class Disassembler():
def __init__(self, filename, raw_type, raw_base, raw_big_endian):
import capstone as CAPSTONE
self.code = {}
self.binary = Binary(filename, raw_type, raw_base, raw_big_endian)
arch, mode = self.binary.get_arch()
if arch is None or mode is None:
raise ExcArch(self.binary.get_arch_string())
self.binary.load_extra()
self.md = CAPSTONE.Cs(arch, mode)
self.md.detail = True
self.arch = arch
self.mode = mode
def check_addr(self, ctx, addr):
addr_exists, is_exec = self.binary.check_addr(addr)
if not ctx.print_data and not is_exec:
raise ExcNotExec(addr)
if not addr_exists:
raise ExcNotAddr(addr)
def load_arch_module(self):
import capstone as CAPSTONE
if self.arch == CAPSTONE.CS_ARCH_X86:
import lib.arch.x86 as ARCH
elif self.arch == CAPSTONE.CS_ARCH_ARM:
import lib.arch.arm as ARCH
elif self.arch == CAPSTONE.CS_ARCH_MIPS:
import lib.arch.mips as ARCH
else:
raise NotImplementedError
return ARCH
def get_addr_from_string(self, opt_addr, raw=False):
if opt_addr is None:
if raw:
return 0
search = ["main", "_main"]
else:
search = [opt_addr]
for s in search:
if s.startswith("0x"):
a = int(opt_addr, 16)
else:
a = self.binary.symbols.get(s, -1)
if a != -1:
return a
raise ExcSymNotFound(search[0])
def print_section_meta(self, name, start, end):
print_no_end(color_section(name.ljust(20)))
print_no_end(" [")
print_no_end(hex(start))
print_no_end(" - ")
print_no_end(hex(end))
print("]")
def dump_asm(self, ctx, lines):
from capstone import CS_OP_IMM
ARCH = self.load_arch_module()
ARCH_UTILS = ARCH.utils
ARCH_OUTPUT = ARCH.output
s_name, s_start, s_end = self.binary.get_section_meta(ctx.entry_addr)
self.print_section_meta(s_name, s_start, s_end)
# WARNING: this assume that on every architectures the jump
# address is the last operand (operands[-1])
# set jumps color
ad = ctx.entry_addr
l = 0
while l < lines and ad < s_end:
i = self.lazy_disasm(ad, s_start)
if i is None:
ad += 1
else:
if ARCH_UTILS.is_jump(i) and i.operands[-1].type == CS_OP_IMM:
pick_color(i.operands[-1].value.imm)
ad += i.size
l += 1
# Here we have loaded all instructions we want to print
if self.binary.type == T_BIN_PE:
self.binary.pe_reverse_stripped_symbols(self)
o = ARCH_OUTPUT.Output(ctx)
# dump
ad = ctx.entry_addr
l = 0
while l < lines and ad < s_end:
i = self.lazy_disasm(ad, s_start)
if i is None:
ad += 1
o.print_bad(ad)
else:
o.print_inst(i)
ad += i.size
l += 1
def dump_data(self, ctx, lines):
N = 128
addr = ctx.entry_addr
s_name, s_start, s_end = self.binary.get_section_meta(ctx.entry_addr)
self.print_section_meta(s_name, s_start, s_end)
addr_ascii_str = -1
ascii_str = []
l = 0
is_new_line = True
while l < lines:
buf = self.binary.section_stream_read(addr, N)
if not buf:
break
i = 0
while i < len(buf):
c = buf[i]
if c in BYTES_PRINTABLE_SET:
if addr_ascii_str == -1:
addr_ascii_str = addr
ascii_str.append(c)
addr += 1
i += 1
continue
stack_char = []
if addr_ascii_str != -1:
if c == 0 and len(ascii_str) >= 2:
if not is_new_line:
print()
l += 1
if l >= lines:
return
print_no_end(color_addr(addr_ascii_str))
print_no_end(color_string("\"" + "".join(map(get_char, ascii_str)) + "\""))
print(", 0")
is_new_line = True
l += 1
if l >= lines:
return
ascii_str = []
addr_ascii_str = -1
addr += 1
i += 1
continue
stack_char = ascii_str
i -= len(ascii_str)
addr -= len(ascii_str)
stack_char.append(c)
for c in stack_char:
if is_new_line:
print_no_end(color_addr(addr))
elif addr % 4 == 0 and addr != ctx.entry_addr:
print()
is_new_line = True
l += 1
if l >= lines:
return
print_no_end(color_addr(addr))
print_no_end("%.2x " % c)
addr += 1
i += 1
is_new_line = False
ascii_str = []
addr_ascii_str = -1
if not is_new_line:
print()
def print_calls(self, ctx):
ARCH = self.load_arch_module()
ARCH_UTILS = ARCH.utils
ARCH_OUTPUT = ARCH.output
s_name, s_start, s_end = self.binary.get_section_meta(ctx.entry_addr)
self.print_section_meta(s_name, s_start, s_end)
o = ARCH_OUTPUT.Output(ctx)
ad = s_start
while ad < s_end:
i = self.lazy_disasm(ad, s_start)
if i is None:
ad += 1
else:
ad += i.size
if ARCH_UTILS.is_call(i):
o.print_inst(i)
def print_symbols(self, print_sections, sym_filter=None):
if sym_filter is not None:
sym_filter = sym_filter.lower()
for addr in self.binary.reverse_symbols:
sy = self.binary.reverse_symbols[addr]
if sym_filter is None or sym_filter in sy.lower():
sec_name, _ = self.binary.is_address(addr)
if sy:
print_no_end(color_addr(addr) + " " +
color_symbol("<" + sy + ">"))
if print_sections and sec_name is not None:
print_no_end(" (" + color_section(sec_name) + ")")
print()
def load_user_sym_file(self, fd):
for l in fd:
arg = l.split()
addr = int(arg[0], 16)
self.binary.reverse_symbols[addr] = arg[1]
self.binary.symbols[arg[1]] = addr
def lazy_disasm(self, addr, stay_in_section=-1):
meta = self.binary.get_section_meta(addr)
if meta is None:
return None
_, start, _ = meta
if stay_in_section != -1 and start != stay_in_section:
return None
if addr in self.code:
return self.code[addr]
# Disassemble by block of N bytes
N = 1024
d = self.binary.section_stream_read(addr, N)
gen = self.md.disasm(d, addr)
first = None
try:
first = next(gen)
self.code[first.address] = first
# Max N instructions (N is in bytes)
for n in range(N):
i = next(gen)
if i.address in self.code:
return first
self.code[i.address] = i
except StopIteration:
pass
return first
def __prefetch_inst(self, inst):
return self.lazy_disasm(inst.address + inst.size)
# Generate a flow graph of the given function (addr)
def get_graph(self, addr):
from capstone import CS_OP_IMM, CS_ARCH_MIPS
ARCH_UTILS = self.load_arch_module().utils
curr = self.lazy_disasm(addr)
if curr == None:
return None
gph = Graph(self, addr)
rest = []
start = time.clock()
prefetch = None
# WARNING: this assume that on every architectures the jump
# address is the last operand (operands[-1])
while 1:
if not gph.exists(curr):
if self.arch == CS_ARCH_MIPS:
prefetch = self.__prefetch_inst(curr)
if ARCH_UTILS.is_uncond_jump(curr) and len(curr.operands) > 0:
if curr.operands[-1].type == CS_OP_IMM:
addr = curr.operands[-1].value.imm
nxt = self.lazy_disasm(addr)
gph.set_next(curr, nxt, prefetch)
rest.append(nxt.address)
else:
# Can't interpret jmp ADDR|reg
gph.add_node(curr, prefetch)
gph.uncond_jumps_set.add(curr.address)
elif ARCH_UTILS.is_cond_jump(curr) and len(curr.operands) > 0:
if curr.operands[-1].type == CS_OP_IMM:
nxt_jump = self.lazy_disasm(curr.operands[-1].value.imm)
if self.arch == CS_ARCH_MIPS:
direct_nxt = \
self.lazy_disasm(prefetch.address + prefetch.size)
else:
direct_nxt = \
self.lazy_disasm(curr.address + curr.size)
gph.set_cond_next(curr, nxt_jump, direct_nxt, prefetch)
rest.append(nxt_jump.address)
rest.append(direct_nxt.address)
else:
# Can't interpret jmp ADDR|reg
gph.add_node(curr, prefetch)
gph.cond_jumps_set.add(curr.address)
elif ARCH_UTILS.is_ret(curr):
gph.add_node(curr, prefetch)
else:
try:
nxt = self.lazy_disasm(curr.address + curr.size)
gph.set_next(curr, nxt)
rest.append(nxt.address)
except:
gph.add_node(curr)
pass
try:
curr = self.lazy_disasm(rest.pop())
except IndexError:
break
if self.binary.type == T_BIN_PE:
self.binary.pe_reverse_stripped_symbols(self)
elapsed = time.clock()
elapsed = elapsed - start
debug__("Graph built in %fs" % elapsed)
return gph
class Disassembler():
def __init__(self, filename, raw_type, raw_base, raw_big_endian):
import capstone as CAPSTONE
self.code = {}
self.binary = Binary(filename, raw_type, raw_base, raw_big_endian)
arch, mode = self.binary.get_arch()
if arch is None or mode is None:
raise ExcArch(self.binary.get_arch_string())
self.binary.load_extra()
self.md = CAPSTONE.Cs(arch, mode)
self.md.detail = True
self.arch = arch
self.mode = mode
def check_addr(self, addr):
addr_exists, is_exec = self.binary.check_addr(addr)
if not is_exec:
raise ExcNotExec(addr)
if not addr_exists:
raise ExcNotAddr(addr)
def load_arch_module(self):
import capstone as CAPSTONE
if self.arch == CAPSTONE.CS_ARCH_X86:
import lib.arch.x86 as ARCH
elif self.arch == CAPSTONE.CS_ARCH_ARM:
import lib.arch.arm as ARCH
elif self.arch == CAPSTONE.CS_ARCH_MIPS:
import lib.arch.mips as ARCH
else:
raise NotImplementedError
return ARCH
def get_addr_from_string(self, opt_addr, raw=False):
if opt_addr is None:
if raw:
return 0
search = ["main", "_main"]
else:
search = [opt_addr]
for s in search:
if s.startswith("0x"):
a = int(opt_addr, 16)
else:
a = self.binary.symbols.get(s, -1)
if a != -1:
return a
raise ExcSymNotFound(search[0])
def print_section_meta(self, name, start, end):
print_no_end(color_section(name))
print_no_end(" [")
print_no_end(hex(start))
print_no_end(" - ")
print_no_end(hex(end))
print("]")
def dump(self, ctx, lines):
from capstone import CS_OP_IMM
ARCH = self.load_arch_module()
ARCH_UTILS = ARCH.utils
ARCH_OUTPUT = ARCH.output
s_name, s_start, s_end = self.binary.get_section_meta(ctx.addr)
self.print_section_meta(s_name, s_start, s_end)
# WARNING: this assume that on every architectures the jump
# address is the last operand (operands[-1])
# set jumps color
i = self.lazy_disasm(ctx.addr, s_start)
l = 0
while i is not None and l < lines:
if ARCH_UTILS.is_jump(i) and i.operands[-1].type == CS_OP_IMM:
pick_color(i.operands[-1].value.imm)
i = self.lazy_disasm(i.address + i.size, s_start)
l += 1
# Here we have loaded all instructions we want to print
if self.binary.type == T_BIN_PE:
self.binary.pe_reverse_stripped_symbols(self)
o = ARCH_OUTPUT.Output(ctx)
# dump
i = self.lazy_disasm(ctx.addr, s_start)
l = 0
while i is not None and l < lines:
o.print_inst(i)
i = self.lazy_disasm(i.address + i.size, s_start)
l += 1
def print_calls(self, ctx):
# Print all calls which are in the section containing ctx.addr
ARCH = self.load_arch_module()
ARCH_UTILS = ARCH.utils
ARCH_OUTPUT = ARCH.output
s_name, s_start, s_end = self.binary.get_section_meta(ctx.addr)
self.print_section_meta(s_name, s_start, s_end)
o = ARCH_OUTPUT.Output(ctx)
i = self.lazy_disasm(s_start, s_start)
while i is not None:
if ARCH_UTILS.is_call(i):
o.print_inst(i)
i = self.lazy_disasm(i.address + i.size, s_start)
def print_symbols(self, print_sections, sym_filter=None):
if sym_filter is not None:
sym_filter = sym_filter.lower()
for addr in self.binary.reverse_symbols:
sy = self.binary.reverse_symbols[addr]
if sym_filter is None or sym_filter in sy.lower():
sec_name, _ = self.binary.is_address(addr)
if sy:
print_no_end(color_addr(addr) + " " +
color_symbol("<" + sy + ">"))
if print_sections and sec_name is not None:
print_no_end(" (" + color_section(sec_name) + ")")
print()
def load_user_sym_file(self, fd):
for l in fd:
arg = l.split()
addr = int(arg[0], 16)
self.binary.reverse_symbols[addr] = arg[1]
self.binary.symbols[arg[1]] = addr
def lazy_disasm(self, addr, stay_in_section=-1):
_, start, _ = self.binary.get_section_meta(addr)
if stay_in_section != -1 and start != stay_in_section:
return None
if addr in self.code:
return self.code[addr]
# Disassemble by block of N bytes
N = 1024
d = self.binary.section_stream_read(addr, N)
gen = self.md.disasm(d, addr)
first = None
try:
first = next(gen)
self.code[first.address] = first
# Max N instructions (N is in bytes)
for n in range(N):
i = next(gen)
if i.address in self.code:
return first
self.code[i.address] = i
except StopIteration:
pass
return first
# Generate a flow graph of the given function (addr)
def get_graph(self, addr):
from capstone import CS_OP_IMM
ARCH_UTILS = self.load_arch_module().utils
curr = self.lazy_disasm(addr)
if curr == None:
return None
gph = Graph(self, addr)
rest = []
start = time.clock()
# WARNING: this assume that on every architectures the jump
# address is the last operand (operands[-1])
while 1:
if not gph.exists(curr):
if ARCH_UTILS.is_uncond_jump(curr) and len(curr.operands) > 0:
if curr.operands[-1].type == CS_OP_IMM:
addr = curr.operands[-1].value.imm
nxt = self.lazy_disasm(addr)
gph.set_next(curr, nxt)
rest.append(nxt.address)
else:
# Can't interpret jmp ADDR|reg
gph.add_node(curr)
gph.uncond_jumps_set.add(curr.address)
elif ARCH_UTILS.is_cond_jump(curr) and len(curr.operands) > 0:
if curr.operands[-1].type == CS_OP_IMM:
nxt_jump = self.lazy_disasm(curr.operands[-1].value.imm)
direct_nxt = self.lazy_disasm(curr.address + curr.size)
gph.set_cond_next(curr, nxt_jump, direct_nxt)
rest.append(nxt_jump.address)
rest.append(direct_nxt.address)
else:
# Can't interpret jmp ADDR|reg
gph.add_node(curr)
gph.cond_jumps_set.add(curr.address)
elif ARCH_UTILS.is_ret(curr):
gph.add_node(curr)
else:
try:
nxt = self.lazy_disasm(curr.address + curr.size)
gph.set_next(curr, nxt)
rest.append(nxt.address)
except:
gph.add_node(curr)
pass
try:
curr = self.lazy_disasm(rest.pop())
except IndexError:
break
if self.binary.type == T_BIN_PE:
self.binary.pe_reverse_stripped_symbols(self)
elapsed = time.clock()
elapsed = elapsed - start
debug__("Graph built in %fs" % elapsed)
return gph