ## Basic Information - **Project Name**: Stealthy-Kernelmode-Injector - **Description**: DrvLoad clone 备份来的... - **Primary Language**: C - **License**: Not specified - **Default Branch**: main - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 0 - **Forks**: 0 - **Created**: 2023-10-03 - **Last Updated**: 2023-12-18 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README # Stealthy Kernel-mode Injector Manual mapper that uses PTE manipulation, Virtual Address Descriptor (VAD) manipulation, and forceful memory allocation to hide executable pages. ## Injector To hide our exectuable pages, the injector supports the following techniques: - Spoof page protection by modifying PTE execute bit or VAD protection flags - Remove created VAD entries after allocation - Forcefully allocate and write our DLL to wherever we want in the process address space Manipulating the VAD or underlying PTEs is desired in order to spoof page protection from API functions which rely on the VAD such as `NtQueryVirtualMemory`. In the case of this injector, we use VAD and PTE manipulation interchangely as they are both used to fake the protections specified by the VAD. Rather than "randomly" allocating memory by using an API function such as `NtAllocateVirtualMemory`, forcefully allocating memory via the VAD is used to write our DLL to an "unconventional" location in a processes address space. Anti-tampering solutions using heuristics to detect unwanted executable pages outside signed modules may avoid searching some memory regions to prevent false positives. This injector will allocate memory behind the thread stack. To execute our DLL, the injector will use `SetWindowsHookEx` to load a valid DLL such as `ntdll.dll`, then use the hook procedure to hijack control flow to call our DLL entry point. ## Driver Stealthy communication between our user-mode client and kernel-mode driver is achieved by placing a series of PatchGuard safe function hooks inside the windows kernel. An initial hook is placed on a function that is callable from user-mode using a syscall. Then, instead of immediately redirecting execution to our driver, the hook will lead to another function within the same kernel image, which we then hook to be redirected to our driver. The reason why we go through this process is to make it difficult to verify the integrity of `ntoskrnl.exe`. Since there are few usable functions inside `ntoskrnl.exe` suitable for communication, a separate driver looking to prevent tampering of `ntoskrnl.exe` could manually verify functions known to be used for user to kernel-mode communication by checking if they lead to an address outside a valid kernel image. However, if we chain hooks within the same kernel image the amount of possible functions an anti-tampering driver would need to verify increases significantly. The driver is meant to be mapped into the kernel and includes some functionality to hide that a mapper was ever loaded. This includes clearing the PiDDBCacheTable, MmUnloadedDrivers, and g_KernelHashBucketList. Strings in the driver and user-mode client are encrypted at compile time using skCrypter ## Showcase The following images show the process memory map of each setting after the DLL is mapped. The memory regions of the mapped DLL are highlighted.
|
|
|
|
|