-
-
[原创]关于爆破迷宫路径的一系列思考
-
发表于: 2024-6-8 20:20
-
前几天学长给了一道maze的题，动态调试过程可谓非常艰难(对我现在的水平而言)，于是学长提供了一种用subprocess的方法
首先通过字符串定位到主函数那里，我们发现cmp那里其实是一个长度校验，判断路径的长度是不是为140
这里为判断路径方向
中间过程太复杂，我们直接跳到后面，我们发现他判断所处位置是不是为0，还有判断最后到达位置对不对的
经过分析我们并不知道迷宫在哪，因此我们才要尝试爆破的方法。
首先第一步，我们要把长度校验给patch掉
重点的patch在后面，我们为了一步一步进行爆破，我们要对每一步进行爆破
但是有一个问题，后面不管走不走对都会输出
我们可以先写一个简单的试验一下
我们发现根本无法判断是不是走对了，这个时候我们再patch一下
把下面那个指向wrong的地址改为指向' what()'的地址
我们这个时候可以再看一下
发现有不一样的回显了，我们根据这个可以写一个脚本，但是这里有一个问题，它在程序里面每走一步不会把走过的给标记，所以我们得自己标记一下，由于没有写这个标记，导致脚本跑了好久也没跑出来。
差不多二十秒的时间跑出来
我自己的虚拟环境是kali 2023.3 内存 4G 处理器2x2
如果师傅们有更好的思路欢迎来交流
https://docs.python.org/zh-cn/3/library/subprocess.html#using-the-subprocess-module
https://www.cnblogs.com/lordtianqiyi/articles/16588649.html
https://bbs.kanxue.com/thread-281796.html
import subprocess
printable=['w','s','a','d']
for i in range(len(printable)):
sh=subprocess.Popen("./mapmap", shell=True, stdout=subprocess.PIPE, stdin=subprocess.PIPE)
sh.stdin.write((printable[i]+'\n').encode())
sh.stdin.flush()
output ,_ =sh.communicate()
print(output)
import subprocess
printable=['w','s','a','d']
for i in range(len(printable)):
sh=subprocess.Popen("./mapmap", shell=True, stdout=subprocess.PIPE, stdin=subprocess.PIPE)
sh.stdin.write((printable[i]+'\n').encode())
sh.stdin.flush()
output ,_ =sh.communicate()
print(output)
import subprocess
def dfs(path, n, x_location, y_location, count, try_direction, visited_paths, coordinates):
if n < count:
for direction in try_direction:
process = subprocess.Popen("./mapmap", shell=True, stdout=subprocess.PIPE, stdin=subprocess.PIPE)
next_path = path + direction
process.stdin.write((next_path + '\n').encode())
process.stdin.flush()
output, _ = process.communicate()
if b'wrong\n' not in output:
new_x_location, new_y_location = x_location, y_location
if direction == 'd':
new_x_location += 1
elif direction == 'a':
new_x_location -= 1
elif direction == 'w':
new_y_location -= 1
elif direction == 's':
new_y_location += 1
if (new_x_location, new_y_location) not in coordinates:
print(path)
coordinates[(new_x_location, new_y_location)] = next_path
dfs(next_path, len(next_path), new_x_location, new_y_location, count, try_direction, visited_paths, coordinates)
elif n == count:
visited_paths.append(path)
coordinates[(x_location, y_location)] = path
# DFS startif __name__ == '__main__':
start_path = ''
try_direction = ['w', 'd', 's', 'a']
count = 140
visited_paths = []
coordinates = {}
x_location = 0
y_location = 0
coordinates[(x_location, y_location)] = start_path
print("\nVisit paths:")
dfs(start_path, len(start_path), x_location, y_location, count, try_direction, visited_paths, coordinates)
print("\nFinal paths:")
for path in visited_paths:
process = subprocess.Popen("./mapmap", shell=True, stdout=subprocess.PIPE, stdin=subprocess.PIPE)
process.stdin.write((path + '\n').encode())
process.stdin.flush()
output, _ = process.communicate()
if b'flag{md5(your input)}\n' in output:
print(path)
import subprocess
def dfs(path, n, x_location, y_location, count, try_direction, visited_paths, coordinates):
if n < count:
for direction in try_direction:
process = subprocess.Popen("./mapmap", shell=True, stdout=subprocess.PIPE, stdin=subprocess.PIPE)
next_path = path + direction
process.stdin.write((next_path + '\n').encode())
process.stdin.flush()
output, _ = process.communicate()
if b'wrong\n' not in output:
new_x_location, new_y_location = x_location, y_location
if direction == 'd':
new_x_location += 1
elif direction == 'a':
[招生]科锐逆向工程师培训(2024年11月15日实地，远程教学同时开班, 第51期)
最后于 2024-6-8 21:48
被5m10v3编辑
，原因：
