Angr

支持 HackTricks

本备忘单的部分内容基于 angr 文档

安装

sudo apt-get install python3-dev libffi-dev build-essential
python3 -m pip install --user virtualenv
python3 -m venv ang
source ang/bin/activate
pip install angr

基本操作

import angr
import monkeyhex # this will format numerical results in hexadecimal
#Load binary
proj = angr.Project('/bin/true')

#BASIC BINARY DATA
proj.arch #Get arch "<Arch AMD64 (LE)>"
proj.arch.name #'AMD64'
proj.arch.memory_endness #'Iend_LE'
proj.entry #Get entrypoint "0x4023c0"
proj.filename #Get filename "/bin/true"

#There are specific options to load binaries
#Usually you won't need to use them but you could
angr.Project('examples/fauxware/fauxware', main_opts={'backend': 'blob', 'arch': 'i386'}, lib_opts={'libc.so.6': {'backend': 'elf'}})

加载和主对象信息

加载的数据

#LOADED DATA
proj.loader #<Loaded true, maps [0x400000:0x5004000]>
proj.loader.min_addr #0x400000
proj.loader.max_addr #0x5004000
proj.loader.all_objects #All loaded
proj.loader.shared_objects #Loaded binaries
"""
OrderedDict([('true', <ELF Object true, maps [0x400000:0x40a377]>),
('libc.so.6',
<ELF Object libc-2.31.so, maps [0x500000:0x6c4507]>),
('ld-linux-x86-64.so.2',
<ELF Object ld-2.31.so, maps [0x700000:0x72c177]>),
('extern-address space',
<ExternObject Object cle##externs, maps [0x800000:0x87ffff]>),
('cle##tls',
<ELFTLSObjectV2 Object cle##tls, maps [0x900000:0x91500f]>)])
"""
proj.loader.all_elf_objects #Get all ELF objects loaded (Linux)
proj.loader.all_pe_objects #Get all binaries loaded (Windows)
proj.loader.find_object_containing(0x400000)#Get object loaded in an address "<ELF Object fauxware, maps [0x400000:0x60105f]>"

主要目标

#Main Object (main binary loaded)
obj = proj.loader.main_object #<ELF Object true, maps [0x400000:0x60721f]>
obj.execstack #"False" Check for executable stack
obj.pic #"True" Check PIC
obj.imports #Get imports
obj.segments #<Regions: [<ELFSegment flags=0x5, relro=0x0, vaddr=0x400000, memsize=0xa74, filesize=0xa74, offset=0x0>, <ELFSegment flags=0x4, relro=0x1, vaddr=0x600e28, memsize=0x1d8, filesize=0x1d8, offset=0xe28>, <ELFSegment flags=0x6, relro=0x0, vaddr=0x601000, memsize=0x60, filesize=0x50, offset=0x1000>]>
obj.find_segment_containing(obj.entry) #Get segment by address
obj.sections #<Regions: [<Unnamed | offset 0x0, vaddr 0x0, size 0x0>, <.interp | offset 0x238, vaddr 0x400238, size 0x1c>, <.note.ABI-tag | offset 0x254, vaddr 0x400254, size 0x20>, <.note.gnu.build-id ...
obj.find_section_containing(obj.entry) #Get section by address
obj.plt['strcmp'] #Get plt address of a funcion (0x400550)
obj.reverse_plt[0x400550] #Get function from plt address ('strcmp')

符号和重定位

strcmp = proj.loader.find_symbol('strcmp') #<Symbol "strcmp" in libc.so.6 at 0x1089cd0>

strcmp.name #'strcmp'
strcmp.owne #<ELF Object libc-2.23.so, maps [0x1000000:0x13c999f]>
strcmp.rebased_addr #0x1089cd0
strcmp.linked_addr #0x89cd0
strcmp.relative_addr #0x89cd0
strcmp.is_export #True, as 'strcmp' is a function exported by libc

#Get strcmp from the main object
main_strcmp = proj.loader.main_object.get_symbol('strcmp')
main_strcmp.is_export #False
main_strcmp.is_import #True
main_strcmp.resolvedby #<Symbol "strcmp" in libc.so.6 at 0x1089cd0>

#Blocks
block = proj.factory.block(proj.entry) #Get the block of the entrypoint fo the binary
block.pp() #Print disassembly of the block
block.instructions #"0xb" Get number of instructions
block.instruction_addrs #Get instructions addresses "[0x401670, 0x401672, 0x401675, 0x401676, 0x401679, 0x40167d, 0x40167e, 0x40167f, 0x401686, 0x40168d, 0x401694]"

动态分析

模拟管理器,状态

#Live States
#This is useful to modify content in a live analysis
state = proj.factory.entry_state()
state.regs.rip #Get the RIP
state.mem[proj.entry].int.resolved #Resolve as a C int (BV)
state.mem[proj.entry].int.concreteved #Resolve as python int
state.regs.rsi = state.solver.BVV(3, 64) #Modify RIP
state.mem[0x1000].long = 4 #Modify mem

#Other States
project.factory.entry_state()
project.factory.blank_state() #Most of its data left uninitialized
project.factory.full_init_statetate() #Execute through any initializers that need to be run before the main binary's entry point
project.factory.call_state() #Ready to execute a given function.

#Simulation manager
#The simulation manager stores all the states across the execution of the binary
simgr = proj.factory.simulation_manager(state) #Start
simgr.step() #Execute one step
simgr.active[0].regs.rip #Get RIP from the last state

调用函数

  • 你可以通过 args 传递参数列表,通过 env 传递环境变量字典到 entry_statefull_init_state。这些结构中的值可以是字符串或位向量,并将被序列化为模拟执行的参数和环境。默认的 args 是一个空列表,因此如果你分析的程序期望至少找到一个 argv[0],你应该始终提供它!

  • 如果你希望 argc 是符号的,可以将一个符号位向量作为 argc 传递给 entry_statefull_init_state 构造函数。不过要小心:如果这样做,你还应该向结果状态添加一个约束,确保你的 argc 值不能大于你传递给 args 的参数数量。

  • 要使用调用状态,你应该使用 .call_state(addr, arg1, arg2, ...) 调用它,其中 addr 是你想要调用的函数的地址,argN 是该函数的第 N 个参数,可以是 Python 整数、字符串、数组或位向量。如果你想分配内存并实际传递一个指向对象的指针,你应该将其包装在 PointerWrapper 中,即 angr.PointerWrapper("point to me!")。这个 API 的结果可能有点不可预测,但我们正在努力改进它。

位向量

#BitVectors
state = proj.factory.entry_state()
bv = state.solver.BVV(0x1234, 32) #Create BV of 32bits with the value "0x1234"
state.solver.eval(bv) #Convert BV to python int
bv.zero_extend(30) #Will add 30 zeros on the left of the bitvector
bv.sign_extend(30) #Will add 30 zeros or ones on the left of the BV extending the sign

符号位向量与约束

x = state.solver.BVS("x", 64) #Symbolic variable BV of length 64
y = state.solver.BVS("y", 64)

#Symbolic oprations
tree = (x + 1) / (y + 2)
tree #<BV64 (x_9_64 + 0x1) / (y_10_64 + 0x2)>
tree.op #'__floordiv__' Access last operation
tree.args #(<BV64 x_9_64 + 0x1>, <BV64 y_10_64 + 0x2>)
tree.args[0].op #'__add__' Access of dirst arg
tree.args[0].args #(<BV64 x_9_64>, <BV64 0x1>)
tree.args[0].args[1].op #'BVV'
tree.args[0].args[1].args #(1, 64)

#Symbolic constraints solver
state = proj.factory.entry_state() #Get a fresh state without constraints
input = state.solver.BVS('input', 64)
operation = (((input + 4) * 3) >> 1) + input
output = 200
state.solver.add(operation == output)
state.solver.eval(input) #0x3333333333333381
state.solver.add(input < 2**32)
state.satisfiable() #False

#Solver solutions
solver.eval(expression) #one possible solution
solver.eval_one(expression) #solution to the given expression, or throw an error if more than one solution is possible.
solver.eval_upto(expression, n) #n solutions to the given expression, returning fewer than n if fewer than n are possible.
solver.eval_atleast(expression, n) #n solutions to the given expression, throwing an error if fewer than n are possible.
solver.eval_exact(expression, n) #n solutions to the given expression, throwing an error if fewer or more than are possible.
solver.min(expression) #minimum possible solution to the given expression.
solver.max(expression) #maximum possible solution to the given expression.

钩子

>>> stub_func = angr.SIM_PROCEDURES['stubs']['ReturnUnconstrained'] # this is a CLASS
>>> proj.hook(0x10000, stub_func())  # hook with an instance of the class

>>> proj.is_hooked(0x10000)            # these functions should be pretty self-explanitory
True
>>> proj.hooked_by(0x10000)
<ReturnUnconstrained>
>>> proj.unhook(0x10000)

>>> @proj.hook(0x20000, length=5)
... def my_hook(state):
...     state.regs.rax = 1

>>> proj.is_hooked(0x20000)
True

此外,您可以使用 proj.hook_symbol(name, hook),将符号的名称作为第一个参数,以挂钩符号所在的地址。

示例

支持 HackTricks

Last updated