`angr` — Analysis and Coordination

Project

angr.project.load_shellcode(shellcode, arch, start_offset=0, load_address=0, thumb=False, **kwargs)

Load a new project based on a snippet of assembly or bytecode.

Parameters

shellcode – The data to load, as either a bytestring of instructions or a string of assembly text
arch – The name of the arch to use, or an archinfo class
start_offset – The offset into the data to start analysis (default 0)
load_address – The address to place the data in memory (default 0)
thumb – Whether this is ARM Thumb shellcode

class angr.project.Project(thing, default_analysis_mode=None, ignore_functions=None, use_sim_procedures=True, exclude_sim_procedures_func=None, exclude_sim_procedures_list=(), arch=None, simos=None, engine=None, load_options: Optional[Dict[str, Any]] = None, translation_cache=True, support_selfmodifying_code=False, store_function=None, load_function=None, analyses_preset=None, concrete_target=None, **kwargs)

Bases: object

This is the main class of the angr module. It is meant to contain a set of binaries and the relationships between them, and perform analyses on them.

Parameters: thing – The path to the main executable object to analyze, or a CLE Loader object.

The following parameters are optional.

Parameters

default_analysis_mode – The mode of analysis to use by default. Defaults to ‘symbolic’.
ignore_functions – A list of function names that, when imported from shared libraries, should never be stepped into in analysis (calls will return an unconstrained value).
use_sim_procedures – Whether to replace resolved dependencies for which simprocedures are available with said simprocedures.
exclude_sim_procedures_func – A function that, when passed a function name, returns whether or not to wrap it with a simprocedure.
exclude_sim_procedures_list – A list of functions to not wrap with simprocedures.
arch – The target architecture (auto-detected otherwise).
simos – a SimOS class to use for this project.
engine – The SimEngine class to use for this project.
translation_cache (bool) – If True, cache translated basic blocks rather than re-translating them.
support_selfmodifying_code (bool) – Whether we aggressively support self-modifying code. When enabled, emulation will try to read code from the current state instead of the original memory, regardless of the current memory protections.
store_function – A function that defines how the Project should be stored. Default to pickling.
load_function – A function that defines how the Project should be loaded. Default to unpickling.
analyses_preset (angr.misc.PluginPreset) – The plugin preset for the analyses provider (i.e. Analyses instance).

Any additional keyword arguments passed will be passed onto cle.Loader.

Variables

analyses – The available analyses.
entry – The program entrypoint.
factory – Provides access to important analysis elements such as path groups and symbolic execution results.
filename – The filename of the executable.
loader – The program loader.
storage – Dictionary of things that should be loaded/stored with the Project.

arch: archinfo.arch.Arch

analyses: angr.analyses.analysis.AnalysesHub

hook(addr, hook=None, length=0, kwargs=None, replace=False)

Hook a section of code with a custom function. This is used internally to provide symbolic summaries of library functions, and can be used to instrument execution or to modify control flow.

When hook is not specified, it returns a function decorator that allows easy hooking. Usage:

# Assuming proj is an instance of angr.Project, we will add a custom hook at the entry
# point of the project.
@proj.hook(proj.entry)
def my_hook(state):
    print("Welcome to execution!")

Parameters

addr – The address to hook.
hook – A angr.project.Hook describing a procedure to run at the given address. You may also pass in a SimProcedure class or a function directly and it will be wrapped in a Hook object for you.
length – If you provide a function for the hook, this is the number of bytes that will be skipped by executing the hook by default.
kwargs – If you provide a SimProcedure for the hook, these are the keyword arguments that will be passed to the procedure’s run method eventually.
replace – Control the behavior on finding that the address is already hooked. If true, silently replace the hook. If false (default), warn and do not replace the hook. If none, warn and replace the hook.

is_hooked(addr) → bool

Returns True if addr is hooked.

Parameters: addr – An address.
Returns: True if addr is hooked, False otherwise.

hooked_by(addr) → Optional[angr.sim_procedure.SimProcedure]

Returns the current hook for addr.

Parameters: addr – An address.
Returns: None if the address is not hooked.

unhook(addr)

Remove a hook.

Parameters: addr – The address of the hook.

hook_symbol(symbol_name, simproc, kwargs=None, replace=None)

Resolve a dependency in a binary. Looks up the address of the given symbol, and then hooks that address. If the symbol was not available in the loaded libraries, this address may be provided by the CLE externs object.

Additionally, if instead of a symbol name you provide an address, some secret functionality will kick in and you will probably just hook that address, UNLESS you’re on powerpc64 ABIv1 or some yet-unknown scary ABI that has its function pointers point to something other than the actual functions, in which case it’ll do the right thing.

Parameters

symbol_name – The name of the dependency to resolve.
simproc – The SimProcedure instance (or function) with which to hook the symbol
kwargs – If you provide a SimProcedure for the hook, these are the keyword arguments that will be passed to the procedure’s run method eventually.
replace – Control the behavior on finding that the address is already hooked. If true, silently replace the hook. If false, warn and do not replace the hook. If none (default), warn and replace the hook.

Returns

The address of the new symbol.

Return type

int

is_symbol_hooked(symbol_name)

Check if a symbol is already hooked.

Parameters: symbol_name (str) – Name of the symbol.
Returns: True if the symbol can be resolved and is hooked, False otherwise.
Return type: bool

unhook_symbol(symbol_name): Remove the hook on a symbol. This function will fail if the symbol is provided by the extern object, as that would result in a state where analysis would be unable to cope with a call to this symbol.

rehook_symbol(new_address, symbol_name, stubs_on_sync): Move the hook for a symbol to a specific address :param new_address: the new address that will trigger the SimProc execution :param symbol_name: the name of the symbol (f.i. strcmp ) :return: None

execute(*args, **kwargs)

This function is a symbolic execution helper in the simple style supported by triton and manticore. It designed to be run after setting up hooks (see Project.hook), in which the symbolic state can be checked.

This function can be run in three different ways:

When run with no parameters, this function begins symbolic execution from the entrypoint.

It can also be run with a “state” parameter specifying a SimState to begin symbolic execution from.

Finally, it can accept any arbitrary keyword arguments, which are all passed to project.factory.full_init_state.

If symbolic execution finishes, this function returns the resulting simulation manager.

terminate_execution(): Terminates a symbolic execution that was started with Project.execute().

class angr.factory.AngrObjectFactory(project, default_engine=None)

Bases: object

This factory provides access to important analysis elements.

snippet(addr, jumpkind=None, **block_opts)

successors(*args, engine=None, **kwargs)

Perform execution using an engine. Generally, return a SimSuccessors object classifying the results of the run.

Parameters

state – The state to analyze
engine – The engine to use. If not provided, will use the project default.
addr – optional, an address to execute at instead of the state’s ip
jumpkind – optional, the jumpkind of the previous exit
inline – This is an inline execution. Do not bother copying the state.

Additional keyword arguments will be passed directly into each engine’s process method.

blank_state(**kwargs)

Returns a mostly-uninitialized state object. All parameters are optional.

Parameters

addr – The address the state should start at instead of the entry point.
initial_prefix – If this is provided, all symbolic registers will hold symbolic values with names prefixed by this string.
fs – A dictionary of file names with associated preset SimFile objects.
concrete_fs – bool describing whether the host filesystem should be consulted when opening files.
chroot – A path to use as a fake root directory, Behaves similarly to a real chroot. Used only when concrete_fs is set to True.
kwargs – Any additional keyword args will be passed to the SimState constructor.

Returns

The blank state.

Return type

SimState

entry_state(**kwargs) → angr.sim_state.SimState

Returns a state object representing the program at its entry point. All parameters are optional.

Parameters

addr – The address the state should start at instead of the entry point.
initial_prefix – If this is provided, all symbolic registers will hold symbolic values with names prefixed by this string.
fs – a dictionary of file names with associated preset SimFile objects.
concrete_fs – boolean describing whether the host filesystem should be consulted when opening files.
chroot – a path to use as a fake root directory, behaves similar to a real chroot. used only when concrete_fs is set to True.
argc – a custom value to use for the program’s argc. May be either an int or a bitvector. If not provided, defaults to the length of args.
args – a list of values to use as the program’s argv. May be mixed strings and bitvectors.
env – a dictionary to use as the environment for the program. Both keys and values may be mixed strings and bitvectors.

Returns

The entry state.

Return type

SimState

full_init_state(**kwargs)

Very much like entry_state(), except that instead of starting execution at the program entry point, execution begins at a special SimProcedure that plays the role of the dynamic loader, calling each of the initializer functions that should be called before execution reaches the entry point.

It can take any of the arguments that can be provided to entry_state, except for addr.

call_state(addr, *args, **kwargs)

Returns a state object initialized to the start of a given function, as if it were called with given parameters.

Parameters

addr – The address the state should start at instead of the entry point.
args – Any additional positional arguments will be used as arguments to the function call.

The following parametrs are optional.

Parameters

base_state – Use this SimState as the base for the new state instead of a blank state.
cc – Optionally provide a SimCC object to use a specific calling convention.
ret_addr – Use this address as the function’s return target.
stack_base – An optional pointer to use as the top of the stack, circa the function entry point
alloc_base – An optional pointer to use as the place to put excess argument data
grow_like_stack – When allocating data at alloc_base, whether to allocate at decreasing addresses
toc – The address of the table of contents for ppc64
initial_prefix – If this is provided, all symbolic registers will hold symbolic values with names prefixed by this string.
fs – A dictionary of file names with associated preset SimFile objects.
concrete_fs – bool describing whether the host filesystem should be consulted when opening files.
chroot – A path to use as a fake root directory, Behaves similarly to a real chroot. Used only when concrete_fs is set to True.
kwargs – Any additional keyword args will be passed to the SimState constructor.

Returns

The state at the beginning of the function.

Return type

SimState

The idea here is that you can provide almost any kind of python type in args and it’ll be translated to a binary format to be placed into simulated memory. Lists (representing arrays) must be entirely elements of the same type and size, while tuples (representing structs) can be elements of any type and size. If you’d like there to be a pointer to a given value, wrap the value in a SimCC.PointerWrapper. Any value that can’t fit in a register will be automatically put in a PointerWrapper.

If stack_base is not provided, the current stack pointer will be used, and it will be updated. If alloc_base is not provided, the current stack pointer will be used, and it will be updated. You might not like the results if you provide stack_base but not alloc_base.

grow_like_stack controls the behavior of allocating data at alloc_base. When data from args needs to be wrapped in a pointer, the pointer needs to point somewhere, so that data is dumped into memory at alloc_base. If you set alloc_base to point to somewhere other than the stack, set grow_like_stack to False so that sequencial allocations happen at increasing addresses.

simulation_manager(thing: Optional[Union[List[angr.sim_state.SimState], angr.sim_state.SimState]] = None, **kwargs) → angr.sim_manager.SimulationManager

Constructs a new simulation manager.

Parameters

thing – Optional - What to put in the new SimulationManager’s active stash (either a SimState or a list of SimStates).
kwargs – Any additional keyword arguments will be passed to the SimulationManager constructor

Returns

The new SimulationManager

Return type

angr.sim_manager.SimulationManager

Many different types can be passed to this method:

If nothing is passed in, the SimulationManager is seeded with a state initialized for the program entry point, i.e. entry_state().
If a SimState is passed in, the SimulationManager is seeded with that state.
If a list is passed in, the list must contain only SimStates and the whole list will be used to seed the SimulationManager.

simgr(*args, **kwargs): Alias for simulation_manager to save our poor fingers

callable(addr, prototype=None, concrete_only=False, perform_merge=True, base_state=None, toc=None, cc=None)

A Callable is a representation of a function in the binary that can be interacted with like a native python function.

Parameters

addr – The address of the function to use
prototype – The prototype of the call to use, as a string or a SimTypeFunction
concrete_only – Throw an exception if the execution splits into multiple states
perform_merge – Merge all result states into one at the end (only relevant if concrete_only=False)
base_state – The state from which to do these runs
toc – The address of the table of contents for ppc64
cc – The SimCC to use for a calling convention

Returns

A Callable object that can be used as a interface for executing guest code like a python function.

Return type

angr.callable.Callable

cc()

Return a SimCC (calling convention) parametrized for this project.

Relevant subclasses of SimFunctionArgument are SimRegArg and SimStackArg, and shortcuts to them can be found on this cc object.

For stack arguments, offsets are relative to the stack pointer on function entry.

block(addr: int, size=None, max_size=None, byte_string=None, vex=None, thumb=False, backup_state=None, extra_stop_points=None, opt_level=None, num_inst=None, traceflags=0, insn_bytes=None, insn_text=None, strict_block_end=None, collect_data_refs=False, cross_insn_opt=True, load_from_ro_regions=False, initial_regs=None) → Block
block(addr: archinfo.arch_soot.SootAddressDescriptor, size=None, max_size=None, byte_string=None, vex=None, thumb=False, backup_state=None, extra_stop_points=None, opt_level=None, num_inst=None, traceflags=0, insn_bytes=None, insn_text=None, strict_block_end=None, collect_data_refs=False, cross_insn_opt=True) → SootBlock

fresh_block(addr, size, backup_state=None)

class angr.block.DisassemblerBlock(addr, insns, thumb, arch)

Bases: object

Helper class to represent a block of dissassembled target architecture instructions

addr

insns

thumb

arch

pp()

class angr.block.DisassemblerInsn

Bases: object

Helper class to represent a disassembled target architecture instruction

property size: int

property address: int

property mnemonic: str

property op_str: str

class angr.block.CapstoneBlock(addr, insns, thumb, arch)

Bases: angr.block.DisassemblerBlock

Deep copy of the capstone blocks, which have serious issues with having extended lifespans outside of capstone itself

class angr.block.CapstoneInsn(capstone_insn)

Bases: angr.block.DisassemblerInsn

Represents a capstone instruction.

insn

property size: int

property address: int

property mnemonic: str

property op_str: str

class angr.block.Block(addr, project=None, arch=None, size=None, byte_string=None, vex=None, thumb=False, backup_state=None, extra_stop_points=None, opt_level=None, num_inst=None, traceflags=0, strict_block_end=None, collect_data_refs=False, cross_insn_opt=True, load_from_ro_regions=False, initial_regs=None)

Bases: angr.serializable.Serializable

Represents a basic block in a binary or a program.

BLOCK_MAX_SIZE = 4096

arch

thumb

addr

size

pp(**kwargs)

set_initial_regs()

static reset_initial_regs()

property vex: pyvex.block.IRSB

property vex_nostmt

property disassembly: angr.block.DisassemblerBlock: Provide a disassembly object using whatever disassembler is available

property capstone

property codenode

property bytes

property instructions

property instruction_addrs

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg)

class angr.block.SootBlock(addr, project=None, arch=None)

Bases: object

Represents a Soot IR basic block.

property soot

property size

property codenode

Plugin Ecosystem

class angr.misc.plugins.PluginHub

Bases: object

A plugin hub is an object which contains many plugins, as well as the notion of a “preset”, or a backer that can provide default implementations of plugins which cater to a certain circumstance.

Objects in angr like the SimState, the Analyses hub, the SimEngine selector, etc all use this model to unify their mechanisms for automatically collecting and selecting components to use. If you’re familiar with design patterns this is a configurable Strategy Pattern.

Each PluginHub subclass should have a corresponding Plugin subclass, and perhaps a PluginPreset subclass if it wants its presets to be able to specify anything more interesting than a list of defaults.

classmethod register_default(name, plugin_cls, preset='default')

classmethod register_preset(name, preset): Register a preset instance with the class of the hub it corresponds to. This allows individual plugin objects to automatically register themselves with a preset by using a classmethod of their own with only the name of the preset to register with.

property plugin_preset: Get the current active plugin preset

property has_plugin_preset: bool: Check whether or not there is a plugin preset in use on this hub right now

use_plugin_preset(preset)

Apply a preset to the hub. If there was a previously active preset, discard it.

Preset can be either the string name of a preset or a PluginPreset instance.

discard_plugin_preset(): Discard the current active preset. Will release any active plugins that could have come from the old preset.

get_plugin(name: str) → SimStatePlugin: Get the plugin named name. If no such plugin is currently active, try to activate a new one using the current preset.

has_plugin(name): Return whether or not a plugin with the name name is curently active.

register_plugin(name: str, plugin): Add a new plugin plugin with name name to the active plugins.

release_plugin(name): Deactivate and remove the plugin with name name.

class angr.misc.plugins.PluginPreset

Bases: object

A plugin preset object contains a mapping from name to a plugin class. A preset can be active on a hub, which will cause it to handle requests for plugins which are not already present on the hub.

Unlike Plugins and PluginHubs, instances of PluginPresets are defined on the module level for individual presets. You should register the preset instance with a hub to allow plugins to easily add themselves to the preset without an explicit reference to the preset itself.

activate(hub): This method is called when the preset becomes active on a hub.

deactivate(hub): This method is called when the preset is discarded from the hub.

add_default_plugin(name, plugin_cls): Add a plugin to the preset.

list_default_plugins(): Return a list of the names of available default plugins.

request_plugin(name: str) → Type[SimStatePlugin]: Return the plugin class which is registered under the name name, or raise NoPlugin if the name isn’t available.

copy(): Return a copy of self.

class angr.misc.plugins.PluginVendor

Bases: angr.misc.plugins.PluginHub

A specialized hub which serves only as a plugin vendor, never having any “active” plugins. It will directly return the plugins provided by the preset instead of instanciating them.

release_plugin(name)

register_plugin(name, plugin)

class angr.misc.plugins.VendorPreset

Bases: angr.misc.plugins.PluginPreset

A specialized preset class for use with the PluginVendor.

Program State

angr.sim_state.arch_overrideable(f)

class angr.sim_state.SimState(project=None, arch=None, plugins=None, mode=None, options=None, add_options=None, remove_options=None, special_memory_filler=None, os_name=None, plugin_preset='default', cle_memory_backer=None, dict_memory_backer=None, permissions_map=None, default_permissions=3, stack_perms=None, stack_end=None, stack_size=None, regioned_memory_cls=None, **kwargs)

Bases: angr.misc.plugins.PluginHub

The SimState represents the state of a program, including its memory, registers, and so forth.

Parameters

project (angr.Project) – The project instance.
arch (archinfo.Arch|str) – The architecture of the state.

Variables

regs – A convenient view of the state’s registers, where each register is a property
mem – A convenient view of the state’s memory, a angr.state_plugins.view.SimMemView
registers – The state’s register file as a flat memory region
memory – The state’s memory as a flat memory region
solver – The symbolic solver and variable manager for this state
inspect – The breakpoint manager, a angr.state_plugins.inspect.SimInspector
log – Information about the state’s history
scratch – Information about the current execution step
posix – MISNOMER: information about the operating system or environment model
fs – The current state of the simulated filesystem
libc – Information about the standard library we are emulating
cgc – Information about the cgc environment
uc_manager – Control of under-constrained symbolic execution
unicorn – Control of the Unicorn Engine

solver: SimSolver

posix: SimSystemPosix

registers: MemoryMixin

regs: SimRegNameView

memory: MemoryMixin

callstack: CallStack

mem: SimMemView

history: SimStateHistory

inspect: SimInspector

jni_references: SimStateJNIReferences

scratch: SimStateScratch

property plugins

property se: Deprecated alias for solver

property ip

Get the instruction pointer expression, trigger SimInspect breakpoints, and generate SimActions. Use _ip to not trigger breakpoints or generate actions.

Returns: an expression

property addr

Get the concrete address of the instruction pointer, without triggering SimInspect breakpoints or generating SimActions. An integer is returned, or an exception is raised if the instruction pointer is symbolic.

Returns: an int

property arch: archinfo.arch.Arch

T: alias of TypeVar(‘T’)

get_plugin(name)

has_plugin(name)

register_plugin(name, plugin, inhibit_init=False)

property javavm_memory

In case of an JavaVM with JNI support, a state can store the memory plugin twice; one for the native and one for the java view of the state.

Returns: The JavaVM view of the memory plugin.

property javavm_registers

In case of an JavaVM with JNI support, a state can store the registers plugin twice; one for the native and one for the java view of the state.

Returns: The JavaVM view of the registers plugin.

simplify(*args): Simplify this state’s constraints.

add_constraints(*args, **kwargs)

Add some constraints to the state.

You may pass in any number of symbolic booleans as variadic positional arguments.

satisfiable(**kwargs): Whether the state’s constraints are satisfiable

downsize(): Clean up after the solver engine. Calling this when a state no longer needs to be solved on will reduce memory usage.

step(**kwargs)

Perform a step of symbolic execution using this state. Any arguments to AngrObjectFactory.successors can be passed to this.

Returns: A SimSuccessors object categorizing the results of the step.

block(*args, **kwargs)

Represent the basic block at this state’s instruction pointer. Any arguments to AngrObjectFactory.block can ba passed to this.

Returns: A Block object describing the basic block of code at this point.

copy(): Returns a copy of the state.

merge(*others, **kwargs)

Merges this state with the other states. Returns the merging result, merged state, and the merge flag.

Parameters

states – the states to merge
merge_conditions – a tuple of the conditions under which each state holds
common_ancestor – a state that represents the common history between the states being merged. Usually it is only available when EFFICIENT_STATE_MERGING is enabled, otherwise weak-refed states might be dropped from state history instances.
plugin_whitelist – a list of plugin names that will be merged. If this option is given and is not None, any plugin that is not inside this list will not be merged, and will be created as a fresh instance in the new state.
common_ancestor_history – a SimStateHistory instance that represents the common history between the states being merged. This is to allow optimal state merging when EFFICIENT_STATE_MERGING is disabled.

Returns

(merged state, merge flag, a bool indicating if any merging occurred)

widen(*others): Perform a widening between self and other states :param others: :return:

reg_concrete(*args, **kwargs): Returns the contents of a register but, if that register is symbolic, raises a SimValueError.

mem_concrete(*args, **kwargs): Returns the contents of a memory but, if the contents are symbolic, raises a SimValueError.

stack_push(thing): Push ‘thing’ to the stack, writing the thing to memory and adjusting the stack pointer.

stack_pop(): Pops from the stack and returns the popped thing. The length will be the architecture word size.

stack_read(offset, length, bp=False)

Reads length bytes, at an offset into the stack.

Parameters

offset – The offset from the stack pointer.
length – The number of bytes to read.
bp – If True, offset from the BP instead of the SP. Default: False.

make_concrete_int(expr)

prepare_callsite(retval, args, cc='wtf')

dbg_print_stack(depth=None, sp=None): Only used for debugging purposes. Return the current stack info in formatted string. If depth is None, the current stack frame (from sp to bp) will be printed out.

set_mode(mode)

property thumb

property with_condition

class angr.sim_state_options.StateOption(name, types, default='_NO_DEFAULT_VALUE', description=None)

Bases: object

Describes a state option.

name

types

default

description

property has_default_value

one_type()

class angr.sim_state_options.SimStateOptions(thing)

Bases: object

A per-state manager of state options. An option can be either a key-valued entry or a Boolean switch (which can be seen as a key-valued entry whose value can only be either True or False).

Parameters: thing – Either a set of Boolean switches to enable, or an existing SimStateOptions instance.

OPTIONS = {'ABSTRACT_MEMORY': <O ABSTRACT_MEMORY[bool]>, 'ABSTRACT_SOLVER': <O ABSTRACT_SOLVER[bool]>, 'ACTION_DEPS': <O ACTION_DEPS[bool]>, 'ALLOW_SEND_FAILURES': <O ALLOW_SEND_FAILURES[bool]>, 'ALL_FILES_EXIST': <O ALL_FILES_EXIST[bool]>, 'APPROXIMATE_FIRST': <O APPROXIMATE_FIRST[bool]>, 'APPROXIMATE_GUARDS': <O APPROXIMATE_GUARDS[bool]>, 'APPROXIMATE_MEMORY_INDICES': <O APPROXIMATE_MEMORY_INDICES[bool]>, 'APPROXIMATE_MEMORY_SIZES': <O APPROXIMATE_MEMORY_SIZES[bool]>, 'APPROXIMATE_SATISFIABILITY': <O APPROXIMATE_SATISFIABILITY[bool]>, 'AST_DEPS': <O AST_DEPS[bool]>, 'AUTO_REFS': <O AUTO_REFS[bool]>, 'AVOID_MULTIVALUED_READS': <O AVOID_MULTIVALUED_READS[bool]>, 'AVOID_MULTIVALUED_WRITES': <O AVOID_MULTIVALUED_WRITES[bool]>, 'BEST_EFFORT_MEMORY_STORING': <O BEST_EFFORT_MEMORY_STORING[bool]>, 'BYPASS_ERRORED_IRCCALL': <O BYPASS_ERRORED_IRCCALL[bool]>, 'BYPASS_ERRORED_IROP': <O BYPASS_ERRORED_IROP[bool]>, 'BYPASS_ERRORED_IRSTMT': <O BYPASS_ERRORED_IRSTMT[bool]>, 'BYPASS_UNSUPPORTED_IRCCALL': <O BYPASS_UNSUPPORTED_IRCCALL[bool]>, 'BYPASS_UNSUPPORTED_IRDIRTY': <O BYPASS_UNSUPPORTED_IRDIRTY[bool]>, 'BYPASS_UNSUPPORTED_IREXPR': <O BYPASS_UNSUPPORTED_IREXPR[bool]>, 'BYPASS_UNSUPPORTED_IROP': <O BYPASS_UNSUPPORTED_IROP[bool]>, 'BYPASS_UNSUPPORTED_IRSTMT': <O BYPASS_UNSUPPORTED_IRSTMT[bool]>, 'BYPASS_UNSUPPORTED_SYSCALL': <O BYPASS_UNSUPPORTED_SYSCALL[bool]>, 'BYPASS_VERITESTING_EXCEPTIONS': <O BYPASS_VERITESTING_EXCEPTIONS[bool]>, 'CACHELESS_SOLVER': <O CACHELESS_SOLVER[bool]>, 'CALLLESS': <O CALLLESS[bool]>, 'CGC_ENFORCE_FD': <O CGC_ENFORCE_FD[bool]>, 'CGC_NON_BLOCKING_FDS': <O CGC_NON_BLOCKING_FDS[bool]>, 'CGC_NO_SYMBOLIC_RECEIVE_LENGTH': <O CGC_NO_SYMBOLIC_RECEIVE_LENGTH[bool]>, 'COMPOSITE_SOLVER': <O COMPOSITE_SOLVER[bool]>, 'CONCRETIZE': <O CONCRETIZE[bool]>, 'CONCRETIZE_SYMBOLIC_FILE_READ_SIZES': <O CONCRETIZE_SYMBOLIC_FILE_READ_SIZES[bool]>, 'CONCRETIZE_SYMBOLIC_WRITE_SIZES': <O CONCRETIZE_SYMBOLIC_WRITE_SIZES[bool]>, 'CONSERVATIVE_READ_STRATEGY': <O CONSERVATIVE_READ_STRATEGY[bool]>, 'CONSERVATIVE_WRITE_STRATEGY': <O CONSERVATIVE_WRITE_STRATEGY[bool]>, 'CONSTRAINT_TRACKING_IN_SOLVER': <O CONSTRAINT_TRACKING_IN_SOLVER[bool]>, 'COPY_STATES': <O COPY_STATES[bool]>, 'CPUID_SYMBOLIC': <O CPUID_SYMBOLIC[bool]>, 'DOWNSIZE_Z3': <O DOWNSIZE_Z3[bool]>, 'DO_CCALLS': <O DO_CCALLS[bool]>, 'DO_RET_EMULATION': <O DO_RET_EMULATION[bool]>, 'EFFICIENT_STATE_MERGING': <O EFFICIENT_STATE_MERGING[bool]>, 'ENABLE_NX': <O ENABLE_NX[bool]>, 'EXCEPTION_HANDLING': <O EXCEPTION_HANDLING[bool]>, 'EXTENDED_IROP_SUPPORT': <O EXTENDED_IROP_SUPPORT[bool]>, 'FAST_MEMORY': <O FAST_MEMORY[bool]>, 'FAST_REGISTERS': <O FAST_REGISTERS[bool]>, 'FILES_HAVE_EOF': <O FILES_HAVE_EOF[bool]>, 'HYBRID_SOLVER': <O HYBRID_SOLVER[bool]>, 'JAVA_IDENTIFY_GETTER_SETTER': <O JAVA_IDENTIFY_GETTER_SETTER[bool]>, 'JAVA_TRACK_ATTRIBUTES': <O JAVA_TRACK_ATTRIBUTES[bool]>, 'KEEP_IP_SYMBOLIC': <O KEEP_IP_SYMBOLIC[bool]>, 'KEEP_MEMORY_READS_DISCRETE': <O KEEP_MEMORY_READS_DISCRETE[bool]>, 'LAZY_SOLVES': <O LAZY_SOLVES[bool]>, 'MEMORY_CHUNK_INDIVIDUAL_READS': <O MEMORY_CHUNK_INDIVIDUAL_READS[bool]>, 'MEMORY_FIND_STRICT_SIZE_LIMIT': <O MEMORY_FIND_STRICT_SIZE_LIMIT[bool]>, 'MEMORY_SYMBOLIC_BYTES_MAP': <O MEMORY_SYMBOLIC_BYTES_MAP[bool]>, 'NO_CROSS_INSN_OPT': <O NO_CROSS_INSN_OPT[bool]>, 'NO_IP_CONCRETIZATION': <O NO_IP_CONCRETIZATION[bool]>, 'NO_SYMBOLIC_JUMP_RESOLUTION': <O NO_SYMBOLIC_JUMP_RESOLUTION[bool]>, 'NO_SYMBOLIC_SYSCALL_RESOLUTION': <O NO_SYMBOLIC_SYSCALL_RESOLUTION[bool]>, 'OPTIMIZE_IR': <O OPTIMIZE_IR[bool]>, 'PRODUCE_ZERODIV_SUCCESSORS': <O PRODUCE_ZERODIV_SUCCESSORS[bool]>, 'REGION_MAPPING': <O REGION_MAPPING[bool]>, 'REPLACEMENT_SOLVER': <O REPLACEMENT_SOLVER[bool]>, 'REVERSE_MEMORY_HASH_MAP': <O REVERSE_MEMORY_HASH_MAP[bool]>, 'REVERSE_MEMORY_NAME_MAP': <O REVERSE_MEMORY_NAME_MAP[bool]>, 'SHORT_READS': <O SHORT_READS[bool]>, 'SIMPLIFY_CONSTRAINTS': <O SIMPLIFY_CONSTRAINTS[bool]>, 'SIMPLIFY_EXIT_GUARD': <O SIMPLIFY_EXIT_GUARD[bool]>, 'SIMPLIFY_EXIT_STATE': <O SIMPLIFY_EXIT_STATE[bool]>, 'SIMPLIFY_EXIT_TARGET': <O SIMPLIFY_EXIT_TARGET[bool]>, 'SIMPLIFY_EXPRS': <O SIMPLIFY_EXPRS[bool]>, 'SIMPLIFY_MEMORY_READS': <O SIMPLIFY_MEMORY_READS[bool]>, 'SIMPLIFY_MEMORY_WRITES': <O SIMPLIFY_MEMORY_WRITES[bool]>, 'SIMPLIFY_REGISTER_READS': <O SIMPLIFY_REGISTER_READS[bool]>, 'SIMPLIFY_REGISTER_WRITES': <O SIMPLIFY_REGISTER_WRITES[bool]>, 'SIMPLIFY_RETS': <O SIMPLIFY_RETS[bool]>, 'SPECIAL_MEMORY_FILL': <O SPECIAL_MEMORY_FILL[bool]>, 'STRICT_PAGE_ACCESS': <O STRICT_PAGE_ACCESS[bool]>, 'STRINGS_ANALYSIS': <O STRINGS_ANALYSIS[bool]>, 'SUPER_FASTPATH': <O SUPER_FASTPATH[bool]>, 'SUPPORT_FLOATING_POINT': <O SUPPORT_FLOATING_POINT[bool]>, 'SYMBION_KEEP_STUBS_ON_SYNC': <O SYMBION_KEEP_STUBS_ON_SYNC[bool]>, 'SYMBION_SYNC_CLE': <O SYMBION_SYNC_CLE[bool]>, 'SYMBOLIC': <O SYMBOLIC[bool]>, 'SYMBOLIC_INITIAL_VALUES': <O SYMBOLIC_INITIAL_VALUES[bool]>, 'SYMBOLIC_MEMORY_NO_SINGLEVALUE_OPTIMIZATIONS': <O SYMBOLIC_MEMORY_NO_SINGLEVALUE_OPTIMIZATIONS[bool]>, 'SYMBOLIC_TEMPS': <O SYMBOLIC_TEMPS[bool]>, 'SYMBOLIC_WRITE_ADDRESSES': <O SYMBOLIC_WRITE_ADDRESSES[bool]>, 'SYMBOL_FILL_UNCONSTRAINED_MEMORY': <O SYMBOL_FILL_UNCONSTRAINED_MEMORY[bool]>, 'SYMBOL_FILL_UNCONSTRAINED_REGISTERS': <O SYMBOL_FILL_UNCONSTRAINED_REGISTERS[bool]>, 'SYNC_CLE_BACKEND_CONCRETE': <O SYNC_CLE_BACKEND_CONCRETE[bool]>, 'TRACK_ACTION_HISTORY': <O TRACK_ACTION_HISTORY[bool]>, 'TRACK_CONSTRAINTS': <O TRACK_CONSTRAINTS[bool]>, 'TRACK_CONSTRAINT_ACTIONS': <O TRACK_CONSTRAINT_ACTIONS[bool]>, 'TRACK_JMP_ACTIONS': <O TRACK_JMP_ACTIONS[bool]>, 'TRACK_MEMORY_ACTIONS': <O TRACK_MEMORY_ACTIONS[bool]>, 'TRACK_MEMORY_MAPPING': <O TRACK_MEMORY_MAPPING[bool]>, 'TRACK_OP_ACTIONS': <O TRACK_OP_ACTIONS[bool]>, 'TRACK_REGISTER_ACTIONS': <O TRACK_REGISTER_ACTIONS[bool]>, 'TRACK_SOLVER_VARIABLES': <O TRACK_SOLVER_VARIABLES[bool]>, 'TRACK_TMP_ACTIONS': <O TRACK_TMP_ACTIONS[bool]>, 'TRUE_RET_EMULATION_GUARD': <O TRUE_RET_EMULATION_GUARD[bool]>, 'UNDER_CONSTRAINED_SYMEXEC': <O UNDER_CONSTRAINED_SYMEXEC[bool]>, 'UNICORN': <O UNICORN[bool]>, 'UNICORN_AGGRESSIVE_CONCRETIZATION': <O UNICORN_AGGRESSIVE_CONCRETIZATION[bool]>, 'UNICORN_HANDLE_CGC_RECEIVE_SYSCALL': <O UNICORN_HANDLE_CGC_RECEIVE_SYSCALL[bool]>, 'UNICORN_HANDLE_CGC_TRANSMIT_SYSCALL': <O UNICORN_HANDLE_CGC_TRANSMIT_SYSCALL[bool]>, 'UNICORN_SYM_REGS_SUPPORT': <O UNICORN_SYM_REGS_SUPPORT[bool]>, 'UNICORN_THRESHOLD_CONCRETIZATION': <O UNICORN_THRESHOLD_CONCRETIZATION[bool]>, 'UNICORN_TRACK_BBL_ADDRS': <O UNICORN_TRACK_BBL_ADDRS[bool]>, 'UNICORN_TRACK_STACK_POINTERS': <O UNICORN_TRACK_STACK_POINTERS[bool]>, 'UNICORN_ZEROPAGE_GUARD': <O UNICORN_ZEROPAGE_GUARD[bool]>, 'UNINITIALIZED_ACCESS_AWARENESS': <O UNINITIALIZED_ACCESS_AWARENESS[bool]>, 'UNSUPPORTED_BYPASS_ZERO_DEFAULT': <O UNSUPPORTED_BYPASS_ZERO_DEFAULT[bool]>, 'USE_SIMPLIFIED_CCALLS': <O USE_SIMPLIFIED_CCALLS[bool]>, 'USE_SYSTEM_TIMES': <O USE_SYSTEM_TIMES[bool]>, 'VALIDATE_APPROXIMATIONS': <O VALIDATE_APPROXIMATIONS[bool]>, 'ZERO_FILL_UNCONSTRAINED_MEMORY': <O ZERO_FILL_UNCONSTRAINED_MEMORY[bool]>, 'ZERO_FILL_UNCONSTRAINED_REGISTERS': <O ZERO_FILL_UNCONSTRAINED_REGISTERS[bool]>, 'jumptable_symbolic_ip_max_targets': <O jumptable_symbolic_ip_max_targets[int]: The maximum number of concrete addresses a symbolic instruction pointer can be concretized to if it is part of a jump table.>, 'symbolic_ip_max_targets': <O symbolic_ip_max_targets[int]: The maximum number of concrete addresses a symbolic instruction pointer can be concretized to.>}

add(boolean_switch)

[COMPATIBILITY] Enable a Boolean switch.

Parameters: boolean_switch (str) – Name of the Boolean switch.
Returns: None

update(boolean_switches)

[COMPATIBILITY] In order to be compatible with the old interface, you can enable a collection of Boolean switches at the same time by doing the following:

>>> state.options.update({sim_options.SYMBOLIC, sim_options.ABSTRACT_MEMORY})

or

>>> state.options.update(sim_options.unicorn)

Parameters: boolean_switches (set) – A collection of Boolean switches to enable.
Returns: None

remove(name)

Drop a state option if it exists, or raise a KeyError if the state option is not set.

[COMPATIBILITY] Remove a Boolean switch.

Parameters: name (str) – Name of the state option.
Returns: NNone

discard(name)

Drop a state option if it exists, or silently return if the state option is not set.

[COMPATIBILITY] Disable a Boolean switch.

Parameters: name (str) – Name of the Boolean switch.
Returns: None

difference(boolean_switches)

[COMPATIBILITY] Make a copy of the current instance, and then discard all options that are in boolean_switches.

Parameters: boolean_switches (set) – A collection of Boolean switches to disable.
Returns: A new SimStateOptions instance.

copy()

Get a copy of the current SimStateOptions instance.

Returns: A new SimStateOptions instance.
Return type: SimStateOptions

tally(exclude_false=True, description=False)

Return a string representation of all state options.

Parameters

exclude_false (bool) – Whether to exclude Boolean switches that are disabled.
description (bool) – Whether to display the description of each option.

Returns

A string representation.

Return type

str

classmethod register_option(name, types, default=None, description=None)

Register a state option.

Parameters

name (str) – Name of the state option.
types – A collection of allowed types of this state option.
default – The default value of this state option.
description (str) – The description of this state option.

Returns

None

classmethod register_bool_option(name, description=None)

Register a Boolean switch as state option. This is equivalent to cls.register_option(name, set([bool]), description=description)

Parameters

name (str) – Name of the state option.
description (str) – The description of this state option.

Returns

None

class angr.state_plugins.plugin.SimStatePlugin

Bases: object

This is a base class for SimState plugins. A SimState plugin will be copied along with the state when the state is branched. They are intended to be used for things such as tracking open files, tracking heap details, and providing storage and persistence for SimProcedures.

STRONGREF_STATE = False

set_state(state): Sets a new state (for example, if the state has been branched)

set_strongref_state(state)

copy(_memo)

Should return a copy of the plugin without any state attached. Should check the memo first, and add itself to memo if it ends up making a new copy.

In order to simplify using the memo, you should annotate implementations of this function with SimStatePlugin.memo

The base implementation of this function constructs a new instance of the plugin’s class without calling its initializer. If you super-call down to it, make sure you instanciate all the fields in your copy method!

Parameters: memo – A dictionary mapping object identifiers (id(obj)) to their copied instance. Use this to avoid infinite recursion and diverged copies.

static memo(f): A decorator function you should apply to copy

merge(others, merge_conditions, common_ancestor=None)

Should merge the state plugin with the provided others. This will be called by state.merge() after copying the target state, so this should mutate the current instance to merge with the others.

Note that when multiple instances of a single plugin object (for example, a file) are referenced in the state, it is important that merge only ever be called once. This should be solved by designating one of the plugin’s referees as the “real owner”, who should be the one to actually merge it. This technique doesn’t work to resolve the similar issue that arises during copying because merging doesn’t produce a new reference to insert.

There will be n others and n+1 merge conditions, since the first condition corresponds to self. To match elements up to conditions, say zip([self] + others, merge_conditions)

When implementing this, make sure that you “deepen” both others and common_ancestor before calling sub-elements’ merge methods, e.g. self.foo.merge([o.foo for o in others], merge_conditions, common_ancestor=common_ancestor.foo if common_ancestor is not None else None).

During static analysis, merge_conditions can be None, in which case you should use state.solver.union(values). TODO: fish please make this less bullshit

There is a utility state.solver.ite_cases which will help with constructing arbitrarily large merged ASTs. Use it like self.bar = self.state.solver.ite_cases(zip(conditions[1:], [o.bar for o in others]), self.bar)

Parameters

others – the other state plugins to merge with
merge_conditions – a symbolic condition for each of the plugins
common_ancestor – a common ancestor of this plugin and the others being merged

Returns

True if the state plugins are actually merged.

Return type

bool

widen(others)

The widening operation for plugins. Widening is a special kind of merging that produces a more general state from several more specific states. It is used only during intensive static analysis. The same behavior regarding copying and mutation from merge should be followed.

Parameters: others – the other state plugin
Returns: True if the state plugin is actually widened.
Return type: bool

classmethod register_default(name, xtr=None)

init_state(): Use this function to perform any initialization on the state at plugin-add time

class angr.state_plugins.inspect.BP(when='before', enabled=None, condition=None, action=None, **kwargs)

Bases: object

A breakpoint.

check(state, when)

Checks state state to see if the breakpoint should fire.

Parameters

state – The state.
when – Whether the check is happening before or after the event.

Returns

A boolean representing whether the checkpoint should fire.

fire(state)

Trigger the breakpoint.

Parameters: state – The state.

class angr.state_plugins.inspect.SimInspector

Bases: angr.state_plugins.plugin.SimStatePlugin

The breakpoint interface, used to instrument execution. For usage information, look here: https://docs.angr.io/core-concepts/simulation#breakpoints

BP_AFTER = 'after'

BP_BEFORE = 'before'

BP_BOTH = 'both'

action(event_type, when, **kwargs): Called from within SimuVEX when events happens. This function checks all breakpoints registered for that event and fires the ones whose conditions match.

make_breakpoint(event_type, *args, **kwargs)

Creates and adds a breakpoint which would trigger on event_type. Additional arguments are passed to the BP constructor.

Returns: The created breakpoint, so that it can be removed later.

b(event_type, *args, **kwargs)

Creates and adds a breakpoint which would trigger on event_type. Additional arguments are passed to the BP constructor.

Returns: The created breakpoint, so that it can be removed later.

add_breakpoint(event_type, bp)

Adds a breakpoint which would trigger on event_type.

Parameters

event_type – The event type to trigger on
bp – The breakpoint

Returns

The created breakpoint.

remove_breakpoint(event_type, bp=None, filter_func=None)

Removes a breakpoint.

Parameters

bp – The breakpoint to remove.
filter_func – A filter function to specify whether each breakpoint should be removed or not.

copy(memo=None, **kwargs)

downsize()

Remove previously stored attributes from this plugin instance to save memory. This method is supposed to be called by breakpoint implementors. A typical workflow looks like the following :

>>> # Add `attr0` and `attr1` to `self.state.inspect`
>>> self.state.inspect(xxxxxx, attr0=yyyy, attr1=zzzz)
>>> # Get new attributes out of SimInspect in case they are modified by the user
>>> new_attr0 = self.state._inspect.attr0
>>> new_attr1 = self.state._inspect.attr1
>>> # Remove them from SimInspect
>>> self.state._inspect.downsize()

merge(others, merge_conditions, common_ancestor=None)

widen(others)

set_state(state)

class angr.state_plugins.libc.SimStateLibc

Bases: angr.state_plugins.plugin.SimStatePlugin

This state plugin keeps track of various libc stuff:

LOCALE_ARRAY = [b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x03 ', b'\x02 ', b'\x02 ', b'\x02 ', b'\x02 ', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x02\x00', b'\x01`', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x08\xd8', b'\x08\xd8', b'\x08\xd8', b'\x08\xd8', b'\x08\xd8', b'\x08\xd8', b'\x08\xd8', b'\x08\xd8', b'\x08\xd8', b'\x08\xd8', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x08\xd5', b'\x08\xd5', b'\x08\xd5', b'\x08\xd5', b'\x08\xd5', b'\x08\xd5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x08\xc5', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x08\xd6', b'\x08\xd6', b'\x08\xd6', b'\x08\xd6', b'\x08\xd6', b'\x08\xd6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x08\xc6', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x04\xc0', b'\x02\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00', b'\x00\x00']

TOLOWER_LOC_ARRAY = [128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 4294967295, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 91, 92, 93, 94, 95, 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255]

TOUPPER_LOC_ARRAY = [128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 4294967295, 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253, 254, 255]

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

property errno

ret_errno(val)

class angr.state_plugins.posix.PosixDevFS

Bases: angr.state_plugins.filesystem.SimMount

get(path)

insert(path, simfile)

delete(path)

lookup(_)

merge(others, conditions, common_ancestor=None)

widen(others)

copy(_)

class angr.state_plugins.posix.PosixProcFS

Bases: angr.state_plugins.filesystem.SimMount

The virtual file system mounted at /proc (as of now, on Linux).

get(path)

insert(path, simfile)

delete(path)

lookup(_)

merge(others, conditions, common_ancestor=None)

widen(others)

copy(_)

class angr.state_plugins.posix.SimSystemPosix(stdin=None, stdout=None, stderr=None, fd=None, sockets=None, socket_queue=None, argv=None, argc=None, environ=None, auxv=None, tls_modules=None, sigmask=None, pid=None, ppid=None, uid=None, gid=None, brk=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

Data storage and interaction mechanisms for states with an environment conforming to posix. Available as state.posix.

SIG_BLOCK = 0

SIG_UNBLOCK = 1

SIG_SETMASK = 2

EPERM = 1

ENOENT = 2

ESRCH = 3

EINTR = 4

EIO = 5

ENXIO = 6

E2BIG = 7

ENOEXEC = 8

EBADF = 9

ECHILD = 10

EAGAIN = 11

ENOMEM = 12

EACCES = 13

EFAULT = 14

ENOTBLK = 15

EBUSY = 16

EEXIST = 17

EXDEV = 18

ENODEV = 19

ENOTDIR = 20

EISDIR = 21

EINVAL = 22

ENFILE = 23

EMFILE = 24

ENOTTY = 25

ETXTBSY = 26

EFBIG = 27

ENOSPC = 28

ESPIPE = 29

EROFS = 30

EMLINK = 31

EPIPE = 32

EDOM = 33

ERANGE = 34

copy(memo=None, **kwargs)

property closed_fds

init_state()

set_brk(new_brk)

set_state(state)

open(name, flags, preferred_fd=None)

Open a symbolic file. Basically open(2).

Parameters

name (string or bytes) – Path of the symbolic file, as a string or bytes.
flags – File operation flags, a bitfield of constants from open(2), as an AST
preferred_fd – Assign this fd if it’s not already claimed.

Returns

The file descriptor number allocated (maps through posix.get_fd to a SimFileDescriptor) or None if the open fails.

mode from open(2) is unsupported at present.

open_socket(ident)

get_fd(fd): Looks up the SimFileDescriptor associated with the given number (an AST). If the number is concrete and does not map to anything, return None. If the number is symbolic, constrain it to an open fd and create a new file for it.

close(fd): Closes the given file descriptor (an AST). Returns whether the operation succeeded (a concrete boolean)

fstat(sim_fd)

sigmask(sigsetsize=None)

Gets the current sigmask. If it’s blank, a new one is created (of sigsetsize).

Parameters: sigsetsize – the size (in bytes of the sigmask set)
Returns: the sigmask

sigprocmask(how, new_mask, sigsetsize, valid_ptr=True)

Updates the signal mask.

Parameters

how – the “how” argument of sigprocmask (see manpage)
new_mask – the mask modification to apply
sigsetsize – the size (in bytes of the sigmask set)
valid_ptr – is set if the new_mask was not NULL

merge(others, merge_conditions, common_ancestor=None)

widen(_)

dump_file_by_path(path, **kwargs)

Returns the concrete content for a file by path.

Parameters

path – file path as string
kwargs – passed to state.solver.eval

Returns

file contents as string

dumps(fd, **kwargs)

Returns the concrete content for a file descriptor.

BACKWARD COMPATIBILITY: if you ask for file descriptors 0 1 or 2, it will return the data from stdin, stdout, or stderr as a flat string.

Parameters: fd – A file descriptor.
Returns: The concrete content.
Return type: str

class angr.state_plugins.filesystem.Stat(st_dev, st_ino, st_nlink, st_mode, st_uid, st_gid, st_rdev, st_size, st_blksize, st_blocks, st_atime, st_atimensec, st_mtime, st_mtimensec, st_ctime, st_ctimensec)

Bases: tuple

Create new instance of Stat(st_dev, st_ino, st_nlink, st_mode, st_uid, st_gid, st_rdev, st_size, st_blksize, st_blocks, st_atime, st_atimensec, st_mtime, st_mtimensec, st_ctime, st_ctimensec)

property st_atime: Alias for field number 10

property st_atimensec: Alias for field number 11

property st_blksize: Alias for field number 8

property st_blocks: Alias for field number 9

property st_ctime: Alias for field number 14

property st_ctimensec: Alias for field number 15

property st_dev: Alias for field number 0

property st_gid: Alias for field number 5

property st_ino: Alias for field number 1

property st_mode: Alias for field number 3

property st_mtime: Alias for field number 12

property st_mtimensec: Alias for field number 13

property st_nlink: Alias for field number 2

property st_rdev: Alias for field number 6

property st_size: Alias for field number 7

property st_uid: Alias for field number 4

class angr.state_plugins.filesystem.SimFilesystem(files=None, pathsep=None, cwd=None, mountpoints=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

angr’s emulated filesystem. Available as state.fs. When constructing, all parameters are optional.

Parameters

files – A mapping from filepath to SimFile
pathsep – The character used to separate path elements, default forward slash.
cwd – The path of the current working directory to use
mountpoints – A mapping from filepath to SimMountpoint

Variables

pathsep – The current pathsep
cwd – The current working directory
unlinks – A list of unlink operations, tuples of filename and simfile. Be careful, this list is shallow-copied from successor to successor, so don’t mutate anything in it without copying.

copy(memo=None, **kwargs)

property unlinks

set_state(state)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

chdir(path): Changes the current directory to the given path

get(path): Get a file from the filesystem. Returns a SimFile or None.

insert(path, simfile): Insert a file into the filesystem. Returns whether the operation was successful.

delete(path)

Remove a file from the filesystem. Returns whether the operation was successful.

This will add a fs_unlink event with the path of the file and also the index into the unlinks list.

mount(path, mount): Add a mountpoint to the filesystem.

unmount(path): Remove a mountpoint from the filesystem.

get_mountpoint(path)

Look up the mountpoint servicing the given path.

Returns: A tuple of the mount and a list of path elements traversing from the mountpoint to the specified file.

class angr.state_plugins.filesystem.SimMount

Bases: angr.state_plugins.plugin.SimStatePlugin

This is the base class for “mount points” in angr’s simulated filesystem. Subclass this class and give it to the filesystem to intercept all file creations and opens below the mountpoint. Since this a SimStatePlugin you may also want to implement set_state, copy, merge, etc.

get(path_elements)

Implement this function to instrument file lookups.

Parameters: path_elements – A list of path elements traversing from the mountpoint to the file
Returns: A SimFile, or None

insert(path_elements, simfile)

Implement this function to instrument file creation.

Parameters

path_elements – A list of path elements traversing from the mountpoint to the file
simfile – The file to insert

Returns

A bool indicating whether the insert occurred

delete(path_elements)

Implement this function to instrument file deletion.

Parameters: path_elements – A list of path elements traversing from the mountpoint to the file
Returns: A bool indicating whether the delete occurred

lookup(sim_file)

Look up the path of a SimFile in the mountpoint

Parameters: sim_file – A SimFile object needs to be looked up
Returns: A string representing the path of the file in the mountpoint Or None if the SimFile does not exist in the mountpoint

class angr.state_plugins.filesystem.SimConcreteFilesystem(pathsep='/')

Bases: angr.state_plugins.filesystem.SimMount

Abstract SimMount allowing the user to import files from some external source into the guest

Parameters: pathsep (str) – The host path separator character, default os.path.sep

get(path_elements)

insert(path_elements, simfile)

delete(path_elements)

lookup(sim_file)

copy(memo=None, **kwargs)

set_state(state)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

class angr.state_plugins.filesystem.SimHostFilesystem(host_path=None, **kwargs)

Bases: angr.state_plugins.filesystem.SimConcreteFilesystem

Simulated mount that makes some piece from the host filesystem available to the guest.

Parameters

host_path (str) – The path on the host to mount
pathsep (str) – The host path separator character, default os.path.sep

copy(memo=None, **kwargs)

angr.state_plugins.solver.timed_function(f)

angr.state_plugins.solver.enable_timing()

angr.state_plugins.solver.disable_timing()

angr.state_plugins.solver.error_converter(f)

angr.state_plugins.solver.concrete_path_bool(f)

angr.state_plugins.solver.concrete_path_not_bool(f)

angr.state_plugins.solver.concrete_path_scalar(f)

angr.state_plugins.solver.concrete_path_tuple(f)

angr.state_plugins.solver.concrete_path_list(f)

class angr.state_plugins.solver.SimSolver(solver=None, all_variables=None, temporal_tracked_variables=None, eternal_tracked_variables=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

This is the plugin you’ll use to interact with symbolic variables, creating them and evaluating them. It should be available on a state as state.solver.

Any top-level variable of the claripy module can be accessed as a property of this object.

reload_solver(constraints=None)

Reloads the solver. Useful when changing solver options.

Parameters: constraints (list) – A new list of constraints to use in the reloaded solver instead of the current one

get_variables(*keys)

Iterate over all variables for which their tracking key is a prefix of the values provided.

Elements are a tuple, the first element is the full tracking key, the second is the symbol.

>>> list(s.solver.get_variables('mem'))
[(('mem', 0x1000), <BV64 mem_1000_4_64>), (('mem', 0x1008), <BV64 mem_1008_5_64>)]

>>> list(s.solver.get_variables('file'))
[(('file', 1, 0), <BV8 file_1_0_6_8>), (('file', 1, 1), <BV8 file_1_1_7_8>), (('file', 2, 0), <BV8 file_2_0_8_8>)]

>>> list(s.solver.get_variables('file', 2))
[(('file', 2, 0), <BV8 file_2_0_8_8>)]

>>> list(s.solver.get_variables())
[(('mem', 0x1000), <BV64 mem_1000_4_64>), (('mem', 0x1008), <BV64 mem_1008_5_64>), (('file', 1, 0), <BV8 file_1_0_6_8>), (('file', 1, 1), <BV8 file_1_1_7_8>), (('file', 2, 0), <BV8 file_2_0_8_8>)]

register_variable(v, key, eternal=True)

Register a value with the variable tracking system

Parameters

v – The BVS to register
key – A tuple to register the variable under

Parma eternal

Whether this is an eternal variable, default True. If False, an incrementing counter will be appended to the key.

describe_variables(v): Given an AST, iterate over all the keys of all the BVS leaves in the tree which are registered.

Unconstrained(name, bits, uninitialized=True, inspect=True, events=True, key=None, eternal=False, **kwargs)

Creates an unconstrained symbol or a default concrete value (0), based on the state options.

Parameters

name – The name of the symbol.
bits – The size (in bits) of the symbol.
uninitialized – Whether this value should be counted as an “uninitialized” value in the course of an analysis.
inspect – Set to False to avoid firing SimInspect breakpoints
events – Set to False to avoid generating a SimEvent for the occasion
key – Set this to a tuple of increasingly specific identifiers (for example, ('mem', 0xffbeff00) or ('file', 4, 0x20) to cause it to be tracked, i.e. accessable through solver.get_variables.
eternal – Set to True in conjunction with setting a key to cause all states with the same ancestry to retrieve the same symbol when trying to create the value. If False, a counter will be appended to the key.

Returns

an unconstrained symbol (or a concrete value of 0).

BVS(name, size, min=None, max=None, stride=None, uninitialized=False, explicit_name=None, key=None, eternal=False, inspect=True, events=True, **kwargs)

Creates a bit-vector symbol (i.e., a variable). Other keyword parameters are passed directly on to the constructor of claripy.ast.BV.

Parameters

name – The name of the symbol.
size – The size (in bits) of the bit-vector.
min – The minimum value of the symbol. Note that this only work when using VSA.
max – The maximum value of the symbol. Note that this only work when using VSA.
stride – The stride of the symbol. Note that this only work when using VSA.
uninitialized – Whether this value should be counted as an “uninitialized” value in the course of an analysis.
explicit_name – Set to True to prevent an identifier from appended to the name to ensure uniqueness.
key – Set this to a tuple of increasingly specific identifiers (for example, ('mem', 0xffbeff00) or ('file', 4, 0x20) to cause it to be tracked, i.e. accessable through solver.get_variables.
eternal – Set to True in conjunction with setting a key to cause all states with the same ancestry to retrieve the same symbol when trying to create the value. If False, a counter will be appended to the key.
inspect – Set to False to avoid firing SimInspect breakpoints
events – Set to False to avoid generating a SimEvent for the occasion

Returns

A BV object representing this symbol.

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

downsize(): Frees memory associated with the constraint solver by clearing all of its internal caches.

property constraints: Returns the constraints of the state stored by the solver.

eval_to_ast(e, n, extra_constraints=(), exact=None)

Evaluate an expression, using the solver if necessary. Returns AST objects.

Parameters

e – the expression
n – the number of desired solutions
extra_constraints – extra constraints to apply to the solver
exact – if False, returns approximate solutions

Returns

a tuple of the solutions, in the form of claripy AST nodes

Return type

tuple

max(e, extra_constraints=(), exact=None)

Return the maximum value of expression e.

:param e : expression (an AST) to evaluate :param extra_constraints: extra constraints (as ASTs) to add to the solver for this solve :param exact : if False, return approximate solutions. :return: the maximum possible value of e (backend object)

min(e, extra_constraints=(), exact=None)

Return the minimum value of expression e.

:param e : expression (an AST) to evaluate :param extra_constraints: extra constraints (as ASTs) to add to the solver for this solve :param exact : if False, return approximate solutions. :return: the minimum possible value of e (backend object)

solution(e, v, extra_constraints=(), exact=None)

Return True if v is a solution of expr with the extra constraints, False otherwise.

Parameters

e – An expression (an AST) to evaluate
v – The proposed solution (an AST)
extra_constraints – Extra constraints (as ASTs) to add to the solver for this solve.
exact – If False, return approximate solutions.

Returns

True if v is a solution of expr, False otherwise

is_true(e, extra_constraints=(), exact=None)

If the expression provided is absolutely, definitely a true boolean, return True. Note that returning False doesn’t necessarily mean that the expression can be false, just that we couldn’t figure that out easily.

Parameters

e – An expression (an AST) to evaluate
extra_constraints – Extra constraints (as ASTs) to add to the solver for this solve.
exact – If False, return approximate solutions.

Returns

True if v is definitely true, False otherwise

is_false(e, extra_constraints=(), exact=None)

If the expression provided is absolutely, definitely a false boolean, return True. Note that returning False doesn’t necessarily mean that the expression can be true, just that we couldn’t figure that out easily.

Parameters

e – An expression (an AST) to evaluate
extra_constraints – Extra constraints (as ASTs) to add to the solver for this solve.
exact – If False, return approximate solutions.

Returns

True if v is definitely false, False otherwise

unsat_core(extra_constraints=())

This function returns the unsat core from the backend solver.

Parameters: extra_constraints – Extra constraints (as ASTs) to add to the solver for this solve.
Returns: The unsat core.

satisfiable(extra_constraints=(), exact=None)

This function does a constraint check and checks if the solver is in a sat state.

Parameters

extra_constraints – Extra constraints (as ASTs) to add to s for this solve
exact – If False, return approximate solutions.

Returns

True if sat, otherwise false

add(*constraints)

Add some constraints to the solver.

Parameters: constraints – Pass any constraints that you want to add (ASTs) as varargs.

T: alias of TypeVar(‘T’, int, bytes)

CastTarget: alias of TypeVar(‘CastTarget’)

eval_upto(e, n: int, cast_to: None = None, **kwargs) → Any

eval_upto(e, n: int, cast_to: CastTarget = None, **kwargs) → CastTarget

Evaluate an expression, using the solver if necessary. Returns primitives as specified by the cast_to parameter. Only certain primitives are supported, check the implementation of _cast_to to see which ones.

Parameters

e – the expression
n – the number of desired solutions
extra_constraints – extra constraints to apply to the solver
exact – if False, returns approximate solutions
cast_to – A type to cast the resulting values to

Returns

a tuple of the solutions, in the form of Python primitives

Return type

tuple

eval(e, cast_to: None = None, **kwargs) → Any

eval(e, cast_to: CastTarget = None, **kwargs) → CastTarget

Evaluate an expression to get any possible solution. The desired output types can be specified using the cast_to parameter. extra_constraints can be used to specify additional constraints the returned values must satisfy.

Parameters

e – the expression to get a solution for
kwargs – Any additional kwargs will be passed down to eval_upto

Raises

SimUnsatError – if no solution could be found satisfying the given constraints

Returns

eval_one(e, **kwargs)

Evaluate an expression to get the only possible solution. Errors if either no or more than one solution is returned. A kwarg parameter default can be specified to be returned instead of failure!

Parameters

e – the expression to get a solution for
default – A value can be passed as a kwarg here. It will be returned in case of failure.
kwargs – Any additional kwargs will be passed down to eval_upto

Raises

SimUnsatError – if no solution could be found satisfying the given constraints
SimValueError – if more than one solution was found to satisfy the given constraints

Returns

The value for e

eval_atmost(e, n, **kwargs)

Evaluate an expression to get at most n possible solutions. Errors if either none or more than n solutions are returned.

Parameters

e – the expression to get a solution for
n – the inclusive upper limit on the number of solutions
kwargs – Any additional kwargs will be passed down to eval_upto

Raises

SimUnsatError – if no solution could be found satisfying the given constraints
SimValueError – if more than n solutions were found to satisfy the given constraints

Returns

The solutions for e

eval_atleast(e, n, **kwargs)

Evaluate an expression to get at least n possible solutions. Errors if less than n solutions were found.

Parameters

e – the expression to get a solution for
n – the inclusive lower limit on the number of solutions
kwargs – Any additional kwargs will be passed down to eval_upto

Raises

SimUnsatError – if no solution could be found satisfying the given constraints
SimValueError – if less than n solutions were found to satisfy the given constraints

Returns

The solutions for e

eval_exact(e, n, **kwargs)

Evaluate an expression to get exactly the n possible solutions. Errors if any number of solutions other than n was found to exist.

Parameters

e – the expression to get a solution for
n – the inclusive lower limit on the number of solutions
kwargs – Any additional kwargs will be passed down to eval_upto

Raises

SimUnsatError – if no solution could be found satisfying the given constraints
SimValueError – if any number of solutions other than n were found to satisfy the given constraints

Returns

The solutions for e

min_int(e, extra_constraints=(), exact=None)

Return the minimum value of expression e.

:param e : expression (an AST) to evaluate :param extra_constraints: extra constraints (as ASTs) to add to the solver for this solve :param exact : if False, return approximate solutions. :return: the minimum possible value of e (backend object)

max_int(e, extra_constraints=(), exact=None)

Return the maximum value of expression e.

:param e : expression (an AST) to evaluate :param extra_constraints: extra constraints (as ASTs) to add to the solver for this solve :param exact : if False, return approximate solutions. :return: the maximum possible value of e (backend object)

unique(e, **kwargs): Returns True if the expression e has only one solution by querying the constraint solver. It does also add that unique solution to the solver’s constraints.

symbolic(e): Returns True if the expression e is symbolic.

single_valued(e): Returns True whether e is a concrete value or is a value set with only 1 possible value. This differs from unique in that this does not query the constraint solver.

simplify(e=None): Simplifies e. If e is None, simplifies the constraints of this state.

variables(e): Returns the symbolic variables present in the AST of e.

class angr.state_plugins.log.SimStateLog(log=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

property actions

add_event(event_type, **kwargs)

add_action(action)

extend_actions(new_actions)

events_of_type(event_type)

actions_of_type(action_type)

property fresh_constraints

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

clear()

class angr.state_plugins.callstack.CallStack(call_site_addr=0, func_addr=0, stack_ptr=0, ret_addr=0, jumpkind='Ijk_Call', next_frame: Optional[angr.state_plugins.callstack.CallStack] = None, invoke_return_variable=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

Stores the address of the function you’re in and the value of SP at the VERY BOTTOM of the stack, i.e. points to the return address.

copy(memo=None, **kwargs)

set_state(state)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

property current_function_address

Address of the current function.

Returns: the address of the function
Return type: int

property current_stack_pointer

Get the value of the stack pointer.

Returns: Value of the stack pointer
Return type: int

property current_return_target

Get the return target.

Returns: The address of return target.
Return type: int

static stack_suffix_to_string(stack_suffix): Convert a stack suffix to a human-readable string representation. :param tuple stack_suffix: The stack suffix. :return: A string representation :rtype: str

property top

Returns the element at the top of the callstack without removing it.

Returns: A CallStack.

push(cf): Push the frame cf onto the stack. Return the new stack.

pop(): Pop the top frame from the stack. Return the new stack.

call(callsite_addr, addr, retn_target=None, stack_pointer=None)

Push a stack frame into the call stack. This method is called when calling a function in CFG recovery.

Parameters

callsite_addr (int) – Address of the call site
addr (int) – Address of the call target
retn_target (int or None) – Address of the return target
stack_pointer (int) – Value of the stack pointer

Returns

None

ret(retn_target=None)

Pop one or many call frames from the stack. This method is called when returning from a function in CFG recovery.

Parameters: retn_target (int) – The target to return to.
Returns: None

dbg_repr()

Debugging representation of this CallStack object.

Returns: Details of this CalLStack
Return type: str

stack_suffix(context_sensitivity_level)

Generate the stack suffix. A stack suffix can be used as the key to a SimRun in CFG recovery.

Parameters: context_sensitivity_level (int) – Level of context sensitivity.
Returns: A tuple of stack suffix.
Return type: tuple

class angr.state_plugins.callstack.CallStackAction(callstack_hash, callstack_depth, action, callframe=None, ret_site_addr=None)

Bases: object

Used in callstack backtrace, which is a history of callstacks along a path, to record individual actions occurred each time the callstack is changed.

class angr.state_plugins.light_registers.SimLightRegisters(reg_map=None, registers=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

copy(memo=None, **kwargs)

set_state(state)

resolve_register(offset, size)

load(offset, size=None, **kwargs)

store(offset, value, size=None, endness=None, **kwargs)

class angr.state_plugins.history.SimStateHistory(parent=None, clone=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

This class keeps track of historically-relevant information for paths.

STRONGREF_STATE = True

init_state()

set_strongref_state(state)

property addr

merge(others, merge_conditions, common_ancestor=None)

widen(others)

copy(memo=None, **kwargs)

trim(): Discard the ancestry of this state.

filter_actions(start_block_addr=None, end_block_addr=None, block_stmt=None, insn_addr=None, read_from=None, write_to=None)

Filter self.actions based on some common parameters.

[start_block_addr, end_block_addr]

Parameters

start_block_addr – Only return actions generated in blocks starting at this address.
end_block_addr – Only return actions generated in blocks ending at this address.
block_stmt – Only return actions generated in the nth statement of each block.
insn_addr – Only return actions generated in the assembly instruction at this address.
read_from – Only return actions that perform a read from the specified location.
write_to – Only return actions that perform a write to the specified location.

Notes: If IR optimization is turned on, reads and writes may not occur in the instruction they originally came from. Most commonly, If a register is read from twice in the same block, the second read will not happen, instead reusing the temp the value is already stored in.

Valid values for read_from and write_to are the string literals ‘reg’ or ‘mem’ (matching any read or write to registers or memory, respectively), any string (representing a read or write to the named register), and any integer (representing a read or write to the memory at this address).

demote(): Demotes this history node, causing it to drop the strong state reference.

reachable()

add_event(event_type, **kwargs)

add_action(action)

extend_actions(new_actions)

subscribe_actions()

property recent_constraints

property recent_actions

property block_count

property lineage

property parents

property events

property actions

property jumpkinds

property jump_guards

property jump_targets

property jump_sources

property descriptions

property bbl_addrs

property ins_addrs

property stack_actions

closest_common_ancestor(other)

Find the common ancestor between this history node and ‘other’.

Parameters: other – the PathHistory to find a common ancestor with.
Returns: the common ancestor SimStateHistory, or None if there isn’t one

constraints_since(other)

Returns the constraints that have been accumulated since other.

Parameters: other – a prior PathHistory object
Returns: a list of constraints

make_child()

class angr.state_plugins.history.TreeIter(start, end=None)

Bases: object

property hardcopy

count(v): Count occurrences of value v in the entire history. Note that the subclass must implement the __reversed__ method, otherwise an exception will be thrown. :param object v: The value to look for :return: The number of occurrences :rtype: int

class angr.state_plugins.history.HistoryIter(start, end=None): Bases: angr.state_plugins.history.TreeIter

class angr.state_plugins.history.LambdaAttrIter(start, f, **kwargs): Bases: angr.state_plugins.history.TreeIter

class angr.state_plugins.history.LambdaIterIter(start, f, reverse=True, **kwargs): Bases: angr.state_plugins.history.LambdaAttrIter

class angr.state_plugins.gdb.GDB(omit_fp=False, adjust_stack=False)

Bases: angr.state_plugins.plugin.SimStatePlugin

Initialize or update a state from gdb dumps of the stack, heap, registers and data (or arbitrary) segments.

Parameters

omit_fp – The frame pointer register is used for something else. (i.e. –omit_frame_pointer)
adjust_stack – Use different stack addresses than the gdb session (not recommended).

set_stack(stack_dump, stack_top)

Stack dump is a dump of the stack from gdb, i.e. the result of the following gdb command :

dump binary memory [stack_dump] [begin_addr] [end_addr]

We set the stack to the same addresses as the gdb session to avoid pointers corruption.

Parameters

stack_dump – The dump file.
stack_top – The address of the top of the stack in the gdb session.

set_heap(heap_dump, heap_base)

Heap dump is a dump of the heap from gdb, i.e. the result of the following gdb command:

dump binary memory [stack_dump] [begin] [end]

Parameters

heap_dump – The dump file.
heap_base – The start address of the heap in the gdb session.

set_data(addr, data_dump): Update any data range (most likely use is the data segments of loaded objects)

set_regs(regs_dump)

Initialize register values within the state

Parameters: regs_dump – The output of info registers in gdb.

copy(memo=None, **kwargs)

class angr.state_plugins.cgc.SimStateCGC

Bases: angr.state_plugins.plugin.SimStatePlugin

This state plugin keeps track of CGC state.

EBADF = 1

EFAULT = 2

EINVAL = 3

ENOMEM = 4

ENOSYS = 5

EPIPE = 6

FD_SETSIZE = 1024

max_allocation = 268435456

copy(memo=None, **kwargs)

peek_input()

discard_input(num_bytes)

peek_output()

discard_output(num_bytes)

addr_invalid(a)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

get_max_sinkhole(length)

Find a sinkhole which is large enough to support length bytes.

This uses first-fit. The first sinkhole (ordered in descending order by their address) which can hold length bytes is chosen. If there are more than length bytes in the sinkhole, a new sinkhole is created representing the remaining bytes while the old sinkhole is removed.

add_sinkhole(address, length)

Add a sinkhole.

Allow the possibility for the program to reuse the memory represented by the address length pair.

angr.state_plugins.trace_additions.l = <Logger angr.state_plugins.trace_additions (WARNING)>

This file contains objects to track additional information during a trace or modify symbolic variables during a trace.

The ChallRespInfo plugin tracks variables in stdin and stdout to enable handling of challenge response It handles atoi/int2str in a special manner since path constraints will usually prevent their values from being modified

The Zen plugin simplifies expressions created from variables in the flag page (losing some accuracy) to avoid situations where they become to complex for z3, but the actual equation doesn’t matter much. This can happen in challenge response if all of the values in the flag page are multiplied together before being printed.

class angr.state_plugins.trace_additions.FormatInfo

Bases: object

copy()

compute(state)

get_type()

class angr.state_plugins.trace_additions.FormatInfoStrToInt(addr, func_name, str_arg_num, base, base_arg, allows_negative)

Bases: angr.state_plugins.trace_additions.FormatInfo

copy()

compute(state)

get_type()

class angr.state_plugins.trace_additions.FormatInfoIntToStr(addr, func_name, int_arg_num, str_dst_num, base, base_arg)

Bases: angr.state_plugins.trace_additions.FormatInfo

copy()

compute(state)

get_type()

class angr.state_plugins.trace_additions.FormatInfoDontConstrain(addr, func_name, check_symbolic_arg)

Bases: angr.state_plugins.trace_additions.FormatInfo

copy()

compute(state)

get_type()

angr.state_plugins.trace_additions.int2base(x, base)

angr.state_plugins.trace_additions.generic_info_hook(state)

angr.state_plugins.trace_additions.end_info_hook(state)

angr.state_plugins.trace_additions.exit_hook(state)

angr.state_plugins.trace_additions.syscall_hook(state)

angr.state_plugins.trace_additions.constraint_hook(state)

class angr.state_plugins.trace_additions.ChallRespInfo

Bases: angr.state_plugins.plugin.SimStatePlugin

This state plugin keeps track of the reads and writes to symbolic addresses

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

static get_byte(var_name)

lookup_original(replacement)

pop_from_backup()

get_stdin_indices(variable)

get_stdout_indices(variable)

get_real_len(input_val, base, result_bv, allows_negative)

get_possible_len(input_val, base, allows_negative)

get_same_length_constraints()

static atoi_dumps(state, require_same_length=True)

static prep_tracer(state, format_infos=None)

angr.state_plugins.trace_additions.zen_hook(state, expr)

angr.state_plugins.trace_additions.zen_memory_write(state)

angr.state_plugins.trace_additions.zen_register_write(state)

class angr.state_plugins.trace_additions.ZenPlugin(max_depth=13)

Bases: angr.state_plugins.plugin.SimStatePlugin

static get_flag_rand_args(expr)

get_expr_depth(expr)

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

get_flag_bytes(ast)

filter_constraints(constraints)

analyze_transmit(state, buf)

static prep_tracer(state)

class angr.state_plugins.globals.SimStateGlobals(backer=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

set_state(state)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

keys()

values()

items()

get(k, alt=None)

pop(k, alt=None)

copy(memo=None, **kwargs)

class angr.state_plugins.uc_manager.SimUCManager(man=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

assign(dst_addr_ast)

Assign a new region for under-constrained symbolic execution.

Parameters: dst_addr_ast – the symbolic AST which address of the new allocated region will be assigned to.
Returns: as ast of memory address that points to a new region

copy(memo=None, **kwargs)

get_alloc_depth(addr)

is_bounded(ast)

Test whether an AST is bounded by any existing constraint in the related solver.

Parameters: ast – an claripy.AST object
Returns: True if there is at least one related constraint, False otherwise

set_state(state)

class angr.state_plugins.scratch.SimStateScratch(scratch=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

Implements the scratch state plugin.

property priv

push_priv(priv)

pop_priv()

set_tyenv(tyenv)

tmp_expr(tmp)

Returns the Claripy expression of a VEX temp value.

Parameters

tmp – the number of the tmp
simplify – simplify the tmp before returning it

Returns

a Claripy expression of the tmp

store_tmp(tmp, content, reg_deps=None, tmp_deps=None, deps=None, **kwargs)

Stores a Claripy expression in a VEX temp value. If in symbolic mode, this involves adding a constraint for the tmp’s symbolic variable.

Parameters

tmp – the number of the tmp
content – a Claripy expression of the content
reg_deps – the register dependencies of the content
tmp_deps – the temporary value dependencies of the content

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

clear()

class angr.state_plugins.preconstrainer.SimStatePreconstrainer(constrained_addrs=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

This state plugin manages the concept of preconstraining - adding constraints which you would like to remove later.

:param constrained_addrs : SimActions for memory operations whose addresses should be constrained during crash analysis

merge(others, merge_conditions, common_ancestor=None)

widen(others)

copy(memo=None, **kwargs)

preconstrain(value, variable)

Add a preconstraint that variable == value to the state.

Parameters

value – The concrete value. Can be a bitvector or a bytestring or an integer.
variable – The BVS to preconstrain.

preconstrain_file(content, simfile, set_length=False)

Preconstrain the contents of a file.

Parameters

content – The content to preconstrain the file to. Can be a bytestring or a list thereof.
simfile – The actual simfile to preconstrain

preconstrain_flag_page(magic_content)

Preconstrain the data in the flag page.

Parameters: magic_content – The content of the magic page as a bytestring.

remove_preconstraints(to_composite_solver=True, simplify=True)

Remove the preconstraints from the state.

If you are using the zen plugin, this will also use that to filter the constraints.

Parameters

to_composite_solver – Whether to convert the replacement solver to a composite solver. You probably want this if you’re switching from tracing to symbolic analysis.
simplify – Whether to simplify the resulting set of constraints.

reconstrain(): Split the solver. If any of the subsolvers time out after a short timeout (10 seconds), re-add the preconstraints associated with each of its variables. Hopefully these constraints still allow us to do meaningful things to the state.

class angr.state_plugins.unicorn_engine.MEM_PATCH

Bases: _ctypes.Structure

struct mem_update_t

address: Structure/Union member

length: Structure/Union member

next: Structure/Union member

class angr.state_plugins.unicorn_engine.TRANSMIT_RECORD

Bases: _ctypes.Structure

struct transmit_record_t

count: Structure/Union member

data: Structure/Union member

class angr.state_plugins.unicorn_engine.TaintEntityEnum

Bases: object

taint_entity_enum_t

TAINT_ENTITY_REG = 0

TAINT_ENTITY_TMP = 1

TAINT_ENTITY_MEM = 2

TAINT_ENTITY_NONE = 3

class angr.state_plugins.unicorn_engine.MemoryValue

Bases: _ctypes.Structure

struct memory_value_t

address: Structure/Union member

is_value_symbolic: Structure/Union member

size: Structure/Union member

value: Structure/Union member

class angr.state_plugins.unicorn_engine.RegisterValue

Bases: _ctypes.Structure

struct register_value_t

offset: Structure/Union member

size: Structure/Union member

value: Structure/Union member

class angr.state_plugins.unicorn_engine.InstrDetails

Bases: _ctypes.Structure

struct sym_instr_details_t

has_memory_dep: Structure/Union member

instr_addr: Structure/Union member

memory_values: Structure/Union member

memory_values_count: Structure/Union member

class angr.state_plugins.unicorn_engine.BlockDetails

Bases: _ctypes.Structure

struct sym_block_details_ret_t

block_addr: Structure/Union member

block_size: Structure/Union member

register_values: Structure/Union member

register_values_count: Structure/Union member

symbolic_instrs: Structure/Union member

symbolic_instrs_count: Structure/Union member

class angr.state_plugins.unicorn_engine.STOP

Bases: object

enum stop_t

STOP_NORMAL = 0

STOP_STOPPOINT = 1

STOP_ERROR = 2

STOP_SYSCALL = 3

STOP_EXECNONE = 4

STOP_ZEROPAGE = 5

STOP_NOSTART = 6

STOP_SEGFAULT = 7

STOP_ZERO_DIV = 8

STOP_NODECODE = 9

STOP_HLT = 10

STOP_VEX_LIFT_FAILED = 11

STOP_SYMBOLIC_CONDITION = 12

STOP_SYMBOLIC_PC = 13

STOP_SYMBOLIC_READ_ADDR = 14

STOP_SYMBOLIC_READ_SYMBOLIC_TRACKING_DISABLED = 15

STOP_SYMBOLIC_WRITE_ADDR = 16

STOP_SYMBOLIC_BLOCK_EXIT_CONDITION = 17

STOP_SYMBOLIC_BLOCK_EXIT_TARGET = 18

STOP_UNSUPPORTED_STMT_PUTI = 19

STOP_UNSUPPORTED_STMT_STOREG = 20

STOP_UNSUPPORTED_STMT_LOADG = 21

STOP_UNSUPPORTED_STMT_CAS = 22

STOP_UNSUPPORTED_STMT_LLSC = 23

STOP_UNSUPPORTED_STMT_DIRTY = 24

STOP_UNSUPPORTED_EXPR_GETI = 25

STOP_UNSUPPORTED_STMT_UNKNOWN = 26

STOP_UNSUPPORTED_EXPR_UNKNOWN = 27

STOP_UNKNOWN_MEMORY_WRITE_SIZE = 28

STOP_SYMBOLIC_MEM_DEP_NOT_LIVE = 29

STOP_SYSCALL_ARM = 30

STOP_SYMBOLIC_MEM_DEP_NOT_LIVE_CURR_BLOCK = 31

STOP_X86_CPUID = 32

stop_message = {0: 'Reached maximum steps', 1: 'Hit a stop point', 2: 'Something wrong', 3: 'Unable to handle syscall', 4: 'Fetching empty page', 5: 'Accessing zero page', 6: 'Failed to start', 7: 'Permissions or mapping error', 8: 'Divide by zero', 9: 'Instruction decoding error', 10: 'hlt instruction encountered', 11: 'Failed to lift block to VEX', 12: 'Symbolic condition for ITE', 13: 'Instruction pointer became symbolic', 14: 'Attempted to read from symbolic address', 15: 'Attempted to read symbolic data from memory but symbolic tracking is disabled', 16: 'Attempted to write to symbolic address', 17: "Guard condition of block's exit statement is symbolic", 18: 'Target of default exit of block is symbolic', 19: 'Symbolic taint propagation for PutI statement not yet supported', 20: 'Symbolic taint propagation for StoreG statement not yet supported', 21: 'Symbolic taint propagation for LoadG statement not yet supported', 22: 'Symbolic taint propagation for CAS statement not yet supported', 23: 'Symbolic taint propagation for LLSC statement not yet supported', 24: 'Symbolic taint propagation for Dirty statement not yet supported', 25: 'Symbolic taint propagation for GetI expression not yet supported', 26: 'Canoo propagate symbolic taint for unsupported VEX statement type', 27: 'Cannot propagate symbolic taint for unsupported VEX expression', 28: "Cannot determine size of memory write; likely because unicorn didn't", 29: 'A symbolic memory dependency on stack is no longer in scope', 30: 'ARM syscalls are currently not supported by SimEngineUnicorn', 31: 'An instruction in current block overwrites a symbolic value needed for re-executing some instruction in same block', 32: 'Block executes cpuid which should be handled in VEX engine'}

symbolic_stop_reasons = [12, 13, 14, 15, 16, 17, 18, 30, 31, 32]

unsupported_reasons = [19, 20, 21, 22, 23, 24, 26, 27, 11]

static name_stop(num)

static get_stop_msg(stop_reason)

class angr.state_plugins.unicorn_engine.StopDetails

Bases: _ctypes.Structure

struct stop_details_t

block_addr: Structure/Union member

block_size: Structure/Union member

stop_reason: Structure/Union member

class angr.state_plugins.unicorn_engine.SimOSEnum

Bases: object

enum simos_t

SIMOS_CGC = 0

SIMOS_LINUX = 1

SIMOS_OTHER = 2

exception angr.state_plugins.unicorn_engine.MemoryMappingError: Bases: Exception

exception angr.state_plugins.unicorn_engine.AccessingZeroPageError: Bases: angr.state_plugins.unicorn_engine.MemoryMappingError

exception angr.state_plugins.unicorn_engine.FetchingZeroPageError: Bases: angr.state_plugins.unicorn_engine.MemoryMappingError

exception angr.state_plugins.unicorn_engine.SegfaultError: Bases: angr.state_plugins.unicorn_engine.MemoryMappingError

exception angr.state_plugins.unicorn_engine.MixedPermissonsError: Bases: angr.state_plugins.unicorn_engine.MemoryMappingError

class angr.state_plugins.unicorn_engine.AggressiveConcretizationAnnotation(addr): Bases: claripy.annotation.SimplificationAvoidanceAnnotation

class angr.state_plugins.unicorn_engine.Uniwrapper(arch, cache_key, thumb=False)

Bases: unicorn.unicorn.Uc

hook_add(htype, callback, user_data=None, begin=1, end=0, arg1=0)

hook_del(h)

mem_map(addr, size, perms=7)

mem_map_ptr(addr, size, perms, ptr)

mem_unmap(addr, size)

mem_reset()

hook_reset()

reset()

class angr.state_plugins.unicorn_engine.Unicorn(syscall_hooks=None, cache_key=None, unicount=None, symbolic_var_counts=None, symbolic_inst_counts=None, concretized_asts=None, always_concretize=None, never_concretize=None, concretize_at=None, concretization_threshold_memory=None, concretization_threshold_registers=None, concretization_threshold_instruction=None, cooldown_symbolic_stop=2, cooldown_unsupported_stop=2, cooldown_nonunicorn_blocks=100, cooldown_stop_point=1, max_steps=1000000)

Bases: angr.state_plugins.plugin.SimStatePlugin

setup the unicorn engine for a state

Initializes the Unicorn plugin for angr. This plugin handles communication with UnicornEngine.

UC_CONFIG = {}

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

set_state(state)

property uc

static delete_uc()

set_last_block_details(details)

set_stops(stop_points)

set_tracking(track_bbls, track_stack)

hook()

uncache_region(addr, length)

clear_page_cache()

setup()

start(step=None)

finish()

destroy()

set_regs(): setting unicorn registers

setup_flags()

setup_gdt(fs, gs)

read_msr(msr=3221225728)

write_msr(val, msr=3221225728)

get_regs(): loading registers from unicorn

class angr.state_plugins.loop_data.SimStateLoopData(back_edge_trip_counts=None, header_trip_counts=None, current_loop=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

This class keeps track of loop-related information for states. Note that we have 2 counters for loop iterations (trip counts): the first recording the number of times one of the back edges (or continue edges) of a loop is taken, whereas the second recording the number of times the loop header (or loop entry) is executed. These 2 counters may differ since compilers usually optimize loops hence completely change the loop structure at the binary level. This is supposed to be used with LoopSeer exploration technique, which monitors loop execution. For the moment, the only thing we want to analyze is loop trip counts, but nothing prevents us from extending this plugin for other loop analyses.

Parameters

back_edge_trip_counts – Dictionary that stores back edge based trip counts for each loop. Keys are address of loop headers.
header_trip_counts – Dictionary that stores header based trip counts for each loop. Keys are address of loop headers.
current_loop – List of currently running loops. Each element is a tuple (loop object, list of loop exits).

merge(others, merge_conditions, common_ancestor=None)

widen(others)

copy(memo=None, **kwargs)

class angr.state_plugins.concrete.Concrete(segment_registers_initialized=False, segment_registers_callback_initialized=False, whitelist=None, fs_register_bp=None, already_sync_objects_addresses=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

copy(_memo)

merge(_others, _merge_conditions, _common_ancestor=None)

widen(_others)

set_state(state)

sync()

Handle the switch between the concrete execution and angr. This method takes care of: 1- Synchronize registers. 2- Set a concrete target to the memory backer so the memory reads are redirected in the concrete process memory. 3- If possible restore the SimProcedures with the real addresses inside the concrete process. 4- Set an inspect point to sync the segments register as soon as they are read during the symbolic execution. 5- Flush all the pages loaded until now.

Returns

class angr.state_plugins.javavm_classloader.SimJavaVmClassloader(initialized_classes=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

JavaVM Classloader is used as an interface for resolving and initializing Java classes.

get_class(class_name, init_class=False, step_func=None)

Get a class descriptor for the class.

Parameters

class_name (str) – Name of class.
init_class (bool) – Whether the class initializer <clinit> should be executed.
step_func (func) – Callback function executed at every step of the simulation manager during the execution of the main <clinit> method

get_superclass(class_): Get the superclass of the class.

get_class_hierarchy(base_class): Walks up the class hierarchy and returns a list of all classes between base class (inclusive) and java.lang.Object (exclusive).

is_class_initialized(class_): Indicates whether the classes initializing method <clinit> was already executed on the state.

init_class(class_, step_func=None)

This method simulates the loading of a class by the JVM, during which parts of the class (e.g. static fields) are initialized. For this, we run the class initializer method <clinit> (if available) and update the state accordingly.

Note: Initialization is skipped, if the class has already been: initialized (or if it’s not loaded in CLE).

property initialized_classes: List of all initialized classes.

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

class angr.state_plugins.jni_references.SimStateJNIReferences(local_refs=None, global_refs=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

Management of the mapping between opaque JNI references and the corresponding Java objects.

lookup(opaque_ref): Lookups the object that was used for creating the reference.

create_new_reference(obj, global_ref=False)

Create a new reference thats maps to the given object.

Parameters

obj – Object which gets referenced.
global_ref (bool) – Whether a local or global reference is created.

clear_local_references(): Clear all local references.

delete_reference(opaque_ref, global_ref=False)

Delete the stored mapping of a reference.

Parameters

opaque_ref – Reference which should be removed.
global_ref (bool) – Whether opaque_ref is a local or global reference.

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

class angr.state_plugins.heap.heap_base.SimHeapBase(heap_base=None, heap_size=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

This is the base heap class that all heap implementations should subclass. It defines a few handlers for common heap functions (the libc memory management functions). Heap implementations are expected to override these functions regardless of whether they implement the SimHeapLibc interface. For an example, see the SimHeapBrk implementation, which is based on the original libc SimProcedure implementations.

Variables

heap_base – the address of the base of the heap in memory
heap_size – the total size of the main memory region managed by the heap in memory
mmap_base – the address of the region from which large mmap allocations will be made

copy(memo)

init_state()

class angr.state_plugins.heap.heap_brk.SimHeapBrk(heap_base=None, heap_size=None)

Bases: angr.state_plugins.heap.heap_base.SimHeapBase

SimHeapBrk represents a trivial heap implementation based on the Unix brk system call. This type of heap stores virtually no metadata, so it is up to the user to determine when it is safe to release memory. This also means that it does not properly support standard heap operations like realloc.

This heap implementation is a holdover from before any more proper implementations were modelled. At the time, various libc (or win32) SimProcedures handled the heap in the same way that this plugin does now. To make future heap implementations plug-and-playable, they should implement the necessary logic themselves, and dependent SimProcedures should invoke a method by the same name as theirs (prepended with an underscore) upon the heap plugin. Depending on the heap implementation, if the method is not supported, an error should be raised.

Out of consideration for the original way the heap was handled, this plugin implements functionality for all relevant SimProcedures (even those that would not normally be supported together in a single heap implementation).

Variables: heap_location – the address of the top of the heap, bounding the allocations made starting from heap_base

copy(memo=None, **kwargs)

allocate(sim_size)

The actual allocation primitive for this heap implementation. Increases the position of the break to allocate space. Has no guards against the heap growing too large.

Parameters: sim_size – a size specifying how much to increase the break pointer by
Returns: a pointer to the previous break position, above which there is now allocated space

release(sim_size)

The memory release primitive for this heap implementation. Decreases the position of the break to deallocate space. Guards against releasing beyond the initial heap base.

Parameters: sim_size – a size specifying how much to decrease the break pointer by (may be symbolic or not)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

class angr.state_plugins.heap.heap_freelist.Chunk(base, sim_state)

Bases: object

The sort of chunk as would typically be found in a freelist-style heap implementation. Provides a representation of a chunk via a view into the memory plugin. Chunks may be adjacent, in different senses, to as many as four other chunks. For any given chunk, two of these chunks are adjacent to it in memory, and are referred to as the “previous” and “next” chunks throughout this implementation. For any given free chunk, there may also be two significant chunks that are adjacent to it in some linked list of free chunks. These chunks are referred to the “backward” and “foward” chunks relative to the chunk in question.

Variables

base – the location of the base of the chunk in memory
state – the program state that the chunk is resident in

get_size(): Returns the actual size of a chunk (as opposed to the entire size field, which may include some flags).

get_data_size(): Returns the size of the data portion of a chunk.

set_size(size): Sets the size of the chunk, preserving any flags.

data_ptr(): Returns the address of the payload of the chunk.

is_free(): Returns a concrete determination as to whether the chunk is free.

next_chunk(): Returns the chunk immediately following (and adjacent to) this one.

prev_chunk(): Returns the chunk immediately prior (and adjacent) to this one.

fwd_chunk(): Returns the chunk following this chunk in the list of free chunks.

set_fwd_chunk(fwd)

Sets the chunk following this chunk in the list of free chunks.

Parameters: fwd – the chunk to follow this chunk in the list of free chunks

bck_chunk(): Returns the chunk backward from this chunk in the list of free chunks.

set_bck_chunk(bck)

Sets the chunk backward from this chunk in the list of free chunks.

Parameters: bck – the chunk to precede this chunk in the list of free chunks

class angr.state_plugins.heap.heap_freelist.SimHeapFreelist(heap_base=None, heap_size=None)

Bases: angr.state_plugins.heap.heap_libc.SimHeapLibc

A freelist-style heap implementation. Distinguishing features of such heaps include chunks containing heap metadata in addition to user data and at least (but often more than) one linked list of free chunks.

chunks(): Returns an iterator over all the chunks in the heap.

allocated_chunks(): Returns an iterator over all the allocated chunks in the heap.

free_chunks(): Returns an iterator over all the free chunks in the heap.

chunk_from_mem(ptr)

Given a pointer to a user payload, return the chunk associated with that payload.

Parameters: ptr – a pointer to the base of a user payload in the heap
Returns: the associated heap chunk

print_heap_state()

print_all_chunks()

class angr.state_plugins.heap.heap_libc.SimHeapLibc(heap_base=None, heap_size=None)

Bases: angr.state_plugins.heap.heap_base.SimHeapBase

A class of heap that implements the major libc heap management functions.

malloc(sim_size)

A somewhat faithful implementation of libc malloc.

Parameters: sim_size – the amount of memory (in bytes) to be allocated
Returns: the address of the allocation, or a NULL pointer if the allocation failed

free(ptr)

A somewhat faithful implementation of libc free.

Parameters: ptr – the location in memory to be freed

calloc(sim_nmemb, sim_size)

A somewhat faithful implementation of libc calloc.

Parameters

sim_nmemb – the number of elements to allocated
sim_size – the size of each element (in bytes)

Returns

the address of the allocation, or a NULL pointer if the allocation failed

realloc(ptr, size)

A somewhat faithful implementation of libc realloc.

Parameters

ptr – the location in memory to be reallocated
size – the new size desired for the allocation

Returns

the address of the allocation, or a NULL pointer if the allocation was freed or if no new allocation was made

angr.state_plugins.heap.heap_ptmalloc.silence_logger()

angr.state_plugins.heap.heap_ptmalloc.unsilence_logger(level)

class angr.state_plugins.heap.heap_ptmalloc.PTChunk(base, sim_state, heap=None)

Bases: angr.state_plugins.heap.heap_freelist.Chunk

A chunk, inspired by the implementation of chunks in ptmalloc. Provides a representation of a chunk via a view into the memory plugin. For the chunk definitions and docs that this was loosely based off of, see glibc malloc/malloc.c, line 1033, as of commit 5a580643111ef6081be7b4c7bd1997a5447c903f. Alternatively, take the following link. https://sourceware.org/git/?p=glibc.git;a=blob;f=malloc/malloc.c;h=67cdfd0ad2f003964cd0f7dfe3bcd85ca98528a7;hb=5a580643111ef6081be7b4c7bd1997a5447c903f#l1033

Variables

base – the location of the base of the chunk in memory
state – the program state that the chunk is resident in
heap – the heap plugin that the chunk is managed by

get_size()

get_data_size()

set_size(size, is_free=None)

Use this to set the size on a chunk. When the chunk is new (such as when a free chunk is shrunk to form an allocated chunk and a remainder free chunk) it is recommended that the is_free hint be used since setting the size depends on the chunk’s freeness, and vice versa.

Parameters

size – size of the chunk
is_free – boolean indicating the chunk’s freeness

set_prev_freeness(is_free)

Sets (or unsets) the flag controlling whether the previous chunk is free.

Parameters: is_free – if True, sets the previous chunk to be free; if False, sets it to be allocated

is_prev_free()

Returns a concrete state of the flag indicating whether the previous chunk is free or not. Issues a warning if that flag is symbolic and has multiple solutions, and then assumes that the previous chunk is free.

Returns: True if the previous chunk is free; False otherwise

prev_size(): Returns the size of the previous chunk, masking off what would be the flag bits if it were in the actual size field. Performs NO CHECKING to determine whether the previous chunk size is valid (for example, when the previous chunk is not free, its size cannot be determined).

is_free()

data_ptr()

next_chunk()

Returns the chunk immediately following (and adjacent to) this one, if it exists.

Returns: The following chunk, or None if applicable

prev_chunk()

Returns the chunk immediately prior (and adjacent) to this one, if that chunk is free. If the prior chunk is not free, then its base cannot be located and this method raises an error.

Returns: If possible, the previous chunk; otherwise, raises an error

fwd_chunk()

Returns the chunk following this chunk in the list of free chunks. If this chunk is not free, then it resides in no such list and this method raises an error.

Returns: If possible, the forward chunk; otherwise, raises an error

set_fwd_chunk(fwd)

bck_chunk()

Returns the chunk backward from this chunk in the list of free chunks. If this chunk is not free, then it resides in no such list and this method raises an error.

Returns: If possible, the backward chunk; otherwise, raises an error

set_bck_chunk(bck)

class angr.state_plugins.heap.heap_ptmalloc.PTChunkIterator(chunk, cond=<function PTChunkIterator.<lambda>>): Bases: object

class angr.state_plugins.heap.heap_ptmalloc.SimHeapPTMalloc(heap_base=None, heap_size=None)

Bases: angr.state_plugins.heap.heap_freelist.SimHeapFreelist

A freelist-style heap implementation inspired by ptmalloc. The chunks used by this heap contain heap metadata in addition to user data. While the real-world ptmalloc is implemented using multiple lists of free chunks (corresponding to their different sizes), this more basic model uses a single list of chunks and searches for free chunks using a first-fit algorithm.

NOTE: The plugin must be registered using register_plugin with name heap in order to function properly.

Variables

heap_base – the address of the base of the heap in memory
heap_size – the total size of the main memory region managed by the heap in memory
mmap_base – the address of the region from which large mmap allocations will be made
free_head_chunk – the head of the linked list of free chunks in the heap

copy(memo=None, **kwargs)

chunks()

allocated_chunks()

free_chunks()

chunk_from_mem(ptr)

Given a pointer to a user payload, return the base of the chunk associated with that payload (i.e. the chunk pointer). Returns None if ptr is null.

Parameters: ptr – a pointer to the base of a user payload in the heap
Returns: a pointer to the base of the associated heap chunk, or None if ptr is null

malloc(sim_size)

free(ptr)

calloc(sim_nmemb, sim_size)

realloc(ptr, size)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

init_state()

angr.state_plugins.heap.utils.concretize(x, solver, sym_handler)

For now a lot of naive concretization is done when handling heap metadata to keep things manageable. This idiom showed up a lot as a result, so to reduce code repetition this function uses a callback to handle the one or two operations that varied across invocations.

Parameters

x – the item to be concretized
solver – the solver to evaluate the item with
sym_handler – the handler to be used when the item may take on more than one value

Returns

a concrete value for the item

class angr.state_plugins.symbolizer.SimSymbolizer

Bases: angr.state_plugins.plugin.SimStatePlugin

The symbolizer state plugin ensures that pointers that are stored in memory are symbolic. This allows for the tracking of and reasoning over these pointers (for example, to reason about memory disclosure).

init_state()

set_symbolization_for_all_pages(): Sets the symbolizer to symbolize pointers to all pages as they are written to memory..

set_symbolized_target_range(base, length)

All pointers to the target range will be symbolized as they are written to memory.

Due to optimizations, the _pages_ containing this range will be set as symbolization targets, not just the range itself.

resymbolize(): Re-symbolizes all pointers in memory. This can be called to symbolize any pointers to target regions that were written (and not mangled beyond recognition) before symbolization was set.

copy(memo=None, **kwargs)

Storage

class angr.state_plugins.view.SimRegNameView

Bases: angr.state_plugins.plugin.SimStatePlugin

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

get(reg_name)

class angr.state_plugins.view.SimMemView(ty=None, addr=None, state=None)

Bases: angr.state_plugins.plugin.SimStatePlugin

This is a convenient interface with which you can access a program’s memory.

The interface works like this:

You first use [array index notation] to specify the address you’d like to load from

If at that address is a pointer, you may access the deref property to return a SimMemView at the address present in memory.

You then specify a type for the data by simply accesing a property of that name. For a list of supported types, look at state.mem.types.

You can then refine the type. Any type may support any refinement it likes. Right now the only refinements supported are that you may access any member of a struct by its member name, and you may index into a string or array to access that element.

If the address you specified initially points to an array of that type, you can say .array(n) to view the data as an array of n elements.

Finally, extract the structured data with .resolved or .concrete. .resolved will return bitvector values, while .concrete will return integer, string, array, etc values, whatever best represents the data.

Alternately, you may store a value to memory, by assigning to the chain of properties that you’ve constructed. Note that because of the way python works, x = s.mem[...].prop; x = val will NOT work, you must say s.mem[...].prop = val.

For example:

>>> s.mem[0x601048].long
<long (64 bits) <BV64 0x4008d0> at 0x601048>
>>> s.mem[0x601048].long.resolved
<BV64 0x4008d0>
>>> s.mem[0x601048].deref
<<untyped> <unresolvable> at 0x4008d0>
>>> s.mem[0x601048].deref.string.concrete
'SOSNEAKY'

set_state(state)

types = {'CharT': char, '_Bool': bool, '__int128': int128_t, '__int256': int256_t, 'basic_string': string_t, 'bool': bool, 'byte': uint8_t, 'char': char, 'double': double, 'dword': uint32_t, 'float': float, 'int': int, 'int16_t': int16_t, 'int32_t': int32_t, 'int64_t': int64_t, 'int8_t': int8_t, 'long': long, 'long double': double, 'long int': long, 'long long': long long, 'long long int': long long, 'long signed': long, 'long unsigned int': unsigned long, 'ptrdiff_t': long, 'qword': uint64_t, 'short': short, 'short int': short, 'signed': int, 'signed char': char, 'signed int': int, 'signed long': long, 'signed long int': long, 'signed long long': long long, 'signed long long int': long long, 'signed short': short, 'signed short int': short, 'size_t': size_t, 'ssize': size_t, 'ssize_t': size_t, 'string': string_t, 'struct iovec': struct iovec, 'struct timespec': struct timespec, 'struct timeval': struct timeval, 'time_t': long, 'uint16_t': uint16_t, 'uint32_t': uint32_t, 'uint64_t': uint64_t, 'uint8_t': uint8_t, 'uintptr_t': unsigned long, 'unsigned': unsigned int, 'unsigned __int128': uint128_t, 'unsigned __int256': uint256_t, 'unsigned char': char, 'unsigned int': unsigned int, 'unsigned long': unsigned long, 'unsigned long int': unsigned long, 'unsigned long long': unsigned long long, 'unsigned long long int': unsigned long long, 'unsigned short': unsigned short, 'unsigned short int': unsigned short, 'va_list': struct va_list, 'void': void, 'word': uint16_t, 'wstring': wstring_t}

state = None

struct: angr.state_plugins.view.StructMode

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

property resolvable

property resolved

property concrete

property deref: angr.state_plugins.view.SimMemView

array(n) → angr.state_plugins.view.SimMemView

store(value)

class angr.state_plugins.view.StructMode(view): Bases: object

class angr.storage.file.Flags

Bases: object

O_RDONLY = 0

O_WRONLY = 1

O_RDWR = 2

O_ACCMODE = 3

O_APPEND = 1024

O_ASYNC = 8192

O_CLOEXEC = 524288

O_CREAT = 64

O_DIRECT = 16384

O_DIRECTORY = 65536

O_DSYNC = 4096

O_EXCL = 128

O_LARGEFILE = 32768

O_NOATIME = 262144

O_NOCTTY = 256

O_NOFOLLOW = 131072

O_NONBLOCK = 2048

O_NDELAY = 2048

O_PATH = 2097152

O_SYNC = 1052672

O_TMPFILE = 4259840

O_TRUNC = 512

class angr.storage.file.SimFileBase(name=None, writable=True, ident=None, concrete=False, **kwargs)

Bases: angr.state_plugins.plugin.SimStatePlugin

SimFiles are the storage mechanisms used by SimFileDescriptors.

Different types of SimFiles can have drastically different interfaces, and as a result there’s not much that can be specified on this base class. All the read and write methods take a pos argument, which may have different semantics per-class. 0 will always be a valid position to use, though, and the next position you should use is part of the return tuple.

Some simfiles are “streams”, meaning that the position that reads come from is determined not by the position you pass in (it will in fact be ignored), but by an internal variable. This is stored as .pos if you care to read it. Don’t write to it. The same lack-of-semantics applies to this field as well.

Variables

name – The name of the file. Purely for cosmetic purposes
ident – The identifier of the file, typically autogenerated from the name and a nonce. Purely for cosmetic purposes, but does appear in symbolic values autogenerated in the file.
seekable – Bool indicating whether seek operations on this file should succeed. If this is True, then pos must be a number of bytes from the start of the file.
writable – Bool indicating whether writing to this file is allowed.
pos – If the file is a stream, this will be the current position. Otherwise, None.
concrete – Whether or not this file contains mostly concrete data. Will be used by some SimProcedures to choose how to handle variable-length operations like fgets.

seekable = False

pos = None

static make_ident(name)

concretize(**kwargs): Return a concretization of the contents of the file. The type of the return value of this method will vary depending on which kind of SimFile you’re using.

read(pos, size, **kwargs)

Read some data from the file.

Parameters

pos – The offset in the file to read from.
size – The size to read. May be symbolic.

Returns

A tuple of the data read (a bitvector of the length that is the maximum length of the read), the actual size of the read, and the new file position pointer.

write(pos, data, size=None, **kwargs)

Write some data to the file.

Parameters

pos – The offset in the file to write to. May be ignored if the file is a stream or device.
data – The data to write as a bitvector
size – The optional size of the data to write. If not provided will default to the length of the data. Must be constrained to less than or equal to the size of the data.

Returns

The new file position pointer.

property size: The number of data bytes stored by the file at present. May be a symbolic value.

copy(memo=None, **kwargs)

class angr.storage.file.SimFile(name=None, content=None, size=None, has_end=None, seekable=True, writable=True, ident=None, concrete=None, **kwargs)

Bases: angr.storage.file.SimFileBase, angr.storage.memory_mixins.DefaultMemory

The normal SimFile is meant to model files on disk. It subclasses SimSymbolicMemory so loads and stores to/from it are very simple.

Parameters

name – The name of the file
content – Optional initial content for the file as a string or bitvector
size – Optional size of the file. If content is not specified, it defaults to zero
has_end – Whether the size boundary is treated as the end of the file or a frontier at which new content will be generated. If unspecified, will pick its value based on options.FILES_HAVE_EOF. Another caveat is that if the size is also unspecified this value will default to False.
seekable – Optional bool indicating whether seek operations on this file should succeed, default True.
writable – Whether writing to this file is allowed
concrete – Whether or not this file contains mostly concrete data. Will be used by some SimProcedures to choose how to handle variable-length operations like fgets.

Variables

has_end – Whether this file has an EOF

property category

set_state(state)

property size

concretize(**kwargs): Return a concretization of the contents of the file, as a flat bytestring.

read(pos, size, **kwargs)

write(pos, data, size=None, events=True, **kwargs)

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(_)

class angr.storage.file.SimFileStream(name=None, content=None, pos=0, **kwargs)

Bases: angr.storage.file.SimFile

A specialized SimFile that uses a flat memory backing, but functions as a stream, tracking its position internally.

The pos argument to the read and write methods will be ignored, and will return None. Instead, there is an attribute pos on the file itself, which will give you what you want.

Parameters

name – The name of the file, for cosmetic purposes
pos – The initial position of the file, default zero
kwargs – Any other keyword arguments will go on to the SimFile constructor.

Variables

pos – The current position in the file.

set_state(state)

read(pos, size, **kwargs)

write(_, data, size=None, **kwargs)

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

class angr.storage.file.SimPackets(name, write_mode=None, content=None, writable=True, ident=None, **kwargs)

Bases: angr.storage.file.SimFileBase

The SimPackets is meant to model inputs whose content is delivered a series of asynchronous chunks. The data is stored as a list of read or write results. For symbolic sizes, state.libc.max_packet_size will be respected. If the SHORT_READS option is enabled, reads will return a symbolic size constrained to be less than or equal to the requested size.

A SimPackets cannot be used for both reading and writing - for socket objects that can be both read and written to you should use a file descriptor to multiplex the read and write operations into two separate file storage mechanisms.

Parameters

name – The name of the file, for cosmetic purposes
write_mode – Whether this file is opened in read or write mode. If this is unspecified it will be autodetected.
content – Some initial content to use for the file. Can be a list of bytestrings or a list of tuples of content ASTs and size ASTs.

Variables

write_mode – See the eponymous parameter
content – A list of packets, as tuples of content ASTs and size ASTs.

set_state(state)

property size

concretize(**kwargs): Returns a list of the packets read or written as bytestrings.

read(pos, size, **kwargs)

Read a packet from the stream.

Parameters

pos (int) – The packet number to read from the sequence of the stream. May be None to append to the stream.
size – The size to read. May be symbolic.
short_reads – Whether to replace the size with a symbolic value constrained to less than or equal to the original size. If unspecified, will be chosen based on the state option.

Returns

A tuple of the data read (a bitvector of the length that is the maximum length of the read) and the actual size of the read.

write(pos, data, size=None, events=True, **kwargs)

Write a packet to the stream.

Parameters

pos (int) – The packet number to write in the sequence of the stream. May be None to append to the stream.
data – The data to write, as a string or bitvector.
size – The optional size to write. May be symbolic; must be constrained to at most the size of data.

Returns

The next packet to use after this

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(_)

class angr.storage.file.SimPacketsStream(name, pos=0, **kwargs)

Bases: angr.storage.file.SimPackets

A specialized SimPackets that tracks its position internally.

The pos argument to the read and write methods will be ignored, and will return None. Instead, there is an attribute pos on the file itself, which will give you what you want.

Parameters

name – The name of the file, for cosmetic purposes
pos – The initial position of the file, default zero
kwargs – Any other keyword arguments will go on to the SimPackets constructor.

Variables

pos – The current position in the file.

read(pos, size, **kwargs)

write(_, data, size=None, **kwargs)

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

class angr.storage.file.SimFileDescriptorBase

Bases: angr.state_plugins.plugin.SimStatePlugin

The base class for implementations of POSIX file descriptors.

All file descriptors should respect the CONCRETIZE_SYMBOLIC_{READ,WRITE}_SIZES state options.

read(pos, size, **kwargs)

Reads some data from the file, storing it into memory.

Parameters

pos – The address to write the read data into memory
size – The requested length of the read

Returns

The real length of the read

write(pos, size, **kwargs)

Writes some data, loaded from the state, into the file.

Parameters

pos – The address to read the data to write from in memory
size – The requested size of the write

Returns

The real length of the write

read_data(size, **kwargs)

Reads some data from the file, returning the data.

Parameters: size – The requested length of the read
Returns: A tuple of the data read and the real length of the read

write_data(data, size=None, **kwargs)

Write some data, provided as an argument into the file.

Parameters

data – A bitvector to write into the file
size – The requested size of the write (may be symbolic)

Returns

The real length of the write

seek(offset, whence='start')

Seek the file descriptor to a different position in the file.

Parameters

offset – The offset to seek to, interpreted according to whence
whence – What the offset is relative to; one of the strings “start”, “current”, or “end”

Returns

A symbolic boolean describing whether the seek succeeded or not

tell(): Return the current position, or None if the concept doesn’t make sense for the given file.

eof(): Return the EOF status. May be a symbolic boolean.

size(): Return the size of the data stored in the file in bytes, or None if the concept doesn’t make sense for the given file.

property read_storage: Return the SimFile backing reads from this fd

property write_storage: Return the SimFile backing writes to this fd

property read_pos

Return the current position of the read file pointer.

If the underlying read file is a stream, this will return the position of the stream. Otherwise, will return the position of the file descriptor in the file.

property write_pos

Return the current position of the read file pointer.

If the underlying read file is a stream, this will return the position of the stream. Otherwise, will return the position of the file descriptor in the file.

concretize(**kwargs)

Return a concretizeation of the data in the underlying file. Has different return types to represent differnt data structures on a per-class basis.

Any arguments passed to this will be passed onto state.solver.eval.

class angr.storage.file.SimFileDescriptor(simfile, flags=0)

Bases: angr.storage.file.SimFileDescriptorBase

A simple file descriptor forwarding reads and writes to a SimFile. Contains information about the current opened state of the file, such as the flags or (if relevant) the current position.

Variables

file – The SimFile described to by this descriptor
flags – The mode that the file descriptor was opened with, a bitfield of flags

read_data(size, **kwargs)

write_data(data, size=None, **kwargs)

seek(offset, whence='start')

eof()

tell()

size()

concretize(**kwargs): Return a concretization of the underlying file. Returns whatever format is preferred by the file.

property read_storage

property write_storage

property read_pos

property write_pos

set_state(state)

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(_)

class angr.storage.file.SimFileDescriptorDuplex(read_file, write_file)

Bases: angr.storage.file.SimFileDescriptorBase

A file descriptor that refers to two file storage mechanisms, one to read from and one to write to. As a result, operations like seek, eof, etc no longer make sense.

Parameters

read_file – The SimFile to read from
write_file – The SimFile to write to

read_data(size, **kwargs)

write_data(data, size=None, **kwargs)

set_state(state)

eof()

tell()

seek(offset, whence='start')

size()

concretize(**kwargs): Return a concretization of the underlying files, as a tuple of (read file, write file).

property read_storage

property write_storage

property read_pos

property write_pos

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(_)

class angr.storage.file.SimPacketsSlots(name, read_sizes, ident=None, **kwargs)

Bases: angr.storage.file.SimFileBase

SimPacketsSlots is the new SimDialogue, if you’ve ever seen that before.

The idea is that in some cases, the only thing you really care about is getting the lengths of reads right, and some of them should be short reads, and some of them should be truncated. You provide to this class a list of read lengths, and it figures out the length of each read, and delivers some content.

This class will NOT respect the position argument you pass it - this storage is not stateless.

seekable = False

concretize(**kwargs)

read(pos, size, **kwargs)

write(pos, data, size=None, **kwargs)

property size

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(_)

angr.storage.memory_object.obj_bit_size(o)

class angr.storage.memory_object.SimMemoryObject(obj, base, endness, length=None, byte_width=8)

Bases: object

A SimMemoryObject is a reference to a byte or several bytes in a specific object in memory. It should be used only by the bottom layer of memory.

is_bytes

base

object

length

endness

size()

property variables

property cache_key

property symbolic

property last_addr

includes(x)

bytes_at(addr, length, allow_concrete=False, endness='Iend_BE')

class angr.storage.memory_object.SimLabeledMemoryObject(obj, base, endness, length=None, byte_width=8, label=None)

Bases: angr.storage.memory_object.SimMemoryObject

label

angr.storage.memory_object.bv_slice(value, offset, size, rev, bw)

Extremely cute utility to pretend you’ve serialized a value to stored bytes, sliced it a la python slicing, and then deserialized those bytes to an integer again.

Parameters

value – The bitvector to slice
offset – The byte offset from the first stored byte to slice from, or a negative offset from the end.
size – The number of bytes to return. If None, return all bytes from the offset to the end. If larger than the number of bytes from the offset to the end, return all bytes from the offset to the end.
rev – Whether the pretend-serialization should be little-endian
bw – The byte width

Returns

The new bitvector

class angr.storage.pcap.PCAP(path, ip_port_tup, init=True)

Bases: object

initialize(path)

recv(length)

copy()

class angr.concretization_strategies.SimConcretizationStrategy(filter=None, exact=True)

Bases: object

Concretization strategies control the resolution of symbolic memory indices in SimuVEX. By subclassing this class and setting it as a concretization strategy (on state.memory.read_strategies and state.memory.write_strategies), SimuVEX’s memory index concretization behavior can be modified.

Initializes the base SimConcretizationStrategy.

Parameters

filter – A function, taking arguments of (SimMemory, claripy.AST) that determins if this strategy can handle resolving the provided AST.
exact – A flag (default: True) that determines if the convenience resolution functions provided by this class use exact or approximate resolution.

concretize(memory, addr, **kwargs): Concretizes the address into a list of values. If this strategy cannot handle this address, returns None.

copy(): Returns a copy of the strategy, if there is data that should be kept separate between states. If not, returns self.

merge(others): Merges this strategy with others (if there is data that should be kept separate between states. If not, is a no-op.

Memory Mixins

class angr.storage.memory_mixins.MemoryMixin(memory_id=None, endness='Iend_BE')

Bases: angr.state_plugins.plugin.SimStatePlugin

SUPPORTS_CONCRETE_LOAD = False

copy(memo)

property category

reg, mem, or file.

Type: Return the category of this SimMemory instance. It can be one of the three following categories

property variable_key_prefix

find(addr, data, max_search, **kwargs)

load(addr, **kwargs)

store(addr, data, **kwargs)

merge(others, merge_conditions, common_ancestor=None) → bool

widen(others)

permissions(addr, permissions=None, **kwargs)

map_region(addr, length, permissions, init_zero=False, **kwargs)

unmap_region(addr, length, **kwargs)

concrete_load(addr, size, writing=False, **kwargs) → memoryview

Set SUPPORTS_CONCRETE_LOAD to True and implement concrete_load if reading concrete bytes is faster in this memory model.

Parameters

addr – The address to load from.
size – Size of the memory read.
writing –

Returns

A memoryview into the loaded bytes.

erase(addr, size=None, **kwargs) → None

Set [addr:addr+size) to uninitialized. In many cases this will be faster than overwriting those locations with new values. This is commonly used during static data flow analysis.

Parameters

addr – The address to start erasing.
size – The number of bytes for erasing.

Returns

None

replace_all(old: claripy.ast.bv.BV, new: claripy.ast.bv.BV)

copy_contents(dst, src, size, condition=None, **kwargs)

Override this method to provide faster copying of large chunks of data.

Parameters

dst – The destination of copying.
src – The source of copying.
size – The size of copying.
condition – The storing condition.
kwargs – Other parameters.

Returns

None

class angr.storage.memory_mixins.DefaultMemory(*args, **kwargs): Bases: angr.storage.memory_mixins.hex_dumper_mixin.HexDumperMixin, angr.storage.memory_mixins.smart_find_mixin.SmartFindMixin, angr.storage.memory_mixins.unwrapper_mixin.UnwrapperMixin, angr.storage.memory_mixins.name_resolution_mixin.NameResolutionMixin, angr.storage.memory_mixins.bvv_conversion_mixin.DataNormalizationMixin, angr.storage.memory_mixins.simplification_mixin.SimplificationMixin, angr.storage.memory_mixins.clouseau_mixin.InspectMixinHigh, angr.storage.memory_mixins.actions_mixin.ActionsMixinHigh, angr.storage.memory_mixins.underconstrained_mixin.UnderconstrainedMixin, angr.storage.memory_mixins.size_resolution_mixin.SizeConcretizationMixin, angr.storage.memory_mixins.size_resolution_mixin.SizeNormalizationMixin, angr.storage.memory_mixins.address_concretization_mixin.AddressConcretizationMixin, angr.storage.memory_mixins.actions_mixin.ActionsMixinLow, angr.storage.memory_mixins.conditional_store_mixin.ConditionalMixin, angr.storage.memory_mixins.convenient_mappings_mixin.ConvenientMappingsMixin, angr.storage.memory_mixins.dirty_addrs_mixin.DirtyAddrsMixin, angr.storage.memory_mixins.paged_memory.stack_allocation_mixin.StackAllocationMixin, angr.storage.memory_mixins.paged_memory.page_backer_mixins.ClemoryBackerMixin, angr.storage.memory_mixins.paged_memory.page_backer_mixins.DictBackerMixin, angr.storage.memory_mixins.paged_memory.privileged_mixin.PrivilegedPagingMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.UltraPagesMixin, angr.storage.memory_mixins.default_filler_mixin.DefaultFillerMixin, angr.storage.memory_mixins.symbolic_merger_mixin.SymbolicMergerMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

class angr.storage.memory_mixins.DefaultListPagesMemory(*args, **kwargs): Bases: angr.storage.memory_mixins.hex_dumper_mixin.HexDumperMixin, angr.storage.memory_mixins.smart_find_mixin.SmartFindMixin, angr.storage.memory_mixins.unwrapper_mixin.UnwrapperMixin, angr.storage.memory_mixins.name_resolution_mixin.NameResolutionMixin, angr.storage.memory_mixins.bvv_conversion_mixin.DataNormalizationMixin, angr.storage.memory_mixins.simplification_mixin.SimplificationMixin, angr.storage.memory_mixins.actions_mixin.ActionsMixinHigh, angr.storage.memory_mixins.underconstrained_mixin.UnderconstrainedMixin, angr.storage.memory_mixins.size_resolution_mixin.SizeConcretizationMixin, angr.storage.memory_mixins.size_resolution_mixin.SizeNormalizationMixin, angr.storage.memory_mixins.clouseau_mixin.InspectMixinHigh, angr.storage.memory_mixins.address_concretization_mixin.AddressConcretizationMixin, angr.storage.memory_mixins.actions_mixin.ActionsMixinLow, angr.storage.memory_mixins.conditional_store_mixin.ConditionalMixin, angr.storage.memory_mixins.convenient_mappings_mixin.ConvenientMappingsMixin, angr.storage.memory_mixins.dirty_addrs_mixin.DirtyAddrsMixin, angr.storage.memory_mixins.paged_memory.stack_allocation_mixin.StackAllocationMixin, angr.storage.memory_mixins.paged_memory.page_backer_mixins.ClemoryBackerMixin, angr.storage.memory_mixins.paged_memory.page_backer_mixins.DictBackerMixin, angr.storage.memory_mixins.paged_memory.privileged_mixin.PrivilegedPagingMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.ListPagesMixin, angr.storage.memory_mixins.default_filler_mixin.DefaultFillerMixin, angr.storage.memory_mixins.symbolic_merger_mixin.SymbolicMergerMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

class angr.storage.memory_mixins.FastMemory(uninitialized_read_handler=None, **kwargs): Bases: angr.storage.memory_mixins.name_resolution_mixin.NameResolutionMixin, angr.storage.memory_mixins.simple_interface_mixin.SimpleInterfaceMixin, angr.storage.memory_mixins.simplification_mixin.SimplificationMixin, angr.storage.memory_mixins.clouseau_mixin.InspectMixinHigh, angr.storage.memory_mixins.conditional_store_mixin.ConditionalMixin, angr.storage.memory_mixins.default_filler_mixin.ExplicitFillerMixin, angr.storage.memory_mixins.default_filler_mixin.DefaultFillerMixin, angr.storage.memory_mixins.slotted_memory.SlottedMemoryMixin

class angr.storage.memory_mixins.AbstractMemory(*args, **kwargs): Bases: angr.storage.memory_mixins.unwrapper_mixin.UnwrapperMixin, angr.storage.memory_mixins.name_resolution_mixin.NameResolutionMixin, angr.storage.memory_mixins.bvv_conversion_mixin.DataNormalizationMixin, angr.storage.memory_mixins.simplification_mixin.SimplificationMixin, angr.storage.memory_mixins.clouseau_mixin.InspectMixinHigh, angr.storage.memory_mixins.actions_mixin.ActionsMixinHigh, angr.storage.memory_mixins.underconstrained_mixin.UnderconstrainedMixin, angr.storage.memory_mixins.size_resolution_mixin.SizeConcretizationMixin, angr.storage.memory_mixins.size_resolution_mixin.SizeNormalizationMixin, angr.storage.memory_mixins.actions_mixin.ActionsMixinLow, angr.storage.memory_mixins.conditional_store_mixin.ConditionalMixin, angr.storage.memory_mixins.regioned_memory.regioned_address_concretization_mixin.RegionedAddressConcretizationMixin, angr.storage.memory_mixins.regioned_memory.regioned_memory_mixin.RegionedMemoryMixin

class angr.storage.memory_mixins.RegionedMemory(related_function_addr=None, **kwargs): Bases: angr.storage.memory_mixins.regioned_memory.region_category_mixin.RegionCategoryMixin, angr.storage.memory_mixins.regioned_memory.region_meta_mixin.MemoryRegionMetaMixin, angr.storage.memory_mixins.regioned_memory.static_find_mixin.StaticFindMixin, angr.storage.memory_mixins.unwrapper_mixin.UnwrapperMixin, angr.storage.memory_mixins.name_resolution_mixin.NameResolutionMixin, angr.storage.memory_mixins.bvv_conversion_mixin.DataNormalizationMixin, angr.storage.memory_mixins.simplification_mixin.SimplificationMixin, angr.storage.memory_mixins.size_resolution_mixin.SizeConcretizationMixin, angr.storage.memory_mixins.size_resolution_mixin.SizeNormalizationMixin, angr.storage.memory_mixins.address_concretization_mixin.AddressConcretizationMixin, angr.storage.memory_mixins.convenient_mappings_mixin.ConvenientMappingsMixin, angr.storage.memory_mixins.dirty_addrs_mixin.DirtyAddrsMixin, angr.storage.memory_mixins.paged_memory.page_backer_mixins.ClemoryBackerMixin, angr.storage.memory_mixins.paged_memory.page_backer_mixins.DictBackerMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.UltraPagesMixin, angr.storage.memory_mixins.default_filler_mixin.DefaultFillerMixin, angr.storage.memory_mixins.regioned_memory.abstract_merger_mixin.AbstractMergerMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

class angr.storage.memory_mixins.LabeledMemory(*args, top_func=None, **kwargs)

Bases: angr.storage.memory_mixins.size_resolution_mixin.SizeNormalizationMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.ListPagesWithLabelsMixin, angr.storage.memory_mixins.default_filler_mixin.DefaultFillerMixin, angr.storage.memory_mixins.top_merger_mixin.TopMergerMixin, angr.storage.memory_mixins.label_merger_mixin.LabelMergerMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

LabeledMemory is used in static analysis. It allows storing values with labels, such as Definition.

class angr.storage.memory_mixins.MultiValuedMemory(*args, skip_missing_values_during_merging=False, **kwargs): Bases: angr.storage.memory_mixins.size_resolution_mixin.SizeNormalizationMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.MVListPagesMixin, angr.storage.memory_mixins.default_filler_mixin.DefaultFillerMixin, angr.storage.memory_mixins.multi_value_merger_mixin.MultiValueMergerMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

class angr.storage.memory_mixins.KeyValueMemory(*args, **kwargs): Bases: angr.storage.memory_mixins.keyvalue_memory.keyvalue_memory_mixin.KeyValueMemoryMixin

class angr.storage.memory_mixins.JavaVmMemory(memory_id='mem', stack=None, heap=None, vm_static_table=None, load_strategies=None, store_strategies=None, max_array_size=1000, **kwargs): Bases: angr.storage.memory_mixins.javavm_memory.javavm_memory_mixin.JavaVmMemoryMixin

class angr.storage.memory_mixins.name_resolution_mixin.NameResolutionMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

This mixin allows you to provide register names as load addresses, and will automatically translate this to an offset and size.

store(addr, data, size=None, **kwargs)

load(addr, size=None, **kwargs)

class angr.storage.memory_mixins.smart_find_mixin.SmartFindMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

find(addr, needle, max_search, default=None, endness=None, chunk_size=None, max_symbolic_bytes=None, condition=None, char_size=1, **kwargs)

class angr.storage.memory_mixins.default_filler_mixin.DefaultFillerMixin(memory_id=None, endness='Iend_BE'): Bases: angr.storage.memory_mixins.MemoryMixin

class angr.storage.memory_mixins.default_filler_mixin.SpecialFillerMixin(special_memory_filler=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

copy(memo)

class angr.storage.memory_mixins.default_filler_mixin.ExplicitFillerMixin(uninitialized_read_handler=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

copy(memo)

class angr.storage.memory_mixins.bvv_conversion_mixin.DataNormalizationMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

Normalizes the data field for a store and the fallback field for a load to be BVs.

store(addr, data, size=None, **kwargs)

load(addr, size=None, fallback=None, **kwargs)

class angr.storage.memory_mixins.hex_dumper_mixin.HexDumperMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

hex_dump(start, size, word_size=4, words_per_row=4, endianness='Iend_BE', symbolic_char='?', unprintable_char='.', solve=False, extra_constraints=None, inspect=False, disable_actions=True)

Returns a hex dump as a string. The solver, if enabled, is called once for every byte potentially making this function very slow. It is meant to be used mainly as a “visualization” for debugging.

Warning: May read and display more bytes than size due to rounding. Particularly, if size is less than, or not a multiple of word_size*words_per_line.

Parameters

start – starting address from which to print
size – number of bytes to display
word_size – number of bytes to group together as one space-delimited unit
words_per_row – number of words to display per row of output
endianness – endianness to use when displaying each word (ASCII representation is unchanged)
symbolic_char – the character to display when a byte is symbolic and has multiple solutions
unprintable_char – the character to display when a byte is not printable
solve – whether or not to attempt to solve (warning: can be very slow)
extra_constraints – extra constraints to pass to the solver is solve is True
inspect – whether or not to trigger SimInspect breakpoints for the memory load
disable_actions – whether or not to disable SimActions for the memory load

Returns

hex dump as a string

class angr.storage.memory_mixins.underconstrained_mixin.UnderconstrainedMixin(*args, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

copy(memo=None, **kwargs)

load(addr, **kwargs)

store(addr, data, **kwargs)

class angr.storage.memory_mixins.simple_interface_mixin.SimpleInterfaceMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

load(addr, size=None, endness=None, condition=None, fallback=None, **kwargs)

store(addr, data, size=None, endness=None, condition=None, **kwargs)

class angr.storage.memory_mixins.actions_mixin.ActionsMixinHigh(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

load(addr, size=None, condition=None, fallback=None, disable_actions=False, action=None, **kwargs)

store(addr, data, size=None, disable_actions=False, action=None, condition=None, **kwargs)

class angr.storage.memory_mixins.actions_mixin.ActionsMixinLow(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

load(addr, action=None, **kwargs)

store(addr, data, action: Optional[angr.state_plugins.sim_action.SimActionData] = None, **kwargs)

class angr.storage.memory_mixins.symbolic_merger_mixin.SymbolicMergerMixin(memory_id=None, endness='Iend_BE'): Bases: angr.storage.memory_mixins.MemoryMixin

class angr.storage.memory_mixins.size_resolution_mixin.SizeNormalizationMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

Provides basic services related to normalizing sizes. After this mixin, sizes will always be a plain int. Assumes that the data is a BV.

load will throw a TypeError if no size is provided
store will default to len(data)//byte_width if no size is provided

load(addr, size=None, **kwargs)

store(addr, data, size=None, **kwargs)

class angr.storage.memory_mixins.size_resolution_mixin.SizeConcretizationMixin(concretize_symbolic_write_size: bool = False, max_concretize_count: Optional[int] = 256, max_symbolic_size: int = 4194304, raise_memory_limit_error: bool = False, size_limit: int = 257, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

This mixin allows memory to process symbolic sizes. It will not touch any sizes which are not ASTs with non-BVV ops. Assumes that the data is a BV.

symbolic load sizes will be concretized as their maximum and a warning will be logged
symbolic store sizes will be dispatched as several conditional stores with concrete sizes

copy(memo)

load(addr, size=None, **kwargs)

store(addr, data, size=None, condition=None, **kwargs)

class angr.storage.memory_mixins.dirty_addrs_mixin.DirtyAddrsMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

store(addr, data, size=None, **kwargs)

class angr.storage.memory_mixins.address_concretization_mixin.MultiwriteAnnotation

Bases: claripy.annotation.Annotation

property eliminatable

property relocateable

class angr.storage.memory_mixins.address_concretization_mixin.AddressConcretizationMixin(read_strategies=None, write_strategies=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

The address concretization mixin allows symbolic reads and writes to be handled sanely by dispatching them as a number of conditional concrete reads/writes. It provides a “concretization strategies” interface allowing the process of serializing symbolic addresses into concrete ones to be specified.

set_state(state)

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None) → bool

concretize_write_addr(addr, strategies=None, condition=None)

Concretizes an address meant for writing.

Parameters

addr – An expression for the address.
strategies – A list of concretization strategies (to override the default).
condition – Any extra constraints that should be observed when determining address satisfiability

Returns

A list of concrete addresses.

concretize_read_addr(addr, strategies=None, condition=None)

Concretizes an address meant for reading.

Parameters

addr – An expression for the address.
strategies – A list of concretization strategies (to override the default).

Returns

A list of concrete addresses.

load(addr, size=None, condition=None, **kwargs)

store(addr, data, size=None, condition=None, **kwargs)

permissions(addr, permissions=None, **kwargs)

map_region(addr, length, permissions, **kwargs)

unmap_region(addr, length, **kwargs)

concrete_load(addr, size, *args, **kwargs)

class angr.storage.memory_mixins.clouseau_mixin.InspectMixinHigh(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

store(addr, data, size=None, condition=None, endness=None, inspect=True, **kwargs)

load(addr, size=None, condition=None, endness=None, inspect=True, **kwargs)

class angr.storage.memory_mixins.conditional_store_mixin.ConditionalMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

load(addr, condition=None, fallback=None, **kwargs)

store(addr, data, size=None, condition=None, **kwargs)

class angr.storage.memory_mixins.label_merger_mixin.LabelMergerMixin(*args, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

A memory mixin for merging labels. Labels come from SimLabeledMemoryObject.

copy(memo=None)

class angr.storage.memory_mixins.simplification_mixin.SimplificationMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

store(addr, data, **kwargs)

class angr.storage.memory_mixins.unwrapper_mixin.UnwrapperMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

This mixin processes SimActionObjects by passing on their .ast field.

store(addr, data, size=None, condition=None, **kwargs)

load(addr, size=None, condition=None, fallback=None, **kwargs)

find(addr, what, max_search, default=None, **kwargs)

copy_contents(dst, src, size, condition=None, **kwargs)

class angr.storage.memory_mixins.convenient_mappings_mixin.ConvenientMappingsMixin(**kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

Implements mappings between names and hashes of symbolic variables and these variables themselves.

copy(memo)

store(addr, data, size=None, **kwargs)

get_symbolic_addrs()

addrs_for_name(n): Returns addresses that contain expressions that contain a variable named n.

addrs_for_hash(h): Returns addresses that contain expressions that contain a variable with the hash of h.

replace_all(old: claripy.ast.bv.BV, new: claripy.ast.bv.BV)

Replaces all instances of expression old with expression new.

Parameters

old – A claripy expression. Must contain at least one named variable (to make it possible to use the name index for speedup).
new – The new variable to replace it with.

class angr.storage.memory_mixins.paged_memory.pages.mv_list_page.MVListPage(memory=None, content=None, sinkhole=None, mo_cmp=None, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.pages.cooperation.MemoryObjectSetMixin, angr.storage.memory_mixins.paged_memory.pages.PageBase

MVListPage allows storing multiple values at the same location, thus allowing weak updates.

Each store() may take a value or multiple values, and a “weak” parameter to specify if this store is a weak update or not. Each load() returns an iterator of all values stored at that location.

copy(memo) → angr.storage.memory_mixins.paged_memory.pages.mv_list_page.MVListPage

load(addr, size=None, endness=None, page_addr=None, memory=None, cooperate=False, **kwargs) → List[Tuple[int, angr.storage.memory_object.SimMemoryObject]]

store(addr, data, size=None, endness=None, memory=None, cooperate=False, weak=False, **kwargs)

erase(addr, size=None, **kwargs) → None

merge(others: List[angr.storage.memory_mixins.paged_memory.pages.mv_list_page.MVListPage], merge_conditions, common_ancestor=None, page_addr: Optional[int] = None, memory=None, changed_offsets: Optional[Set[int]] = None)

changed_bytes(other: angr.storage.memory_mixins.paged_memory.pages.mv_list_page.MVListPage, page_addr: Optional[int] = None)

content_gen(index)

class angr.storage.memory_mixins.paged_memory.pages.multi_values.MultiValues(offset_to_values=None)

Bases: object

values: Dict[int, Set[claripy.ast.base.Base]]

add_value(offset, value) → None

one_value() → Optional[claripy.ast.base.Base]

merge(mv: angr.storage.memory_mixins.paged_memory.pages.multi_values.MultiValues) → angr.storage.memory_mixins.paged_memory.pages.multi_values.MultiValues

class angr.storage.memory_mixins.top_merger_mixin.TopMergerMixin(*args, top_func=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

A memory mixin for merging values in memory to TOP.

copy(memo=None)

class angr.storage.memory_mixins.multi_value_merger_mixin.MultiValueMergerMixin(*args, element_limit=5, top_func=None, phi_maker=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

copy(memo=None)

class angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin(page_size=4096, default_permissions=3, permissions_map=None, page_kwargs=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

A bottom-level storage mechanism. Dispatches reads to individual pages, the type of which is the PAGE_TYPE class variable.

SUPPORTS_CONCRETE_LOAD = True

PAGE_TYPE: Type[PageType] = None

copy(memo=None, **kwargs)

load(addr: int, size: Optional[int] = None, endness=None, **kwargs)

store(addr: int, data, size: Optional[int] = None, endness=None, **kwargs)

erase(addr, size=None, **kwargs) → None

merge(others: Iterable[angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin], merge_conditions, common_ancestor=None) → bool

permissions(addr, permissions=None, **kwargs)

map_region(addr, length, permissions, init_zero=False, **kwargs)

unmap_region(addr, length, **kwargs)

concrete_load(addr, size, writing=False, with_bitmap=False, **kwargs)

changed_bytes(other) → Set[int]

changed_pages(other) → Dict[int, Optional[Set[int]]]

copy_contents(dst, src, size, condition=None, **kwargs)

flush_pages(white_list): Flush all pages not included in the white_list by removing their pages. Note, this will not wipe them from memory if they were backed by a memory_backer, it will simply reset them to their initial state. Returns the list of pages that were cleared consisting of (addr, length) tuples. :param white_list: white list of regions in the form of (start, end) to exclude from the flush :return: a list of memory page ranges that were flushed :rtype: list

class angr.storage.memory_mixins.paged_memory.paged_memory_mixin.LabeledPagesMixin(page_size=4096, default_permissions=3, permissions_map=None, page_kwargs=None, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

load_with_labels(addr: int, size: Optional[int] = None, endness=None, **kwargs) → Tuple[claripy.ast.base.Base, Tuple[Tuple[int, int, int, Any]]]

class angr.storage.memory_mixins.paged_memory.paged_memory_mixin.ListPagesMixin(page_size=4096, default_permissions=3, permissions_map=None, page_kwargs=None, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

PAGE_TYPE: alias of angr.storage.memory_mixins.paged_memory.pages.list_page.ListPage

class angr.storage.memory_mixins.paged_memory.paged_memory_mixin.MVListPagesMixin(*args, skip_missing_values_during_merging=False, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

PAGE_TYPE: alias of angr.storage.memory_mixins.paged_memory.pages.mv_list_page.MVListPage

copy(memo=None, **kwargs)

class angr.storage.memory_mixins.paged_memory.paged_memory_mixin.ListPagesWithLabelsMixin(page_size=4096, default_permissions=3, permissions_map=None, page_kwargs=None, **kwargs): Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.LabeledPagesMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.ListPagesMixin

class angr.storage.memory_mixins.paged_memory.paged_memory_mixin.MVListPagesWithLabelsMixin(*args, skip_missing_values_during_merging=False, **kwargs): Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.LabeledPagesMixin, angr.storage.memory_mixins.paged_memory.paged_memory_mixin.MVListPagesMixin

class angr.storage.memory_mixins.paged_memory.paged_memory_mixin.UltraPagesMixin(page_size=4096, default_permissions=3, permissions_map=None, page_kwargs=None, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

PAGE_TYPE: alias of angr.storage.memory_mixins.paged_memory.pages.ultra_page.UltraPage

class angr.storage.memory_mixins.paged_memory.page_backer_mixins.NotMemoryview(obj, offset, size): Bases: object

class angr.storage.memory_mixins.paged_memory.page_backer_mixins.ClemoryBackerMixin(cle_memory_backer: Union[None, cle.loader.Loader, cle.memory.Clemory] = None, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

copy(memo)

class angr.storage.memory_mixins.paged_memory.page_backer_mixins.DictBackerMixin(dict_memory_backer=None, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

copy(memo)

class angr.storage.memory_mixins.paged_memory.stack_allocation_mixin.StackAllocationMixin(stack_end=None, stack_size=None, stack_perms=None, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

This mixin adds automatic allocation for a stack region based on the stack_end and stack_size parameters.

copy(memo)

allocate_stack_pages(addr: int, size: int, **kwargs)

Pre-allocates pages for the stack without triggering any logic related to reading from them.

Parameters

addr – The highest address that should be mapped
size – The number of bytes to be allocated. byte 1 is the one at addr, byte 2 is the one before that, and so on.

Returns

A list of the new page objects

class angr.storage.memory_mixins.paged_memory.privileged_mixin.PrivilegedPagingMixin(page_size=4096, default_permissions=3, permissions_map=None, page_kwargs=None, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.paged_memory_mixin.PagedMemoryMixin

A mixin for paged memory models which will raise SimSegfaultExceptions if STRICT_PAGE_ACCESS is enabled and a segfault condition is detected.

Segfault conditions include: - getting a page for reading which is non-readable - getting a page for writing which is non-writable - creating a page

The latter condition means that this should be inserted under any mixins which provide other implementations of _initialize_page.

class angr.storage.memory_mixins.paged_memory.pages.PageBase(*args, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.pages.history_tracking_mixin.HistoryTrackingMixin, angr.storage.memory_mixins.paged_memory.pages.refcount_mixin.RefcountMixin, angr.storage.memory_mixins.paged_memory.pages.cooperation.CooperationBase, angr.storage.memory_mixins.paged_memory.pages.ispo_mixin.ISPOMixin, angr.storage.memory_mixins.paged_memory.pages.permissions_mixin.PermissionsMixin, angr.storage.memory_mixins.MemoryMixin

This is a fairly succinct definition of the contract between PagedMemoryMixin and its constituent pages:

Pages must implement the MemoryMixin model for loads, stores, copying, merging, etc
However, loading/storing may not necessarily use the same data domain as PagedMemoryMixin. In order to do more efficient loads/stores across pages, we use the CooperationBase interface which allows the page class to determine how to generate and unwrap the objects which are actually stored.
To support COW, we use the RefcountMixin and the ISPOMixin (which adds the contract element that memory=self be passed to every method call)
Pages have permissions associated with them, stored in the PermissionsMixin.

Read the docstrings for each of the constituent classes to understand the nuances of their functionalities

class angr.storage.memory_mixins.paged_memory.pages.refcount_mixin.RefcountMixin(**kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

This mixin adds a locked reference counter and methods to manipulate it, to facilitate copy-on-write optimizations.

copy(memo)

acquire_unique(): Call this function to return a version of this page which can be used for writing, which may or may not be the same object as before. If you use this you must immediately replace the shared reference you previously had with the new unique copy.

acquire_shared() → None: Call this function to indicate that this page has had a reference added to it and must be copied before it can be acquired uniquely again. Creating the object implicitly starts it with one shared reference.

release_shared() → None: Call this function to indicate that this page has had a shared reference to it released

class angr.storage.memory_mixins.paged_memory.pages.permissions_mixin.PermissionsMixin(permissions=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

This mixin adds a permissions field and properties for extracting the read/write/exec permissions. It does NOT add permissions checking.

copy(memo)

property perm_read

property perm_write

property perm_exec

class angr.storage.memory_mixins.paged_memory.pages.history_tracking_mixin.HistoryTrackingMixin(*args, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.pages.refcount_mixin.RefcountMixin, angr.storage.memory_mixins.MemoryMixin

Tracks the history of memory writes.

store(addr, data, size=None, **kwargs)

copy(memo)

acquire_unique()

parents()

changed_bytes(other, **kwargs) → Optional[Set[int]]

all_bytes_changed_in_history() → Set[int]

class angr.storage.memory_mixins.paged_memory.pages.ispo_mixin.ISPOMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

An implementation of the International Stateless Persons Organisation, a mixin which should be applied as a bottom layer for memories which have no state and must redirect certain operations to a parent memory. Main usecase is for memory region classes which are stored within other memories, such as pages.

set_state(state)

class angr.storage.memory_mixins.paged_memory.pages.cooperation.CooperationBase

Bases: object

Any given subclass of this class which is not a subclass of MemoryMixin should have the property that any subclass it which is a subclass of MemoryMixin should all work with the same datatypes

class angr.storage.memory_mixins.paged_memory.pages.cooperation.MemoryObjectMixin

Bases: angr.storage.memory_mixins.paged_memory.pages.cooperation.CooperationBase

Uses SimMemoryObjects in region storage. With this, load will return a list of tuple (address, MO) and store will take a MO.

class angr.storage.memory_mixins.paged_memory.pages.cooperation.MemoryObjectSetMixin

Bases: angr.storage.memory_mixins.paged_memory.pages.cooperation.CooperationBase

Uses sets of SimMemoryObjects in region storage.

class angr.storage.memory_mixins.paged_memory.pages.cooperation.BasicClaripyCooperation

Bases: angr.storage.memory_mixins.paged_memory.pages.cooperation.CooperationBase

Mix this (along with PageBase) into a storage class which supports loading and storing claripy bitvectors and it will be able to work as a page in the paged memory model.

class angr.storage.memory_mixins.paged_memory.pages.list_page.ListPage(memory=None, content=None, sinkhole=None, mo_cmp=None, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.pages.cooperation.MemoryObjectMixin, angr.storage.memory_mixins.paged_memory.pages.PageBase

This class implements a page memory mixin with lists as the main content store.

copy(memo)

load(addr, size=None, endness=None, page_addr=None, memory=None, cooperate=False, **kwargs)

store(addr, data, size=None, endness=None, memory=None, cooperate=False, **kwargs)

erase(addr, size=None, **kwargs) → None

merge(others: List[angr.storage.memory_mixins.paged_memory.pages.list_page.ListPage], merge_conditions, common_ancestor=None, page_addr: Optional[int] = None, memory=None, changed_offsets: Optional[Set[int]] = None)

changed_bytes(other: angr.storage.memory_mixins.paged_memory.pages.list_page.ListPage, page_addr: Optional[int] = None)

class angr.storage.memory_mixins.paged_memory.pages.ultra_page.UltraPage(memory=None, init_zero=False, **kwargs)

Bases: angr.storage.memory_mixins.paged_memory.pages.cooperation.MemoryObjectMixin, angr.storage.memory_mixins.paged_memory.pages.PageBase

Default page implementation

SUPPORTS_CONCRETE_LOAD = True

classmethod new_from_shared(data, memory=None, **kwargs)

copy(memo)

load(addr, size=None, page_addr=None, endness=None, memory=None, cooperate=False, **kwargs)

store(addr, data: Union[int, angr.storage.memory_object.SimMemoryObject], size: Optional[int] = None, endness=None, memory=None, page_addr=None, cooperate=False, **kwargs)

merge(others: List[angr.storage.memory_mixins.paged_memory.pages.ultra_page.UltraPage], merge_conditions, common_ancestor=None, page_addr: Optional[int] = None, memory=None, changed_offsets: Optional[Set[int]] = None)

concrete_load(addr, size, **kwargs)

changed_bytes(other, page_addr=None) → Set[int]

replace_all_with_offsets(offsets: Iterable[int], old: claripy.ast.bv.BV, new: claripy.ast.bv.BV, memory=None)

class angr.storage.memory_mixins.regioned_memory.regioned_memory_mixin.RegionedMemoryMixin(write_targets_limit: int = 2048, read_targets_limit: int = 4096, stack_region_map: Optional[angr.storage.memory_mixins.regioned_memory.region_data.RegionMap] = None, generic_region_map: Optional[angr.storage.memory_mixins.regioned_memory.region_data.RegionMap] = None, stack_size: int = 65536, cle_memory_backer: Optional = None, dict_memory_backer: Optional[Dict] = None, regioned_memory_cls: Optional[type] = None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

Regioned memory. This mixin manages multiple memory regions. Each address is represented as a tuple of (region ID, offset into the region), which is called a regioned address.

Converting absolute addresses into regioned addresses: We map an absolute address to a region by looking up which region this address belongs to in the region map. Currently this is only enabled for stack. Heap support has not landed yet.

When start analyzing a function, the user should call set_stack_address_mapping() to create a new region mapping. Likewise, when exiting from a function, the user should cancel the previous mapping by calling unset_stack_address_mapping().

copy(memo=None, **kwargs)

load(addr, size: Optional[Union[claripy.ast.bv.BV, int]] = None, endness=None, condition: Optional[claripy.ast.bool.Bool] = None, **kwargs)

store(addr, data, size: Optional[int] = None, endness=None, **kwargs)

merge(others: Iterable[angr.storage.memory_mixins.regioned_memory.regioned_memory_mixin.RegionedMemoryMixin], merge_conditions, common_ancestor=None) → bool

find(addr: Union[int, claripy.ast.bits.Bits], data, max_search, **kwargs)

set_state(state)

replace_all(old: claripy.ast.bv.BV, new: claripy.ast.bv.BV)

set_stack_address_mapping(absolute_address: int, region_id: str, related_function_address: Optional[int] = None)

Create a new mapping between an absolute address (which is the base address of a specific stack frame) and a region ID.

Parameters

absolute_address – The absolute memory address.
region_id – The region ID.
related_function_address – Related function address.

unset_stack_address_mapping(absolute_address: int)

Remove a stack mapping.

Parameters: absolute_address – An absolute memory address that is the base address of the stack frame to destroy.

stack_id(function_address: int) → str

Return a memory region ID for a function. If the default region ID exists in the region mapping, an integer will appended to the region name. In this way we can handle recursive function calls, or a function that appears more than once in the call frame.

This also means that stack_id() should only be called when creating a new stack frame for a function. You are not supposed to call this function every time you want to map a function address to a stack ID.

Parameters: function_address – Address of the function.
Returns: ID of the new memory region.

set_stack_size(size: int)

class angr.storage.memory_mixins.regioned_memory.region_data.AddressWrapper(region: str, region_base_addr: int, address, is_on_stack: bool, function_address: Optional[int])

Bases: object

AddressWrapper is used in SimAbstractMemory, which provides extra meta information for an address (or a ValueSet object) that is normalized from an integer/BVV/StridedInterval.

Constructor for the class AddressWrapper.

Parameters

region – Name of the memory regions it belongs to.
region_base_addr – Base address of the memory region
address – An address (not a ValueSet object).
is_on_stack – Whether this address is on a stack region or not.
function_address – Related function address (if any).

region

region_base_addr

address

is_on_stack

function_address

to_valueset(state)

Convert to a ValueSet instance

Parameters: state – A state
Returns: The converted ValueSet instance

class angr.storage.memory_mixins.regioned_memory.region_data.RegionDescriptor(region_id, base_address, related_function_address=None)

Bases: object

Descriptor for a memory region ID.

region_id

base_address

related_function_address

class angr.storage.memory_mixins.regioned_memory.region_data.RegionMap(is_stack)

Bases: object

Mostly used in SimAbstractMemory, RegionMap stores a series of mappings between concrete memory address ranges and memory regions, like stack frames and heap regions.

Constructor

Parameters: is_stack – Whether this is a region map for stack frames or not. Different strategies apply for stack regions.

property is_empty

property stack_base

property region_ids

copy(memo=None, **kwargs)

map(absolute_address, region_id, related_function_address=None)

Add a mapping between an absolute address and a region ID. If this is a stack region map, all stack regions beyond (lower than) this newly added regions will be discarded.

Parameters

absolute_address – An absolute memory address.
region_id – ID of the memory region.
related_function_address – A related function address, mostly used for stack regions.

unmap_by_address(absolute_address)

Removes a mapping based on its absolute address.

Parameters: absolute_address – An absolute address

absolutize(region_id, relative_address)

Convert a relative address in some memory region to an absolute address.

Parameters

region_id – The memory region ID
relative_address – The relative memory offset in that memory region

Returns

An absolute address if converted, or an exception is raised when region id does not exist.

relativize(absolute_address, target_region_id=None)

Convert an absolute address to the memory offset in a memory region.

Note that if an address belongs to heap region is passed in to a stack region map, it will be converted to an offset included in the closest stack frame, and vice versa for passing a stack address to a heap region. Therefore you should only pass in address that belongs to the same category (stack or non-stack) of this region map.

Parameters: absolute_address – An absolute memory address
Returns: A tuple of the closest region ID, the relative offset, and the related function address.

class angr.storage.memory_mixins.regioned_memory.region_category_mixin.RegionCategoryMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.MemoryMixin

property category

class angr.storage.memory_mixins.regioned_memory.static_find_mixin.StaticFindMixin(memory_id=None, endness='Iend_BE')

Bases: angr.storage.memory_mixins.smart_find_mixin.SmartFindMixin

Implements data finding for abstract memory.

find(addr, data, max_search, default=None, endness=None, chunk_size=None, max_symbolic_bytes=None, condition=None, char_size=1, **kwargs)

class angr.storage.memory_mixins.regioned_memory.abstract_address_descriptor.AbstractAddressDescriptor

Bases: object

AbstractAddressDescriptor describes a list of region+offset tuples. It provides a convenient way for accessing the cardinality (the total number of addresses) without enumerating or creating all addresses in static mode.

property cardinality

add_regioned_address(region: str, addr: claripy.vsa.strided_interval.StridedInterval)

clear()

class angr.storage.memory_mixins.regioned_memory.region_meta_mixin.MemoryRegionMetaMixin(related_function_addr=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

copy(memo=None, **kwargs)

property is_stack

property related_function_addr

get_abstract_locations(addr, size)

Get a list of abstract locations that is within the range of [addr, addr + size]

This implementation is pretty slow. But since this method won’t be called frequently, we can live with the bad implementation for now.

Parameters

addr – Starting address of the memory region.
size – Size of the memory region, in bytes.

Returns

A list of covered AbstractLocation objects, or an empty list if there is none.

store(addr, data, bbl_addr=None, stmt_id=None, ins_addr=None, endness=None, **kwargs)

load(addr, size=None, bbl_addr=None, stmt_idx=None, ins_addr=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None) → bool

widen(others)

dbg_print(indent=0): Print out debugging information

class angr.storage.memory_mixins.regioned_memory.abstract_merger_mixin.AbstractMergerMixin(memory_id=None, endness='Iend_BE'): Bases: angr.storage.memory_mixins.MemoryMixin

class angr.storage.memory_mixins.regioned_memory.regioned_address_concretization_mixin.RegionedAddressConcretizationMixin(read_strategies=None, write_strategies=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

set_state(state)

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None) → bool

class angr.storage.memory_mixins.slotted_memory.SlottedMemoryMixin(width=None, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

set_state(state)

copy(memo)

merge(others, merge_conditions, common_ancestor=None)

load(addr, size=None, endness=None, **kwargs)

store(addr, data, size=None, endness=None, **kwargs)

changed_bytes(other)

class angr.storage.memory_mixins.keyvalue_memory.keyvalue_memory_mixin.TypedVariable(type_, value)

Bases: object

type

value

class angr.storage.memory_mixins.keyvalue_memory.keyvalue_memory_mixin.KeyValueMemoryMixin(*args, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

load(key, none_if_missing=False, **kwargs)

store(key, data, type_=None, **kwargs)

copy(memo=None, **kwargs)

class angr.storage.memory_mixins.javavm_memory.javavm_memory_mixin.JavaVmMemoryMixin(memory_id='mem', stack=None, heap=None, vm_static_table=None, load_strategies=None, store_strategies=None, max_array_size=1000, **kwargs)

Bases: angr.storage.memory_mixins.MemoryMixin

static get_new_uuid(): Generate a unique id within the scope of the JavaVM memory. This, for example, is used for distinguishing memory objects of the same type (e.g. multiple instances of the same class).

store(addr, data, frame=0)

load(addr, frame=0, none_if_missing=False)

push_stack_frame()

pop_stack_frame()

property stack

store_array_element(array, idx, value)

store_array_elements(array, start_idx, data)

Stores either a single element or a range of elements in the array.

Parameters

array – Reference to the array.
start_idx – Starting index for the store.
data – Either a single value or a list of values.

load_array_element(array, idx)

load_array_elements(array, start_idx, no_of_elements)

Loads either a single element or a range of elements from the array.

Parameters

array – Reference to the array.
start_idx – Starting index for the load.
no_of_elements – Number of elements to load.

concretize_store_idx(idx, strategies=None)

Concretizes a store index.

Parameters

idx – An expression for the index.
strategies – A list of concretization strategies (to override the default).
min_idx – Minimum value for a concretized index (inclusive).
max_idx – Maximum value for a concretized index (exclusive).

Returns

A list of concrete indexes.

concretize_load_idx(idx, strategies=None)

Concretizes a load index.

Parameters

idx – An expression for the index.
strategies – A list of concretization strategies (to override the default).
min_idx – Minimum value for a concretized index (inclusive).
max_idx – Maximum value for a concretized index (exclusive).

Returns

A list of concrete indexes.

set_state(state)

copy(memo=None, **kwargs)

merge(others, merge_conditions, common_ancestor=None)

widen(others)

Concretization Strategies

class angr.concretization_strategies.single.SimConcretizationStrategySingle(filter=None, exact=True)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that ensures a single solution for an address.

Initializes the base SimConcretizationStrategy.

Parameters

filter – A function, taking arguments of (SimMemory, claripy.AST) that determins if this strategy can handle resolving the provided AST.
exact – A flag (default: True) that determines if the convenience resolution functions provided by this class use exact or approximate resolution.

class angr.concretization_strategies.eval.SimConcretizationStrategyEval(limit, **kwargs)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that resolves an address into some limited number of solutions. Always handles the concretization, but only returns a maximum of limit number of solutions. Therefore, should only be used as the fallback strategy.

class angr.concretization_strategies.norepeats.SimConcretizationStrategyNorepeats(repeat_expr, repeat_constraints=None, **kwargs)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that resolves addresses, without repeating.

copy()

merge(others)

class angr.concretization_strategies.solutions.SimConcretizationStrategySolutions(limit, **kwargs)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that resolves an address into some limited number of solutions.

class angr.concretization_strategies.nonzero_range.SimConcretizationStrategyNonzeroRange(limit, **kwargs)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that resolves a range in a non-zero location.

class angr.concretization_strategies.range.SimConcretizationStrategyRange(limit, **kwargs)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that resolves addresses to a range.

class angr.concretization_strategies.max.SimConcretizationStrategyMax(max_addr: Optional[int] = None)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that returns the maximum address.

class angr.concretization_strategies.norepeats_range.SimConcretizationStrategyNorepeatsRange(repeat_expr, min=None, granularity=None, **kwargs)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that resolves a range, with no repeats.

copy()

merge(others)

class angr.concretization_strategies.nonzero.SimConcretizationStrategyNonzero(filter=None, exact=True)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that returns any non-zero solution.

Initializes the base SimConcretizationStrategy.

Parameters

filter – A function, taking arguments of (SimMemory, claripy.AST) that determins if this strategy can handle resolving the provided AST.
exact – A flag (default: True) that determines if the convenience resolution functions provided by this class use exact or approximate resolution.

class angr.concretization_strategies.any.SimConcretizationStrategyAny(filter=None, exact=True)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that returns any single solution.

Initializes the base SimConcretizationStrategy.

Parameters

filter – A function, taking arguments of (SimMemory, claripy.AST) that determins if this strategy can handle resolving the provided AST.
exact – A flag (default: True) that determines if the convenience resolution functions provided by this class use exact or approximate resolution.

class angr.concretization_strategies.controlled_data.SimConcretizationStrategyControlledData(limit, fixed_addrs, **kwargs)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that constraints the address to controlled data. Controlled data consists of symbolic data and the addresses given as arguments. memory.

class angr.concretization_strategies.unlimited_range.SimConcretizationStrategyUnlimitedRange(limit, **kwargs)

Bases: angr.concretization_strategies.SimConcretizationStrategy

Concretization strategy that resolves addresses to a range without checking if the number of possible addresses is within the limit.

Simulation Manager

class angr.sim_manager.SimulationManager(project, active_states=None, stashes=None, hierarchy=None, resilience=None, save_unsat=False, auto_drop=None, errored=None, completion_mode=<built-in function any>, techniques=None, **kwargs)

Bases: object

The Simulation Manager is the future future.

Simulation managers allow you to wrangle multiple states in a slick way. States are organized into “stashes”, which you can step forward, filter, merge, and move around as you wish. This allows you to, for example, step two different stashes of states at different rates, then merge them together.

Stashes can be accessed as attributes (i.e. .active). A mulpyplexed stash can be retrieved by prepending the name with mp_, e.g. .mp_active. A single state from the stash can be retrieved by prepending the name with one_, e.g. .one_active.

Note that you shouldn’t usually be constructing SimulationManagers directly - there is a convenient shortcut for creating them in Project.factory: see angr.factory.AngrObjectFactory.

The most important methods you should look at are step, explore, and use_technique.

Parameters

project (angr.project.Project) – A Project instance.
stashes – A dictionary to use as the stash store.
active_states – Active states to seed the “active” stash with.
hierarchy – A StateHierarchy object to use to track the relationships between states.
resilience – A set of errors to catch during stepping to put a state in the errore list. You may also provide the values False, None (default), or True to catch, respectively, no errors, all angr-specific errors, and a set of many common errors.
save_unsat – Set to True in order to introduce unsatisfiable states into the unsat stash instead of discarding them immediately.
auto_drop – A set of stash names which should be treated as garbage chutes.
completion_mode – A function describing how multiple exploration techniques with the complete hook set will interact. By default, the builtin function any.
techniques – A list of techniques that should be pre-set to use with this manager.

Variables

errored – Not a stash, but a list of ErrorRecords. Whenever a step raises an exception that we catch, the state and some information about the error are placed in this list. You can adjust the list of caught exceptions with the resilience parameter.
stashes – All the stashes on this instance, as a dictionary.
completion_mode – A function describing how multiple exploration techniques with the complete hook set will interact. By default, the builtin function any.

ALL = '_ALL'

DROP = '_DROP'

property errored

property stashes: DefaultDict[str, List[angr.sim_state.SimState]]

mulpyplex(*stashes)

Mulpyplex across several stashes.

Parameters: stashes – the stashes to mulpyplex
Returns: a mulpyplexed list of states from the stashes in question, in the specified order

copy(deep=False): Make a copy of this simulation manager. Pass deep=True to copy all the states in it as well.

use_technique(tech)

Use an exploration technique with this SimulationManager.

Techniques can be found in angr.exploration_techniques.

Parameters: tech (ExplorationTechnique) – An ExplorationTechnique object that contains code to modify this SimulationManager’s behavior.
Returns: The technique that was added, for convenience

remove_technique(tech)

Remove an exploration technique from a list of active techniques.

Parameters: tech (ExplorationTechnique) – An ExplorationTechnique object.

explore(stash='active', n=None, find=None, avoid=None, find_stash='found', avoid_stash='avoid', cfg=None, num_find=1, **kwargs)

Tick stash “stash” forward (up to “n” times or until “num_find” states are found), looking for condition “find”, avoiding condition “avoid”. Stores found states into “find_stash’ and avoided states into “avoid_stash”.

The “find” and “avoid” parameters may be any of:

An address to find
A set or list of addresses to find
A function that takes a state and returns whether or not it matches.

If an angr CFG is passed in as the “cfg” parameter and “find” is either a number or a list or a set, then any states which cannot possibly reach a success state without going through a failure state will be preemptively avoided.

run(stash='active', n=None, until=None, **kwargs)

Run until the SimulationManager has reached a completed state, according to the current exploration techniques. If no exploration techniques that define a completion state are being used, run until there is nothing left to run.

Parameters

stash – Operate on this stash
n – Step at most this many times
until – If provided, should be a function that takes a SimulationManager and returns True or False. Stepping will terminate when it is True.

Returns

The simulation manager, for chaining.

Return type

SimulationManager

complete(): Returns whether or not this manager has reached a “completed” state.

step(stash='active', n=None, selector_func=None, step_func=None, successor_func=None, until=None, filter_func=None, **run_args)

Step a stash of states forward and categorize the successors appropriately.

The parameters to this function allow you to control everything about the stepping and categorization process.

Parameters

stash – The name of the stash to step (default: ‘active’)
selector_func – If provided, should be a function that takes a state and returns a boolean. If True, the state will be stepped. Otherwise, it will be kept as-is.
step_func – If provided, should be a function that takes a SimulationManager and returns a SimulationManager. Will be called with the SimulationManager at every step. Note that this function should not actually perform any stepping - it is meant to be a maintenance function called after each step.
successor_func – If provided, should be a function that takes a state and return its successors. Otherwise, project.factory.successors will be used.
filter_func – If provided, should be a function that takes a state and return the name of the stash, to which the state should be moved.
until – (DEPRECATED) If provided, should be a function that takes a SimulationManager and returns True or False. Stepping will terminate when it is True.
n – (DEPRECATED) The number of times to step (default: 1 if “until” is not provided)

Additionally, you can pass in any of the following keyword args for project.factory.successors:

Parameters

jumpkind – The jumpkind of the previous exit
addr – An address to execute at instead of the state’s ip.
stmt_whitelist – A list of stmt indexes to which to confine execution.
last_stmt – A statement index at which to stop execution.
thumb – Whether the block should be lifted in ARM’s THUMB mode.
backup_state – A state to read bytes from instead of using project memory.
opt_level – The VEX optimization level to use.
insn_bytes – A string of bytes to use for the block instead of the project.
size – The maximum size of the block, in bytes.
num_inst – The maximum number of instructions.
traceflags – traceflags to be passed to VEX. Default: 0

Returns

The simulation manager, for chaining.

Return type

SimulationManager

step_state(state, successor_func=None, **run_args): Don’t use this function manually - it is meant to interface with exploration techniques.

filter(state, filter_func=None): Don’t use this function manually - it is meant to interface with exploration techniques.

selector(state, selector_func=None): Don’t use this function manually - it is meant to interface with exploration techniques.

successors(state, successor_func=None, **run_args): Don’t use this function manually - it is meant to interface with exploration techniques.

prune(filter_func=None, from_stash='active', to_stash='pruned')

Prune unsatisfiable states from a stash.

This function will move all unsatisfiable states in the given stash into a different stash.

Parameters

filter_func – Only prune states that match this filter.
from_stash – Prune states from this stash. (default: ‘active’)
to_stash – Put pruned states in this stash. (default: ‘pruned’)

Returns

The simulation manager, for chaining.

Return type

SimulationManager

populate(stash, states)

Populate a stash with a collection of states.

Parameters

stash – A stash to populate.
states – A list of states with which to populate the stash.

move(from_stash, to_stash, filter_func=None)

Move states from one stash to another.

Parameters

from_stash – Take matching states from this stash.
to_stash – Put matching states into this stash.
filter_func – Stash states that match this filter. Should be a function that takes a state and returns True or False. (default: stash all states)

Returns

The simulation manager, for chaining.

Return type

SimulationManager

stash(filter_func=None, from_stash='active', to_stash='stashed')

Stash some states. This is an alias for move(), with defaults for the stashes.

Parameters

filter_func – Stash states that match this filter. Should be a function that takes a state and returns True or False. (default: stash all states)
from_stash – Take matching states from this stash. (default: ‘active’)
to_stash – Put matching states into this stash. (default: ‘stashed’)

Returns

The simulation manager, for chaining.

Return type

SimulationManager

unstash(filter_func=None, to_stash='active', from_stash='stashed')

Unstash some states. This is an alias for move(), with defaults for the stashes.

Parameters

filter_func – Unstash states that match this filter. Should be a function that takes a state and returns True or False. (default: unstash all states)
from_stash – take matching states from this stash. (default: ‘stashed’)
to_stash – put matching states into this stash. (default: ‘active’)

Returns

The simulation manager, for chaining.

Return type

SimulationManager

drop(filter_func=None, stash='active')

Drops states from a stash. This is an alias for move(), with defaults for the stashes.

Parameters

filter_func – Drop states that match this filter. Should be a function that takes a state and returns True or False. (default: drop all states)
stash – Drop matching states from this stash. (default: ‘active’)

Returns

The simulation manager, for chaining.

Return type

SimulationManager

apply(state_func=None, stash_func=None, stash='active', to_stash=None)

Applies a given function to a given stash.

Parameters

state_func – A function to apply to every state. Should take a state and return a state. The returned state will take the place of the old state. If the function doesn’t return a state, the old state will be used. If the function returns a list of states, they will replace the original states.
stash_func – A function to apply to the whole stash. Should take a list of states and return a list of states. The resulting list will replace the stash. If both state_func and stash_func are provided state_func is applied first, then stash_func is applied on the results.
stash – A stash to work with.
to_stash – If specified, this stash will be used to store the resulting states instead.

Returns

The simulation manager, for chaining.

Return type

SimulationManager

split(stash_splitter=None, stash_ranker=None, state_ranker=None, limit=8, from_stash='active', to_stash='stashed')

Split a stash of states into two stashes depending on the specified options.

The stash from_stash will be split into two stashes depending on the other options passed in. If to_stash is provided, the second stash will be written there.

stash_splitter overrides stash_ranker, which in turn overrides state_ranker. If no functions are provided, the states are simply split according to the limit.

The sort done with state_ranker is ascending.

Parameters

stash_splitter – A function that should take a list of states and return a tuple of two lists (the two resulting stashes).
stash_ranker – A function that should take a list of states and return a sorted list of states. This list will then be split according to “limit”.
state_ranker – An alternative to stash_splitter. States will be sorted with outputs of this function, which are to be used as a key. The first “limit” of them will be kept, the rest split off.
limit – For use with state_ranker. The number of states to keep. Default: 8
from_stash – The stash to split (default: ‘active’)
to_stash – The stash to write to (default: ‘stashed’)

Returns

The simulation manager, for chaining.

Return type

SimulationManager

merge(merge_func=None, merge_key=None, stash='active')

Merge the states in a given stash.

Parameters

stash – The stash (default: ‘active’)
merge_func – If provided, instead of using state.merge, call this function with the states as the argument. Should return the merged state.
merge_key – If provided, should be a function that takes a state and returns a key that will compare equal for all states that are allowed to be merged together, as a first aproximation. By default: uses PC, callstack, and open file descriptors.

Returns

The simulation manager, for chaining.

Return type

SimulationManager

class angr.sim_manager.ErrorRecord(state, error, traceback)

Bases: object

A container class for a state and an error that was thrown during its execution. You can find these in SimulationManager.errored.

Variables

state – The state that encountered an error, at the point in time just before the erroring step began.
error – The error that was thrown.
traceback – The traceback for the error that was thrown.

debug(): Launch a postmortem debug shell at the site of the error.

reraise()

class angr.state_hierarchy.StateHierarchy

Bases: object

get_ref(obj)

clear_ref(ref)

add_state(s)

add_history(h)

simplify()

full_simplify()

lineage(h): Returns the lineage of histories leading up to h.

all_successors(h)

history_successors(h)

history_predecessors(h)

history_contains(h)

unreachable_state(state)

unreachable_history(h)

most_mergeable(states)

Find the “most mergeable” set of states from those provided.

Parameters: states – a list of states
Returns: a tuple of: (a list of states to merge, those states’ common history, a list of states to not merge yet)

Exploration Techniques

class angr.exploration_techniques.ExplorationTechniqueMeta(name, bases, attrs): Bases: type

class angr.exploration_techniques.ExplorationTechnique

Bases: object

An otiegnqwvk is a set of hooks for a simulation manager that assists in the implementation of new techniques in symbolic exploration.

TODO: choose actual name for the functionality (techniques? strategies?)

Any number of these methods may be overridden by a subclass. To use an exploration technique, call simgr.use_technique with an instance of the technique.

Warning

There is currently installed a compatibility layer for the previous version of this API. This layer requires that implementations of ExplorationTechnique use the same argument names as the original methods, or else it will mangle the behvior.

setup(simgr)

Perform any initialization on this manager you might need to do.

Parameters: simgr (angr.SimulationManager) – The simulation manager to which you have just been added

step(simgr, stash='active', **kwargs)

Hook the process of stepping a stash forward. Should call simgr.step(stash, **kwargs) in order to do the actual processing.

Parameters

simgr (angr.SimulationManager) –
stash (str) –

filter(simgr, state, **kwargs)

Perform filtering on which stash a state should be inserted into.

If the state should be filtered, return the name of the stash to move the state to. If you want to modify the state before filtering it, return a tuple of the stash to move the state to and the modified state. To defer to the original categorization procedure, return the result of simgr.filter(state, **kwargs)

If the user provided a filter_func in their step or run command, it will appear here.

Parameters

simgr (angr.SimulationManager) –
state (angr.SimState) –

selector(simgr, state, **kwargs)

Determine if a state should participate in the current round of stepping. Return True if the state should be stepped, and False if the state should not be stepped. To defer to the original selection procedure, return the result of simgr.selector(state, **kwargs).

If the user provided a selector_func in their step or run command, it will appear here.

Parameters

simgr (angr.SimulationManager) –
state (angr.SimState) –

step_state(simgr, state, **kwargs)

Determine the categorization of state successors into stashes. The result should be a dict mapping stash names to the list of successor states that fall into that stash, or None as a stash name to use the original stash name.

If you would like to directly work with a SimSuccessors object, you can obtain it with simgr.successors(state, **kwargs). This is not recommended, as it denies other hooks the opportunity to look at the successors. Therefore, the usual technique is to call simgr.step_state(state, **kwargs) and then mutate the returned dict before returning it yourself.

..note:: This takes precedence over the filter hook - filter is only applied to states returned from here in the None stash.

Parameters

simgr (angr.SimulationManager) –
state (angr.SimState) –

successors(simgr, state, **kwargs)

Perform the process of stepping a state forward, returning a SimSuccessors object.

To defer to the original succession procedure, return the result of simgr.successors(state, **kwargs). Be careful about not calling this method (e.g. calling project.factory.successors manually) as it denies other hooks the opportunity to instrument the step. Instead, you can mutate the kwargs for the step before calling the original, and mutate the result before returning it yourself.

If the user provided a successor_func in their step or run command, it will appear here.

Parameters

simgr (angr.SimulationManager) –
state (angr.SimState) –

complete(simgr)

Return whether or not this manager has reached a “completed” state, i.e. SimulationManager.run() should halt.

This is the one hook which is not subject to the nesting rules of hooks. You should not call simgr.complete, you should make your own decision and return True or False. Each of the techniques’ completion checkers will be called and the final result will be compted with simgr.completion_mode.

Parameters: simgr (angr.SimulationManager) –

class angr.exploration_techniques.dfs.DFS(deferred_stash='deferred')

Bases: angr.exploration_techniques.ExplorationTechnique

Depth-first search.

Will only keep one path active at a time, any others will be stashed in the ‘deferred’ stash. When we run out of active paths to step, we take the longest one from deferred and continue.

setup(simgr)

step(simgr, stash='active', **kwargs)

class angr.exploration_techniques.explorer.Explorer(find=None, avoid=None, find_stash='found', avoid_stash='avoid', cfg=None, num_find=1, avoid_priority=False)

Bases: angr.exploration_techniques.ExplorationTechnique

Search for up to “num_find” paths that satisfy condition “find”, avoiding condition “avoid”. Stashes found paths into “find_stash’ and avoided paths into “avoid_stash”.

The “find” and “avoid” parameters may be any of:

An address to find
A set or list of addresses to find
A function that takes a path and returns whether or not it matches.

If an angr CFG is passed in as the “cfg” parameter and “find” is either a number or a list or a set, then any paths which cannot possibly reach a success state without going through a failure state will be preemptively avoided.

If either the “find” or “avoid” parameter is a function returning a boolean, and a path triggers both conditions, it will be added to the find stash, unless “avoid_priority” is set to True.

setup(simgr)

step(simgr, stash='active', **kwargs)

filter(simgr, state, **kwargs)

complete(simgr)

class angr.exploration_techniques.lengthlimiter.LengthLimiter(max_length, drop=False)

Bases: angr.exploration_techniques.ExplorationTechnique

Length limiter on paths.

step(simgr, stash='active', **kwargs)

class angr.exploration_techniques.manual_mergepoint.ManualMergepoint(address, wait_counter=10)

Bases: angr.exploration_techniques.ExplorationTechnique

setup(simgr)

filter(simgr, state, **kwargs)

mark_nofilter(simgr, stash)

mark_okfilter(simgr, stash)

step(simgr, stash='active', **kwargs)

class angr.exploration_techniques.spiller.PickledStatesBase

Bases: object

The base class of pickled states

sort(): Sort pickled states.

add(prio, sid)

Add a newly pickled state.

Parameters

prio (int) – Priority of the state.
sid (str) – Persistent ID of the state.

Returns

None

pop_n(n)

Pop the top N states.

Parameters: n (int) – Number of states to take.
Returns: A list of states.

class angr.exploration_techniques.spiller.PickledStatesList

Bases: angr.exploration_techniques.spiller.PickledStatesBase

List-backed pickled state storage.

sort()

add(prio, sid)

pop_n(n)

class angr.exploration_techniques.spiller.PickledStatesDb(db_str='sqlite:///:memory:')

Bases: angr.exploration_techniques.spiller.PickledStatesBase

Database-backed pickled state storage.

sort()

add(prio, sid, taken=False, stash='spilled')

pop_n(n, stash='spilled')

get_recent_n(n, stash='spilled')

count()

class angr.exploration_techniques.spiller.Spiller(src_stash='active', min=5, max=10, staging_stash='spill_stage', staging_min=10, staging_max=20, pickle_callback=None, unpickle_callback=None, post_pickle_callback=None, priority_key=None, vault=None, states_collection=None)

Bases: angr.exploration_techniques.ExplorationTechnique

Automatically spill states out. It can spill out states to a different stash, spill them out to ANA, or first do the former and then (after enough states) the latter.

Initializes the spiller.

Parameters

max – the number of states that are not spilled
src_stash – the stash from which to spill states (default: active)
staging_stash – the stash to which to spill states (default: “spill_stage”)
staging_max – the number of states that can be in the staging stash before things get spilled to ANA (default: None. If staging_stash is set, then this means unlimited, and ANA will not be used).
priority_key – a function that takes a state and returns its numerical priority (MAX_INT is lowest priority). By default, self.state_priority will be used, which prioritizes by object ID.
vault – an angr.Vault object to handle storing and loading of states. If not provided, an angr.vaults.VaultShelf will be created with a temporary file.

step(simgr, stash='active', **kwargs)

static state_priority(state)

class angr.exploration_techniques.spiller_db.PickledState(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

priority

taken

stash

timestamp

class angr.exploration_techniques.threading.Threading(threads=8)

Bases: angr.exploration_techniques.ExplorationTechnique

Enable multithreading.

This is only useful in paths where a lot of time is taken inside z3, doing constraint solving. This is because of python’s GIL, which says that only one thread at a time may be executing python code.

step(simgr, stash='active', **kwargs)

class angr.exploration_techniques.veritesting.Veritesting(**options)

Bases: angr.exploration_techniques.ExplorationTechnique

Enable veritesting. This technique, described in a paper[1] from CMU, attempts to address the problem of state explosions in loops by performing smart merging.

[1] https://users.ece.cmu.edu/~aavgerin/papers/veritesting-icse-2014.pdf

step_state(simgr, state, successor_func=None, **kwargs)

class angr.exploration_techniques.tracer.TracingMode

Bases: object

Variables

Strict – Strict mode, the default mode, where an exception is raised immediately if tracer’s path deviates from the provided trace.
Permissive – Permissive mode, where tracer attempts to force the path back to the provided trace when a deviation happens. This does not always work, especially when the cause of deviation is related to input that will later be used in exploit generation. But, it might work magically sometimes.
CatchDesync – CatchDesync mode, catch desync because of sim_procedures. It might be a sign of something interesting.

Strict = 'strict'

Permissive = 'permissive'

CatchDesync = 'catch_desync'

exception angr.exploration_techniques.tracer.TracerDesyncError(msg, deviating_addr=None, deviating_trace_idx=None): Bases: angr.errors.AngrTracerError

class angr.exploration_techniques.tracer.RepHook(mnemonic)

Bases: object

run(state)

class angr.exploration_techniques.tracer.Tracer(trace=None, resiliency=False, keep_predecessors=1, crash_addr=None, copy_states=False, fast_forward_to_entry=True, mode='strict', aslr=True, follow_unsat=False)

Bases: angr.exploration_techniques.ExplorationTechnique

An exploration technique that follows an angr path with a concrete input. The tracing result is the state at the last address of the trace, which can be found in the ‘traced’ stash.

If the given concrete input makes the program crash, you should provide crash_addr, and the crashing state will be found in the ‘crashed’ stash.

Parameters

trace – The basic block trace.
resiliency – Should we continue to step forward even if qemu and angr disagree?
keep_predecessors – Number of states before the final state we should log.
crash_addr – If the trace resulted in a crash, provide the crashing instruction pointer here, and the ‘crashed’ stash will be populated with the crashing state.
copy_states – Whether COPY_STATES should be enabled for the tracing state. It is off by default because most tracing workloads benefit greatly from not performing copying. You want to enable it if you want to see the missed states. It will be re-added for the last 2% of the trace in order to set the predecessors list correctly. If you turn this on you may want to enable the LAZY_SOLVES option.
mode – Tracing mode.
aslr – Whether there are aslr slides. if not, tracer uses trace address as state address.
follow_unsat – Whether unsatisfiable states should be treated as potential successors or not.

Variables

predecessors – A list of states in the history before the final state.

setup(simgr)

complete(simgr)

filter(simgr, state, **kwargs)

step(simgr, stash='active', **kwargs)

step_state(simgr, state, **kwargs)

classmethod crash_windup(state, crash_addr)

class angr.exploration_techniques.driller_core.DrillerCore(trace, fuzz_bitmap=None)

Bases: angr.exploration_techniques.ExplorationTechnique

An exploration technique that symbolically follows an input looking for new state transitions.

It has to be used with Tracer exploration technique. Results are put in ‘diverted’ stash.

:param trace : The basic block trace. :param fuzz_bitmap: AFL’s bitmap of state transitions. Defaults to saying every transition is worth satisfying.

setup(simgr)

step(simgr, stash='active', **kwargs)

class angr.exploration_techniques.slicecutor.Slicecutor(annotated_cfg, force_taking_exit=False, force_sat: bool = False)

Bases: angr.exploration_techniques.ExplorationTechnique

The Slicecutor is an exploration that executes provided code slices.

All parameters except annotated_cfg are optional.

Parameters

annotated_cfg – The AnnotatedCFG that provides the code slice.
force_taking_exit – Set to True if you want to create a successor based on our slice in case of unconstrained successors.
force_sat – If a branch specified by the slice is unsatisfiable, set this option to True if you want to force it to be satisfiable and be taken anyway.

setup(simgr)

filter(simgr, state, **kwargs)

step_state(simgr, state, **kwargs)

successors(simgr, state, **kwargs)

class angr.exploration_techniques.director.BaseGoal(sort)

Bases: object

REQUIRE_CFG_STATES = False

check(cfg, state, peek_blocks)

Parameters

cfg (angr.analyses.CFGEmulated) – An instance of CFGEmulated.
state (angr.SimState) – The state to check.
peek_blocks (int) – Number of blocks to peek ahead from the current point.

Returns

True if we can determine that this condition is definitely satisfiable if the path is taken, False otherwise.

Return type

bool

check_state(state)

Check if the current state satisfies the goal.

Parameters: state (angr.SimState) – The state to check.
Returns: True if it satisfies the goal, False otherwise.
Return type: bool

class angr.exploration_techniques.director.ExecuteAddressGoal(addr)

Bases: angr.exploration_techniques.director.BaseGoal

A goal that prioritizes states reaching (or are likely to reach) certain address in some specific steps.

check(cfg, state, peek_blocks)

Check if the specified address will be executed

Parameters

cfg –
state –
peek_blocks (int) –

Returns

Return type

bool

check_state(state)

Check if the current address is the target address.

Parameters: state (angr.SimState) – The state to check.
Returns: True if the current address is the target address, False otherwise.
Return type: bool

class angr.exploration_techniques.director.CallFunctionGoal(function, arguments)

Bases: angr.exploration_techniques.director.BaseGoal

A goal that prioritizes states reaching certain function, and optionally with specific arguments. Note that constraints on arguments (and on function address as well) have to be identifiable on an accurate CFG. For example, you may have a CallFunctionGoal saying “call printf with the first argument being ‘Hello, world’”, and CFGEmulated must be able to figure our the first argument to printf is in fact “Hello, world”, not some symbolic strings that will be constrained to “Hello, world” during symbolic execution (or simulation, however you put it).

REQUIRE_CFG_STATES = True

check(cfg, state, peek_blocks)

Check if the specified function will be reached with certain arguments.

Parameters

cfg –
state –
peek_blocks –

Returns

check_state(state)

Check if the specific function is reached with certain arguments

Parameters: state (angr.SimState) – The state to check
Returns: True if the function is reached with certain arguments, False otherwise.
Return type: bool

class angr.exploration_techniques.director.Director(peek_blocks=100, peek_functions=5, goals=None, cfg_keep_states=False, goal_satisfied_callback=None, num_fallback_states=5)

Bases: angr.exploration_techniques.ExplorationTechnique

An exploration technique for directed symbolic execution.

A control flow graph (using CFGEmulated) is built and refined during symbolic execution. Each time the execution reaches a block that is outside of the CFG, the CFG recovery will be triggered with that state, with a maximum recovery depth (100 by default). If we see a basic block during state stepping that is not yet in the control flow graph, we go back to control flow graph recovery and “peek” more blocks forward.

When stepping a simulation manager, all states are categorized into three different categories:

Might reach the destination within the peek depth. Those states are prioritized.
Will not reach the destination within the peek depth. Those states are de-prioritized. However, there is a little chance for those states to be explored as well in order to prevent over-fitting.

Constructor.

step(simgr, stash='active', **kwargs)

Parameters

simgr –
stash –
kwargs –

Returns

add_goal(goal)

Add a goal.

Parameters: goal (BaseGoal) – The goal to add.
Returns: None

class angr.exploration_techniques.oppologist.Oppologist

Bases: angr.exploration_techniques.ExplorationTechnique

The Oppologist is an exploration technique that forces uncooperative code through qemu.

successors(simgr, state, **kwargs)

class angr.exploration_techniques.loop_seer.LoopSeer(cfg=None, functions=None, loops=None, use_header=False, bound=None, bound_reached=None, discard_stash='spinning', limit_concrete_loops=True)

Bases: angr.exploration_techniques.ExplorationTechnique

This exploration technique monitors exploration and maintains all loop-related data (well, currently it is just the loop trip counts, but feel free to add something else).

Parameters

cfg – Normalized CFG is required.
functions – Function(s) containing the loop(s) to be analyzed.
loops – Specific group of Loop(s) to be analyzed, if this is None we run the LoopFinder analysis.
use_header – Whether to use header based trip counter to compare with the bound limit.
bound – Limit the number of iterations a loop may be executed.
bound_reached – If provided, should be a function that takes the LoopSeer and the succ_state. Will be called when loop execution reach the given bound. Default to moving states that exceed the loop limit to a discard stash.
discard_stash – Name of the stash containing states exceeding the loop limit.
limit_concrete_loops – If False, do not limit a loop back-edge if it is the only successor (Defaults to True to maintain the original behavior)

setup(simgr)

filter(simgr, state, **kwargs)

successors(simgr, state, **kwargs)

class angr.exploration_techniques.cacher.Cacher(when=None, dump_cache=True, load_cache=True, container=None, lookup=None, dump_func=None, load_func=None)

Bases: angr.exploration_techniques.ExplorationTechnique

An exploration technique that caches states during symbolic execution.

DO NOT USE THIS - THIS IS FOR ARCHIVAL PURPOSES ONLY

Parameters

dump_cache – Whether to dump data to cache.
load_cache – Whether to load data from cache.
container – Data container.
when – If provided, should be a function that takes a SimulationManager and returns a Boolean, or the address of the state to be cached.
lookup – A function that returns True if cache hit and False otherwise.
dump_func – If provided, should be a function that defines how Cacher should cache the SimulationManager. Default to caching the active stash.
load_func – If provided, should be a function that defines how Cacher should uncache the SimulationManager. Default to uncaching the stash to be stepped.

setup(simgr)

step(simgr, stash='active', **kwargs)

class angr.exploration_techniques.stochastic.StochasticSearch(start_state, restart_prob=0.0001)

Bases: angr.exploration_techniques.ExplorationTechnique

Stochastic Search.

Will only keep one path active at a time, any others will be discarded. Before each pass through, weights are randomly assigned to each basic block. These weights form a probability distribution for determining which state remains after splits. When we run out of active paths to step, we start again from the start state.

Parameters

start_state – The initial state from which exploration stems.
restart_prob – The probability of randomly restarting the search (default 0.0001).

step(simgr, stash='active', **kwargs)

class angr.exploration_techniques.unique.UniqueSearch(similarity_func=None, deferred_stash='deferred')

Bases: angr.exploration_techniques.ExplorationTechnique

Unique Search.

Will only keep one path active at a time, any others will be deferred. The state that is explored depends on how unique it is relative to the other deferred states. A path’s uniqueness is determined by its average similarity between the other (deferred) paths. Similarity is calculated based on the supplied similarity_func, which by default is: The (L2) distance between the counts of the state addresses in the history of the path.

Parameters

similarity_func – How to calculate similarity between two states.
deferred_stash – Where to store the deferred states.

setup(simgr)

step(simgr, stash='active', **kwargs)

static similarity(state_a, state_b): The (L2) distance between the counts of the state addresses in the history of the path. :param state_a: The first state to compare :param state_b: The second state to compare

static sequence_matcher_similarity(state_a, state_b): The difflib.SequenceMatcher ratio between the state addresses in the history of the path. :param state_a: The first state to compare :param state_b: The second state to compare

class angr.exploration_techniques.tech_builder.TechniqueBuilder(setup=None, step_state=None, step=None, successors=None, filter=None, selector=None, complete=None)

Bases: angr.exploration_techniques.ExplorationTechnique

This meta technique could be used to hook a couple of simulation manager methods without actually creating a new exploration technique, for example:

class SomeComplexAnalysis(Analysis):

def do_something():
simgr = self.project.factory.simulation_manager() simgr.use_tech(ProxyTechnique(step_state=self._step_state)) simgr.run()

def _step_state(self, state):
# Do stuff! pass

In the above example, the _step_state method can access all the necessary stuff, hidden in the analysis instance, without passing that instance to a one-shot-styled exploration technique.

angr.exploration_techniques.common.condition_to_lambda(condition, default=False)

Translates an integer, set, list or function into a lambda that checks if state’s current basic block matches some condition.

Parameters

condition – An integer, set, list or lambda to convert to a lambda.
default – The default return value of the lambda (in case condition is None). Default: false.

Returns

A tuple of two items: a lambda that takes a state and returns the set of addresses that it matched from the condition, and a set that contains the normalized set of addresses to stop at, or None if no addresses were provided statically.

class angr.exploration_techniques.symbion.Symbion(find=None, memory_concretize=None, register_concretize=None, timeout=0, find_stash='found')

Bases: angr.exploration_techniques.ExplorationTechnique

The Symbion exploration technique uses the SimEngineConcrete available to step a SimState.

Parameters

find – address or list of addresses that we want to reach, these will be translated into breakpoints inside the concrete process using the ConcreteTarget interface provided by the user inside the SimEngineConcrete.
memory_concretize – list of tuples (address, symbolic variable) that are going to be written in the concrete process memory.
register_concretize – list of tuples (reg_name, symbolic variable) that are going to be written
timeout – how long we should wait the concrete target to reach the breakpoint

setup(simgr)

step(simgr, stash='active', **kwargs)

step_state(simgr, *args, **kwargs)

complete(simgr)

class angr.exploration_techniques.memory_watcher.MemoryWatcher(min_memory=512, memory_stash='lowmem')

Bases: angr.exploration_techniques.ExplorationTechnique

Memory Watcher

Parameters

min_memory (int,optional) – Minimum amount of free memory in MB before stopping execution (default: 95% memory use)
memory_stash (str, optional) – What to call the low memory stash (default: ‘lowmem’)

At each step, keep an eye on how much memory is left on the system. Stash off states to effectively stop execution if we’re below a given threshold.

setup(simgr)

step(simgr, stash='active', **kwargs)

class angr.exploration_techniques.bucketizer.Bucketizer

Bases: angr.exploration_techniques.ExplorationTechnique

Loop bucketization: Pick log(n) paths out of n possible paths, and stash (or drop) everything else.

successors(simgr, state, **kwargs)

Simulation Engines

class angr.engines.UberEngine(*args, **kwargs): Bases: angr.engines.failure.SimEngineFailure, angr.engines.syscall.SimEngineSyscall, angr.engines.hook.HooksMixin, angr.engines.unicorn.SimEngineUnicorn, angr.engines.vex.heavy.super_fastpath.SuperFastpathMixin, angr.engines.vex.heavy.actions.TrackActionsMixin, angr.engines.vex.heavy.inspect.SimInspectMixin, angr.engines.vex.heavy.resilience.HeavyResilienceMixin, angr.engines.soot.engine.SootMixin, angr.engines.vex.heavy.heavy.HeavyVEXMixin

class angr.engines.engine.SimEngineBase(project=None, **kwargs)

Bases: object

Even more basey of a base class for SimEngine. Used as a base by mixins which want access to the project but for which having method process (contained in SimEngine) doesn’t make sense

class angr.engines.engine.SimEngine(project=None, **kwargs)

Bases: angr.engines.engine.SimEngineBase

A SimEngine is a class which understands how to perform execution on a state. This is a base class.

abstract process(state, **kwargs)

The main entry point for an engine. Should take a state and return a result.

Parameters: state – The state to proceed from
Returns: The result. Whatever you want ;)

class angr.engines.engine.TLSMixin(*args, **kwargs)

Bases: object

Mix this class into any class that defines __tls to make all of the attributes named in that list into thread-local properties.

MAGIC MAGIC MAGIC

class angr.engines.engine.TLSProperty(name): Bases: object

class angr.engines.engine.SuccessorsMixin(*args, **kwargs)

Bases: angr.engines.engine.SimEngine

A mixin for SimEngine which implements process to perform common operations related to symbolic execution and dispatches to a process_successors method to fill a SimSuccessors object with the results.

process(state, *args, **kwargs)

Perform execution with a state.

You should only override this method in a subclass in order to provide the correct method signature and docstring. You should override the _process method to do your actual execution.

Parameters

state – The state with which to execute. This state will be copied before modification.
inline – This is an inline execution. Do not bother copying the state.
force_addr – Force execution to pretend that we’re working at this concrete address

Returns

A SimSuccessors object categorizing the execution’s successor states

process_successors(successors, **kwargs)

Implement this function to fill out the SimSuccessors object with the results of stepping state.

In order to implement a model where multiple mixins can potentially handle a request, a mixin may implement this method and then perform a super() call if it wants to pass on handling to the next mixin.

Keep in mind python’s method resolution order when composing multiple classes implementing this method. In short: left-to-right, depth-first, but deferring any base classes which are shared by multiple subclasses (the merge point of a diamond pattern in the inheritance graph) until the last point where they would be encountered in this depth-first search. For example, if you have classes A, B(A), C(B), D(A), E(C, D), then the method resolution order will be E, C, B, D, A.

Parameters

state – The state to manipulate
successors – The successors object to fill out
kwargs – Any extra arguments. Do not fail if you are passed unexpected arguments.

class angr.engines.successors.SimSuccessors(addr, initial_state)

Bases: object

This class serves as a categorization of all the kinds of result states that can come from a SimEngine run.

Variables

addr (int) – The address at which execution is taking place, as a python int
initial_state – The initial state for which execution produced these successors
engine – The engine that produced these successors
sort – A string identifying the type of engine that produced these successors
processed (bool) – Whether or not the processing succeeded
description (str) – A textual description of the execution step

The successor states produced by this run are categorized into several lists:

Variables

artifacts (dict) – Any analysis byproducts (for example, an IRSB) that were produced during execution
successors – The “normal” successors. IP may be symbolic, but must have reasonable number of solutions
unsat_successors – Any successor which is unsatisfiable after its guard condition is added.
all_successors – successors + unsat_successors
flat_successors – The normal successors, but any symbolic IPs have been concretized. There is one state in this list for each possible value an IP may be concretized to for each successor state.
unconstrained_successors – Any state for which during the flattening process we find too many solutions.

A more detailed description of the successor lists may be found here: https://docs.angr.io/core-concepts/simulation#simsuccessors

classmethod failure()

property is_empty

add_successor(state, target, guard, jumpkind, add_guard=True, exit_stmt_idx=None, exit_ins_addr=None, source=None)

Add a successor state of the SimRun. This procedure stores method parameters into state.scratch, does some housekeeping, and calls out to helper functions to prepare the state and categorize it into the appropriate successor lists.

Parameters

state (SimState) – The successor state.
target – The target (of the jump/call/ret).
guard – The guard expression.
jumpkind (str) – The jumpkind (call, ret, jump, or whatnot).
add_guard (bool) – Whether to add the guard constraint (default: True).
exit_stmt_idx (int) – The ID of the exit statement, an integer by default. ‘default’ stands for the default exit, and None means it’s not from a statement (for example, from a SimProcedure).
exit_ins_addr (int) – The instruction pointer of this exit, which is an integer by default.
source (int) – The source of the jump (i.e., the address of the basic block).

class angr.engines.procedure.ProcedureMixin

Bases: object

A mixin for SimEngine which adds the process_procedure method for calling a SimProcedure and adding its results to a SimSuccessors.

process_procedure(state, successors, procedure, ret_to=None, arguments=None, **kwargs)

class angr.engines.procedure.ProcedureEngine(*args, **kwargs)

Bases: angr.engines.procedure.ProcedureMixin, angr.engines.engine.SuccessorsMixin

A SimEngine that you may use if you only care about processing SimProcedures. Requires the procedure kwarg to be passed to process.

process_successors(successors, procedure=None, **kwargs)

class angr.engines.hook.HooksMixin(*args, **kwargs)

Bases: angr.engines.engine.SuccessorsMixin, angr.engines.procedure.ProcedureMixin

A SimEngine mixin which adds a SimSuccessors handler which will look into the project’s hooks and run the hook at the current address.

Will respond to the following parameters provided to the step stack:

procedure: A SimProcedure instance to force-run instead of consulting the current hooks
ret_to: An address to force-return-to at the end of the procedure

process_successors(successors, procedure=None, **kwargs)

class angr.engines.syscall.SimEngineSyscall(*args, **kwargs)

Bases: angr.engines.engine.SuccessorsMixin, angr.engines.procedure.ProcedureMixin

A SimEngine mixin which adds a successors handling step that checks if a syscall was just requested and if so handles it as a step.

process_successors(successors, **kwargs)

class angr.engines.failure.SimEngineFailure(*args, **kwargs)

Bases: angr.engines.engine.SuccessorsMixin, angr.engines.procedure.ProcedureMixin

process_successors(successors, **kwargs)

class angr.engines.soot.engine.SootMixin(*args, **kwargs)

Bases: angr.engines.engine.SuccessorsMixin, angr.engines.procedure.ProcedureMixin

Execution engine based on Soot.

lift_soot(addr=None, the_binary=None, **kwargs)

process_successors(successors, **kwargs)

get_unconstrained_simprocedure()

classmethod setup_callsite(state, args, ret_addr, ret_var=None)

static setup_arguments(state, args)

static prepare_return_state(state, ret_value=None)

static terminate_execution(statement, state, successors)

static prepare_native_return_state(native_state)

Hook target for native function call returns.

Recovers and stores the return value from native memory and toggles the state, s.t. execution continues in the Soot engine.

class angr.engines.unicorn.SimEngineUnicorn(*args, **kwargs)

Bases: angr.engines.engine.SuccessorsMixin

Concrete execution in the Unicorn Engine, a fork of qemu.

Responds to the following parameters in the step stack:

step: How many basic blocks we want to execute
extra_stop_points: A collection of addresses at which execution should halt

process_successors(successors, **kwargs)

class angr.engines.concrete.SimEngineConcrete(project)

Bases: angr.engines.engine.SuccessorsMixin

Concrete execution using a concrete target provided by the user.

process_successors(successors, extra_stop_points=None, memory_concretize=None, register_concretize=None, timeout=0, *args, **kwargs)

to_engine(state, extra_stop_points, memory_concretize, register_concretize, timeout)

Handle the concrete execution of the process This method takes care of: 1- Set the breakpoints on the addresses provided by the user 2- Concretize the symbolic variables and perform the write inside the concrete process 3- Continue the program execution.

Parameters

state – The state with which to execute
extra_stop_points – list of a addresses where to stop the concrete execution and return to the simulated one
memory_concretize – list of tuples (address, symbolic variable) that are going to be written in the concrete process memory.
register_concretize – list of tuples (reg_name, symbolic variable) that are going to be written
timeout – how long we should wait the concrete target to reach the breakpoint

Returns

None

static check_concrete_target_methods(concrete_target): Check if the concrete target methods return the correct type of data :return: True if the concrete target is compliant

Simulation Logging

class angr.state_plugins.sim_action.SimAction(state, region_type)

Bases: angr.state_plugins.sim_event.SimEvent

A SimAction represents a semantic action that an analyzed program performs.

Initializes the SimAction.

Parameters: state – the state that’s the SimAction is taking place in.

TMP = 'tmp'

REG = 'reg'

MEM = 'mem'

property all_objects

property is_symbolic

property tmp_deps

property reg_deps

copy()

downsize(): Clears some low-level details (that take up memory) out of the SimAction.

class angr.state_plugins.sim_action.SimActionExit(state, target, condition=None, exit_type=None)

Bases: angr.state_plugins.sim_action.SimAction

An Exit action represents a (possibly conditional) jump.

CONDITIONAL = 'conditional'

DEFAULT = 'default'

property all_objects

property is_symbolic

class angr.state_plugins.sim_action.SimActionConstraint(state, constraint, condition=None)

Bases: angr.state_plugins.sim_action.SimAction

A constraint action represents an extra constraint added during execution of a path.

property all_objects

property is_symbolic

class angr.state_plugins.sim_action.SimActionOperation(state, op, exprs, result)

Bases: angr.state_plugins.sim_action.SimAction

An action representing an operation between variables and/or constants.

property all_objects

property is_symbolic

class angr.state_plugins.sim_action.SimActionData(state, region_type, action, tmp=None, addr=None, size=None, data=None, condition=None, fallback=None, fd=None)

Bases: angr.state_plugins.sim_action.SimAction

A Data action represents a read or a write from memory, registers or a file.

READ = 'read'

WRITE = 'write'

OPERATE = 'operate'

downsize()

property all_objects

property is_symbolic

property tmp_deps

property reg_deps

property storage

angr.state_plugins.sim_action_object.ast_stripping_op(f, *args, **kwargs)

angr.state_plugins.sim_action_object.ast_preserving_op(f, *args, **kwargs)

angr.state_plugins.sim_action_object.ast_stripping_decorator(f)

class angr.state_plugins.sim_action_object.SimActionObject(ast, reg_deps=None, tmp_deps=None, deps=None, state=None)

Bases: object

A SimActionObject tracks an AST and its dependencies.

to_claripy()

copy()

SDiv(*args, **kwargs)

SMod(*args, **kwargs)

intersection(*args, **kwargs)

union(*args, **kwargs)

widen(*args, **kwargs)

angr.state_plugins.sim_action_object.make_methods()

class angr.state_plugins.sim_event.SimEvent(state, event_type, **kwargs): Bases: object

Procedures

class angr.sim_procedure.SimProcedure(project=None, cc=None, prototype=None, symbolic_return=None, returns=None, is_syscall=False, is_stub=False, num_args=None, display_name=None, library_name=None, is_function=None, **kwargs)

Bases: object

A SimProcedure is a wonderful object which describes a procedure to run on a state.

You may subclass SimProcedure and override run(), replacing it with mutating self.state however you like, and then either returning a value or jumping away somehow.

A detailed discussion of programming SimProcedures may be found at https://docs.angr.io/extending-angr/simprocedures

Parameters: arch – The architecture to use for this procedure

The following parameters are optional:

Parameters

symbolic_return – Whether the procedure’s return value should be stubbed into a single symbolic variable constratined to the real return value
returns – Whether the procedure should return to its caller afterwards
is_syscall – Whether this procedure is a syscall
num_args – The number of arguments this procedure should extract
display_name – The name to use when displaying this procedure
library_name – The name of the library from which the function we’re emulating comes
cc – The SimCC to use for this procedure
sim_kwargs – Additional keyword arguments to be passed to run()
is_function – Whether this procedure emulates a function

The following class variables should be set if necessary when implementing a new SimProcedure:

Variables

NO_RET – Set this to true if control flow will never return from this function
ADDS_EXITS – Set this to true if you do any control flow other than returning
IS_FUNCTION – Does this procedure simulate a function? True by default
ARGS_MISMATCH – Does this procedure have a different list of arguments than what is provided in the function specification? This may happen when we manually extract arguments in the run() method of a SimProcedure. False by default.
local_vars – If you use self.call(), set this to a list of all the local variable names in your class. They will be restored on return.

The following instance variables are available when working with simprocedures from the inside or the outside:

Variables

project – The associated angr project
arch – The associated architecture
addr – The linear address at which the procedure is executing
cc – The calling convention in use for engaging with the ABI
canonical – The canonical version of this SimProcedure. Procedures are deepcopied for many reasons, including to be able to store state related to a specific run and to be able to hook continuations.
kwargs – Any extra keyword arguments used to construct the procedure; will be passed to run
display_name – See the eponymous parameter
library_name – See the eponymous parameter
abi – If this is a syscall simprocedure, which ABI are we using to map the syscall numbers?
symbolic_return – See the eponymous parameter
syscall_number – If this procedure is a syscall, the number will be populated here.
returns – See eponymous parameter and NO_RET cvar
is_syscall – See eponymous parameter
is_function – See eponymous parameter and cvar
is_stub – See eponymous parameter
is_continuation – Whether this procedure is the original or a continuation resulting from self.call()
continuations – A mapping from name to each known continuation
run_func – The name of the function implementing the procedure. “run” by default, but different in continuations.
num_args – The number of arguments to the procedure. If not provided in the parameter, extracted from the definition of self.run

The following instance variables are only used in a copy of the procedure that is actually executing on a state:

Variables

state – The SimState we should be mutating to perform the procedure
successors – The SimSuccessors associated with the current step
arguments – The function arguments, deserialized from the state
arg_session – The ArgSession that was used to parse arguments out of the state, in case you need it for varargs
use_state_arguments – Whether we’re using arguments extracted from the state or manually provided
ret_to – The current return address
ret_expr – The computed return value
call_ret_expr – The return value from having used self.call()
inhibit_autoret – Whether we should avoid automatically adding an exit for returning once the run function ends
arg_session – The ArgSession object that was used to extract the runtime argument values. Useful for if you want to extract variadic args.

state: SimState

execute(state, successors=None, arguments=None, ret_to=None)

Call this method with a SimState and a SimSuccessors to execute the procedure.

Alternately, successors may be none if this is an inline call. In that case, you should provide arguments to the function.

make_continuation(name)

NO_RET = False

ADDS_EXITS = False

IS_FUNCTION = True

ARGS_MISMATCH = False

ALT_NAMES = None

local_vars: Tuple[str, ...] = ()

run(*args, **kwargs): Implement the actual procedure here!

static_exits(blocks, **kwargs)

Get new exits by performing static analysis and heuristics. This is a fast and best-effort approach to get new exits for scenarios where states are not available (e.g. when building a fast CFG).

Parameters: blocks (list) – Blocks that are executed before reaching this SimProcedure.
Returns: A list of dicts. Each dict should contain the following entries: ‘address’, ‘jumpkind’, and ‘namehint’.
Return type: list

property should_add_successors

set_args(args)

va_arg(ty, index=None)

inline_call(procedure, *arguments, **kwargs)

Call another SimProcedure in-line to retrieve its return value. Returns an instance of the procedure with the ret_expr property set.

Parameters

procedure – The class of the procedure to execute
arguments – Any additional positional args will be used as arguments to the procedure call
sim_kwargs – Any additional keyword args will be passed as sim_kwargs to the procedure construtor

ret(expr=None): Add an exit representing a return from this function. If this is not an inline call, grab a return address from the state and jump to it. If this is not an inline call, set a return expression with the calling convention.

call(addr, args, continue_at, cc=None, prototype=None)

Add an exit representing calling another function via pointer.

Parameters

addr – The address of the function to call
args – The list of arguments to call the function with
continue_at – Later, when the called function returns, execution of the current procedure will continue in the named method.
cc – Optional: use this calling convention for calling the new function. Default is to use the current convention.
prototype – Optional: The prototype to use for the call. Will default to all-ints.

jump(addr): Add an exit representing jumping to an address.

exit(exit_code): Add an exit representing terminating the program.

ty_ptr(ty)

property is_java

property argument_types

property return_type

class angr.procedures.stubs.format_parser.FormatString(parser, components)

Bases: object

Describes a format string.

Takes a list of components which are either just strings or a FormatSpecifier.

SCANF_DELIMITERS = [b'\t', b'\n', b'\x0b', b'\r', b' ']

property state

replace(va_arg)

Implement printf - based on the stored format specifier information, format the values from the arg getter function args into a string.

Parameters: va_arg – A function which takes a type and returns the next argument of that type
Returns: The result formatted string

interpret(va_arg, addr=None, simfd=None)

implement scanf - extract formatted data from memory or a file according to the stored format specifiers and store them into the pointers extracted from args.

Parameters

va_arg – A function which, given a type, returns the next argument of that type
addr – The address in the memory to extract data from, or…
simfd – A file descriptor to use for reading data from

Returns

The number of arguments parsed

class angr.procedures.stubs.format_parser.FormatSpecifier(string, length_spec, pad_chr, size, signed)

Bases: object

Describes a format specifier within a format string.

string

size

signed

length_spec

pad_chr

property spec_type

class angr.procedures.stubs.format_parser.FormatParser(project=None, cc=None, prototype=None, symbolic_return=None, returns=None, is_syscall=False, is_stub=False, num_args=None, display_name=None, library_name=None, is_function=None, **kwargs)

Bases: angr.sim_procedure.SimProcedure

For SimProcedures relying on printf-style format strings.

ARGS_MISMATCH = True

basic_spec = {b'A': double, b'E': double, b'F': double, b'G': double, b'X': unsigned int, b'a': double, b'c': char, b'd': int, b'e': double, b'f': double, b'g': double, b'i': int, b'n': unsigned int*, b'o': unsigned int, b'p': unsigned int*, b's': char*, b'u': unsigned int, b'x': unsigned int}

int_sign = {'signed': [b'd', b'i'], 'unsigned': [b'o', b'u', b'x', b'X']}

int_len_mod = {b'h': (short, unsigned short), b'hh': (char, char), b'j': (long long, unsigned long long), b'l': (long, unsigned long), b'll': (long long, unsigned long long), b't': (long, long), b'z': (size_t, size_t)}

other_types = {('string',): <function FormatParser.<lambda>>}

flags = ['#', '0', '\\-', ' ', '\\+', "\\'", 'I']

extract_components(fmt: List) → List

Extract the actual formats from the format string fmt.

Parameters: fmt – A list of format chars.
Returns: a FormatString object

state: SimState

arg_session: Union[None, ArgSession, int]

class angr.procedures.stubs.format_parser.ScanfFormatParser(project=None, cc=None, prototype=None, symbolic_return=None, returns=None, is_syscall=False, is_stub=False, num_args=None, display_name=None, library_name=None, is_function=None, **kwargs)

Bases: angr.procedures.stubs.format_parser.FormatParser

For SimProcedures relying on scanf-style format strings.

basic_spec = {b'A': float, b'E': float, b'F': float, b'G': float, b'X': unsigned int, b'a': float, b'c': char, b'd': int, b'e': float, b'f': float, b'g': float, b'i': int, b'n': unsigned int*, b'o': unsigned int, b'p': unsigned int*, b's': char*, b'u': unsigned int, b'x': unsigned int}

float_spec = [b'e', b'E', b'f', b'F', b'g', b'G', b'a', b'A']

float_len_mod = {b'l': <class 'angr.sim_type.SimTypeDouble'>, b'll': <class 'angr.sim_type.SimTypeDouble'>}

state: SimState

arg_session: Union[None, ArgSession, int]

class angr.procedures.definitions.SimLibrary

Bases: object

A SimLibrary is the mechanism for describing a dynamic library’s API, its functions and metadata.

Any instance of this class (or its subclasses) found in the angr.procedures.definitions package will be automatically picked up and added to angr.SIM_LIBRARIES via all its names.

Variables

fallback_cc – A mapping from architecture to the default calling convention that should be used if no other information is present. Contains some sane defaults for linux.
fallback_proc – A SimProcedure class that should be used to provide stub procedures. By default, ReturnUnconstrained.

copy()

Make a copy of this SimLibrary, allowing it to be mutated without affecting the global version.

Returns: A new SimLibrary object with the same library references but different dict/list references

update(other)

Augment this SimLibrary with the information from another SimLibrary

Parameters: other – The other SimLibrary

property name: The first common name of this library, e.g. libc.so.6, or ‘??????’ if none are known.

set_library_names(*names)

Set some common names of this library by which it may be referred during linking

Parameters: names – Any number of string library names may be passed as varargs.

set_default_cc(arch_name, cc_cls)

Set the default calling convention used for this library under a given architecture

Parameters: arch_name – The string name of the architecture, i.e. the .name field from archinfo.
Parm cc_cls: The SimCC class (not an instance!) to use

set_non_returning(*names)

Mark some functions in this class as never returning, i.e. loops forever or terminates execution

Parameters: names – Any number of string function names may be passed as varargs

set_prototype(name, proto)

Set the prototype of a function in the form of a SimTypeFunction containing argument and return types

Parameters

name – The name of the function as a string
proto – The prototype of the function as a SimTypeFunction

set_prototypes(protos)

Set the prototypes of many functions

Parameters: protos – Dictionary mapping function names to SimTypeFunction objects

set_c_prototype(c_decl)

Set the prototype of a function in the form of a C-style function declaration.

Parameters: c_decl (str) – The C-style declaration of the function.
Returns: A tuple of (function name, function prototype)
Return type: tuple

add(name, proc_cls, **kwargs)

Add a function implementation fo the library.

Parameters

name – The name of the function as a string
proc_cls – The implementation of the function as a SimProcedure _class_, not instance
kwargs – Any additional parameters to the procedure class constructor may be passed as kwargs

add_all_from_dict(dictionary, **kwargs)

Batch-add function implementations to the library.

Parameters

dictionary – A mapping from name to procedure class, i.e. the first two arguments to add()
kwargs – Any additional kwargs will be passed to the constructors of _each_ procedure class

add_alias(name, *alt_names)

Add some duplicate names for a given function. The original function’s implementation must already be registered.

Parameters

name – The name of the function for which an implementation is already present
alt_names – Any number of alternate names may be passed as varargs

get(name, arch)

Get an implementation of the given function specialized for the given arch, or a stub procedure if none exists.

Parameters

name – The name of the function as a string
arch – The architecure to use, as either a string or an archinfo.Arch instance

Returns

A SimProcedure instance representing the function as found in the library

get_stub(name, arch)

Get a stub procedure for the given function, regardless of if a real implementation is available. This will apply any metadata, such as a default calling convention or a function prototype.

By stub, we pretty much always mean a ReturnUnconstrained SimProcedure with the appropriate display name and metadata set. This will appear in state.history.descriptions as <SimProcedure display_name (stub)>

Parameters

name – The name of the function as a string
arch – The architecture to use, as either a string or an archinfo.Arch instance

Returns

A SimProcedure instance representing a plausable stub as could be found in the library.

get_prototype(name: str, arch=None) → Optional[angr.sim_type.SimTypeFunction]

Get a prototype of the given function name, optionally specialize the prototype to a given architecture.

Parameters

name – Name of the function.
arch – The architecture to specialize to.

Returns

Prototype of the function, or None if the prototype does not exist.

has_metadata(name)

Check if a function has either an implementation or any metadata associated with it

Parameters: name – The name of the function as a string
Returns: A bool indicating if anything is known about the function

has_implementation(name)

Check if a function has an implementation associated with it

Parameters: name – The name of the function as a string
Returns: A bool indicating if an implementation of the function is available

has_prototype(func_name)

Check if a function has a prototype associated with it.

Parameters: func_name (str) – The name of the function.
Returns: A bool indicating if a prototype of the function is available.
Return type: bool

class angr.procedures.definitions.SimCppLibrary

Bases: angr.procedures.definitions.SimLibrary

SimCppLibrary is a specialized version of SimLibrary that will demangle C++ function names before looking for an implementation or prototype for it.

get(name, arch)

Get an implementation of the given function specialized for the given arch, or a stub procedure if none exists. Demangle the function name if it is a mangled C++ name.

Parameters

name (str) – The name of the function as a string
arch – The architecure to use, as either a string or an archinfo.Arch instance

Returns

A SimProcedure instance representing the function as found in the library

get_stub(name, arch)

Get a stub procedure for the given function, regardless of if a real implementation is available. This will apply any metadata, such as a default calling convention or a function prototype. Demangle the function name if it is a mangled C++ name.

Parameters

name (str) – The name of the function as a string
arch – The architecture to use, as either a string or an archinfo.Arch instance

Returns

A SimProcedure instance representing a plausable stub as could be found in the library.

get_prototype(name: str, arch=None) → Optional[angr.sim_type.SimTypeFunction]

Get a prototype of the given function name, optionally specialize the prototype to a given architecture. The function name will be demangled first.

Parameters

name – Name of the function.
arch – The architecture to specialize to.

Returns

Prototype of the function, or None if the prototype does not exist.

has_metadata(name)

Check if a function has either an implementation or any metadata associated with it. Demangle the function name if it is a mangled C++ name.

Parameters: name – The name of the function as a string
Returns: A bool indicating if anything is known about the function

has_implementation(name)

Check if a function has an implementation associated with it. Demangle the function name if it is a mangled C++ name.

Parameters: name (str) – A mangled function name.
Returns: bool

has_prototype(func_name)

Check if a function has a prototype associated with it. Demangle the function name if it is a mangled C++ name.

Parameters: name (str) – A mangled function name.
Returns: bool

class angr.procedures.definitions.SimSyscallLibrary

Bases: angr.procedures.definitions.SimLibrary

SimSyscallLibrary is a specialized version of SimLibrary for dealing not with a dynamic library’s API but rather an operating system’s syscall API. Because this interface is inherently lower-level than a dynamic library, many parts of this class has been changed to store data based on an “ABI name” (ABI = application binary interface, like an API but for when there’s no programming language) instead of an architecture. An ABI name is just an arbitrary string with which a calling convention and a syscall numbering is associated.

All the SimLibrary methods for adding functions still work, but now there’s an additional layer on top that associates them with numbers.

copy()

update(other)

minimum_syscall_number(abi)

Parameters: abi – The abi to evaluate
Returns: The smallest syscall number known for the given abi

maximum_syscall_number(abi)

Parameters: abi – The abi to evaluate
Returns: The largest syscall number known for the given abi

add_number_mapping(abi, number, name)

Associate a syscall number with the name of a function present in the underlying SimLibrary

Parameters

abi – The abi for which this mapping applies
number – The syscall number
name – The name of the function

add_number_mapping_from_dict(abi, mapping)

Batch-associate syscall numbers with names of functions present in the underlying SimLibrary

Parameters

abi – The abi for which this mapping applies
mapping – A dict mapping syscall numbers to function names

set_abi_cc(abi, cc_cls)

Set the default calling convention for an abi

Parameters

abi – The name of the abi
cc_cls – A SimCC _class_, not an instance, that should be used for syscalls using the abi

set_prototype(abi: str, name: str, proto: angr.sim_type.SimTypeFunction) → None

Set the prototype of a function in the form of a SimTypeFunction containing argument and return types

Parameters

abi – ABI of the syscall.
name – The name of the syscall as a string
proto – The prototype of the syscall as a SimTypeFunction

set_prototypes(abi: str, protos: Dict[str, angr.sim_type.SimTypeFunction]) → None

Set the prototypes of many syscalls.

Parameters

abi – ABI of the syscalls.
protos – Dictionary mapping syscall names to SimTypeFunction objects

get(number, arch, abi_list=())

The get() function for SimSyscallLibrary looks a little different from its original version.

Instead of providing a name, you provide a number, and you additionally provide a list of abi names that are applicable. The first abi for which the number is present in the mapping will be chosen. This allows for the easy abstractions of architectures like ARM or MIPS linux for which there are many ABIs that can be used at any time by using syscall numbers from various ranges. If no abi knows about the number, the stub procedure with the name “sys_%d” will be used.

Parameters

number – The syscall number
arch – The architecture being worked with, as either a string name or an archinfo.Arch
abi_list – A list of ABI names that could be used

Returns

A SimProcedure representing the implementation of the given syscall, or a stub if no implementation is available

get_stub(number, arch, abi_list=())

Pretty much the intersection of SimLibrary.get_stub() and SimSyscallLibrary.get().

Parameters

number – The syscall number
arch – The architecture being worked with, as either a string name or an archinfo.Arch
abi_list – A list of ABI names that could be used

Returns

A SimProcedure representing a plausable stub that could model the syscall

get_prototype(abi: str, name: str, arch=None) → Optional[angr.sim_type.SimTypeFunction]

Get a prototype of the given syscall name and its ABI, optionally specialize the prototype to a given architecture.

Parameters

abi – ABI of the prototype to get.
name – Name of the syscall.
arch – The architecture to specialize to.

Returns

Prototype of the syscall, or None if the prototype does not exist.

has_metadata(number, arch, abi_list=())

Pretty much the intersection of SimLibrary.has_metadata() and SimSyscallLibrary.get().

Parameters

number – The syscall number
arch – The architecture being worked with, as either a string name or an archinfo.Arch
abi_list – A list of ABI names that could be used

Returns

A bool of whether or not any implementation or metadata is known about the given syscall

has_implementation(number, arch, abi_list=())

Pretty much the intersection of SimLibrary.has_implementation() and SimSyscallLibrary.get().

Parameters

number – The syscall number
arch – The architecture being worked with, as either a string name or an archinfo.Arch
abi_list – A list of ABI names that could be used

Returns

A bool of whether or not an implementation of the syscall is available

has_prototype(abi: str, name: str) → bool

Check if a function has a prototype associated with it. Demangle the function name if it is a mangled C++ name.

Parameters

abi – Name of the ABI.
name – The syscall name.

Returns

bool

angr.procedures.definitions.load_win32api_definitions()

angr.procedures.definitions.load_all_definitions()

Calling Conventions and Types

class angr.calling_conventions.PointerWrapper(value, buffer=False): Bases: object

class angr.calling_conventions.AllocHelper(ptrsize)

Bases: object

alloc(size)

dump(val, state, loc=None)

translate(val, base)

apply(state, base)

size()

classmethod calc_size(val, arch)

classmethod stack_loc(val, arch, offset=0)

angr.calling_conventions.refine_locs_with_struct_type(arch: archinfo.arch.Arch, locs: List, arg_type: angr.sim_type.SimType, offset: int = 0, treat_bot_as_int=True)

class angr.calling_conventions.SerializableIterator

Bases: object

getstate()

setstate(state)

class angr.calling_conventions.SerializableListIterator(lst)

Bases: angr.calling_conventions.SerializableIterator

getstate()

setstate(state)

class angr.calling_conventions.SerializableCounter(start, stride, mapping=<function SerializableCounter.<lambda>>)

Bases: angr.calling_conventions.SerializableIterator

getstate()

setstate(state)

class angr.calling_conventions.SimFunctionArgument(size, is_fp=False)

Bases: object

Represent a generic function argument.

Variables

size (int) – The size of the argument, in number of bytes.
is_fp (bool) – Whether loads from this location should return a floating point bitvector

check_value_set(value, arch)

check_value_get(value)

set_value(state, value, **kwargs)

get_value(state, **kwargs)

refine(size, arch=None, offset=None, is_fp=None)

get_footprint(): Return a list of SimRegArg and SimStackArgs that are the base components used for this location

class angr.calling_conventions.SimRegArg(reg_name: str, size: int, reg_offset=0, is_fp=False, clear_entire_reg=False)

Bases: angr.calling_conventions.SimFunctionArgument

Represents a function argument that has been passed in a register.

Variables

reg_name (string) – The name of the represented register.
size (int) – The size of the data to store, in number of bytes.
reg_offset – The offset into the register to start storing data.
clear_entire_reg – Whether a store to this register should zero the unused parts of the register.
is_fp (bool) – Whether loads from this location should return a floating point bitvector

get_footprint()

check_offset(arch)

set_value(state, value, **kwargs)

get_value(state, **kwargs)

refine(size, arch=None, offset=None, is_fp=None)

sse_extend()

class angr.calling_conventions.SimStackArg(stack_offset, size, is_fp=False)

Bases: angr.calling_conventions.SimFunctionArgument

Represents a function argument that has been passed on the stack.

Variables

stack_offset (int) – The position of the argument relative to the stack pointer after the function prelude.
size (int) – The size of the argument, in number of bytes.
is_fp (bool) – Whether loads from this location should return a floating point bitvector

get_footprint()

set_value(state, value, stack_base=None, **kwargs)

get_value(state, stack_base=None, **kwargs)

refine(size, arch=None, offset=None, is_fp=None)

class angr.calling_conventions.SimComboArg(locations, is_fp=False)

Bases: angr.calling_conventions.SimFunctionArgument

An argument which spans multiple storage locations. Locations should be given least-significant first.

get_footprint()

set_value(state, value, **kwargs)

get_value(state, **kwargs)

class angr.calling_conventions.SimStructArg(struct: angr.sim_type.SimStruct, locs: Dict[str, angr.calling_conventions.SimFunctionArgument])

Bases: angr.calling_conventions.SimFunctionArgument

An argument which de/serializes a struct from a list of storage locations

Variables

struct – The simtype describing the structure
locs – The storage locations to use

get_footprint()

get_value(state, **kwargs)

set_value(state, value, **kwargs)

class angr.calling_conventions.SimArrayArg(locs)

Bases: angr.calling_conventions.SimFunctionArgument

get_footprint()

get_value(state, **kwargs)

set_value(state, value, **kwargs)

class angr.calling_conventions.SimReferenceArgument(ptr_loc, main_loc)

Bases: angr.calling_conventions.SimFunctionArgument

A function argument which is passed by reference.

Variables

ptr_loc – The location the reference’s pointer is stored
main_loc – A SimStackArgument describing how to load the argument’s value as if it were stored at offset zero on the stack. It will be passed stack_base=ptr_loc.get_value(state)

get_footprint()

get_value(state, **kwargs)

set_value(state, value, **kwargs)

class angr.calling_conventions.ArgSession(cc)

Bases: object

A class to keep track of the state accumulated in laying parameters out into memory

cc

fp_iter

int_iter

both_iter

getstate()

setstate(state)

class angr.calling_conventions.UsercallArgSession(cc)

Bases: object

An argsession for use with SimCCUsercall

cc

real_args

getstate()

setstate(state)

class angr.calling_conventions.SimCC(arch: archinfo.arch.Arch)

Bases: object

A calling convention allows you to extract from a state the data passed from function to function by calls and returns. Most of the methods provided by SimCC that operate on a state assume that the program is just after a call but just before stack frame allocation, though this may be overridden with the stack_base parameter to each individual method.

This is the base class for all calling conventions.

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = []

FP_ARG_REGS: List[str] = []

STACKARG_SP_BUFF = 0

STACKARG_SP_DIFF = 0

CALLER_SAVED_REGS: List[str] = []

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = None

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = None

OVERFLOW_RETURN_VAL: Optional[angr.calling_conventions.SimFunctionArgument] = None

FP_RETURN_VAL: Optional[angr.calling_conventions.SimFunctionArgument] = None

ARCH = None

CALLEE_CLEANUP = False

STACK_ALIGNMENT = 1

property int_args

Iterate through all the possible arg positions that can only be used to store integer or pointer values.

Returns an iterator of SimFunctionArguments

property memory_args

Iterate through all the possible arg positions that can be used to store any kind of argument.

Returns an iterator of SimFunctionArguments

property fp_args

Iterate through all the possible arg positions that can only be used to store floating point values.

Returns an iterator of SimFunctionArguments

is_fp_arg(arg)

This should take a SimFunctionArgument instance and return whether or not that argument is a floating-point argument.

Returns True for MUST be a floating point arg,: False for MUST NOT be a floating point arg, None for when it can be either.

class ArgSession(cc)

Bases: object

A class to keep track of the state accumulated in laying parameters out into memory

cc

fp_iter

int_iter

both_iter

getstate()

setstate(state)

arg_session(ret_ty: Optional[angr.sim_type.SimType])

Return an arg session.

A session provides the control interface necessary to describe how integral and floating-point arguments are laid out into memory. The default behavior is that there are a finite list of int-only and fp-only argument slots, and an infinite number of generic slots, and when an argument of a given type is requested, the most slot available is used. If you need different behavior, subclass ArgSession.

You need to provide the return type of the function in order to kick off an arg layout session.

return_in_implicit_outparam(ty)

stack_space(args)

Parameters: args – A list of SimFunctionArguments
Returns: The number of bytes that should be allocated on the stack to store all these args, NOT INCLUDING the return address.

return_val(ty, perspective_returned=False): The location the return value is stored, based on its type.

property return_addr: The location the return address is stored.

next_arg(session: angr.calling_conventions.ArgSession, arg_type: angr.sim_type.SimType)

static is_fp_value(val)

static guess_prototype(args, prototype=None)

Come up with a plausible SimTypeFunction for the given args (as would be passed to e.g. setup_callsite).

You can pass a variadic function prototype in the base_type parameter and all its arguments will be used, only guessing types for the variadic arguments.

arg_locs(prototype)

get_args(state, prototype, stack_base=None)

set_return_val(state, val, ty, stack_base=None, perspective_returned=False)

setup_callsite(state, ret_addr, args, prototype, stack_base=None, alloc_base=None, grow_like_stack=True)

This function performs the actions of the caller getting ready to jump into a function.

Parameters

state – The SimState to operate on
ret_addr – The address to return to when the called function finishes
args – The list of arguments that that the called function will see
prototype – The signature of the call you’re making. Should include variadic args concretely.
stack_base – An optional pointer to use as the top of the stack, circa the function entry point
alloc_base – An optional pointer to use as the place to put excess argument data
grow_like_stack – When allocating data at alloc_base, whether to allocate at decreasing addresses

The idea here is that you can provide almost any kind of python type in args and it’ll be translated to a binary format to be placed into simulated memory. Lists (representing arrays) must be entirely elements of the same type and size, while tuples (representing structs) can be elements of any type and size. If you’d like there to be a pointer to a given value, wrap the value in a PointerWrapper.

If stack_base is not provided, the current stack pointer will be used, and it will be updated. If alloc_base is not provided, the stack base will be used and grow_like_stack will implicitly be True.

grow_like_stack controls the behavior of allocating data at alloc_base. When data from args needs to be wrapped in a pointer, the pointer needs to point somewhere, so that data is dumped into memory at alloc_base. If you set alloc_base to point to somewhere other than the stack, set grow_like_stack to False so that sequential allocations happen at increasing addresses.

teardown_callsite(state, return_val=None, prototype=None, force_callee_cleanup=False)

This function performs the actions of the callee as it’s getting ready to return. It returns the address to return to.

Parameters

state – The state to mutate
return_val – The value to return
prototype – The prototype of the given function
force_callee_cleanup – If we should clean up the stack allocation for the arguments even if it’s not the callee’s job to do so

TODO: support the stack_base parameter from setup_callsite…? Does that make sense in this context? Maybe it could make sense by saying that you pass it in as something like the “saved base pointer” value?

static find_cc(arch: archinfo.arch.Arch, args: List[angr.calling_conventions.SimFunctionArgument], sp_delta: int) → Optional[angr.calling_conventions.SimCC]

Pinpoint the best-fit calling convention and return the corresponding SimCC instance, or None if no fit is found.

Parameters

arch – An ArchX instance. Can be obtained from archinfo.
args – A list of arguments. It may be updated by the first matched calling convention to remove non-argument arguments.
sp_delta – The change of stack pointer before and after the call is made.

Returns

A calling convention instance, or None if none of the SimCC subclasses seems to fit the arguments provided.

get_arg_info(state, prototype): This is just a simple wrapper that collects the information from various locations prototype is as passed to self.arg_locs and self.get_args :param angr.SimState state: The state to evaluate and extract the values from :return: A list of tuples, where the nth tuple is (type, name, location, value) of the nth argument

class angr.calling_conventions.SimLyingRegArg(name, size=8)

Bases: angr.calling_conventions.SimRegArg

A register that LIES about the types it holds

get_value(state, **kwargs)

set_value(state, val, **kwargs)

refine(size, arch=None, offset=None, is_fp=None)

class angr.calling_conventions.SimCCUsercall(arch, arg_locs, ret_loc)

Bases: angr.calling_conventions.SimCC

ArgSession: alias of angr.calling_conventions.UsercallArgSession

next_arg(session, arg_type)

return_val(ty, **kwargs)

class angr.calling_conventions.SimCCCdecl(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = []

FP_ARG_REGS: List[str] = []

STACKARG_SP_DIFF = 4

CALLER_SAVED_REGS: List[str] = ['eax', 'ecx', 'edx']

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <eax>

OVERFLOW_RETURN_VAL: Optional[angr.calling_conventions.SimFunctionArgument] = <edx>

FP_RETURN_VAL: Optional[angr.calling_conventions.SimFunctionArgument] = <st0>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = [0x0]

ARCH: alias of archinfo.arch_x86.ArchX86

next_arg(session, arg_type)

STRUCT_RETURN_THRESHOLD = 32

return_val(ty, perspective_returned=False)

return_in_implicit_outparam(ty)

class angr.calling_conventions.SimCCMicrosoftCdecl(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCCdecl

Parameters: arch – The Archinfo arch for this CC

STRUCT_RETURN_THRESHOLD = 64

class angr.calling_conventions.SimCCStdcall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCMicrosoftCdecl

Parameters: arch – The Archinfo arch for this CC

CALLEE_CLEANUP = True

class angr.calling_conventions.SimCCMicrosoftFastcall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['ecx', 'edx']

STACKARG_SP_DIFF = 4

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <eax>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = [0x0]

ARCH: alias of archinfo.arch_x86.ArchX86

class angr.calling_conventions.MicrosoftAMD64ArgSession(cc): Bases: object

class angr.calling_conventions.SimCCMicrosoftAMD64(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['rcx', 'rdx', 'r8', 'r9']

FP_ARG_REGS: List[str] = ['xmm0', 'xmm1', 'xmm2', 'xmm3']

STACKARG_SP_DIFF = 8

STACKARG_SP_BUFF = 32

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <rax>

OVERFLOW_RETURN_VAL: Optional[angr.calling_conventions.SimFunctionArgument] = <rdx>

FP_RETURN_VAL: Optional[angr.calling_conventions.SimFunctionArgument] = <xmm0>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = [0x0]

ARCH: alias of archinfo.arch_amd64.ArchAMD64

STACK_ALIGNMENT = 16

ArgSession: alias of angr.calling_conventions.MicrosoftAMD64ArgSession

next_arg(session, arg_type)

return_in_implicit_outparam(ty)

class angr.calling_conventions.SimCCSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

The base class of all syscall CCs.

Parameters: arch – The Archinfo arch for this CC

ERROR_REG: angr.calling_conventions.SimRegArg = None

SYSCALL_ERRNO_START = None

static syscall_num(state) → int

linux_syscall_update_error_reg(state, expr)

set_return_val(state, val, ty, **kwargs)

class angr.calling_conventions.SimCCX86LinuxSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['ebx', 'ecx', 'edx', 'esi', 'edi', 'ebp']

FP_ARG_REGS: List[str] = []

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <eax>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_x86.ArchX86

static syscall_num(state)

class angr.calling_conventions.SimCCX86WindowsSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = []

FP_ARG_REGS: List[str] = []

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <eax>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_x86.ArchX86

static syscall_num(state)

class angr.calling_conventions.SimCCSystemVAMD64(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['rdi', 'rsi', 'rdx', 'rcx', 'r8', 'r9']

FP_ARG_REGS: List[str] = ['xmm0', 'xmm1', 'xmm2', 'xmm3', 'xmm4', 'xmm5', 'xmm6', 'xmm7']

STACKARG_SP_DIFF = 8

CALLER_SAVED_REGS: List[str] = ['rdi', 'rsi', 'rdx', 'rcx', 'r8', 'r9', 'r10', 'r11', 'rax']

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = [0x0]

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <rax>

OVERFLOW_RETURN_VAL: Optional[angr.calling_conventions.SimFunctionArgument] = <rdx>

FP_RETURN_VAL: Optional[angr.calling_conventions.SimFunctionArgument] = <xmm0>

OVERFLOW_FP_RETURN_VAL = <xmm1>

ARCH: alias of archinfo.arch_amd64.ArchAMD64

STACK_ALIGNMENT = 16

next_arg(session, arg_type)

return_val(ty: Optional[angr.sim_type.SimType], perspective_returned=False)

return_in_implicit_outparam(ty)

class angr.calling_conventions.SimCCAMD64LinuxSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['rdi', 'rsi', 'rdx', 'r10', 'r8', 'r9']

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <rax>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_amd64.ArchAMD64

static syscall_num(state)

class angr.calling_conventions.SimCCAMD64WindowsSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = []

FP_ARG_REGS: List[str] = []

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <rax>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_amd64.ArchAMD64

static syscall_num(state)

class angr.calling_conventions.SimCCARM(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['r0', 'r1', 'r2', 'r3']

FP_ARG_REGS: List[str] = []

CALLER_SAVED_REGS: List[str] = []

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <lr>

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <r0>

ARCH: alias of archinfo.arch_arm.ArchARM

class angr.calling_conventions.SimCCARMLinuxSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['r0', 'r1', 'r2', 'r3']

FP_ARG_REGS: List[str] = []

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <r0>

ARCH: alias of archinfo.arch_arm.ArchARM

static syscall_num(state)

class angr.calling_conventions.SimCCAArch64(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['x0', 'x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7']

FP_ARG_REGS: List[str] = []

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <lr>

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <x0>

ARCH: alias of archinfo.arch_aarch64.ArchAArch64

class angr.calling_conventions.SimCCAArch64LinuxSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['x0', 'x1', 'x2', 'x3', 'x4', 'x5', 'x6', 'x7']

FP_ARG_REGS: List[str] = []

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <x0>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_aarch64.ArchAArch64

static syscall_num(state)

class angr.calling_conventions.SimCCO32(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['a0', 'a1', 'a2', 'a3']

FP_ARG_REGS: List[str] = []

STACKARG_SP_BUFF = 16

CALLER_SAVED_REGS: List[str] = ['t9', 'gp']

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ra>

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <v0>

ARCH: alias of archinfo.arch_mips32.ArchMIPS32

class angr.calling_conventions.SimCCO32LinuxSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['a0', 'a1', 'a2', 'a3']

FP_ARG_REGS: List[str] = []

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <v0>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_mips32.ArchMIPS32

ERROR_REG: angr.calling_conventions.SimRegArg = <a3>

SYSCALL_ERRNO_START = -1133

static syscall_num(state)

class angr.calling_conventions.SimCCO64(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['a0', 'a1', 'a2', 'a3']

CALLER_SAVED_REGS: List[str] = ['t9', 'gp']

FP_ARG_REGS: List[str] = []

STACKARG_SP_BUFF = 32

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ra>

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <v0>

ARCH: alias of archinfo.arch_mips64.ArchMIPS64

class angr.calling_conventions.SimCCO64LinuxSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['a0', 'a1', 'a2', 'a3', 'a4', 'a5', 'a6', 'a7']

FP_ARG_REGS: List[str] = []

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <v0>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_mips64.ArchMIPS64

ERROR_REG: angr.calling_conventions.SimRegArg = <a3>

SYSCALL_ERRNO_START = -1133

static syscall_num(state)

class angr.calling_conventions.SimCCPowerPC(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['r3', 'r4', 'r5', 'r6', 'r7', 'r8', 'r9', 'r10']

FP_ARG_REGS: List[str] = []

STACKARG_SP_BUFF = 8

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <lr>

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <r3>

ARCH: alias of archinfo.arch_ppc32.ArchPPC32

class angr.calling_conventions.SimCCPowerPCLinuxSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['r3', 'r4', 'r5', 'r6', 'r7', 'r8', 'r9', 'r10']

FP_ARG_REGS: List[str] = []

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <r3>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_ppc32.ArchPPC32

ERROR_REG: angr.calling_conventions.SimRegArg = <cr0_0>

SYSCALL_ERRNO_START = -515

static syscall_num(state)

class angr.calling_conventions.SimCCPowerPC64(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['r3', 'r4', 'r5', 'r6', 'r7', 'r8', 'r9', 'r10']

FP_ARG_REGS: List[str] = []

STACKARG_SP_BUFF = 112

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <lr>

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <r3>

ARCH: alias of archinfo.arch_ppc64.ArchPPC64

class angr.calling_conventions.SimCCPowerPC64LinuxSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['r3', 'r4', 'r5', 'r6', 'r7', 'r8', 'r9', 'r10']

FP_ARG_REGS: List[str] = []

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <r3>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_ppc64.ArchPPC64

ERROR_REG: angr.calling_conventions.SimRegArg = <cr0_0>

SYSCALL_ERRNO_START = -515

static syscall_num(state)

class angr.calling_conventions.SimCCSoot(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARCH: alias of archinfo.arch_soot.ArchSoot

ARG_REGS: List[str] = []

setup_callsite(state, ret_addr, args, prototype, stack_base=None, alloc_base=None, grow_like_stack=True)

static guess_prototype(args, prototype=None)

class angr.calling_conventions.SimCCUnknown(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Represent an unknown calling convention.

Parameters: arch – The Archinfo arch for this CC

class angr.calling_conventions.SimCCS390X(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCC

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['r2', 'r3', 'r4', 'r5', 'r6']

FP_ARG_REGS: List[str] = ['f0', 'f2', 'f4', 'f6']

STACKARG_SP_BUFF = 160

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <r14>

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <r2>

ARCH: alias of archinfo.arch_s390x.ArchS390X

class angr.calling_conventions.SimCCS390XLinuxSyscall(arch: archinfo.arch.Arch)

Bases: angr.calling_conventions.SimCCSyscall

Parameters: arch – The Archinfo arch for this CC

ARG_REGS: List[str] = ['r2', 'r3', 'r4', 'r5', 'r6', 'r7']

FP_ARG_REGS: List[str] = []

RETURN_VAL: angr.calling_conventions.SimFunctionArgument = <r2>

RETURN_ADDR: angr.calling_conventions.SimFunctionArgument = <ip_at_syscall>

ARCH: alias of archinfo.arch_s390x.ArchS390X

static syscall_num(state)

angr.calling_conventions.register_default_cc(arch, cc)

angr.calling_conventions.register_syscall_cc(arch, os, cc)

class angr.sim_variable.SimVariable(ident=None, name=None, region: Optional[int] = None, category=None)

Bases: angr.serializable.Serializable

Parameters

ident – A unique identifier provided by user or the program. Usually a string.
name (str) – Name of this variable.

ident

name

region: Optional[int]

category: Optional[str]

renamed

candidate_names

copy()

loc_repr(arch: archinfo.Arch): The representation that shows up in a GUI

property is_function_argument

class angr.sim_variable.SimConstantVariable(ident=None, value=None, region=None)

Bases: angr.sim_variable.SimVariable

value

loc_repr(arch)

copy() → angr.sim_variable.SimConstantVariable

class angr.sim_variable.SimTemporaryVariable(tmp_id)

Bases: angr.sim_variable.SimVariable

tmp_id

loc_repr(arch)

copy() → angr.sim_variable.SimTemporaryVariable

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, **kwargs)

class angr.sim_variable.SimRegisterVariable(reg_offset, size, ident=None, name=None, region=None, category=None)

Bases: angr.sim_variable.SimVariable

reg: int

size: int

property bits

loc_repr(arch)

copy() → angr.sim_variable.SimRegisterVariable

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, **kwargs)

class angr.sim_variable.SimMemoryVariable(addr, size, ident=None, name=None, region=None, category=None)

Bases: angr.sim_variable.SimVariable

addr

size

loc_repr(arch)

property bits

copy() → angr.sim_variable.SimMemoryVariable

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, **kwargs)

class angr.sim_variable.SimStackVariable(offset, size, base='sp', base_addr=None, ident=None, name=None, region=None, category=None)

Bases: angr.sim_variable.SimMemoryVariable

base

offset

base_addr

loc_repr(arch)

copy() → angr.sim_variable.SimStackVariable

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, **kwargs)

class angr.sim_variable.SimVariableSet

Bases: collections.abc.MutableSet

A collection of SimVariables.

add(item)

add_register_variable(reg_var)

add_memory_variable(mem_var)

discard(item)

discard_register_variable(reg_var)

discard_memory_variable(mem_var)

add_memory_variables(addrs, size)

copy()

complement(other)

Calculate the complement of self and other.

Parameters: other – Another SimVariableSet instance.
Returns: The complement result.

contains_register_variable(reg_var)

contains_memory_variable(mem_var)

class angr.sim_type.SimType(label=None)

Bases: object

SimType exists to track type information for SimProcedures.

Parameters: label – the type label.

base = True

property size: The size of the type in bits.

property alignment: The alignment of the type in bytes.

with_arch(arch)

c_repr(name=None, full=0, memo=None, indent=0)

copy()

extract_claripy(bits)

Given a bitvector bits which was loaded from memory in a big-endian fashion, return a more appropriate or structured representation of the data.

A type must have an arch associated in order to use this method.

class angr.sim_type.TypeRef(name, ty)

Bases: angr.sim_type.SimType

A TypeRef is a reference to a type with a name. This allows for interactivity in type analysis, by storing a type and having the option to update it later and have all references to it automatically update as well.

property name: This is a read-only property because it is desirable to store typerefs in a mapping from name to type, and we want the mapping to be in the loop for any updates.

property size

property alignment

with_arch(arch)

c_repr(name=None, full=0, memo=None, indent=0)

copy()

class angr.sim_type.NamedTypeMixin(*args, name: Optional[str] = None, **kwargs)

Bases: object

SimType classes with this mixin in the class hierarchy allows setting custom class names. A typical use case is to represent same or similar type classes with different qualified names, such as “std::basic_string” vs “std::__cxx11::basic_string”. In such cases, .name stores the qualified name, and .unqualified_name() returns the unqualified name of the type.

property name: str

unqualified_name(lang: str = 'c++') → str

class angr.sim_type.SimTypeBottom(label=None)

Bases: angr.sim_type.SimType

SimTypeBottom basically represents a type error.

copy()

class angr.sim_type.SimTypeTop(size=None, label=None)

Bases: angr.sim_type.SimType

SimTypeTop represents any type (mostly used with a pointer for void*).

copy()

class angr.sim_type.SimTypeReg(size, label=None)

Bases: angr.sim_type.SimType

SimTypeReg is the base type for all types that are register-sized.

Parameters

label – the type label.
size – the size of the type (e.g. 32bit, 8bit, etc.).

extract(state, addr, concrete=False)

store(state, addr, value)

copy()

class angr.sim_type.SimTypeNum(size, signed=True, label=None)

Bases: angr.sim_type.SimType

SimTypeNum is a numeric type of arbitrary length

Parameters

size – The size of the integer, in bits
signed – Whether the integer is signed or not
label – A label for the type

extract(state, addr, concrete=False)

store(state, addr, value)

copy()

class angr.sim_type.SimTypeInt(signed=True, label=None)

Bases: angr.sim_type.SimTypeReg

SimTypeInt is a type that specifies a signed or unsigned C integer.

Parameters

signed – True if signed, False if unsigned
label – The type label

c_repr(name=None, full=0, memo=None, indent=0)

property size

extract(state, addr, concrete=False)

copy()

class angr.sim_type.SimTypeShort(signed=True, label=None)

Bases: angr.sim_type.SimTypeInt

Parameters

signed – True if signed, False if unsigned
label – The type label

class angr.sim_type.SimTypeLong(signed=True, label=None)

Bases: angr.sim_type.SimTypeInt

Parameters

signed – True if signed, False if unsigned
label – The type label

class angr.sim_type.SimTypeLongLong(signed=True, label=None)

Bases: angr.sim_type.SimTypeInt

Parameters

signed – True if signed, False if unsigned
label – The type label

class angr.sim_type.SimTypeChar(signed=True, label=None)

Bases: angr.sim_type.SimTypeReg

SimTypeChar is a type that specifies a character; this could be represented by a byte, but this is meant to be interpreted as a character.

Parameters: label – the type label.

store(state, addr, value)

extract(state, addr, concrete=False)

copy()

class angr.sim_type.SimTypeBool(signed=True, label=None)

Bases: angr.sim_type.SimTypeChar

Parameters: label – the type label.

store(state, addr, value)

extract(state, addr, concrete=False)

class angr.sim_type.SimTypeFd(label=None)

Bases: angr.sim_type.SimTypeReg

SimTypeFd is a type that specifies a file descriptor.

Parameters: label – the type label

copy()

class angr.sim_type.SimTypePointer(pts_to, label=None, offset=0)

Bases: angr.sim_type.SimTypeReg

SimTypePointer is a type that specifies a pointer to some other type.

Parameters

label – The type label.
pts_to – The type to which this pointer points.

c_repr(name=None, full=0, memo=None, indent=0)

make(pts_to)

property size

copy()

class angr.sim_type.SimTypeReference(refs, label=None)

Bases: angr.sim_type.SimTypeReg

SimTypeReference is a type that specifies a reference to some other type.

c_repr(name=None, full=0, memo=None, indent=0)

make(refs)

property size

copy()

class angr.sim_type.SimTypeFixedSizeArray(elem_type, length)

Bases: angr.sim_type.SimType

SimTypeFixedSizeArray is a literal (i.e. not a pointer) fixed-size array.

c_repr(name=None, full=0, memo=None, indent=0)

extract(state, addr, concrete=False)

store(state, addr, values)

property size

property alignment

copy()

class angr.sim_type.SimTypeArray(elem_type, length=None, label=None)

Bases: angr.sim_type.SimType

SimTypeArray is a type that specifies a pointer to an array; while it is a pointer, it has a semantic difference.

Parameters

label – The type label.
elem_type – The type of each element in the array.
length – An expression of the length of the array, if known.

c_repr(name=None, full=0, memo=None, indent=0)

property size

property alignment

copy()

class angr.sim_type.SimTypeString(length=None, label=None, name: Optional[str] = None)

Bases: angr.sim_type.NamedTypeMixin, angr.sim_type.SimTypeArray

SimTypeString is a type that represents a C-style string, i.e. a NUL-terminated array of bytes.

Parameters

label – The type label.
length – An expression of the length of the string, if known.

extract(state: angr.sim_state.SimState, addr, concrete=False)

property size

property alignment

copy()

class angr.sim_type.SimTypeWString(length=None, label=None, name: Optional[str] = None)

Bases: angr.sim_type.NamedTypeMixin, angr.sim_type.SimTypeArray

A wide-character null-terminated string, where each character is 2 bytes.

extract(state, addr, concrete=False)

property size

property alignment

copy()

class angr.sim_type.SimTypeFunction(args, returnty, label=None, arg_names=None, variadic=False)

Bases: angr.sim_type.SimType

SimTypeFunction is a type that specifies an actual function (i.e. not a pointer) with certain types of arguments and a certain return value.

Parameters

label – The type label
args – A tuple of types representing the arguments to the function
returnty – The return type of the function, or none for void
variadic – Whether the function accepts varargs

base = False

c_repr(name=None, full=0, memo=None, indent=0)

property size

copy()

class angr.sim_type.SimTypeCppFunction(args, returnty, label=None, arg_names: Optional[Tuple[str]] = None, ctor: bool = False, dtor: bool = False)

Bases: angr.sim_type.SimTypeFunction

SimTypeCppFunction is a type that specifies an actual C++-style function with information about arguments, return value, and more C++-specific properties.

Variables

ctor – Whether the function is a constructor or not.
dtor – Whether the function is a destructor or not.

copy()

returnty: Optional[angr.sim_type.SimType]

class angr.sim_type.SimTypeLength(signed=False, addr=None, length=None, label=None)

Bases: angr.sim_type.SimTypeLong

SimTypeLength is a type that specifies the length of some buffer in memory.

…I’m not really sure what the original design of this class was going for

Parameters

signed – Whether the value is signed or not
label – The type label.
addr – The memory address (expression).
length – The length (expression).

property size

copy()

class angr.sim_type.SimTypeFloat(size=32)

Bases: angr.sim_type.SimTypeReg

An IEEE754 single-precision floating point number

sort = FLOAT

signed = True

extract(state, addr, concrete=False)

store(state, addr, value)

copy()

class angr.sim_type.SimTypeDouble(align_double=True)

Bases: angr.sim_type.SimTypeFloat

An IEEE754 double-precision floating point number

sort = DOUBLE

property alignment

copy()

class angr.sim_type.SimStruct(fields: Union[Dict[str, angr.sim_type.SimType], collections.OrderedDict], name=None, pack=False, align=None)

Bases: angr.sim_type.NamedTypeMixin, angr.sim_type.SimType

property packed

property offsets: Dict[str, int]

extract(state, addr, concrete=False)

c_repr(name=None, full=0, memo=None, indent=0)

property size

property alignment

store(state, addr, value)

copy()

class angr.sim_type.SimStructValue(struct, values=None)

Bases: object

A SimStruct type paired with some real values

Parameters

struct – A SimStruct instance describing the type of this struct
values – A mapping from struct fields to values

property struct

copy()

class angr.sim_type.SimUnion(members, name=None, label=None)

Bases: angr.sim_type.NamedTypeMixin, angr.sim_type.SimType

Parameters

members – The members of the union, as a mapping name -> type
name – The name of the union

fields = ('members', 'name')

property size

property alignment

extract(state, addr, concrete=False)

c_repr(name=None, full=0, memo=None, indent=0)

copy()

class angr.sim_type.SimUnionValue(union, values=None)

Bases: object

A SimStruct type paired with some real values

Parameters

union – A SimUnion instance describing the type of this union
values – A mapping from union members to values

copy()

class angr.sim_type.SimCppClass(members: Optional[Dict[str, angr.sim_type.SimType]] = None, function_members: Optional[Dict[str, angr.sim_type.SimTypeCppFunction]] = None, vtable_ptrs=None, name: Optional[str] = None, pack: bool = False, align=None)

Bases: angr.sim_type.SimStruct

property members

extract(state, addr, concrete=False)

store(state, addr, value)

copy()

class angr.sim_type.SimCppClassValue(class_type, values)

Bases: object

A SimCppClass type paired with some real values

copy()

class angr.sim_type.SimTypeNumOffset(size, signed=True, label=None, offset=0)

Bases: angr.sim_type.SimTypeNum

like SimTypeNum, but supports an offset of 1 to 7 to a byte aligned address to allow structs with bitfields

extract(state: angr.sim_state.SimState, addr, concrete=False)

store(state, addr, value)

copy()

angr.sim_type.register_types(types)

Pass in some types and they will be registered to the global type store.

The argument may be either a mapping from name to SimType, or a plain SimType. The plain SimType must be either a struct or union type with a name present.

>>> register_types(parse_types("typedef int x; typedef float y;"))
>>> register_types(parse_type("struct abcd { int ab; float cd; }"))

angr.sim_type.do_preprocess(defn, include_path=()): Run a string through the C preprocessor that ships with pycparser but is weirdly inaccessible?

angr.sim_type.parse_signature(defn, preprocess=True, predefined_types=None, arch=None): Parse a single function prototype and return its type

angr.sim_type.parse_defns(defn, preprocess=True, predefined_types=None, arch=None): Parse a series of C definitions, returns a mapping from variable name to variable type object

angr.sim_type.parse_types(defn, preprocess=True, predefined_types=None, arch=None): Parse a series of C definitions, returns a mapping from type name to type object

angr.sim_type.parse_file(defn, preprocess=True, predefined_types=None, arch=None): Parse a series of C definitions, returns a tuple of two type mappings, one for variable definitions and one for type definitions.

angr.sim_type.parse_type(defn, preprocess=True, predefined_types=None, arch=None)

Parse a simple type expression into a SimType

>>> parse_type('int *')

angr.sim_type.parse_type_with_name(defn, preprocess=True, predefined_types=None, arch=None)

Parse a simple type expression into a SimType, returning the a tuple of the type object and any associated name that might be found in the place a name would go in a type declaration.

>>> parse_type_with_name('int *foo')

angr.sim_type.normalize_cpp_function_name(name: str) → str

angr.sim_type.parse_cpp_file(cpp_decl, with_param_names: bool = False)

class angr.type_backend.TypedValue(ty, value): Bases: claripy.backend_object.BackendObject

class angr.type_backend.TypeBackend

Bases: claripy.backends.Backend

apply_annotation(obj, a)

static default_op(expr)

class angr.type_backend.TypeAnnotation(ty)

Bases: claripy.annotation.Annotation

property eliminatable

property relocatable

class angr.callable.Callable(project, addr, prototype=None, concrete_only=False, perform_merge=True, base_state=None, toc=None, cc=None)

Bases: object

Callable is a representation of a function in the binary that can be interacted with like a native python function.

If you set perform_merge=True (the default), the result will be returned to you, and you can get the result state with callable.result_state.

Otherwise, you can get the resulting simulation manager at callable.result_path_group.

Parameters

project – The project to operate on
addr – The address of the function to use

The following parameters are optional:

Parameters

prototype – The signature of the calls you would like to make. This really shouldn’t be optional.
concrete_only – Throw an exception if the execution splits into multiple paths
perform_merge – Merge all result states into one at the end (only relevant if concrete_only=False)
base_state – The state from which to do these runs
toc – The address of the table of contents for ppc64
cc – The SimCC to use for a calling convention

set_base_state(state): Swap out the state you’d like to use to perform the call :param state: The state to use to perform the call

perform_call(*args, prototype=None)

call_c(c_args)

Call this Callable with a string of C-style arguments.

Parameters: c_args (str) – C-style arguments.
Returns: The return value from the call.
Return type: claripy.Ast

Knowledge Base

Representing the artifacts of a project.

class angr.knowledge_base.knowledge_base.KnowledgeBase(project, obj=None, name=None)

Bases: object

Represents a “model” of knowledge about an artifact.

Contains things like a CFG, data references, etc.

functions: FunctionManager

variables: VariableManager

structured_code: StructuredCodeManager

defs: KeyDefinitionManager

cfgs: CFGManager

types: TypesStore

property callgraph

property unresolved_indirect_jumps

property resolved_indirect_jumps

has_plugin(name)

get_plugin(name)

register_plugin(name, plugin)

release_plugin(name)

K: alias of TypeVar(‘K’, bound=angr.knowledge_plugins.plugin.KnowledgeBasePlugin)

get_knowledge(requested_plugin_cls: Type[K]) → Optional[K]

Type inference safe method to request a knowledge base plugin Explicitly passing the type of the requested plugin achieves two things: 1. Every location using this plugin can be easily found with an IDE by searching explicit references to the type 2. Basic type inference can deduce the result type and properly type check usages of it

If there isn’t already an instance of this class None will be returned to make it clear to the caller that there is no existing knowledge of this type yet. The code that initially creates this knowledge should use the register_plugin method to register the initial knowledge state :param requested_plugin_cls: :return: Instance of the requested plugin class or null if it is not a known plugin

request_knowledge(requested_plugin_cls: Type[K]) → K

class angr.knowledge_plugins.patches.Patch(addr, new_bytes, comment: Optional[str] = None): Bases: object

class angr.knowledge_plugins.patches.PatchManager(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin

A placeholder-style implementation for a binary patch manager. This class should be significantly changed in the future when all data about loaded binary objects are loaded into angr knowledge base from CLE. As of now, it only stores byte-level replacements. Other angr components may choose to use or not use information provided by this manager. In other words, it is not transparent.

Patches should not overlap, but it’s user’s responsibility to check for and avoid overlapping patches.

add_patch(addr, new_bytes, comment: Optional[str] = None)

add_patch_obj(patch: angr.knowledge_plugins.patches.Patch)

remove_patch(addr)

patch_addrs()

get_patch(addr)

Get patch at the given address.

Parameters: addr (int) – The address of the patch.
Returns: The patch if there is one starting at the address, or None if there isn’t any.
Return type: Patch or None

get_all_patches(addr, size)

Retrieve all patches that cover a region specified by [addr, addr+size).

Parameters

addr (int) – The address of the beginning of the region.
size (int) – Size of the region.

Returns

A list of patches.

Return type

list

keys()

items()

values()

copy()

static overlap(a0, a1, b0, b1)

apply_patches_to_binary(binary_bytes: Optional[bytes] = None, patches: Optional[List[angr.knowledge_plugins.patches.Patch]] = None) → bytes

class angr.knowledge_plugins.plugin.KnowledgeBasePlugin

Bases: object

copy()

static register_default(name, cls)

class angr.knowledge_plugins.cfg.cfg_model.CFGModel(ident, cfg_manager=None, is_arm=False)

Bases: angr.serializable.Serializable

This class describes a Control Flow Graph for a specific range of code.

ident

is_arm

graph

jump_tables: Dict[int, angr.knowledge_plugins.cfg.indirect_jump.IndirectJump]

memory_data: Dict[int, angr.knowledge_plugins.cfg.memory_data.MemoryData]

insn_addr_to_memory_data: Dict[int, angr.knowledge_plugins.cfg.memory_data.MemoryData]

property project

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, cfg_manager=None, loader=None)

copy()

add_node(block_id: int, node: angr.knowledge_plugins.cfg.cfg_node.CFGNode) → None

remove_node(block_id: int, node: angr.knowledge_plugins.cfg.cfg_node.CFGNode) → None

get_node(block_id)

Get a single node from node key.

Parameters: block_id (BlockID) – Block ID of the node.
Returns: The CFGNode
Return type: CFGNode

get_any_node(addr: int, is_syscall: Optional[bool] = None, anyaddr: bool = False, force_fastpath: bool = False) → Optional[angr.knowledge_plugins.cfg.cfg_node.CFGNode]

Get an arbitrary CFGNode (without considering their contexts) from our graph.

Parameters

addr – Address of the beginning of the basic block. Set anyaddr to True to support arbitrary address.
is_syscall – Whether you want to get the syscall node or any other node. This is due to the fact that syscall SimProcedures have the same address as the targer it returns to. None means get either, True means get a syscall node, False means get something that isn’t a syscall node.
anyaddr – If anyaddr is True, then addr doesn’t have to be the beginning address of a basic block. By default the entire graph.nodes() will be iterated, and the first node containing the specific address is returned, which can be slow.
force_fastpath – If force_fastpath is True, it will only perform a dict lookup in the _nodes_by_addr dict.

Returns

A CFGNode if there is any that satisfies given conditions, or None otherwise

get_all_nodes(addr: int, is_syscall: Optional[bool] = None, anyaddr: bool = False) → List[angr.knowledge_plugins.cfg.cfg_node.CFGNode]

Get all CFGNodes whose address is the specified one.

Parameters

addr – Address of the node
is_syscall – True returns the syscall node, False returns the normal CFGNode, None returns both

Returns

all CFGNodes

nodes()

An iterator of all nodes in the graph.

Returns: The iterator.
Return type: iterator

get_predecessors(cfgnode: angr.knowledge_plugins.cfg.cfg_node.CFGNode, excluding_fakeret: bool = True, jumpkind: Optional[str] = None) → List[angr.knowledge_plugins.cfg.cfg_node.CFGNode]

Get predecessors of a node in the control flow graph.

Parameters

cfgnode – The node.
excluding_fakeret – True if you want to exclude all predecessors that is connected to the node with a fakeret edge.
jumpkind – Only return predecessors with the specified jumpkind. This argument will be ignored if set to None.

Returns

A list of predecessors

get_successors(node: angr.knowledge_plugins.cfg.cfg_node.CFGNode, excluding_fakeret: bool = True, jumpkind: Optional[str] = None) → List[angr.knowledge_plugins.cfg.cfg_node.CFGNode]

Get successors of a node in the control flow graph.

Parameters

node (CFGNode) – The node.
excluding_fakeret (bool) – True if you want to exclude all successors that is connected to the node with a fakeret edge.
jumpkind (str or None) – Only return successors with the specified jumpkind. This argument will be ignored if set to None.

Returns

A list of successors

Return type

list

get_successors_and_jumpkinds(node, excluding_fakeret=True)

Get a list of tuples where the first element is the successor of the CFG node and the second element is the jumpkind of the successor.

Parameters

node (CFGNode) – The node.
excluding_fakeret (bool) – True if you want to exclude all successors that are fall-through successors.

Returns

A list of successors and their corresponding jumpkinds.

Return type

list

get_successors_and_jumpkind(node, excluding_fakeret=True)

Get a list of tuples where the first element is the successor of the CFG node and the second element is the jumpkind of the successor.

Parameters

node (CFGNode) – The node.
excluding_fakeret (bool) – True if you want to exclude all successors that are fall-through successors.

Returns

A list of successors and their corresponding jumpkinds.

Return type

list

get_predecessors_and_jumpkinds(node: angr.knowledge_plugins.cfg.cfg_node.CFGNode, excluding_fakeret: bool = True) → List[Tuple[angr.knowledge_plugins.cfg.cfg_node.CFGNode, str]]

Get a list of tuples where the first element is the predecessor of the CFG node and the second element is the jumpkind of the predecessor.

Parameters

node – The node.
excluding_fakeret – True if you want to exclude all predecessors that are fall-through predecessors.

Returns

A list of predecessors and their corresponding jumpkinds.

get_predecessors_and_jumpkind(node: angr.knowledge_plugins.cfg.cfg_node.CFGNode, excluding_fakeret: bool = True) → List[Tuple[angr.knowledge_plugins.cfg.cfg_node.CFGNode, str]]

Get a list of tuples where the first element is the predecessor of the CFG node and the second element is the jumpkind of the predecessor.

Parameters

node – The node.
excluding_fakeret – True if you want to exclude all predecessors that are fall-through predecessors.

Returns

A list of predecessors and their corresponding jumpkinds.

get_all_predecessors(cfgnode, depth_limit=None)

Get all predecessors of a specific node on the control flow graph.

Parameters

cfgnode (CFGNode) – The CFGNode object
depth_limit (int) – Optional depth limit for the depth-first search

Returns

A list of predecessors in the CFG

Return type

list

get_all_successors(cfgnode, depth_limit=None)

Get all successors of a specific node on the control flow graph.

Parameters

cfgnode (CFGNode) – The CFGNode object
depth_limit (int) – Optional depth limit for the depth-first search

Returns

A list of successors in the CFG

Return type

list

get_branching_nodes(): Returns all nodes that has an out degree >= 2

get_exit_stmt_idx(src_block, dst_block)

Get the corresponding exit statement ID for control flow to reach destination block from source block. The exit statement ID was put on the edge when creating the CFG. Note that there must be a direct edge between the two blocks, otherwise an exception will be raised.

Returns: The exit statement ID

class angr.knowledge_plugins.cfg.memory_data.MemoryDataSort

Bases: object

Unspecified = None

Unknown = 'unknown'

Integer = 'integer'

PointerArray = 'pointer-array'

String = 'string'

UnicodeString = 'unicode'

SegmentBoundary = 'segment-boundary'

CodeReference = 'code reference'

GOTPLTEntry = 'GOT PLT Entry'

ELFHeader = 'elf-header'

FloatingPoint = 'fp'

class angr.knowledge_plugins.cfg.memory_data.MemoryData(address, size, sort, pointer_addr=None, max_size=None)

Bases: angr.serializable.Serializable

MemoryData describes the syntactic content of a single address of memory.

addr

size

sort

max_size

pointer_addr

content

property address

copy()

Make a copy of the MemoryData.

Returns: A copy of the MemoryData instance.
Return type: MemoryData

fill_content(loader)

Load data to fill self.content.

Parameters: loader – The project loader.
Returns: None

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, **kwargs)

class angr.knowledge_plugins.cfg.cfg_manager.CFGManager(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin

new_model(prefix)

copy()

get_most_accurate() → Optional[angr.knowledge_plugins.cfg.cfg_model.CFGModel]

Returns: The most accurate CFG present in the CFGManager, or None if it does not hold any.

class angr.knowledge_plugins.cfg.cfg_node.CFGNodeCreationFailure(exc_info=None, to_copy=None)

Bases: object

This class contains additional information for whenever creating a CFGNode failed. It includes a full traceback and the exception messages.

short_reason

long_reason

traceback

class angr.knowledge_plugins.cfg.cfg_node.CFGNode(addr, size, cfg, simprocedure_name=None, no_ret=False, function_address=None, block_id=None, irsb=None, soot_block=None, instruction_addrs=None, thumb=False, byte_string=None, is_syscall=None, name=None)

Bases: angr.serializable.Serializable

This class stands for each single node in CFG.

Note: simprocedure_name is not used to recreate the SimProcedure object. It’s only there for better __repr__.

addr

size

simprocedure_name

no_ret

function_address

thumb

byte_string

is_syscall

instruction_addrs

irsb

soot_block

has_return

block_id

property name

property successors

property predecessors

successors_and_jumpkinds(excluding_fakeret=True)

predecessors_and_jumpkinds(excluding_fakeret=True)

get_data_references(kb=None)

Get the known data references for this CFGNode via the knowledge base.

Parameters: kb – Which knowledge base to use; uses the global KB by default if none is provided
Returns: Generator yielding xrefs to this CFGNode’s block.
Return type: iter

property accessed_data_references

Property providing a view of all the known data references for this CFGNode via the global knowledge base

Returns: Generator yielding xrefs to this CFGNode’s block.
Return type: iter

property is_simprocedure

property callstack_key

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, cfg=None)

copy()

merge(other): Merges this node with the other, returning a new node that spans the both.

to_codenode()

property block

syscall_name

class angr.knowledge_plugins.cfg.cfg_node.CFGENode(addr, size, cfg, simprocedure_name=None, no_ret=False, function_address=None, block_id=None, irsb=None, instruction_addrs=None, thumb=False, byte_string=None, is_syscall=None, name=None, input_state=None, final_states=None, syscall_name=None, looping_times=0, depth=None, callstack_key=None, creation_failure_info=None)

Bases: angr.knowledge_plugins.cfg.cfg_node.CFGNode

The CFGNode that is used in CFGEmulated.

input_state

looping_times

depth

creation_failure_info

final_states

return_target

syscall

property callstack_key

property creation_failed

downsize(): Drop saved states.

copy()

class angr.knowledge_plugins.cfg.indirect_jump.IndirectJumpType

Bases: object

Jumptable_AddressLoadedFromMemory = 0

Jumptable_AddressComputed = 1

Vtable = 3

Unknown = 255

class angr.knowledge_plugins.cfg.indirect_jump.IndirectJump(addr: int, ins_addr: int, func_addr: int, jumpkind: str, stmt_idx: int, resolved_targets: Optional[List[int]] = None, jumptable: bool = False, jumptable_addr: Optional[int] = None, jumptable_size: Optional[int] = None, jumptable_entry_size: Optional[int] = None, jumptable_entries: Optional[List[int]] = None, type_: Optional[int] = 255)

Bases: angr.serializable.Serializable

addr

ins_addr

func_addr

jumpkind

stmt_idx

resolved_targets

jumptable

jumptable_addr

jumptable_size

jumptable_entry_size

jumptable_entries

type

class angr.knowledge_plugins.types.TypesStore(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin, collections.UserDict

A kb plugin that stores a mapping from name to TypeRef. It will return types from angr.sim_type.ALL_TYPES as a default.

copy()

iter_own(): Iterate over all the names which are stored in this object - i.e. values() without ALL_TYPES

rename(old, new)

unique_type_name() → str

class angr.knowledge_plugins.propagations.Propagations(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin

exists(func_loc)

Internal function to check if a func, specified as a CodeLocation exists in our known propagations

Parameters: func_loc – CodeLocation of function
Returns: Bool

update(func_loc, replacements)

Add the replacements to known propagations

Parameters

func_loc – CodeLocation of function
replacements – Dict of replacements

get(func_loc)

Gets the replacements for a specified function location. If the replacement does not exist in the known propagations, it returns None.

Parameters: func_loc – CodeLocation of function
Returns: Dict or None

copy()

class angr.knowledge_plugins.comments.Comments(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin, dict

copy()

class angr.knowledge_plugins.data.Data(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin

copy()

class angr.knowledge_plugins.indirect_jumps.IndirectJumps(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin, dict

copy()

update_resolved_addrs(indirect_address: int, resolved_addresses: List[int])

class angr.knowledge_plugins.labels.Labels(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin

items()

get(addr): Get a label as string for a given address Same as .labels[x]

lookup(name): Returns an address to a given label To show all available labels, iterate over .labels or list(b.kb.labels)

copy()

get_unique_label(label)

Get a unique label name from the given label name.

Parameters: label (str) – The desired label name.
Returns: A unique label name.

class angr.knowledge_plugins.functions.function_manager.FunctionDict(backref, *args, **kwargs)

Bases: sortedcontainers.sorteddict.SortedDict

FunctionDict is a dict where the keys are function starting addresses and map to the associated Function.

get(addr)

floor_addr(addr)

ceiling_addr(addr)

class angr.knowledge_plugins.functions.function_manager.FunctionManager(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin, collections.abc.Mapping

This is a function boundaries management tool. It takes in intermediate results during CFG generation, and manages a function map of the binary.

copy()

clear()

get_by_addr(addr) → angr.knowledge_plugins.functions.function.Function

contains_addr(addr)

Decide if an address is handled by the function manager.

Note: this function is non-conformant with python programming idioms, but its needed for performance reasons.

Parameters: addr (int) – Address of the function.

ceiling_func(addr)

Return the function who has the least address that is greater than or equal to addr.

Parameters: addr (int) – The address to query.
Returns: A Function instance, or None if there is no other function after addr.
Return type: Function or None

floor_func(addr)

Return the function who has the greatest address that is less than or equal to addr.

Parameters: addr (int) – The address to query.
Returns: A Function instance, or None if there is no other function before addr.
Return type: Function or None

function(addr=None, name=None, create=False, syscall=False, plt=None) → Optional[angr.knowledge_plugins.functions.function.Function]

Get a function object from the function manager.

Pass either addr or name with the appropriate values.

Parameters

addr (int) – Address of the function.
name (str) – Name of the function.
create (bool) – Whether to create the function or not if the function does not exist.
syscall (bool) – True to create the function as a syscall, False otherwise.
plt (bool or None) – True to find the PLT stub, False to find a non-PLT stub, None to disable this restriction.

Returns

The Function instance, or None if the function is not found and create is False.

Return type

Function or None

dbg_draw(prefix='dbg_function_')

rebuild_callgraph()

class angr.knowledge_plugins.functions.function.Function(function_manager, addr, name=None, syscall=None, is_simprocedure=None, binary_name=None, is_plt=None, returning=None, alignment=False)

Bases: angr.serializable.Serializable

A representation of a function and various information about it.

Function constructor. If the optional parameters are not provided, they will be automatically determined upon the creation of a Function object.

Parameters: addr – The address of the function.

The following parameters are optional.

Parameters

name (str) – The name of the function.
syscall (bool) – Whether this function is a syscall or not.
is_simprocedure (bool) – Whether this function is a SimProcedure or not.
binary_name (str) – Name of the binary where this function is.
is_plt (bool) – If this function is a PLT entry.
returning (bool) – If this function returns.
alignment (bool) – If this function acts as an alignment filler. Such functions usually only contain nops.

transition_graph

normalized

addr

startpoint

alignment

bp_on_stack

retaddr_on_stack

sp_delta

prototype

is_prototype_guessed: bool

prepared_registers

prepared_stack_variables

registers_read_afterwards

info

tags

ran_cca

is_syscall

is_simprocedure

is_plt

is_default_name

from_signature

binary_name

calling_convention

property name

property project

property returning

property blocks

An iterator of all local blocks in the current function.

Returns: angr.lifter.Block instances.

property block_addrs

An iterator of all local block addresses in the current function.

Returns: block addresses.

property block_addrs_set

Return a set of block addresses for a better performance of inclusion tests.

Returns: A set of block addresses.
Return type: set

get_block(addr: int, size: Optional[int] = None, byte_string: Optional[bytes] = None)

Getting a block out of the current function.

Parameters

addr (int) – The address of the block.
size (int) – The size of the block. This is optional. If not provided, angr will load
byte_string –

Returns

property nodes: Generator[angr.codenode.CodeNode, None, None]

get_node(addr)

property has_unresolved_jumps

property has_unresolved_calls

property operations: All of the operations that are done by this functions.

property code_constants: All of the constants that are used by this functions’s code.

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, **kwargs)

Parameters: cmsg –
Return Function: The function instantiated out of the cmsg data.

string_references(minimum_length=2, vex_only=False)

All of the constant string references used by this function.

Parameters

minimum_length – The minimum length of strings to find (default is 1)
vex_only – Only analyze VEX IR, don’t interpret the entry state to detect additional constants.

Returns

A list of tuples of (address, string) where is address is the location of the string in memory.

property local_runtime_values

Tries to find all runtime values of this function which do not come from inputs. These values are generated by starting from a blank state and reanalyzing the basic blocks once each. Function calls are skipped, and back edges are never taken so these values are often unreliable, This function is good at finding simple constant addresses which the function will use or calculate.

Returns: a set of constants

property num_arguments

property endpoints

property endpoints_with_type

property ret_sites

property jumpout_sites

property retout_sites

property callout_sites

property size

property binary: Get the object this function belongs to. :return: The object this function belongs to.

property offset: int

the function’s binary offset (i.e., non-rebased address)

Type: return

property symbol: Union[None, cle.backends.symbol.Symbol]

the function’s Symbol, if any

Type: return

add_jumpout_site(node)

Add a custom jumpout site.

Parameters: node – The address of the basic block that control flow leaves during this transition.
Returns: None

add_retout_site(node)

Add a custom retout site.

Retout (returning to outside of the function) sites are very rare. It mostly occurs during CFG recovery when we incorrectly identify the beginning of a function in the first iteration, and then correctly identify that function later in the same iteration (function alignments can lead to this bizarre case). We will mark all edges going out of the header of that function as a outside edge, because all successors now belong to the incorrectly-identified function. This identification error will be fixed in the second iteration of CFG recovery. However, we still want to keep track of jumpouts/retouts during the first iteration so other logic in CFG recovery still work.

Parameters: node – The address of the basic block that control flow leaves the current function after a call.
Returns: None

mark_nonreturning_calls_endpoints()

Iterate through all call edges in transition graph. For each call a non-returning function, mark the source basic block as an endpoint.

This method should only be executed once all functions are recovered and analyzed by CFG recovery, so we know whether each function returns or not.

Returns: None

get_call_sites() → Iterable[int]

Gets a list of all the basic blocks that end in calls.

Returns: A view of the addresses of the blocks that end in calls.

get_call_target(callsite_addr)

Get the target of a call.

Parameters: callsite_addr – The address of a basic block that ends in a call.
Returns: The target of said call, or None if callsite_addr is not a callsite.

get_call_return(callsite_addr)

Get the hypothetical return address of a call.

Parameters: callsite_addr – The address of the basic block that ends in a call.
Returns: The likely return target of said call, or None if callsite_addr is not a callsite.

property graph

Get a local transition graph. A local transition graph is a transition graph that only contains nodes that belong to the current function. All edges, except for the edges going out from the current function or coming from outside the current function, are included.

The generated graph is cached in self._local_transition_graph.

Returns: A local transition graph.
Return type: networkx.DiGraph

graph_ex(exception_edges=True)

Get a local transition graph with a custom configuration. A local transition graph is a transition graph that only contains nodes that belong to the current function. This method allows user to exclude certain types of edges together with the nodes that are only reachable through such edges, such as exception edges.

The generated graph is not cached.

Parameters: exception_edges (bool) – Should exception edges and the nodes that are only reachable through exception edges be kept.
Returns: A local transition graph with a special configuration.
Return type: networkx.DiGraph

transition_graph_ex(exception_edges=True)

Get a transition graph with a custom configuration. This method allows user to exclude certain types of edges together with the nodes that are only reachable through such edges, such as exception edges.

The generated graph is not cached.

Parameters: exception_edges (bool) – Should exception edges and the nodes that are only reachable through exception edges be kept.
Returns: A local transition graph with a special configuration.
Return type: networkx.DiGraph

subgraph(ins_addrs)

Generate a sub control flow graph of instruction addresses based on self.graph

Parameters: ins_addrs (iterable) – A collection of instruction addresses that should be included in the subgraph.
Return networkx.DiGraph: A subgraph.

instruction_size(insn_addr)

Get the size of the instruction specified by insn_addr.

Parameters: insn_addr (int) – Address of the instruction
Return int: Size of the instruction in bytes, or None if the instruction is not found.

addr_to_instruction_addr(addr)

Obtain the address of the instruction that covers @addr.

Parameters: addr (int) – An address.
Returns: Address of the instruction that covers @addr, or None if this addr is not covered by any instruction of this function.
Return type: int or None

dbg_print(): Returns a representation of the list of basic blocks in this function.

dbg_draw(filename): Draw the graph and save it to a PNG file.

property arguments

property has_return

property callable

normalize()

Make sure all basic blocks in the transition graph of this function do not overlap. You will end up with a CFG that IDA Pro generates.

This method does not touch the CFG result. You may call CFG{Emulated, Fast}.normalize() for that matter.

Returns: None

find_declaration()

Find the most likely function declaration from the embedded collection of prototypes, set it to self.prototype, and update self.calling_convention with the declaration.

Returns: None

property demangled_name

apply_definition(definition, calling_convention=None)

Parameters

definition (str) –
calling_convention (Optional[Union[SimCC, Type[SimCC]]]) –

Return None

functions_called() → Set[angr.knowledge_plugins.functions.function.Function]

Returns: The set of all functions that can be reached from the function represented by self.

copy()

pp(**kwargs): Pretty-print the function disassembly.

class angr.knowledge_plugins.functions.function_parser.FunctionParser

Bases: object

The implementation of the serialization methods for the <Function> class.

static serialize(function): :return :

static parse_from_cmsg(cmsg, function_manager=None, project=None, all_func_addrs=None)

Parameters: cmsg – The data to instanciate the <Function> from.
Return Function

class angr.knowledge_plugins.functions.soot_function.SootFunction(function_manager, addr, name=None, syscall=None)

Bases: angr.knowledge_plugins.functions.function.Function

A representation of a function and various information about it.

Function constructor for Soot

Parameters

addr – The address of the function.
name – (Optional) The name of the function.
syscall – (Optional) Whether this function is a syscall or not.

transition_graph

normalized

addr

is_syscall

is_plt

is_simprocedure

binary_name

bp_on_stack

retaddr_on_stack

sp_delta

calling_convention

prototype

alignment

prepared_registers

prepared_stack_variables

registers_read_afterwards

startpoint

info

tags

normalize()

is_default_name

from_signature

is_prototype_guessed: bool

ran_cca

class angr.knowledge_plugins.variables.variable_access.VariableAccessSort

Bases: object

Provides enums for variable access types.

WRITE = 0

READ = 1

REFERENCE = 2

class angr.knowledge_plugins.variables.variable_access.VariableAccess(variable, access_type, location, offset, atom_hash=None)

Bases: angr.serializable.Serializable

Describes a variable access.

variable: SimVariable

access_type: int

location: angr.code_location.CodeLocation

offset: Optional[int]

atom_hash: Optional[int]

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, variable_by_ident: Optional[Dict[str, SimVariable]] = None, **kwargs) → VariableAccess

class angr.knowledge_plugins.variables.variable_manager.VariableType

Bases: object

Describes variable types.

REGISTER = 0

MEMORY = 1

class angr.knowledge_plugins.variables.variable_manager.LiveVariables(register_region, stack_region)

Bases: object

A collection of live variables at a program point.

register_region

stack_region

class angr.knowledge_plugins.variables.variable_manager.VariableManagerInternal(manager, func_addr=None)

Bases: angr.serializable.Serializable

Manage variables for a function. It is meant to be used internally by VariableManager.

set_manager(manager: angr.knowledge_plugins.variables.variable_manager.VariableManager)

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, variable_manager=None, func_addr=None, **kwargs) → angr.knowledge_plugins.variables.variable_manager.VariableManagerInternal

next_variable_ident(sort)

add_variable(sort, start, variable)

set_variable(sort, start, variable: angr.sim_variable.SimVariable)

write_to(variable, offset, location, overwrite=False, atom=None)

read_from(variable, offset, location, overwrite=False, atom=None)

reference_at(variable, offset, location, overwrite=False, atom=None)

make_phi_node(block_addr, *variables)

Create a phi variable for variables at block block_addr.

Parameters

block_addr (int) – The address of the current block.
variables – Variables that the phi variable represents.

Returns

The created phi variable.

set_live_variables(addr, register_region, stack_region)

find_variables_by_insn(ins_addr, sort)

find_variable_by_stmt(block_addr, stmt_idx, sort)

find_variables_by_stmt(block_addr: int, stmt_idx: int, sort: str) → List[Tuple[angr.sim_variable.SimVariable, int]]

find_variable_by_atom(block_addr, stmt_idx, atom)

find_variables_by_atom(block_addr, stmt_idx, atom) → Set[Tuple[angr.sim_variable.SimVariable, int]]

find_variables_by_stack_offset(offset: int) → Set[angr.sim_variable.SimVariable]

get_variable_accesses(variable: angr.sim_variable.SimVariable, same_name: bool = False) → List[angr.knowledge_plugins.variables.variable_access.VariableAccess]

get_variables(sort=None, collapse_same_ident=False) → List[Union[angr.sim_variable.SimStackVariable, angr.sim_variable.SimRegisterVariable]]

Get a list of variables.

Parameters

sort (str or None) – Sort of the variable to get.
collapse_same_ident – Whether variables of the same identifier should be collapsed or not.

Returns

A list of variables.

Return type

list

get_global_variables(addr)

Get global variable by the address of the variable.

Parameters: addr (int) – Address of the variable.
Returns: A set of variables or an empty set if no variable exists.

is_phi_variable(var)

Test if var is a phi variable.

Parameters: var (SimVariable) – The variable instance.
Returns: True if var is a phi variable, False otherwise.
Return type: bool

get_phi_subvariables(var)

Get sub-variables that phi variable var represents.

Parameters: var (SimVariable) – The variable instance.
Returns: A set of sub-variables, or an empty set if var is not a phi variable.
Return type: set

get_phi_variables(block_addr)

Get a dict of phi variables and their corresponding variables.

Parameters: block_addr (int) – Address of the block.
Returns: A dict of phi variables of an empty dict if there are no phi variables at the block.
Return type: dict

input_variables(exclude_specials=True)

Get all variables that have never been written to.

Returns: A list of variables that are never written to.

assign_variable_names(labels=None, types=None)

Assign default names to all SSA variables.

Parameters: labels – Known labels in the binary.
Returns: None

assign_unified_variable_names(labels=None, reset: bool = False)

Assign default names to all unified variables.

Parameters

labels – Known labels in the binary.
reset – Reset all variable names or not.

Returns

None

set_variable_type(var: angr.sim_variable.SimVariable, ty: angr.sim_type.SimType, name: Optional[str] = None) → None

get_variable_type(var) → Optional[angr.sim_type.SimType]

remove_types()

unify_variables() → None: Map SSA variables to a unified variable. Fill in self._unified_variables.

set_unified_variable(variable: angr.sim_variable.SimVariable, unified: angr.sim_variable.SimVariable) → None

Set the unified variable for a given SSA variable.

Parameters

variable – The SSA variable.
unified – The unified variable.

Returns

None

unified_variable(variable: angr.sim_variable.SimVariable) → Optional[angr.sim_variable.SimVariable]

Return the unified variable for a given SSA variable,

Parameters: variable – The SSA variable.
Returns: The unified variable, or None if there is no such SSA variable.

class angr.knowledge_plugins.variables.variable_manager.VariableManager(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin

Manage variables.

has_function_manager(key: int) → bool

get_function_manager(func_addr) → angr.knowledge_plugins.variables.variable_manager.VariableManagerInternal

initialize_variable_names() → None

get_variable_accesses(variable: angr.sim_variable.SimVariable, same_name: bool = False) → List[angr.knowledge_plugins.variables.variable_access.VariableAccess]

Get a list of all references to the given variable.

Parameters

variable – The variable.
same_name – Whether to include all variables with the same variable name, or just based on the variable identifier.

Returns

All references to the variable.

copy()

static convert_variable_list(vlist: List[cle.backends.elf.variable.Variable], manager: angr.knowledge_plugins.variables.variable_manager.VariableManagerInternal)

load_from_dwarf(cu_list: Optional[List[cle.backends.elf.compilation_unit.CompilationUnit]] = None)

class angr.knowledge_plugins.structured_code.manager.StructuredCodeManager(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin

discard(key)

available_flavors(item)

copy()

class angr.knowledge_plugins.key_definitions.atoms.Atom

Bases: object

This class represents a data storage location manipulated by IR instructions.

It could either be a Tmp (temporary variable), a Register, a MemoryLocation.

property size: int

static from_argument(argument: angr.calling_conventions.SimFunctionArgument, registers: Dict[str, Tuple[int, int]])

Instanciate an Atom from a given argument.

Parameters

argument – The argument to create a new atom from.
registers – A mapping representing the registers of a given architecture.

class angr.knowledge_plugins.key_definitions.atoms.GuardUse(target)

Bases: angr.knowledge_plugins.key_definitions.atoms.Atom

Implements a guard use.

target

property size: int

class angr.knowledge_plugins.key_definitions.atoms.FunctionCall(target, callsite)

Bases: angr.knowledge_plugins.key_definitions.atoms.Atom

target

callsite

property single_target

property size

class angr.knowledge_plugins.key_definitions.atoms.ConstantSrc(const)

Bases: angr.knowledge_plugins.key_definitions.atoms.Atom

const

property size

class angr.knowledge_plugins.key_definitions.atoms.Tmp(tmp_idx: int, size: int)

Bases: angr.knowledge_plugins.key_definitions.atoms.Atom

Represents a variable used by the IR to store intermediate values.

tmp_idx

property size: int

class angr.knowledge_plugins.key_definitions.atoms.Register(reg_offset: int, size: int)

Bases: angr.knowledge_plugins.key_definitions.atoms.Atom

Represents a given CPU register.

As an IR abstracts the CPU design to target different architectures, registers are represented as a separated memory space. Thus a register is defined by its offset from the base of this memory and its size.

Variables

reg_offset (int) – The offset from the base to define its place in the memory bloc.
size (int) – The size, in number of bytes.

reg_offset

property bits: int

property size: int

class angr.knowledge_plugins.key_definitions.atoms.MemoryLocation(addr: Union[angr.engines.light.data.SpOffset, angr.knowledge_plugins.key_definitions.heap_address.HeapAddress, int], size: int, endness: Optional[str] = None)

Bases: angr.knowledge_plugins.key_definitions.atoms.Atom

Represents a memory slice.

It is characterized by its address and its size.

Parameters

addr (int) – The address of the beginning memory location slice.
size (int) – The size of the represented memory location, in bytes.

addr: Union[angr.engines.light.data.SpOffset, int, claripy.ast.bv.BV]

endness

property is_on_stack: bool: True if this memory location is located on the stack.

property bits: int

property size: int

property symbolic: bool

class angr.knowledge_plugins.key_definitions.constants.ObservationPointType(value)

Bases: enum.IntEnum

Enum to replace the previously generic constants This makes it possible to annotate where they are expected by typing something as ObservationPointType instead of Literal[0,1]

OP_BEFORE = 0

OP_AFTER = 1

class angr.knowledge_plugins.key_definitions.definition.Definition(atom: angr.knowledge_plugins.key_definitions.atoms.Atom, codeloc: angr.code_location.CodeLocation, dummy: bool = False, tags: Optional[Set[angr.knowledge_plugins.key_definitions.tag.Tag]] = None)

Bases: object

An atom definition.

Variables

atom – The atom being defined.
codeloc – Where this definition is created in the original binary code.
data – A concrete value (or many concrete values) that the atom holds when the definition is created.
dummy – Tell whether the definition should be considered dummy or not. During simplification by AILment, definitions marked as dummy will not be removed.
tags – A set of tags containing information about the definition gathered during analyses.

atom: angr.knowledge_plugins.key_definitions.atoms.Atom

codeloc: angr.code_location.CodeLocation

dummy: bool

tags

property offset: int

property size: int

data

class angr.knowledge_plugins.key_definitions.environment.Environment(environment: Optional[Dict[Union[str, angr.knowledge_plugins.key_definitions.undefined.Undefined], Set[claripy.ast.base.Base]]] = None)

Bases: object

Represent the environment in which a program runs. It’s a mapping of variable names, to claripy.ast.Base that should contain possible addresses, or <UNDEFINED>, at which their respective values are stored.

Note: The <Environment> object does not store the values associated with variables themselves.

get(names: Set[str]) → Tuple[Set[claripy.ast.base.Base], bool]

Parameters: names – Potential values for the name of the environment variable to get the pointers of.
Returns: The potential addresses of the values the environment variable can take; And a boolean value telling whether all the names were known of the internal representation (i.e. will be False if one of the queried variable was not found).

set(name: Union[str, angr.knowledge_plugins.key_definitions.undefined.Undefined], pointers: Set[claripy.ast.base.Base])

Parameters

name – Name of the environment variable to which we will associate the pointers.
pointers – New addresses where the new values of the environment variable are located.

merge(*others: angr.knowledge_plugins.key_definitions.environment.Environment) → Tuple[angr.knowledge_plugins.key_definitions.environment.Environment, bool]

class angr.knowledge_plugins.key_definitions.heap_address.HeapAddress(value: Union[int, angr.knowledge_plugins.key_definitions.undefined.Undefined])

Bases: object

The representation of an address on the heap.

property value

class angr.knowledge_plugins.key_definitions.key_definition_manager.RDAObserverControl(func_addr: int, call_site_block_addrs: Iterable[int], call_site_ins_addrs: Iterable[int])

Bases: object

rda_observe_callback(ob_type, **kwargs)

class angr.knowledge_plugins.key_definitions.key_definition_manager.KeyDefinitionManager(kb: KnowledgeBase)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin

KeyDefinitionManager manages and caches reaching definition models for each function.

For each function, by default we cache the entire reaching definitions model with observed results at the following locations: - Before each call instruction: (‘insn’, address of the call instruction, OP_BEFORE) - After returning from each call: (‘node’, address of the block that ends with a call, OP_AFTER)

has_model(func_addr: int)

get_model(func_addr: int)

copy() → angr.knowledge_plugins.key_definitions.key_definition_manager.KeyDefinitionManager

class angr.knowledge_plugins.key_definitions.live_definitions.DefinitionAnnotation(definition)

Bases: claripy.annotation.Annotation

definition

property relocatable

property eliminatable

class angr.knowledge_plugins.key_definitions.live_definitions.LiveDefinitions(arch: archinfo.arch.Arch, track_tmps: bool = False, canonical_size=8, register_definitions=None, stack_definitions=None, memory_definitions=None, heap_definitions=None)

Bases: object

A LiveDefinitions instance contains definitions and uses for register, stack, memory, and temporary variables, uncovered during the analysis.

INITIAL_SP_32BIT = 2147418112

INITIAL_SP_64BIT = 140737488289792

project

arch

track_tmps

register_definitions

stack_definitions

memory_definitions

heap_definitions

tmps: Dict[int, Set[angr.knowledge_plugins.key_definitions.definition.Definition]]

register_uses

stack_uses

heap_uses

memory_uses

uses_by_codeloc: Dict[angr.code_location.CodeLocation, Set[angr.knowledge_plugins.key_definitions.definition.Definition]]

tmp_uses: Dict[int, Set[angr.code_location.CodeLocation]]

copy() → angr.knowledge_plugins.key_definitions.live_definitions.LiveDefinitions

static top(bits: int)

Get a TOP value.

Parameters: bits – Width of the TOP value (in bits).
Returns: The TOP value.

static is_top(expr) → bool

Check if the given expression is a TOP value.

Parameters: expr – The given expression.
Returns: True if the expression is TOP, False otherwise.

stack_address(offset: int) → claripy.ast.base.Base

static is_stack_address(addr: claripy.ast.base.Base) → bool

static get_stack_offset(addr: claripy.ast.base.Base) → Optional[int]

static annotate_with_def(symvar: claripy.ast.base.Base, definition: angr.knowledge_plugins.key_definitions.definition.Definition)

Parameters

symvar –
definition –

Returns

static extract_defs(symvar: claripy.ast.base.Base) → Generator[angr.knowledge_plugins.key_definitions.definition.Definition, None, None]

get_sp() → int: Return the concrete value contained by the stack pointer.

get_stack_address(offset: claripy.ast.base.Base) → Optional[int]

stack_offset_to_stack_addr(offset) → int

merge(*others) → Tuple[angr.knowledge_plugins.key_definitions.live_definitions.LiveDefinitions, bool]

kill_definitions(atom: angr.knowledge_plugins.key_definitions.atoms.Atom) → None

Overwrite existing definitions w.r.t ‘atom’ with a dummy definition instance. A dummy definition will not be removed during simplification.

Parameters: atom –
Returns: None

kill_and_add_definition(atom: angr.knowledge_plugins.key_definitions.atoms.Atom, code_loc: angr.code_location.CodeLocation, data: angr.storage.memory_mixins.paged_memory.pages.multi_values.MultiValues, dummy=False, tags: Optional[Set[angr.knowledge_plugins.key_definitions.tag.Tag]] = None, endness=None, annotated=False) → Optional[angr.storage.memory_mixins.paged_memory.pages.multi_values.MultiValues]

add_use(atom: angr.knowledge_plugins.key_definitions.atoms.Atom, code_loc: angr.code_location.CodeLocation) → None

add_use_by_def(definition: angr.knowledge_plugins.key_definitions.definition.Definition, code_loc: angr.code_location.CodeLocation) → None

get_definitions(atom: angr.knowledge_plugins.key_definitions.atoms.Atom) → Iterable[angr.knowledge_plugins.key_definitions.definition.Definition]

get_definitions_from_atoms(atoms: Iterable[angr.knowledge_plugins.key_definitions.atoms.Atom]) → Iterable[angr.knowledge_plugins.key_definitions.definition.Definition]

get_value_from_definition(definition: angr.knowledge_plugins.key_definitions.definition.Definition) → angr.storage.memory_mixins.paged_memory.pages.multi_values.MultiValues

get_value_from_atom(atom: angr.knowledge_plugins.key_definitions.atoms.Atom) → Optional[angr.storage.memory_mixins.paged_memory.pages.multi_values.MultiValues]

class angr.knowledge_plugins.key_definitions.rd_model.ReachingDefinitionsModel(func_addr: Optional[int] = None)

Bases: object

copy() → angr.knowledge_plugins.key_definitions.rd_model.ReachingDefinitionsModel

Classes to structure the different types of <Tag>s that can be attached to <Definition>s.

Tag
- FunctionTag
  
  ParameterTag
  
  LocalVariableTag
  
  ReturnValueTag
- InitialValueTag

class angr.knowledge_plugins.key_definitions.tag.Tag(metadata: object = None)

Bases: object

A tag for a Definition that can carry different kinds of metadata.

class angr.knowledge_plugins.key_definitions.tag.FunctionTag(function: Optional[int] = None, metadata: object = None)

Bases: angr.knowledge_plugins.key_definitions.tag.Tag

A tag for a definition created (or used) in the context of a function.

class angr.knowledge_plugins.key_definitions.tag.SideEffectTag(function: Optional[int] = None, metadata: object = None)

Bases: angr.knowledge_plugins.key_definitions.tag.FunctionTag

A tag for a definition created or used as a side-effect of a function.

Example: The <MemoryLocation> pointed by rdi during a sprintf.

class angr.knowledge_plugins.key_definitions.tag.ParameterTag(function: Optional[int] = None, metadata: object = None)

Bases: angr.knowledge_plugins.key_definitions.tag.FunctionTag

A tag for a definition of a parameter.

class angr.knowledge_plugins.key_definitions.tag.LocalVariableTag(function: Optional[int] = None, metadata: object = None)

Bases: angr.knowledge_plugins.key_definitions.tag.FunctionTag

A tag for a definition of a local variable of a function.

class angr.knowledge_plugins.key_definitions.tag.ReturnValueTag(function: Optional[int] = None, metadata: object = None)

Bases: angr.knowledge_plugins.key_definitions.tag.FunctionTag

A tag for a definiton of a return value of a function.

class angr.knowledge_plugins.key_definitions.tag.InitialValueTag(metadata: object = None)

Bases: angr.knowledge_plugins.key_definitions.tag.Tag

A tag for a definiton of an initial value

class angr.knowledge_plugins.key_definitions.tag.UnknownSizeTag(metadata: object = None)

Bases: angr.knowledge_plugins.key_definitions.tag.Tag

A tag for a definiton of an initial value

class angr.knowledge_plugins.key_definitions.undefined.Undefined

Bases: object

A TOP-like value indicating an unknown data source. Should live next to raw integers in DataSets.

class angr.knowledge_plugins.key_definitions.unknown_size.UnknownSize

Bases: object

A value indicating an unknown size for elements of DataSets. Should “behave” like an integer.

class angr.knowledge_plugins.key_definitions.uses.Uses

Bases: object

add_use(definition: Definition, codeloc: angr.code_location.CodeLocation)

Add a use for a given definition.

Parameters

definition (angr.analyses.reaching_definitions.definition.Definition) – The definition that is used.
codeloc – The code location where the use occurs.

get_uses(definition: Definition) → Set[angr.code_location.CodeLocation]

Retrieve the uses of a given definition.

Parameters: definition – The definition for which we get the uses.

remove_use(definition: Definition, codeloc: CodeLocation) → None

Remove one use of a given definition.

Parameters

definition – The definition of which to remove the uses.
codeloc – The code location where the use is.

Returns

None

remove_uses(definition: Definition)

Remove all uses of a given definition.

Parameters: definition – The definition of which to remove the uses.
Returns: None

get_uses_by_location(codeloc: angr.code_location.CodeLocation) → Set

Retrieve all definitions that are used at a given location.

Parameters: codeloc – The code location.
Returns: A set of definitions that are used at the given location.

copy()

Copy the instance.

Return angr.angr.analyses.reaching_definitions.uses.Uses: Return a new <Uses> instance containing the same data.

merge(other) → bool

Merge an instance of <Uses> into the current instance.

Parameters: other (angr.angr.analyses.reaching_definitions.uses.Uses) – The other <Uses> from which the data will be added to the current instance.
Returns: True if any merge occurred, False otherwise

angr.knowledge_plugins.sync.sync_controller.import_binsync()

angr.knowledge_plugins.sync.sync_controller.make_state(f): Build a writeable State instance and pass to f as the state kwarg if the state kwarg is None. Function f should have have at least two kwargs, user and state.

angr.knowledge_plugins.sync.sync_controller.make_ro_state(f): Build a read-only State instance and pass to f as the state kwarg if the state kwarg is None. Function f should have have at least two kwargs, user and state.

angr.knowledge_plugins.sync.sync_controller.init_checker(f)

class angr.knowledge_plugins.sync.sync_controller.SyncController(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin

SyncController interfaces with a binsync client to push changes upwards and pull changes downwards.

Variables: client (binsync.Client) – The binsync client.

connect(user, path, bin_hash='', init_repo=False, ssh_agent_pid=None, ssh_auth_sock=None, remote_url=None)

property connected

commit()

update()

copy()

pull()

property has_remote

users()

status()

tally(users=None)

push_function(func: angr.knowledge_plugins.functions.function.Function, user=None, state=None)

Push a function upwards.

Parameters: func (Function) – The angr Function object to push upwards.
Returns: True if updates are made. False otherwise.
Return type: bool

push_comment(addr, comment, decompiled=False, user=None, state=None)

push_comments(comments: List[binsync.data.Comment], user=None, state=None)

Push a bunch of comments upwards.

Parameters: comments (list) – A list of BinSync Comments
Returns: bool

push_stack_variables(stack_variables: List[angr.sim_variable.SimStackVariable], var_manager: angr.knowledge_plugins.variables.variable_manager.VariableManagerInternal, user=None, state=None)

Parameters

stack_variables –
var_manager –

Returns

push_stack_variable(func_addr, offset, name, type_, size_, user=None, state=None)

pull_function(addr, user=None, state=None) → Optional[binsync.data.Function]

Pull a function downwards.

Parameters

addr (int) – Address of the function.
user (str) – Name of the user.

Returns

The binsync.data.Function object if pulling succeeds, or None if pulling fails.

pull_comment(addr, user=None, state=None) → Optional[binsync.data.Comment]

Pull a comment downwards.

Parameters

addr (int) – Address of the comment.
user (str) – Name of the user.

Returns

a Comment object from BinSync, or None

pull_comments(func_addr, user=None, state=None)

Pull comments downwards.

Parameters

start_addr (int) – Where we want to pull comments.
end_addr (int) – Where we want to stop pulling comments (exclusive).

Returns

An iterator.

Return type

Iterable

pull_patches(user=None, state=None)

Pull patches.

Parameters: user (str) – Name of the user to patches from.
Returns: An iterator
Return type: Iterable

pull_stack_variables(func_addr, user=None, state=None)

Pull stack variables from a function.

@param func_addr: Function address to pull from @param user: @param state: @return:

get_func_addr_from_addr(addr)

class angr.knowledge_plugins.xrefs.xref.XRef(ins_addr=None, block_addr=None, stmt_idx=None, insn_op_idx=None, memory_data=None, dst: Optional[int] = None, xref_type=None)

Bases: angr.serializable.Serializable

XRef describes a reference to a MemoryData instance (if a MemoryData instance is available) or just an address.

ins_addr

insn_op_idx

block_addr

stmt_idx

memory_data

type

dst

property type_string

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, bits=None, **kwargs)

copy()

insn_op_type

class angr.knowledge_plugins.xrefs.xref_types.XRefType

Bases: object

Offset = 0

Read = 1

Write = 2

static to_string(ty)

class angr.knowledge_plugins.xrefs.xref_manager.XRefManager(kb)

Bases: angr.knowledge_plugins.plugin.KnowledgeBasePlugin, angr.serializable.Serializable

copy()

add_xref(xref)

add_xrefs(xrefs)

get_xrefs_by_ins_addr(ins_addr)

get_xrefs_by_dst(dst)

get_xrefs_by_dst_region(start, end): Get a set of XRef objects that point to a given address region bounded by start and end. Will only return absolute xrefs, not relative ones (like SP offsets)

get_xrefs_by_ins_addr_region(start, end) → Set[angr.knowledge_plugins.xrefs.xref.XRef]: Get a set of XRef objects that originate at a given address region bounded by start and end. Useful for finding references from a basic block or function.

serialize_to_cmessage()

classmethod parse_from_cmessage(cmsg, cfg_model=None, kb=None, **kwargs)

class angr.code_location.CodeLocation(block_addr: int, stmt_idx: Optional[int], sim_procedure=None, ins_addr: Optional[int] = None, context: Optional[Tuple[int]] = None, block_idx: Optional[int] = None, **kwargs)

Bases: object

Stands for a specific program point by specifying basic block address and statement ID (for IRSBs), or SimProcedure name (for SimProcedures).

Constructor.

Parameters

block_addr – Address of the block
stmt_idx – Statement ID. None for SimProcedures
sim_procedure (class) – The corresponding SimProcedure class.
ins_addr – The instruction address.
context – A tuple that represents the context of this CodeLocation in contextful mode, or None in contextless mode.
kwargs – Optional arguments, will be stored, but not used in __eq__ or __hash__.

block_addr: int

stmt_idx: int

sim_procedure

ins_addr: Optional[int]

context: Optional[Tuple[int]]

block_idx

info: Optional[Dict]

property short_repr

class angr.keyed_region.StoredObject(start, obj, size)

Bases: object

start

obj

size: Union[UnknownSize, int]

property obj_id

class angr.keyed_region.RegionObject(start, size, objects=None)

Bases: object

Represents one or more objects occupying one or more bytes in KeyedRegion.

start

size

stored_objects

property is_empty

property end

property internal_objects

includes(offset)

split(split_at)

add_object(obj)

set_object(obj)

copy()

class angr.keyed_region.KeyedRegion(tree=None, phi_node_contains=None, canonical_size=8)

Bases: object

KeyedRegion keeps a mapping between stack offsets and all objects covering that offset. It assumes no variable in this region overlap with another variable in this region.

Registers and function frames can all be viewed as a keyed region.

copy()

merge(other, replacements=None)

Merge another KeyedRegion into this KeyedRegion.

Parameters: other (KeyedRegion) – The other instance to merge with.
Returns: None

merge_to_top(other, replacements=None, top=None)

Merge another KeyedRegion into this KeyedRegion, but mark all variables with different values as TOP.

Parameters

other – The other instance to merge with.
replacements –

Returns

self

replace(replacements)

Replace variables with other variables.

Parameters: replacements (dict) – A dict of variable replacements.
Returns: self

dbg_repr(): Get a debugging representation of this keyed region. :return: A string of debugging output.

add_variable(start, variable)

Add a variable to this region at the given offset.

Parameters

start (int) –
variable (SimVariable) –

Returns

None

add_object(start, obj, object_size)

Add/Store an object to this region at the given offset.

Parameters

start –
obj –
object_size (int) – Size of the object

Returns

set_variable(start, variable)

Add a variable to this region at the given offset, and remove all other variables that are fully covered by this variable.

Parameters

start (int) –
variable (SimVariable) –

Returns

None

set_object(start, obj, object_size)

Add an object to this region at the given offset, and remove all other objects that are fully covered by this object.

Parameters

start –
obj –
object_size –

Returns

get_base_addr(addr)

Get the base offset (the key we are using to index objects covering the given offset) of a specific offset.

Parameters: addr (int) –
Returns
Return type: int or None

get_variables_by_offset(start)

Find variables covering the given region offset.

Parameters: start (int) –
Returns: A set of variables.
Return type: set

get_objects_by_offset(start)

Find objects covering the given region offset.

Parameters: start –
Returns

get_all_variables()

Get all variables covering the current region.

Returns: A set of all variables.

Serialization

class angr.serializable.Serializable

Bases: object

The base class of all protobuf-serializable classes in angr.

serialize_to_cmessage()

Serialize the class object and returns a protobuf cmessage object.

Returns: A protobuf cmessage object.
Return type: protobuf.cmessage

serialize()

Serialize the class object and returns a bytes object.

Returns: A bytes object.
Return type: bytes

classmethod parse_from_cmessage(cmsg, **kwargs)

Parse a protobuf cmessage and create a class object.

Parameters: cmsg – The probobuf cmessage object.
Returns: A unserialized class object.
Return type: cls

classmethod parse(s, **kwargs)

Parse a bytes object and create a class object.

Parameters: s (bytes) – A bytes object.
Returns: A class object.
Return type: cls

class angr.vaults.VaultPickler(vault, file, *args, assigned_objects=(), **kwargs)

Bases: _pickle.Pickler

A persistence-aware pickler. It will check for persistence of any objects except for those with IDs in ‘assigned_objects’.

persistent_id(obj)

class angr.vaults.VaultUnpickler(vault, file, *args, **kwargs)

Bases: _pickle.Unpickler

persistent_load(pid)

class angr.vaults.Vault

Bases: collections.abc.MutableMapping

The vault is a serializer for angr.

keys(): Should return the IDs stored by the vault.

is_stored(i): Checks if the provided id is already in the vault.

load(oid)

store(o)

dumps(o)

Returns a serialized string representing the object, post-deduplication.

Parameters: o – the object

loads(s)

Deserializes a string representation of the object.

Parameters: s – the string

static close()

class angr.vaults.VaultDict(d=None)

Bases: angr.vaults.Vault

A Vault that uses a dictionary for storage.

is_stored(i)

keys()

class angr.vaults.VaultDir(d=None)

Bases: angr.vaults.Vault

A Vault that uses a directory for storage.

keys()

class angr.vaults.VaultShelf(path=None)

Bases: angr.vaults.VaultDict

A Vault that uses a shelve.Shelf for storage.

close()

class angr.vaults.VaultDirShelf(d=None)

Bases: angr.vaults.VaultDict

A Vault that uses a directory for storage, where every object is stored into a single shelve.Shelf instance. VaultDir creates a file for each object. VaultDirShelf creates only one file for a stored object and everything else it references.

store(o)

load(oid)

keys()

Analysis

angr.analyses.register_analysis(cls, name)

class angr.analyses.analysis.AnalysisLogEntry(message, exc_info=False): Bases: object

class angr.analyses.analysis.AnalysesHub(project)

Bases: angr.misc.plugins.PluginVendor

This class contains functions for all the registered and runnable analyses,

reload_analyses(**kwargs)

class angr.analyses.analysis.AnalysisFactory(project: Project, analysis_cls: Type[A])

Bases: Generic[A]

prep(fail_fast=False, kb: Optional[KnowledgeBase] = None, progress_callback: Optional[Callable] = None, show_progressbar: bool = False) → Type[A]

class angr.analyses.analysis.StatusBar(width: Optional[int] = 40)

Bases: progressbar.widgets.WidgetBase

Implements a progressbar component for displaying raw text.

class angr.analyses.analysis.Analysis

Bases: object

This class represents an analysis on the program.

Variables

project – The project for this analysis.
kb (KnowledgeBase) – The knowledgebase object.
_progress_callback – A callback function for receiving the progress of this analysis. It only takes one argument, which is a float number from 0.0 to 100.0 indicating the current progress.
_show_progressbar (bool) – If a progressbar should be shown during the analysis. It’s independent from _progress_callback.
_progressbar (progressbar.ProgressBar) – The progress bar object.

project: Project = None

kb: KnowledgeBase = None

errors = []

named_errors = {}

class angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis(order_jobs=False, allow_merging=False, allow_widening=False, status_callback=None, graph_visitor: Optional[GraphVisitor[NodeType]] = None)

Bases: Generic[AnalysisState, NodeType]

This is my very first attempt to build a static forward analysis framework that can serve as the base of multiple static analyses in angr, including CFG analysis, VFG analysis, DDG, etc.

In short, ForwardAnalysis performs a forward data-flow analysis by traversing a graph, compute on abstract values, and store results in abstract states. The user can specify what graph to traverse, how a graph should be traversed, how abstract values and abstract states are defined, etc.

ForwardAnalysis has a few options to toggle, making it suitable to be the base class of several different styles of forward data-flow analysis implementations.

ForwardAnalysis supports a special mode when no graph is available for traversal (for example, when a CFG is being initialized and constructed, no other graph can be used). In that case, the graph traversal functionality is disabled, and the optimal graph traversal order is not guaranteed. The user can provide a job sorting method to sort the jobs in queue and optimize traversal order.

Feel free to discuss with me (Fish) if you have any suggestions or complaints.

Constructor

Parameters

order_jobs (bool) – If all jobs should be ordered or not.
allow_merging (bool) – If job merging is allowed.
allow_widening (bool) – If job widening is allowed.
graph_visitor (GraphVisitor or None) – A graph visitor to provide successors.

Returns

None

property should_abort: Should the analysis be terminated. :return: True/False

property graph: networkx.classes.digraph.DiGraph

property jobs

abort(): Abort the analysis :return: None

has_job(job: angr.analyses.cfg.cfg_job_base.CFGJobBase) → bool

Checks whether there exists another job which has the same job key. :param job: The job to check.

Returns: True if there exists another job with the same key, False otherwise.

downsize()

class angr.analyses.forward_analysis.job_info.JobInfo(key: angr.analyses.cfg.cfg_job_base.BlockID, job: angr.analyses.cfg.cfg_job_base.CFGJobBase)

Bases: object

Stores information of each job.

property job: angr.analyses.cfg.cfg_job_base.CFGJobBase

Get the latest available job.

Returns: The latest available job.

property merged_jobs

property widened_jobs

add_job(job, merged=False, widened=False): Appended a new job to this JobInfo node. :param job: The new job to append. :param bool merged: Whether it is a merged job or not. :param bool widened: Whether it is a widened job or not.

class angr.analyses.forward_analysis.visitors.call_graph.CallGraphVisitor(callgraph)

Bases: angr.analyses.forward_analysis.visitors.graph.GraphVisitor

Parameters: callgraph (networkx.DiGraph) –

successors(node)

predecessors(node)

sort_nodes(nodes=None)

class angr.analyses.forward_analysis.visitors.function_graph.FunctionGraphVisitor(func, graph=None)

Bases: angr.analyses.forward_analysis.visitors.graph.GraphVisitor

Parameters: func (knowledge.Function) –

successors(node)

predecessors(node)

sort_nodes(nodes=None)

class angr.analyses.forward_analysis.visitors.graph.GraphVisitor

Bases: Generic[NodeType]

A graph visitor takes a node in the graph and returns its successors. Typically it visits a control flow graph, and returns successors of a CFGNode each time. This is the base class of all graph visitors.

successors(node: NodeType) → List[NodeType]

Get successors of a node. The node should be in the graph.

Parameters: node – The node to work with.
Returns: A list of successors.
Return type: list

predecessors(node: NodeType) → List[NodeType]

Get predecessors of a node. The node should be in the graph.

Parameters: node – The node to work with.
Returns: A list of predecessors.
Return type: list

sort_nodes(nodes: Optional[Collection[NodeType]] = None) → List[NodeType]

Get a list of all nodes sorted in an optimal traversal order.

Parameters: nodes (iterable) – A collection of nodes to sort. If none, all nodes in the graph will be used to sort.
Returns: A list of sorted nodes.
Return type: list

nodes() → Iterator[NodeType]

Return an iterator of nodes following an optimal traversal order.

Returns

nodes_iter(**kwargs)

reset()

Reset the internal node traversal state. Must be called prior to visiting future nodes.

Returns: None

next_node() → Optional[NodeType]

Get the next node to visit.

Returns: A node in the graph.

all_successors(node: NodeType, skip_reached_fixedpoint=False) → Set[NodeType]

Returns all successors to the specific node.

Parameters: node – A node in the graph.
Returns: A set of nodes that are all successors to the given node.
Return type: set

revisit_successors(node: NodeType, include_self=True) → None

Revisit a node in the future. As a result, the successors to this node will be revisited as well.

Parameters: node – The node to revisit in the future.
Returns: None

revisit_node(node: NodeType) → None

Revisit a node in the future. Do not include its successors immediately.

Parameters: node – The node to revisit in the future.
Returns: None

reached_fixedpoint(node: NodeType) → None

Mark a node as reached fixed-point. This node as well as all its successors will not be visited in the future.

Parameters: node – The node to mark as reached fixed-point.
Returns: None

class angr.analyses.forward_analysis.visitors.loop.LoopVisitor(loop)

Bases: angr.analyses.forward_analysis.visitors.graph.GraphVisitor

Parameters: loop (angr.analyses.loopfinder.Loop) – The loop to visit.

successors(node)

predecessors(node)

sort_nodes(nodes=None)

class angr.analyses.forward_analysis.visitors.single_node_graph.SingleNodeGraphVisitor(node)

Bases: angr.analyses.forward_analysis.visitors.graph.GraphVisitor

Parameters: node – The single node that should be in the graph.

node

node_returned

reset()

next_node()

successors(node)

predecessors(node)

sort_nodes(nodes=None)

class angr.analyses.backward_slice.BackwardSlice(cfg, cdg, ddg, targets=None, cfg_node=None, stmt_id=None, control_flow_slice=False, same_function=False, no_construct=False)

Bases: angr.analyses.analysis.Analysis

Represents a backward slice of the program.

Create a backward slice from a specific statement based on provided control flow graph (CFG), control dependence graph (CDG), and data dependence graph (DDG).

The data dependence graph can be either CFG-based, or Value-set analysis based. A CFG-based DDG is much faster to generate, but it only reflects those states while generating the CFG, and it is neither sound nor accurate. The VSA based DDG (called VSA_DDG) is based on static analysis, which gives you a much better result.

Parameters

cfg – The control flow graph.
cdg – The control dependence graph.
ddg – The data dependence graph.
targets – A list of “target” that specify targets of the backward slices. Each target can be a tuple in form of (cfg_node, stmt_idx), or a CodeLocation instance.
cfg_node – Deprecated. The target CFGNode to reach. It should exist in the CFG.
stmt_id – Deprecated. The target statement to reach.
control_flow_slice – True/False, indicates whether we should slice only based on CFG. Sometimes when acquiring DDG is difficult or impossible, you can just create a slice on your CFG. Well, if you don’t even have a CFG, then…
no_construct – Only used for testing and debugging to easily create a BackwardSlice object.

dbg_repr(max_display=10)

Debugging output of this slice.

Parameters: max_display – The maximum number of SimRun slices to show.
Returns: A string representation.

dbg_repr_run(run_addr)

Debugging output of a single SimRun slice.

Parameters: run_addr – Address of the SimRun.
Returns: A string representation.

annotated_cfg(start_point=None): Returns an AnnotatedCFG based on slicing result.

is_taint_related_to_ip(simrun_addr, stmt_idx, taint_type, simrun_whitelist=None)

Query in taint graph to check if a specific taint will taint the IP in the future or not. The taint is specified with the tuple (simrun_addr, stmt_idx, taint_type).

Parameters

simrun_addr – Address of the SimRun.
stmt_idx – Statement ID.
taint_type – Type of the taint, might be one of the following: ‘reg’, ‘tmp’, ‘mem’.
simrun_whitelist – A list of SimRun addresses that are whitelisted, i.e. the tainted exit will be ignored if it is in those SimRuns.

Returns

True/False

is_taint_impacting_stack_pointers(simrun_addr, stmt_idx, taint_type, simrun_whitelist=None)

Query in taint graph to check if a specific taint will taint the stack pointer in the future or not. The taint is specified with the tuple (simrun_addr, stmt_idx, taint_type).

Parameters

simrun_addr – Address of the SimRun.
stmt_idx – Statement ID.
taint_type – Type of the taint, might be one of the following: ‘reg’, ‘tmp’, ‘mem’.
simrun_whitelist – A list of SimRun addresses that are whitelisted.

Returns

True/False.

exception angr.analyses.bindiff.UnmatchedStatementsException: Bases: Exception

class angr.analyses.bindiff.Difference(diff_type, value_a, value_b): Bases: object

class angr.analyses.bindiff.ConstantChange(offset, value_a, value_b): Bases: object

angr.analyses.bindiff.differing_constants(block_a, block_b)

Compares two basic blocks and finds all the constants that differ from the first block to the second.

Parameters

block_a – The first block to compare.
block_b – The second block to compare.

Returns

Returns a list of differing constants in the form of ConstantChange, which has the offset in the block and the respective constants.

angr.analyses.bindiff.compare_statement_dict(statement_1, statement_2)

class angr.analyses.bindiff.NormalizedBlock(block, function): Bases: object

class angr.analyses.bindiff.NormalizedFunction(function: Function): Bases: object

class angr.analyses.bindiff.FunctionDiff(function_a: Function, function_b: Function, bindiff=None)

Bases: object

This class computes the a diff between two functions.

Parameters

function_a – The first angr Function object to diff.
function_b – The second angr Function object.
bindiff – An optional Bindiff object. Used for some extra normalization during basic block comparison.

property probably_identical

Whether or not these two functions are identical.

Type: returns

property identical_blocks

A list of block matches which appear to be identical

Type: returns

property differing_blocks

A list of block matches which appear to differ

Type: returns

property blocks_with_differing_constants

A list of block matches which appear to differ

Type: return

property block_matches

property unmatched_blocks

static get_normalized_block(addr, function)

Parameters

addr – Where to start the normalized block.
function – A function containing the block address.

Returns

A normalized basic block.

block_similarity(block_a, block_b)

Parameters

block_a – The first block address.
block_b – The second block address.

Returns

The similarity of the basic blocks, normalized for the base address of the block and function call addresses.

blocks_probably_identical(block_a, block_b, check_constants=False)

Parameters

block_a – The first block address.
block_b – The second block address.
check_constants – Whether or not to require matching constants in blocks.

Returns

Whether or not the blocks appear to be identical.

class angr.analyses.bindiff.BinDiff(other_project, enable_advanced_backward_slicing=False, cfg_a=None, cfg_b=None)

Bases: angr.analyses.analysis.Analysis

This class computes the a diff between two binaries represented by angr Projects

Parameters: other_project – The second project to diff

functions_probably_identical(func_a_addr, func_b_addr, check_consts=False)

Compare two functions and return True if they appear identical.

Parameters

func_a_addr – The address of the first function (in the first binary).
func_b_addr – The address of the second function (in the second binary).

Returns

Whether or not the functions appear to be identical.

property identical_functions

A list of function matches that appear to be identical

Type: returns

property differing_functions

A list of function matches that appear to differ

Type: returns

differing_functions_with_consts()

Returns: A list of function matches that appear to differ including just by constants

property differing_blocks

A list of block matches that appear to differ

Type: returns

property identical_blocks: return A list of all block matches that appear to be identical

property blocks_with_differing_constants

A dict of block matches with differing constants to the tuple of constants

Type: return

property unmatched_functions

get_function_diff(function_addr_a, function_addr_b)

Parameters

function_addr_a – The address of the first function (in the first binary)
function_addr_b – The address of the second function (in the second binary)

Returns

the FunctionDiff of the two functions

class angr.analyses.boyscout.BoyScout(cookiesize=1)

Bases: angr.analyses.analysis.Analysis

Try to determine the architecture and endieness of a binary blob

class angr.analyses.calling_convention.CallSiteFact(return_value_used)

Bases: object

Store facts about each call site.

class angr.analyses.calling_convention.UpdateArgumentsOption

Bases: object

Enums for controlling the argument updating behavior in _adjust_cc.

DoNotUpdate = 0

AlwaysUpdate = 1

UpdateWhenCCHasNoArgs = 2

class angr.analyses.calling_convention.DummyFunctionHandler

Bases: angr.analyses.reaching_definitions.function_handler.FunctionHandler

A function handler that is used during reaching definition analysis.

handle_local_function(state: ReachingDefinitionsState, function_address: int, call_stack: Optional[List], maximum_local_call_depth: int, visited_blocks: Set[int], dep_graph: DepGraph, src_ins_addr: Optional[int] = None, codeloc: Optional[CodeLocation] = None) → Tuple[bool, ReachingDefinitionsState, Set[int], DepGraph]

class angr.analyses.calling_convention.CallingConventionAnalysis(func: Function, cfg: Optional[CFGModel] = None, analyze_callsites: bool = False)

Bases: angr.analyses.analysis.Analysis

Analyze the calling convention of a function and guess a probable prototype.

The calling convention of a function can be inferred at both its call sites and the function itself. At call sites, we consider all register and stack variables that are not alive after the function call as parameters to this function. In the function itself, we consider all register and stack variables that are read but without initialization as parameters. Then we synthesize the information from both locations and make a reasonable inference of calling convention of this function.

Variables

_function – The function to recover calling convention for.
_variable_manager – A handy accessor to the variable manager.
_cfg – A reference of the CFGModel of the current binary. It is used to discover call sites of the current function in order to perform analysis at call sites.
analyze_callsites – True if we should analyze all call sites of the current function to determine the calling convention and arguments. This can be time-consuming if there are many call sites to analyze.
cc – The recovered calling convention for the function.

class angr.analyses.complete_calling_conventions.CompleteCallingConventionsAnalysis(recover_variables=False, low_priority=False, force=False, cfg: Optional[angr.knowledge_plugins.cfg.cfg_model.CFGModel] = None, analyze_callsites: bool = False, skip_signature_matched_functions: bool = False, max_function_blocks: Optional[int] = None, max_function_size: Optional[int] = None, workers: int = 0, cc_callback: Optional[Callable] = None, prioritize_func_addrs: Optional[Iterable[int]] = None, auto_start: bool = True)

Bases: angr.analyses.analysis.Analysis

Implements full-binary calling convention analysis. During the initial analysis of a binary, you may set recover_variables to True so that it will perform variable recovery on each function before performing calling convention analysis.

Parameters

recover_variables – Recover variables on each function before performing calling convention analysis.
low_priority – Run in the background - periodically release GIL.
force – Perform calling convention analysis on functions even if they have calling conventions or prototypes already specified (or previously recovered).
cfg – The control flow graph model, which will be passed to CallingConventionAnalysis.
analyze_callsites – Consider artifacts at call sites when performing calling convention analysis.
skip_signature_matched_functions – Do not perform calling convention analysis on functions that match against existing FLIRT signatures.
max_function_blocks – Do not perform calling convention analysis on functions with more than the specified number of blocks. Setting it to None disables this check.
max_function_size – Do not perform calling convention analysis on functions whose sizes are more than max_function_size. Setting it to None disables this check.
workers – Number of multiprocessing workers.

work()

prioritize_functions(func_addrs: Iterable[int])

Prioritize the analysis of specified functions.

Parameters: func_addrs – A collection of function addresses to analyze first.

static function_needs_variable_recovery(func)

Check if running variable recovery on the function is the only way to determine the calling convention of the this function.

We do not need to run variable recovery to determine the calling convention of a function if: - The function is a SimProcedure. - The function is a PLT stub. - The function is a library function and we already know its prototype.

Parameters: func – The function object.
Returns: True if we must run VariableRecovery before we can determine what the calling convention of this function is. False otherwise.
Return type: bool

exception angr.analyses.soot_class_hierarchy.SootClassHierarchyError(msg): Bases: Exception

exception angr.analyses.soot_class_hierarchy.NoConcreteDispatch(msg): Bases: angr.analyses.soot_class_hierarchy.SootClassHierarchyError

class angr.analyses.soot_class_hierarchy.SootClassHierarchy

Bases: angr.analyses.analysis.Analysis

Generate complete hierarchy.

init_hierarchy()

has_super_class(cls)

is_subclass_including(cls_child, cls_parent)

is_subclass(cls_child, cls_parent)

is_visible_method(cls, method)

is_visible_class(cls_from, cls_to)

get_super_classes(cls)

get_super_classes_including(cls)

get_implementers(interface)

get_sub_interfaces_including(interface)

get_sub_interfaces(interface)

get_sub_classes(cls)

get_sub_classes_including(cls)

resolve_abstract_dispatch(cls, method)

resolve_concrete_dispatch(cls, method)

resolve_special_dispatch(method, container)

resolve_invoke(invoke_expr, method, container)

class angr.analyses.cfg.cfb.CFBlanketView(cfb)

Bases: object

A view into the control-flow blanket.

class angr.analyses.cfg.cfb.MemoryRegion(addr, size, type_, object_, cle_region): Bases: object

class angr.analyses.cfg.cfb.Unknown(addr, size, bytes_=None, object_=None, segment=None, section=None): Bases: object

class angr.analyses.cfg.cfb.CFBlanket(exclude_region_types=None)

Bases: angr.analyses.analysis.Analysis

A Control-Flow Blanket is a representation for storing all instructions, data entries, and bytes of a full program.

Region types: - section - segment - extern - tls - kernel

property regions: Return all memory regions.

floor_addr(addr)

floor_item(addr)

floor_items(addr=None, reverse=False)

ceiling_addr(addr)

ceiling_item(addr)

ceiling_items(addr=None, reverse=False, include_first=True)

add_obj(addr, obj): Adds an object obj to the blanket at the specified address addr

add_function(func): Add a function func and all blocks of this function to the blanket.

dbg_repr()

The debugging representation of this CFBlanket.

Returns: The debugging representation of this CFBlanket.
Return type: str

exception angr.analyses.cfg.cfg.OutdatedError: Bases: Exception

class angr.analyses.cfg.cfg.CFG(**kwargs)

Bases: angr.analyses.cfg.cfg_fast.CFGFast

tl;dr: CFG is just a wrapper around CFGFast for compatibility issues. It will be fully replaced by CFGFast in future releases. Feel free to use CFG if you intend to use CFGFast. Please use CFGEmulated if you have to use the old, slow, dynamically-generated version of CFG.

For multiple historical reasons, angr’s CFG is accurate but slow, which does not meet what most people expect. We developed CFGFast for light-speed CFG recovery, and renamed the old CFG class to CFGEmulated. For compability concerns, CFG was kept as an alias to CFGEmulated.

However, so many new users of angr would load up a binary and generate a CFG immediately after running “pip install angr”, and draw the conclusion that “angr’s CFG is so slow - angr must be unusable!” Therefore, we made the hard decision: CFG will be an alias to CFGFast, instead of CFGEmulated.

To ease the transition of your existing code and script, the following changes are made:

A CFG class, which is a sub class of CFGFast, is created.
You will see both a warning message printed out to stderr and an exception raised by angr if you are passing CFG any parameter that only CFGEmulated supports. This exception is not a sub class of AngrError, so you wouldn’t capture it with your old code by mistake.
In the near future, this wrapper class will be removed completely, and CFG will be a simple alias to CFGFast.

We expect most interfaces are the same between CFGFast and CFGEmulated. Apparently some functionalities (like context-sensitivity, and state keeping) only exist in CFGEmulated, which is when you want to use CFGEmulated instead.

class angr.analyses.cfg.cfg_emulated.CFGJob(*args, **kwargs)

Bases: angr.analyses.cfg.cfg_job_base.CFGJobBase

property block_id

property is_syscall

class angr.analyses.cfg.cfg_emulated.PendingJob(caller_func_addr, returning_source, state, src_block_id, src_exit_stmt_idx, src_exit_ins_addr, call_stack)

Bases: object

A PendingJob is whatever will be put into our pending_exit list. A pending exit is an entry that created by the returning of a call or syscall. It is “pending” since we cannot immediately figure out whether this entry will be executed or not. If the corresponding call/syscall intentially doesn’t return, then the pending exit will be removed. If the corresponding call/syscall returns, then the pending exit will be removed as well (since a real entry is created from the returning and will be analyzed later). If the corresponding call/syscall might return, but for some reason (for example, an unsupported instruction is met during the analysis) our analysis does not return properly, then the pending exit will be picked up and put into remaining_jobs list.

Parameters

returning_source – Address of the callee function. It might be None if address of the callee is not resolvable.
state – The state after returning from the callee function. Of course there is no way to get a precise state without emulating the execution of the callee, but at least we can properly adjust the stack and registers to imitate the real returned state.
call_stack – A callstack.

class angr.analyses.cfg.cfg_emulated.CFGEmulated(context_sensitivity_level=1, start=None, avoid_runs=None, enable_function_hints=False, call_depth=None, call_tracing_filter=None, initial_state=None, starts=None, keep_state=False, indirect_jump_target_limit=100000, resolve_indirect_jumps=True, enable_advanced_backward_slicing=False, enable_symbolic_back_traversal=False, indirect_jump_resolvers=None, additional_edges=None, no_construct=False, normalize=False, max_iterations=1, address_whitelist=None, base_graph=None, iropt_level=None, max_steps=None, state_add_options=None, state_remove_options=None, model=None)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis, angr.analyses.cfg.cfg_base.CFGBase

This class represents a control-flow graph.

All parameters are optional.

Parameters

context_sensitivity_level – The level of context-sensitivity of this CFG (see documentation for further details). It ranges from 0 to infinity. Default 1.
avoid_runs – A list of runs to avoid.
enable_function_hints – Whether to use function hints (constants that might be used as exit targets) or not.
call_depth – How deep in the call stack to trace.
call_tracing_filter – Filter to apply on a given path and jumpkind to determine if it should be skipped when call_depth is reached.
initial_state – An initial state to use to begin analysis.
starts (iterable) – A collection of starting points to begin analysis. It can contain the following three different types of entries: an address specified as an integer, a 2-tuple that includes an integer address and a jumpkind, or a SimState instance. Unsupported entries in starts will lead to an AngrCFGError being raised.
keep_state – Whether to keep the SimStates for each CFGNode.
resolve_indirect_jumps – Whether to enable the indirect jump resolvers for resolving indirect jumps
enable_advanced_backward_slicing – Whether to enable an intensive technique for resolving indirect jumps
enable_symbolic_back_traversal – Whether to enable an intensive technique for resolving indirect jumps
indirect_jump_resolvers (list) – A custom list of indirect jump resolvers. If this list is None or empty, default indirect jump resolvers specific to this architecture and binary types will be loaded.
additional_edges – A dict mapping addresses of basic blocks to addresses of successors to manually include and analyze forward from.
no_construct (bool) – Skip the construction procedure. Only used in unit-testing.
normalize (bool) – If the CFG as well as all Function graphs should be normalized or not.
max_iterations (int) – The maximum number of iterations that each basic block should be “executed”. 1 by default. Larger numbers of iterations are usually required for complex analyses like loop analysis.
address_whitelist (iterable) – A list of allowed addresses. Any basic blocks outside of this collection of addresses will be ignored.
base_graph (networkx.DiGraph) – A basic control flow graph to follow. Each node inside this graph must have the following properties: addr and size. CFG recovery will strictly follow nodes and edges shown in the graph, and discard any contorl flow that does not follow an existing edge in the base graph. For example, you can pass in a Function local transition graph as the base graph, and CFGEmulated will traverse nodes and edges and extract useful information.
iropt_level (int) – The optimization level of VEX IR (0, 1, 2). The default level will be used if iropt_level is None.
max_steps (int) – The maximum number of basic blocks to recover forthe longest path from each start before pausing the recovery procedure.
state_add_options – State options that will be added to the initial state.
state_remove_options – State options that will be removed from the initial state.

tag: Optional[str] = 'CFGEmulated'

copy() → angr.analyses.cfg.cfg_emulated.CFGEmulated

Make a copy of the CFG.

Returns: A copy of the CFG instance.

resume(starts=None, max_steps=None)

Resume a paused or terminated control flow graph recovery.

Parameters

starts (iterable) – A collection of new starts to resume from. If starts is None, we will resume CFG recovery from where it was paused before.
max_steps (int) – The maximum number of blocks on the longest path starting from each start before pausing the recovery.

Returns

None

remove_cycles(): Forces graph to become acyclic, removes all loop back edges and edges between overlapped loop headers and their successors.

downsize()

Remove saved states from all CFGNodes to reduce memory usage.

Returns: None

unroll_loops(max_loop_unrolling_times)

Unroll loops for each function. The resulting CFG may still contain loops due to recursion, function calls, etc.

Parameters: max_loop_unrolling_times (int) – The maximum iterations of unrolling.
Returns: None

force_unroll_loops(max_loop_unrolling_times)

Unroll loops globally. The resulting CFG does not contain any loop, but this method is slow on large graphs.

Parameters: max_loop_unrolling_times (int) – The maximum iterations of unrolling.
Returns: None

immediate_dominators(start, target_graph=None)

Get all immediate dominators of sub graph from given node upwards.

Parameters

start (str) – id of the node to navigate forwards from.
target_graph (networkx.classes.digraph.DiGraph) – graph to analyse, default is self.graph.

Returns

each node of graph as index values, with element as respective node’s immediate dominator.

Return type

dict

immediate_postdominators(end, target_graph=None)

Get all immediate postdominators of sub graph from given node upwards.

Parameters

start (str) – id of the node to navigate forwards from.
target_graph (networkx.classes.digraph.DiGraph) – graph to analyse, default is self.graph.

Returns

each node of graph as index values, with element as respective node’s immediate dominator.

Return type

dict

remove_fakerets()

Get rid of fake returns (i.e., Ijk_FakeRet edges) from this CFG

Returns: None

get_topological_order(cfg_node)

Get the topological order of a CFG Node.

Parameters: cfg_node – A CFGNode instance.
Returns: An integer representing its order, or None if the CFGNode does not exist in the graph.

get_subgraph(starting_node, block_addresses)

Get a sub-graph out of a bunch of basic block addresses.

Parameters

starting_node (CFGNode) – The beginning of the subgraph
block_addresses (iterable) – A collection of block addresses that should be included in the subgraph if there is a path between starting_node and a CFGNode with the specified address, and all nodes on the path should also be included in the subgraph.

Returns

A new CFG that only contain the specific subgraph.

Return type

CFGEmulated

get_function_subgraph(start, max_call_depth=None)

Get a sub-graph of a certain function.

Parameters

start – The function start. Currently it should be an integer.
max_call_depth – Call depth limit. None indicates no limit.

Returns

A CFG instance which is a sub-graph of self.graph

property context_sensitivity_level

property graph

property unresolvables

Get those SimRuns that have non-resolvable exits.

Returns: A set of SimRuns
Return type: set

property deadends

Get all CFGNodes that has an out-degree of 0

Returns: A list of CFGNode instances
Return type: list

class angr.analyses.cfg.cfg_base.CFGBase(sort, context_sensitivity_level, normalize=False, binary=None, force_segment=False, base_state=None, resolve_indirect_jumps=True, indirect_jump_resolvers=None, indirect_jump_target_limit=100000, detect_tail_calls=False, low_priority=False, sp_tracking_track_memory=True, model=None)

Bases: angr.analyses.analysis.Analysis

The base class for control flow graphs.

Parameters

sort (str) – ‘fast’ or ‘emulated’.
context_sensitivity_level (int) – The level of context-sensitivity of this CFG (see documentation for further details). It ranges from 0 to infinity.
normalize (bool) – Whether the CFG as well as all Function graphs should be normalized.
binary (cle.backends.Backend) – The binary to recover CFG on. By default the main binary is used.
force_segment (bool) – Force CFGFast to rely on binary segments instead of sections.
base_state (angr.SimState) – A state to use as a backer for all memory loads.
resolve_indirect_jumps (bool) – Whether to try to resolve indirect jumps. This is necessary to resolve jump targets from jump tables, etc.
indirect_jump_resolvers (list) – A custom list of indirect jump resolvers. If this list is None or empty, default indirect jump resolvers specific to this architecture and binary types will be loaded.
indirect_jump_target_limit (int) – Maximum indirect jump targets to be recovered.
detect_tail_calls (bool) – Aggressive tail-call optimization detection. This option is only respected in make_functions().
sp_tracking_track_memory (bool) – Whether or not to track memory writes if tracking the stack pointer. This increases the accuracy of stack pointer tracking, especially for architectures without a base pointer. Only used if detect_tail_calls is enabled.
model (None or CFGModel) – The CFGModel instance to write to. A new CFGModel instance will be created and registered with the knowledge base if model is None.

Returns

None

tag: Optional[str] = None

property model: Get the CFGModel instance. :return: The CFGModel instance that this analysis currently uses. :rtype: CFGModel

property normalized

property context_sensitivity_level

property functions

A reference to the FunctionManager in the current knowledge base.

Returns: FunctionManager with all functions
Return type: angr.knowledge_plugins.FunctionManager

make_copy(copy_to)

Copy self attributes to the new object.

Parameters: copy_to (CFGBase) – The target to copy to.
Returns: None

copy()

output()

generate_index()

Generate an index of all nodes in the graph in order to speed up get_any_node() with anyaddr=True.

Returns: None

get_predecessors(**kwargs)

get_successors(**kwargs)

get_successors_and_jumpkind(**kwargs)

get_all_predecessors(**kwargs)

get_all_successors(**kwargs)

get_node(**kwargs)

get_any_node(**kwargs)

get_all_nodes(**kwargs)

nodes(**kwargs)

nodes_iter(**kwargs)

get_loop_back_edges()

get_branching_nodes(**kwargs)

get_exit_stmt_idx(**kwargs)

property graph: networkx.classes.digraph.DiGraph

remove_edge(block_from, block_to)

is_thumb_addr(addr)

normalize()

Normalize the CFG, making sure that there are no overlapping basic blocks.

Note that this method will not alter transition graphs of each function in self.kb.functions. You may call normalize() on each Function object to normalize their transition graphs.

Returns: None

mark_function_alignments()

Find all potential function alignments and mark them.

Note that it is not always correct to simply remove them, because these functions may not be actual alignments but part of an actual function, and is incorrectly marked as an individual function because of failures in resolving indirect jumps. An example is in the test binary x86_64/dir_gcc_-O0 0x40541d (indirect jump at 0x4051b0). If the indirect jump cannot be correctly resolved, removing function 0x40541d will cause a missing label failure in reassembler.

Returns: None

make_functions()

Revisit the entire control flow graph, create Function instances accordingly, and correctly put blocks into each function.

Although Function objects are crated during the CFG recovery, they are neither sound nor accurate. With a pre-constructed CFG, this method rebuilds all functions bearing the following rules:

A block may only belong to one function.

Small functions lying inside the startpoint and the endpoint of another function will be merged with the other function

Tail call optimizations are detected.

PLT stubs are aligned by 16.

Returns: None

exception angr.analyses.cfg.cfg_fast.ContinueScanningNotification

Bases: RuntimeError

A notification raised by _next_code_addr_core() to indicate no code address is found and _next_code_addr_core() should be invoked again.

class angr.analyses.cfg.cfg_fast.ARMDecodingMode

Bases: object

Enums indicating decoding mode for ARM code.

ARM = 0

THUMB = 1

class angr.analyses.cfg.cfg_fast.DecodingAssumption(addr: int, size: int, mode: int)

Bases: object

Describes the decoding mode (ARM/THUMB) for a given basic block identified by its address.

add_data_seg(addr: int, size: int) → None

class angr.analyses.cfg.cfg_fast.FunctionReturn(callee_func_addr, caller_func_addr, call_site_addr, return_to)

Bases: object

FunctionReturn describes a function call in a specific location and its return location. Hashable and equatable

callee_func_addr

caller_func_addr

call_site_addr

return_to

class angr.analyses.cfg.cfg_fast.PendingJobs(functions, deregister_job_callback)

Bases: object

A collection of pending jobs during CFG recovery.

add_job(job)

pop_job(returning=True)

Pop a job from the pending jobs list.

When returning == True, we prioritize the jobs whose functions are known to be returning (function.returning is True). As an optimization, we are sorting the pending jobs list according to job.function.returning.

Parameters: returning (bool) – Only pop a pending job if the corresponding function returns.
Returns: A pending job if we can find one, or None if we cannot find any that satisfies the requirement.
Return type: angr.analyses.cfg.cfg_fast.CFGJob

cleanup()

Remove those pending exits if: a) they are the return exits of non-returning SimProcedures b) they are the return exits of non-returning syscalls b) they are the return exits of non-returning functions

Returns: None

add_returning_function(func_addr)

Mark a function as returning.

Parameters: func_addr (int) – Address of the function that returns.
Returns: None

add_nonreturning_function(func_addr)

Mark a function as not returning.

Parameters: func_addr (int) – Address of the function that does not return.
Returns: None

clear_updated_functions()

Clear the updated_functions set.

Returns: None

class angr.analyses.cfg.cfg_fast.FunctionEdge

Bases: object

Describes an edge in functions’ transition graphs. Base class for all types of edges.

apply(cfg)

src_func_addr

stmt_idx

ins_addr

class angr.analyses.cfg.cfg_fast.FunctionTransitionEdge(src_node, dst_addr, src_func_addr, to_outside=False, dst_func_addr=None, stmt_idx=None, ins_addr=None, is_exception=False)

Bases: angr.analyses.cfg.cfg_fast.FunctionEdge

Describes a transition edge in functions’ transition graphs.

src_node

dst_addr

src_func_addr

to_outside

dst_func_addr

is_exception

apply(cfg)

class angr.analyses.cfg.cfg_fast.FunctionCallEdge(src_node, dst_addr, ret_addr, src_func_addr, syscall=False, stmt_idx=None, ins_addr=None)

Bases: angr.analyses.cfg.cfg_fast.FunctionEdge

Describes a call edge in functions’ transition graphs.

src_node

dst_addr

ret_addr

syscall

apply(cfg)

class angr.analyses.cfg.cfg_fast.FunctionFakeRetEdge(src_node, dst_addr, src_func_addr, confirmed=None)

Bases: angr.analyses.cfg.cfg_fast.FunctionEdge

Describes a FakeReturn (also called fall-through) edge in functions’ transition graphs.

src_node

dst_addr

confirmed

apply(cfg)

class angr.analyses.cfg.cfg_fast.FunctionReturnEdge(ret_from_addr, ret_to_addr, dst_func_addr)

Bases: angr.analyses.cfg.cfg_fast.FunctionEdge

Describes a return (from a function call or a syscall) edge in functions’ transition graphs.

ret_from_addr

ret_to_addr

dst_func_addr

apply(cfg)

class angr.analyses.cfg.cfg_fast.CFGJob(addr: int, func_addr: int, jumpkind: str, ret_target: Optional[int] = None, last_addr: Optional[int] = None, src_node: Optional[angr.knowledge_plugins.cfg.cfg_node.CFGNode] = None, src_ins_addr: Optional[int] = None, src_stmt_idx: Optional[int] = None, returning_source=None, syscall: bool = False, func_edges: Optional[List] = None, job_type='Normal', gp: Optional[int] = None)

Bases: object

Defines a job to work on during the CFG recovery

JOB_TYPE_NORMAL = 'Normal'

JOB_TYPE_FUNCTION_PROLOGUE = 'Function-prologue'

JOB_TYPE_COMPLETE_SCANNING = 'Complete-scanning'

JOB_TYPE_IFUNC_HINTS = 'ifunc-hints'

addr

func_addr

jumpkind

ret_target

last_addr

src_node

src_ins_addr

src_stmt_idx

returning_source

syscall

job_type

gp

add_function_edge(edge)

apply_function_edges(cfg, clear=False)

class angr.analyses.cfg.cfg_fast.CFGFast(binary=None, objects=None, regions=None, pickle_intermediate_results=False, symbols=True, function_prologues=True, resolve_indirect_jumps=True, force_segment=False, force_smart_scan=True, force_complete_scan=False, indirect_jump_target_limit=100000, data_references=True, cross_references=False, normalize=False, start_at_entry=True, function_starts=None, extra_memory_regions=None, data_type_guessing_handlers=None, arch_options=None, indirect_jump_resolvers=None, base_state=None, exclude_sparse_regions=True, skip_specific_regions=True, heuristic_plt_resolving=None, detect_tail_calls=False, low_priority=False, cfb=None, model=None, use_patches=False, elf_eh_frame=True, exceptions=True, nodecode_window_size=512, nodecode_threshold=0.3, nodecode_step=16483, start=None, end=None, collect_data_references=None, extra_cross_references=None, **extra_arch_options)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis, angr.analyses.cfg.cfg_base.CFGBase

We find functions inside the given binary, and build a control-flow graph in very fast manners: instead of simulating program executions, keeping track of states, and performing expensive data-flow analysis, CFGFast will only perform light-weight analyses combined with some heuristics, and with some strong assumptions.

In order to identify as many functions as possible, and as accurate as possible, the following operation sequence is followed:

# Active scanning

If the binary has “function symbols” (TODO: this term is not accurate enough), they are starting points of the code scanning

If the binary does not have any “function symbol”, we will first perform a function prologue scanning on the entire binary, and start from those places that look like function beginnings

Otherwise, the binary’s entry point will be the starting point for scanning

# Passive scanning

After all active scans are done, we will go through the whole image and scan all code pieces

Due to the nature of those techniques that are used here, a base address is often not required to use this analysis routine. However, with a correct base address, CFG recovery will almost always yield a much better result. A custom analysis, called GirlScout, is specifically made to recover the base address of a binary blob. After the base address is determined, you may want to reload the binary with the new base address by creating a new Project object, and then re-recover the CFG.

Parameters

binary – The binary to recover CFG on. By default the main binary is used.
objects – A list of objects to recover the CFG on. By default it will recover the CFG of all loaded objects.
regions (iterable) – A list of tuples in the form of (start address, end address) describing memory regions that the CFG should cover.
pickle_intermediate_results (bool) – If we want to store the intermediate results or not.
symbols (bool) – Get function beginnings from symbols in the binary.
function_prologues (bool) – Scan the binary for function prologues, and use those positions as function beginnings
resolve_indirect_jumps (bool) – Try to resolve indirect jumps. This is necessary to resolve jump targets from jump tables, etc.
force_segment (bool) – Force CFGFast to rely on binary segments instead of sections.
force_complete_scan (bool) – Perform a complete scan on the binary and maximize the number of identified code blocks.
data_references (bool) – Enables the collection of references to data used by individual instructions. This does not collect ‘cross-references’, particularly those that involve multiple instructions. For that, see cross_references
cross_references (bool) – Whether CFGFast should collect “cross-references” from the entire program or not. This will populate the knowledge base with references to and from each recognizable address constant found in the code. Note that, because this performs constant propagation on the entire program, it may be much slower and consume more memory. This option implies data_references=True.
normalize (bool) – Normalize the CFG as well as all function graphs after CFG recovery.
start_at_entry (bool) – Begin CFG recovery at the entry point of this project. Setting it to False prevents CFGFast from viewing the entry point as one of the starting points of code scanning.
function_starts (list) – A list of extra function starting points. CFGFast will try to resume scanning from each address in the list.
extra_memory_regions (list) – A list of 2-tuple (start-address, end-address) that shows extra memory regions. Integers falling inside will be considered as pointers.
indirect_jump_resolvers (list) – A custom list of indirect jump resolvers. If this list is None or empty, default indirect jump resolvers specific to this architecture and binary types will be loaded.
base_state – A state to use as a backer for all memory loads
detect_tail_calls (bool) – Enable aggressive tail-call optimization detection.
elf_eh_frame (bool) – Retrieve function starts (and maybe sizes later) from the .eh_frame of ELF binaries.
start (int) – (Deprecated) The beginning address of CFG recovery.
end (int) – (Deprecated) The end address of CFG recovery.
arch_options (CFGArchOptions) – Architecture-specific options.
extra_arch_options (dict) – Any key-value pair in kwargs will be seen as an arch-specific option and will be used to set the option value in self._arch_options.

Extra parameters that angr.Analysis takes:

Parameters

progress_callback – Specify a callback function to get the progress during CFG recovery.
show_progressbar (bool) – Should CFGFast show a progressbar during CFG recovery or not.

Returns

None

PRINTABLES = b'0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ!"#$%&\'()*+,-./:;<=>?@[\\]^_`{|}~ \t\n\r'

SPECIAL_THUNKS = {'AMD64': {b'\xe8\x07\x00\x00\x00\xf3\x90\x0f\xae\xe8\xeb\xf9H\x89\x04$\xc3': ('jmp', 'rax'), b'\xe8\x07\x00\x00\x00\xf3\x90\x0f\xae\xe8\xeb\xf9H\x8dd$\x08\xc3': ('ret',)}}

tag: Optional[str] = 'CFGFast'

property graph

property memory_data

property jump_tables

property insn_addr_to_memory_data

do_full_xrefs(overlay_state=None)

Perform xref recovery on all functions.

Parameters: overlay (SimState) – An overlay state for loading constant data.
Returns: None

copy()

output()

generate_code_cover(**kwargs)

indirect_jumps: Dict[int, IndirectJump]

class angr.analyses.cfg.cfg_arch_options.CFGArchOptions(arch, **options)

Bases: object

Stores architecture-specific options and settings, as well as the detailed explanation of those options and settings.

Suppose ao is the CFGArchOptions object, and there is an option called ret_jumpkind_heuristics, you can access it by ao.ret_jumpkind_heuristics and set its value via ao.ret_jumpkind_heuristics = True

Variables

OPTIONS (dict) – A dict of all default options for different architectures.
arch (archinfo.Arch) – The architecture object.
_options (dict) – Values of all CFG options that are specific to the current architecture.

Constructor.

Parameters

arch (archinfo.Arch) – The architecture instance.
options (dict) – Architecture-specific options, which will be used to initialize this object.

OPTIONS = {'ARMCortexM': {'pattern_match_ifuncs': (<class 'bool'>, True), 'ret_jumpkind_heuristics': (<class 'bool'>, True), 'switch_mode_on_nodecode': (<class 'bool'>, False)}, 'ARMEL': {'pattern_match_ifuncs': (<class 'bool'>, True), 'ret_jumpkind_heuristics': (<class 'bool'>, True), 'switch_mode_on_nodecode': (<class 'bool'>, True)}, 'ARMHF': {'pattern_match_ifuncs': (<class 'bool'>, True), 'ret_jumpkind_heuristics': (<class 'bool'>, True), 'switch_mode_on_nodecode': (<class 'bool'>, True)}}

arch = None

class angr.analyses.cfg.cfg_job_base.BlockID(addr, callsite_tuples, jump_type)

Bases: object

A context-sensitive key for a SimRun object.

callsite_repr()

static new(addr, callstack_suffix, jumpkind)

property func_addr

class angr.analyses.cfg.cfg_job_base.FunctionKey(addr, callsite_tuples)

Bases: object

A context-sensitive key for a function.

callsite_repr()

static new(addr, callsite_tuples)

class angr.analyses.cfg.cfg_job_base.CFGJobBase(addr, state, context_sensitivity_level, block_id=None, src_block_id=None, src_exit_stmt_idx=None, src_ins_addr=None, jumpkind=None, call_stack=None, is_narrowing=False, skip=False, final_return_address=None)

Bases: object

Describes an entry in CFG or VFG. Only used internally by the analysis.

property call_stack

call_stack_copy()

get_call_stack_suffix()

property func_addr

property current_stack_pointer

class angr.analyses.cfg.indirect_jump_resolvers.amd64_elf_got.AMD64ElfGotResolver(project)

Bases: angr.analyses.cfg.indirect_jump_resolvers.resolver.IndirectJumpResolver

filter(cfg, addr, func_addr, block, jumpkind)

resolve(cfg, addr, func_addr, block, jumpkind)

class angr.analyses.cfg.indirect_jump_resolvers.arm_elf_fast.ArmElfFastResolver(project)

Bases: angr.analyses.cfg.indirect_jump_resolvers.resolver.IndirectJumpResolver

filter(cfg, addr, func_addr, block, jumpkind)

resolve(cfg, addr, func_addr, block, jumpkind)

Resolves the indirect jump in ARM ELF binaries where all internal function calls are performed in the following manner:

ldr r3, [pc+#0x124] ; load a constant from the constant_pool blx r3

Parameters

cfg – A CFG instance.
addr (int) – Address of the IRSB.
func_addr (int) – Address of the function.
block – The IRSB.
jumpkind (str) – The jumpkind.

Returns

Return type

tuple

class angr.analyses.cfg.indirect_jump_resolvers.x86_pe_iat.X86PeIatResolver(project)

Bases: angr.analyses.cfg.indirect_jump_resolvers.resolver.IndirectJumpResolver

filter(cfg, addr, func_addr, block, jumpkind)

resolve(cfg, addr, func_addr, block, jumpkind)

class angr.analyses.cfg.indirect_jump_resolvers.mips_elf_fast.OverwriteTmpValueCallback(gp_value)

Bases: object

Overwrites temporary values during resolution

overwrite_tmp_value(state)

class angr.analyses.cfg.indirect_jump_resolvers.mips_elf_fast.MipsElfFastResolver(project)

Bases: angr.analyses.cfg.indirect_jump_resolvers.resolver.IndirectJumpResolver

Indirect Jump Resolver for MIPs

filter(cfg, addr, func_addr, block, jumpkind)

resolve(cfg, addr, func_addr, block, jumpkind)

Wrapper for _resolve that slowly increments the max_depth used by Blade for finding sources until we can resolve the addr or we reach the default max_depth

Parameters

cfg – A CFG instance.
addr (int) – IRSB address.
func_addr (int) – The function address.
block (pyvex.IRSB) – The IRSB.
jumpkind (str) – The jumpkind.

Returns

If it was resolved and targets alongside it

Return type

tuple

class angr.analyses.cfg.indirect_jump_resolvers.x86_elf_pic_plt.X86ElfPicPltResolver(project)

Bases: angr.analyses.cfg.indirect_jump_resolvers.resolver.IndirectJumpResolver

In X86 ELF position-independent code, PLT stubs uses ebx to resolve library calls, where ebx stores the address to the beginning of the GOT. We resolve the target by forcing ebx to be the beginning of the GOT and simulate the execution in fast path mode.

filter(cfg, addr, func_addr, block, jumpkind)

resolve(cfg, addr, func_addr, block, jumpkind)

angr.analyses.cfg.indirect_jump_resolvers.default_resolvers.default_indirect_jump_resolvers(obj, project)

exception angr.analyses.cfg.indirect_jump_resolvers.jumptable.NotAJumpTableNotification: Bases: angr.errors.AngrError

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.UninitReadMeta

Bases: object

uninit_read_base = 201326592

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.AddressTransferringTypes

Bases: object

Assignment = 0

SignedExtension = 1

UnsignedExtension = 2

Truncation = 3

Or1 = 4

ShiftLeft = 5

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.JumpTargetBaseAddr(stmt_loc, stmt, tmp, base_addr=None, tmp_1=None)

Bases: object

property base_addr_available

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.JumpTableProcessorState(arch)

Bases: object

The state used in JumpTableProcessor.

arch

is_jumptable

stmts_to_instrument

regs_to_initialize

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.RegOffsetAnnotation(reg_offset: angr.engines.light.data.RegisterOffset)

Bases: claripy.annotation.Annotation

reg_offset

property relocatable

property eliminatable

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.JumpTableProcessor(project, bp_sp_diff=256)

Bases: angr.engines.light.engine.SimEngineLightVEXMixin, angr.engines.light.engine.SimEngineLight

Implements a simple and stupid data dependency tracking for stack and register variables.

Also determines which statements to instrument during static execution of the slice later. For example, the following example is not uncommon in non-optimized binaries:

    mov  [rbp+var_54], 1
loc_4051a6:
    cmp  [rbp+var_54], 6
    ja   loc_405412 (default)
loc_4051b0:
    mov  eax, [rbp+var_54]
    mov  rax, qword [rax*8+0x223a01]
    jmp  rax

We want to instrument the first instruction and replace the constant 1 with a symbolic variable, otherwise we will not be able to recover all jump targets later in block 0x4051b0.

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.StoreHook

Bases: object

static hook(state)

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.LoadHook

Bases: object

hook_before(state)

hook_after(state)

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.PutHook

Bases: object

static hook(state)

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.RegisterInitializerHook(reg_offset, reg_bits, value)

Bases: object

hook(state)

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.BSSHook(project, bss_regions)

Bases: object

bss_memory_read_hook(state)

bss_memory_write_hook(state)

class angr.analyses.cfg.indirect_jump_resolvers.jumptable.JumpTableResolver(project)

Bases: angr.analyses.cfg.indirect_jump_resolvers.resolver.IndirectJumpResolver

A generic jump table resolver.

This is a fast jump table resolution. For performance concerns, we made the following assumptions:

The final jump target comes from the memory.
The final jump target must be directly read out of the memory, without any further modification or altering.

filter(cfg, addr, func_addr, block, jumpkind)

resolve(cfg, addr, func_addr, block, jumpkind)

Resolves jump tables.

Parameters

cfg – A CFG instance.
addr (int) – IRSB address.
func_addr (int) – The function address.
block (pyvex.IRSB) – The IRSB.

Returns

A bool indicating whether the indirect jump is resolved successfully, and a list of resolved targets

Return type

tuple

angr.analyses.cfg.indirect_jump_resolvers.const_resolver.exists_in_replacements(replacements, block_loc, tmp_var)

class angr.analyses.cfg.indirect_jump_resolvers.const_resolver.ConstantResolver(project)

Bases: angr.analyses.cfg.indirect_jump_resolvers.resolver.IndirectJumpResolver

filter(cfg, addr, func_addr, block, jumpkind)

resolve(cfg, addr, func_addr, block, jumpkind)

This function does the actual resolve. Our process is easy: Propagate all values inside the function specified, then extract the tmp_var used for the indirect jump from the basic block. Use the tmp var to locate the constant value stored in the replacements. If not present, returns False tuple.

Parameters

cfg – CFG with specified function
addr – Address of indirect jump
func_addr – Address of function of indirect jump
block – Block of indirect jump (Block object)
jumpkind – VEX jumpkind (Ijk_Boring or Ijk_Call)

Returns

Bool tuple with replacement address

class angr.analyses.cfg.indirect_jump_resolvers.resolver.IndirectJumpResolver(project, timeless=False, base_state=None)

Bases: object

filter(cfg, addr, func_addr, block, jumpkind)

Check if this resolution method may be able to resolve the indirect jump or not.

Parameters

addr (int) – Basic block address of this indirect jump.
func_addr (int) – Address of the function that this indirect jump belongs to.
block – The basic block. The type is determined by the backend being used. It’s pyvex.IRSB if pyvex is used as the backend.
jumpkind (str) – The jumpkind.

Returns

True if it is possible for this resolution method to resolve the specific indirect jump, False otherwise.

Return type

bool

resolve(cfg, addr, func_addr, block, jumpkind)

Resolve an indirect jump.

Parameters

cfg – The CFG analysis object.
addr (int) – Basic block address of this indirect jump.
func_addr (int) – Address of the function that this indirect jump belongs to.
block – The basic block. The type is determined by the backend being used. It’s pyvex.IRSB if pyvex is used as the backend.
jumpkind (str) – The jumpkind.

Returns

A tuple of a boolean indicating whether the resolution is successful or not, and a list of resolved targets (ints).

Return type

tuple

class angr.analyses.cfg.cfg_utils.SCCPlaceholder(scc_id)

Bases: object

scc_id

class angr.analyses.cfg.cfg_utils.CFGUtils

Bases: object

A helper class with some static methods and algorithms implemented, that in fact, might take more than just normal CFGs.

static find_merge_points(function_addr, function_endpoints, graph)

Given a local transition graph of a function, find all merge points inside, and then perform a quasi-topological sort of those merge points.

A merge point might be one of the following cases: - two or more paths come together, and ends at the same address. - end of the current function

Parameters

function_addr (int) – Address of the function.
function_endpoints (list) – Endpoints of the function. They typically come from Function.endpoints.
graph (networkx.DiGraph) – A local transition graph of a function. Normally it comes from Function.graph.

Returns

A list of ordered addresses of merge points.

Return type

list

static find_widening_points(function_addr, function_endpoints, graph)

Given a local transition graph of a function, find all widening points inside.

Correctly choosing widening points is very important in order to not lose too much information during static analysis. We mainly consider merge points that has at least one loop back edges coming in as widening points.

Parameters

function_addr (int) – Address of the function.
function_endpoints (list) – Endpoints of the function, typically coming from Function.endpoints.
graph (networkx.DiGraph) – A local transition graph of a function, normally Function.graph.

Returns

A list of addresses of widening points.

Return type

list

static reverse_post_order_sort_nodes(graph, nodes=None)

Sort a given set of nodes in reverse post ordering.

Parameters

graph (networkx.DiGraph) – A local transition graph of a function.
nodes (iterable) – A collection of nodes to sort.

Returns

A list of sorted nodes.

Return type

list

static quasi_topological_sort_nodes(graph, nodes=None)

Sort a given set of nodes from a graph based on the following rules:

# - if A -> B and not B -> A, then we have A < B # - if A -> B and B -> A, then the ordering is undefined

Following the above rules gives us a quasi-topological sorting of nodes in the graph. It also works for cyclic graphs.

Parameters

graph (networkx.DiGraph) – A local transition graph of the function.
nodes (iterable) – A list of nodes to sort. None if you want to sort all nodes inside the graph.

Returns

A list of ordered nodes.

Return type

list

class angr.analyses.cfg.cfg_fast_soot.CFGFastSoot(support_jni=False, **kwargs)

Bases: angr.analyses.cfg.cfg_fast.CFGFast

normalize()

make_functions()

Revisit the entire control flow graph, create Function instances accordingly, and correctly put blocks into each function.

Although Function objects are crated during the CFG recovery, they are neither sound nor accurate. With a pre-constructed CFG, this method rebuilds all functions bearing the following rules:

A block may only belong to one function.

Small functions lying inside the startpoint and the endpoint of another function will be merged with the other function

Tail call optimizations are detected.

PLT stubs are aligned by 16.

Returns: None

indirect_jumps: Dict[int, IndirectJump]

class angr.analyses.cfg.segment_list.Segment(start, end, sort)

Bases: object

Representing a memory block. This is not the “Segment” in ELF memory model

Parameters

start (int) – Start address.
end (int) – End address.
sort (str) – Type of the segment, can be code, data, etc.

Returns

None

start

end

sort

property size

Calculate the size of the Segment.

Returns: Size of the Segment.
Return type: int

copy()

Make a copy of the Segment.

Returns: A copy of the Segment instance.
Return type: angr.analyses.cfg_fast.Segment

class angr.analyses.cfg.segment_list.SegmentList

Bases: object

SegmentList describes a series of segmented memory blocks. You may query whether an address belongs to any of the blocks or not, and obtain the exact block(segment) that the address belongs to.

search(addr: int) → int

Checks which segment that the address addr should belong to, and, returns the offset of that segment. Note that the address may not actually belong to the block.

Parameters: addr – The address to search
Returns: The offset of the segment.

next_free_pos(address)

Returns the next free position with respect to an address, including that address itself

Parameters: address – The address to begin the search with (including itself)
Returns: The next free position

next_pos_with_sort_not_in(address, sorts, max_distance=None)

Returns the address of the next occupied block whose sort is not one of the specified ones.

Parameters

address (int) – The address to begin the search with (including itself).
sorts – A collection of sort strings.
max_distance – The maximum distance between address and the next position. Search will stop after we come across an occupied position that is beyond address + max_distance. This check will be disabled if max_distance is set to None.

Returns

The next occupied position whose sort is not one of the specified ones, or None if no such position exists.

Return type

int or None

is_occupied(address)

Check if an address belongs to any segment

Parameters: address – The address to check
Returns: True if this address belongs to a segment, False otherwise

occupied_by_sort(address: int) → Optional[str]

Check if an address belongs to any segment, and if yes, returns the sort of the segment

Parameters: address – The address to check
Returns: Sort of the segment that occupies this address

occupied_by(address: int) → Optional[Tuple[int, int, str]]

Check if an address belongs to any segment, and if yes, returns the beginning, the size, and the sort of the segment.

Parameters: address – The address to check

occupy(address, size, sort)

Include a block, specified by (address, size), in this segment list.

Parameters

address (int) – The starting address of the block.
size (int) – Size of the block.
sort (str) – Type of the block.

Returns

None

release(address: int, size: int) → None

Remove a block, specified by (address, size), in this segment list.

Parameters

address – The starting address of the block.
size – Size of the block.

copy()

Make a copy of the SegmentList.

Returns: A copy of the SegmentList instance.
Return type: angr.analyses.cfg_fast.SegmentList

property occupied_size

The sum of sizes of all blocks

Returns: An integer

property has_blocks

Returns if this segment list has any block or not. !is_empty

Returns: True if it’s not empty, False otherwise

class angr.analyses.cdg.CDG(cfg, start=None, no_construct=False)

Bases: angr.analyses.analysis.Analysis

Implements a control dependence graph.

Constructor.

Parameters

cfg – The control flow graph upon which this control dependence graph will build
start – The starting point to begin constructing the control dependence graph
no_construct – Skip the construction step. Only used in unit-testing.

property graph

get_post_dominators(): Return the post-dom tree

get_dependants(run): Return a list of nodes that are control dependent on the given node in the control dependence graph

get_guardians(run): Return a list of nodes on whom the specific node is control dependent in the control dependence graph

class angr.analyses.datagraph_meta.DataGraphMeta

Bases: object

get_irsb_at(addr)

pp(imarks=False): Pretty print the graph. @imarks determine whether the printed graph represents instructions (coarse grained) for easier navigation, or exact statements.

class angr.analyses.code_tagging.CodeTags

Bases: object

HAS_XOR = 'HAS_XOR'

HAS_BITSHIFTS = 'HAS_BITSHIFTS'

HAS_SQL = 'HAS_SQL'

LARGE_SWITCH = 'LARGE_SWITCH'

class angr.analyses.code_tagging.CodeTagging(func)

Bases: angr.analyses.analysis.Analysis

analyze()

has_xor()

Detects if there is any xor operation in the function.

Returns: Tags

has_bitshifts()

Detects if there is any bitwise operation in the function.

Returns: Tags.

has_sql(): Detects if there is any reference to strings that look like SQL queries.

class angr.angrdb.db.AngrDB(project=None)

Bases: object

AngrDB provides a storage solution for an angr project, its knowledge bases, and some other types of data. It is designed to use an SQL-based database as the storage backend.

ALL_TABLES = ['objects']

VERSION = 1

static open_db(db_str='sqlite:///:memory:')

static session_scope(Session)

static save_info(session, key, value)

Save an information entry to the database.

Parameters

session –
key –
value –

Returns

static get_info(session, key)

Get an information entry from the database.

Parameters

session –
key –

Returns

update_dbinfo(session, extra_info: Optional[Dict[str, str]] = None)

Update the information in database.

Parameters: session –
Returns

get_dbinfo(session, extra_info: Optional[Dict[str, str]] = None)

Get database information.

Parameters: session –
Returns: A dict of information entries.

db_compatible(version)

Checks if the given database version is compatible with the current AngrDB class.

Parameters: version (int) – The version of the database.
Returns: True if compatible, False otherwise.
Return type: bool

dump(db_path, kbs: Optional[List[KnowledgeBase]] = None, extra_info: Optional[Dict[str, Any]] = None)

load(db_path: str, kb_names: Optional[List[str]] = None, other_kbs: Optional[Dict[str, KnowledgeBase]] = None, extra_info: Optional[Dict[str, Any]] = None)

class angr.angrdb.models.DbInformation(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Stores information related to the current database. Basically a key-value store.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

key

value

class angr.angrdb.models.DbObject(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Models a binary object.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

main_object

path

content

backend

backend_args

class angr.angrdb.models.DbKnowledgeBase(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Models a knowledge base.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

name

cfgs

funcs

xrefs

comments

labels

var_collections

structured_code

class angr.angrdb.models.DbCFGModel(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Models a CFGFast instance.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

kb_id

kb

ident

blob

class angr.angrdb.models.DbFunction(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Models a Function instance.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

kb_id

kb

addr

blob

class angr.angrdb.models.DbVariableCollection(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Models a VariableManagerInternal instance.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

kb_id

kb

func_addr

ident

blob

class angr.angrdb.models.DbStructuredCode(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Models a StructuredCode instance.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

kb_id

kb

func_addr

flavor

expr_comments

stmt_comments

configuration

class angr.angrdb.models.DbXRefs(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Models an XRefManager instance.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

kb_id

kb

blob

class angr.angrdb.models.DbComment(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Models a comment.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

kb_id

kb

addr

comment

type

class angr.angrdb.models.DbLabel(**kwargs)

Bases: sqlalchemy.orm.decl_api.Base

Models a label.

A simple constructor that allows initialization from kwargs.

Sets attributes on the constructed instance using the names and values in kwargs.

Only keys that are present as attributes of the instance’s class are allowed. These could be, for example, any mapped columns or relationships.

id

kb_id

kb

addr

name

class angr.angrdb.serializers.cfg_model.CFGModelSerializer

Bases: object

Serialize/unserialize a CFGModel.

static dump(session, db_kb, ident, cfg_model)

Parameters

session –
db_kb (DbKnowledgeBase) – The database object for KnowledgeBase.
ident (str) – Identifier of the CFG model.
cfg_model (CFGModel) – The CFG model to dump.

Returns

None

static load(session, db_kb, ident, cfg_manager, loader=None)

class angr.angrdb.serializers.comments.CommentsSerializer

Bases: object

Serialize/unserialize comments to/from a database session.

static dump(session, db_kb, comments)

Parameters

session –
db_kb (DbKnowledgeBase) –
comments (Comments) –

Returns

None

static load(session, db_kb, kb)

Parameters

session –
db_kb (DbKnowledgeBase) –
kb (KnowledgeBase) –

Returns

class angr.angrdb.serializers.funcs.FunctionManagerSerializer

Bases: object

Serialize/unserialize a function manager and its functions.

static dump(session, db_kb: DbKnowledgeBase, func_manager: angr.knowledge_plugins.functions.function_manager.FunctionManager)

Parameters

session –
db_kb (DbKnowledgeBase) –
func_manager (FunctionManager) –

Returns

static load(session, db_kb: DbKnowledgeBase, kb: KnowledgeBase)

Parameters

session –
db_kb (DbKnowledgeBase) –
kb (KnowledgeBase) –

Returns

A loaded function manager.

class angr.angrdb.serializers.kb.KnowledgeBaseSerializer

Bases: object

Serialize/unserialize a KnowledgeBase object.

static dump(session, kb)

Parameters

session – The database session object.
kb (KnowledgeBase) – The KnowledgeBase instance to serialize.

Returns

None

static load(session, project, name)

Parameters: session –
Returns

class angr.angrdb.serializers.labels.LabelsSerializer

Bases: object

Serialize/unserialize labels to/from a database session.

static dump(session, db_kb, labels)

Parameters

session –
db_kb (DbKnowledgeBase) –
labels (Labels) –

Returns

None

static load(session, db_kb, kb)

Parameters

session –
db_kb (DbKnowledgeBase) –
kb (KnowledgeBase) –

Returns

class angr.angrdb.serializers.loader.LoaderSerializer

Bases: object

Serialize/unserialize a CLE Loader object into/from an angr DB.

backend2name = {<class 'cle.backends.elf.elf.ELF'>: 'elf', <class 'cle.backends.blob.Blob'>: 'blob', <class 'cle.backends.elf.elfcore.ELFCore'>: 'elfcore', <class 'cle.backends.pe.pe.PE'>: 'pe', <class 'cle.backends.cgc.cgc.CGC'>: 'cgc', <class 'cle.backends.cgc.backedcgc.BackedCGC'>: 'backedcgc', <class 'cle.backends.ihex.Hex'>: 'hex', <class 'cle.backends.minidump.Minidump'>: 'minidump', <class 'cle.backends.macho.macho.MachO'>: 'mach-o', <class 'cle.backends.named_region.NamedRegion'>: 'named_region', <class 'cle.backends.java.jar.Jar'>: 'jar', <class 'cle.backends.java.apk.Apk'>: 'apk', <class 'cle.backends.xbe.XBE'>: 'xbe', <class 'cle.backends.static_archive.StaticArchive'>: 'AR'}

static dump(session, loader)

static load(session)

class angr.angrdb.serializers.xrefs.XRefsSerializer

Bases: object

Serialize/unserialize an XRefs object to/from a database session.

static dump(session, db_kb, xrefs)

Parameters

session –
db_kb (DbKnowledgeBase) –
xrefs (XRefManager) –

Returns

static load(session, db_kb, kb, cfg_model=None)

Parameters

session –
db_kb (DbKnowledgeBase) –
kb (KnowledgeBase) –
cfg_model (CFGModel) –

Returns

class angr.angrdb.serializers.variables.VariableManagerSerializer

Bases: object

Serialize/unserialize a variable manager and its variables.

static dump(session, db_kb: DbKnowledgeBase, var_manager: angr.knowledge_plugins.variables.variable_manager.VariableManager)

static dump_internal(session, db_kb: DbKnowledgeBase, internal_manager: angr.knowledge_plugins.variables.variable_manager.VariableManagerInternal, func_addr: int, ident=None)

static load(session, db_kb: DbKnowledgeBase, kb: KnowledgeBase, ident=None)

static load_internal(db_varcoll, variable_manager: angr.knowledge_plugins.variables.variable_manager.VariableManager) → angr.knowledge_plugins.variables.variable_manager.VariableManagerInternal

class angr.angrdb.serializers.structured_code.StructuredCodeManagerSerializer

Bases: object

Serialize/unserialize a structured code manager.

static dump(session, db_kb: DbKnowledgeBase, code_manager: angr.knowledge_plugins.structured_code.manager.StructuredCodeManager)

Parameters

session –
db_kb –
code_manager –

Returns

static dict_strkey_to_intkey(d: Dict[str, Any]) → Dict[int, Any]

static load(session, db_kb: DbKnowledgeBase, kb: KnowledgeBase) → angr.knowledge_plugins.structured_code.manager.StructuredCodeManager

Parameters

session –
db_kb –
kb –

Returns

A loaded structured code manager

class angr.analyses.decompiler.structurer.RecursiveStructurer(region, cond_proc=None, func: Optional[Function] = None)

Bases: angr.analyses.analysis.Analysis

Recursively structure a region and all of its subregions.

class angr.analyses.decompiler.structurer.Structurer(region, parent_map=None, condition_processor=None, func: Optional[Function] = None, is_loop_body=False)

Bases: angr.analyses.analysis.Analysis

Structure a region.

The current function graph is provided so that we can detect certain edge cases, for example, jump table entries no longer exist due to empty node removal during structuring or prior steps.

exception angr.analyses.decompiler.ail_simplifier.HasCallNotification

Bases: Exception

Notifies the existence of a call statement.

class angr.analyses.decompiler.ail_simplifier.AILBlockTempCollector(**kwargs)

Bases: angr.analyses.decompiler.ailblock_walker.AILBlockWalker

Collects any temporaries used in a block.

class angr.analyses.decompiler.ail_simplifier.AILSimplifier(func, func_graph=None, remove_dead_memdefs=False, stack_arg_offsets: Optional[Set[Tuple[int, int]]] = None, unify_variables=False)

Bases: angr.analyses.analysis.Analysis

Perform function-level simplifications.

class angr.analyses.decompiler.ailblock_walker.AILBlockWalker(stmt_handlers=None, expr_handlers=None)

Bases: object

Walks all statements and expressions of an AIL node.

walk(block: ailment.block.Block)

walk_statement(stmt: ailment.statement.Statement)

walk_expression(expr: ailment.expression.Expression, stmt_idx: Optional[int] = None, stmt: Optional[int] = None, block: Optional[ailment.block.Block] = None)

exception angr.analyses.decompiler.ailgraph_walker.RemoveNodeNotice: Bases: Exception

class angr.analyses.decompiler.ailgraph_walker.AILGraphWalker(graph, handler, replace_nodes: bool = False)

Bases: object

Walks an AIL graph and optionally replaces each node with a new node.

walk()

class angr.analyses.decompiler.block_simplifier.BlockSimplifier(block: Optional[Block], func_addr: Optional[int] = None, remove_dead_memdefs=False, stack_pointer_tracker=None, peephole_optimizations: Optional[Iterable[Union[Type[angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationStmtBase], Type[angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase]]]] = None, stack_arg_offsets: Optional[Set[Tuple[int, int]]] = None)

Bases: angr.analyses.analysis.Analysis

Simplify an AIL block.

Parameters: block – The AIL block to simplify. Setting it to None to skip calling self._analyze(), which is useful in test cases.

class angr.analyses.decompiler.callsite_maker.CallSiteMaker(block, reaching_definitions=None, stack_pointer_tracker=None, ail_manager=None)

Bases: angr.analyses.analysis.Analysis

Add calling convention, declaration, and args to a call site.

class angr.analyses.decompiler.ccall_rewriters.rewriter_base.CCallRewriterBase(ccall: ailment.expression.VEXCCallExpression)

Bases: object

The base class for CCall rewriters.

result: Optional[ailment.expression.Expression]

class angr.analyses.decompiler.ccall_rewriters.amd64_ccalls.AMD64CCallRewriter(ccall: ailment.expression.VEXCCallExpression)

Bases: angr.analyses.decompiler.ccall_rewriters.rewriter_base.CCallRewriterBase

Implements ccall rewriter for AMD64.

class angr.analyses.decompiler.clinic.Clinic(func, remove_dead_memdefs=False, exception_edges=False, sp_tracker_track_memory=True, optimization_passes=None, cfg=None, peephole_optimizations: Optional[Iterable[Union[Type[PeepholeOptimizationStmtBase], Type[PeepholeOptimizationExprBase]]]] = None, must_struct: Optional[Set[str]] = None, variable_kb=None, reset_variable_names=False, cache: Optional[DecompilationCache] = None)

Bases: angr.analyses.analysis.Analysis

A Clinic deals with AILments.

block(addr, size)

Get the converted block at the given specific address with the given size.

Parameters

addr (int) –
size (int) –

Returns

dbg_repr()

Returns

class angr.analyses.decompiler.condition_processor.ConditionProcessor(condition_mapping=None)

Bases: object

Convert between claripy AST and AIL expressions. Also calculates reaching conditions of all nodes on a graph.

clear()

recover_reaching_conditions(region, with_successors=False, jump_tables=None)

recover_reaching_conditions_for_jumptables(region, jump_tables, edge_conditions)

remove_claripy_bool_asts(node, memo=None)

classmethod get_last_statement(block): This is the buggy version of get_last_statements, because, you know, there can always be more than one last statement due to the existence of branching statements (like, If-then-else). All methods using get_last_statement() should switch to get_last_statements() and properly handle multiple last statements.

classmethod get_last_statements(block)

EXC_COUNTER = 1000

convert_claripy_bool_ast(cond, memo=None)

Convert recovered reaching conditions from claripy ASTs to ailment Expressions

Returns: None

convert_claripy_bool_ast_core(cond, memo)

claripy_ast_from_ail_condition(condition) → claripy.ast.base.Base

static claripy_ast_to_sympy_expr(ast, memo=None)

static sympy_expr_to_claripy_ast(expr, memo: Dict)

static simplify_condition(cond, depth_limit=8, variables_limit=8)

static simplify_condition_deprecated(cond)

class angr.analyses.decompiler.decompilation_options.DecompilationOption(name, description, value_type, cls, param, value_range=None, category='General', default_value=None, clears_cache=True): Bases: object

angr.analyses.decompiler.decompilation_options.O: alias of angr.analyses.decompiler.decompilation_options.DecompilationOption

class angr.analyses.decompiler.decompilation_cache.DecompilationCache

Bases: object

Caches key data structures that can be used later for refining decompilation results, such as retyping variables.

class angr.analyses.decompiler.decompiler.Decompiler(func, cfg=None, options=None, optimization_passes=None, sp_tracker_track_memory=True, variable_kb=None, peephole_optimizations: Optional[Iterable[Union[Type[PeepholeOptimizationStmtBase], Type[PeepholeOptimizationExprBase]]]] = None, vars_must_struct: Optional[Set[str]] = None, flavor='pseudocode', expr_comments=None, stmt_comments=None, decompile=True)

Bases: angr.analyses.analysis.Analysis

reflow_variable_types(type_constraints: Set, var_to_typevar: Dict, codegen)

Re-run type inference on an existing variable recovery result, then rerun codegen to generate new results.

Returns

static options_to_params(options)

Convert decompilation options to a dict of params.

Parameters: options (List[Tuple[DecompilationOption, Any]]) – The decompilation options.
Returns: A dict of keyword arguments.
Return type: dict

class angr.analyses.decompiler.empty_node_remover.EmptyNodeRemover(node)

Bases: object

Rewrites a node and its children to remove empty nodes.

The following optimizations are performed at the same time: - Convert if (A) { } else { … } to if(!A) { … } else { }

class angr.analyses.decompiler.graph_region.GraphRegion(head, graph, successors: Optional[list], graph_with_successors: networkx.classes.digraph.DiGraph, cyclic)

Bases: object

GraphRegion represents a region of nodes.

Variables

head – The head of the region.
graph – The region graph.
successors – A set of successors of nodes in the graph. These successors do not belong to the current region.
graph_with_successors – The region graph that includes successor nodes.

head

graph

successors

graph_with_successors: networkx.classes.digraph.DiGraph

cyclic

recursive_copy()

property addr

static dbg_get_repr(obj, ident=0)

dbg_print(ident=0)

replace_region(sub_region, replace_with)

angr.analyses.decompiler.optimization_passes.get_optimization_passes(arch, platform)

angr.analyses.decompiler.optimization_passes.get_default_optimization_passes(arch: Union[archinfo.arch.Arch, str], platform: Optional[str])

class angr.analyses.decompiler.optimization_passes.const_derefs.BlockWalker(project: Project)

Bases: angr.analyses.decompiler.ailblock_walker.AILBlockWalker

walk(block: ailment.block.Block)

class angr.analyses.decompiler.optimization_passes.const_derefs.ConstantDereferencesSimplifier(func, **kwargs)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

Makes the following simplifications:

*(*(const_addr))  ==>  *(value) iff  *const_addr == value

ARCHES = ['X86', 'AMD64', 'ARMEL', 'ARMHF', 'ARMCortexM']

PLATFORMS = ['linux']

STAGE: int = 1

class angr.analyses.decompiler.optimization_passes.eager_returns.EagerReturnsSimplifier(func, blocks_by_addr=None, blocks_by_addr_and_idx=None, graph=None, node_idx_start=0, max_level=2, min_indegree=4)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

Some compilers (if not all) generate only one returning block for a function regardless of how many returns there are in the source code. This oftentimes result in irreducible graphs and reduce the readability of the decompiled code. This optimization pass will make the function return eagerly by duplicating the return site of a function multiple times and assigning one copy of the return site to each of its sources when certain thresholds are met.

Note that this simplifier may reduce the readability of the generated code in certain cases, especially if the graph is already reducible without applying this simplifier.

Variables

max_level (int) – Number of times that we repeat the process of making returns eager.
min_indegree (int) – The minimum in-degree of the return site to be duplicated.
node_idx – The next node index. Each duplicated return site gets assigned a unique index, otherwise those duplicates will be considered as the same block in the graph because they have the same hash.

ARCHES = ['X86', 'AMD64', 'ARMCortexM', 'ARMHF', 'ARMEL']

PLATFORMS = ['cgc', 'linux']

STAGE: int = 0

exception angr.analyses.decompiler.optimization_passes.optimization_pass.MultipleBlocksException: Bases: Exception

class angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPassStage

Bases: object

AFTER_SINGLE_BLOCK_SIMPLIFICATION = 0

AFTER_GLOBAL_SIMPLIFICATION = 1

AFTER_VARIABLE_RECOVERY = 2

class angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass(func, blocks_by_addr=None, blocks_by_addr_and_idx=None, graph=None)

Bases: angr.analyses.analysis.Analysis

ARCHES = []

PLATFORMS = []

STAGE: int = None

property blocks_by_addr: Dict[int, Set[ailment.block.Block]]

property blocks_by_addr_and_idx: Dict[Tuple[int, Optional[int]], ailment.block.Block]

analyze()

angr.analyses.decompiler.optimization_passes.stack_canary_simplifier.s2u(s, bits)

class angr.analyses.decompiler.optimization_passes.stack_canary_simplifier.StackCanarySimplifier(func, **kwargs)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

Removes stack canary checks from AIL graphs.

ARCHES = ['X86', 'AMD64']

PLATFORMS = ['cgc', 'linux']

STAGE: int = 1

class angr.analyses.decompiler.optimization_passes.base_ptr_save_simplifier.BasePointerSaveSimplifier(func, **kwargs)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

ARCHES = ['X86', 'AMD64', 'ARMEL', 'ARMHF', 'ARMCortexM', 'MIPS32', 'MIPS64']

PLATFORMS = ['cgc', 'linux']

STAGE: int = 1

class angr.analyses.decompiler.optimization_passes.div_simplifier.DivSimplifierAILEngine: Bases: angr.analyses.decompiler.optimization_passes.engine_base.SimplifierAILEngine

class angr.analyses.decompiler.optimization_passes.div_simplifier.DivSimplifier(func, **kwargs)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

ARCHES = ['X86', 'AMD64']

PLATFORMS = ['linux', 'windows']

STAGE: int = 1

class angr.analyses.decompiler.optimization_passes.multi_simplifier.MultiSimplifierAILEngine: Bases: angr.analyses.decompiler.optimization_passes.engine_base.SimplifierAILEngine

class angr.analyses.decompiler.optimization_passes.multi_simplifier.MultiSimplifier(func, **kwargs)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

ARCHES = ['X86', 'AMD64']

PLATFORMS = ['linux', 'windows']

STAGE: int = 1

class angr.analyses.decompiler.optimization_passes.mod_simplifier.ModSimplifierAILEngine: Bases: angr.analyses.decompiler.optimization_passes.engine_base.SimplifierAILEngine

class angr.analyses.decompiler.optimization_passes.mod_simplifier.ModSimplifier(func, **kwargs)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

ARCHES = ['X86', 'AMD64']

PLATFORMS = ['linux', 'windows']

STAGE: int = 1

class angr.analyses.decompiler.optimization_passes.engine_base.SimplifierAILState(arch, variables=None)

Bases: object

copy()

merge(*others)

store_variable(old, new)

get_variable(old)

remove_variable(old)

filter_variables(atom)

class angr.analyses.decompiler.optimization_passes.engine_base.SimplifierAILEngine

Bases: angr.engines.light.engine.SimEngineLightAILMixin, angr.engines.light.engine.SimEngineLight

process(state, *args, **kwargs)

angr.analyses.decompiler.optimization_passes.register_save_area_simplifier.s2u(s, bits)

class angr.analyses.decompiler.optimization_passes.register_save_area_simplifier.RegisterSaveAreaSimplifier(func, **kwargs)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

ARCHES = ['X86', 'AMD64', 'ARM', 'ARMEL', 'ARMHF', 'ARMCortexM']

PLATFORMS = ['cgc', 'linux']

STAGE: int = 1

class angr.analyses.decompiler.optimization_passes.ret_addr_save_simplifier.RetAddrSaveSimplifier(func, **kwargs)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

Removes code in function prologues and epilogues for saving and restoring return address registers (ra, lr, etc.)

ARCHES = ['MIPS32', 'MIPS64']

PLATFORMS = ['linux']

STAGE: int = 1

class angr.analyses.decompiler.optimization_passes.x86_gcc_getpc_simplifier.X86GccGetPcSimplifier(func, **kwargs)

Bases: angr.analyses.decompiler.optimization_passes.optimization_pass.OptimizationPass

Simplifies __x86.get_pc_thunk calls.

ARCHES = ['X86']

PLATFORMS = ['linux']

STAGE: int = 0

class angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationStmtBase(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: object

name = 'Peephole Optimization - Statement'

description = 'Peephole Optimization - Statement'

stmt_classes = None

project: angr.project.Project

kb: angr.knowledge_base.knowledge_base.KnowledgeBase

func_addr: Optional[int]

optimize(stmt)

class angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: object

name = 'Peephole Optimization - Expression'

description = 'Peephole Optimization - Expression'

expr_classes = None

project: angr.project.Project

kb: angr.knowledge_base.knowledge_base.KnowledgeBase

func_addr: Optional[int]

optimize(expr)

static is_bool_expr(ail_expr)

class angr.analyses.decompiler.peephole_optimizations.a_sub_a_sub_n.ASubASubN(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'expr - (expr - N) => N'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.bool_expr_xor_1.BoolExprXor1(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'bool_expr ^ 1 => !bool_expr (a)'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.basepointeroffset_and_mask.BasePointerOffsetAndMask(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'Ptr & mask => Ptr'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.basepointeroffset_add_n.BasePointerOffsetAddN(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = '(Ptr - M) + N => Ptr - (M - N)'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.extended_byte_and_mask.ExtendedByteAndMask(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'extended byte & 0xff..ff => extended byte'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.coalesce_same_cascading_ifs.CoalesceSameCascadingIfs(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationStmtBase

name = 'Coalescing cascading If constructs'

stmt_classes = (<class 'ailment.statement.ConditionalJump'>,)

optimize(stmt: ailment.statement.ConditionalJump)

class angr.analyses.decompiler.peephole_optimizations.constant_derefs.ConstantDereferences(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

Dereferences constant memory loads from read-only memory regions.

name = 'Dereference constant references'

expr_classes = (<class 'ailment.expression.Load'>,)

optimize(expr: ailment.expression.Load)

class angr.analyses.decompiler.peephole_optimizations.conv_a_sub0_shr_and.ConvASub0ShrAnd(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'Conv(M->1, (expr >> N) & 1) => expr < 0'

expr_classes = (<class 'ailment.expression.Convert'>,)

optimize(expr: ailment.expression.Convert)

class angr.analyses.decompiler.peephole_optimizations.conv_shl_shr.ConvShlShr(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = '(expr << P) >> Q => (expr & mask) >> R'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.eager_eval.EagerEvaluation(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

Eagerly evaluates certain types of expressions.

name = 'Eager expression evaluation'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.one_sub_bool.OneSubBool(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = '1 - bool_expr => !bool_expr'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.remove_empty_if_body.RemoveEmptyIfBody(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationStmtBase

name = 'Remove empty If bodies'

stmt_classes = (<class 'ailment.statement.ConditionalJump'>,)

optimize(stmt: ailment.statement.ConditionalJump)

class angr.analyses.decompiler.peephole_optimizations.remove_cascading_conversions.RemoveCascadingConversions(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'Remove adjacent conversions'

expr_classes = (<class 'ailment.expression.Convert'>,)

optimize(expr: ailment.expression.Convert)

class angr.analyses.decompiler.peephole_optimizations.remove_redundant_bitmasks.RemoveRedundantBitmasks(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'Remove redundant bitmasks'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.remove_redundant_ite_branch.RemoveRedundantITEBranches(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'Remove redundant ITE branches'

expr_classes = (<class 'ailment.expression.ITE'>,)

optimize(expr: ailment.expression.ITE)

class angr.analyses.decompiler.peephole_optimizations.remove_redundant_nots.RemoveRedundantNots(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'Remove redundant Nots'

expr_classes = (<class 'ailment.expression.UnaryOp'>,)

optimize(expr: ailment.expression.UnaryOp)

class angr.analyses.decompiler.peephole_optimizations.remove_redundant_shifts.RemoveRedundantShifts(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'Remove redundant bitshifts'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.remove_redundant_conversions.RemoveRedundantConversions(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'Remove redundant conversions around binary operators'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.rewrite_mips_gp_loads.RewriteMipsGpLoads(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

Rewrite $gp-based loads to their actual values on MIPS.

name = 'MIPS GP-based Loads Rewriter'

expr_classes = (<class 'ailment.expression.Load'>,)

optimize(expr: ailment.expression.Load)

class angr.analyses.decompiler.peephole_optimizations.simplify_pc_relative_loads.SimplifyPcRelativeLoads(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

Simplifying pc-relative loads.

name = 'Simplify PC-relative loads'

expr_classes = (<class 'ailment.expression.BinaryOp'>,)

optimize(expr: ailment.expression.BinaryOp)

class angr.analyses.decompiler.peephole_optimizations.single_bit_xor.SingleBitXor(project: angr.project.Project, kb: angr.knowledge_base.knowledge_base.KnowledgeBase, func_addr: Optional[int] = None)

Bases: angr.analyses.decompiler.peephole_optimizations.base.PeepholeOptimizationExprBase

name = 'bool_expr ^ 1 => !bool_expr (b)'

expr_classes = (<class 'ailment.expression.Convert'>,)

optimize(expr: ailment.expression.Convert)

class angr.analyses.decompiler.region_identifier.RegionIdentifier(func, cond_proc=None, graph=None)

Bases: angr.analyses.analysis.Analysis

Identifies regions within a function.

static slice_graph(graph, node, frontier, include_frontier=False)

Generate a slice of the graph from the head node to the given frontier.

Parameters

graph (networkx.DiGraph) – The graph to work on.
node – The starting node in the graph.
frontier – A list of frontier nodes.
include_frontier (bool) – Whether the frontier nodes are included in the slice or not.

Returns

A subgraph.

Return type

networkx.DiGraph

class angr.analyses.decompiler.region_simplifiers.cascading_cond_transformer.CascadingConditionTransformer(node)

Bases: angr.analyses.decompiler.sequence_walker.SequenceWalker

Identifies and transforms if { … } else { if { … } else { … } } to if { … } else if { … } else if { … }.

class angr.analyses.decompiler.region_simplifiers.cascading_ifs.CascadingIfsRemover(node)

Bases: angr.analyses.decompiler.sequence_walker.SequenceWalker

Coalesce cascading If constructs. Transforming the following construct:

if (cond_a) {
    if (cond_b) {
        true_body
    } else { }
} else { }

into:

if (cond_a and cond_b) {
    true_body
} else { }

static is_empty_node(node)

class angr.analyses.decompiler.region_simplifiers.expr_folding.LocationBase: Bases: object

class angr.analyses.decompiler.region_simplifiers.expr_folding.StatementLocation(block_addr, block_idx, stmt_idx)

Bases: angr.analyses.decompiler.region_simplifiers.expr_folding.LocationBase

block_addr

block_idx

stmt_idx

class angr.analyses.decompiler.region_simplifiers.expr_folding.ExpressionLocation(block_addr, block_idx, stmt_idx, expr_idx)

Bases: angr.analyses.decompiler.region_simplifiers.expr_folding.LocationBase

block_addr

block_idx

stmt_idx

expr_idx

class angr.analyses.decompiler.region_simplifiers.expr_folding.ConditionLocation(cond_node_addr)

Bases: angr.analyses.decompiler.region_simplifiers.expr_folding.LocationBase

node_addr

class angr.analyses.decompiler.region_simplifiers.expr_folding.ConditionalBreakLocation(node_addr)

Bases: angr.analyses.decompiler.region_simplifiers.expr_folding.LocationBase

node_addr

class angr.analyses.decompiler.region_simplifiers.expr_folding.ExpressionUseFinder: Bases: angr.analyses.decompiler.ailblock_walker.AILBlockWalker

class angr.analyses.decompiler.region_simplifiers.expr_folding.ExpressionCounter(node, variable_manager)

Bases: angr.analyses.decompiler.sequence_walker.SequenceWalker

Find all expressions that are assigned once and only used once.

class angr.analyses.decompiler.region_simplifiers.expr_folding.ExpressionReplacer(assignments: Dict, uses: Dict): Bases: angr.analyses.decompiler.ailblock_walker.AILBlockWalker

class angr.analyses.decompiler.region_simplifiers.expr_folding.ExpressionFolder(assignments: Dict, uses: Dict, node): Bases: angr.analyses.decompiler.sequence_walker.SequenceWalker

class angr.analyses.decompiler.region_simplifiers.goto.GotoSimplifier(node)

Bases: angr.analyses.decompiler.sequence_walker.SequenceWalker

Remove unnecessary Jump statements.

class angr.analyses.decompiler.region_simplifiers.if_.IfSimplifier(node)

Bases: angr.analyses.decompiler.sequence_walker.SequenceWalker

Remove unnecessary jump or conditional jump statements if they jump to the successor right afterwards.

class angr.analyses.decompiler.region_simplifiers.ifelse.IfElseFlattener(node, functions)

Bases: angr.analyses.decompiler.sequence_walker.SequenceWalker

Remove unnecessary else branches and make the else node a direct successor of the previous If node if the If node always returns.

class angr.analyses.decompiler.region_simplifiers.loop.LoopSimplifier(node)

Bases: angr.analyses.decompiler.sequence_walker.SequenceWalker

Simplifies loops.

class angr.analyses.decompiler.region_simplifiers.node_address_finder.NodeAddressFinder(node)

Bases: angr.analyses.decompiler.sequence_walker.SequenceWalker

Walk the entire node and collect all addresses of nodes.

class angr.analyses.decompiler.region_simplifiers.region_simplifier.RegionSimplifier(func, region, variable_kb=None)

Bases: angr.analyses.analysis.Analysis

Simplifies a given region.

class angr.analyses.decompiler.sequence_walker.SequenceWalker(handlers=None, exception_on_unsupported=False)

Bases: object

Walks a SequenceNode and all its nodes, recursively.

walk(sequence)

class angr.analyses.decompiler.structured_codegen.base.PositionMappingElement(start, length, obj)

Bases: object

start: int

length: int

obj

class angr.analyses.decompiler.structured_codegen.base.PositionMapping

Bases: object

DUPLICATION_CHECK = True

items()

add_mapping(start_pos, length, obj)

get_node(pos: int)

get_element(pos: int) → Optional[angr.analyses.decompiler.structured_codegen.base.PositionMappingElement]

class angr.analyses.decompiler.structured_codegen.base.InstructionMappingElement(ins_addr, posmap_pos)

Bases: object

ins_addr: int

posmap_pos: int

class angr.analyses.decompiler.structured_codegen.base.InstructionMapping

Bases: object

items()

add_mapping(ins_addr, posmap_pos)

get_nearest_pos(ins_addr: int) → Optional[int]

class angr.analyses.decompiler.structured_codegen.base.BaseStructuredCodeGenerator(flavor=None)

Bases: object

reapply_options(options)

regenerate_text() → None

angr.analyses.decompiler.structured_codegen.c.unpack_typeref(ty)

class angr.analyses.decompiler.structured_codegen.c.CConstruct(codegen)

Bases: object

Represents a program construct in C. Acts as the base class for all other representation constructions.

codegen: StructuredCodeGenerator

c_repr(indent=0, pos_to_node=None, pos_to_addr=None, addr_to_pos=None)

Creates the C reperesentation of the code and displays it by constructing a large string. This function is called by each program function that needs to be decompiled. The map_pos_to_node and map_pos_to_addr act as position maps for the location of each variable and statment to be tracked for later GUI operations. The map_pos_to_addr also contains expressions that are nested inside of statments.

Parameters

indent – # of indents (int)
pos_to_nodemap_pos_to_ast –

Returns

c_repr_chunks(indent=0, asexpr=False)

static indent_str(indent=0)

class angr.analyses.decompiler.structured_codegen.c.CFunction(addr, name, functy: angr.sim_type.SimTypeFunction, arg_list: List[angr.analyses.decompiler.structured_codegen.c.CVariable], statements, variables_in_use, variable_manager, demangled_name=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CConstruct

Represents a function in C.

addr

name

functy

arg_list

statements

variables_in_use

variable_manager: VariableManagerInternal

demangled_name

variable_list_repr_chunks(indent=0)

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CStatement(codegen)

Bases: angr.analyses.decompiler.structured_codegen.c.CConstruct

Represents a statement in C.

class angr.analyses.decompiler.structured_codegen.c.CExpression(**kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CConstruct

Base class for C expressions.

property type

set_type(v)

class angr.analyses.decompiler.structured_codegen.c.CStatements(statements, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

Represents a sequence of statements in C.

statements

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CAILBlock(block, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

Represents a block of AIL statements.

block

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CLoop(codegen)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

Represents a loop in C.

class angr.analyses.decompiler.structured_codegen.c.CWhileLoop(condition, body, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CLoop

Represents a while loop in C.

condition

body

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CDoWhileLoop(condition, body, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CLoop

Represents a do-while loop in C.

condition

body

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CForLoop(initializer, condition, iterator, body, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

Represents a for-loop in C.

initializer

condition

iterator

body

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CIfElse(condition_and_nodes: List[Tuple[angr.analyses.decompiler.structured_codegen.c.CExpression, Optional[angr.analyses.decompiler.structured_codegen.c.CStatement]]], else_node=None, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

Represents an if-else construct in C.

condition_and_nodes

else_node

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CIfBreak(condition, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

Represents an if-break statement in C.

condition

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CBreak(tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

Represents a break statement in C.

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CContinue(tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

Represents a continue statement in C.

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CSwitchCase(switch, cases, default, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

Represents a switch-case statement in C.

switch

cases

default

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CAssignment(lhs, rhs, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

a = b

lhs

rhs

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CFunctionCall(callee_target, callee_func, args, returning=True, ret_expr=None, tags=None, is_expr: bool = False, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement, angr.analyses.decompiler.structured_codegen.c.CExpression

func(arg0, arg1)

Variables: callee_func (Function) – The function getting called.

callee_target

callee_func: Optional[Function]

args

returning

ret_expr

tags

is_expr

property type

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CReturn(retval, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

retval

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CGoto(target, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

target

tags

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CUnsupportedStatement(stmt, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CStatement

A wrapper for unsupported AIL statement.

stmt

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CStructField(struct_type, offset, field, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

struct_type

offset

field

tags

property type

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CPlaceholder(placeholder, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

placeholder: str

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CVariable(variable, unified_variable=None, offset=None, variable_type=None, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

Read value from a variable.

variable: angr.sim_variable.SimVariable

unified_variable: Optional[angr.sim_variable.SimVariable]

offset: Optional[int]

variable_type: Optional[angr.sim_type.SimType]

tags

property type

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CUnaryOp(op, operand, variable, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

Unary operations.

op

operand

variable

tags

property type

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CBinaryOp(op, lhs, rhs, variable, tags: Optional[dict] = None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

Binary operations.

op

lhs

rhs

variable

tags

property type

property op_precedence

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CTypeCast(src_type, dst_type, expr, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

src_type

dst_type

expr

tags

property type

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CConstant(value, type_, reference_values=None, reference_variable=None, tags: Optional[Dict] = None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

value

reference_values

reference_variable

tags

property type

static str_to_c_str(_str)

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CRegister(reg, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

reg

property type

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CITE(cond, iftrue, iffalse, tags=None, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

cond

iftrue

iffalse

tags

property type

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CDirtyExpression(dirty, **kwargs)

Bases: angr.analyses.decompiler.structured_codegen.c.CExpression

Ideally all dirty expressions should be handled and converted to proper conversions during conversion from VEX to AIL. Eventually this class should not be used at all.

dirty

property type

c_repr_chunks(indent=0, asexpr=False)

class angr.analyses.decompiler.structured_codegen.c.CClosingObject(opening_symbol)

Bases: object

A class to represent all objects that can be closed by it’s correspodning character. Examples: (), {}, []

opening_symbol

class angr.analyses.decompiler.structured_codegen.c.CStructuredCodeGenerator(func, sequence, indent=0, cfg=None, variable_kb=None, func_args: Optional[List[angr.sim_variable.SimVariable]] = None, binop_depth_cutoff: int = 10, show_casts=True, braces_on_own_lines=True, use_compound_assignments=True, flavor=None, stmt_comments=None, expr_comments=None)

Bases: angr.analyses.decompiler.structured_codegen.base.BaseStructuredCodeGenerator, angr.analyses.analysis.Analysis

reapply_options(options)

cleanup(): Remove existing rendering results.

regenerate_text() → None: Re-render text and re-generate all sorts of mapping information.

render_text(cfunc: angr.analyses.decompiler.structured_codegen.c.CFunction) → Tuple[str, angr.analyses.decompiler.structured_codegen.base.PositionMapping, angr.analyses.decompiler.structured_codegen.base.PositionMapping, angr.analyses.decompiler.structured_codegen.base.InstructionMapping, Dict[Any, Set[Any]]]

default_simtype_from_size(n: int) → angr.sim_type.SimType

angr.analyses.decompiler.structured_codegen.c.StructuredCodeGenerator: alias of angr.analyses.decompiler.structured_codegen.c.CStructuredCodeGenerator

class angr.analyses.decompiler.structured_codegen.dwarf_import.ImportedLine(addr): Bases: object

class angr.analyses.decompiler.structured_codegen.dwarf_import.ImportSourceCode(function, flavor='source', source_root=None, encoding='utf-8')

Bases: angr.analyses.decompiler.structured_codegen.base.BaseStructuredCodeGenerator, angr.analyses.analysis.Analysis

regenerate_text()

class angr.analyses.decompiler.structured_codegen.dummy.DummyStructuredCodeGenerator(flavor: str, expr_comments=None, stmt_comments=None, configuration=None)

Bases: angr.analyses.decompiler.structured_codegen.base.BaseStructuredCodeGenerator

A dummy structured code generator that only stores user-specified information.

exception angr.analyses.decompiler.structurer_nodes.EmptyBlockNotice: Bases: Exception

class angr.analyses.decompiler.structurer_nodes.MultiNode(nodes)

Bases: object

nodes

copy()

property addr

class angr.analyses.decompiler.structurer_nodes.BaseNode

Bases: object

static test_empty_node(node)

static test_empty_condition_node(cond_node)

class angr.analyses.decompiler.structurer_nodes.SequenceNode(addr: Optional[int], nodes=None)

Bases: angr.analyses.decompiler.structurer_nodes.BaseNode

addr

nodes

add_node(node)

insert_node(pos, node)

remove_node(node)

node_position(node)

copy()

dbg_repr(indent=0)

class angr.analyses.decompiler.structurer_nodes.CodeNode(node, reaching_condition)

Bases: angr.analyses.decompiler.structurer_nodes.BaseNode

node

reaching_condition

property addr

dbg_repr(indent=0)

copy()

class angr.analyses.decompiler.structurer_nodes.ConditionNode(addr, reaching_condition, condition, true_node, false_node=None)

Bases: angr.analyses.decompiler.structurer_nodes.BaseNode

addr

reaching_condition

condition

true_node

false_node

dbg_repr(indent=0)

node

class angr.analyses.decompiler.structurer_nodes.CascadingConditionNode(addr, condition_and_nodes: List[Tuple[Any, angr.analyses.decompiler.structurer_nodes.BaseNode]], else_node: Optional[angr.analyses.decompiler.structurer_nodes.BaseNode] = None)

Bases: angr.analyses.decompiler.structurer_nodes.BaseNode

addr

condition_and_nodes

else_node

class angr.analyses.decompiler.structurer_nodes.LoopNode(sort, condition, sequence_node, addr=None, continue_addr=None, initializer=None, iterator=None)

Bases: angr.analyses.decompiler.structurer_nodes.BaseNode

sort

condition

sequence_node

initializer

iterator

copy()

property addr

property continue_addr

class angr.analyses.decompiler.structurer_nodes.BreakNode(addr, target)

Bases: angr.analyses.decompiler.structurer_nodes.BaseNode

addr

target

class angr.analyses.decompiler.structurer_nodes.ContinueNode(addr, target)

Bases: angr.analyses.decompiler.structurer_nodes.BaseNode

addr

target

class angr.analyses.decompiler.structurer_nodes.ConditionalBreakNode(addr, condition, target)

Bases: angr.analyses.decompiler.structurer_nodes.BreakNode

condition

class angr.analyses.decompiler.structurer_nodes.SwitchCaseNode(switch_expr, cases, default_node, addr=None)

Bases: angr.analyses.decompiler.structurer_nodes.BaseNode

switch_expr

cases

default_node

addr

angr.analyses.decompiler.utils.remove_last_statement(node)

angr.analyses.decompiler.utils.append_statement(node, stmt)

angr.analyses.decompiler.utils.replace_last_statement(node, old_stmt, new_stmt)

angr.analyses.decompiler.utils.extract_jump_targets(stmt)

Extract concrete goto targets from a Jump or a ConditionalJump statement.

Parameters: stmt – The statement to analyze.
Returns: A list of known concrete jump targets.
Return type: list

angr.analyses.decompiler.utils.switch_extract_cmp_bounds(last_stmt: ailment.statement.ConditionalJump) → Optional[Tuple[Any, int, int]]

Check the last statement of the switch-case header node, and extract lower+upper bounds for the comparison.

Parameters: last_stmt – The last statement of the switch-case header node.
Returns: A tuple of (comparison expression, lower bound, upper bound), or None

angr.analyses.decompiler.utils.get_ast_subexprs(claripy_ast)

angr.analyses.decompiler.utils.insert_node(parent, insert_idx, node, node_idx, label=None, insert_location=None)

angr.analyses.decompiler.utils.to_ail_supergraph(transition_graph: networkx.classes.digraph.DiGraph) → networkx.classes.digraph.DiGraph

Takes an AIL graph and converts it into a AIL graph that treats calls and redundant jumps as parts of a bigger block instead of transitions. Calls to returning functions do not terminate basic blocks.

Based on region_identifier super_graph

Returns: A converted super transition graph

class angr.analyses.ddg.AST(op, *operands)

Bases: object

A mini implementation for AST

class angr.analyses.ddg.ProgramVariable(variable, location, initial=False, arch=None)

Bases: object

Describes a variable in the program at a specific location.

Variables

variable (SimVariable) – The variable.
location (CodeLocation) – Location of the variable.

property short_repr

class angr.analyses.ddg.DDGJob(cfg_node, call_depth): Bases: object

class angr.analyses.ddg.LiveDefinitions

Bases: object

A collection of live definitions with some handy interfaces for definition killing and lookups.

Constructor.

branch()

Create a branch of the current live definition collection.

Returns: A new LiveDefinition instance.
Return type: angr.analyses.ddg.LiveDefinitions

copy()

Make a hard copy of self.

Returns: A new LiveDefinition instance.
Return type: angr.analyses.ddg.LiveDefinitions

add_def(variable, location, size_threshold=32)

Add a new definition of variable.

Parameters

variable (SimVariable) – The variable being defined.
location (CodeLocation) – Location of the varaible being defined.
size_threshold (int) – The maximum bytes to consider for the variable.

Returns

True if the definition was new, False otherwise

Return type

bool

add_defs(variable, locations, size_threshold=32)

Add a collection of new definitions of a variable.

Parameters

variable (SimVariable) – The variable being defined.
locations (iterable) – A collection of locations where the variable was defined.
size_threshold (int) – The maximum bytes to consider for the variable.

Returns

True if any of the definition was new, False otherwise

Return type

bool

kill_def(variable, location, size_threshold=32)

Add a new definition for variable and kill all previous definitions.

Parameters

variable (SimVariable) – The variable to kill.
location (CodeLocation) – The location where this variable is defined.
size_threshold (int) – The maximum bytes to consider for the variable.

Returns

None

lookup_defs(variable, size_threshold=32)

Find all definitions of the variable.

Parameters

variable (SimVariable) – The variable to lookup for.
size_threshold (int) – The maximum bytes to consider for the variable. For example, if the variable is 100 byte long, only the first size_threshold bytes are considered.

Returns

A set of code locations where the variable is defined.

Return type

set

items()

An iterator that returns all live definitions.

Returns: The iterator.
Return type: iter

itervariables()

An iterator that returns all live variables.

Returns: The iterator.
Return type: iter

class angr.analyses.ddg.DDGViewItem(ddg, variable, simplified=False)

Bases: object

property depends_on

property dependents

class angr.analyses.ddg.DDGViewInstruction(cfg, ddg, insn_addr, simplified=False)

Bases: object

property definitions

Get all definitions located at the current instruction address.

Returns: A list of ProgramVariable instances.
Return type: list

class angr.analyses.ddg.DDGView(cfg, ddg, simplified=False)

Bases: object

A view of the data dependence graph.

class angr.analyses.ddg.DDG(cfg, start=None, call_depth=None, block_addrs=None)

Bases: angr.analyses.analysis.Analysis

This is a fast data dependence graph directly generated from our CFG analysis result. The only reason for its existence is the speed. There is zero guarantee for being sound or accurate. You are supposed to use it only when you want to track the simplest data dependence, and you do not care about soundness or accuracy.

For a better data dependence graph, please consider performing a better static analysis first (like Value-set Analysis), and then construct a dependence graph on top of the analysis result (for example, the VFG in angr).

The DDG is based on a CFG, which should ideally be a CFGEmulated generated with the following options:

keep_state=True to keep all input states

state_add_options=angr.options.refs to store memory, register, and temporary value accesses

You may want to consider a high value for context_sensitivity_level as well when generating the CFG.

Also note that since we are using states from CFG, any improvement in analysis performed on CFG (like a points-to analysis) will directly benefit the DDG.

Parameters

cfg – Control flow graph. Please make sure each node has an associated state with it, e.g. by passing the keep_state=True and state_add_options=angr.options.refs arguments to CFGEmulated.
start – An address, Specifies where we start the generation of this data dependence graph.
call_depth – None or integers. A non-negative integer specifies how deep we would like to track in the call tree. None disables call_depth limit.
block_addrs (iterable or None) – A collection of block addresses that the DDG analysis should be performed on.

property graph

A networkx DiGraph instance representing the dependence relations between statements. :rtype: networkx.DiGraph

Type: returns

property data_graph

Get the data dependence graph.

Returns: A networkx DiGraph instance representing data dependence.
Return type: networkx.DiGraph

property simplified_data_graph: return:

property ast_graph

pp(): Pretty printing.

dbg_repr(): Representation for debugging.

get_predecessors(code_location)

Returns all predecessors of the code location.

Parameters: code_location – A CodeLocation instance.
Returns: A list of all predecessors.

function_dependency_graph(func)

Get a dependency graph for the function func.

Parameters: func – The Function object in CFG.function_manager.
Returns: A networkx.DiGraph instance.

data_sub_graph(pv, simplified=True, killing_edges=False, excluding_types=None)

Get a subgraph from the data graph or the simplified data graph that starts from node pv.

Parameters

pv (ProgramVariable) – The starting point of the subgraph.
simplified (bool) – When True, the simplified data graph is used, otherwise the data graph is used.
killing_edges (bool) – Are killing edges included or not.
excluding_types (iterable) – Excluding edges whose types are among those excluded types.

Returns

A subgraph.

Return type

networkx.MultiDiGraph

find_definitions(variable, location=None, simplified_graph=True)

Find all definitions of the given variable.

Parameters

variable (SimVariable) –
simplified_graph (bool) – True if you just want to search in the simplified graph instead of the normal graph. Usually the simplified graph suffices for finding definitions of register or memory variables.

Returns

A collection of all variable definitions to the specific variable.

Return type

list

find_consumers(var_def, simplified_graph=True)

Find all consumers to the specified variable definition.

Parameters

var_def (ProgramVariable) – The variable definition.
simplified_graph (bool) – True if we want to search in the simplified graph, False otherwise.

Returns

A collection of all consumers to the specified variable definition.

Return type

list

find_killers(var_def, simplified_graph=True)

Find all killers to the specified variable definition.

Parameters

var_def (ProgramVariable) – The variable definition.
simplified_graph (bool) – True if we want to search in the simplified graph, False otherwise.

Returns

A collection of all killers to the specified variable definition.

Return type

list

find_sources(var_def, simplified_graph=True)

Find all sources to the specified variable definition.

Parameters

var_def (ProgramVariable) – The variable definition.
simplified_graph (bool) – True if we want to search in the simplified graph, False otherwise.

Returns

A collection of all sources to the specified variable definition.

Return type

list

class angr.analyses.flirt.FlirtAnalysis(sig: Optional[Union[angr.flirt.FlirtSignature, str]] = None)

Bases: angr.analyses.analysis.Analysis

FlirtAnalysis accomplishes two purposes:

If a FLIRT signature file is specified, it will match the given signature file against the current binary and rename recognized functions accordingly.
If no FLIRT signature file is specified, it will use strings to determine possible libraries embedded in the current binary, and then match all possible signatures for the architecture.

class angr.engines.light.data.ArithmeticExpression(op, operands)

Bases: object

Add = 0

Sub = 1

Or = 2

And = 4

RShift = 8

LShift = 16

Mul = 32

Xor = 64

CONST_TYPES = (<class 'int'>, <class 'ailment.expression.Const'>)

op

operands

static try_unpack_const(expr)

class angr.engines.light.data.RegisterOffset(bits, reg, offset)

Bases: object

reg

offset

property bits

property symbolic

class angr.engines.light.data.SpOffset(bits, offset, is_base=False)

Bases: angr.engines.light.data.RegisterOffset

is_base

class angr.engines.light.engine.SimEngineLightMixin(*args, logger=None, **kwargs)

Bases: object

sp_offset(offset: int)

extract_offset_to_sp(spoffset_expr: claripy.ast.base.Base) → Optional[int]

Extract the offset to the original stack pointer.

Parameters: spoffset_expr – The claripy AST to parse.
Returns: The offset to the original stack pointer, or None if spoffset_expr is not a supported type of SpOffset expression.

class angr.engines.light.engine.SimEngineLight

Bases: angr.engines.light.engine.SimEngineLightMixin, angr.engines.engine.SimEngine

process(state, *args, **kwargs)

class angr.engines.light.engine.SimEngineLightVEXMixin(*args, logger=None, **kwargs): Bases: angr.engines.light.engine.SimEngineLightMixin

class angr.engines.light.engine.SimEngineLightAILMixin(*args, logger=None, **kwargs): Bases: angr.engines.light.engine.SimEngineLightMixin

angr.engines.light.engine.SimEngineLightVEX: alias of angr.engines.light.engine.SimEngineLightVEXMixin

angr.engines.light.engine.SimEngineLightAIL: alias of angr.engines.light.engine.SimEngineLightAILMixin

class angr.analyses.propagator.values.Top(size)

Bases: object

size

property bits

class angr.analyses.propagator.values.Bottom: Bases: object

class angr.analyses.propagator.vex_vars.VEXVariable: Bases: object

class angr.analyses.propagator.vex_vars.VEXMemVar(addr, size)

Bases: object

addr

size

class angr.analyses.propagator.vex_vars.VEXReg(offset, size)

Bases: angr.analyses.propagator.vex_vars.VEXVariable

offset

size

class angr.analyses.propagator.vex_vars.VEXTmp(tmp)

Bases: angr.analyses.propagator.vex_vars.VEXVariable

tmp

class angr.analyses.propagator.call_expr_finder.CallExprFinder

Bases: angr.analyses.decompiler.ailblock_walker.AILBlockWalker

Walks an AIL expression to find if it contains a call expression anywhere.

class angr.analyses.propagator.engine_base.SimEnginePropagatorBase(stack_pointer_tracker=None, project=None, propagate_tmps=True, arch=None)

Bases: angr.engines.light.engine.SimEngineLight

process(state, *args, **kwargs)

class angr.analyses.propagator.engine_vex.SimEnginePropagatorVEX(stack_pointer_tracker=None, project=None, propagate_tmps=True, arch=None)

Bases: angr.analyses.propagator.top_checker_mixin.TopCheckerMixin, angr.engines.light.engine.SimEngineLightVEXMixin, angr.analyses.propagator.engine_base.SimEnginePropagatorBase

state: PropagatorVEXState

class angr.analyses.propagator.engine_ail.SimEnginePropagatorAIL(stack_pointer_tracker=None, project=None, propagate_tmps=True, arch=None)

Bases: angr.engines.light.engine.SimEngineLightAILMixin, angr.analyses.propagator.engine_base.SimEnginePropagatorBase

The AIl engine for Propagator.

state: PropagatorAILState

extract_offset_to_sp(expr: Union[claripy.ast.base.Base, ailment.expression.StackBaseOffset]) → Optional[int]

is_using_outdated_def(expr: ailment.expression.Expression, avoid: Optional[ailment.expression.Expression] = None) → bool

class angr.analyses.propagator.outdated_definition_walker.OutdatedDefinitionWalker(expr, state: PropagatorAILState, avoid: Optional[ailment.expression.Expression] = None)

Bases: angr.analyses.decompiler.ailblock_walker.AILBlockWalker

Walks an AIL expression to find outdated definitions.

class angr.analyses.propagator.propagator.PropagatorState(arch, project=None, replacements=None, only_consts=False, prop_count=None, equivalence=None, store_tops=True)

Bases: object

Describes the base state used in Propagator.

arch

gpr_size

project

static top(bits: int) → claripy.ast.base.Base

Get a TOP value.

Parameters: size – Width of the TOP value (in bits).
Returns: The TOP value.

static is_top(expr) → bool

Check if the given expression is a TOP value.

Parameters: expr – The given expression.
Returns: True if the expression is TOP, False otherwise.

copy() → angr.analyses.propagator.propagator.PropagatorState

merge(*others)

add_replacement(codeloc, old, new)

Add a replacement record: Replacing expression old with new at program location codeloc. If the self._only_consts flag is set to true, only constant values will be set.

Parameters

codeloc (CodeLocation) – The code location.
old – The expression to be replaced.
new – The expression to replace with.

Returns

None

filter_replacements()

class angr.analyses.propagator.propagator.PropagatorVEXState(arch, project=None, registers=None, local_variables=None, replacements=None, only_consts=False, prop_count=None, do_binops=True, store_tops=True)

Bases: angr.analyses.propagator.propagator.PropagatorState

Describes the state used in the VEX engine of Propagator.

do_binops

copy() → angr.analyses.propagator.propagator.PropagatorVEXState

merge(*others: angr.analyses.propagator.propagator.PropagatorVEXState) → Tuple[angr.analyses.propagator.propagator.PropagatorVEXState, bool]

store_local_variable(offset, size, value, endness)

load_local_variable(offset, size, endness)

store_register(offset, size, value)

load_register(offset, size)

class angr.analyses.propagator.propagator.Equivalence(codeloc, atom0, atom1)

Bases: object

Describes an equivalence relationship between two atoms.

codeloc

atom0

atom1

class angr.analyses.propagator.propagator.PropagatorAILState(arch, project=None, replacements=None, only_consts=False, prop_count=None, equivalence=None, stack_variables=None, registers=None)

Bases: angr.analyses.propagator.propagator.PropagatorState

Describes the state used in the AIL engine of Propagator.

last_store

copy()

merge(*others) → Tuple[angr.analyses.propagator.propagator.PropagatorAILState, bool]

store_temp(tmp_idx: int, value: angr.analyses.propagator.prop_value.PropValue)

load_tmp(tmp_idx: int) → Optional[angr.analyses.propagator.prop_value.PropValue]

store_register(reg: ailment.expression.Register, value: angr.analyses.propagator.prop_value.PropValue) → None

store_stack_variable(sp_offset: int, new: angr.analyses.propagator.prop_value.PropValue, endness=None) → None

load_register(reg: ailment.expression.Register) → Optional[angr.analyses.propagator.prop_value.PropValue]

load_stack_variable(sp_offset: int, size, endness=None) → Optional[angr.analyses.propagator.prop_value.PropValue]

add_replacement(codeloc, old, new)

filter_replacements()

add_equivalence(codeloc, old, new)

class angr.analyses.propagator.propagator.PropagatorAnalysis(func=None, block=None, func_graph=None, base_state=None, max_iterations=3, load_callback=None, stack_pointer_tracker=None, only_consts=False, completed_funcs=None, do_binops=True, store_tops=True, vex_cross_insn_opt=False)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis, angr.analyses.analysis.Analysis

PropagatorAnalysis propagates values (either constant values or variables) and expressions inside a block or across a function.

PropagatorAnalysis supports both VEX and AIL. The VEX propagator only performs constant propagation. The AIL propagator performs both constant propagation and copy propagation of depth-N expressions.

PropagatorAnalysis performs certain arithmetic operations between constants, including but are not limited to:

addition
subtraction
multiplication
division
xor

It also performs the following memory operations:

Loading values from a known address
Writing values to a stack variable

class angr.analyses.propagator.prop_value.Detail(size: int, expr: Optional[ailment.expression.Expression], def_at: CodeLocation)

Bases: object

A companion class used together with PropValue. It describes stored information at each offset (in bytes).

size

expr

def_at

class angr.analyses.propagator.prop_value.PropValue(value: claripy.ast.bits.Bits, offset_and_details: Optional[Dict[int, angr.analyses.propagator.prop_value.Detail]] = None)

Bases: object

Describes immutable basic value type that is used in Propagator.

value

offset_and_details

property needs_details

property one_expr: Optional[ailment.expression.Expression]: Get the expression that starts at offset 0 and covers the entire PropValue. Returns None if there are no expressions or multiple expressions.

property one_defat: Optional[CodeLocation]: Get the definition location of the expression that starts at offset 0 and covers the entire PropValue. Returns None if there are no expressions or multiple expressions.

to_label()

with_details(size: int, expr: ailment.expression.Expression, def_at: CodeLocation) → PropValue

all_exprs() → Generator[ailment.expression.Expression, None, None]

non_zero_exprs() → Generator[ailment.expression.Expression, None, None]

static chop_value(value: claripy.ast.bits.Bits, begin_offset, end_offset) → claripy.ast.bits.Bits

value_and_labels() → Generator[Tuple[int, claripy.ast.bits.Bits, int, Optional[Dict]], None, None]

static from_value_and_labels(value: claripy.ast.bits.Bits, labels: Iterable[Tuple[int, int, int, Dict[str, Any]]]) → angr.analyses.propagator.prop_value.PropValue

static from_value_and_details(value: claripy.ast.bits.Bits, size: int, expr: ailment.expression.Expression, def_at: CodeLocation)

static extract_ail_expression(start: int, bits: int, expr: Optional[ailment.expression.Expression]) → Optional[ailment.expression.Expression]

static extend_ail_expression(bits: int, expr: Optional[ailment.expression.Expression]) → Optional[ailment.expression.Expression]

class angr.analyses.propagator.top_checker_mixin.TopCheckerMixin(*args, logger=None, **kwargs): Bases: angr.engines.light.engine.SimEngineLightMixin

angr.analyses.reaching_definitions.get_all_definitions(region: MultiValuedMemory) → Set[Definition]

class angr.analyses.reaching_definitions.call_trace.CallSite(caller_func_addr: int, block_addr: Optional[int], callee_func_addr: int)

Bases: object

Describes a call site on a CFG.

caller_func_addr

callee_func_addr

block_addr

class angr.analyses.reaching_definitions.call_trace.CallTrace(target: int)

Bases: object

Describes a series of functions calls to get from one function (current_function_address()) to another function or a basic block (self.target).

target

callsites: List[angr.analyses.reaching_definitions.call_trace.CallSite]

current_function_address() → int

step_back(caller_func_addr: int, block_addr: Optional[int], callee_func_addr) → angr.analyses.reaching_definitions.call_trace.CallTrace

includes_function(func_addr: int) → bool

copy() → angr.analyses.reaching_definitions.call_trace.CallTrace

class angr.analyses.reaching_definitions.engine_vex.SimEngineRDVEX(project, call_stack, maximum_local_call_depth, functions=None, function_handler=None)

Bases: angr.engines.light.engine.SimEngineLightVEXMixin, angr.engines.light.engine.SimEngineLight

Implements the VEX execution engine for reaching definition analysis.

process(state, *args, **kwargs)

class angr.analyses.reaching_definitions.reaching_definitions.ReachingDefinitionsAnalysis(subject: Union[angr.analyses.reaching_definitions.subject.Subject, ailment.block.Block, angr.block.Block, angr.knowledge_plugins.functions.function.Function] = None, func_graph=None, max_iterations=3, track_tmps=False, track_calls=None, track_consts=False, observation_points: Iterable[ObservationPoint] = None, init_state: angr.analyses.reaching_definitions.rd_state.ReachingDefinitionsState = None, cc=None, function_handler: Optional[FunctionHandler] = None, call_stack: Optional[List[int]] = None, maximum_local_call_depth=5, observe_all=False, visited_blocks=None, dep_graph: Optional[DepGraph] = None, observe_callback=None, canonical_size=8)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis[angr.analyses.reaching_definitions.rd_state.ReachingDefinitionsState, NodeType], angr.analyses.analysis.Analysis

ReachingDefinitionsAnalysis is a text-book implementation of a static data-flow analysis that works on either a function or a block. It supports both VEX and AIL. By registering observers to observation points, users may use this analysis to generate use-def chains, def-use chains, and reaching definitions, and perform other traditional data-flow analyses such as liveness analysis.

I’ve always wanted to find a better name for this analysis. Now I gave up and decided to live with this name for the foreseeable future (until a better name is proposed by someone else).
Aliasing is definitely a problem, and I forgot how aliasing is resolved in this implementation. I’ll leave this as a post-graduation TODO.
Some more documentation and examples would be nice.

Parameters

subject – The subject of the analysis: a function, or a single basic block
func_graph – Alternative graph for function.graph.
max_iterations (int) – The maximum number of iterations before the analysis is terminated.
track_tmps (bool) – Whether or not temporary variables should be taken into consideration during the analysis.
track_calls (bool) – Whether or not calls will show up as elements in the def-use graph.
observation_points (iterable) – A collection of tuples of (“node”|”insn”, ins_addr, OP_TYPE) defining where reaching definitions should be copied and stored. OP_TYPE can be OP_BEFORE or OP_AFTER.
init_state – An optional initialization state. The analysis creates and works on a copy. Default to None: the analysis then initialize its own abstract state, based on the given <Subject>.
cc (SimCC) – Calling convention of the function.
function_handler (FunctionHandler) – The function handler to update the analysis state and results on function calls.
call_stack – An ordered list of Function addresses representing the call stack leading to the analysed subject, from older to newer calls. Setting it to None to limit the analysis to a single function and disable call stack tracking; In that case, all contexts in CodeLocation will be None, which makes CodeLocation objects contextless.
maximum_local_call_depth (int) – Maximum local function recursion depth.
observe_all (Boolean) – Observe every statement, both before and after.
visited_blocks – A set of previously visited blocks.
dep_graph – An initial dependency graph to add the result of the analysis to. Set it to None to skip dependency graph generation.
canonical_size – The sizes (in bytes) that objects with an UNKNOWN_SIZE are treated as for operations where sizes are necessary.

property observed_results: Dict[Tuple[str, int, int], angr.knowledge_plugins.key_definitions.live_definitions.LiveDefinitions]

property all_definitions

property all_uses

property one_result

property dep_graph

property visited_blocks

get_reaching_definitions(**kwargs)

get_reaching_definitions_by_insn(ins_addr, op_type)

get_reaching_definitions_by_node(node_addr, op_type)

node_observe(node_addr: int, state: angr.analyses.reaching_definitions.rd_state.ReachingDefinitionsState, op_type: angr.knowledge_plugins.key_definitions.constants.ObservationPointType) → None

Parameters

node_addr – Address of the node.
state – The analysis state.
op_type – Type of the bbservation point. Must be one of the following: OP_BEFORE, OP_AFTER.

insn_observe(insn_addr: int, stmt: Union[ailment.statement.Statement, pyvex.stmt.IRStmt], block: Union[angr.block.Block, ailment.block.Block], state: angr.analyses.reaching_definitions.rd_state.ReachingDefinitionsState, op_type: angr.knowledge_plugins.key_definitions.constants.ObservationPointType) → None

Parameters

insn_addr – Address of the instruction.
stmt – The statement.
block – The current block.
state – The abstract analysis state.
op_type – Type of the observation point. Must be one of the following: OP_BEORE, OP_AFTER.

property subject

class angr.analyses.reaching_definitions.dep_graph.DepGraph(graph: Optional[networkx.classes.digraph.DiGraph] = None)

Bases: object

The representation of a dependency graph: a directed graph, where nodes are definitions, and edges represent uses.

Mostly a wrapper around a <networkx.DiGraph>.

Parameters: graph – A graph where nodes are definitions, and edges represent uses.

property graph: networkx.classes.digraph.DiGraph

add_node(node: angr.knowledge_plugins.key_definitions.definition.Definition) → None

Parameters: node – The definition to add to the definition-use graph.

add_edge(source: angr.knowledge_plugins.key_definitions.definition.Definition, destination: angr.knowledge_plugins.key_definitions.definition.Definition, **labels) → None

The edge to add to the definition-use graph. Will create nodes that are not yet present.

Parameters

source – The “source” definition, used by the “destination”.
destination – The “destination” definition, using the variable defined by “source”.
labels – Optional keyword arguments to represent edge labels.

nodes() → networkx.classes.reportviews.NodeView

predecessors(node: angr.knowledge_plugins.key_definitions.definition.Definition) → networkx.classes.reportviews.NodeView

Parameters: node – The definition to get the predecessors of.

transitive_closure(definition: angr.knowledge_plugins.key_definitions.definition.Definition) → networkx.classes.digraph.DiGraph

Compute the “transitive closure” of a given definition. Obtained by transitively aggregating the ancestors of this definition in the graph.

Note: Each definition is memoized to avoid any kind of recomputation across the lifetime of this object.

Parameters: definition – The Definition to get transitive closure for.
Returns: A graph of the transitive closure of the given definition.

contains_atom(atom: angr.knowledge_plugins.key_definitions.atoms.Atom) → bool

add_dependencies_for_concrete_pointers_of(values: Iterable[Union[claripy.ast.base.Base, int]], definition: angr.knowledge_plugins.key_definitions.definition.Definition, cfg: angr.knowledge_plugins.cfg.cfg_model.CFGModel, loader: cle.loader.Loader)

When a given definition holds concrete pointers, make sure the <MemoryLocation>s they point to are present in the dependency graph; Adds them if necessary.

Parameters

values –
definition – The definition which has data that can contain concrete pointers.
cfg – The CFG, containing information about memory data.
loader –

class angr.analyses.reaching_definitions.heap_allocator.HeapAllocator(canonical_size: int)

Bases: object

A simple modelisation to help represent heap memory management during a <ReachingDefinitionsAnalysis>: - Act as if allocations were always done in consecutive memory segments; - Take care of the size not to screw potential pointer arithmetic (avoid overlapping segments).

The content of the heap itself is modeled using a <KeyedRegion> attribute in the <LiveDefinitions> state; This class serves to generate consistent heap addresses to be used by the aforementionned.

Note: This has NOT been made to help detect heap vulnerabilities.

Parameters: canonical_size – The concrete size an <UNKNOWN_SIZE> defaults to.

allocate(size: Union[int, angr.knowledge_plugins.key_definitions.unknown_size.UnknownSize]) → angr.knowledge_plugins.key_definitions.heap_address.HeapAddress

Gives an address for a new memory chunck of <size> bytes.

Parameters: size – The requested size for the chunck, in number of bytes.
Returns: The address of the chunck.

free(address: Union[angr.knowledge_plugins.key_definitions.undefined.Undefined, angr.knowledge_plugins.key_definitions.heap_address.HeapAddress])

Mark the chunck pointed by <address> as freed.

Parameters: address – The address of the chunck to free.

property allocated_addresses

The list of addresses that are currently allocated on the heap.

Type: return

class angr.analyses.reaching_definitions.function_handler.FunctionHandler

Bases: object

An abstract base class for function handlers.

To work properly, we expect function handlers to:

Be related to a <ReachingDefinitionsAnalysis>;
Provide a handle_local_function method.

hook(analysis) → angr.analyses.reaching_definitions.function_handler.FunctionHandler

A <FunctionHandler> needs information about the context in which it is executed. A <ReachingDefinitionsAnalysis> would “hook” into a handler by calling: <FunctionHandler>.hook(self).

Parameters: analysis (angr.analyses.ReachingDefinitionsAnalysis) – A RDA using this <FunctionHandler>.
Return FunctionHandler

handle_local_function(state: ReachingDefinitionsState, function_address: int, call_stack: Optional[List], maximum_local_call_depth: int, visited_blocks: Set[int], dep_graph: DepGraph, src_ins_addr: Optional[int] = None, codeloc: Optional[CodeLocation] = None) → Tuple[bool, ReachingDefinitionsState, Set[int], DepGraph]

Parameters

state – The state at the entry of the function, i.e. the function’s input state.
function_address – The address of the function to handle.
call_stack –
maximum_local_call_depth –
visited_blocks – A set of the addresses of the previously visited blocks.
dep_graph – A definition-use graph, where nodes represent definitions, and edges represent uses.
codeloc – The code location of the call to the analysed function.

handle_unknown_call(state: ReachingDefinitionsState, src_codeloc: Optional[CodeLocation] = None) → Tuple[bool, ReachingDefinitionsState]: Called when the RDA encounters a function call to somewhere really weird. E.g. the function address was invalid (not even TOP), or the address of the function is outside of the main object, but also not a known symbol :param state: :param src_codeloc: :return:

handle_indirect_call(state: ReachingDefinitionsState, src_codeloc: Optional[CodeLocation] = None) → Tuple[bool, ReachingDefinitionsState]: The RDA encountered a function call with multiple possible values, or TOP as a target :param state: :param src_codeloc: :return:

handle_external_function_fallback(state: ReachingDefinitionsState, src_codeloc: Optional[CodeLocation] = None) → Tuple[bool, ReachingDefinitionsState]: Fallback for a call to an external function, that has no specific implementation :param state: :param src_codeloc: :return:

handle_external_function_symbol(state: ReachingDefinitionsState, symbol: cle.backends.symbol.Symbol, src_codeloc: Optional[CodeLocation] = None) → Tuple[bool, ReachingDefinitionsState]: The generic handler for external functions with a known symbol This is different from The default behavior using hasattr/getattr supports existing code, but you can also implement the check if the external function is supported in another way, e.g. similar to SimProcedures :param state: :param symbol: :param src_codeloc: :return:

handle_external_function_name(state: ReachingDefinitionsState, ext_func_name: str, src_codeloc: Optional[CodeLocation] = None) → Tuple[bool, ReachingDefinitionsState]

class angr.analyses.reaching_definitions.rd_state.ReachingDefinitionsState(arch: archinfo.arch.Arch, subject: angr.analyses.reaching_definitions.subject.Subject, track_tmps: bool = False, track_calls: bool = False, track_consts: bool = False, analysis: Optional[ReachingDefinitionsAnalysis] = None, rtoc_value=None, live_definitions: Optional[angr.knowledge_plugins.key_definitions.live_definitions.LiveDefinitions] = None, canonical_size: int = 8, heap_allocator: angr.analyses.reaching_definitions.heap_allocator.HeapAllocator = None, environment: angr.knowledge_plugins.key_definitions.environment.Environment = None)

Bases: object

Represents the internal state of the ReachingDefinitionsAnalysis.

It contains a data class LiveDefinitions, which stores both definitions and uses for register, stack, memory, and temporary variables, uncovered during the analysis.

Parameters

subject – The subject being analysed.
track_tmps – Only tells whether or not temporary variables should be taken into consideration when representing the state of the analysis. Should be set to true when the analysis has counted uses and definitions for temporary variables, false otherwise.
analysis – The analysis that generated the state represented by this object.
rtoc_value – When the targeted architecture is ppc64, the initial function needs to know the rtoc_value.
live_definitions –
canonical_size – The sizes (in bytes) that objects with an UNKNOWN_SIZE are treated as for operations where sizes are necessary.
heap_allocator – Mechanism to model the management of heap memory.
environment – Representation of the environment of the analysed program.

Variables

arch – The architecture targeted by the program.

arch

analysis

live_definitions

all_definitions: Set[angr.knowledge_plugins.key_definitions.definition.Definition]

heap_allocator

current_codeloc: Optional[angr.code_location.CodeLocation]

codeloc_uses: Set[angr.knowledge_plugins.key_definitions.definition.Definition]

top(bits: int)

is_top(*args)

heap_address(offset: int) → claripy.ast.base.Base

static is_heap_address(addr: claripy.ast.base.Base) → bool

static get_heap_offset(addr: claripy.ast.base.Base) → Optional[int]

stack_address(offset: int) → claripy.ast.base.Base

is_stack_address(addr: claripy.ast.base.Base) → bool

get_stack_offset(addr: claripy.ast.base.Base) → Optional[int]

annotate_with_def(symvar: claripy.ast.base.Base, definition: angr.knowledge_plugins.key_definitions.definition.Definition)

Parameters

symvar –
definition –

Returns

extract_defs(symvar: claripy.ast.base.Base) → Generator[angr.knowledge_plugins.key_definitions.definition.Definition, None, None]

property tmp_definitions

property tmp_uses

property register_uses

property register_definitions

property stack_definitions

property stack_uses

property heap_definitions

property heap_uses

property memory_uses

property memory_definitions

property uses_by_codeloc

get_sp() → int

get_stack_address(offset: claripy.ast.base.Base) → int

property environment

property dep_graph

copy() → angr.analyses.reaching_definitions.rd_state.ReachingDefinitionsState

merge(*others) → Tuple[angr.analyses.reaching_definitions.rd_state.ReachingDefinitionsState, bool]

kill_definitions(atom: angr.knowledge_plugins.key_definitions.atoms.Atom) → None

Overwrite existing definitions w.r.t ‘atom’ with a dummy definition instance. A dummy definition will not be removed during simplification.

Parameters

atom –
code_loc (CodeLocation) –
data (object) –

Returns

None

kill_and_add_definition(atom: angr.knowledge_plugins.key_definitions.atoms.Atom, code_loc: angr.code_location.CodeLocation, data: angr.storage.memory_mixins.paged_memory.pages.multi_values.MultiValues, dummy=False, tags: Optional[Set[angr.knowledge_plugins.key_definitions.tag.Tag]] = None, endness=None, annotated: bool = False) → Optional[angr.storage.memory_mixins.paged_memory.pages.multi_values.MultiValues]

add_use(atom: angr.knowledge_plugins.key_definitions.atoms.Atom, code_loc) → None

add_use_by_def(definition: angr.knowledge_plugins.key_definitions.definition.Definition, code_loc: angr.code_location.CodeLocation) → None

get_definitions(atom: angr.knowledge_plugins.key_definitions.atoms.Atom) → Iterable[angr.knowledge_plugins.key_definitions.definition.Definition]

mark_guard(code_loc: angr.code_location.CodeLocation, target)

mark_call(code_loc: angr.code_location.CodeLocation, target)

mark_const(code_loc: angr.code_location.CodeLocation, const)

class angr.analyses.reaching_definitions.subject.SubjectType(value)

Bases: enum.Enum

An enumeration.

Function = 1

Block = 2

CallTrace = 3

class angr.analyses.reaching_definitions.subject.Subject(content, func_graph=None, cc=None)

Bases: object

The thing being analysed, and the way (visitor) to analyse it.

Parameters

content (Union[ailment.Block, angr.Block, Function]) – Thing to be analysed.
func_graph (networkx.DiGraph) – Alternative graph for function.graph.
cc (SimCC) – Calling convention of the function.

property cc

property content

property func_graph

property type

property visitor: Union[angr.analyses.forward_analysis.visitors.function_graph.FunctionGraphVisitor, angr.analyses.forward_analysis.visitors.single_node_graph.SingleNodeGraphVisitor]

class angr.analyses.reaching_definitions.engine_ail.SimEngineRDAIL(project, call_stack, maximum_local_call_depth, function_handler: Optional[angr.analyses.reaching_definitions.function_handler.FunctionHandler] = None)

Bases: angr.engines.light.engine.SimEngineLightAILMixin, angr.engines.light.engine.SimEngineLight

arch: archinfo.arch.Arch

state: angr.analyses.reaching_definitions.rd_state.ReachingDefinitionsState

process(state, *args, **kwargs)

sp_offset(offset: int)

class angr.analyses.reaching_definitions.external_codeloc.ExternalCodeLocation: Bases: angr.code_location.CodeLocation

class angr.analyses.cfg_slice_to_sink.cfg_slice_to_sink.CFGSliceToSink(target, transitions=None)

Bases: object

The representation of a slice of a CFG.

Parameters

target (angr.knowledge_plugins.functions.function.Function) – The targeted sink, to which every path in the slice leads.
transitions (Dict[int,List[int]]) – A mapping representing transitions in the graph. Indexes are source addresses and values a list of destination addresses, for which there exists a transition in the slice from source to destination.

property transitions

The transitions in the slice.

Type: return Dict[int,List[int]]

property transitions_as_tuples

The list of transitions as pairs of (source, destination).

Type: return List[Tuple[int,int]]

property target: return angr.knowledge_plugins.functions.function.Function: The targeted sink function, from which the slice is constructed.

property nodes: List[int]

The complete list of addresses present in the slice.

Type: return

property entrypoints

Entrypoints are all source addresses that are not the destination address of any transition.

Return List[int]: The list of entrypoints addresses.

add_transitions(transitions)

Add the given transitions to the current slice.

Parameters: transitions (Dict[int,List[int]]) – The list of transitions to be added to self.transitions.
Return Dict[int,List[int]]: Return the updated list of transitions.

is_empty()

Test if a given slice does not contain any transition.

Return bool: True if the <CFGSliceToSink> instance does not contain any transitions. False otherwise.

path_between(source: int, destination: int, visited: Optional[Set[Any]] = None) → bool

Check the existence of a path in the slice between two given node adresses.

Parameters

source – The source address.
destination – The destination address.
visited – Used to avoid infinite recursion if loops are present in the slice.

Returns

True if there is a path between the source and the destination in the CFG, False if not, or if we have been unable to decide (because of loops).

angr.analyses.cfg_slice_to_sink.graph.slice_callgraph(callgraph, cfg_slice_to_sink)

Slice a callgraph, keeping only the nodes present in the <CFGSliceToSink> representation, and th transitions for which a path exists.

Note that this function mutates the graph passed as an argument.

Parameters

callgraph (networkx.MultiDiGraph) – The callgraph to update.
cfg_slice_to_sink (CFGSliceToSink) – The representation of the slice, containing the data to update the callgraph from.

angr.analyses.cfg_slice_to_sink.graph.slice_cfg_graph(graph, cfg_slice_to_sink)

Slice a CFG graph, keeping only the transitions and nodes present in the <CFGSliceToSink> representation.

Note that this function mutates the graph passed as an argument.

Parameters

graph (networkx.DiGraph) – The graph to slice.
cfg_slice_to_sink (CFGSliceToSink) – The representation of the slice, containing the data to update the CFG from.

Return networkx.DiGraph

The sliced graph.

angr.analyses.cfg_slice_to_sink.graph.slice_function_graph(function_graph, cfg_slice_to_sink)

Slice a function graph, keeping only the nodes present in the <CFGSliceToSink> representation.

Because the <CFGSliceToSink> is build from the CFG, and the function graph is NOT a subgraph of the CFG, edges of the function graph will no be present in the <CFGSliceToSink> transitions. However, we use the fact that if there is an edge between two nodes in the function graph, then there must exist a path between these two nodes in the slice; Proof idea: - The <CFGSliceToSink> is backward and recursively constructed; - If a node is in the slice, then all its predecessors will be (transitively); - If there is an edge between two nodes in the function graph, there is a path between them in the CFG; - So: The origin node is a transitive predecessor of the destination one, hence if destination is in the slice, then origin will be too.

In consequence, in the end, removing the only nodes not present in the slice, and their related transitions gives us the expected result: a function graph representing (a higher view of) the flow in the slice.

Note that this function mutates the graph passed as an argument.

Parameters

graph (networkx.DiGraph) – The graph to slice.
cfg_slice_to_sink (CFGSliceToSink) – The representation of the slice, containing the data to update the CFG from.

Return networkx.DiGraph

The sliced graph.

Some utilitary functions to manage our representation of transitions:: A dictionary, indexed by int (source addresses), which values are list of ints (target addresses).

angr.analyses.cfg_slice_to_sink.transitions.merge_transitions(transitions, existing_transitions)

Merge two dictionaries of transitions together.

Parameters

transitions (Dict[int,List[int]]) – Some transitions.
existing_transitions (Dict[int,List[int]]) – Other transitions.

Return Dict[int,List[int]]

The merge of the two parameters.

class angr.analyses.stack_pointer_tracker.BottomType

Bases: object

The bottom value for register values.

class angr.analyses.stack_pointer_tracker.Constant(val)

Bases: object

Represents a constant value.

val

class angr.analyses.stack_pointer_tracker.Register(offset, bitlen)

Bases: object

Represent a register.

offset

bitlen

class angr.analyses.stack_pointer_tracker.OffsetVal(reg, offset)

Bases: object

Represent a value with an offset added.

property reg

property offset

class angr.analyses.stack_pointer_tracker.FrozenStackPointerTrackerState(regs, memory, is_tracking_memory)

Bases: object

Abstract state for StackPointerTracker analysis with registers and memory values being in frozensets.

regs

memory

is_tracking_memory

unfreeze()

merge(other)

class angr.analyses.stack_pointer_tracker.StackPointerTrackerState(regs, memory, is_tracking_memory)

Bases: object

Abstract state for StackPointerTracker analysis.

regs

memory

is_tracking_memory

give_up_on_memory_tracking()

store(addr, val)

load(addr)

get(reg)

put(reg, val)

copy()

freeze()

merge(other)

exception angr.analyses.stack_pointer_tracker.CouldNotResolveException

Bases: Exception

An exception used in StackPointerTracker analysis to represent internal resolving failures.

class angr.analyses.stack_pointer_tracker.StackPointerTracker(func: angr.knowledge_plugins.functions.function.Function, reg_offsets: Set[int], track_memory=True)

Bases: angr.analyses.analysis.Analysis, angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis

Track the offset of stack pointer at the end of each basic block of a function.

offset_after(addr, reg)

offset_before(addr, reg)

offset_after_block(block_addr, reg)

offset_before_block(block_addr, reg)

property inconsistent

inconsistent_for(reg)

class angr.analyses.variable_recovery.annotations.StackLocationAnnotation(offset)

Bases: claripy.annotation.Annotation

property eliminatable

property relocatable

class angr.analyses.variable_recovery.annotations.VariableSourceAnnotation(block_addr, stmt_idx, ins_addr)

Bases: claripy.annotation.Annotation

property eliminatable

property relocatable

static from_state(state)

angr.analyses.variable_recovery.variable_recovery_base.parse_stack_pointer(sp)

Convert multiple supported forms of stack pointer representations into stack offsets.

Parameters: sp – A stack pointer representation.
Returns: A stack pointer offset.
Return type: int

class angr.analyses.variable_recovery.variable_recovery_base.VariableAnnotation(addr_and_variables: List[Tuple[int, angr.sim_variable.SimVariable]])

Bases: claripy.annotation.Annotation

addr_and_variables

property relocatable

property eliminatable

class angr.analyses.variable_recovery.variable_recovery_base.VariableRecoveryBase(func, max_iterations, store_live_variables: bool)

Bases: angr.analyses.analysis.Analysis

The base class for VariableRecovery and VariableRecoveryFast.

get_variable_definitions(block_addr)

Get variables that are defined at the specified block.

Parameters: block_addr (int) – Address of the block.
Returns: A set of variables.

initialize_dominance_frontiers()

class angr.analyses.variable_recovery.variable_recovery_base.VariableRecoveryStateBase(block_addr, analysis, arch, func, stack_region=None, register_region=None, global_region=None, typevars=None, type_constraints=None, delayed_type_constraints=None, stack_offset_typevars=None, project=None)

Bases: object

The base abstract state for variable recovery analysis.

static top(bits) → claripy.ast.bv.BV

static is_top(thing) → bool

static extract_variables(expr: claripy.ast.base.Base) → Generator[Tuple[int, Union[angr.sim_variable.SimVariable, angr.engines.light.data.SpOffset]], None, None]

static annotate_with_variables(expr: claripy.ast.base.Base, addr_and_variables: Iterable[Tuple[int, Union[angr.sim_variable.SimVariable, angr.engines.light.data.SpOffset]]]) → claripy.ast.base.Base

stack_address(offset: int) → claripy.ast.base.Base

static is_stack_address(addr: claripy.ast.base.Base) → bool

is_global_variable_address(addr: claripy.ast.base.Base) → bool

get_stack_offset(addr: claripy.ast.base.Base) → Optional[int]

stack_addr_from_offset(offset: int) → int

property func_addr

property dominance_frontiers

property variable_manager

property variables

get_variable_definitions(block_addr)

Get variables that are defined at the specified block.

Parameters: block_addr (int) – Address of the block.
Returns: A set of variables.

add_type_constraint(constraint)

Add a new type constraint.

Parameters: constraint –
Returns

downsize() → None

Remove unnecessary members.

Returns: None

static downsize_region(region: angr.storage.memory_mixins.MultiValuedMemory) → angr.storage.memory_mixins.MultiValuedMemory

Get rid of unnecessary references in region so that it won’t avoid garbage collection on those referenced objects.

Parameters: region – A MultiValuedMemory region.
Returns: None

class angr.analyses.variable_recovery.variable_recovery_fast.VariableRecoveryFastState(block_addr, analysis, arch, func, stack_region=None, register_region=None, global_region=None, typevars=None, type_constraints=None, delayed_type_constraints=None, stack_offset_typevars=None, project=None)

Bases: angr.analyses.variable_recovery.variable_recovery_base.VariableRecoveryStateBase

The abstract state of variable recovery analysis.

Variables

stack_region (KeyedRegion) – The stack store.
register_region (KeyedRegion) – The register store.

copy()

merge(others: Tuple[angr.analyses.variable_recovery.variable_recovery_fast.VariableRecoveryFastState], successor=None) → Tuple[angr.analyses.variable_recovery.variable_recovery_fast.VariableRecoveryFastState, bool]

Merge two abstract states.

For any node A whose dominance frontier that the current node (at the current program location) belongs to, we create a phi variable V’ for each variable V that is defined in A, and then replace all existence of V with V’ in the merged abstract state.

Parameters: others – Other abstract states to merge.
Returns: The merged abstract state.

class angr.analyses.variable_recovery.variable_recovery_fast.VariableRecoveryFast(func, func_graph=None, max_iterations=2, low_priority=False, track_sp=True, func_args: Optional[List[angr.sim_variable.SimVariable]] = None, store_live_variables=False)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis, angr.analyses.variable_recovery.variable_recovery_base.VariableRecoveryBase

Recover “variables” from a function by keeping track of stack pointer offsets and pattern matching VEX statements.

If calling conventions are recovered prior to running VariableRecoveryFast, variables can be recognized more accurately. However, it is not a requirement.

Parameters

func (knowledge.Function) – The function to analyze.
max_iterations (int) –
clinic –

class angr.analyses.variable_recovery.variable_recovery.VariableRecoveryState(block_addr, analysis, arch, func, concrete_states, stack_region=None, register_region=None)

Bases: angr.analyses.variable_recovery.variable_recovery_base.VariableRecoveryStateBase

The abstract state of variable recovery analysis.

Variables: variable_manager (angr.knowledge.variable_manager.VariableManager) – The variable manager.

property concrete_states

get_concrete_state(addr)

Parameters: addr –
Returns

copy()

register_callbacks(concrete_states)

Parameters: concrete_states –
Returns

merge(others: Tuple[angr.analyses.variable_recovery.variable_recovery.VariableRecoveryState], successor=None) → Tuple[angr.analyses.variable_recovery.variable_recovery.VariableRecoveryState, bool]

Merge two abstract states.

Parameters: others – Other abstract states to merge.
Returns: The merged abstract state.
Return type: VariableRecoveryState, and a boolean that indicates if any merge has happened.

class angr.analyses.variable_recovery.variable_recovery.VariableRecovery(func, max_iterations=20, store_live_variables=False)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis, angr.analyses.variable_recovery.variable_recovery_base.VariableRecoveryBase

Recover “variables” from a function using forced execution.

While variables play a very important role in programming, it does not really exist after compiling. However, we can still identify and recovery their counterparts in binaries. It is worth noting that not every variable in source code can be identified in binaries, and not every recognized variable in binaries have a corresponding variable in the original source code. In short, there is no guarantee that the variables we identified/recognized in a binary are the same variables in its source code.

This analysis uses heuristics to identify and recovers the following types of variables: - Register variables. - Stack variables. - Heap variables. (not implemented yet) - Global variables. (not implemented yet)

This analysis takes a function as input, and performs a data-flow analysis on nodes. It runs concrete execution on every statement and hooks all register/memory accesses to discover all places that are accessing variables. It is slow, but has a more accurate analysis result. For a fast but inaccurate variable recovery, you may consider using VariableRecoveryFast.

This analysis follows SSA, which means every write creates a new variable in registers or memory (statck, heap, etc.). Things may get tricky when overlapping variable (in memory, as you cannot really have overlapping accesses to registers) accesses exist, and in such cases, a new variable will be created, and this new variable will overlap with one or more existing varaibles. A decision procedure (which is pretty much TODO) is required at the end of this analysis to resolve the conflicts between overlapping variables.

Parameters: func (knowledge.Function) – The function to analyze.

class angr.analyses.variable_recovery.engine_ail.SimEngineVRAIL(*args, **kwargs)

Bases: angr.engines.light.engine.SimEngineLightAILMixin, angr.analyses.variable_recovery.engine_base.SimEngineVRBase

state: VariableRecoveryFastState

class angr.analyses.variable_recovery.engine_vex.SimEngineVRVEX(project, kb)

Bases: angr.engines.light.engine.SimEngineLightVEXMixin, angr.analyses.variable_recovery.engine_base.SimEngineVRBase

Implements the VEX engine for variable recovery analysis.

state: VariableRecoveryStateBase

class angr.analyses.variable_recovery.engine_base.RichR(data: claripy.ast.base.Base, variable=None, typevar: Optional[angr.analyses.typehoon.typevars.TypeVariable] = None, type_constraints=None)

Bases: object

A rich representation of calculation results. The variable recovery data domain.

data: claripy.ast.base.Base

variable

typevar

type_constraints

property bits

class angr.analyses.variable_recovery.engine_base.SimEngineVRBase(project, kb)

Bases: angr.engines.light.engine.SimEngineLight

The base class for variable recovery analyses. Contains methods for basic interactions with the state, like loading and storing data.

state: VariableRecoveryStateBase

property func_addr

process(state, *args, **kwargs)

class angr.analyses.typehoon.lifter.TypeLifter(bits: int)

Bases: object

Lift SimTypes to type constants.

bits

lift(ty: angr.sim_type.SimType)

class angr.analyses.typehoon.simple_solver.RecursiveType(typevar, offset): Bases: object

class angr.analyses.typehoon.simple_solver.SimpleSolver(bits: int, constraints)

Bases: object

SimpleSolver is, literally, a simple, unification-based type constraint solver.

solve()

determine()

class angr.analyses.typehoon.translator.SimTypeTempRef(typevar)

Bases: angr.sim_type.SimType

c_repr()

class angr.analyses.typehoon.translator.TypeTranslator(arch=None)

Bases: object

Translate type variables to SimType equivalence.

struct_name()

tc2simtype(tc)

simtype2tc(simtype: angr.sim_type.SimType) → angr.analyses.typehoon.typeconsts.TypeConstant

backpatch(st, translated)

Parameters

st (sim_type.SimType) –
translated (dict) –

Returns

class angr.analyses.typehoon.typevars.TypeConstraint

Bases: object

pp_str(mapping: Dict[angr.analyses.typehoon.typevars.TypeVariable, Any]) → str

class angr.analyses.typehoon.typevars.Equivalence(type_a, type_b)

Bases: angr.analyses.typehoon.typevars.TypeConstraint

type_a

type_b

pp_str(mapping: Dict[angr.analyses.typehoon.typevars.TypeVariable, Any]) → str

class angr.analyses.typehoon.typevars.Existence(type_)

Bases: angr.analyses.typehoon.typevars.TypeConstraint

type_

pp_str(mapping: Dict[angr.analyses.typehoon.typevars.TypeVariable, Any]) → str

replace(replacements)

class angr.analyses.typehoon.typevars.Subtype(sub_type, super_type)

Bases: angr.analyses.typehoon.typevars.TypeConstraint

super_type

sub_type

pp_str(mapping: Dict[angr.analyses.typehoon.typevars.TypeVariable, Any]) → str

replace(replacements)

class angr.analyses.typehoon.typevars.Add(type_0, type_1, type_r)

Bases: angr.analyses.typehoon.typevars.TypeConstraint

Describes the constraint that type_r == type0 + type1

type_0

type_1

type_r

pp_str(mapping: Dict[angr.analyses.typehoon.typevars.TypeVariable, Any]) → str

replace(replacements)

class angr.analyses.typehoon.typevars.Sub(type_0, type_1, type_r)

Bases: angr.analyses.typehoon.typevars.TypeConstraint

Describes the constraint that type_r == type0 - type1

type_0

type_1

type_r

pp_str(mapping: Dict[angr.analyses.typehoon.typevars.TypeVariable, Any]) → str

replace(replacements)

class angr.analyses.typehoon.typevars.TypeVariable(idx: Optional[int] = None)

Bases: object

idx: int

pp_str(mapping: Dict[angr.analyses.typehoon.typevars.TypeVariable, Any]) → str

class angr.analyses.typehoon.typevars.DerivedTypeVariable(type_var, label, idx=None)

Bases: angr.analyses.typehoon.typevars.TypeVariable

type_var

label

pp_str(mapping: Dict[angr.analyses.typehoon.typevars.TypeVariable, Any]) → str

replace(replacements)

class angr.analyses.typehoon.typevars.TypeVariables

Bases: object

merge(tvs)

copy()

add_type_variable(var: SimVariable, codeloc, typevar: angr.analyses.typehoon.typevars.TypeVariable)

get_type_variable(var, codeloc)

has_type_variable_for(var: SimVariable, codeloc)

class angr.analyses.typehoon.typevars.BaseLabel: Bases: object

class angr.analyses.typehoon.typevars.FuncIn(loc)

Bases: angr.analyses.typehoon.typevars.BaseLabel

loc

class angr.analyses.typehoon.typevars.FuncOut(loc)

Bases: angr.analyses.typehoon.typevars.BaseLabel

loc

class angr.analyses.typehoon.typevars.Load: Bases: angr.analyses.typehoon.typevars.BaseLabel

class angr.analyses.typehoon.typevars.Store: Bases: angr.analyses.typehoon.typevars.BaseLabel

class angr.analyses.typehoon.typevars.AddN(n)

Bases: angr.analyses.typehoon.typevars.BaseLabel

n

class angr.analyses.typehoon.typevars.SubN(n)

Bases: angr.analyses.typehoon.typevars.BaseLabel

n

class angr.analyses.typehoon.typevars.ConvertTo(to_bits)

Bases: angr.analyses.typehoon.typevars.BaseLabel

to_bits

class angr.analyses.typehoon.typevars.ReinterpretAs(to_type, to_bits)

Bases: angr.analyses.typehoon.typevars.BaseLabel

to_type

to_bits

class angr.analyses.typehoon.typevars.HasField(bits, offset)

Bases: angr.analyses.typehoon.typevars.BaseLabel

bits

offset

class angr.analyses.typehoon.typehoon.Typehoon(constraints, ground_truth=None, var_mapping: Optional[Dict[SimVariable, TypeVariable]] = None, prioritize_char_array_over_struct: bool = True, must_struct: Optional[Set[TypeVariable]] = None)

Bases: angr.analyses.analysis.Analysis

A spiritual tribute to the long-standing typehoon project that @jmg (John Grosen) worked on during his days in the angr team. Now I feel really bad of asking the poor guy to work directly on VEX IR without any fancy static analysis support as we have right now…

Typehoon analysis implements a pushdown system that simplifies and solves type constraints. Our type constraints are largely an implementation of the paper Polymorphic Type Inference for Machine Code by Noonan, Loginov, and Cok from GrammaTech (with missing functionality support and bugs, of course). Type constraints are collected by running VariableRecoveryFast (maybe VariableRecovery later as well) on a function, and then solved using this analysis.

User may specify ground truth, which will override all types at certain program points during constraint solving.

Parameters

constraints –
ground_truth – A set of SimType-style solutions for some or all type variables. They will be respected during type solving.
var_mapping –
prioritize_char_array_over_struct –
must_struct –

update_variable_types(func_addr: Union[int, str], var_to_typevar)

pp_constraints() → None: Pretty-print constraints between variables using the variable mapping.

pp_solution() → None: Pretty-print solutions using the variable mapping.

All type constants used in type inference. They can be mapped, translated, or rewritten to C-style types.

class angr.analyses.typehoon.typeconsts.TypeConstant

Bases: object

SIZE = None

pp_str(mapping) → str

property size: int

class angr.analyses.typehoon.typeconsts.TopType: Bases: angr.analyses.typehoon.typeconsts.TypeConstant

class angr.analyses.typehoon.typeconsts.BottomType: Bases: angr.analyses.typehoon.typeconsts.TypeConstant

class angr.analyses.typehoon.typeconsts.Int: Bases: angr.analyses.typehoon.typeconsts.TypeConstant

class angr.analyses.typehoon.typeconsts.Char

Bases: angr.analyses.typehoon.typeconsts.Int

SIZE = 1

class angr.analyses.typehoon.typeconsts.Int1

Bases: angr.analyses.typehoon.typeconsts.Int

SIZE = 1

class angr.analyses.typehoon.typeconsts.Int8

Bases: angr.analyses.typehoon.typeconsts.Int

SIZE = 1

class angr.analyses.typehoon.typeconsts.Int16

Bases: angr.analyses.typehoon.typeconsts.Int

SIZE = 2

class angr.analyses.typehoon.typeconsts.Int32

Bases: angr.analyses.typehoon.typeconsts.Int

SIZE = 4

class angr.analyses.typehoon.typeconsts.Int64

Bases: angr.analyses.typehoon.typeconsts.Int

SIZE = 8

class angr.analyses.typehoon.typeconsts.Int128

Bases: angr.analyses.typehoon.typeconsts.Int

SIZE = 16

class angr.analyses.typehoon.typeconsts.Pointer(basetype)

Bases: angr.analyses.typehoon.typeconsts.TypeConstant

new(basetype)

class angr.analyses.typehoon.typeconsts.Pointer32(basetype)

Bases: angr.analyses.typehoon.typeconsts.Pointer, angr.analyses.typehoon.typeconsts.Int32

32-bit pointers.

class angr.analyses.typehoon.typeconsts.Pointer64(basetype)

Bases: angr.analyses.typehoon.typeconsts.Pointer, angr.analyses.typehoon.typeconsts.Int64

64-bit pointers.

class angr.analyses.typehoon.typeconsts.Array(element, count=None): Bases: angr.analyses.typehoon.typeconsts.TypeConstant

class angr.analyses.typehoon.typeconsts.Struct(fields=None): Bases: angr.analyses.typehoon.typeconsts.TypeConstant

class angr.analyses.typehoon.typeconsts.TypeVariableReference(typevar): Bases: angr.analyses.typehoon.typeconsts.TypeConstant

angr.analyses.typehoon.typeconsts.int_type(bits)

class angr.analyses.identifier.identify.FuncInfo: Bases: object

class angr.analyses.identifier.identify.Identifier(cfg=None, require_predecessors=True, only_find=None)

Bases: angr.analyses.analysis.Analysis

run(only_find=None)

can_call_same_name(addr, name)

get_func_info(func)

static constrain_all_zero(before_state, state, regs)

identify_func(function)

check_tests(cfg_func, match_func)

map_callsites()

do_trace(addr_trace, reverse_accesses, func_info)

get_call_args(func, callsite)

static get_reg_name(arch, reg_offset)

Parameters

arch – the architecture
reg_offset – Tries to find the name of a register given the offset in the registers.

Returns

The register name

find_stack_vars_x86(func)

static make_initial_state(project, stack_length)

Returns: an initial state with a symbolic stack and good options for rop

static make_symbolic_state(project, reg_list, stack_length=80): converts an input state into a state with symbolic registers :return: the symbolic state

class angr.analyses.loopfinder.Loop(entry, entry_edges, break_edges, continue_edges, body_nodes, graph, subloops): Bases: object

class angr.analyses.loopfinder.LoopFinder(functions=None, normalize=True)

Bases: angr.analyses.analysis.Analysis

Extracts all the loops from all the functions in a binary.

class angr.analyses.loop_analysis.VariableTypes

Bases: object

Iterator = 'Iterator'

HasNext = 'HasNext'

Next = 'Next'

class angr.analyses.loop_analysis.AnnotatedVariable(variable, type_)

Bases: object

variable

type

class angr.analyses.loop_analysis.Condition(op, val0, val1)

Bases: object

Equal = '=='

NotEqual = '!='

classmethod from_opstr(opstr)

class angr.analyses.loop_analysis.SootBlockProcessor(state, block, loop, defuse)

Bases: object

process()

class angr.analyses.loop_analysis.LoopAnalysisState(block)

Bases: object

copy()

merge(state)

add_loop_exit_stmt(stmt_idx, condition=None)

class angr.analyses.loop_analysis.LoopAnalysis(loop, defuse)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis, angr.analyses.analysis.Analysis

Analyze a loop and recover important information about the loop (e.g., invariants, induction variables) in a static manner.

exception angr.analyses.veritesting.VeritestingError: Bases: Exception

class angr.analyses.veritesting.CallTracingFilter(project, depth, blacklist=None)

Bases: object

Filter to apply during CFG creation on a given state and jumpkind to determine if it should be skipped at a certain depth

whitelist = {<class 'angr.procedures.libc.strcmp.strcmp'>, <class 'angr.procedures.libc.fgetc.fgetc'>, <class 'angr.procedures.posix.read.read'>, <class 'angr.procedures.glibc.__ctype_b_loc.__ctype_b_loc'>, <class 'angr.procedures.libc.atoi.atoi'>, <class 'angr.procedures.libc.strlen.strlen'>, <class 'angr.procedures.cgc.receive.receive'>, <class 'angr.procedures.cgc.transmit.transmit'>}

cfg_cache = {}

filter(call_target_state, jumpkind)

The call will be skipped if it returns True.

Parameters

call_target_state – The new state of the call target.
jumpkind – The Jumpkind of this call.

Returns

True if we want to skip this call, False otherwise.

class angr.analyses.veritesting.Veritesting(input_state, boundaries=None, loop_unrolling_limit=10, enable_function_inlining=False, terminator=None, deviation_filter=None)

Bases: angr.analyses.analysis.Analysis

An exploration technique made for condensing chunks of code to single (nested) if-then-else constraints via CFG accurate to conduct Static Symbolic Execution SSE (conversion to single constraint)

SSE stands for Static Symbolic Execution, and we also implemented an extended version of Veritesting (Avgerinos, Thanassis, et al, ICSE 2014).

Parameters

input_state – The initial state to begin the execution with.
boundaries – Addresses where execution should stop.
loop_unrolling_limit – The maximum times that Veritesting should unroll a loop for.
enable_function_inlining – Whether we should enable function inlining and syscall inlining.
terminator – A callback function that takes a state as parameter. Veritesting will terminate if this function returns True.
deviation_filter – A callback function that takes a state as parameter. Veritesting will put the state into “deviated” stash if this function returns True.

cfg_cache = {}

all_stashes = ('successful', 'errored', 'deadended', 'deviated', 'unconstrained')

is_not_in_cfg(s)

Returns if s.addr is not a proper node in our CFG.

Parameters: s (SimState) – The SimState instance to test.
Returns bool: False if our CFG contains p.addr, True otherwise.

is_overbound(state)

Filter out all states that run out of boundaries or loop too many times.

param SimState state: SimState instance to check returns bool: True if outside of mem/loop_ctr boundary

class angr.analyses.vfg.VFGJob(*args, **kwargs)

Bases: angr.analyses.cfg.cfg_job_base.CFGJobBase

A job descriptor that contains local variables used during VFG analysis.

property block_id

callstack_repr(kb=None)

class angr.analyses.vfg.PendingJob(block_id, state, call_stack, src_block_id, src_stmt_idx, src_ins_addr)

Bases: object

block_id

state

call_stack

src_block_id

src_stmt_idx

src_ins_addr

class angr.analyses.vfg.AnalysisTask

Bases: object

An analysis task describes a task that should be done before popping this task out of the task stack and discard it.

property done

class angr.analyses.vfg.FunctionAnalysis(function_address, return_address)

Bases: angr.analyses.vfg.AnalysisTask

Analyze a function, generate fix-point states from all endpoints of that function, and then merge them to one state.

property done

class angr.analyses.vfg.CallAnalysis(address, return_address, function_analysis_tasks=None, mergeable_plugins=None)

Bases: angr.analyses.vfg.AnalysisTask

Analyze a call by analyze all functions this call might be calling, collect all final states generated by analyzing those functions, and merge them into one state.

property done

register_function_analysis(task)

add_final_job(job)

merge_jobs()

class angr.analyses.vfg.VFGNode(addr, key, state=None)

Bases: object

A descriptor of nodes in a Value-Flow Graph

Constructor.

Parameters

addr (int) –
key (BlockID) –
state (SimState) –

append_state(s, is_widened_state=False): Appended a new state to this VFGNode. :param s: The new state to append :param is_widened_state: Whether it is a widened state or not.

class angr.analyses.vfg.VFG(cfg=None, context_sensitivity_level=2, start=None, function_start=None, interfunction_level=0, initial_state=None, avoid_runs=None, remove_options=None, timeout=None, max_iterations_before_widening=8, max_iterations=40, widening_interval=3, final_state_callback=None, status_callback=None, record_function_final_states=False)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis, angr.analyses.analysis.Analysis

This class represents a control-flow graph with static analysis result.

Perform abstract interpretation analysis starting from the given function address. The output is an invariant at the beginning (or the end) of each basic block.

Steps:

Generate a CFG first if CFG is not provided.

Identify all merge points (denote the set of merge points as Pw) in the CFG.

Cut those loop back edges (can be derived from Pw) so that we gain an acyclic CFG.

Identify all variables that are 1) from memory loading 2) from initial values, or 3) phi functions. Denote
the set of those variables as S_{var}.

Start real AI analysis and try to compute a fix point of each merge point. Perform widening/narrowing only on
variables in S_{var}.

Parameters

cfg – The control-flow graph to base this analysis on. If none is provided, we will construct a CFGEmulated.
context_sensitivity_level – The level of context-sensitivity of this VFG. It ranges from 0 to infinity. Default 2.
function_start – The address of the function to analyze.
interfunction_level – The level of interfunction-ness to be
initial_state – A state to use as the initial one
avoid_runs – A list of runs to avoid
remove_options – State options to remove from the initial state. It only works when initial_state is None
timeout (int) –

property function_initial_states

property function_final_states

get_any_node(addr): Get any VFG node corresponding to the basic block at @addr. Note that depending on the context sensitivity level, there might be multiple nodes corresponding to different contexts. This function will return the first one it encounters, which might not be what you want.

get_all_nodes(addr) → Generator[angr.analyses.vfg.VFGNode, None, None]

irsb_from_node(node)

copy()

class angr.analyses.vsa_ddg.DefUseChain(def_loc, use_loc, variable)

Bases: object

Stand for a def-use chain. it is generated by the DDG itself.

Constructor.

Parameters

def_loc –
use_loc –
variable –

Returns

class angr.analyses.vsa_ddg.VSA_DDG(vfg=None, start_addr=None, interfunction_level=0, context_sensitivity_level=2, keep_data=False)

Bases: angr.analyses.analysis.Analysis

A Data dependency graph based on VSA states. That means we don’t (and shouldn’t) expect any symbolic expressions.

Constructor.

Parameters

vfg – An already constructed VFG. If not specified, a new VFG will be created with other specified parameters. vfg and start_addr cannot both be unspecified.
start_addr – The address where to start the analysis (typically, a function’s entry point).
interfunction_level – See VFG analysis.
context_sensitivity_level – See VFG analysis.
keep_data – Whether we keep set of addresses as edges in the graph, or just the cardinality of the sets, which can be used as a “weight”.

get_predecessors(code_location)

Returns all predecessors of code_location.

Parameters: code_location – A CodeLocation instance.
Returns: A list of all predecessors.

get_all_nodes(simrun_addr, stmt_idx): Get all DDG nodes matching the given basic block address and statement index.

class angr.analyses.vtable.Vtable(vaddr, size, func_addrs=None)

Bases: object

This contains the addr, size and function addresses of a Vtable

class angr.analyses.vtable.VtableFinder

Bases: angr.analyses.analysis.Analysis

This analysis locates Vtables in a binary based on heuristics taken from - “Reconstruction of Class Hierarchies for Decompilation of C++ Programs”

is_cross_referenced(addr)

is_function(addr)

analyze()

create_extract_vtable(start_addr, sec_size)

class angr.analyses.find_objects_static.PossibleObject(size, addr, class_name=None)

Bases: object

This holds the address and class name of possible class instances. The address that it holds in mapped outside the binary so it is only valid in this analysis. TO DO: map the address to its uses in the registers/memory locations in the instructions

class angr.analyses.find_objects_static.NewFunctionHandler(max_addr=None, new_func_addr=None, project=None)

Bases: angr.analyses.reaching_definitions.function_handler.FunctionHandler

This handles calls to the function new(), by recording the size parameter passed to it and also assigns a new

address outside the mapped binary to the newly created space(possible object).

It also tracks if the function called right after new() is passed the same ‘this’ pointer and is a constructor, if so we mark it as an instance of the class the constructor belongs to.(only for non stripped binaries)

hook(analysis)

handle_local_function(state, function_address, call_stack, maximum_local_call_depth, visited_blocks, dep_graph, src_ins_addr=None, codeloc=None)

class angr.analyses.find_objects_static.StaticObjectFinder

Bases: angr.analyses.analysis.Analysis

This analysis tries to find objects on the heap based on calls to new(), and subsequent calls to constructors with: the ‘this’ pointer

class angr.analyses.class_identifier.ClassIdentifier

Bases: angr.analyses.analysis.Analysis

This is a class identifier for non stripped or partially stripped binaries, it identifies classes based on the demangled function names, and also assigns functions to their respective classes based on their names. It also uses the results from the VtableFinder analysis to assign the corresponding vtable to the classes.

self.classes contains a mapping between class names and SimCppClass objects

e.g. A::tool() and A::qux() belong to the class A

class angr.analyses.disassembly.DisassemblyPiece

Bases: object

addr = None

ident = nan

render(formatting=None)

getpiece(formatting, column)

width(formatting)

height(formatting)

static color(string, coloring, formatting)

highlight(string, formatting=None)

class angr.analyses.disassembly.FunctionStart(func)

Bases: angr.analyses.disassembly.DisassemblyPiece

Constructor.

Parameters: func (angr.knowledge.Function) – The function instance.

height(formatting)

class angr.analyses.disassembly.Label(addr, name): Bases: angr.analyses.disassembly.DisassemblyPiece

class angr.analyses.disassembly.IROp(addr: int, seq: int, obj: <module 'pyvex.stmt' from '/home/vsts/work/1/s/angr_venv/lib/python3.6/site-packages/pyvex/stmt.py'>, irsb: pyvex.block.IRSB)

Bases: angr.analyses.disassembly.DisassemblyPiece

addr: int

seq: int

obj: <module 'pyvex.stmt' from '/home/vsts/work/1/s/angr_venv/lib/python3.6/site-packages/pyvex/stmt.py'>

irsb: pyvex.block.IRSB

class angr.analyses.disassembly.BlockStart(block, parentfunc, project): Bases: angr.analyses.disassembly.DisassemblyPiece

class angr.analyses.disassembly.Hook(block): Bases: angr.analyses.disassembly.DisassemblyPiece

class angr.analyses.disassembly.Instruction(insn, parentblock, project=None)

Bases: angr.analyses.disassembly.DisassemblyPiece

property mnemonic

reload_format()

disect_instruction()

static split_op_string(insn_str)

class angr.analyses.disassembly.SootExpression(expr): Bases: angr.analyses.disassembly.DisassemblyPiece

class angr.analyses.disassembly.SootExpressionTarget(target_stmt_idx): Bases: angr.analyses.disassembly.SootExpression

class angr.analyses.disassembly.SootExpressionStaticFieldRef(field): Bases: angr.analyses.disassembly.SootExpression

class angr.analyses.disassembly.SootExpressionInvoke(invoke_type, expr)

Bases: angr.analyses.disassembly.SootExpression

Virtual = 'virtual'

Static = 'static'

Special = 'special'

class angr.analyses.disassembly.SootStatement(block_addr, raw_stmt)

Bases: angr.analyses.disassembly.DisassemblyPiece

property stmt_idx

class angr.analyses.disassembly.Opcode(parentinsn): Bases: angr.analyses.disassembly.DisassemblyPiece

class angr.analyses.disassembly.Operand(op_num, children, parentinsn)

Bases: angr.analyses.disassembly.DisassemblyPiece

property cs_operand

static build(operand_type, op_num, children, parentinsn)

class angr.analyses.disassembly.ConstantOperand(op_num, children, parentinsn): Bases: angr.analyses.disassembly.Operand

class angr.analyses.disassembly.RegisterOperand(op_num, children, parentinsn)

Bases: angr.analyses.disassembly.Operand

property register

class angr.analyses.disassembly.MemoryOperand(op_num, children, parentinsn): Bases: angr.analyses.disassembly.Operand

class angr.analyses.disassembly.OperandPiece

Bases: angr.analyses.disassembly.DisassemblyPiece

addr = None

parentop = None

ident = None

class angr.analyses.disassembly.Register(reg, prefix): Bases: angr.analyses.disassembly.OperandPiece

class angr.analyses.disassembly.Value(val, render_with_sign)

Bases: angr.analyses.disassembly.OperandPiece

property project

class angr.analyses.disassembly.Comment(addr, text)

Bases: angr.analyses.disassembly.DisassemblyPiece

height(formatting)

class angr.analyses.disassembly.FuncComment(func): Bases: angr.analyses.disassembly.DisassemblyPiece

class angr.analyses.disassembly.Disassembly(function: Optional[angr.knowledge_plugins.functions.function.Function] = None, ranges: Optional[Sequence[Tuple[int, int]]] = None, include_ir: bool = False)

Bases: angr.analyses.analysis.Analysis

Produce formatted machine code disassembly.

func_lookup(block)

parse_block(block: angr.codenode.BlockNode) → None: Parse instructions for a given block node

render(formatting=None, show_edges: bool = True, show_addresses: bool = True, show_bytes: bool = False, ascii_only: Optional[bool] = None) → str

Render the disassembly to a string, with optional edges and addresses.

Color will be added by default, if enabled. To disable color pass an empty formatting dict.

angr.analyses.disassembly_utils.decode_instruction(arch, instr)

exception angr.analyses.reassembler.BinaryError: Bases: Exception

exception angr.analyses.reassembler.InstructionError: Bases: angr.analyses.reassembler.BinaryError

exception angr.analyses.reassembler.ReassemblerFailureNotice: Bases: angr.analyses.reassembler.BinaryError

angr.analyses.reassembler.string_escape(s)

angr.analyses.reassembler.fill_reg_map()

angr.analyses.reassembler.split_operands(s)

angr.analyses.reassembler.is_hex(s)

class angr.analyses.reassembler.Label(binary, name, original_addr=None)

Bases: object

g_label_ctr = count(0)

property operand_str

property offset

static new_label(binary, name=None, function_name=None, original_addr=None, data_label=False)

class angr.analyses.reassembler.DataLabel(binary, original_addr, name=None)

Bases: angr.analyses.reassembler.Label

property operand_str

class angr.analyses.reassembler.FunctionLabel(binary, function_name, original_addr, plt=False)

Bases: angr.analyses.reassembler.Label

property function_name

property operand_str

class angr.analyses.reassembler.ObjectLabel(binary, symbol_name, original_addr, plt=False)

Bases: angr.analyses.reassembler.Label

property symbol_name

property operand_str

class angr.analyses.reassembler.NotypeLabel(binary, symbol_name, original_addr, plt=False)

Bases: angr.analyses.reassembler.Label

property symbol_name

property operand_str

class angr.analyses.reassembler.SymbolManager(binary, cfg)

Bases: object

SymbolManager manages all symbols in the binary.

Constructor.

Parameters

binary (Reassembler) – The Binary analysis instance.
cfg (angr.analyses.CFG) – The CFG analysis instance.

Returns

None

get_unique_symbol_name(symbol_name)

new_label(addr, name=None, is_function=None, force=False)

label_got(addr, label)

Mark a certain label as assigned (to an instruction or a block of data).

Parameters

addr (int) – The address of the label.
label (angr.analyses.reassembler.Label) – The label that is just assigned.

Returns

None

class angr.analyses.reassembler.Operand(binary, insn_addr, insn_size, capstone_operand, operand_str, mnemonic, operand_offset, syntax=None)

Bases: object

Constructor.

Parameters

binary (Reassembler) – The Binary analysis.
insn_addr (int) – Address of the instruction.
capstone_operand –
operand_str (str) – the string representation of this operand
mnemonic (str) – Mnemonic of the instruction that this operand belongs to.
operand_offset (int) – offset of the operand into the instruction.
syntax (str) – Provide a way to override the default syntax coming from binary.

Returns

None

assembly()

property is_immediate

property symbolized

class angr.analyses.reassembler.Instruction(binary, addr, size, insn_bytes, capstone_instr)

Bases: object

High-level representation of an instruction in the binary

Parameters

binary (Reassembler) – The Binary analysis
addr (int) – Address of the instruction
size (int) – Size of the instruction
insn_bytes (str) – Instruction bytes
capstone_instr – Capstone Instr object.

Returns

None

assign_labels()

dbg_comments()

assembly(comments=False, symbolized=True)

Returns

class angr.analyses.reassembler.BasicBlock(binary, addr, size, x86_getpc_retsite: bool = False)

Bases: object

BasicBlock represents a basic block in the binary.

Constructor.

Parameters

binary (Reassembler) – The Binary analysis.
addr (int) – Address of the block
size (int) – Size of the block

Returns

None

assign_labels()

assembly(comments=False, symbolized=True)

instruction_addresses()

class angr.analyses.reassembler.Procedure(binary, function=None, addr=None, size=None, name=None, section='.text', asm_code=None)

Bases: object

Procedure in the binary.

Constructor.

Parameters

binary (Reassembler) – The Binary analysis.
function (angr.knowledge.Function) – The function it represents
addr (int) – Address of the function. Not required if function is provided.
size (int) – Size of the function. Not required if function is provided.
section (str) – Which section this function comes from.

Returns

None

property name: Get function name from the labels of the very first block. :return: Function name if there is any, None otherwise :rtype: string

property is_plt: If this function is a PLT entry or not. :return: True if this function is a PLT entry, False otherwise :rtype: bool

assign_labels()

assembly(comments=False, symbolized=True)

Get the assembly manifest of the procedure.

Parameters

comments –
symbolized –

Returns

A list of tuples (address, basic block assembly), ordered by basic block addresses

Return type

list

instruction_addresses()

Get all instruction addresses in the binary.

Returns: A list of sorted instruction addresses.
Return type: list

class angr.analyses.reassembler.ProcedureChunk(project, addr, size)

Bases: angr.analyses.reassembler.Procedure

Procedure chunk.

Constructor.

Parameters

project –
addr –
size –

Returns

class angr.analyses.reassembler.Data(binary, memory_data=None, section=None, section_name=None, name=None, size=None, sort=None, addr=None, initial_content=None)

Bases: object

property content

shrink(new_size)

Reduce the size of this block

Parameters: new_size (int) – The new size
Returns: None

desymbolize()

We believe this was a pointer and symbolized it before. Now we want to desymbolize it.

The following actions are performed: - Reload content from memory - Mark the sort as ‘unknown’

Returns: None

assign_labels()

assembly(comments=False, symbolized=True)

class angr.analyses.reassembler.Relocation(addr, ref_addr, sort): Bases: object

class angr.analyses.reassembler.Reassembler(syntax='intel', remove_cgc_attachments=True, log_relocations=True)

Bases: angr.analyses.analysis.Analysis

High-level representation of a binary with a linear representation of all instructions and data regions. After calling “symbolize”, it essentially acts as a binary reassembler.

Tested on CGC, x86 and x86-64 binaries.

Discliamer: The reassembler is an empirical solution. Don’t be surprised if it does not work on some binaries.

property instructions

Get a list of all instructions in the binary

Returns: A list of (address, instruction)
Return type: tuple

property relocations

property inserted_asm_before_label

property inserted_asm_after_label

property main_executable_regions: return:

property main_nonexecutable_regions: return:

section_alignment(section_name)

Get the alignment for the specific section. If the section is not found, 16 is used as default.

Parameters: section_name (str) – The section.
Returns: The alignment in bytes.
Return type: int

main_executable_regions_contain(addr)

Parameters: addr –
Returns

main_executable_region_limbos_contain(addr)

Sometimes there exists a pointer that points to a few bytes before the beginning of a section, or a few bytes after the beginning of the section. We take care of that here.

Parameters: addr (int) – The address to check.
Returns: A 2-tuple of (bool, the closest base address)
Return type: tuple

main_nonexecutable_regions_contain(addr)

Parameters: addr (int) – The address to check.
Returns: True if the address is inside a non-executable region, False otherwise.
Return type: bool

main_nonexecutable_region_limbos_contain(addr, tolerance_before=64, tolerance_after=64)

Sometimes there exists a pointer that points to a few bytes before the beginning of a section, or a few bytes after the beginning of the section. We take care of that here.

Parameters: addr (int) – The address to check.
Returns: A 2-tuple of (bool, the closest base address)
Return type: tuple

register_instruction_reference(insn_addr, ref_addr, sort, operand_offset)

register_data_reference(data_addr, ref_addr)

add_label(name, addr)

Add a new label to the symbol manager.

Parameters

name (str) – Name of the label.
addr (int) – Address of the label.

Returns

None

insert_asm(addr, asm_code, before_label=False)

Insert some assembly code at the specific address. There must be an instruction starting at that address.

Parameters

addr (int) – Address of insertion
asm_code (str) – The assembly code to insert

Returns

None

append_procedure(name, asm_code)

Add a new procedure with specific name and assembly code.

Parameters

name (str) – The name of the new procedure.
asm_code (str) – The assembly code of the procedure

Returns

None

append_data(name, initial_content, size, readonly=False, sort='unknown')

Append a new data entry into the binary with specific name, content, and size.

Parameters

name (str) – Name of the data entry. Will be used as the label.
initial_content (bytes) – The initial content of the data entry.
size (int) – Size of the data entry.
readonly (bool) – If the data entry belongs to the readonly region.
sort (str) – Type of the data.

Returns

None

remove_instruction(ins_addr)

Parameters: ins_addr –
Returns

randomize_procedures()

Returns

symbolize()

assembly(comments=False, symbolized=True)

remove_cgc_attachments()

Remove CGC attachments.

Returns: True if CGC attachments are found and removed, False otherwise
Return type: bool

remove_unnecessary_stuff()

Remove unnecessary functions and data

Returns: None

remove_unnecessary_stuff_glibc()

fast_memory_load(addr, size, data_type, endness='Iend_LE')

Load memory bytes from loader’s memory backend.

Parameters

addr (int) – The address to begin memory loading.
size (int) – Size in bytes.
data_type – Type of the data.
endness (str) – Endianness of this memory load.

Returns

Data read out of the memory.

Return type

int or bytes or str or None

class angr.analyses.congruency_check.CongruencyCheck(throw=False)

Bases: angr.analyses.analysis.Analysis

This is an analysis to ensure that angr executes things identically with different execution backends (i.e., unicorn vs vex).

Initializes a CongruencyCheck analysis.

Parameters: throw – whether to raise an exception if an incongruency is found.

set_state_options(left_add_options=None, left_remove_options=None, right_add_options=None, right_remove_options=None): Checks that the specified state options result in the same states over the next depth states.

set_states(left_state, right_state): Checks that the specified paths stay the same over the next depth states.

set_simgr(simgr)

run(depth=None)

Checks that the paths in the specified path group stay the same over the next depth bytes.

The path group should have a “left” and a “right” stash, each with a single path.

compare_path_group(pg)

compare_states(sl, sr): Compares two states for similarity.

compare_paths(pl, pr)

class angr.analyses.static_hooker.StaticHooker(library, binary=None)

Bases: angr.analyses.analysis.Analysis

This analysis works on statically linked binaries - it finds the library functions statically linked into the binary and hooks them with the appropriate simprocedures.

Right now it only works on unstripped binaries, but hey! There’s room to grow!

class angr.analyses.binary_optimizer.ConstantPropagation(constant, constant_assignment_loc, constant_consuming_loc): Bases: object

class angr.analyses.binary_optimizer.RedundantStackVariable(argument, stack_variable, stack_variable_consuming_locs): Bases: object

class angr.analyses.binary_optimizer.RegisterReallocation(stack_variable, register_variable, stack_variable_sources, stack_variable_consumers, prologue_addr, prologue_size, epilogue_addr, epilogue_size)

Bases: object

Constructor.

Parameters

stack_variable (SimStackVariable) –
register_variable (SimRegisterVariable) –
stack_variable_sources (list) –
stack_variable_consumers (list) –
prologue_addr (int) –
prologue_size (int) –
epilogue_addr (int) –
epilogue_size (int) –

class angr.analyses.binary_optimizer.DeadAssignment(pv)

Bases: object

Constructor.

Parameters: pv (angr.analyses.ddg.ProgramVariable) – The assignment to remove.

class angr.analyses.binary_optimizer.BinaryOptimizer(cfg, techniques)

Bases: angr.analyses.analysis.Analysis

This is a collection of binary optimization techniques we used in Mechanical Phish during the finals of Cyber Grand Challange. It focuses on dealing with some serious speed-impacting code constructs, and sort of worked on some CGC binaries compiled with O0. Use this analysis as a reference of how to use data dependency graph and such.

There is no guarantee that BinaryOptimizer will ever work on non-CGC binaries. Feel free to give us PR or MR, but please do not ask for support of non-CGC binaries.

BLOCKS_THRESHOLD = 500

optimize()

class angr.analyses.callee_cleanup_finder.CalleeCleanupFinder(starts=None, hook_all=False)

Bases: angr.analyses.analysis.Analysis

analyze(addr)

class angr.analyses.dominance_frontier.DominanceFrontier(func, exception_edges=False)

Bases: angr.analyses.analysis.Analysis

Computes the dominance frontier of all nodes in a function graph, and provides an easy-to-use interface for querying the frontier information.

class angr.analyses.init_finder.SimEngineInitFinderVEX(project, replacements, overlay, pointers_only=False)

Bases: angr.engines.light.engine.SimEngineLightVEXMixin, angr.engines.light.engine.SimEngineLight

static is_concrete(expr) → bool

class angr.analyses.init_finder.InitializationFinder(func=None, func_graph=None, block=None, max_iterations=1, replacements=None, overlay=None, pointers_only=False)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis, angr.analyses.analysis.Analysis

Finds possible initializations for global data sections and generate an overlay to be used in other analyses later on.

class angr.analyses.xrefs.SimEngineXRefsVEX(xref_manager, project=None, replacements=None)

Bases: angr.engines.light.engine.SimEngineLightVEXMixin, angr.engines.light.engine.SimEngineLight

add_xref(xref_type, from_loc, to_loc)

static extract_value_if_concrete(expr) → Optional[int]

Extract the concrete value from expr if it is a concrete claripy AST.

Parameters: expr – A claripy AST.
Returns: A concrete value or None if nothing concrete can be extracted.

class angr.analyses.xrefs.XRefsAnalysis(func=None, func_graph=None, block=None, max_iterations=1, replacements=None)

Bases: angr.analyses.forward_analysis.forward_analysis.ForwardAnalysis, angr.analyses.analysis.Analysis

XRefsAnalysis recovers in-depth x-refs (cross-references) in disassembly code.

Here is an example:

.text:
000023C8                 LDR     R2, =time_now
000023CA                 LDR     R3, [R2]
000023CC                 ADDS    R3, #1
000023CE                 STR     R3, [R2]
000023D0                 BX      LR

.bss:
1FFF36F4 time_now        % 4

You will have the following x-refs for time_now:

23c8 - offset
23ca - read access
23ce - write access

class angr.analyses.proximity_graph.ProxiNodeTypes

Bases: object

Node Type Enums

Empty = 0

String = 1

Function = 2

FunctionCall = 3

Integer = 4

Unknown = 5

Variable = 6

class angr.analyses.proximity_graph.BaseProxiNode(type_: int, ref_at: Optional[Set[int]] = None)

Bases: object

Base class for all nodes in a proximity graph.

class angr.analyses.proximity_graph.FunctionProxiNode(func, ref_at: Optional[Set[int]] = None)

Bases: angr.analyses.proximity_graph.BaseProxiNode

Proximity node showing current and expanded function calls in graph.

class angr.analyses.proximity_graph.VariableProxiNode(addr, name, ref_at: Optional[Set[int]] = None)

Bases: angr.analyses.proximity_graph.BaseProxiNode

Variable arg node

class angr.analyses.proximity_graph.StringProxiNode(addr, content, ref_at: Optional[Set[int]] = None)

Bases: angr.analyses.proximity_graph.BaseProxiNode

String arg node

class angr.analyses.proximity_graph.CallProxiNode(callee, ref_at: Optional[Set[int]] = None, args: Optional[Tuple[angr.analyses.proximity_graph.BaseProxiNode]] = None)

Bases: angr.analyses.proximity_graph.BaseProxiNode

Call node

class angr.analyses.proximity_graph.IntegerProxiNode(value: int, ref_at: Optional[Set[int]] = None)

Bases: angr.analyses.proximity_graph.BaseProxiNode

Int arg node

class angr.analyses.proximity_graph.UnknownProxiNode(dummy_value: str)

Bases: angr.analyses.proximity_graph.BaseProxiNode

Unknown arg node

class angr.analyses.proximity_graph.ProximityGraphAnalysis(func: Function, cfg_model: CFGModel, xrefs: XRefManager, decompilation: Optional[Decompiler] = None, expand_funcs: Optional[Set[int]] = None)

Bases: angr.analyses.analysis.Analysis

Generate a proximity graph.

Defines analysis that will generate a dynamic data-dependency graph

class angr.analyses.data_dep.data_dependency_analysis.NodalAnnotation(node: BaseDepNode)

Bases: claripy.annotation.Annotation

Allows a node to be stored as an annotation to a BV in a DefaultMemory instance

property relocatable: bool: Can not be relocated in a simplification

property eliminatable: Can not be eliminated in a simplification

class angr.analyses.data_dep.data_dependency_analysis.DataDependencyGraphAnalysis(end_state: SimState, start_from: Optional[int] = None, end_at: Optional[int] = None, block_addrs: Optional[List[int]] = None)

Bases: angr.analyses.analysis.Analysis

This is a DYNAMIC data dependency graph that utilizes a given SimState to produce a DDG graph that is accurate to the path the program took during execution.

This analysis utilizes the SimActionData objects present in the provided SimState’s action history to generate the dependency graph.

Parameters

end_state – Simulation state used to extract all SimActionData
start_from – An address or None, Specifies where to start generation of DDG
end_at – An address or None, Specifies where to end generation of DDG
block_addrs (iterable or None) – List of block addresses that the DDG analysis should be run on

property graph: Optional[networkx.classes.digraph.DiGraph]

property simplified_graph: Optional[networkx.classes.digraph.DiGraph]

property sub_graph: Optional[networkx.classes.digraph.DiGraph]

get_data_dep(g_node: BaseDepNode, include_tmp_nodes: bool, backwards: bool) → Optional[networkx.classes.digraph.DiGraph]

class angr.analyses.data_dep.sim_act_location.SimActLocation(bbl_addr: int, ins_addr: int, stmt_idx: int)

Bases: object

Structure-like class used to bundle the instruction address and statement index of a given SimAction in order to uniquely identify a given SimAction

class angr.analyses.data_dep.sim_act_location.ParsedInstruction(ins_addr: int, min_stmt_idx: int, max_stmt_idx: int)

Bases: object

Used by parser to facilitate linking with recent ancestors in an efficient manner

class angr.analyses.data_dep.dep_nodes.DepNodeTypes

Bases: object

Enumeration of types of BaseDepNode supported by this analysis

Memory = 1

Register = 2

Tmp = 3

Constant = 4

class angr.analyses.data_dep.dep_nodes.BaseDepNode(type_: int, sim_act: SimActionData)

Bases: object

Base class for all nodes in a data-dependency graph

value_tuple() → Tuple[BV, int]

Returns: A tuple containing the node’s value as a BV and as an evaluated integer

property ast: BV

property type: int: Getter :return: An integer defined in DepNodeTypes, represents the subclass type of this DepNode.

class angr.analyses.data_dep.dep_nodes.ConstantDepNode(sim_act: SimActionData, value: int)

Bases: angr.analyses.data_dep.dep_nodes.BaseDepNode

Used to create a DepNode that will hold a constant, numeric value Uniquely identified by its value

class angr.analyses.data_dep.dep_nodes.MemDepNode(sim_act: SimActionData, addr: int)

Bases: angr.analyses.data_dep.dep_nodes.BaseDepNode

Used to represent SimActions of type MEM

property width: int

classmethod cast_to_mem(base_dep_node: angr.analyses.data_dep.dep_nodes.BaseDepNode): Casts a BaseDepNode into a MemDepNode

class angr.analyses.data_dep.dep_nodes.VarDepNode(type_: int, sim_act: SimActionData, reg: int, arch_name: str = '')

Bases: angr.analyses.data_dep.dep_nodes.BaseDepNode

Abstract class for representing SimActions of TYPE reg or tmp

property display_name: str

class angr.analyses.data_dep.dep_nodes.TmpDepNode(sim_act: SimActionData, reg: int, arch_name: str = '')

Bases: angr.analyses.data_dep.dep_nodes.VarDepNode

Used to represent SimActions of type TMP

class angr.analyses.data_dep.dep_nodes.RegDepNode(sim_act: SimActionData, reg: int, arch_name: str = '')

Bases: angr.analyses.data_dep.dep_nodes.VarDepNode

Base class for representing SimActions of TYPE reg

property reg_size: int

class angr.blade.Blade(graph, dst_run, dst_stmt_idx, direction='backward', project=None, cfg=None, ignore_sp=False, ignore_bp=False, ignored_regs=None, max_level=3, base_state=None, stop_at_calls=False, cross_insn_opt=False)

Bases: object

Blade is a light-weight program slicer that works with networkx DiGraph containing CFGNodes. It is meant to be used in angr for small or on-the-fly analyses.

Parameters

graph (networkx.DiGraph) – A graph representing the control flow graph. Note that it does not take angr.analyses.CFGEmulated or angr.analyses.CFGFast.
dst_run (int) – An address specifying the target SimRun.
dst_stmt_idx (int) – The target statement index. -1 means executing until the last statement.
direction (str) – ‘backward’ or ‘forward’ slicing. Forward slicing is not yet supported.
project (angr.Project) – The project instance.
cfg (angr.analyses.CFGBase) – the CFG instance. It will be made mandatory later.
ignore_sp (bool) – Whether the stack pointer should be ignored in dependency tracking. Any dependency from/to stack pointers will be ignored if this options is True.
ignore_bp (bool) – Whether the base pointer should be ignored or not.
max_level (int) – The maximum number of blocks that we trace back for.
stop_at_calls (int) – Limit slicing within a single function. Do not proceed when encounters a call edge.

Returns

None

property slice

dbg_repr(arch=None)

class angr.slicer.SimLightState(temps=None, regs=None, stack_offsets=None, options=None): Bases: object

class angr.slicer.SimSlicer(arch, statements, target_tmps=None, target_regs=None, target_stack_offsets=None, inslice_callback=None, inslice_callback_infodict=None)

Bases: object

A super lightweight intra-IRSB slicing class.

class angr.annocfg.AnnotatedCFG(project, cfg=None, detect_loops=False)

Bases: object

AnnotatedCFG is a control flow graph with statement whitelists and exit whitelists to describe a slice of the program.

Constructor.

Parameters

project – The angr Project instance
cfg – Control flow graph.
detect_loops –

from_digraph(digraph)

Initialize this AnnotatedCFG object with a networkx.DiGraph consisting of the following form of nodes:

Tuples like (block address, statement ID)

Those nodes are connected by edges indicating the execution flow.

Parameters: digraph (networkx.DiGraph) – A networkx.DiGraph object

get_addr(run)

add_block_to_whitelist(block)

add_statements_to_whitelist(block, stmt_ids)

add_exit_to_whitelist(run_from, run_to)

set_last_statement(block_addr, stmt_id)

add_loop(loop_tuple): A loop tuple contains a series of IRSB addresses that form a loop. Ideally it always starts with the first IRSB that we meet during the execution.

should_take_exit(addr_from, addr_to)

should_execute_statement(addr, stmt_id)

get_run(addr)

get_whitelisted_statements(addr)

Returns: True if all statements are whitelisted

get_last_statement_index(addr)

Get the statement index of the last statement to execute in the basic block specified by addr.

Parameters: addr (int) – Address of the basic block.
Returns: The statement index of the last statement to be executed in the block. Usually if the default exit is taken, it will be the last statement to execute. If the block is not in the slice or we should never take any exit going to this block, None is returned.
Return type: int or None

get_loops()

get_targets(source_addr)

dbg_repr()

dbg_print_irsb(irsb_addr, project=None): Pretty-print an IRSB with whitelist information

keep_path(path): Given a path, returns True if the path should be kept, False if it should be cut.

merge_points(path)

successor_func(path)

Callback routine that takes in a path, and returns all feasible successors to path group. This callback routine should be passed to the keyword argument “successor_func” of PathGroup.step().

Parameters: path – A Path instance.
Returns: A list of all feasible Path successors.

angr.codenode.repr_addr(addr)

class angr.codenode.CodeNode(addr, size, graph=None, thumb=False)

Bases: object

addr

size

thumb

successors()

predecessors()

is_hook = None

class angr.codenode.BlockNode(addr, size, bytestr=None, **kwargs)

Bases: angr.codenode.CodeNode

is_hook = False

bytestr

class angr.codenode.SootBlockNode(addr, size, stmts, **kwargs)

Bases: angr.codenode.BlockNode

stmts

class angr.codenode.HookNode(addr, size, sim_procedure, **kwargs)

Bases: angr.codenode.CodeNode

Parameters: sim_procedure (type) – the the sim_procedure class

is_hook = True

sim_procedure

class angr.codenode.SyscallNode(addr, size, sim_procedure, **kwargs)

Bases: angr.codenode.HookNode

Parameters: sim_procedure (type) – the the sim_procedure class

is_hook = False

sim_procedure

SimOS

Manage OS-level configuration.

angr.simos.register_simos(name, cls)

class angr.simos.simos.SimOS(project: angr.project.Project, name=None)

Bases: object

A class describing OS/arch-level configuration.

configure_project(): Configure the project to set up global settings (like SimProcedures).

state_blank(addr=None, initial_prefix=None, brk=None, stack_end=None, stack_size=8388608, stdin=None, thread_idx=None, permissions_backer=None, **kwargs)

Initialize a blank state.

All parameters are optional.

Parameters

addr – The execution start address.
initial_prefix –
stack_end – The end of the stack (i.e., the byte after the last valid stack address).
stack_size – The number of bytes to allocate for stack space
brk – The address of the process’ break.

Returns

The initialized SimState.

Any additional arguments will be passed to the SimState constructor

state_entry(**kwargs)

state_full_init(**kwargs)

state_call(addr, *args, **kwargs)

prepare_call_state(calling_state, initial_state=None, preserve_registers=(), preserve_memory=())

This function prepares a state that is executing a call instruction. If given an initial_state, it copies over all of the critical registers to it from the calling_state. Otherwise, it prepares the calling_state for action.

This is mostly used to create minimalistic for CFG generation. Some ABIs, such as MIPS PIE and x86 PIE, require certain information to be maintained in certain registers. For example, for PIE MIPS, this function transfer t9, gp, and ra to the new state.

prepare_function_symbol(symbol_name, basic_addr=None)

Prepare the address space with the data necessary to perform relocations pointing to the given symbol

Returns a 2-tuple. The first item is the address of the function code, the second is the address of the relocation target.

handle_exception(successors, engine, exception)

Perform exception handling. This method will be called when, during execution, a SimException is thrown. Currently, this can only indicate a segfault, but in the future it could indicate any unexpected exceptional behavior that can’t be handled by ordinary control flow.

The method may mutate the provided SimSuccessors object in any way it likes, or re-raise the exception.

Parameters

successors – The SimSuccessors object currently being executed on
engine – The engine that was processing this step
exc_type – The value of sys.exc_info()[0] from the error, the type of the exception that was raised
exc_value – The value of sys.exc_info()[1] from the error, the actual exception object
exc_traceback – The value of sys.exc_info()[2] from the error, the traceback from the exception

syscall(state, allow_unsupported=True)

syscall_abi(state) → str

syscall_cc(state) → Optional[angr.calling_conventions.SimCCSyscall]

is_syscall_addr(addr)

syscall_from_addr(addr, allow_unsupported=True)

syscall_from_number(number, allow_unsupported=True, abi=None)

setup_gdt(state, gdt)

Write the GlobalDescriptorTable object in the current state memory

Parameters

state – state in which to write the GDT
gdt – GlobalDescriptorTable object

Returns

generate_gdt(fs, gs, fs_size=4294967295, gs_size=4294967295)

Generate a GlobalDescriptorTable object and populate it using the value of the gs and fs register

Parameters

fs – value of the fs segment register
gs – value of the gs segment register
fs_size – size of the fs segment register
gs_size – size of the gs segment register

Returns

gdt a GlobalDescriptorTable object

class angr.simos.simos.GlobalDescriptorTable(addr, limit, table, gdt_sel, cs_sel, ds_sel, es_sel, ss_sel, fs_sel, gs_sel): Bases: object

class angr.simos.linux.SimLinux(project, **kwargs)

Bases: angr.simos.userland.SimUserland

OS-specific configuration for *nix-y OSes.

configure_project()

syscall_abi(state)

state_blank(fs=None, concrete_fs=False, chroot=None, cwd=None, pathsep=b'/', thread_idx=None, init_libc=False, **kwargs)

state_entry(args=None, env=None, argc=None, **kwargs)

set_entry_register_values(state)

state_full_init(**kwargs)

prepare_function_symbol(symbol_name, basic_addr=None)

Prepare the address space with the data necessary to perform relocations pointing to the given symbol.

Returns a 2-tuple. The first item is the address of the function code, the second is the address of the relocation target.

initialize_segment_register_x64(state, concrete_target)

Set the fs register in the angr to the value of the fs register in the concrete process

Parameters

state – state which will be modified
concrete_target – concrete target that will be used to read the fs register

Returns

None

initialize_gdt_x86(state, concrete_target)

Create a GDT in the state memory and populate the segment registers. Rehook the vsyscall address using the real value in the concrete process memory

Parameters

state – state which will be modified
concrete_target – concrete target that will be used to read the fs register

Returns

get_segment_register_name()

class angr.simos.cgc.SimCGC(project, **kwargs)

Bases: angr.simos.userland.SimUserland

Environment configuration for the CGC DECREE platform

state_blank(flag_page=None, allocate_stack_page_count=256, **kwargs)

Parameters

flag_page – Flag page content, either a string or a list of BV8s
allocate_stack_page_count – Number of pages to pre-allocate for stack

state_entry(add_options=None, **kwargs)

class angr.simos.userland.SimUserland(project, syscall_library=None, syscall_addr_alignment=4, **kwargs)

Bases: angr.simos.simos.SimOS

This is a base class for any SimOS that wants to support syscalls.

It uses the CLE kernel object to provide addresses for syscalls. Syscalls will be emulated as a jump to one of these addresses, where a SimProcedure from the syscall library provided at construction time will be executed.

configure_project(abi_list=None)

syscall_cc(state) → angr.calling_conventions.SimCCSyscall

syscall(state, allow_unsupported=True)

Given a state, return the procedure corresponding to the current syscall. This procedure will have .syscall_number, .display_name, and .addr set.

Parameters

state – The state to get the syscall number from
allow_unsupported – Whether to return a “dummy” sycall instead of raising an unsupported exception

syscall_abi(state): Optionally, override this function to determine which abi is being used for the state’s current syscall.

is_syscall_addr(addr): Return whether or not the given address corresponds to a syscall implementation.

syscall_from_addr(addr, allow_unsupported=True)

Get a syscall SimProcedure from an address.

Parameters

addr – The address to convert to a syscall SimProcedure
allow_unsupported – Whether to return a dummy procedure for an unsupported syscall instead of raising an exception.

Returns

The SimProcedure for the syscall, or None if the address is not a syscall address.

syscall_from_number(number, allow_unsupported=True, abi=None)

Get a syscall SimProcedure from its number.

Parameters

number – The syscall number
allow_unsupported – Whether to return a “stub” syscall for unsupported numbers instead of throwing an error
abi – The name of the abi to use. If None, will assume that the abis have disjoint numbering schemes and pick the right one.

Returns

The SimProcedure for the syscall

class angr.simos.windows.SecurityCookieInit(value)

Bases: enum.Enum

An enumeration.

NONE = 0

RANDOM = 1

STATIC = 2

SYMBOLIC = 3

class angr.simos.windows.SimWindows(project)

Bases: angr.simos.simos.SimOS

Environment for the Windows Win32 subsystem. Does not support syscalls currently.

configure_project()

state_entry(args=None, env=None, argc=None, **kwargs)

state_blank(thread_idx=None, **kwargs)

handle_exception(successors, engine, exception)

initialize_segment_register_x64(state, concrete_target)

Set the gs register in the angr to the value of the fs register in the concrete process

Parameters

state – state which will be modified
concrete_target – concrete target that will be used to read the fs register

Returns

None

initialize_gdt_x86(state, concrete_target)

Create a GDT in the state memory and populate the segment registers.

Parameters

state – state which will be modified
concrete_target – concrete target that will be used to read the fs register

Returns

the created GlobalDescriptorTable object

get_segment_register_name()

class angr.simos.javavm.SimJavaVM(*args, **kwargs)

Bases: angr.simos.simos.SimOS

state_blank(addr=None, **kwargs)

state_entry(args=None, **kwargs)

Create an entry state.

Parameters: args – List of SootArgument values (optional).

static generate_symbolic_cmd_line_arg(state, max_length=1000): Generates a new symbolic cmd line argument string. :return: The string reference.

state_call(addr, *args, **kwargs)

Create a native or a Java call state.

Parameters

addr – Soot or native addr of the invoke target.
args – List of SootArgument values.

static get_default_value_by_type(type_, state)

Java specify defaults values for primitive and reference types. This method returns the default value for a given type.

Parameters

type (str) – Name of type.
state (SimState) – Current SimState.

Returns

Default value for this type.

static cast_primitive(state, value, to_type)

Cast the value of primtive types.

Parameters

value – Bitvector storing the primitive value.
to_type – Name of the targeted type.

Returns

Resized value.

static init_static_field(state, field_class_name, field_name, field_type)

Initialize the static field with an allocated, but not initialized, object of the given type.

Parameters

state – State associated to the field.
field_class_name – Class containing the field.
field_name – Name of the field.
field_type – Type of the field and the new object.

static get_cmd_line_args(state)

get_addr_of_native_method(soot_method)

Get address of the implementation from a native declared Java function.

Parameters: soot_method – Method descriptor of a native declared function.
Returns: CLE address of the given method.

get_native_type(java_type)

Maps the Java type to a SimTypeReg representation of its native counterpart. This type can be used to indicate the (well-defined) size of native JNI types.

Returns: A SymTypeReg with the JNI size of the given type.

get_method_native_type(method)

property native_arch

Arch of the native simos.

Type: return

get_native_cc()

Returns: SimCC object for the native simos.

angr.simos.javavm.prepare_native_return_state(native_state)

Hook target for native function call returns.

Recovers and stores the return value from native memory and toggles the state, s.t. execution continues in the Soot engine.

Note: Redirection needed for pickling.

Function Signature Matching

class angr.flirt.FlirtSignature(arch: str, platform: str, sig_name: str, sig_path: str, unique_strings: Optional[Set[str]] = None, compiler: Optional[str] = None, compiler_version: Optional[str] = None, os_name: Optional[str] = None, os_version: Optional[str] = None)

Bases: object

This class describes a FLIRT signature.

angr.flirt.FS: alias of angr.flirt.FlirtSignature

angr.flirt.load_signatures(path: str) → None

Recursively load all FLIRT signatures under a specific path.

Parameters: path – Location of FLIRT signatures.

angr.flirt.build_sig.get_basic_info(ar_path: str) → Dict[str, str]: Get basic information of the archive file.

angr.flirt.build_sig.get_unique_strings(ar_path: str) → List[str]: For Linux libraries, this method requires ar (from binutils), nm (from binutils), and strings.

angr.flirt.build_sig.run_pelf(pelf_path: str, ar_path: str, output_path: str)

angr.flirt.build_sig.run_sigmake(sigmake_path: str, sig_name: str, pat_path: str, sig_path: str)

angr.flirt.build_sig.process_exc_file(exc_path: str)

We are doing the stupidest thing possible: For each batch of conflicts, we pick the most likely result baed on a set of predefined rules.

TODO: Add caller-callee-based de-duplication.

angr.flirt.build_sig.main()

Utils

angr.utils.looks_like_sql(s: str) → bool

Determine if string s looks like an SQL query.

Parameters: s (str) – The string to detect.
Returns: True if the string looks like an SQL, False otherwise.

angr.utils.algo.binary_insert(lst: List, elem: Any, key: Callable, lo: int = 0, hi: Optional[int] = None) → None

Insert an element into a sorted list, and keep the list sorted.

The major difference from bisect.bisect_left is that this function supports a key method, so user doesn’t have to create the key array for each insertion.

Parameters

lst (list) – The list. Must be pre-ordered.
element (object) – An element to insert into the list.
key (func) – A method to get the key for each element in the list.
lo (int) – Lower bound of the search.
hi (int) – Upper bound of the search.

Returns

None

angr.utils.constants.is_alignment_mask(n)

class angr.utils.cowdict.ChainMapCOW(*args, collapse_threshold=None)

Bases: collections.ChainMap

Implements a copy-on-write version of ChainMap that supports auto-collapsing.

copy()

clean()

class angr.utils.cowdict.DefaultChainMapCOW(default_factory, *args, collapse_threshold=None)

Bases: angr.utils.cowdict.ChainMapCOW

Implements a copy-on-write version of ChainMap with default values that supports auto-collapsing.

clean()

angr.utils.enums_conv.cfg_jumpkind_to_pb(jk)

angr.utils.enums_conv.func_edge_type_to_pb(jk)

angr.utils.enums_conv.cfg_jumpkind_from_pb(pb)

angr.utils.enums_conv.func_edge_type_from_pb(pb)

angr.utils.graph.shallow_reverse(g) → networkx.classes.digraph.DiGraph

Make a shallow copy of a directional graph and reverse the edges. This is a workaround to solve the issue that one cannot easily make a shallow reversed copy of a graph in NetworkX 2, since networkx.reverse(copy=False) now returns a GraphView, and GraphViews are always read-only.

Parameters: g (networkx.DiGraph) – The graph to reverse.
Returns: A new networkx.DiGraph that has all nodes and all edges of the original graph, with edges reversed.

angr.utils.graph.dfs_back_edges(graph, start_node)

Do a DFS traversal of the graph, and return with the back edges.

Note: This is just a naive recursive implementation, feel free to replace it. I couldn’t find anything in networkx to do this functionality. Although the name suggest it, but dfs_labeled_edges is doing something different.

Parameters

graph – The graph to traverse.
node – The node where to start the traversal

Returns

An iterator of ‘backward’ edges

angr.utils.graph.subgraph_between_nodes(graph, source, frontier, include_frontier=False)

For a directed graph, return a subgraph that includes all nodes going from a source node to a target node.

Parameters

graph (networkx.DiGraph) – The directed graph.
source – The source node.
frontier (list) – A collection of target nodes.
include_frontier (bool) – Should nodes in frontier be included in the subgraph.

Returns

A subgraph.

Return type

networkx.DiGraph

angr.utils.graph.dominates(idom, dominator_node, node)

angr.utils.graph.compute_dominance_frontier(graph, domtree)

Compute a dominance frontier based on the given post-dominator tree.

This implementation is based on figure 2 of paper An Efficient Method of Computing Static Single Assignment Form by Ron Cytron, etc.

Parameters

graph – The graph where we want to compute the dominance frontier.
domtree – The dominator tree

Returns

A dict of dominance frontier

class angr.utils.graph.TemporaryNode(label)

Bases: object

A temporary node.

Used as the start node and end node in post-dominator tree generation. Also used in some test cases.

class angr.utils.graph.ContainerNode(obj)

Bases: object

A container node.

Only used in dominator tree generation. We did this so we can set the index property without modifying the original object.

index

property obj

class angr.utils.graph.Dominators(graph, entry_node, successors_func=None, reverse=False)

Bases: object

dom: networkx.classes.digraph.DiGraph

class angr.utils.graph.PostDominators(graph, entry_node, successors_func=None)

Bases: angr.utils.graph.Dominators

property post_dom: networkx.classes.digraph.DiGraph

dom: networkx.classes.digraph.DiGraph

angr.utils.library.get_function_name(s)

Get the function name from a C-style function declaration string.

Parameters: s (str) – A C-style function declaration string.
Returns: The function name.
Return type: str

angr.utils.library.register_kernel_types()

angr.utils.library.convert_cproto_to_py(c_decl) → Tuple[str, SimTypeFunction, str]

Convert a C-style function declaration string to its corresponding SimTypes-based Python representation.

Parameters: c_decl (str) – The C-style function declaration string.
Returns: A tuple of the function name, the prototype, and a string representing the SimType-based Python representation.

angr.utils.library.convert_cppproto_to_py(cpp_decl: str, with_param_names: bool = False) → Tuple[Optional[str], Optional[angr.sim_type.SimTypeCppFunction], Optional[str]]

Pre-process a C++-style function declaration string to its corresponding SimTypes-based Python representation.

Parameters: cpp_decl – The C++-style function declaration string.
Returns: A tuple of the function name, the prototype, and a string representing the SimType-based Python representation.

angr.utils.library.parsedcprotos2py(parsed_cprotos: List[Tuple[str, SimTypeFunction, str]], fd_spots=frozenset({}), remove_sys_prefix=False) → str

Parse a list of C function declarations and output to Python code that can be embedded into angr.procedures.definitions.

>>> # parse the list of glibc C prototypes and output to a file
>>> from angr.procedures.definitions import glibc
>>> with open("glibc_protos", "w") as f: f.write(cprotos2py(glibc._libc_c_decls))

Parameters: parsed_cprotos – A list of tuples where each tuple is (function name, parsed C function prototype, the original function declaration).
Returns: A Python string.

angr.utils.library.cprotos2py(cprotos: List[str], fd_spots=frozenset({}), remove_sys_prefix=False) → str

Parse a list of C function declarations and output to Python code that can be embedded into angr.procedures.definitions.

>>> # parse the list of glibc C prototypes and output to a file
>>> from angr.procedures.definitions import glibc
>>> with open("glibc_protos", "w") as f: f.write(cprotos2py(glibc._libc_c_decls))

Parameters: cprotos – A list of C prototype strings.
Returns: A Python string.

angr.utils.library.get_cpp_function_name(demangled_name, specialized=True, qualified=True)

angr.utils.timing.timethis(func)

angr.utils.formatting.setup_terminal(): Check if we are running in a TTY. If so, make sure the terminal supports ANSI escape sequences. If not, disable colorized output. Sets global ansi_color_enabled to True if colorized output should be enabled by default.

angr.utils.formatting.ansi_color(s: str, color: Optional[str]) → str

Colorize string s by wrapping in ANSI escape sequence for given color.

This function does not consider whether escape sequences are functional or not; it is up to the caller to determine if its appropriate. Check global ansi_color_enabled value in this module.

angr.utils.formatting.add_edge_to_buffer(buf: Sequence[str], ref: Sequence[str], start: int, end: int, formatter: Optional[Callable[str, str]] = None, dashed: bool = False, ascii_only: Optional[bool] = None)

Draw an edge by adding Unicode box and arrow glyphs to beginning of each line in a list of lines.

Parameters

buf – Output buffer, used to render formatted edges.
ref – Reference buffer, used to calculate edge depth.
start – Start line.
end – End line, where arrow points.
formatter – Optional callback function used to format the edge before writing it to output buffer.
dashed – Render edge line dashed instead of solid.
ascii_only – Render edge using ASCII characters only. If unspecified, guess by stdout encoding.

Returns

Errors

exception angr.errors.AngrError: Bases: Exception

exception angr.errors.AngrValueError: Bases: angr.errors.AngrError, ValueError

exception angr.errors.AngrLifterError: Bases: angr.errors.AngrError

exception angr.errors.AngrExitError: Bases: angr.errors.AngrError

exception angr.errors.AngrPathError: Bases: angr.errors.AngrError

exception angr.errors.AngrVaultError: Bases: angr.errors.AngrError

exception angr.errors.PathUnreachableError: Bases: angr.errors.AngrPathError

exception angr.errors.SimulationManagerError: Bases: angr.errors.AngrError

exception angr.errors.AngrInvalidArgumentError: Bases: angr.errors.AngrError

exception angr.errors.AngrSurveyorError: Bases: angr.errors.AngrError

exception angr.errors.AngrAnalysisError: Bases: angr.errors.AngrError

exception angr.errors.AngrBladeError: Bases: angr.errors.AngrError

exception angr.errors.AngrBladeSimProcError: Bases: angr.errors.AngrBladeError

exception angr.errors.AngrAnnotatedCFGError: Bases: angr.errors.AngrError

exception angr.errors.AngrBackwardSlicingError: Bases: angr.errors.AngrError

exception angr.errors.AngrGirlScoutError: Bases: angr.errors.AngrError

exception angr.errors.AngrCallableError: Bases: angr.errors.AngrSurveyorError

exception angr.errors.AngrCallableMultistateError: Bases: angr.errors.AngrCallableError

exception angr.errors.AngrSyscallError: Bases: angr.errors.AngrError

exception angr.errors.AngrSimOSError: Bases: angr.errors.AngrError

exception angr.errors.AngrAssemblyError: Bases: angr.errors.AngrError

exception angr.errors.AngrIncongruencyError: Bases: angr.errors.AngrAnalysisError

exception angr.errors.AngrForwardAnalysisError: Bases: angr.errors.AngrError

exception angr.errors.AngrSkipJobNotice: Bases: angr.errors.AngrForwardAnalysisError

exception angr.errors.AngrDelayJobNotice: Bases: angr.errors.AngrForwardAnalysisError

exception angr.errors.AngrJobMergingFailureNotice: Bases: angr.errors.AngrForwardAnalysisError

exception angr.errors.AngrJobWideningFailureNotice: Bases: angr.errors.AngrForwardAnalysisError

exception angr.errors.AngrCFGError: Bases: angr.errors.AngrError

exception angr.errors.AngrVFGError: Bases: angr.errors.AngrError

exception angr.errors.AngrVFGRestartAnalysisNotice: Bases: angr.errors.AngrVFGError

exception angr.errors.AngrDataGraphError: Bases: angr.errors.AngrAnalysisError

exception angr.errors.AngrDDGError: Bases: angr.errors.AngrAnalysisError

exception angr.errors.AngrLoopAnalysisError: Bases: angr.errors.AngrAnalysisError

exception angr.errors.AngrExplorationTechniqueError: Bases: angr.errors.AngrError

exception angr.errors.AngrExplorerError: Bases: angr.errors.AngrExplorationTechniqueError

exception angr.errors.AngrDirectorError: Bases: angr.errors.AngrExplorationTechniqueError

exception angr.errors.AngrTracerError: Bases: angr.errors.AngrExplorationTechniqueError

exception angr.errors.AngrVariableRecoveryError: Bases: angr.errors.AngrAnalysisError

exception angr.errors.AngrDBError: Bases: angr.errors.AngrError

exception angr.errors.AngrCorruptDBError: Bases: angr.errors.AngrDBError

exception angr.errors.AngrIncompatibleDBError: Bases: angr.errors.AngrDBError

exception angr.errors.TracerEnvironmentError: Bases: angr.errors.AngrError

exception angr.errors.SimError

Bases: Exception

bbl_addr = None

stmt_idx = None

ins_addr = None

executed_instruction_count = None

guard = None

record_state(state)

exception angr.errors.SimStateError: Bases: angr.errors.SimError

exception angr.errors.SimMergeError: Bases: angr.errors.SimStateError

exception angr.errors.SimMemoryError: Bases: angr.errors.SimStateError

exception angr.errors.SimMemoryMissingError(missing_addr, missing_size, *args): Bases: angr.errors.SimMemoryError

exception angr.errors.SimAbstractMemoryError: Bases: angr.errors.SimMemoryError

exception angr.errors.SimRegionMapError: Bases: angr.errors.SimMemoryError

exception angr.errors.SimMemoryLimitError: Bases: angr.errors.SimMemoryError

exception angr.errors.SimMemoryAddressError: Bases: angr.errors.SimMemoryError

exception angr.errors.SimFastMemoryError: Bases: angr.errors.SimMemoryError

exception angr.errors.SimEventError: Bases: angr.errors.SimStateError

exception angr.errors.SimPosixError: Bases: angr.errors.SimStateError

exception angr.errors.SimFilesystemError: Bases: angr.errors.SimError

exception angr.errors.SimSymbolicFilesystemError: Bases: angr.errors.SimFilesystemError

exception angr.errors.SimFileError: Bases: angr.errors.SimMemoryError, angr.errors.SimFilesystemError

exception angr.errors.SimHeapError: Bases: angr.errors.SimStateError

exception angr.errors.SimUnsupportedError: Bases: angr.errors.SimError

exception angr.errors.SimSolverError: Bases: angr.errors.SimError

exception angr.errors.SimSolverModeError: Bases: angr.errors.SimSolverError

exception angr.errors.SimSolverOptionError: Bases: angr.errors.SimSolverError

exception angr.errors.SimValueError: Bases: angr.errors.SimSolverError

exception angr.errors.SimUnsatError: Bases: angr.errors.SimValueError

exception angr.errors.SimOperationError: Bases: angr.errors.SimError

exception angr.errors.UnsupportedIROpError: Bases: angr.errors.SimOperationError, angr.errors.SimUnsupportedError

exception angr.errors.SimExpressionError: Bases: angr.errors.SimError

exception angr.errors.UnsupportedIRExprError: Bases: angr.errors.SimExpressionError, angr.errors.SimUnsupportedError

exception angr.errors.SimCCallError: Bases: angr.errors.SimExpressionError

exception angr.errors.UnsupportedCCallError: Bases: angr.errors.SimCCallError, angr.errors.SimUnsupportedError

exception angr.errors.SimUninitializedAccessError(expr_type, expr): Bases: angr.errors.SimExpressionError

exception angr.errors.SimStatementError: Bases: angr.errors.SimError

exception angr.errors.UnsupportedIRStmtError: Bases: angr.errors.SimStatementError, angr.errors.SimUnsupportedError

exception angr.errors.UnsupportedDirtyError: Bases: angr.errors.UnsupportedIRStmtError, angr.errors.SimUnsupportedError

exception angr.errors.SimMissingTempError: Bases: angr.errors.SimValueError, IndexError

exception angr.errors.SimEngineError: Bases: angr.errors.SimError

exception angr.errors.SimIRSBError: Bases: angr.errors.SimEngineError

exception angr.errors.SimTranslationError: Bases: angr.errors.SimEngineError

exception angr.errors.SimProcedureError: Bases: angr.errors.SimEngineError

exception angr.errors.SimProcedureArgumentError: Bases: angr.errors.SimProcedureError

exception angr.errors.SimShadowStackError: Bases: angr.errors.SimProcedureError

exception angr.errors.SimFastPathError: Bases: angr.errors.SimEngineError

exception angr.errors.SimIRSBNoDecodeError: Bases: angr.errors.SimIRSBError

exception angr.errors.AngrUnsupportedSyscallError: Bases: angr.errors.AngrSyscallError, angr.errors.SimProcedureError, angr.errors.SimUnsupportedError

angr.errors.UnsupportedSyscallError: alias of angr.errors.AngrUnsupportedSyscallError

exception angr.errors.SimReliftException(state): Bases: angr.errors.SimEngineError

exception angr.errors.SimSlicerError: Bases: angr.errors.SimError

exception angr.errors.SimActionError: Bases: angr.errors.SimError

exception angr.errors.SimCCError: Bases: angr.errors.SimError

exception angr.errors.SimUCManagerError: Bases: angr.errors.SimError

exception angr.errors.SimUCManagerAllocationError: Bases: angr.errors.SimUCManagerError

exception angr.errors.SimUnicornUnsupport: Bases: angr.errors.SimError

exception angr.errors.SimUnicornError: Bases: angr.errors.SimError

exception angr.errors.SimUnicornSymbolic: Bases: angr.errors.SimError

exception angr.errors.SimEmptyCallStackError: Bases: angr.errors.SimError

exception angr.errors.SimStateOptionsError: Bases: angr.errors.SimError

exception angr.errors.SimException: Bases: angr.errors.SimError

exception angr.errors.SimSegfaultException(addr, reason, original_addr=None): Bases: angr.errors.SimException, angr.errors.SimMemoryError

angr.errors.SimSegfaultError: alias of angr.errors.SimSegfaultException

exception angr.errors.SimZeroDivisionException: Bases: angr.errors.SimException, angr.errors.SimOperationError

exception angr.errors.AngrNoPluginError: Bases: angr.errors.AngrError

exception angr.errors.SimConcreteMemoryError: Bases: angr.errors.AngrError

exception angr.errors.SimConcreteRegisterError: Bases: angr.errors.AngrError

exception angr.errors.SimConcreteBreakpointError: Bases: angr.errors.AngrError

exception angr.errors.UnsupportedNodeTypeError: Bases: angr.errors.AngrError, NotImplementedError

Distributed analysis

class angr.distributed.server.Server(project, spill_yard=None, db=None, max_workers=None, max_states=10, staging_max=10, bucketizer=True, recursion_limit=1000, worker_exit_callback=None, techniques=None, add_options=None, remove_options=None)

Bases: object

Server implements the analysis server with a series of control interfaces exposed.

Variables

project – An instance of angr.Project.
spill_yard (str) – A directory to store spilled states.
db (str) – Path of the database that stores information about spilled states.
max_workers (int) – Maximum number of workers. Each worker starts a new process.
max_states (int) – Maximum number of active states for each worker.
staging_max (int) – Maximum number of inactive states that are kept into memory before spilled onto the disk and potentially be picked up by another worker.
bucketizer (bool) – Use the Bucketizer exploration strategy.
_worker_exit_callback – A method that will be called upon the exit of each worker.

inc_active_workers()

dec_active_workers()

stop()

property active_workers

property stopped

on_worker_exit(worker_id, stashes)

run()

class angr.distributed.worker.BadStatesDropper(vault, db)

Bases: angr.exploration_techniques.ExplorationTechnique

Dumps and drops states that are not “active”.

step(simgr, stash='active', **kwargs)

class angr.distributed.worker.ExplorationStatusNotifier(server_state: Dict)

Bases: angr.exploration_techniques.ExplorationTechnique

Force the exploration to stop if the server.stop is True.

step(simgr, stash='active', **kwargs)

class angr.distributed.worker.Worker(worker_id, server, server_state, recursion_limit=None, techniques=None, add_options=None, remove_options=None)

Bases: object

Worker implements a worker thread/process for conducting a task.

start()

run()

angr — Analysis and Coordination

Project

Plugin Ecosystem

Program State

Storage

Memory Mixins

Concretization Strategies

Simulation Manager

Exploration Techniques

Simulation Engines

Simulation Logging

Procedures

Calling Conventions and Types

Knowledge Base

Serialization

Analysis

SimOS

Function Signature Matching

Utils

Errors

Distributed analysis

`angr` — Analysis and Coordination