Hi,guys

 

Do you know any hook api library with its open source code ?

 

I have tried mhook library,not quiet good with named 'RtlAllocateHeap' api.

 

So I am wandering if you guys may know some better library that I can use it to hook that api.

Currently I am re-coding a plugin,you may see it within Christmas

 

 

Thanks in advance.

I personally use Detours 3.0 in most of my projects since it is free. The 32bit source code is included with the project as well. 

 

Some others are:

https://github.com/TsudaKageyu/minhook

https://code.google.com/p/ihook/

http://newgre.net/ncodehook

 

There are also a ton of homebrewed implementations you can find on various game hacking websites like UnknownCheats, MPGH, and so on.

Hi, atom0s

 

 

Detours library use RtlAllocateheap api to allocate memory for a temp bridge to get orignal command for its hook routine. So it may cause stack overflow.

 

ihook use malloc api to allocate,so it is the same situation as RtlAllocateheap do.

 

Hope that minhook and ncodehook will be different.  

Thanks for your reply.

You could always adjust the code to use a different allocation API that would prevent that issue as well.

writing your own is easy....

https://bitbucket.org/NtQuery/scyllahide/src/540ec6e2d29b9e9fd4355c7a9c7e3a1099da298a/HookLibrary/Hook.cpp?at=master

@atom0s

em,currently I use VirtualAlloc to allocate a larger enough memory,write some self code for a temp use. not quiet perfect but good .

 

@Aguila

 

'0xE9 jmp' is not well for those special target,re-redirection command.I use '0x68' and '0xC3' for a temp hook.

Thanks for your pieces of code, got some better overview.

I don't see any difference between E9 and 68, except that 68 needs one byte more.

ff 25 xx xx xx xx works nicely too, and stops some code that 'morphs' api code into its own.. like oh.. starforce and stuff...

ah you mean position independent jump code. yeh that is a small advantage, but typically people check for any api modifications.

@evlncrn8

em, FF25 is very nice for my case,

 

@Aguila

yes, re-redirection is annoying part of its anti dump. For its second  redirection command,it use memcpy api to copy the opcode.

So if I use jmp code, it may point to somewhere unknown.That's the most bad thing of it.

In most cases anti-cheats / anti-tampers and such will check at least the first 5-6 bytes of an API to determine if its been altered. Since most API start with the same stub of:

mov EDI, EDI,

push EBP

mov EBP, ESP

 

That being said, sometimes you will need to do a mid-function hook instead to bypass the checks. Unless of course they take snapshots of the full API and its branches and do comparisons that way.

It depends on how thorough the detections are.

@atoms

this situation happens after its all check thread started.

My point is hook that api before all that thread.So that will not happen in my case.

The code will restore all hooked api when some special routine done.

@atom0s Are you have Detours 3.0 full version? 

Edited February 4, 20169 yr by Perplex

I use the free version. http://research.microsoft.com/en-us/projects/detours/

@atom0s The free version has many limitations, Even the 32-bit version is limited.

Quote

Detours Express 3.0 is available for immediate download under a no-fee, click-through license for research, non-commercial, and non-production use.  Detours Express is limited to 32-bit processes on x86 processors.

 

1 minute ago, Perplex said:

@atom0s The free version has many limitations, Even the 32-bit version is limited.

 

For me, it fits my needs. I don't really do much with 64bit targets so it is not a concern of mine.

Usually I use hardware breakpoints in my injected libraries. Memory breakpoints also work but I never implemented it. Obviously this doesn't work if your target screws with hardware breakpoints. You can also place an int3 byte and do things that way.

https://bitbucket.org/mrexodia/enigmahwid/src/b3eeee1772c369b1a37b7dfa68d5a4b7d9589ec0/main.cpp?at=master&fileviewer=file-view-default