Tutorial A5 Breaking AES-256 Bootloader

From ChipWhisperer Wiki
Jump to: navigation, search

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.

A5: Breaking AES-256 Bootloader
Target Architecture XMEGA
Hardware Crypto No
Software Release

V3

V4

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:

       |<-------- Encrypted block (16 bytes) ---------->|
       |                                                |
+------+------+------+------+------+------+ .... +------+------+------+
| 0x00 |    Signature (4 Bytes)    |  Data (12 Bytes)   |   CRC-16    |
+------+------+------+------+------+------+ .... +------+------+------+

This frame has four parts:

  • 0x00: 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:

            +------+
CRC-OK:     | 0xA1 |
            +------+

            +------+
CRC Failed: | 0xA4 |
            +------+

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:

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:

  1. 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.
  2. 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.
  3. Perform the MixColumns and ShiftRows operation on the hypothetical key determined above, recovering the 13th round key.
  4. Using the AES-256 key schedule, reverse the 13th and 14th round keys to determine the original AES-256 encryption key.

Setting up the Hardware

This tutorial uses the CW1173 ChipWhisperer-Lite hardware. This hardware does not require any special setup - it should be ready to go out-of-the-box.

Note that you don't need hardware to complete the tutorial. Instead, you can download example traces from the ChipWhisperer Site. Just look for the traces titled AVR: AES256 Bootloader (ChipWhisperer Tutorial #A5).

Building/Programming the Bootloader

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!


The firmware that implements the bootloader is available inside the ChipWhisperer folder at chipwhisperer\hardware\victims\firmware\bootloader-aes256. If you've uploaded the firmware for any of the other tutorials, the process is identical:

  1. Open a command prompt/terminal window and navigate to this folder. Enter the command make PLATFORM=X, where X is the name of your target. For instance, use PLATFORM=CW303 on the ChipWhisperer Lite. Ensure that the program is successfully compiled. The output should end with a line like
    Built for platform CW-Lite XMEGA
  2. Open the ChipWhisperer Capture software and connect to your hardware. Open the programmer window (Tools > CW-Lite XMEGA Programmer), find the .hex file that you just made, and Erase/Program/Verify FLASH.

The firmware is now loaded onto your hardware, and you can continue onto the capture process.

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 chipwhisperer_projects - if you don't know where this is, the File > Preferences window will tell you. The system looks in the folder chipwhisperer_projects\chipwhisperer\capture\targets for new targets, so you can save your file here.
  • Alternatively, all of the normal targets are stored in chipwhisperer\software\chipwhisperer\capture\targets, 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 .py file somewhere convenient. Then, perform the following steps to finish the capture:

  1. Run the capture script, which will open a ChipWhisperer Capture window with everything connected for us.
  2. 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 a1 (CRC failed) or a4 (CRC OK). If you see any other responses, something is wrong.
    image
  3. 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.
  4. Press the Capture Many button to record the 100 traces. You'll see the new traces plotted on-screen.
  5. 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:

  1. Open the ChipWhisperer Analyzer program and load the .cwp file with the 13th and 14th round traces. This can be either the aes256_round1413_key0_100.cwp file downloaded or the capture you performed.
  2. View and manipulate the trace data with the following steps:
    1. Switch to the Trace Output Plot tab
    2. Switch to the Results parameter setting tab
    3. Choose the traces to be plotted and press the Redraw button to draw them
    4. Right-click on the waveform to change options, or left-click and drag to zoom
    5. Use the toolbar to quickly reset the zoom back to original
      image
      Notice that the traces are synchronized for the first 7000 samples, but become unsynchronized later. This fact will be important later in the tutorial.
  3. Set up the attack in the Attack script:
    1. Make a copy of the attack_cpa.py script, call it something new (such as attack_aesdec14.py)
    2. Adjust the model from SBox_output to InvSBox_output. This is done by finding the following line in the script:
      from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, SBox_output
      and change that line to:
      from chipwhisperer.analyzer.attacks.models.AES128_8bit import AES128_8bit, InvBox_output
    3. and then also change this further down where we set the leakage model:
      leak_model = AES128_8bit(InvSBox_output)
    4. 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 for 4200. The default settings will also work fine! To do this adjust the following line to look as follows:
      attack.setPointRange((2900, 4200))
  4. 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 ea 79 79 20 c8 71 44 7d 46 62 5f 51 85 c1 3b cb.
  5. Finally, run the attack by switching to the Results Table tab and then hitting the Run button while your script is selected.
    A5 run script round14.png

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:

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:

image

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):

image

In any case, we've determined that the correct 14th round key is ea 79 79 20 c8 71 44 7d 46 62 5f 51 85 c1 3b cb.

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.

  1. Open the ChipWhisperer Analyzer software again and reopen the project file (if closed).
  2. 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.)
    image
  3. Resynchronize the traces, see the separate 'Preprocessing' tutorial (NB: only in slides right now!)

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!

A5 pp resync end.png

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:

# 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]))

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:

  1. Add the above function to your custom script file.
  2. Change the setAnalysisAlgorithm in the script to use your custom functions by making the following call:
    leak_model = AES128_8bit(AES256_Round13_Model)
  3. 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 setPointRange() function call to
    self.attack.setPointRange((8000,10990))
  4. Start the attack! Wait for the attack to complete, and you will determine the 13th round key:
    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 C6 BD 4E 50 AB CA 75 77 79 87 96 CA 1C 7F C5 82, if you recovered a different key replace the knownkey value with yours):

>>> 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

Our hard work has rewarded us with the 13th round key, which is c6 6a a6 12 4a ba 4d 04 4a 22 03 54 5b 28 0e 63.

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:

image

Open this tool and paste the 13/14 round keys, which are

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

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:

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

Peek into supersecret.h, 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

#!/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")

Appendix B: Capture Script

Note you need to manually CONNECT to the CW-Lite & AES Bootloader target before running this. To do this:

  1. Set the 'Scope Module' as 'ChipWhisperer/OpenADC'
  2. Set the 'Target Module' as 'AES Bootloader' (you need to have that target on your system)
"""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"

Appendix C: AES-256 14th Round Key Script

Full attack script, copy/paste into a file then run from within ChipWhisperer-Analyzer:

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, 3400))

self.results_table.setAnalysisSource(attack)
self.correlation_plot.setAnalysisSource(attack)
self.output_plot.setAnalysisSource(attack)
self.pge_plot.setAnalysisSource(attack)
attack.processTraces()

Appendix D: AES-256 13th Round Key Script

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.ref_points = (9100, 9300)
resync_traces.input_window = (8900, 9500)

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()