Skip to content

Latest commit

 

History

History
149 lines (130 loc) · 5.8 KB

week9.md

File metadata and controls

149 lines (130 loc) · 5.8 KB

Week 9

Day58: 学习UW CSE501 静态分析课程

  • 优化选项:
    • dead code elimination
    • partial redundancy elimination
    • function inlining
    • strength reduction
    • loop transformations
    • constant propagation
  • special edge:
    • back edge: 指向一个之前遍历过的block
    • critical edge: 既不是唯一离开source的边, 也不是唯一进入到target的边
  • dataflow framework: <G, L, F, M>
    • G = flow graph
    • L = (semi-)lattice
    • F/M = flow / transfer functions
  • reaching definition:
    • dataflow equations:
      • IN[b] = OUT[b1] U ... U OUT[bn]
      • OUT[b] = (IN[b] - KILL[b]) U GEN[b]
      • IN[entry] = 0000000
    • solving equations:
    Input: flow graph (CFG)
    // boundary condition
    OUT[Entry] = 0...0
    // initial conditions
    for each basic block B other than entry
     OUT[B] = 0...0
    // iterate
    while (any out[] changes value) {
     for each basic block B other than entry {
     IN[B] = U (OUT[p]), for all predecessor block p of B
     OUT[B] = (IN[B] – KILL[B]) U GEN[B]
     }
    }
    
  • live variable
    • transfer function for live variable:
      • x = y + z
      • generates new live variable: USE[s] = {y, z}
      • kills previously live variable: DEF[s] = x
      • variables that were not killed are propagated: OUT[s] - DEF[s]
      • so: IN[s] = USE[s] | (OUT[s] - DEF[s])
    • setup
      • boundary condition: IN[exit] = None
      • initial conditions: IN[B] = None
      • meet operation: OUT[B] = | IN[Successors]
  • Must Reach: a definition D must reach a program point P if
    • D appears at least once along all paths that leads to P
    • D is not redefined along any path after the last appearance of D and before P
  • constant propagation: lattice
    • undefined: variable has not been initialized
    • NAC: variable definitely has a value( we just don't known what )
    • meet rules:
      • constant & constant = constant (if equal)
      • constant & constant = NAC (if not equal)
      • constant & undefined = constant
      • constant & NAC = NAC
  • maximal fixed point
  • meet over paths: 可能是无穷个

Day59: 学习UW CSE501 指针分析

  • 应用:

    • 别名分析: 确定两个指针是否都指向相同的内存区域
    • 编译优化
    • 并行: 将串行代码转换成并行代码
    • shape analysis: 找到堆上数据结构的属性
    • 检测内存问题: 泄漏, 空指针引用等安全问题
  • Point Language:

    • assume x and y are pointers
    • y = &x -> means y points to x
    • y = x -> means if x points to z then y points to z
    • *y = x -> means if y points to z and z is a pointer, and if x points to w then z now points to w
    • y = *x -> means if x points to z and z is a pointer, and if z points to w then y not points to w
    • points-to(x): set of variables that pointer variable x may point to
  • Andersen as graph closure

    • one node for each memory location
    • each node contains a points-to set
    • solve equations by computing transitive closure of graph, and add edges according to constraints
  • worklist algorithm

W = { nodes with non-empty points-to sets } while W is not empty { v = choose from W for each constraint v in x add edge x -> v, and add x to W if edge is new for each a in points-to(v) do { for each constraint p in *v add edge a -> p, and add a to W if edge is new for each constraint *v in q add edge q -> a, and add q to W if edge is new } for each edge v -> q do { points-to(q) = points-to(q) | points-to(v), and add q to W if points-to(q) changed } } ```

Day60-61: 二进制相似度聚类Golang实现

  • 使用binding连接radare2
  • 获取程序的字符串信息
  • 对字符串进行Base64编码后计算SHA256
  • 通过radare2获取二进制的基本块数据
  • 使用capstone将二进制数据转换成汇编代码
  • 拿到汇编代码后, 生成基本块
  • 拿到基本块后, 生成基本的控制流
  • 对基本块进行简单的符号化(去除偏移和立即数等)

Day62-63: 学习MOBISEC安全课程

  • 04 - Intro to Android Architecture and Security

Binder是Android用于RPC和进程间通信的机制, Android利用Binder来将普通进程内调用的API转换到特权进程/服务实现的特权API.

Android系统启动完毕后会广播一个带 ACTION_BOOT_COMPLETED action的 Intent, 因此app可以通过接收改该 Intent 来做开机启动, 也就可以用于持久化

SYSTEM_ALERT_WINDOW: 可以在其他APP上显示一个窗口, 这会导致许多UI界面的攻击, 比如UI混淆, 点击劫持, 钓鱼等.

  • 05 - Real-World Android Apps

sharedUserId安全问题: 相同证书的APP可以申请使用相同的Linux User ID, 而具有相同的Linux User Id可以共享该ID的所有内容, 也可以访问彼此的内部隐私存储和其他组件等.

  • 08 - Reverse Engineering

Android逆向方法流:

  1. 大概了解app的功能: 模拟器里启动app, 观察初始的UI
  2. 找到app的攻击面:
    • 从入口点开始入手做攻击面分析
    • 检查app的各项组件(activities, broadcast, intent, receiver), 这些组件是否暴露给外部的app使用
    • 检查app如何与外部进行交互, 比如文件系统, 网络, 组件间通信等.
  3. app如何跟网络端点交互
    • 寻找网络端点的IP, URL等. 虽然有可能经过混淆
    • 寻找网络相关API的调用代码
    • 在模拟器里运行并监视其网络活动.
  4. app是如何存储隐私信息
    • 隐私信息包括有, 用户帐号证书, 用户隐私数据, 需要安全权限才能访问的数据等.
  5. 检查某个函数是否存在滥用
    • app是怎么使用函数X的
    • app是否有安全地使用该函数
    • 攻击者该如何到达该函数?