The question that might come naturally after reading this title is:
Why should you know how to code a shellcode runner for your malware analysis?
Malware analysis is a dark art. Every time you conduct a routine analysis, you get a surprise.
Often, you need to extract hidden shell codes and analyze them separately.
While analyzing it, we might meet some unexpected conditions that complicate the understanding of the malware.
In such cases, you can find it easier to execute the shellcode in a separate thread.
In summary, the project will be a simple runner that:
takes a shellcode dump as input
runs it in a separate thread
stops execution by telling us where to set our breakpoint
For more details, you can look at this post on How to embed a shellcode payload into an executable.
What is the Shellcode?
As a quick refresh, I just want to explain in a few words what is a shellcode.
The term “shellcode” refers to machine code used either as the payload in the exploitation of a software vulnerability or, in this case, as a code to inject into a running process.
It has this cool and scary name because it typically starts a command shell from which the attacker can control the compromised machine, but any machine code can be used as the payload.
Typically, the payload is in the form of a binary piece of executable code, which is injected into the target machine’s process, making sure that through certain techniques it is then executed.
Some of the main requirements are:
It has to be very small
It has to avoid some “bad characters”
It has to be position independent
They are the consequence of the fact that the attacker doesn’t know the exact memory region where the victim system puts the payload.
Now, after this quick refresh, we are ready to start coding the shellcode runner!
The preliminary part
You are here for one reason, and you surely want to dive into and start to code it.
But before you can do that, we need to set up everything.
First of all, we’re going to write the required imports and definitions
Anyway, this is a lot easier said than done.
So let’s start getting our hands dirty with the code now!
#include <windows.h>
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#define MAX_SHELLCODE_LEN 1024
Main method
Then let’s look inside the main function step by step
HANDLE threadHandle;
size_t shellcode_len;
LPVOID shellcode_alloc_mem;
unsigned char shellcode[MAX_SHELLCODE_LEN];
HANDLE h_file = CreateFile(argv[1], GENERIC_READ ,0,NULL,OPEN_EXISTING,FILE_ATTRIBUTE_NORMAL,0);
BOOL retval = ReadFile(h_file, shellcode, MAX_SHELLCODE_LEN, &shellcode_len, NULL);
The above lines are extremely simple to explain, they:
Declare the variables we will need in the following steps
Read the file containing the shellcode
Pass it via prompt (argv[1]) and save it as an array of unsigned chars.
Since this is only for our analysis and understanding, we don’t need to do any checks on the input.
shellcode_alloc_mem = VirtualAlloc(0, shellcode_len, MEM_COMMIT | MEM_RESERVE, PAGE_EXECUTE_READWRITE);
RtlMoveMemory(shellcode_alloc_mem, shellcode, shellcode_len);
At this point, a memory area with all permissions is allocated where we’re going to move the shellcode with the help of the RtlMoveMemory function.
printf("%-20s : 0x%-16p\n ", "the shellcode payload addr, put a breakpoint here and then press a key",
(void *)shellcode_alloc_mem);
printf("Press a key to continue:\n");
getchar();
This is part of the code that indicates the area of memory where the shellcode resides and stops execution while waiting for a key press event.
You need to take note of this address because this is exactly where we are when we need to attach the debugger to our runner.
In other words, the runner stops its execution at this point and we need to attach the debugger to the process putting the breakpoints we need for our analysis.
if(shellcode_alloc_mem){
threadHandle = CreateThread(0, 0, (LPTHREAD_START_ROUTINE)shellcode_alloc_mem, 0, 0, 0);
WaitForSingleObject(threadHandle, INFINITE);
}
And finally, here is the part where we launch a new thread containing the shellcode and then wait for the end with the WaitForSingleObject method.
Put all together
At this point, to get a clearer picture, let’s see the full code.
#include <windows.h>
#include <stdio.h>
#include <string.h>
#include <stdlib.h>
#define MAX_SHELLCODE_LEN 1024
int main(int argc, char *argv[]) {
HANDLE threadHandle;
size_t shellcode_len;
LPVOID shellcode_alloc_mem;
unsigned char shellcode[MAX_SHELLCODE_LEN];
HANDLE h_file = CreateFile(argv[1], GENERIC_READ ,0,NULL,OPEN_EXISTING,FILE_ATTRIBUTE_NORMAL,0);
BOOL retval = ReadFile(h_file, shellcode, MAX_SHELLCODE_LEN, &shellcode_len, NULL);
shellcode_alloc_mem = VirtualAlloc(0, shellcode_len, MEM_COMMIT | MEM_RESERVE, PAGE_EXECUTE_READWRITE);
RtlMoveMemory(shellcode_alloc_mem, shellcode, shellcode_len);
printf("%-20s : 0x%-16p\n ", "the shellcode payload addr, put a breakpoint here and then press a key", (void *) shellcode_alloc_mem);
printf("Press a key to continue:\n");
getchar();
if(shellcode_alloc_mem){
threadHandle = CreateThread(0, 0, (LPTHREAD_START_ROUTINE)shellcode_alloc_mem, 0, 0, 0);
WaitForSingleObject(threadHandle, INFINITE);
}
}
We’re finally at the end, now all we have to do is compile.
Open NativeTools and go to the folder containing our file, which we will call “shellcode_runner.cpp“, then run the following command:
cl.exe /Tcshellcode_runner.cpp /link /OUT:shellcode_runner.exe
And now, assuming we have saved our shellcode in a file called shellcode.bin in the same folder as our program, let’s try to launch the runner with the following command.
If the folder is different, you just need to specify the path as a command line argument.
shellcode_runner.exe shellcode.bin
I hope the post was useful and that you enjoyed it, in the meantime, I send you goodbye to the next article!
Further readings
Now that you know how to code a shellcode runner for your malware analysis I suggest you do some readings: