RageGuard - Serversided/Clientsided Antihack Base[C++]
Hello, I thought I might as well make a nice project for the learning experience. I'll be starting development on a full open-source antihack in C++/Erlang. The idea of the project will be to create a good base for an antihack, using the most professional, efficient, and secure practices possible. I do not wish to adapt the project's libraries unless it benefits these 3 reasons, regardless of the community's experience with these libraries.
The warning for using this project is:
This project is NOT an anti-hack, it is an anti-hack base. If you use this alone/without modifications, it WILL be exploited. It is an educational project to create a very efficient anti-hack base for other servers to build upon/customize.
The first step will be a packet encryption method that will not increase latency significantly. I will be working on this tonight and I will post some code/info on the method I would like to use in the morning.
I'm not familiar with google repositories, but I am familiar with SVN repositories. I would appreciate it if someone could create a google repo, but I will do it if needed.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
thx mootie it will help us to start understand how to build a anti hack.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
you should just put them up on codeplex or sourceforge. I can't stand the way google does their svn, but if the people want one, they can do it. I'm not well versed in the google repository much either.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
I'll send you a message on MSN with some info.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
Sorry, I got a bit distracted with an exchange event. Anyways, I just finished the base for the packet crypt. I decided to write a simple one to start off with, this shouldn't hinder performance in the least.
Code:
#include "StdAfx.h"
#include "Crypt.h"
#define CY_CIPHERMAX 256
DWORD CEncrypt::m_dwCipher[CY_CIPHERMAX] =
{
0x8B8624FB , 0x2AA8EAE3 , 0xE6AB92E1 , 0x800AAF82 ,
0xEEFBB098 , 0xEE69BABA , 0x16B2EE28 , 0x3F6961B9 ,
0x9DA62F6B , 0x1BE40D43 , 0xB813230A , 0xE094FD3D ,
0xDF8F0EEE , 0xAF806231 , 0x3E1829EF , 0xEA02EE36 ,
0x3D1EDE9A , 0xE1838DFA , 0x1232860F , 0xF83E688E ,
0x2E3A09F0 , 0x6A86461E , 0xE0F03422 , 0x899A400A ,
0x3A3E2A8F , 0x9ED1E8FE , 0xEDD83BEF , 0x0FE4F6BE ,
0xD99DBAEE , 0x088E432F , 0xA182E18F , 0x0E32FE0A ,
0x94D21931 , 0x3E8F8A29 , 0xE4F6B08A , 0xFF6BB063 ,
0xE286FA68 , 0x98816A83 , 0xFF3E3F38 , 0x69D9B881 ,
0x6E00BABF , 0xB2DB02A6 , 0x69ABDA91 , 0xE402938E ,
0x68EAE8E8 , 0x0088A92A , 0xBAAE2FEE , 0xA88FA8EB ,
0x80B4FBF8 , 0x1D8AEA6A , 0x4BDA9FA3 , 0x8EBA9088 ,
0x4EDF6BEA , 0x8E691208 , 0xE394BF6E , 0xDD24EEB9 ,
0x12E0DEFE , 0x8DE2D46B , 0x9A8B0880 , 0x2D6EB6F6 ,
0x8E8EAEE0 , 0x916BA423 , 0x4913F19A , 0xEAD8F0E9 ,
0x119382EE , 0x82AEFBB3 , 0x4AB81318 , 0xA21008B9 ,
0x3A48A3A8 , 0xA9AD8094 , 0x2E09D49D , 0xD68E8E2D ,
0x6EDD8886 , 0x8D064B4D , 0x81AAEEAE , 0x0EE4AEBE ,
0x38E2F9E6 , 0xA69F3848 , 0x6E8A898E , 0xBB3FD1E4 ,
0x289E4EAE , 0xAE9FA41B , 0xE8B11B9A , 0x8BA60FFA ,
0xADEA1011 , 0x88AF2230 , 0x0A8E8EDA , 0xA8F1AFAE ,
0x4FDAB6BD , 0x823A3A2E , 0xAEA11988 , 0x8038BAA3 ,
0xA344D908 , 0x80EF4A6A , 0x18084812 , 0x14B34ADE ,
0xE39ADF8A , 0x0EDEE381 , 0xEDE9B92A , 0x99F99E42 ,
0xB64A33EA , 0xFDEE8AE8 , 0x8D1BBD28 , 0x98A4284D ,
0xB4EE8A22 , 0xAEA8B819 , 0x8E88FA9A , 0x88BA9E44 ,
0x46D24E48 , 0x8F0EF4AA , 0x83A6DAF3 , 0xF23432F1 ,
0xF3BEE688 , 0x201BB6F3 , 0x88DEDEE6 , 0x6E89B2AA ,
0x43E4FA80 , 0xFA398AAE , 0x98B8818E , 0x4103A8E8 ,
0xAF02E2A3 , 0x8BE0DE22 , 0xF2AB6FBD , 0x2648F233 ,
0xAEAD4A84 , 0x848633E6 , 0x2841EDE1 , 0xA8D1E113 ,
0xE8DF819E , 0xDAA41F83 , 0xB80083A0 , 0xE8166B1F ,
0xA69E3683 , 0x969208A1 , 0x2261EAF1 , 0x8EAF4A91 ,
0x3F33D041 , 0x0F28A938 , 0xFEDE69FD , 0x49E4FE2F ,
0x9B36D6EE , 0xADEDFA4F , 0x8DF988DE , 0x8EAEE682 ,
0x121EE1B1 , 0x18EA0E2F , 0x8D3FD036 , 0xB2E0A981 ,
0x2D881833 , 0xB9FA1F81 , 0x8418E8A2 , 0x8088AE86 ,
0x83A183EE , 0x089EB98B , 0xAB2AE83A , 0x00AEAE3A ,
0x1DFBFA44 , 0x9B3180B8 , 0x89EEB9E4 , 0x193490AE ,
0xF28B86D3 , 0xEE8E8ED9 , 0xDA8EA209 , 0x348AAE11 ,
0xAA19AE20 , 0x62AABB4D , 0x1BEB4B11 , 0xE81F9B28 ,
0x0DEE8A98 , 0x961EAD9B , 0x9FA40E94 , 0x4D62944D ,
0x24EDA3E8 , 0x1EB23E29 , 0x0B90AE09 , 0x4BF9EFFA ,
0xA6B4D012 , 0x34EE8690 , 0xE8E818EA , 0x4B4BD4BD ,
0x4BF9A92A , 0xA00AEB4A , 0x8AEE83BD , 0xB4E22E18 ,
0x4E24FA8A , 0xE9ADEAD8 , 0x49DEB6FB , 0x2E343EA0 ,
0xE0ADE148 , 0x1A80B2B9 , 0x11AFE92B , 0x02EA936E ,
0xAF1BEF6D , 0x22AB2263 , 0x394B04EA , 0xEA2EE166 ,
0xB1EA0931 , 0xA3A6A18E , 0x38B9E068 , 0x38B23204 ,
0x0F99283E , 0xA149B622 , 0x8A41F46E , 0x8814AEAF ,
0xF1A9B4E6 , 0xB4FEEA46 , 0xBDABB86A , 0x8EB9FBEF ,
0x0D9F33E6 , 0xA4268434 , 0x32EE98E1 , 0x181A432E ,
0xADE21F88 , 0xDA40FEA2 , 0x9FB3122B , 0x4EAFDE0E ,
0x681BEF1B , 0xBDF22A99 , 0x8AE3EAFF , 0xA10032FE ,
0xB6A21B80 , 0xD6A48AF1 , 0xEDA32DAA , 0x2FA8E96F ,
0x201F3E2E , 0xBEDD1A21 , 0xD04EABD9 , 0xDE088AFF ,
0x3DEAE986 , 0xEA0ADBB3 , 0x9AEBD9FD , 0xEAAA0E14 ,
0x86AA3B0A , 0x9FB8230D , 0xB3812F8D , 0x010A9E46 ,
0x01B90848 , 0x9DA989A8 , 0xA3992600 , 0x98BE8E0E ,
0xE8BDEAAE , 0x88A2682E , 0xF3EB48FD , 0x291DF6DF ,
0x430FF8E4 , 0x49DA80F0 , 0x2AEF28DD , 0x2AB1D33A ,
0xD8460B01 , 0xF68BE18E , 0xA0FD1D28 , 0x1A20AABE ,
0xE1DE8A0E , 0xE94B40AE , 0xFEBEB8A6 , 0x90E364B6
};
//========================================================================
//CEncrypt
//========================================================================
CEncrypt::CEncrypt(void)
{
}
CEncrypt::~CEncrypt(void)
{
}
void CEncrypt::Encrypt( LPVOID pDestination , LPCVOID pSource , DWORD dwSize , DWORD dwCipher )
{
/* Getting the Quotient and the Remainder of the data size */
UINT nQuo = dwSize / 4;
UINT nRem = dwSize % 4;
/* Casting the source data and the destination buffer as DWORD pointers */
DWORD* pdwSource = (DWORD*)pSource;
DWORD* pdwDestination = (DWORD*)pDestination;
/* Casting the unique cipher(should vary for different functions) */
LPBYTE pBYTE = (LPBYTE)&dwCipher;
/* XORing the segments of the cipher together to create a pseudo-random index */
DWORD index = (DWORD)( pBYTE[0] ^ pBYTE[1] ^ pBYTE[2] ^ pBYTE[3] );
/* XORing the main data */
UINT i;
for( i = 0 ; i < nQuo ; ++i )
{
/* XORing the source data with one of the 256 DWORDs above[using the modded the index(created using the original cipher provided)] */
/* Then, XORing the data the with the original Cipher provided */
pdwDestination[i] = pdwSource[i] ^ m_dwCipher[ ( i + index ) % CY_CIPHERMAX ] ^ dwCipher;
}
/* XORing the remainder of the data using the same method */
if( nRem == 1 )
{
( (LPBYTE)&pdwDestination[i] )[0] = ( (LPBYTE)&pdwSource[i] )[0] ^ ( (LPBYTE)&m_dwCipher[ ( i + index ) % CY_CIPHERMAX ] )[0];
}
else if( nRem == 2 )
{
( (LPBYTE)&pdwDestination[i] )[0] = ( (LPBYTE)&pdwSource[i] )[0] ^ ( (LPBYTE)&m_dwCipher[ ( i + index ) % CY_CIPHERMAX ] )[0];
( (LPBYTE)&pdwDestination[i] )[1] = ( (LPBYTE)&pdwSource[i] )[1] ^ ( (LPBYTE)&m_dwCipher[ ( i + index ) % CY_CIPHERMAX ] )[1];
}
else if( nRem == 3 )
{
( (LPBYTE)&pdwDestination[i] )[0] = ( (LPBYTE)&pdwSource[i] )[0] ^ ( (LPBYTE)&m_dwCipher[ ( i + index ) % CY_CIPHERMAX ] )[0];
( (LPBYTE)&pdwDestination[i] )[1] = ( (LPBYTE)&pdwSource[i] )[1] ^ ( (LPBYTE)&m_dwCipher[ ( i + index ) % CY_CIPHERMAX ] )[1];
( (LPBYTE)&pdwDestination[i] )[2] = ( (LPBYTE)&pdwSource[i] )[2] ^ ( (LPBYTE)&m_dwCipher[ ( i + index ) % CY_CIPHERMAX ] )[2];
}
}
DWORD* CEncrypt::GetCipher()
{
return m_dwCipher;
}
If anyone could get a SVN up, I would appreciate it. I'll get one up tonight if no one else does :S
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
:P maybe combine svn's ? 1 anticheat in c++ and 1 in c# would that be an idea xP?
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
SVN: http://svn3.xp-dev.com/svn/rageguard/
Trac: RageGuard
If you wanna be a member, ask me.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
Your encrypt function is kinda weird.. First of all you're using a function to encrypt any kind of data, which implies you always will have to know what kind of data it is and how big is the variable. This could be useful if you just want to decrypt a part of the packet (perhaps to do a pre-check of an eventual crc routine) but is also completely doable by treating the packet as a byte array.
However if your intention is to simply encrypt/decrypt packets you could treat a packet as a raw array of bytes and let the client handle the variable types.
Why are you using an exchange table with dword? This will just make it harder to apply it to variables with sizes less than 4. Use something like the AES S-Table that uses bytes.
Another point is that you are XoRing too much (XoR over XoR over XoR).. Try to explore other operations such as the bitwise operations, that will make the cipher much more secure and less reversible ^^.
Oh yeah, one more thing.. Why not simply code this into the source? Its not like its that hard and it would consume much less processing power.. ._.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
Quote:
Originally Posted by
.dark.
Why not simply code this into the source? Its not like its that hard and it would consume much less processing power.. ._.
It can be coded into the source with minimal effort (in theory), but the reason many use outside .DLLs or executables is because its easier to manage it as a standalone rather than integrated.
Both ways have their pros and cons, it eventually up to the programmer.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
Quote:
Originally Posted by
.dark.
Your encrypt function is kinda weird.. First of all you're using a function to encrypt any kind of data, which implies you always will have to know what kind of data it is and how big is the variable. This could be useful if you just want to decrypt a part of the packet (perhaps to do a pre-check of an eventual crc routine) but is also completely doable by treating the packet as a byte array.
However if your intention is to simply encrypt/decrypt packets you could treat a packet as a raw array of bytes and let the client handle the variable types.
Why are you using an exchange table with dword? This will just make it harder to apply it to variables with sizes less than 4. Use something like the AES S-Table that uses bytes.
Another point is that you are XoRing too much (XoR over XoR over XoR).. Try to explore other operations such as the bitwise operations, that will make the cipher much more secure and less reversible ^^.
Oh yeah, one more thing.. Why not simply code this into the source? Its not like its that hard and it would consume much less processing power.. ._.
I like how you say that "I always need to know what kind of data it is". It shows how little you know about C++ data storage(buffers). Any data can be cast into a buffer, its simple to get the size of it as well. I'd like to see an encryption method that encrypts integers/arrays(completely useless lol). XORing is pretty much the quickest bitwise function due to the uses symmetrical encryption.
Adding more bitwise functions won't "make it harder to reverse" lol. You forget, this is an open source project. It's a base, you're supposed to add onto it. Also, this crypt is fine for the purpose. Also...You do realize it handles remainders -_-. A DWORD cipher table will not hinder its ability to encrypt data less than 4 bytes long.
It is coded into the source. You're obviously going to need to modify the source to use any anti-hack. Why not add some features in as well as work on the external module? Please think out your argument and actually look at the code before you start criticizing it for what you do not understand.
---------- Post added at 11:17 PM ---------- Previous post was at 11:16 PM ----------
Quote:
Originally Posted by
MisterKid
:P maybe combine svn's ? 1 anticheat in c++ and 1 in c# would that be an idea xP?
Why? There is no reason to code anything in C# over C++ for this project.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
Quote:
Originally Posted by
mootie
I like how you say that "I always need to know what kind of data it is". It shows how little you know about C++ data storage(buffers). Any data can be cast into a buffer, its simple to get the size of it as well. I'd like to see an encryption method that encrypts integers/arrays(completely useless lol). XORing is pretty much the quickest bitwise function due to the uses symmetrical encryption.
Adding more bitwise functions won't "make it harder to reverse" lol. You forget, this is an open source project. It's a base, you're supposed to add onto it. Also, this crypt is fine for the purpose. Also...You do realize it handles remainders -_-. A DWORD cipher table will not hinder its ability to encrypt data less than 4 bytes long.
It is coded into the source. You're obviously going to need to modify the source to use any anti-hack. Why not add some features in as well as work on the external module? Please think out your argument and actually look at the code before you start criticizing it for what you do not understand.
[...]
I think you are not getting what .dark meant with saying 'DWORD cipher tables are harder to apply'.
They really don't hinder any ability to encrypt data but by using them you are contradicting your philosophy of keeping the code as fast and efficient as possible.
You need to cast several times, you create new variables, thus adding more memory to be allocated and freed. In addition to that for keeping the ability to encrypt data less than 4 bytes long you need to add 3 if-checks with more or less static code. Remind that nQuo and nRem also have to be calculated.
So in the end using a DWORD table doesn't hinder you, it just makes the code more complicated, bigger and less efficient.
Btw, I don't understant you insulting .dark he wouldn't know about what he talks after he gave you tips and really constructive criticism. Even if he is not right or doesn't share the same opinion (I don't say he wouldn't be), insulting people who try to help and appearently know at least a decent amount about the topic won't help the packetfilter become a community project benefiting of being open source.
Anyways, I think it's a great idea to do the basics of a packetfilter for servers to build an own pf, especially now as no developer really works on an up-to-date version of a packetfilter.
PS:
Did you test the code you wrote here?
As you define ' i ' only in your for-loop you shouldn't be able to use it later on in the if clauses dealing with remainers.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
Quote:
Originally Posted by
Tom94
Did you test the code you wrote here?
As you define ' i ' only in your for-loop you shouldn't be able to use it later on in the if clauses dealing with remainers.
Ah, you're right. I didn't compile it until after I posted. Feel free to modify it if you wish to make it more efficient/secure, but as I see, its a pretty fast script.
Re: RageGuard - Serversided/Clientsided Antihack Base[C++]
The idea behind open source is more than one contribution. You have to learn to listen as well as doing. This project can be a good jumping off point for anyone to learn to make an antihack.
If you aren't willing to listen to some people who may actually know a thing or two, then you are going to have people just waiting around for you to finish it alone, and then what do they learn?
People please, for me, do not start flaming this thread or any other one for that matter. It serves no purpose other than making yourselves look very childish.
