def build_graph(start_addr,
type_graph,
simplify=False,
dontmodstack=True,
loadint=False,
verbose=False):
machine = guess_machine(addr=start_addr)
dis_engine, ira = machine.dis_engine, machine.ira
class IRADelModCallStack(ira):
def call_effects(self, addr, instr):
assignblks, extra = super(IRADelModCallStack,
self).call_effects(addr, instr)
if not dontmodstack:
return assignblks, extra
out = []
for assignblk in assignblks:
dct = dict(assignblk)
dct = {
dst: src
for (dst, src) in dct.iteritems() if dst != self.sp
}
out.append(AssignBlock(dct, assignblk.instr))
return out, extra
if verbose:
print "Arch", dis_engine
fname = idc.GetInputFile()
if verbose:
print fname
bs = bin_stream_ida()
mdis = dis_engine(bs)
ir_arch = IRADelModCallStack(mdis.loc_db)
# populate symbols with ida names
for addr, name in idautils.Names():
if name is None:
continue
if (mdis.loc_db.get_offset_location(addr)
or mdis.loc_db.get_name_location(name)):
# Symbol alias
continue
mdis.loc_db.add_location(name, addr)
if verbose:
print "start disasm"
if verbose:
print hex(start_addr)
asmcfg = mdis.dis_multiblock(start_addr)
entry_points = set([mdis.loc_db.get_offset_location(start_addr)])
if verbose:
print "generating graph"
open('asm_flow.dot', 'w').write(asmcfg.dot())
print "generating IR... %x" % start_addr
ircfg = ir_arch.new_ircfg_from_asmcfg(asmcfg)
if verbose:
print "IR ok... %x" % start_addr
for irb in ircfg.blocks.itervalues():
irs = []
for assignblk in irb:
new_assignblk = {
expr_simp(dst): expr_simp(src)
for dst, src in assignblk.iteritems()
}
irs.append(AssignBlock(new_assignblk, instr=assignblk.instr))
ircfg.blocks[irb.loc_key] = IRBlock(irb.loc_key, irs)
if verbose:
out = ircfg.dot()
open(os.path.join(tempfile.gettempdir(), 'graph.dot'), 'wb').write(out)
title = "Miasm IR graph"
if simplify:
dead_simp(ir_arch, ircfg)
ircfg.simplify(expr_simp)
modified = True
while modified:
modified = False
modified |= dead_simp(ir_arch, ircfg)
modified |= remove_empty_assignblks(ircfg)
modified |= merge_blocks(ircfg, entry_points)
title += " (simplified)"
if type_graph == TYPE_GRAPH_IR:
graph = GraphMiasmIR(ircfg, title, None)
graph.Show()
return
head = list(entry_points)[0]
class IRAOutRegs(ira):
def get_out_regs(self, block):
regs_todo = super(IRAOutRegs, self).get_out_regs(block)
out = {}
for assignblk in block:
for dst in assignblk:
reg = self.ssa_var.get(dst, None)
if reg is None:
continue
if reg in regs_todo:
out[reg] = dst
return set(out.values())
# Add dummy dependency to uncover out regs affectation
for loc in ircfg.leaves():
irblock = ircfg.blocks.get(loc)
if irblock is None:
continue
regs = {}
for reg in ir_arch.get_out_regs(irblock):
regs[reg] = reg
assignblks = list(irblock)
new_assiblk = AssignBlock(regs, assignblks[-1].instr)
assignblks.append(new_assiblk)
new_irblock = IRBlock(irblock.loc_key, assignblks)
ircfg.blocks[loc] = new_irblock
ir_arch = IRAOutRegs(mdis.loc_db)
ir_arch.ssa_var = {}
modified = True
ssa_forbidden_regs = set(
[ir_arch.pc, ir_arch.IRDst, ir_arch.arch.regs.exception_flags])
head = list(entry_points)[0]
heads = set([head])
all_ssa_vars = {}
propagate_expr = PropagateExpr()
ssa = SSADiGraph(ircfg)
ssa.immutable_ids.update(ssa_forbidden_regs)
ssa.ssa_variable_to_expr.update(all_ssa_vars)
ssa.transform(head)
all_ssa_vars.update(ssa.ssa_variable_to_expr)
ir_arch.ssa_var.update(ssa.ssa_variable_to_expr)
if simplify:
while modified:
ssa = SSADiGraph(ircfg)
ssa.immutable_ids.update(ssa_forbidden_regs)
ssa.ssa_variable_to_expr.update(all_ssa_vars)
ssa.transform(head)
all_ssa_vars.update(ssa.ssa_variable_to_expr)
ir_arch.ssa_var.update(ssa.ssa_variable_to_expr)
while modified:
modified = False
modified |= propagate_expr.propagate(ssa, head)
modified |= ircfg.simplify(expr_simp)
simp_modified = True
while simp_modified:
simp_modified = False
simp_modified |= dead_simp(ir_arch, ircfg)
simp_modified |= remove_empty_assignblks(ircfg)
simp_modified |= load_from_int(ircfg, bs,
is_addr_ro_variable)
modified |= simp_modified
ssa = SSADiGraph(ircfg)
ssa.immutable_ids.update(ssa_forbidden_regs)
ssa.ssa_variable_to_expr.update(all_ssa_vars)
ssa.transform(head)
all_ssa_vars.update(ssa.ssa_variable_to_expr)
if type_graph == TYPE_GRAPH_IRSSA:
graph = GraphMiasmIR(ssa.graph, title, None)
graph.Show()
return
if type_graph == TYPE_GRAPH_IRSSAUNSSA:
cfg_liveness = DiGraphLivenessSSA(ssa.graph)
cfg_liveness.init_var_info(ir_arch)
cfg_liveness.compute_liveness()
UnSSADiGraph(ssa, head, cfg_liveness)
if simplify:
modified = True
while modified:
modified = False
modified |= ssa.graph.simplify(expr_simp)
simp_modified = True
while simp_modified:
simp_modified = False
simp_modified |= dead_simp(ir_arch, ssa.graph)
simp_modified |= remove_empty_assignblks(ssa.graph)
simp_modified |= merge_blocks(ssa.graph, heads)
modified |= simp_modified
graph = GraphMiasmIR(ssa.graph, title, None)
graph.Show()
if args.defuse:
reachings = ReachingDefinitions(ircfg_a)
open('graph_defuse.dot', 'w').write(DiGraphDefUse(reachings).dot())
out = ircfg.dot()
open('graph_irflow_raw.dot', 'w').write(out)
out = ircfg_a.dot()
open('graph_irflow.dot', 'w').write(out)
if args.simplify > 1:
ircfg_a.simplify(expr_simp)
modified = True
while modified:
modified = False
modified |= dead_simp(ir_arch_a, ircfg_a)
modified |= remove_empty_assignblks(ircfg_a)
modified |= merge_blocks(ircfg_a, entry_points)
open('graph_irflow_reduced.dot', 'w').write(ircfg_a.dot())
if args.ssa:
if len(entry_points) != 1:
raise RuntimeError("Your graph should have only one head")
head = list(entry_points)[0]
ssa = SSADiGraph(ircfg_a)
ssa.transform(head)
open("ssa.dot", "wb").write(ssa.graph.dot())
def build_graph(verbose=False, simplify=False, ssa=False, ssa_simplify=False):
start_addr = idc.ScreenEA()
machine = guess_machine(addr=start_addr)
dis_engine, ira = machine.dis_engine, machine.ira
if verbose:
print "Arch", dis_engine
fname = idc.GetInputFile()
if verbose:
print fname
bs = bin_stream_ida()
mdis = dis_engine(bs)
ir_arch = ira(mdis.loc_db)
# populate symbols with ida names
for addr, name in idautils.Names():
if name is None:
continue
if (mdis.loc_db.get_offset_location(addr)
or mdis.loc_db.get_name_location(name)):
# Symbol alias
continue
mdis.loc_db.add_location(name, addr)
if verbose:
print "start disasm"
if verbose:
print hex(addr)
asmcfg = mdis.dis_multiblock(start_addr)
entry_points = set([mdis.loc_db.get_offset_location(start_addr)])
if verbose:
print "generating graph"
open('asm_flow.dot', 'w').write(asmcfg.dot())
print "generating IR... %x" % start_addr
ircfg = ir_arch.new_ircfg_from_asmcfg(asmcfg)
if verbose:
print "IR ok... %x" % start_addr
for irb in ircfg.blocks.itervalues():
irs = []
for assignblk in irb:
new_assignblk = {
expr_simp(dst): expr_simp(src)
for dst, src in assignblk.iteritems()
}
irs.append(AssignBlock(new_assignblk, instr=assignblk.instr))
ircfg.blocks[irb.loc_key] = IRBlock(irb.loc_key, irs)
if verbose:
out = ircfg.dot()
open(os.path.join(tempfile.gettempdir(), 'graph.dot'), 'wb').write(out)
title = "Miasm IR graph"
if simplify:
dead_simp(ir_arch, ircfg)
ircfg.simplify(expr_simp)
modified = True
while modified:
modified = False
modified |= dead_simp(ir_arch, ircfg)
modified |= remove_empty_assignblks(ircfg)
modified |= merge_blocks(ircfg, entry_points)
title += " (simplified)"
graph = GraphMiasmIR(ircfg, title, None)
if ssa:
if len(entry_points) != 1:
raise RuntimeError("Your graph should have only one head")
head = list(entry_points)[0]
ssa = SSADiGraph(ircfg)
ssa.transform(head)
title += " (SSA)"
graph = GraphMiasmIR(ssa.graph, title, None)
if ssa_simplify:
class IRAOutRegs(ira):
def get_out_regs(self, block):
regs_todo = super(self.__class__, self).get_out_regs(block)
out = {}
for assignblk in block:
for dst in assignblk:
reg = self.ssa_var.get(dst, None)
if reg is None:
continue
if reg in regs_todo:
out[reg] = dst
return set(out.values())
# Add dummy dependency to uncover out regs affectation
for loc in ircfg.leaves():
irblock = ircfg.blocks.get(loc)
if irblock is None:
continue
regs = {}
for reg in ir_arch.get_out_regs(irblock):
regs[reg] = reg
assignblks = list(irblock)
new_assiblk = AssignBlock(regs, assignblks[-1].instr)
assignblks.append(new_assiblk)
new_irblock = IRBlock(irblock.loc_key, assignblks)
ircfg.blocks[loc] = new_irblock
ir_arch = IRAOutRegs(mdis.loc_db)
def is_addr_ro_variable(bs, addr, size):
"""
Return True if address at @addr is a read-only variable.
WARNING: Quick & Dirty
@addr: integer representing the address of the variable
@size: size in bits
"""
try:
_ = bs.getbytes(addr, size / 8)
except IOError:
return False
return True
ir_arch.ssa_var = {}
index = 0
modified = True
ssa_forbidden_regs = set(
[ir_arch.pc, ir_arch.IRDst, ir_arch.arch.regs.exception_flags])
head = list(entry_points)[0]
heads = set([head])
all_ssa_vars = set()
propagate_expr = PropagateExpr()
while modified:
ssa = SSADiGraph(ircfg)
ssa.immutable_ids.update(ssa_forbidden_regs)
ssa.transform(head)
all_ssa_vars.update(ssa._ssa_variable_to_expr)
ssa_regs = [reg for reg in ssa.expressions if reg.is_id()]
ssa_forbidden_regs.update(ssa_regs)
ir_arch.ssa_var.update(ssa._ssa_variable_to_expr)
while modified:
index += 1
modified = False
modified |= propagate_expr.propagate(ssa, head)
modified |= ircfg.simplify(expr_simp)
simp_modified = True
while simp_modified:
index += 1
simp_modified = False
simp_modified |= dead_simp(ir_arch, ircfg)
index += 1
simp_modified |= remove_empty_assignblks(ircfg)
simp_modified |= merge_blocks(ircfg, heads)
simp_modified |= load_from_int(ircfg, bs,
is_addr_ro_variable)
modified |= simp_modified
index += 1
merge_blocks(ircfg, heads)
ssa = SSADiGraph(ircfg)
ssa.immutable_ids.update(ssa_forbidden_regs)
ssa.transform(head)
all_ssa_vars.update(ssa._ssa_variable_to_expr)
ssa._ssa_variable_to_expr = all_ssa_vars
dead_simp(ir_arch, ssa.graph)
title += " (SSA Simplified)"
graph = GraphMiasmIR(ssa.graph, title, None)
graph.Show()
ircfg_a.simplify(expr_simp)
modified = True
while modified:
modified = False
modified |= dead_simp(ir_arch_a, ircfg_a)
modified |= remove_empty_assignblks(ircfg_a)
modified |= merge_blocks(ircfg_a, entry_points)
open('graph_irflow_reduced.dot', 'w').write(ircfg_a.dot())
if args.ssa and not args.propagexpr:
if len(entry_points) != 1:
raise RuntimeError("Your graph should have only one head")
head = list(entry_points)[0]
ssa = SSADiGraph(ircfg_a)
ssa.transform(head)
open("ssa.dot", "wb").write(ssa.graph.dot())
if args.propagexpr:
class IRAOutRegs(ira):
def get_out_regs(self, block):
regs_todo = super(self.__class__, self).get_out_regs(block)
out = {}
for assignblk in block:
for dst in assignblk:
reg = self.ssa_var.get(dst, None)
if reg is None:
continue
def build_graph(verbose=False, simplify=False, ssa=False, ssa_simplify=False):
start_addr = idc.ScreenEA()
machine = guess_machine(addr=start_addr)
dis_engine, ira = machine.dis_engine, machine.ira
if verbose:
print "Arch", dis_engine
fname = idc.GetInputFile()
if verbose:
print fname
bs = bin_stream_ida()
mdis = dis_engine(bs)
ir_arch = ira(mdis.loc_db)
# populate symbols with ida names
for addr, name in idautils.Names():
if name is None:
continue
if (mdis.loc_db.get_offset_location(addr) or
mdis.loc_db.get_name_location(name)):
# Symbol alias
continue
mdis.loc_db.add_location(name, addr)
if verbose:
print "start disasm"
if verbose:
print hex(addr)
asmcfg = mdis.dis_multiblock(start_addr)
entry_points = set([mdis.loc_db.get_offset_location(start_addr)])
if verbose:
print "generating graph"
open('asm_flow.dot', 'w').write(asmcfg.dot())
print "generating IR... %x" % start_addr
ircfg = ir_arch.new_ircfg_from_asmcfg(asmcfg)
if verbose:
print "IR ok... %x" % start_addr
for irb in ircfg.blocks.itervalues():
irs = []
for assignblk in irb:
new_assignblk = {
expr_simp(dst): expr_simp(src)
for dst, src in assignblk.iteritems()
}
irs.append(AssignBlock(new_assignblk, instr=assignblk.instr))
ircfg.blocks[irb.loc_key] = IRBlock(irb.loc_key, irs)
if verbose:
out = ircfg.dot()
open(os.path.join(tempfile.gettempdir(), 'graph.dot'), 'wb').write(out)
title = "Miasm IR graph"
if simplify:
dead_simp(ir_arch, ircfg)
ircfg.simplify(expr_simp)
modified = True
while modified:
modified = False
modified |= dead_simp(ir_arch, ircfg)
modified |= remove_empty_assignblks(ircfg)
modified |= merge_blocks(ircfg, entry_points)
title += " (simplified)"
graph = GraphMiasmIR(ircfg, title, None)
if ssa:
if len(entry_points) != 1:
raise RuntimeError("Your graph should have only one head")
head = list(entry_points)[0]
ssa = SSADiGraph(ircfg)
ssa.transform(head)
title += " (SSA)"
graph = GraphMiasmIR(ssa.graph, title, None)
if ssa_simplify:
class IRAOutRegs(ira):
def get_out_regs(self, block):
regs_todo = super(self.__class__, self).get_out_regs(block)
out = {}
for assignblk in block:
for dst in assignblk:
reg = self.ssa_var.get(dst, None)
if reg is None:
continue
if reg in regs_todo:
out[reg] = dst
return set(out.values())
# Add dummy dependecy to uncover out regs affectation
for loc in ircfg.leaves():
irblock = ircfg.blocks.get(loc)
if irblock is None:
continue
regs = {}
for reg in ir_arch.get_out_regs(irblock):
regs[reg] = reg
assignblks = list(irblock)
new_assiblk = AssignBlock(regs, assignblks[-1].instr)
assignblks.append(new_assiblk)
new_irblock = IRBlock(irblock.loc_key, assignblks)
ircfg.blocks[loc] = new_irblock
ir_arch = IRAOutRegs(mdis.loc_db)
def is_addr_ro_variable(bs, addr, size):
"""
Return True if address at @addr is a read-only variable.
WARNING: Quick & Dirty
@addr: integer representing the address of the variable
@size: size in bits
"""
try:
_ = bs.getbytes(addr, size/8)
except IOError:
return False
return True
ir_arch.ssa_var = {}
index = 0
modified = True
ssa_forbidden_regs = set([
ir_arch.pc,
ir_arch.IRDst,
ir_arch.arch.regs.exception_flags
])
head = list(entry_points)[0]
heads = set([head])
all_ssa_vars = set()
propagate_expr = PropagateExpr()
while modified:
ssa = SSADiGraph(ircfg)
ssa.immutable_ids.update(ssa_forbidden_regs)
ssa.transform(head)
all_ssa_vars.update(ssa._ssa_variable_to_expr)
ssa_regs = [reg for reg in ssa.expressions if reg.is_id()]
ssa_forbidden_regs.update(ssa_regs)
ir_arch.ssa_var.update(ssa._ssa_variable_to_expr)
while modified:
index += 1
modified = False
modified |= propagate_expr.propagate(ssa, head)
modified |= ircfg.simplify(expr_simp)
simp_modified = True
while simp_modified:
index += 1
simp_modified = False
simp_modified |= dead_simp(ir_arch, ircfg)
index += 1
simp_modified |= remove_empty_assignblks(ircfg)
simp_modified |= merge_blocks(ircfg, heads)
simp_modified |= load_from_int(ircfg, bs, is_addr_ro_variable)
modified |= simp_modified
index += 1
merge_blocks(ircfg, heads)
ssa = SSADiGraph(ircfg)
ssa.immutable_ids.update(ssa_forbidden_regs)
ssa.transform(head)
all_ssa_vars.update(ssa._ssa_variable_to_expr)
ssa._ssa_variable_to_expr = all_ssa_vars
dead_simp(ir_arch, ssa.graph)
title += " (SSA Simplified)"
graph = GraphMiasmIR(ssa.graph, title, None)
graph.Show()
ircfg_a.simplify(expr_simp)
modified = True
while modified:
modified = False
modified |= dead_simp(ir_arch_a, ircfg_a)
modified |= remove_empty_assignblks(ircfg_a)
modified |= merge_blocks(ircfg_a, entry_points)
open('graph_irflow_reduced.dot', 'w').write(ircfg_a.dot())
if args.ssa and not args.propagexpr:
if len(entry_points) != 1:
raise RuntimeError("Your graph should have only one head")
head = list(entry_points)[0]
ssa = SSADiGraph(ircfg_a)
ssa.transform(head)
open("ssa.dot", "wb").write(ssa.graph.dot())
if args.propagexpr:
class IRAOutRegs(ira):
def get_out_regs(self, block):
regs_todo = super(self.__class__, self).get_out_regs(block)
out = {}
for assignblk in block:
for dst in assignblk: