SLAE ASSIGNMENT 7 | Custom Crypter for our execve-SHELLCODE - LINUX X86

SLAE ASSIGNMENT 7 | Custom Crypter for our execve-SHELLCODE - LINUX X86

June 02, 2020   assembly, crypter, pentester academy, reverse engineering, security tube, shellcode, SLAE, SLAE x86,

This blog post has been created for completing the requirements of SecurityTube Linux Assembly Expert Certification: http://securitytube-training.com/online-courses/securitytube-linux-assembly-expert/ Student ID: SLAE - 1342

Hello Shellcoders,
     It has been two week since I wrote my last article. Well if you haven’t seen my Assignment 6 blog post on SLAE ASSIGNMENT 6 | POLYMORPHIC SHELLCODE'S - LINUX X86, then you can check it [Here]. Today in this post we will be writing the Crypter which will encrypt our execve-stack x86 Shellcode which is the final assignment for the SLAE exam.
So the requirement is:

1. Create a custom crypter like the one shown in the "crypters" video.
2. Free to use any existing encryption schema.
3. Can use any programming language.

I will be using AES CBC Mode Encryption to encrypt the shellcode. I will be implementing the crypter in python3.

If you have any doubt, then please feel free to ping me on twitter [@Pwsecspirit].

Here is the python code of our Crypter:

Let's run it and see the help menu.

So, we can use -e or --encrypt to encrypt the shellcode. Let's encrypt it.

Amazing we have our encrypted shellcode data string. Let pass it to the decrypt option.

Amazing it decrypted and executed the shellcode as expected. I believe that's it for this assignment.

Thanks,

Continue Reading

SLAE ASSIGNMENT 6 | POLYMORPHIC SHELLCODE'S - LINUX X86

SLAE ASSIGNMENT 6 | POLYMORPHIC SHELLCODE'S - LINUX X86

May 15, 2020   assembly, pentester academy, polymorphic shellcode, reverse engineering, security tube, shell-stormshell code, shellcode, SLAE, SLAE x86,

This blog post has been created for completing the requirements of SecurityTube Linux Assembly Expert Certification: http://securitytube-training.com/online-courses/securitytube-linux-assembly-expert/ Student ID: SLAE - 1342

Hello Shellcoders,
     Welcome to Pwsec land. Well if you haven’t read my last blog post in which I completed writing my SLAE Assignment 5 blog post on “METASPLOIT SHELLCODE'S ANALYSIS - LINUX X86”, then you can check it [Here]. Today in this post we will be creating three Polymorphic Linux x86 Shellcodes which we will take from the shell-storm website. So the requirement for the sixth SLAE exam assignment is:

1. Take up 3 shellcode from shell-storm and create a polymorphic version of them to beat the pattern matching. 
2. The polymorphic versions can not be larger than 50% of the existing shellcode.
3. Bonus point for making it shorter in length than original

I have chosen these three shellcodes.

• http://shell-storm.org/shellcode/files/shellcode-842.php
• http://shell-storm.org/shellcode/files/shellcode-542.php
• http://shell-storm.org/shellcode/files/shellcode-758.php

So I have selected these three shellcodes for the completion of this assignment.

Let’s first understand what a polymorphism is by wikipedia :- In computer terminology, polymorphic code is a code that uses a polymorphic engine to mutate while keeping the original algorithm intact. That is, the code changes itself each time it runs, but the function of the code (its semantics) will not change at all. For example, 1+3 and 6-2 both achieve the same result while using different values and operations. This technique is sometimes used by computer viruses, shellcodes and computer worms to hide their presence.

Also I have written comments with my polymorphic version of the shellcode.

Polymorphic shellcode 1

So this is a shellcode which will try to read the first 4096 bytes from /etc/passwd file. The author of this shellcode is geyslan. You can find the shellcode [here].

Let’s run it,

So the original shellcode length is 57 Bytes  currently, and as you can see it printed out the contents of /etc/passwd.

Here is the polymorphic version of this shellcode.

Let's run it,

Now the shellcode length is: 59 Bytes
[Note:Please refer to this video if you also don’t know how to calculate the percentage increase.]

Size Increased by: 3.5%

Polymorphic shellcode 2

So this is a shellcode which will try to create a directory with the name “Hacked”. The author of this shellcode is zillion. You can find the shellcode [here]

Let’s run it,

So the original shellcode length is 36 Bytes currently, and as you can see it just created a new directory with the name “hacked”.

Here is the polymorphic version of this shellcode.

Let’s run it,

Now the shellcode length is: 35 Bytes

Size Decreased by: 2.7%

Polymorphic shellcode 3 

So this is a shellcode which will try to read try to read the file /etc/shadow using /bin/cat command. It will be using execve function to execute this command. The author of this shellcode is antrhacks. You can find the shellcode [here]

Let’s run it,

So the original shellcode length is 42 Bytes currently, and as you can see it printed out the contents of /etc/shadow file.

Here is the polymorphic version of this shellcode.

Let’s run it,

Now the shellcode length is: 51 Bytes

Size Increased by: 21.4%

Pretty easy, huh? If you still have any doubt, then please feel free to ping me on twitter [@Pwsecspirit]

Thanks,

Continue Reading

SLAE ASSIGNMENT 5 | METASPLOIT SHELLCODE'S ANALYSIS - LINUX X86

SLAE ASSIGNMENT 5 | METASPLOIT SHELLCODE'S ANALYSIS - LINUX X86

May 07, 2020   assembly, metasploit shellcode. metasploit shellcode analysis, pentester academy, reverse engineering, security tube, shellcoding, SLAE, SLAE x86,

This blog post has been created for completing the requirements of SecurityTube Linux Assembly
Expert Certification: http://securitytube-training.com/online-courses/securitytube-linux-assembly-expert/
Student ID: SLAE - 1342

Hello Shellcoders,
          Welcome to Pwsec land. So, I think it’s been more than a week and I should take my words back that I said in my last blog post “I guess I get the motivation to complete the assignment only during the weekend.” well it turns out a TV / Web series is enough to ruin this kind of mentality. But nevertheless, in my last blog post I wrote SLAE assignments 4, if you haven’t seen my Assignment 4 blog post on CUSTOM ENCODER/DECODER - LINUX X86, then you can check it [Here]. Today in this post we will be analyzing three Metasploit Linux x86 Shellcodes. Well, I must say before yesterday I had no clue how I am gonna do it. But now I must say, its always worth to give anything a try. So the requirement for the fifth SLAE exam assignment is:

• Take up at least 3 shellcode samples created using Metasploit for Linux/x86. 
• Use GDB/Ndisasm/Libemu to dissect the functionality of the shellcode. 
• Present your analysis.

Ahh!! It seems hard isn’t? Hahaha, I must say it is and it is not. It just depends on you. So, I listed out all the Linux/x86 payloads using msfvenom.

So I have selected these three payloads for the completion of this assignment. Also we will be needing one option of msfvenom command.

Let’s begin with the first one i.e Linux/x86/chmod.

Analyzing linux/x86/chmod shellcode

It has two basic options.

1. The file name/file path itself. 
2. The mode for the file, the default is 666 which read read-write by user-group-other.

So if you have used linux before and used chmod command to make your file executable that is the exact thing we are looking at. But if you don’t know how it works then this might be a good learning lesson for you too. Let’s generate the shellcode

Here is my c code:

I compiled it using gcc

gcc -fno-stack-protector -z execstack shellcode_chmod.c -o shellcode_chmod

First lets check the file permission on “spirit.txt” file.

Right now only cyborg user can write to the file spirit.txt , but let’s see what will happen after running the shellcode.

Ahh! As you can see now cyborg-cyborg group-other have the permission of read-write. Which means now anyone other than cyborg user can read the content of the file too. Let’s open it in GDB and analyze it.

Let’s put the breakpoint on the call instruction. Because it is the one which will call our shellcode.

Now we can run the program using “r” or “run” command.

Alright! Now lets use “stepi” gdb command which is responsible for step into the next machine instruction. Now after it we must be inside but which holds our shellcode.

If you have read my previous articles or know ASM already then you might understand that int 0x80 is responsible for the interrupt call and passes the control to the kernel. Now we are just going to break the first “int 0x80” address and before making the call we will just check the register value from which we can simply analyze.

So currently:

• The value of EAX is 0xf 
• The value of EBX is the pointer to “spirit.txt” string
• The value of ECX is 0x1b6 

Let’s check what 0xf is in decimal

It is 15, now we can see in the syscall list what is the value of 15.

Alright so it is nothing but the _NR_chmod, now we must read the man 2 page of chmod.

Okay! So the chmod function takes two arguments, where the first argument is the file name/path and the second argument is for the mode.

Now let’s check what is 0x1b6 in decimal

It is 438, This is octal for 666. Now all of this make sense, isn’t?

The spirit.txt inside EBX which is the first argument and 0x1b6 inside ECX which is the 2nd argument of chmod function. Also, the next int 0x80 is nothing but just the syscall for the exit. Because the shellcode should exit gracefully haha.

Analyzing linux/x86/read_file shellcode

It has two basic options. 

1. The file descriptor 
2. The path to the file 

Let’s generate the shellcode

Here is my c code:

I compiled it using gcc

gcc -fno-stack-protector -z execstack shellcode_read_file.c -o shellcode_read_file

Let’s open it in GDB and analyze it.

Let’s put the breakpoint on the call instruction. Because it is the one which will call our shellcode.

Now we can run the program using “r” or “run” command. After it just use “stepi” gdb command which is responsible for step into the next machine instruction. Then after it we must be inside buf which holds our shellcode.

So this tie we have total of four “int 0x80” instruction. Let’s put a breakpoint on each one of it.

Also if you are lazy, you can use gdb command “rununtil int” , then it will run until it see the int instruction. But I am good to go with the breakpoint.

[Note:If you have read my previous articles or know ASM already then you might understand that int 0x80 is responsible for the interrupt call and passes the control to the kernel. Now we are just going to break the first “int 0x80” address and before making the call we will just check the register value from which we can simply analyze.]

Let’s continue the program and check the register value.

So currently:

• The value of EAX is 0x5 which is nothing but 5 in decimal.
• The value of EBX is the pointer to “spirit.txt” string.

Now we can see in the syscall list what is the value of 5.

Alright so it is nothing but the _NR_open, now we must read the man 2 page of open.

Okay! So the open function generally takes two arguments, where the first argument is the path name and the second argument is for the flags. Now all of this make sense, isn’t?

The spirit.txt inside EBX which is the first argument. Now time to analyze the next int 0x80 instruction. By typing “c” or “continue” which will continue the program.

So currently: 

• The value of EAX is 0x3 which is nothing but 3 in decimal.
• The value of EBX is 0x3.
• The value of ECX point to address 0xbfffefcc which holds the buffer.
• The value of EDX is 0x1000.

Let’s see the value of 0x1000 in decimal.

Now we can see in the syscall list what is the value of 3.

Alright so it is nothing but the _NR_read, now we must read the man 2 page of read.

Okay! So the read function takes three arguments, where the first argument is the file descriptor, the second argument holds the value of buffer and the third argument holds the value of the size of the buffer. Now all of this makes sense, isn’t?

Now time to analyze the next int 0x80 instruction. By typing “c” or “continue” which will continue the program.

So currently: 

• The value of EAX is 0x4 which is nothing but 4 in decimal.
• The value of EBX is 0x1.
• The value of ECX point to address 0xbfffefcc which holds the buffer.
• The value of EDX is 0x16

Let’s see the value of 0x16 in decimal

Now we can see in the syscall list what is the value of 4.

Alright so it is nothing but the _NR_write, now we must read the man 2 page of write.

Okay! So the read function takes three arguments, where the first argument is the file descriptor, the second argument holds the value of buffer and the third argument holds the value of the size of the buffer. Now all of this makes sense, isn’t?

As EBX value is 0x1 which is stdout, ECX value contains the buffer inside spirit.txt file and EDX is 0x22 which is 22 in decimal, which is the total length of the buffer.

That is actually how we really do it in any other programming language. If you have coded previously in python etc. then you might know we first open the file with a different mode. Lets say read mode then we read the file content and print it out which is nothing but the write func().

Also, the next int 0x80 for nothing but just the syscall for the exit. Because the shellcode should exit gracefully haha.

Lets run the shellcode

Analyzing Linux/x86/shell_reverse_tcp Shellcode

It has three basic options. 

1. CMD which tells what command to execute once the connection will establish.
2. LHOST which is nothing but the listener IP address.
3. LPORT which holds the value of port number on which the attacker is currently listening.

Let’s generate the shellcode

Here is my c code

I compiled it using gcc

gcc -fno-stack-protector -z execstack shellcode_reverse_tcp.c -o shellcode_reverse

Let’s open it in GDB and analyze it.

Let’s put the breakpoint on the call instruction. Because it is the one which will call our shellcode.

Now we can run the program using “r” or “run” command. After it just use “stepi” gdb command which is responsible for step into the next machine instruction. Then after it we must be inside buf which holds our shellcode.

So this tie we have total of four “int 0x80” instruction. Let’s put a breakpoint on each one of it.

Also if you are lazy, you can use gdb command “rununtil int” , then it will run until it see the int instruction. But I am good to go with the breakpoint.

[Note:If you have read my previous articles or know ASM already then you might understand that int 0x80 is responsible for the interrupt call and passes the control to the kernel. Now we are just going to break the first “int 0x80” address and before making the call we will just check the register value from which we can simply analyze.]

Let’s continue the program and check the register value.

So currently:

• The value of EAX is 0x66.
• The value of EBX is 0x1 which is 1 in decimal.
• The value of ECX is 0xbfffefc0 which point to 0x2

Let us first check what is the value of 0x66 in decimal.

Alright! It is 102, Now we can see in the syscall list what is the value of 102.

Alright so it is nothing but the _NR_socketcall, now we must read the man 2 page of socketcall.

Okay! So the socketcall function generally takes two arguments, where the first argument is the call number. Also again I have already discussed all of this in my previous blog posts. But these call number are the /usr/include/linux/net.h call number for socketcall.

Right now the EBX is 0x1. So currently, the SYS_SOCKET is being called. Now after this instruction there will be a loop.

So in this the loop will keep running till the ECX value became -1. Once ECX will be -1 the SIGN flag will get set which will send the control to next instruction and get out of the loop.

So currently: 

• The value of EAX is 0x3f.
• The value of EBX is 0x3 which is 3 in decimal.
• The value of ECX is 0x2 which is 2 in decimal.

Lets check what 0x3f is in decimal

It is 63, Now we can see in the syscall list what is the value of 63.

So this is syscall for dup2, lets check its man 2 page.

We can see the dup2 generally takes two arguments where the first one is old file descriptor and the second one is for the new file descriptor. To know more just read the description. This will run 3 times ,

• 0-->Standard Input-->STDIN_FILENO-->stdin
• 1-->Standard Output-->STDOUT_FILENO-->stdout
• 2-->Standard Error-->STDERR_FILENO-->stderr

[Ref:https://en.wikipedia.org/wiki/File_descriptor]

Now once we get out of the loop. 

It pushes the IP address and port on the stack and points it to ECX. So ECX will be the pointer that holds the value of IP address and port.

Now time to analyze the next int 0x80 instruction. By typing “c” or “continue” which will continue the program.

So currently:

• The value of EAX is 0x66 which is 102 in decimal.
• The value of EBX is 0x3 which is 3 in decimal.
• The value of ECX point to address 0xbfffefb0.

Well now we know 0x66 is 102 in decimal. Now we can see in the syscall list what is the value of 102.

Okay! So the socketcall function generally takes two arguments, where the first argument is the call number. Also again I have already discussed all of this in my previous blog posts. But these call number are the /usr/include/linux/net.h call number for socketcall.

Right now the EBX is 0x3 which is 3 in decimal, 

so currently, the SYS_CONNECT is being called.

And from here forward everything is similar just like Assignment 2 reverse tcp shellcode. The shell execv function which is responsible to call /bin/sh.

Let's give it a try.

Pretty easy, huh? If you still have any doubt, then please feel free to ping me on twitter [@Pwsecspirit]

Thanks,

Continue Reading