Last active
December 10, 2015 14:28
-
-
Save zhuowei/4447343 to your computer and use it in GitHub Desktop.
An attempt to deobfuscate Minecraft Pi by identifying VTables using the readable typeinfo embedded in the executable
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
"""Identify vtables from a C++ ELF application. Tailored for Minecraft-pi. | |
Requires the pyelftools library from https://bitbucket.org/eliben/pyelftools | |
Copyright (C) 2013 Zhuowei Zhang | |
This program 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 2 | |
of the License, or (at your option) any later version. | |
This program 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 General Public License for more details. | |
You should have received a copy of the GNU General Public License | |
along with this program; if not, write to the Free Software | |
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.""" | |
from __future__ import print_function | |
from sys import argv | |
import struct | |
from elftools.elf.elffile import * | |
from elftools.common.utils import * | |
baseClassList = ["Entity", "Tile", "GameMode", "Screen"] | |
def parseTypeInfoName(lastString): | |
# Typeinfo are formatted in the format: (ASCII int with length)TypeName | |
#e.g. 3Pig | |
#Thanks, https://gist.github.com/2174255 | |
digits = [] | |
try: | |
for a in range(0, len(lastString)): | |
b = lastString[a] | |
if not b.isdigit(): | |
strLength = len(lastString) - a | |
supposedLength = int("".join(digits)) | |
#print(strLength, supposedLength) | |
if strLength == supposedLength and supposedLength > 1: | |
return lastString[a:] | |
else: | |
return None | |
else: | |
digits.append(b) | |
except: | |
return None | |
def findTypeInfoNames(mcElfFile): | |
dataSection = mcElfFile.get_section_by_name(".rodata") | |
print("rodata section found:", dataSection) | |
dataStream = dataSection.data(); | |
#print(dataStream) | |
lastString = "" | |
stringBeginLoc = -1; | |
typeInfos = {} | |
dataBegin = dataSection['sh_offset'] | |
dataAddr = dataSection['sh_addr'] | |
print(hex(dataBegin), hex(dataAddr)) | |
dataDiff = dataAddr - dataBegin | |
print("data diff is:", dataDiff) | |
for a in range(0, len(dataStream)): | |
b = dataStream[a] | |
c = ord(b) | |
#print(b) | |
if c == 0 and len(lastString) > 0: | |
#print(lastString) | |
typeInfo = parseTypeInfoName(lastString) | |
if typeInfo != None: | |
typeInfos[stringBeginLoc + dataBegin + dataDiff] = typeInfo | |
print(hex((stringBeginLoc + dataBegin + dataDiff)), typeInfo) | |
lastString = "" | |
elif c >= 32 and c <= 126: | |
if len(lastString) == 0: | |
stringBeginLoc = a | |
lastString = lastString + b | |
else: | |
lastString = "" | |
return typeInfos | |
def findTypeinfoPointers(mcElfFile, typeInfoNames): | |
dataSection = mcElfFile.get_section_by_name(".rodata") | |
print("Rodata section found:", dataSection) | |
dataStream = dataSection.data() | |
dataBegin = dataSection['sh_offset'] | |
dataAddr = dataSection['sh_addr'] | |
dataDiff = dataAddr - dataBegin | |
print("data diff is:", dataDiff) | |
typeInfos = {} | |
vtables = [] | |
print(len(dataStream)) | |
print("<" + str(len(dataStream) / 4) + "I") | |
allIntegers = struct.unpack("<" + str(len(dataStream) / 4) + "Ixx", dataStream) | |
# print(allIntegers) | |
for loc in range(0, len(allIntegers)): | |
i = allIntegers[loc] | |
typeInfo = None | |
#print(hex((loc * 4) + dataBegin), hex(i)) | |
try: | |
typeInfo = typeInfoNames[i] | |
except KeyError: | |
continue | |
if typeInfo != None: | |
print("Found one typeinfo!", hex(i), hex((loc * 4) + dataBegin - 4), typeInfo) | |
typeInfos[(loc * 4) + dataBegin - 4 + dataDiff] = typeInfo | |
# print(typeInfos) | |
for loc in range(0, len(allIntegers)): | |
i = allIntegers[loc] | |
typeInfo = None | |
try: | |
typeInfo = typeInfos[i] | |
except KeyError: | |
continue | |
if typeInfo != None and (allIntegers[loc - 2] == 0 and allIntegers[loc - 1] == 0): | |
print("Found one vtable!", hex(i), typeInfo) | |
stillReading = True | |
vtableOffset = loc | |
vtableData = [0] | |
while stillReading: | |
vtableVal = allIntegers[vtableOffset] | |
if (vtableVal == 0): | |
break | |
vtableData.append(vtableVal) | |
vtableOffset = vtableOffset + 1 | |
vtables.append({"type": typeInfo, "data": vtableData, "offset": ((loc - 1) * 4)}) | |
# print(vtables) | |
for x in vtables: | |
print(hex(x["offset"] + dataBegin), x["type"]) | |
return vtables | |
def generateNewName(type, i): | |
return type + "::func_" + str(i); | |
def findBaseClassName(existingName, newName): | |
existingClassName = existingName[0:existingName.find("::")] | |
if existingClassName in baseClassList: | |
return existingName | |
else: | |
return newName | |
def generateMethodDict(vtables, textBegin, textEnd, dataBegin): | |
methodDict = {} | |
vcopy = vtables[:] | |
for vtable in vtables: | |
for i in range(1, len(vtable["data"])): | |
functionPointer = vtable["data"][i] | |
if functionPointer < textBegin or functionPointer > textEnd: | |
print("Found invalid vtable entry: ", hex(functionPointer), vtable["type"], hex(vtable["offset"] + dataBegin)) | |
vcopy.remove(vtable) | |
break | |
#print(hex(functionPointer), i, vtable["type"]) | |
existingNames = None | |
try: | |
existingNames = methodDict[functionPointer] | |
except KeyError: | |
existingNames = [] | |
methodDict[functionPointer] = existingNames | |
newName = generateNewName(vtable["type"], i) | |
#if len(existingNames) > 0: | |
# newName = findBaseClassName(existingNames, newName) | |
existingNames.append(newName) | |
for x in vcopy: | |
print(hex(x["offset"] + dataBegin), x["type"]) | |
return methodDict | |
def annotateDecomp(mcElfFile, decompInput, decompOutput, vtables): | |
textSection = mcElfFile.get_section_by_name(".text") | |
textBegin = textSection['sh_addr'] | |
textEnd = textBegin + textSection['sh_size'] | |
dataSection = mcElfFile.get_section_by_name(".rodata") | |
dataBegin = dataSection['sh_offset'] | |
dataAddr = dataSection['sh_addr'] | |
dataEnd = dataAddr + dataSection['sh_size'] | |
print(hex(textBegin), hex(textEnd)) | |
methodDict = generateMethodDict(vtables, textBegin, dataEnd, dataBegin) | |
for line in decompInput: | |
try: | |
addr = int(line[0:line.find(":\t")].lstrip(), 16) | |
methodNames = methodDict[addr] | |
#print(hex(addr)) | |
print("\n{0:8x} <{1}>".format(addr, ", ".join(methodNames)), file=decompOutput) | |
except KeyError: | |
pass | |
except ValueError: | |
pass | |
decompOutput.write(line) | |
#print(methodDict) | |
vtables = None | |
if len(argv) != 4: | |
print("Usage: python identvtable.py minecraft-pi originaldecomp.txt annotateddecomp.txt") | |
exit() | |
with open(argv[1], "rb") as mcExecutable: | |
mcElfFile = ELFFile(mcExecutable) | |
typeInfos = findTypeInfoNames(mcElfFile) | |
#print(typeInfos) | |
vtables = findTypeinfoPointers(mcElfFile, typeInfos) | |
with open(argv[2], "r") as decompInput: | |
with open(argv[3], "w") as decompOutput: | |
annotateDecomp(mcElfFile, decompInput, decompOutput, vtables) | |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment