-
-
Save bruce30262/0c0bfd51e193cc631d78f1505bf805a4 to your computer and use it in GitHub Desktop.
[flareon2018 leve12] tiny disassembler for subleq VM
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
#!/usr/bin/env python | |
# -*- coding: utf-8 -*- | |
from pwn import * | |
from collections import OrderedDict | |
import Queue | |
import json | |
import inst | |
import sys | |
q = Queue.Queue() | |
data = "" | |
code = dict() | |
# dump1: the subleq vm | |
with open("./dump1", "rb") as f: | |
data = f.read() | |
assert len(data) == 0x8400 | |
def Word(addr): | |
global data | |
return u16(data[addr:addr+2:]) | |
def add_next_pos(addr): | |
global q | |
if addr not in code: | |
q.put(addr) | |
def subleq(now, f1, f2, f3): | |
global q, code | |
now &= 0xffff | |
j = dict() | |
idx1 = 0x223 + ((f1<<1)&0xffff) | |
idx1 &= 0xffff | |
idx2 = 0x223 + ((f2<<1)&0xffff) | |
idx2 &= 0xffff | |
s = "subleq [{:#x}] [{:#x}] [{:#x}]".format(idx1, idx2, f3+0x223) | |
j["op1"] = idx1 | |
j["op2"] = idx2 | |
if f3: | |
next_pos = (((now+2)<<1)&0xffff) + 0x223 | |
next_pos &= 0xffff | |
if idx1 == idx2: | |
j['jmp_type'] = "unc" # unconditional jmp | |
j['jmp'] = Word(next_pos) | |
s += "; JMP {:#x}".format(j['jmp']) | |
j['asm'] = s | |
code[now] = json.dumps(j) | |
add_next_pos(Word(next_pos)) | |
else: | |
j['jmp_type'] = "con" # conditional | |
j['jmp'] = (Word(next_pos), now+3) # true, false | |
s += "; if [{:#x}] <= 0; JMP {:#x}; else {:#x}".format(idx2, ((Word(next_pos)<<1)+0x223)&0xffff, (((now+3)<<1)+0x223)&0xffff) | |
j['asm'] = s | |
code[now] = json.dumps(j) | |
add_next_pos(Word(next_pos)) | |
add_next_pos(now+3) | |
else: | |
j['jmp_type'] = "next" # unconditional jmp | |
j['asm'] = s | |
code[now] = json.dumps(j) | |
add_next_pos(now+3) | |
# starting point of the disassembler | |
start = int(sys.argv[1], 16) | |
q.put(start) | |
while not q.empty(): | |
now = q.get() | |
cur = (now<<1)&0xffff | |
f1 = Word((0x223+cur)&0xffff) | |
f2 = Word((0x223+cur+2)&0xffff) | |
f3 = Word((0x223+cur+4)&0xffff) | |
subleq(now, f1, f2, f3) | |
od = OrderedDict(sorted(code.items())) | |
dis = [] | |
black = [0xff, 0x102, 0x105] # black list of the disassembling address | |
for pos, code in od.iteritems(): | |
if pos not in black: | |
dis.append((pos, code)) | |
token, rev_token = dict(), dict() | |
cur_token = None | |
Z = '0x223' | |
def trans(code): | |
""" get token for each subleq inst""" | |
def get_token(op): | |
global token, cur_token, rev_token | |
ret = None | |
if op == 0x223: | |
return Z | |
if op not in token: | |
ret = chr(cur_token) | |
token[op] = ret | |
rev_token[ret] = op | |
cur_token += 1 | |
else: | |
ret = token[op] | |
return ret | |
tk3 = None | |
tk1, tk2 = get_token(code["op1"]), get_token(code['op2']) | |
if code["jmp_type"] == "unc": | |
tk3 = get_token(code['jmp']) | |
if not tk3: | |
return (tk1, tk2) | |
else: | |
return (tk1, tk2, tk3) | |
now = 0 | |
while now < len(dis): | |
matched = False | |
for ins in inst.instruction: # traverse all instructions | |
sz = ins.sz | |
token.clear() | |
rev_token.clear() | |
cur_token = ord('a') | |
patt = [] | |
if now+sz-1 >= len(dis): continue | |
for idx in xrange(sz): | |
code = json.loads(dis[now+idx][1]) | |
patt.append(trans(code)) | |
if (ins.check(patt)): | |
print("{:#x}: {}".format((0x223+(dis[now][0]<<1))&0xffff, ins.get_inst(rev_token))) | |
matched = True | |
now += sz | |
break | |
if not matched: | |
print("{:#x}: {}".format((0x223+(dis[now][0]<<1))&0xffff, json.loads(dis[now][1])["asm"])) | |
now += 1 |
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
#!/usr/bin/env python | |
# -*- coding: utf-8 -*- | |
a = "a" | |
Z = "0x223" | |
b = "b" | |
c = "c" | |
class Inst(object): | |
def __init__(self, name, patt, sz): | |
""" | |
name: instruction name | |
patt: subleq pattern ( an array ) | |
sz: subleq instruction number | |
""" | |
self.name = name | |
self.patt = patt | |
self.sz = sz | |
def check(self, input_patt): | |
""" | |
check if a input pattern ( an array ) equals this instruction | |
""" | |
res = input_patt == self.patt | |
#print input_patt | |
#print self.patt | |
return res | |
class Jmp(Inst): | |
def __init__(self, patt): | |
super(Jmp, self).__init__("JMP", patt, 1) | |
def get_inst(self, rev_token): | |
jmp_addr = rev_token["a"] | |
return "JMP {:#x}".format((0x223+(jmp_addr<<1))&0xffff) | |
class Move(Inst): | |
def __init__(self, patt): | |
super(Move, self).__init__("MOVE", patt, 4) | |
def get_inst(self, rev_token): | |
dst = rev_token["a"] | |
src = rev_token["b"] | |
return "MOVE [{:#x}], [{:#x}]".format(dst, src) | |
class CLR(Inst): | |
def __init__(self, patt): | |
super(CLR, self).__init__("CLR", patt, 4) | |
def get_inst(self, rev_token): | |
dst = rev_token["a"] | |
return "CLR [{:#x}]".format(dst) | |
class Add(Inst): | |
def __init__(self, patt): | |
super(Add, self).__init__("ADD", patt, 3) | |
def get_inst(self, rev_token): | |
dst = rev_token["b"] | |
src = rev_token["a"] | |
return "ADD [{:#x}], [{:#x}]".format(dst, src) | |
instruction = [] | |
instruction.append(Move([(a, a), (b, Z), (Z, a), (Z, Z)])) | |
#instruction.append(CLR([(a, a), (Z, Z), (Z, a), (Z, Z)])) | |
instruction.append(Add([(a, Z), (Z, b), (Z, Z)])) | |
instruction.append(Jmp([(Z, Z, a)])) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment