Approved_users, administrator
366
edits
Changes
Created page with "{{Warningbox|This tutorial has been updated for ChipWhisperer 4.0.0 release. If you are using 3.x.x see the "V3" link in the sidebar.}} {{Infobox tutorial |name..."
{{Warningbox|This tutorial has been updated for ChipWhisperer 4.0.0 release. If you are using 3.x.x see the "V3" link in the sidebar.}}
{{Infobox tutorial
|name = A5: Breaking AES-256 Bootloader
|image =
|caption =
|software versions =
|capture hardware = CW-Lite, CW-Lite 2-Part, CW-Pro
|Target Device =
|Target Architecture = XMEGA/Arm
|Hardware Crypto = No
|Purchase Hardware =
}}
This tutorial will take you through a complete attack on an encrypted bootloader using AES-256. This demonstrates how to using side-channel power analysis on practical systems, along with discussing how to perform analysis with custom scripts.
Whilst the tutorial assumes you will be performing the entire capture of traces along with the attack, it is possible to download the traces if you don't have the hardware, in which case skip section [[#Setting up the Hardware]] and [[#Capturing the Traces]].
== Background ==
In the world of microcontrollers, a bootloader is a special piece of firmware that is made to let the user upload new programs into memory. This is especially useful for devices with complex code that may need to be patched or otherwise updated in the future - a bootloader makes it possible for the user to upload a patched version of the firmware onto the micro. The bootloader receives information from a communication line (a USB port, serial port, ethernet port, WiFi connection, etc...) and stores this data into program memory. Once the full firmware has been received, the micro can happily run its updated code.
There is one big security issue to worry about with bootloaders. A company may want to stop their customers from writing their own firmware and uploading it onto the micro. For example, this might be for protection reasons - hackers might be able to access parts of the device that weren't meant to be accessed. One way of stopping this is to add encryption. The company can add their own secret signature to the firmware code and encrypt it with a secret key. Then, the bootloader can decrypt the incoming firmware and confirm that the incoming firmware is correctly signed. Users will not know the secret key or the signature tied to the firmware, so they won't be able to "fake" their own.
This tutorial will work with a simple AES-256 bootloader. The victim will receive data through a serial connection, decrypt the command, and confirm that the included signature is correct. Then, it will only save the code into memory if the signature check succeeded. To make this system more robust against attacks, the bootloader will use cipher-block chaining (CBC mode). Our goal is to find the secret key and the CBC initialization vector so that we could successfully fake our own firmware.
=== Bootloader Communications Protocol ===
The bootloader's communications protocol operates over a serial port at 38400 baud rate. The bootloader is always waiting for new data to be sent in this example; in real life one would typically force the bootloader to enter through a command sequence.
Commands sent to the bootloader look as follows:
<pre>
|<-------- Encrypted block (16 bytes) ---------->|
| |
+------+------+------+------+------+------+ .... +------+------+------+
| 0x00 | Signature (4 Bytes) | Data (12 Bytes) | CRC-16 |
+------+------+------+------+------+------+ .... +------+------+------+
</pre>
This frame has four parts:
* <code>0x00</code>: 1 byte of fixed header
* Signature: A secret 4 byte constant. The bootloader will confirm that this signature is correct after decrypting the frame.
* Data: 12 bytes of the incoming firmware. This system forces us to send the code 12 bytes at a time; more complete bootloaders may allow longer variable-length frames.
* CRC-16: A 16-bit checksum using the CRC-CCITT polynomial (0x1021). The LSB of the CRC is sent first, followed by the MSB. The bootloader will reply over the serial port, describing whether or not this CRC check was valid.
As described in the diagram, the 16 byte block is not sent as plaintext. Instead, it is encrypted using AES-256 in CBC mode. This encryption method will be described in the next section.
The bootloader responds to each command with a single byte indicating if the CRC-16 was OK or not:
<pre>
+------+
CRC-OK: | 0xA1 |
+------+
+------+
CRC Failed: | 0xA4 |
+------+
</pre>
Then, after replying to the command, the bootloader veries that the signature is correct. If it matches the expected manufacturer's signature, the 12 bytes of data will be written to flash memory. Otherwise, the data is discarded.
=== Details of AES-256 CBC ===
The system uses the AES algorithm in Cipher Block Chaining (CBC) mode. In general one avoids using encryption 'as-is' (i.e. Electronic Code Book), since it means any piece of plaintext always maps to the same piece of ciphertext. Cipher Block Chaining ensures that if you encrypted the same thing a bunch of times it would always encrypt to a new piece of ciphertext.
You can see another reference on the design of the encryption side; we'll be only talking about the decryption side here. In this case AES-256 CBC mode is used as follows, where the details of the AES-256 Decryption block will be discussed in detail later:
[[File:aes256_cbc.png|image]]
This diagram shows that the output of the decryption is no longer used directly as the plaintext. Instead, the output is XORed with a 16 byte mask, which is usually taken from the previous ciphertext. Also, the first decryption block has no previous ciphertext to use, so a secret initialization vector (IV) is used instead. If we are going to decrypt the entire ciphertext (including block 0) or correctly generate our own ciphertext, we'll need to find this IV along with the AES key.
=== Attacking AES-256 ===
The system in this tutorial uses AES-256 encryption, which has a 256 bit (32 byte) key - twice as large as the 16 byte key we've attacked in previous tutorials. This means that our regular AES-128 CPA attacks won't quite work. However, extending these attacks to AES-256 is fairly straightforward: the theory is explained in detail in [[Extending AES-128 Attacks to AES-256]].
As the theory page explains, our AES-256 attack will have 4 steps:
# Perform a standard attack (as in AES-128 decryption) to determine the first 16 bytes of the key, corresponding to the 14th round encryption key.
# Using the known 14th round key, calculate the hypothetical outputs of each S-Box from the 13th round using the ciphertext processed by the 14th round, and determine the 16 bytes of the 13th round key manipulated by inverse MixColumns.
# Perform the MixColumns and ShiftRows operation on the hypothetical key determined above, recovering the 13th round key.
# Using the AES-256 key schedule, reverse the 13th and 14th round keys to determine the original AES-256 encryption key.
== Building the Firmware ==
{{Warningbox|Are you following this tutorial at a training event? If so ONLY use the provided hex-file with secret key already embedded, do not rebuild the firmware!}}
For this example, we'll be using the <code>bootloader-aes256</code> project.
{{CollapsibleSection
|intro = === Building for CWLite with XMEGA Target ===
|content= Building for XMEGA}}
{{CollapsibleSection
|intro = === Building for CWLite with Arm Target ===
|content= Building for Arm}}
{{CollapsibleSection
|intro = === Building for Other Targets ===
|content= Building for Other Targets}}
== Setting up the Hardware ==
Note that you '''don't need hardware''' to complete the tutorial. Instead, you can download [https://www.assembla.com/spaces/chipwhisperer/wiki/Example_Captures example traces from the ChipWhisperer Site]. Just look for the traces titled ''AVR: AES256 Bootloader (ChipWhisperer Tutorial #A5)''.
{{CollapsibleSection
|intro = === CW1173 (Lite) Hardware Setup ===
|content= CWLite HW Setup}}
{{CollapsibleSection
|intro = === CW1200 (Pro) Hardware Setup ===
|content= CW1200 HW Setup}}
{{CollapsibleSection
|intro = === CW308 (UFO) Hardware Setup ===
|content= CW308 HW Setup}}
== Programming the Target ==
{{CollapsibleSection
|intro = === Programming the XMEGA Target ===
|content= Programming XMEGA}}
{{CollapsibleSection
|intro = === Programming the STM32F3 (CW303 Arm) Target ===
|content= Programming Arm}}
{{CollapsibleSection
|intro = === Programming Other Targets ===
|content= Programming Other}}
== Capturing the Traces ==
Once the hardware is ready, we can capture some traces for our attack using the ChipWhisperer Capture software. If you somehow got to the 5th ''Advanced Tutorial'' without getting this software ready, you can follow the helpful guide at [[Installing ChipWhisperer]].
The first thing we need to do is add a new target to the ChipWhisperer system. (None of the existing ones know about the bootloader's data format, nor do they recognize the CRC responses that are sent back to us.) The code for this target is included in [[#Appendix A: Target Code]]. Copy/paste this into a Python file (call it whatever you want) and save it in a place where ChipWhisperer will look for it. There are two folders that you can use:
* Your computer should have a folder called <code>chipwhisperer_projects</code> - if you don't know where this is, the ''File > Preferences'' window will tell you. The system looks in the folder <code>chipwhisperer_projects\chipwhisperer\capture\targets</code> for new targets, so you can save your file here.
* Alternatively, all of the normal targets are stored in <code>chipwhisperer\software\chipwhisperer\capture\targets</code>, so you can also save the file here. Note that this may not be possible if you don't have access to these folders (ex: your account doesn't have admin access).
Next is the capture script. In some of the previous tutorials, we entered all of the capture settings by hand. Since we are civilized humans armed with technology, we can use a script to do all of this setup for us. A pre-written Python script is provided at [[#Appendix B: Capture Script]]. Take a look at this code and notice what it does:
* it fills in the scope, target, and trace format that we'll use;
* it connects to the hardware; and
* it loads all of the hardware parameters for us. Nice!
Copy this script into a <code>.py</code> file somewhere convenient. Then, perform the following steps to finish the capture:
# Run the capture script, which will open a ChipWhisperer Capture window with everything connected for us.
# Open the terminal (''Tools > Terminal'') and connect to the board. While the terminal is open, press the ''Capture 1'' button. A single byte of data should appear in the terminal. This byte will either be <code>a1</code> (CRC failed) or <code>a4</code> (CRC OK). If you see any other responses, something is wrong.
#: [[File:Tutorial-A5-Capture.PNG|image]]
# Once you're happy with this, open the General Settings tab and set the Number of Traces. You should need around 100 traces to break AES.
# Press the ''Capture Many'' button to record the 100 traces. You'll see the new traces plotted on-screen.
# Once the program is finished capturing the traces, save the project. Put it somewhere memorable and give it a nice name.
== Finding the Encryption Key ==
Now that we have our traces, we can go ahead and perform the attack. As described in the background theory, we'll have to do two attacks - one to get the 14th round key, and another (using the first result) to get the 13th round key. Then, we'll do some post-processing to finally get the 256 bit encryption key.
=== 14th Round Key ===
We can attack the 14th round key with a standard, no-frills CPA attack:
# Open the ChipWhisperer Analyzer program and load the <code>.cwp</code> file with the 13th and 14th round traces. This can be either the <code>aes256_round1413_key0_100.cwp</code> file downloaded or the capture you performed.
# View and manipulate the trace data with the following steps:
## Switch to the ''Trace Output Plot'' tab
## Switch to the ''Results'' parameter setting tab
## Choose the traces to be plotted and press the ''Redraw'' button to draw them
## Right-click on the waveform to change options, or left-click and drag to zoom
## Use the toolbar to quickly reset the zoom back to original
##: [[File:Tutorial-A5-Plot-Traces.PNG|image]]
##: Notice that the traces are synchronized for the first 7000 samples (if you're attacking an XMEGA target), but become unsynchronized later. This fact will be important later in the tutorial.
# Set up the attack in the ''Attack'' script:
## Make a copy of the ''attack_cpa.py'' script, call it something new (such as ''attack_aesdec14.py'')
## Adjust the model from ''SBox_output'' to ''InvSBox_output''. This is done by finding the following line in the script:
##: <pre>from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, SBox_output</pre>
##: and change that line to:
##: <pre>from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, InvSBox_output</pre>
## and then also change this further down where we set the leakage model:
##: <pre>leak_model = AES128_8bit(InvSBox_output)</pre>
## If you're finding the attack very slow, narrow down the attack a bit. Normally, this requires a bit of investigation to determine which ranges of the trace are important. Here, you can use the range from 2900 to 4200 for XMEGA and from 1400 to 2600 for STM32F3 (CW303 Arm). The default settings will also work fine, so if you're using another target, you may want to attack the full range. To narrow the attack, edit the <code>attack.setPointRange(0, -1)</code> with the new range.
##: <pre>attack.setPointRange((2900, 4200))</pre>
# Note that we do ''not'' know the secret encryption key, so we cannot highlight the correct key automatically. If you want to fix this, the ''Results'' settings tab has a Highlighted Key setting. Change this to Override mode and enter the key <code>ea 79 79 20 c8 71 44 7d 46 62 5f 51 85 c1 3b cb</code>.
# Finally, run the attack by switching to the ''Results Table'' tab and then hitting the ''Run'' button while your script is selected.
#: [[File:A5_run_script_round14.png|400px]]
There are a few ways to check the results of the attack. First, the results table will show the best guesses for each subkey. With the highlight override enabled, the red bytes should be the best guesses for every single subkey:
[[File:Tutorial-A5-Results-Right-Key.PNG|image]]
However, the correct key will still rise to the top even if the wrong bytes are highlighted. The coloring and correlation coefficients in the results table should still make it clear that the top guess is the best one:
[[File:Tutorial-A5-Results-Wrong-Key.PNG|image]]
{{warningbox|The default capture stores the WRONG knownkey, so you will have highlighted bytes that are not the correct key. We are looking instead for a large delta between the best-guess and all other guesses. For example for Byte 0 we have the most likely as 0.8141, and 2nd best guess as 0.3551. If our best guess was 0.8141 and 2nd best guess was 0.7981 this would indicate we likely haven't broken the key.}}
Finally, the ''Output vs Point Plot'' shows the correlation against all of the sample points. The spikes on this plot show exactly where the attack was successful (ie: where the sensitive data was leaked):
[[File:Aes14round points.png|image]]
In any case, we've determined that the correct 14th round key is <code>ea 79 79 20 c8 71 44 7d 46 62 5f 51 85 c1 3b cb</code>.
''NOTE: if you're stuck, a full listing of the attack script is given in [[#Appendix C: AES-256 14th Round Key Script]].''
=== 13th Round Key ===
Unfortunately, we cannot use the GUI to attack the 13th round key. The system has no built-in model for round 13 of the AES-256 algorithm. Instead, we can write our own script and insert a custom model into the system. See [[#Appendix D: AES-256 13th Round Key Script]] for complete script used here.
==== Resynchronizing Traces (XMEGA only) ====
# Open the ChipWhisperer Analyzer software again and reopen the project file (if closed).
# Recall from the 14th round attack that the trace data becomes unsynchronized around sample 7000. This is due to a non-constant AES implementation: the code does not always take the same amount of time to run for every input. (It's actually possible to do a timing attack on this AES implementation! We'll stick with our CPA attack for now.)
#: [[File:syncproblems.png|image]]
# Resynchronize the traces, see the separate 'Preprocessing' tutorial (NB: only in slides right now!)
{{warningbox|Make sure you get a nice aligned last section of the traces, as in the below figure. You may need to adjust the "input window" or "reference points" slightly. If you do not see the nice alignment the remaining attack will fail!
[[File:A5_pp_resync_end.png|400px]]}}
==== Continuing the Attack ====
The next step is to program our own leakage model. The following Python code models the Hamming weight model of the 13th round S-box:
<syntaxhighlight lang="python">
# Imports for AES256 Attack
from chipwhisperer.analyzer.attacks.models.AES128_8bit import AESLeakageHelper
class AES256_Round13_Model(AESLeakageHelper):
def leakage(self, pt, ct, guess, bnum):
#You must but YOUR recovered 14th round key here - this example may not be accurate!
calc_round_key = [0xea, 0x79, 0x79, 0x20, 0xc8, 0x71, 0x44, 0x7d, 0x46, 0x62, 0x5f, 0x51, 0x85, 0xc1, 0x3b, 0xcb]
xored = [calc_round_key[i] ^ pt[i] for i in range(0, 16)]
block = xored
block = self.inv_shiftrows(block)
block = self.inv_subbytes(block)
block = self.inv_mixcolumns(block)
block = self.inv_shiftrows(block)
result = block
return self.inv_sbox((result[bnum] ^ guess[bnum]))
</syntaxhighlight>
You can look back at the C code of the AES-256 decryption to see how this is implementing the decryption code. Note that because of the Inverse MixColumns operation, we need the entire input ciphertext -- otherwise, we would only need to operate on one byte of the ciphertext.
The last step is to perform the attack using this model:
# Add the above function to your custom script file.
# Change the <code>setAnalysisAlgorithm</code> in the script to use your custom functions by making the following call:
#:<pre>leak_model = AES128_8bit(AES256_Round13_Model)</pre>
# As we did in the 14th round attack, reducing the point range can speed up the attack. For example, to use a smaller range of points, try changing the <code>setPointRange()</code> function call to
#:<pre>self.attack.setPointRange((8000,10990))</pre> on XMEGA or <pre>self.attack.setPointRange((6500,8500))</pre> for the STM32.
# Start the attack! Wait for the attack to complete, and you will determine the 13th round key:
#: [[File:Tutorial-A5-Results-Round-13.PNG|image]]
Note you can check [[#Appendix C AES-256 13th Round Key Script]] for the complete contents of the attack script.
Finally, we need to convert this hypothetical key into the actual value of the 13th round key. We can do this by passing the key through ShiftRows and MixColumns to remove the effect of these two functions. This is easy to do in the Python console (assuming we had the recovered key <code>C6 BD 4E 50 AB CA 75 77 79 87 96 CA 1C 7F C5 82</code>, if you recovered a different key replace the <code>knownkey</code> value with yours):
<pre>
>>> from chipwhisperer.analyzer.attacks.models.aes.funcs import shiftrows,mixcolumns
>>> knownkey = [0xC6, 0xBD, 0x4E, 0x50, 0xAB, 0xCA, 0x75, 0x77, 0x79, 0x87, 0x96, 0xCA, 0x1C, 0x7F, 0xC5, 0x82]
>>> key = shiftrows(knownkey)
>>> key = mixcolumns(key)
>>> print " ".join(["%02x" % i for i in key])
c6 6a a6 12 4a ba 4d 04 4a 22 03 54 5b 28 0e 63
</pre>
Our hard work has rewarded us with the 13th round key, which is <code>c6 6a a6 12 4a ba 4d 04 4a 22 03 54 5b 28 0e 63</code>.
=== Recovering the Encryption Key ===
Finally, we have enough information to recover the initial encryption key. In AES-256, the initial key is used in the key expansion routine to generate 15 round keys, and we know the key for round 13 and 14. All we need to do now is reverse the key scheduling algorithm to calculate the ''0/1 Round Key'' from the ''13/14 Round Key''.
In the ChipWhisperer Analyzer software, a key schedule calculator is provided in ''Tools > AES Key Schedule'':
[[File:keyschedule_tool.png|image]]
Open this tool and paste the 13/14 round keys, which are
<pre>
c6 6a a6 12 4a ba 4d 04 4a 22 03 54 5b 28 0e 63 ea 79 79 20 c8 71 44 7d 46 62 5f 51 85 c1 3b cb
</pre>
Tell the tool that this key is the 13/14 round key; it will automatically display the entire key schedule and the initial encryption key. You should find the initial encryption key is:
<pre>
94 28 5d 4d 6d cf ec 08 d8 ac dd f6 be 25 a4 99 c4 d9 d0 1e c3 40 7e d7 d5 28 d4 09 e9 f0 88 a1
</pre>
Peek into <code>supersecret.h</code>, confirm that this is the right key, and celebrate!
== Next Steps ==
If you want to go further with this tutorial, [[Tutorial A5-Bonus Breaking AES-256 Bootloader]] continues working with the same firmware to find the remaining secrets in the bootloader (the IV and the signature).
== Appendix A: Target Code ==
<syntaxhighlight lang=python>
#!/usr/bin/python
# -*- coding: utf-8 -*-
#
# Copyright (c) 2013-2016, NewAE Technology Inc
# All rights reserved.
#
# Authors: Colin O'Flynn, Greg d'Eon
#
# Find this and more at newae.com - this file is part of the chipwhisperer
# project, http://www.assembla.com/spaces/chipwhisperer
#
# This file is part of chipwhisperer.
#
# chipwhisperer is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# chipwhisperer is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU Lesser General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with chipwhisperer. If not, see <http://www.gnu.org/licenses/>.
#=================================================
import sys
import time
import chipwhisperer.capture.ui.CWCaptureGUI as cwc
from chipwhisperer.common.api.CWCoreAPI import CWCoreAPI
from chipwhisperer.capture.targets.SimpleSerial import SimpleSerial
from chipwhisperer.common.scripts.base import UserScriptBase
from chipwhisperer.capture.targets._base import TargetTemplate
from chipwhisperer.common.utils import pluginmanager
from chipwhisperer.capture.targets.simpleserial_readers.cwlite import SimpleSerial_ChipWhispererLite
from chipwhisperer.common.utils.parameter import setupSetParam
# Class Crc
#############################################################
# These CRC routines are copy-pasted from pycrc, which are:
# Copyright (c) 2006-2013 Thomas Pircher <tehpeh@gmx.net>
#
class Crc(object):
"""
A base class for CRC routines.
"""
def __init__(self, width, poly):
"""The Crc constructor.
The parameters are as follows:
width
poly
reflect_in
xor_in
reflect_out
xor_out
"""
self.Width = width
self.Poly = poly
self.MSB_Mask = 0x1 << (self.Width - 1)
self.Mask = ((self.MSB_Mask - 1) << 1) | 1
self.XorIn = 0x0000
self.XorOut = 0x0000
self.DirectInit = self.XorIn
self.NonDirectInit = self.__get_nondirect_init(self.XorIn)
if self.Width < 8:
self.CrcShift = 8 - self.Width
else:
self.CrcShift = 0
def __get_nondirect_init(self, init):
"""
return the non-direct init if the direct algorithm has been selected.
"""
crc = init
for i in range(self.Width):
bit = crc & 0x01
if bit:
crc ^= self.Poly
crc >>= 1
if bit:
crc |= self.MSB_Mask
return crc & self.Mask
def bit_by_bit(self, in_data):
"""
Classic simple and slow CRC implementation. This function iterates bit
by bit over the augmented input message and returns the calculated CRC
value at the end.
"""
# If the input data is a string, convert to bytes.
if isinstance(in_data, str):
in_data = [ord(c) for c in in_data]
register = self.NonDirectInit
for octet in in_data:
for i in range(8):
topbit = register & self.MSB_Mask
register = ((register << 1) & self.Mask) | ((octet >> (7 - i)) & 0x01)
if topbit:
register ^= self.Poly
for i in range(self.Width):
topbit = register & self.MSB_Mask
register = ((register << 1) & self.Mask)
if topbit:
register ^= self.Poly
return register ^ self.XorOut
class BootloaderTarget(TargetTemplate):
_name = 'AES Bootloader'
def __init__(self):
TargetTemplate.__init__(self)
ser_cons = pluginmanager.getPluginsInDictFromPackage("chipwhisperer.capture.targets.simpleserial_readers", True, False)
self.ser = ser_cons[SimpleSerial_ChipWhispererLite._name]
self.keylength = 16
self.input = ""
self.crc = Crc(width=16, poly=0x1021)
self.setConnection(self.ser)
def setKeyLen(self, klen):
""" Set key length in BITS """
self.keylength = klen / 8
def keyLen(self):
""" Return key length in BYTES """
return self.keylength
def getConnection(self):
return self.ser
def setConnection(self, con):
self.ser = con
self.params.append(self.ser.getParams())
self.ser.connectStatus.connect(self.connectStatus.emit)
self.ser.selectionChanged()
def con(self, scope=None):
if not scope or not hasattr(scope, "qtadc"): Warning(
"You need a scope with OpenADC connected to use this Target")
self.ser.con(scope)
# 'x' flushes everything & sets system back to idle
self.ser.write("xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx")
self.ser.flush()
self.connectStatus.setValue(True)
def close(self):
if self.ser != None:
self.ser.close()
return
def init(self):
pass
def setModeEncrypt(self):
return
def setModeDecrypt(self):
return
def convertVarToString(self, var):
if isinstance(var, str):
return var
sep = ""
s = sep.join(["%c" % b for b in var])
return s
def loadEncryptionKey(self, key):
pass
def loadInput(self, inputtext):
self.input = inputtext
def readOutput(self):
# No actual output
return [0] * 16
def isDone(self):
return True
def checkEncryptionKey(self, kin):
return kin
def go(self):
# Starting byte is 0x00
message = [0x00]
# Append 16 bytes of data
message.extend(self.input)
# Append 2 bytes of CRC for input only (not including 0x00)
crcdata = self.crc.bit_by_bit(self.input)
message.append(crcdata >> 8)
message.append(crcdata & 0xff)
# Write message
message = self.convertVarToString(message)
for i in range(0, 5):
self.ser.flush()
self.ser.write(message)
time.sleep(0.1)
data = self.ser.read(1)
if len(data) > 0:
resp = ord(data[0])
if resp == 0xA4:
# Encryption run OK
break
if resp != 0xA1:
raise IOError("Bad Response %x" % resp)
if len(data) > 0:
if resp != 0xA4:
raise IOError("Failed to communicate, last response: %x" % resp)
else:
raise IOError("Failed to communicate, no response")
</syntaxhighlight>
== Appendix B: Capture Script ==
Note you need to manually CONNECT to the CW-Lite & AES Bootloader target before running this. To do this:
# Set the 'Scope Module' as 'ChipWhisperer/OpenADC'
# Set the 'Target Module' as 'AES Bootloader' (you need to have that target on your system)
<syntaxhighlight lang=python>
"""Setup script for CWLite/1200 with XMEGA (CW303/CW308-XMEGA/CWLite target)
specifically for Tutorial A5: the AES-256 bootloader attack
"""
try:
scope = self.scope
except NameError:
pass
scope.gain.gain = 45
scope.adc.samples = 11000
scope.adc.offset = 0
scope.adc.basic_mode = "rising_edge"
scope.clock.clkgen_freq = 7370000
scope.clock.adc_src = "clkgen_x4"
scope.trigger.triggers = "tio4"
scope.io.tio1 = "serial_rx"
scope.io.tio2 = "serial_tx"
scope.io.hs2 = "clkgen"
</syntaxhighlight>
== Appendix C: AES-256 14th Round Key Script ==
Full attack script, copy/paste into a file then run from within ChipWhisperer-Analyzer:
<syntaxhighlight lang=python>
import chipwhisperer as cw
from chipwhisperer.analyzer.attacks.cpa import CPA
from chipwhisperer.analyzer.attacks.cpa_algorithms.progressive import CPAProgressive
from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, InvSBox_output
#self.project = cw.openProject("2017-mar23-xmega-aes.cwp")
traces = self.project.traceManager()
attack = CPA()
leak_model = AES128_8bit(InvSBox_output)
attack.setAnalysisAlgorithm(CPAProgressive, leak_model)
attack.setTraceSource(traces)
attack.setTraceStart(0)
attack.setTracesPerAttack(-1)
attack.setIterations(1)
attack.setReportingInterval(10)
attack.setTargetSubkeys([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15])
attack.setPointRange((2900, 4200))
self.results_table.setAnalysisSource(attack)
self.correlation_plot.setAnalysisSource(attack)
self.output_plot.setAnalysisSource(attack)
self.pge_plot.setAnalysisSource(attack)
attack.processTraces()
</syntaxhighlight>
== Appendix D: AES-256 13th Round Key Script ==
<syntaxhighlight lang=python>
import chipwhisperer as cw
from chipwhisperer.analyzer.attacks.cpa import CPA
from chipwhisperer.analyzer.attacks.cpa_algorithms.progressive import CPAProgressive
from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, AESLeakageHelper
from chipwhisperer.analyzer.preprocessing.resync_sad import ResyncSAD
class AES256_Round13_Model(AESLeakageHelper):
def leakage(self, pt, ct, guess, bnum):
#You must but YOUR recovered 14th round key here - this example may not be accurate!
calc_round_key = [0xea, 0x79, 0x79, 0x20, 0xc8, 0x71, 0x44, 0x7d, 0x46, 0x62, 0x5f, 0x51, 0x85, 0xc1, 0x3b, 0xcb]
xored = [calc_round_key[i] ^ pt[i] for i in range(0, 16)]
block = xored
block = self.inv_shiftrows(block)
block = self.inv_subbytes(block)
block = self.inv_mixcolumns(block)
block = self.inv_shiftrows(block)
result = block
return self.inv_sbox((result[bnum] ^ guess[bnum]))
traces = self.project.traceManager()
resync_traces = ResyncSAD(traces)
resync_traces.enabled = True
resync_traces.ref_trace = 0
resync_traces.target_window = (9100, 9300)
resync_traces.max_shift = 200
attack = CPA()
leak_model = AES128_8bit(AES256_Round13_Model)
attack.setAnalysisAlgorithm(CPAProgressive, leak_model)
attack.setTraceSource(resync_traces)
attack.setTraceStart(0)
attack.setTracesPerAttack(-1)
attack.setIterations(1)
attack.setReportingInterval(10)
attack.setTargetSubkeys([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15])
attack.setPointRange((0, -1))
self.results_table.setAnalysisSource(attack)
self.correlation_plot.setAnalysisSource(attack)
self.output_plot.setAnalysisSource(attack)
self.pge_plot.setAnalysisSource(attack)
attack.processTraces()
</syntaxhighlight>
== Links ==
{{Template:Tutorials}}
[[Category:Tutorials]]
{{Infobox tutorial
|name = A5: Breaking AES-256 Bootloader
|image =
|caption =
|software versions =
|capture hardware = CW-Lite, CW-Lite 2-Part, CW-Pro
|Target Device =
|Target Architecture = XMEGA/Arm
|Hardware Crypto = No
|Purchase Hardware =
}}
This tutorial will take you through a complete attack on an encrypted bootloader using AES-256. This demonstrates how to using side-channel power analysis on practical systems, along with discussing how to perform analysis with custom scripts.
Whilst the tutorial assumes you will be performing the entire capture of traces along with the attack, it is possible to download the traces if you don't have the hardware, in which case skip section [[#Setting up the Hardware]] and [[#Capturing the Traces]].
== Background ==
In the world of microcontrollers, a bootloader is a special piece of firmware that is made to let the user upload new programs into memory. This is especially useful for devices with complex code that may need to be patched or otherwise updated in the future - a bootloader makes it possible for the user to upload a patched version of the firmware onto the micro. The bootloader receives information from a communication line (a USB port, serial port, ethernet port, WiFi connection, etc...) and stores this data into program memory. Once the full firmware has been received, the micro can happily run its updated code.
There is one big security issue to worry about with bootloaders. A company may want to stop their customers from writing their own firmware and uploading it onto the micro. For example, this might be for protection reasons - hackers might be able to access parts of the device that weren't meant to be accessed. One way of stopping this is to add encryption. The company can add their own secret signature to the firmware code and encrypt it with a secret key. Then, the bootloader can decrypt the incoming firmware and confirm that the incoming firmware is correctly signed. Users will not know the secret key or the signature tied to the firmware, so they won't be able to "fake" their own.
This tutorial will work with a simple AES-256 bootloader. The victim will receive data through a serial connection, decrypt the command, and confirm that the included signature is correct. Then, it will only save the code into memory if the signature check succeeded. To make this system more robust against attacks, the bootloader will use cipher-block chaining (CBC mode). Our goal is to find the secret key and the CBC initialization vector so that we could successfully fake our own firmware.
=== Bootloader Communications Protocol ===
The bootloader's communications protocol operates over a serial port at 38400 baud rate. The bootloader is always waiting for new data to be sent in this example; in real life one would typically force the bootloader to enter through a command sequence.
Commands sent to the bootloader look as follows:
<pre>
|<-------- Encrypted block (16 bytes) ---------->|
| |
+------+------+------+------+------+------+ .... +------+------+------+
| 0x00 | Signature (4 Bytes) | Data (12 Bytes) | CRC-16 |
+------+------+------+------+------+------+ .... +------+------+------+
</pre>
This frame has four parts:
* <code>0x00</code>: 1 byte of fixed header
* Signature: A secret 4 byte constant. The bootloader will confirm that this signature is correct after decrypting the frame.
* Data: 12 bytes of the incoming firmware. This system forces us to send the code 12 bytes at a time; more complete bootloaders may allow longer variable-length frames.
* CRC-16: A 16-bit checksum using the CRC-CCITT polynomial (0x1021). The LSB of the CRC is sent first, followed by the MSB. The bootloader will reply over the serial port, describing whether or not this CRC check was valid.
As described in the diagram, the 16 byte block is not sent as plaintext. Instead, it is encrypted using AES-256 in CBC mode. This encryption method will be described in the next section.
The bootloader responds to each command with a single byte indicating if the CRC-16 was OK or not:
<pre>
+------+
CRC-OK: | 0xA1 |
+------+
+------+
CRC Failed: | 0xA4 |
+------+
</pre>
Then, after replying to the command, the bootloader veries that the signature is correct. If it matches the expected manufacturer's signature, the 12 bytes of data will be written to flash memory. Otherwise, the data is discarded.
=== Details of AES-256 CBC ===
The system uses the AES algorithm in Cipher Block Chaining (CBC) mode. In general one avoids using encryption 'as-is' (i.e. Electronic Code Book), since it means any piece of plaintext always maps to the same piece of ciphertext. Cipher Block Chaining ensures that if you encrypted the same thing a bunch of times it would always encrypt to a new piece of ciphertext.
You can see another reference on the design of the encryption side; we'll be only talking about the decryption side here. In this case AES-256 CBC mode is used as follows, where the details of the AES-256 Decryption block will be discussed in detail later:
[[File:aes256_cbc.png|image]]
This diagram shows that the output of the decryption is no longer used directly as the plaintext. Instead, the output is XORed with a 16 byte mask, which is usually taken from the previous ciphertext. Also, the first decryption block has no previous ciphertext to use, so a secret initialization vector (IV) is used instead. If we are going to decrypt the entire ciphertext (including block 0) or correctly generate our own ciphertext, we'll need to find this IV along with the AES key.
=== Attacking AES-256 ===
The system in this tutorial uses AES-256 encryption, which has a 256 bit (32 byte) key - twice as large as the 16 byte key we've attacked in previous tutorials. This means that our regular AES-128 CPA attacks won't quite work. However, extending these attacks to AES-256 is fairly straightforward: the theory is explained in detail in [[Extending AES-128 Attacks to AES-256]].
As the theory page explains, our AES-256 attack will have 4 steps:
# Perform a standard attack (as in AES-128 decryption) to determine the first 16 bytes of the key, corresponding to the 14th round encryption key.
# Using the known 14th round key, calculate the hypothetical outputs of each S-Box from the 13th round using the ciphertext processed by the 14th round, and determine the 16 bytes of the 13th round key manipulated by inverse MixColumns.
# Perform the MixColumns and ShiftRows operation on the hypothetical key determined above, recovering the 13th round key.
# Using the AES-256 key schedule, reverse the 13th and 14th round keys to determine the original AES-256 encryption key.
== Building the Firmware ==
{{Warningbox|Are you following this tutorial at a training event? If so ONLY use the provided hex-file with secret key already embedded, do not rebuild the firmware!}}
For this example, we'll be using the <code>bootloader-aes256</code> project.
{{CollapsibleSection
|intro = === Building for CWLite with XMEGA Target ===
|content= Building for XMEGA}}
{{CollapsibleSection
|intro = === Building for CWLite with Arm Target ===
|content= Building for Arm}}
{{CollapsibleSection
|intro = === Building for Other Targets ===
|content= Building for Other Targets}}
== Setting up the Hardware ==
Note that you '''don't need hardware''' to complete the tutorial. Instead, you can download [https://www.assembla.com/spaces/chipwhisperer/wiki/Example_Captures example traces from the ChipWhisperer Site]. Just look for the traces titled ''AVR: AES256 Bootloader (ChipWhisperer Tutorial #A5)''.
{{CollapsibleSection
|intro = === CW1173 (Lite) Hardware Setup ===
|content= CWLite HW Setup}}
{{CollapsibleSection
|intro = === CW1200 (Pro) Hardware Setup ===
|content= CW1200 HW Setup}}
{{CollapsibleSection
|intro = === CW308 (UFO) Hardware Setup ===
|content= CW308 HW Setup}}
== Programming the Target ==
{{CollapsibleSection
|intro = === Programming the XMEGA Target ===
|content= Programming XMEGA}}
{{CollapsibleSection
|intro = === Programming the STM32F3 (CW303 Arm) Target ===
|content= Programming Arm}}
{{CollapsibleSection
|intro = === Programming Other Targets ===
|content= Programming Other}}
== Capturing the Traces ==
Once the hardware is ready, we can capture some traces for our attack using the ChipWhisperer Capture software. If you somehow got to the 5th ''Advanced Tutorial'' without getting this software ready, you can follow the helpful guide at [[Installing ChipWhisperer]].
The first thing we need to do is add a new target to the ChipWhisperer system. (None of the existing ones know about the bootloader's data format, nor do they recognize the CRC responses that are sent back to us.) The code for this target is included in [[#Appendix A: Target Code]]. Copy/paste this into a Python file (call it whatever you want) and save it in a place where ChipWhisperer will look for it. There are two folders that you can use:
* Your computer should have a folder called <code>chipwhisperer_projects</code> - if you don't know where this is, the ''File > Preferences'' window will tell you. The system looks in the folder <code>chipwhisperer_projects\chipwhisperer\capture\targets</code> for new targets, so you can save your file here.
* Alternatively, all of the normal targets are stored in <code>chipwhisperer\software\chipwhisperer\capture\targets</code>, so you can also save the file here. Note that this may not be possible if you don't have access to these folders (ex: your account doesn't have admin access).
Next is the capture script. In some of the previous tutorials, we entered all of the capture settings by hand. Since we are civilized humans armed with technology, we can use a script to do all of this setup for us. A pre-written Python script is provided at [[#Appendix B: Capture Script]]. Take a look at this code and notice what it does:
* it fills in the scope, target, and trace format that we'll use;
* it connects to the hardware; and
* it loads all of the hardware parameters for us. Nice!
Copy this script into a <code>.py</code> file somewhere convenient. Then, perform the following steps to finish the capture:
# Run the capture script, which will open a ChipWhisperer Capture window with everything connected for us.
# Open the terminal (''Tools > Terminal'') and connect to the board. While the terminal is open, press the ''Capture 1'' button. A single byte of data should appear in the terminal. This byte will either be <code>a1</code> (CRC failed) or <code>a4</code> (CRC OK). If you see any other responses, something is wrong.
#: [[File:Tutorial-A5-Capture.PNG|image]]
# Once you're happy with this, open the General Settings tab and set the Number of Traces. You should need around 100 traces to break AES.
# Press the ''Capture Many'' button to record the 100 traces. You'll see the new traces plotted on-screen.
# Once the program is finished capturing the traces, save the project. Put it somewhere memorable and give it a nice name.
== Finding the Encryption Key ==
Now that we have our traces, we can go ahead and perform the attack. As described in the background theory, we'll have to do two attacks - one to get the 14th round key, and another (using the first result) to get the 13th round key. Then, we'll do some post-processing to finally get the 256 bit encryption key.
=== 14th Round Key ===
We can attack the 14th round key with a standard, no-frills CPA attack:
# Open the ChipWhisperer Analyzer program and load the <code>.cwp</code> file with the 13th and 14th round traces. This can be either the <code>aes256_round1413_key0_100.cwp</code> file downloaded or the capture you performed.
# View and manipulate the trace data with the following steps:
## Switch to the ''Trace Output Plot'' tab
## Switch to the ''Results'' parameter setting tab
## Choose the traces to be plotted and press the ''Redraw'' button to draw them
## Right-click on the waveform to change options, or left-click and drag to zoom
## Use the toolbar to quickly reset the zoom back to original
##: [[File:Tutorial-A5-Plot-Traces.PNG|image]]
##: Notice that the traces are synchronized for the first 7000 samples (if you're attacking an XMEGA target), but become unsynchronized later. This fact will be important later in the tutorial.
# Set up the attack in the ''Attack'' script:
## Make a copy of the ''attack_cpa.py'' script, call it something new (such as ''attack_aesdec14.py'')
## Adjust the model from ''SBox_output'' to ''InvSBox_output''. This is done by finding the following line in the script:
##: <pre>from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, SBox_output</pre>
##: and change that line to:
##: <pre>from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, InvSBox_output</pre>
## and then also change this further down where we set the leakage model:
##: <pre>leak_model = AES128_8bit(InvSBox_output)</pre>
## If you're finding the attack very slow, narrow down the attack a bit. Normally, this requires a bit of investigation to determine which ranges of the trace are important. Here, you can use the range from 2900 to 4200 for XMEGA and from 1400 to 2600 for STM32F3 (CW303 Arm). The default settings will also work fine, so if you're using another target, you may want to attack the full range. To narrow the attack, edit the <code>attack.setPointRange(0, -1)</code> with the new range.
##: <pre>attack.setPointRange((2900, 4200))</pre>
# Note that we do ''not'' know the secret encryption key, so we cannot highlight the correct key automatically. If you want to fix this, the ''Results'' settings tab has a Highlighted Key setting. Change this to Override mode and enter the key <code>ea 79 79 20 c8 71 44 7d 46 62 5f 51 85 c1 3b cb</code>.
# Finally, run the attack by switching to the ''Results Table'' tab and then hitting the ''Run'' button while your script is selected.
#: [[File:A5_run_script_round14.png|400px]]
There are a few ways to check the results of the attack. First, the results table will show the best guesses for each subkey. With the highlight override enabled, the red bytes should be the best guesses for every single subkey:
[[File:Tutorial-A5-Results-Right-Key.PNG|image]]
However, the correct key will still rise to the top even if the wrong bytes are highlighted. The coloring and correlation coefficients in the results table should still make it clear that the top guess is the best one:
[[File:Tutorial-A5-Results-Wrong-Key.PNG|image]]
{{warningbox|The default capture stores the WRONG knownkey, so you will have highlighted bytes that are not the correct key. We are looking instead for a large delta between the best-guess and all other guesses. For example for Byte 0 we have the most likely as 0.8141, and 2nd best guess as 0.3551. If our best guess was 0.8141 and 2nd best guess was 0.7981 this would indicate we likely haven't broken the key.}}
Finally, the ''Output vs Point Plot'' shows the correlation against all of the sample points. The spikes on this plot show exactly where the attack was successful (ie: where the sensitive data was leaked):
[[File:Aes14round points.png|image]]
In any case, we've determined that the correct 14th round key is <code>ea 79 79 20 c8 71 44 7d 46 62 5f 51 85 c1 3b cb</code>.
''NOTE: if you're stuck, a full listing of the attack script is given in [[#Appendix C: AES-256 14th Round Key Script]].''
=== 13th Round Key ===
Unfortunately, we cannot use the GUI to attack the 13th round key. The system has no built-in model for round 13 of the AES-256 algorithm. Instead, we can write our own script and insert a custom model into the system. See [[#Appendix D: AES-256 13th Round Key Script]] for complete script used here.
==== Resynchronizing Traces (XMEGA only) ====
# Open the ChipWhisperer Analyzer software again and reopen the project file (if closed).
# Recall from the 14th round attack that the trace data becomes unsynchronized around sample 7000. This is due to a non-constant AES implementation: the code does not always take the same amount of time to run for every input. (It's actually possible to do a timing attack on this AES implementation! We'll stick with our CPA attack for now.)
#: [[File:syncproblems.png|image]]
# Resynchronize the traces, see the separate 'Preprocessing' tutorial (NB: only in slides right now!)
{{warningbox|Make sure you get a nice aligned last section of the traces, as in the below figure. You may need to adjust the "input window" or "reference points" slightly. If you do not see the nice alignment the remaining attack will fail!
[[File:A5_pp_resync_end.png|400px]]}}
==== Continuing the Attack ====
The next step is to program our own leakage model. The following Python code models the Hamming weight model of the 13th round S-box:
<syntaxhighlight lang="python">
# Imports for AES256 Attack
from chipwhisperer.analyzer.attacks.models.AES128_8bit import AESLeakageHelper
class AES256_Round13_Model(AESLeakageHelper):
def leakage(self, pt, ct, guess, bnum):
#You must but YOUR recovered 14th round key here - this example may not be accurate!
calc_round_key = [0xea, 0x79, 0x79, 0x20, 0xc8, 0x71, 0x44, 0x7d, 0x46, 0x62, 0x5f, 0x51, 0x85, 0xc1, 0x3b, 0xcb]
xored = [calc_round_key[i] ^ pt[i] for i in range(0, 16)]
block = xored
block = self.inv_shiftrows(block)
block = self.inv_subbytes(block)
block = self.inv_mixcolumns(block)
block = self.inv_shiftrows(block)
result = block
return self.inv_sbox((result[bnum] ^ guess[bnum]))
</syntaxhighlight>
You can look back at the C code of the AES-256 decryption to see how this is implementing the decryption code. Note that because of the Inverse MixColumns operation, we need the entire input ciphertext -- otherwise, we would only need to operate on one byte of the ciphertext.
The last step is to perform the attack using this model:
# Add the above function to your custom script file.
# Change the <code>setAnalysisAlgorithm</code> in the script to use your custom functions by making the following call:
#:<pre>leak_model = AES128_8bit(AES256_Round13_Model)</pre>
# As we did in the 14th round attack, reducing the point range can speed up the attack. For example, to use a smaller range of points, try changing the <code>setPointRange()</code> function call to
#:<pre>self.attack.setPointRange((8000,10990))</pre> on XMEGA or <pre>self.attack.setPointRange((6500,8500))</pre> for the STM32.
# Start the attack! Wait for the attack to complete, and you will determine the 13th round key:
#: [[File:Tutorial-A5-Results-Round-13.PNG|image]]
Note you can check [[#Appendix C AES-256 13th Round Key Script]] for the complete contents of the attack script.
Finally, we need to convert this hypothetical key into the actual value of the 13th round key. We can do this by passing the key through ShiftRows and MixColumns to remove the effect of these two functions. This is easy to do in the Python console (assuming we had the recovered key <code>C6 BD 4E 50 AB CA 75 77 79 87 96 CA 1C 7F C5 82</code>, if you recovered a different key replace the <code>knownkey</code> value with yours):
<pre>
>>> from chipwhisperer.analyzer.attacks.models.aes.funcs import shiftrows,mixcolumns
>>> knownkey = [0xC6, 0xBD, 0x4E, 0x50, 0xAB, 0xCA, 0x75, 0x77, 0x79, 0x87, 0x96, 0xCA, 0x1C, 0x7F, 0xC5, 0x82]
>>> key = shiftrows(knownkey)
>>> key = mixcolumns(key)
>>> print " ".join(["%02x" % i for i in key])
c6 6a a6 12 4a ba 4d 04 4a 22 03 54 5b 28 0e 63
</pre>
Our hard work has rewarded us with the 13th round key, which is <code>c6 6a a6 12 4a ba 4d 04 4a 22 03 54 5b 28 0e 63</code>.
=== Recovering the Encryption Key ===
Finally, we have enough information to recover the initial encryption key. In AES-256, the initial key is used in the key expansion routine to generate 15 round keys, and we know the key for round 13 and 14. All we need to do now is reverse the key scheduling algorithm to calculate the ''0/1 Round Key'' from the ''13/14 Round Key''.
In the ChipWhisperer Analyzer software, a key schedule calculator is provided in ''Tools > AES Key Schedule'':
[[File:keyschedule_tool.png|image]]
Open this tool and paste the 13/14 round keys, which are
<pre>
c6 6a a6 12 4a ba 4d 04 4a 22 03 54 5b 28 0e 63 ea 79 79 20 c8 71 44 7d 46 62 5f 51 85 c1 3b cb
</pre>
Tell the tool that this key is the 13/14 round key; it will automatically display the entire key schedule and the initial encryption key. You should find the initial encryption key is:
<pre>
94 28 5d 4d 6d cf ec 08 d8 ac dd f6 be 25 a4 99 c4 d9 d0 1e c3 40 7e d7 d5 28 d4 09 e9 f0 88 a1
</pre>
Peek into <code>supersecret.h</code>, confirm that this is the right key, and celebrate!
== Next Steps ==
If you want to go further with this tutorial, [[Tutorial A5-Bonus Breaking AES-256 Bootloader]] continues working with the same firmware to find the remaining secrets in the bootloader (the IV and the signature).
== Appendix A: Target Code ==
<syntaxhighlight lang=python>
#!/usr/bin/python
# -*- coding: utf-8 -*-
#
# Copyright (c) 2013-2016, NewAE Technology Inc
# All rights reserved.
#
# Authors: Colin O'Flynn, Greg d'Eon
#
# Find this and more at newae.com - this file is part of the chipwhisperer
# project, http://www.assembla.com/spaces/chipwhisperer
#
# This file is part of chipwhisperer.
#
# chipwhisperer is free software: you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation, either version 3 of the License, or
# (at your option) any later version.
#
# chipwhisperer is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU Lesser General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with chipwhisperer. If not, see <http://www.gnu.org/licenses/>.
#=================================================
import sys
import time
import chipwhisperer.capture.ui.CWCaptureGUI as cwc
from chipwhisperer.common.api.CWCoreAPI import CWCoreAPI
from chipwhisperer.capture.targets.SimpleSerial import SimpleSerial
from chipwhisperer.common.scripts.base import UserScriptBase
from chipwhisperer.capture.targets._base import TargetTemplate
from chipwhisperer.common.utils import pluginmanager
from chipwhisperer.capture.targets.simpleserial_readers.cwlite import SimpleSerial_ChipWhispererLite
from chipwhisperer.common.utils.parameter import setupSetParam
# Class Crc
#############################################################
# These CRC routines are copy-pasted from pycrc, which are:
# Copyright (c) 2006-2013 Thomas Pircher <tehpeh@gmx.net>
#
class Crc(object):
"""
A base class for CRC routines.
"""
def __init__(self, width, poly):
"""The Crc constructor.
The parameters are as follows:
width
poly
reflect_in
xor_in
reflect_out
xor_out
"""
self.Width = width
self.Poly = poly
self.MSB_Mask = 0x1 << (self.Width - 1)
self.Mask = ((self.MSB_Mask - 1) << 1) | 1
self.XorIn = 0x0000
self.XorOut = 0x0000
self.DirectInit = self.XorIn
self.NonDirectInit = self.__get_nondirect_init(self.XorIn)
if self.Width < 8:
self.CrcShift = 8 - self.Width
else:
self.CrcShift = 0
def __get_nondirect_init(self, init):
"""
return the non-direct init if the direct algorithm has been selected.
"""
crc = init
for i in range(self.Width):
bit = crc & 0x01
if bit:
crc ^= self.Poly
crc >>= 1
if bit:
crc |= self.MSB_Mask
return crc & self.Mask
def bit_by_bit(self, in_data):
"""
Classic simple and slow CRC implementation. This function iterates bit
by bit over the augmented input message and returns the calculated CRC
value at the end.
"""
# If the input data is a string, convert to bytes.
if isinstance(in_data, str):
in_data = [ord(c) for c in in_data]
register = self.NonDirectInit
for octet in in_data:
for i in range(8):
topbit = register & self.MSB_Mask
register = ((register << 1) & self.Mask) | ((octet >> (7 - i)) & 0x01)
if topbit:
register ^= self.Poly
for i in range(self.Width):
topbit = register & self.MSB_Mask
register = ((register << 1) & self.Mask)
if topbit:
register ^= self.Poly
return register ^ self.XorOut
class BootloaderTarget(TargetTemplate):
_name = 'AES Bootloader'
def __init__(self):
TargetTemplate.__init__(self)
ser_cons = pluginmanager.getPluginsInDictFromPackage("chipwhisperer.capture.targets.simpleserial_readers", True, False)
self.ser = ser_cons[SimpleSerial_ChipWhispererLite._name]
self.keylength = 16
self.input = ""
self.crc = Crc(width=16, poly=0x1021)
self.setConnection(self.ser)
def setKeyLen(self, klen):
""" Set key length in BITS """
self.keylength = klen / 8
def keyLen(self):
""" Return key length in BYTES """
return self.keylength
def getConnection(self):
return self.ser
def setConnection(self, con):
self.ser = con
self.params.append(self.ser.getParams())
self.ser.connectStatus.connect(self.connectStatus.emit)
self.ser.selectionChanged()
def con(self, scope=None):
if not scope or not hasattr(scope, "qtadc"): Warning(
"You need a scope with OpenADC connected to use this Target")
self.ser.con(scope)
# 'x' flushes everything & sets system back to idle
self.ser.write("xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx")
self.ser.flush()
self.connectStatus.setValue(True)
def close(self):
if self.ser != None:
self.ser.close()
return
def init(self):
pass
def setModeEncrypt(self):
return
def setModeDecrypt(self):
return
def convertVarToString(self, var):
if isinstance(var, str):
return var
sep = ""
s = sep.join(["%c" % b for b in var])
return s
def loadEncryptionKey(self, key):
pass
def loadInput(self, inputtext):
self.input = inputtext
def readOutput(self):
# No actual output
return [0] * 16
def isDone(self):
return True
def checkEncryptionKey(self, kin):
return kin
def go(self):
# Starting byte is 0x00
message = [0x00]
# Append 16 bytes of data
message.extend(self.input)
# Append 2 bytes of CRC for input only (not including 0x00)
crcdata = self.crc.bit_by_bit(self.input)
message.append(crcdata >> 8)
message.append(crcdata & 0xff)
# Write message
message = self.convertVarToString(message)
for i in range(0, 5):
self.ser.flush()
self.ser.write(message)
time.sleep(0.1)
data = self.ser.read(1)
if len(data) > 0:
resp = ord(data[0])
if resp == 0xA4:
# Encryption run OK
break
if resp != 0xA1:
raise IOError("Bad Response %x" % resp)
if len(data) > 0:
if resp != 0xA4:
raise IOError("Failed to communicate, last response: %x" % resp)
else:
raise IOError("Failed to communicate, no response")
</syntaxhighlight>
== Appendix B: Capture Script ==
Note you need to manually CONNECT to the CW-Lite & AES Bootloader target before running this. To do this:
# Set the 'Scope Module' as 'ChipWhisperer/OpenADC'
# Set the 'Target Module' as 'AES Bootloader' (you need to have that target on your system)
<syntaxhighlight lang=python>
"""Setup script for CWLite/1200 with XMEGA (CW303/CW308-XMEGA/CWLite target)
specifically for Tutorial A5: the AES-256 bootloader attack
"""
try:
scope = self.scope
except NameError:
pass
scope.gain.gain = 45
scope.adc.samples = 11000
scope.adc.offset = 0
scope.adc.basic_mode = "rising_edge"
scope.clock.clkgen_freq = 7370000
scope.clock.adc_src = "clkgen_x4"
scope.trigger.triggers = "tio4"
scope.io.tio1 = "serial_rx"
scope.io.tio2 = "serial_tx"
scope.io.hs2 = "clkgen"
</syntaxhighlight>
== Appendix C: AES-256 14th Round Key Script ==
Full attack script, copy/paste into a file then run from within ChipWhisperer-Analyzer:
<syntaxhighlight lang=python>
import chipwhisperer as cw
from chipwhisperer.analyzer.attacks.cpa import CPA
from chipwhisperer.analyzer.attacks.cpa_algorithms.progressive import CPAProgressive
from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, InvSBox_output
#self.project = cw.openProject("2017-mar23-xmega-aes.cwp")
traces = self.project.traceManager()
attack = CPA()
leak_model = AES128_8bit(InvSBox_output)
attack.setAnalysisAlgorithm(CPAProgressive, leak_model)
attack.setTraceSource(traces)
attack.setTraceStart(0)
attack.setTracesPerAttack(-1)
attack.setIterations(1)
attack.setReportingInterval(10)
attack.setTargetSubkeys([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15])
attack.setPointRange((2900, 4200))
self.results_table.setAnalysisSource(attack)
self.correlation_plot.setAnalysisSource(attack)
self.output_plot.setAnalysisSource(attack)
self.pge_plot.setAnalysisSource(attack)
attack.processTraces()
</syntaxhighlight>
== Appendix D: AES-256 13th Round Key Script ==
<syntaxhighlight lang=python>
import chipwhisperer as cw
from chipwhisperer.analyzer.attacks.cpa import CPA
from chipwhisperer.analyzer.attacks.cpa_algorithms.progressive import CPAProgressive
from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, AESLeakageHelper
from chipwhisperer.analyzer.preprocessing.resync_sad import ResyncSAD
class AES256_Round13_Model(AESLeakageHelper):
def leakage(self, pt, ct, guess, bnum):
#You must but YOUR recovered 14th round key here - this example may not be accurate!
calc_round_key = [0xea, 0x79, 0x79, 0x20, 0xc8, 0x71, 0x44, 0x7d, 0x46, 0x62, 0x5f, 0x51, 0x85, 0xc1, 0x3b, 0xcb]
xored = [calc_round_key[i] ^ pt[i] for i in range(0, 16)]
block = xored
block = self.inv_shiftrows(block)
block = self.inv_subbytes(block)
block = self.inv_mixcolumns(block)
block = self.inv_shiftrows(block)
result = block
return self.inv_sbox((result[bnum] ^ guess[bnum]))
traces = self.project.traceManager()
resync_traces = ResyncSAD(traces)
resync_traces.enabled = True
resync_traces.ref_trace = 0
resync_traces.target_window = (9100, 9300)
resync_traces.max_shift = 200
attack = CPA()
leak_model = AES128_8bit(AES256_Round13_Model)
attack.setAnalysisAlgorithm(CPAProgressive, leak_model)
attack.setTraceSource(resync_traces)
attack.setTraceStart(0)
attack.setTracesPerAttack(-1)
attack.setIterations(1)
attack.setReportingInterval(10)
attack.setTargetSubkeys([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15])
attack.setPointRange((0, -1))
self.results_table.setAnalysisSource(attack)
self.correlation_plot.setAnalysisSource(attack)
self.output_plot.setAnalysisSource(attack)
self.pge_plot.setAnalysisSource(attack)
attack.processTraces()
</syntaxhighlight>
== Links ==
{{Template:Tutorials}}
[[Category:Tutorials]]