Decoding Malware Payload encoded in a PNG part 2 – “W.H.O.bat”

This post is a sequel to the post covering the sample “Bank Statement.bat.” I had received this message before the Bank Statement message, but I found the sample in the previous post was less obfuscated and easier to reverse engineer.

In this post, I will cover the different ways that this sample hid the decoding routes and how I was able to gather the data to run the same decoding script I used before to extract the payload from the PNG data within this sample.

Detailed Analysis

The metadata between the two samples is different but still tries to represent this .NET compiled binary is from “Apple Inc.” In this dump below, you see that this sample attempts to represent itself as an iTunes Visualizer.

Architecture:     IMAGE_FILE_MACHINE_I386
Subsystem:        IMAGE_SUBSYSTEM_WINDOWS_GUI
Compilation Date: 2020-Apr-16 11:34:55
Comments:         iTunes Visualizer Host
CompanyName:      Apple Inc.
FileDescription:  iTunes Visualizer Host
FileVersion:      4.4.3.0
InternalName:     Vi8BESIfUtQA5qX.exe
LegalCopyright:   © 2000-2020 Apple Inc. All rights reserved.
OriginalFilename: Vi8BESIfUtQA5qX.exe
ProductName:      iTunes Visualizer Host
ProductVersion:   4.4.3.0
Assembly Version: 1.4.0.0

Matching compiler(s):
    Microsoft Visual C# v7.0 / Basic .NET
    .NET executable -> Microsoft

As with the previous sample there is an PNG file embedded in the binary. however the images is 20 pixels larger in each dimension.

DECIMAL       HEXADECIMAL     DESCRIPTION
--------------------------------------------------------------------------------
0             0x0             Microsoft executable, portable (PE)
26921         0x6929          PNG image, 300 x 300, 8-bit/color RGBA, non-interlaced
26999         0x6977          Zlib compressed data, compressed
404804        0x62D44         Copyright string: "CopyrightAttribute"

Visually looking at the PNG data, it looks similar to the PNG data from the Bank Account.bat sample. Seeing this, I started to think I may be able to use the same method I used previously to decode the payload. As a first attempt, I ran the script as-is, and as I expected, it didn’t correctly decode the file. I was already assuming at least that this sample would use a different key.

I started to look at the sample in dnSpy to find the key and the decoding methods in this binary. The first thing I noticed is that this .NET file either had more obfuscation or was just obfuscated differently than the previous binary I investigated. I was able to follow the flow from the entry point to where the sample starts a new process. There is not  whole lot else interesting in the code after this point in the method.

Process execution

After running the sample using the dnSpy debugger to decode the arguments of the Process.Start method call; I found that the sample executes “installUtil.exe” a .NET utility with the /u and the path to the location of the sample.

Decoded method call arguments

Pulling up the documentation for installUtil.exe utility I found the following:

"Installutil.exe uses reflection to inspect the specified assemblies and to find all Installer types that have the System.ComponentModel.RunInstallerAttribute attribute set to true. The tool then executes either the Installer.Install or the Installer.Uninstall method on each instance of the Installer type. Installutil.exe performs installation in a transactional manner; that is, if one of the assemblies fails to install, it rolls back the installations of all other assemblies. Uninstall is not transactional."

https://docs.microsoft.com/en-us/dotnet/framework/tools/installutil-exe-installer-tool

Note: I ran de4dot in between to make life a little easier to parse. It did not note any specific obfuscators.

In short, the /u installUtil.exe option runs the Uninstall method of the binary in the argument, in this case, the sample we are investigating. I searched the sample’s code for “Install” and found the following Uninstall method. This method looks very similar to the method that executed the PNG parsing function on the Bank Account.bat malware. For example, this method has similarly named variables and a similar flow to the PNG decoding method in the other sample.

When attempting to extract the data from the variables and reverse the methods, I found that smethod_0 and other related methods are heavily obfuscated and very hard to analyze statically. I switched to dynamic analysis and executed this sample in dnSpy. I used the following options to run in it using installUtil.exe and set a breakpoint in the Uninstall method.

After running the code, I hit the breakpoint I expected in the Uninstall function. Then I stepped over the “text” and “location” variables having their values assigned, revealing the PNG resources and the password in a similar format to the “Bank Account.bat” sample. Unfortunately, the process crashes when attempting to extract the code that is used to unpack the PNG. This crash is not an issue; I was able to retrieve the data I needed.

Decoded variables in dnSpy

After only changing the extracted PNG file, XOR key, and final PNG pixel data value in the script I created for “Bank Statement.bat” Success, I was able to extract the payload.

The extracted payload looks to be very similar to the Bank Statement.bat payload. They both have the same filename in the metadata “ReZer0V2.exe.” However, some of the metadata is different, indicating they may be different versions of the payload.

Wrap up

My hunch was correct about these samples using the same encoding method for the PNG payload. I still have not reversed the payload yet, but there are some links to other work on this payload in my other post for the Bank Statement.bat sample. I enjoyed working on this sample, the different methods used to hide the decoding routine of the PNG data were a fun challenge.

Sample Download

https://malshare.com/sample.php?action=detail&hash=ad9462489dfac401daf38efb2b5acbbf

IOCs

MD5: ad9462489dfac401daf38efb2b5acbbf
SHA1: ee1b10bf9523d89586f5ba6bf2d44ed0dce5c13a
SHA256: e161ec8af4ae4b055ca4cd2f405c041f643894f403f35bc3cbc25064328682ef

Full Decode Script

import png
import struct

def print_list(thelist, quantity):

	if (quantity == 0):
		quantity = len(thelist)

	#print ("Decimal: ", end="")
	#for i in range(0, quantity):
	#	print (thelist[i], ", ", end="")
	#print ("")
	print ("Hex: ", end="")
	for i in range(0, quantity):
		print (hex(thelist[i]), ", ", end="")
	print ("")
	print ("ASCII: ", end="")
	for i in range(0, quantity):
		print (chr(thelist[i]), ", ", end="")
	print ("")

############

filename = "be8ff-2.bmp"
# 0x0 , 0x58 , 0x0 , 0x41 , 0x0 , 0x64 , 0x0 , 0x67 , 0x0 , 0x57 , 0x0 , 0x6b , 0x0 , 0x4b
plain_key = "EMe2A6he"
key = bytearray(plain_key.encode("utf-16be"))
print_len = 50
print ("Key: ")
print_list(key, 0)
print ("Key len: ", len(key))

##
#### Load PNG
bmp_full_data = png.Reader(filename=filename).read()
bmp_img_data = list(bmp_full_data[2])

##
#### Reverse Start
data_array = []
output_array = []

print ("")
print ("Loading Image data")
print ("IMG height: ", len(bmp_img_data))
row_count = 0
#print ("Row: ", i, len(bmp_img_data[i]), end='')
while (row_count < len(bmp_img_data[0])-4):
	#print (row_count, " ", end="")
		
	for i in range(0,len(bmp_img_data)):
		# AARRGGBB
		R = bmp_img_data[i][row_count]    # 05
		G = bmp_img_data[i][row_count+1]  # 16
		B = bmp_img_data[i][row_count+2]  # 01
		A = bmp_img_data[i][row_count+3]  # 00

		data_array.append(B) 
		data_array.append(G)
		data_array.append(R)
		data_array.append(A)

	row_count += 4

	#print (".. row loaded")

print ("1st bytes")
print_list(data_array[:4], 4)
first_bytes_value = struct.unpack("<I", bytearray(data_array[0:4]))[0]
decode_data = data_array[4:]

##
#### XOR_DEC Start
key_counter = 0

print ("Data Length: ", len(data_array))

print ("")
print ("Pre XORed data")
print_list(decode_data, 50)

outfile = open ("test-prexor.bin", 'wb')
outfile.write(bytearray(decode_data))
outfile.close()

print ("")
print ("XORing Image data")

# below is either B5 or 00 ^ 112
#lastdata = 0x67
static_xor_val = 0x67
lastdata = decode_data[len(decode_data)-1]
key_modifier = lastdata ^ static_xor_val
#key_modifier = 0xb5 ^ 112  # 0xc5
#key_modifier = 0

print("key: ", lastdata, " len: ", len(decode_data)-1, "found key: ", hex(decode_data[len(decode_data)-1]), " mod key: ", key_modifier)

key_counter = 0

for xor_i in range(0,len(decode_data)):

	key_value = key[key_counter]
	output_array.append(decode_data[xor_i] ^ key_modifier ^ key_value)

	if (key_counter < len(plain_key)-1):
		key_counter += 1
	else:
		key_counter = 0

print ("Final Output")
print_list(output_array, print_len)

outfile = open ("test-postxor.bin", 'wb')
outfile.write(bytearray(output_array))
outfile.close()

Decoding Malware Payload encoded in a PNG – “Bank Statement.bat”

When looking through my Spam folder, I have run across a few messages with “.bat” files attached to them. Most messages have had different content in the message to entice a victim to open the attachment. I started to investigate each of the attachments and found they were Windows Binaries, and at least two had PNG files in the resources. After doing this initial triage, I wanted to see if the payload of these pieces of malware is encoded in this PNG data and how it was encoded.

Initial spam message and quick look at the attachment.

I started with a sample named “Bank Statement.bat” with the .NET code that is the least obfuscated and will visit another sample in a later post. In this post, I will reverse engineer the .NET code and uncover the process to extract out the payload encoded in a PNG file embedded in the binary.

Detailed Analysis

First thing, I took a look at the properties of the attached file and determined it was a .NET compiled binary with some suspicious properties such as having a copyright field listing “Apple, Inc.” Some more of the metadata details are shown below.

Architecture:     IMAGE_FILE_MACHINE_I386
Subsystem:        IMAGE_SUBSYSTEM_WINDOWS_GUI
Compilation Date: 2020-Apr-20 14:27:35
Comments:         QuartzCore 227
CompanyName:      Apple Inc
FileDescription:  QuartzCore
FileVersion:      3.0.0.0
InternalName:     Ly2kW4nOksU0vgv.exe
LegalCopyright:   © 2020 Apple Inc. All rights reserved.
OriginalFilename: Ly2kW4nOksU0vgv.exe
ProductName:      QuartzCore
ProductVersion:   3.0.0.0
Assembly Version: 5.4.1.0

Matching compiler(s):
    Microsoft Visual C# v7.0 / Basic .NET
    Microsoft Visual C++ 8.0
    .NET executable -> Microsoft

Next I ran a binwalk to see if there are is any other obvious hidden content within this file and found there is a PNG file embedded within the binary.

DECIMAL       HEXADECIMAL     DESCRIPTION
--------------------------------------------------------------------------------
0             0x0             Microsoft executable, portable (PE)
19329         0x4B81          PNG image, 290 x 290, 8-bit/color RGBA, non-interlaced
19407         0x4BCF          Zlib compressed data, compressed
357216        0x57360         Copyright string: "CopyrightAttribute"

PNG file screenshot

I opened the file in ilSpy and extracted the PNG file from the resources of the binary. When looking at the extracted PNG file I found visually it looks like encoded data. After seeing this image I started to investigate the original binary file to find routines used to decode the PNG file into what I assumed is the payload of the malware. I started to look at the file further in dnSpy and started at the entry point of the binary.

Entry point

Starting at the entry point method and following the flow through a few more methods, finally finding the start of the decoder functionality. The method below shows the initial routines that load the decoder.

encoded PNG decoder library

The first items I noticed were the variables text and test2 are references to the PNG resource data. The next variable of note is test3 which looks like it could be a password. This method also contains a blob of encoded data (shown in the HexToString() call on line 9) that has various bytes swapped. Once the blob of data is decoded and returned to its original values then transformed into a string that is next decoded from Base64 into is DLL. The DLL when loaded is named CoreFunctions.dll.

After CoreFunctions.dll is loaded the method “CoreFunctions.Main” is executed. There are four parameters passed to this method, the first two references the PNG data, third what looks like a password, and finally the path to the full binary file. These are the variables I made a note of earlier. This method runs a few routines that decode the PNG data. Next, let’s walk through these method calls:

  1. Read_R reads the PNG file resource into a bitmap object.
  2. Reverse creates an array of each column’s BRGA (Blue, Red, Green, Alpha) color values.
  3. XOR_DEC decodes the values using XOR rotating through the key “XAdgWkK” that is XOR’ed against the last byte of the PNG data.

The image below shows the calls to these methods. They are high lighted in red by the breakpoints.

Once the PNG resource data is decoded into its executable binary data, it is loaded and executed in memory without writing any data to disk.

I have written a python script (that is at the end of this post), that recreates the decoding process and takes in the export of the resource’s PNG data and the key to decodes the payload.

Once this process is completed the decoded payload is named “ReZer0V2” in the metadata of the binary data. I have not done much analysis on the main payload yet other than executing the sample in a sandbox. The sandbox run can be viewed at the following Anyrun link:

https://app.any.run/tasks/577824dc-7d69-4551-86df-9892dc48c49e

I may do further analysis of this sample however this appears to be a few posts out there about this payload:

Wrap up

I found this an interesting sample to dissect and understand the method used to encode the PNG data and in the future to see if it can be used to decode a second sample I have with a similarly encoded PNG file. The follow-up post about that sample “W.H.O.bat” will be posted up soon. A theory I have about this sample is that it was sent out prematurely and was not fully obfuscated nor was the phishing content of the message fully completed for the campaign, however, it is just a guess.

Sample Download

https://malshare.com/sample.php?action=detail&hash=09cc3eff1d2d8503722bb195ec45d885

IOCs

SHA256: 9253368d34d7342b7c40c42d2df8a862b55bff9e197b92c18a8cdf46a3279c37
SHA1: 9e104d7c818df8e3c47609852580e3f94eb6be53
MD5: 09cc3eff1d2d8503722bb195ec45d885

Decoding Script

import png
import struct

def print_list(thelist, quantity):

	if (quantity == 0):
		quantity = len(thelist)

	#print ("Decimal: ", end="")
	#for i in range(0, quantity):
	#	print (thelist[i], ", ", end="")
	#print ("")
	print ("Hex: ", end="")
	for i in range(0, quantity):
		print (hex(thelist[i]), ", ", end="")
	print ("")
	print ("ASCII: ", end="")
	for i in range(0, quantity):
		print (chr(thelist[i]), ", ", end="")
	print ("")

############
filename = "79fb5.bmp"
# 0x0 , 0x58 , 0x0 , 0x41 , 0x0 , 0x64 , 0x0 , 0x67 , 0x0 , 0x57 , 0x0 , 0x6b , 0x0 , 0x4b
plain_key = "XAdgWkK"
key = bytearray(plain_key.encode("utf-16be"))
print_len = 50
print ("Key: ")
print_list(key, 0)
print ("Key len: ", len(key))

##
#### Load PNG
bmp_full_data = png.Reader(filename=filename).read()
bmp_img_data = list(bmp_full_data[2])

##
#### Reverse Start
data_array = []
output_array = []

print ("")
print ("Loading Image data")
print ("IMG height: ", len(bmp_img_data))
row_count = 0
#print ("Row: ", i, len(bmp_img_data[i]), end='')
while (row_count < len(bmp_img_data[0])-4):
	#print (row_count, " ", end="")
		
	for i in range(0,len(bmp_img_data)):
		# AARRGGBB
		R = bmp_img_data[i][row_count]    # 05
		G = bmp_img_data[i][row_count+1]  # 16
		B = bmp_img_data[i][row_count+2]  # 01
		A = bmp_img_data[i][row_count+3]  # 00

		data_array.append(B) 
		data_array.append(G)
		data_array.append(R)
		data_array.append(A)

	row_count += 4

	#print (".. row loaded")

print ("1st bytes")
print_list(data_array[:4], 4)
first_bytes_value = struct.unpack("<I", bytearray(data_array[0:4]))[0]
decode_data = data_array[4:]

##
#### XOR_DEC Start
key_counter = 0

print ("Data Length: ", len(data_array))

print ("")
print ("Pre XORed data")
print_list(decode_data, 50)

outfile = open ("test-prexor.bin", 'wb')
outfile.write(bytearray(decode_data))
outfile.close()

print ("")
print ("XORing Image data")

# below is either B5 or 00 ^ 112
key_modifier = 0xb5 ^ 112  # 0xc5
#key_modifier = decode_data[len(decode_data)-1] ^ 112
#key_modifier = 0

print(len(decode_data)-1, hex(decode_data[len(decode_data)-1]), key_modifier)

key_counter = 0

for xor_i in range(0,len(decode_data)):

	key_value = key[key_counter]
	output_array.append(decode_data[xor_i] ^ key_modifier ^ key_value)

	if (key_counter < len(plain_key)-1):
		key_counter += 1
	else:
		key_counter = 0


print ("Final Output")
print_list(output_array, print_len)

outfile = open ("test-postxor.bin", 'wb')
outfile.write(bytearray(output_array))
outfile.close()

Link: Exploring Key Features of Cisco ISE Release 2.6

In July I wrote for the CDW blog about the new version of the Cisco Identity Services Engine (ISE) software.

Exploring Key Features of Cisco ISE Release 2.6

The latest version of this cybersecurity tool offers unique device identification and an IoT protocol.

BSidesNH 2019 Recap

Badge

Back on May 18th, I attended the inaugural BsidesNH event. It was a fantastic one-day event. The day started pretty early for me driving down from Maine arriving at Southern NH University. I arrived to pick up the fantastic badge made out of an old 3.5″ disk. After grabbing some coffee and a snack I settled into the auditorium and for a day of great talks. There were a few that stood out to me from the day that I will talk about.

The second talk of the day was Ghost in the Shell: When AppSec Goes Wrong by Tony Martin. Tony first talked about covered some basics of web application security. He framed these issues around the research he has done into various NAS devices and vulnerabilities he has discovered. Including the ability to create shadow users that have administrative access to devices but are not visible through the administrative interfaces of the device.

After lunch was Chinese and Russian Hacking Communities presented by Winnona DeSombre and Dan Byrnes, Intelligence Analyst from Recorded Future. They covered operations and cultures of Chinese and Russian underground groups. This was a very entertaining presentation and a summary of the information contained in the report: Thieves and Geeks: Russian and Chinese Hacking Communities.

The second to last talk of the day was Hunting for Lateral Movement: Offense, Defense, and Corgis presented by Ryan Nolette. He covered the ways attackers move around and infiltrate further into a network…Corgies. A great quote that stuck with me from his talk was: “If you teach an analyst how to think they will punch above their weight.” I feel this quote not only applies to security analysts but all levels of IT professionals.

BsidesNH was a well run and enjoyable event and a great addition to the Security events in New England. Thanks to all of the organizers and sponsors. I look forward to attending next year!