angr 9.2.103__py3-none-manylinux2014_aarch64.whl

angr/exploration_techniques/unique.py

@@ -0,0 +1,104 @@
+from difflib import SequenceMatcher
+from collections import Counter
+
+from . import ExplorationTechnique
+
+
+class UniqueSearch(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.
+    """
+
+    def __init__(self, similarity_func=None, deferred_stash="deferred"):
+        """
+        :param similarity_func: How to calculate similarity between two states.
+        :param deferred_stash:  Where to store the deferred states.
+        """
+        super().__init__()
+        self.similarity_func = similarity_func or UniqueSearch.similarity
+        self.deferred_stash = deferred_stash
+        self.uniqueness = {}
+        self.num_deadended = 0
+
+    def setup(self, simgr):
+        if self.deferred_stash not in simgr.stashes:
+            simgr.stashes[self.deferred_stash] = []
+
+    def step(self, simgr, stash="active", **kwargs):
+        simgr = simgr.step(stash=stash, **kwargs)
+
+        old_states = simgr.stashes[self.deferred_stash][:]
+        new_states = simgr.stashes[stash][:]
+        simgr.move(from_stash=stash, to_stash=self.deferred_stash)
+
+        def update_average(state, new, mem=1.0):
+            """
+            param state: The state to update the average for.
+            param new:   The new value to be accumulated into the average.
+            param mem:   Memory parameter to determine how to weight the past average.
+            """
+            prev, size = self.uniqueness[state]
+            new_average = float(prev * (size**mem) + new) / ((size**mem) + 1)
+            self.uniqueness[state] = new_average, size + 1
+
+        for state_a in new_states:
+            self.uniqueness[state_a] = 0, 0
+            for state_b in old_states:
+                # Update similarity averages between new and old states
+                similarity = self.similarity_func(state_a, state_b)
+                update_average(state_a, similarity)
+                update_average(state_b, similarity)
+            for state_b in (s for s in new_states if s is not state_a):
+                # Update similarity averages between new states
+                similarity = self.similarity_func(state_a, state_b)
+                update_average(state_a, similarity)
+
+        for state_a in simgr.stashes[self.deferred_stash]:
+            for state_b in simgr.deadended[self.num_deadended :]:
+                # Update similarity averages between all states and newly deadended states
+                similarity = self.similarity_func(state_a, state_b)
+                update_average(state_a, similarity)
+        self.num_deadended = len(simgr.deadended)
+
+        if self.uniqueness:
+            unique_state = min(self.uniqueness.items(), key=lambda e: e[1])[0]
+            del self.uniqueness[unique_state]
+
+            simgr.move(from_stash=self.deferred_stash, to_stash=stash, filter_func=lambda s: s is unique_state)
+
+        return simgr
+
+    @staticmethod
+    def 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
+        """
+        count_a = Counter(state_a.history.bbl_addrs)
+        count_b = Counter(state_b.history.bbl_addrs)
+        normal_distance = (
+            sum(
+                (count_a.get(addr, 0) - count_b.get(addr, 0)) ** 2
+                for addr in set(list(count_a.keys()) + list(count_b.keys()))
+            )
+            ** 0.5
+        )
+        return 1.0 / (1 + normal_distance)
+
+    @staticmethod
+    def 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
+        """
+        addrs_a = tuple(state_a.history.bbl_addrs)
+        addrs_b = tuple(state_b.history.bbl_addrs)
+        return SequenceMatcher(a=addrs_a, b=addrs_b).ratio()

angr/exploration_techniques/veritesting.py

@@ -0,0 +1,36 @@
+from . import ExplorationTechnique
+
+from ..sim_options import EFFICIENT_STATE_MERGING
+
+
+class Veritesting(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
+    """
+
+    def __init__(self, **options):
+        super().__init__()
+        self.options = options
+
+    def step_state(self, simgr, state, successor_func=None, **kwargs):
+        if EFFICIENT_STATE_MERGING not in state.options:
+            state.options.add(EFFICIENT_STATE_MERGING)
+
+        vt = self.project.analyses.Veritesting(state, **self.options)
+        if vt.result and vt.final_manager:
+            simgr2 = vt.final_manager
+            simgr2.stash(from_stash="deviated", to_stash="active")
+            simgr2.stash(from_stash="successful", to_stash="active")
+
+            return {
+                "active": simgr2.active,
+                "unconstrained": simgr2.stashes.get("unconstrained", []),
+                "unsat": simgr2.stashes.get("unsat", []),
+                "pruned": simgr2.stashes.get("pruned", []),
+                "errored": simgr2.errored,
+            }
+
+        return simgr.step_state(state, successor_func=successor_func, **kwargs)

angr/factory.py

@@ -0,0 +1,385 @@
+from typing import overload
+import logging
+import archinfo
+from archinfo.arch_soot import ArchSoot, SootAddressDescriptor
+
+from .sim_state import SimState
+from .calling_conventions import default_cc, SimRegArg, SimStackArg, PointerWrapper, SimCCUnknown
+from .callable import Callable
+from .errors import AngrAssemblyError, AngrError
+from .engines import UberEngine, ProcedureEngine, SimEngineConcrete, SimEngine
+from .sim_type import SimTypeFunction, SimTypeInt
+from .codenode import HookNode, SyscallNode
+from .block import Block, SootBlock
+from .sim_manager import SimulationManager
+
+try:
+    from .engines import UberEnginePcode
+    from .engines.pcode import register_pcode_arch_default_cc
+except ImportError:
+    UberEnginePcode = None
+
+
+l = logging.getLogger(name=__name__)
+
+
+class AngrObjectFactory:
+    """
+    This factory provides access to important analysis elements.
+    """
+
+    def __init__(self, project, default_engine: type[SimEngine] | None = None):
+        if default_engine is None:
+            if isinstance(project.arch, archinfo.ArchPcode) and UberEnginePcode is not None:
+                l.warning("Creating project with the experimental 'UberEnginePcode' engine")
+                default_engine_n = UberEnginePcode
+            else:
+                default_engine_n = UberEngine
+        else:
+            default_engine_n = default_engine
+
+        if isinstance(project.arch, archinfo.ArchPcode):
+            register_pcode_arch_default_cc(project.arch)
+
+        self.project = project
+        self._default_cc = default_cc(
+            project.arch.name, platform=project.simos.name if project.simos is not None else None, default=SimCCUnknown
+        )
+        self.default_engine = default_engine_n(project)
+        self.procedure_engine = ProcedureEngine(project)
+
+        if project.concrete_target:
+            self.concrete_engine = SimEngineConcrete(project)
+        else:
+            self.concrete_engine = None
+
+    def snippet(self, addr, jumpkind=None, **block_opts):
+        if self.project.is_hooked(addr) and jumpkind != "Ijk_NoHook":
+            hook = self.project._sim_procedures[addr]
+            size = hook.kwargs.get("length", 0)
+            return HookNode(addr, size, self.project.hooked_by(addr))
+        elif self.project.simos.is_syscall_addr(addr):
+            syscall = self.project.simos.syscall_from_addr(addr)
+            size = syscall.kwargs.get("length", 0)
+            return SyscallNode(addr, size, syscall)
+        else:
+            return self.block(addr, **block_opts).codenode  # pylint: disable=no-member
+
+    def successors(self, *args, engine=None, **kwargs):
+        """
+        Perform execution using an engine. Generally, return a SimSuccessors object classifying the results of the run.
+
+        :param state:           The state to analyze
+        :param engine:          The engine to use. If not provided, will use the project default.
+        :param addr:            optional, an address to execute at instead of the state's ip
+        :param jumpkind:        optional, the jumpkind of the previous exit
+        :param 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.
+        """
+        if engine is not None:
+            return engine.process(*args, **kwargs)
+        return self.default_engine.process(*args, **kwargs)
+
+    def blank_state(self, **kwargs):
+        """
+        Returns a mostly-uninitialized state object. All parameters are optional.
+
+        :param addr:            The address the state should start at instead of the entry point.
+        :param initial_prefix:  If this is provided, all symbolic registers will hold symbolic values with names
+                                prefixed by this string.
+        :param fs:              A dictionary of file names with associated preset SimFile objects.
+        :param concrete_fs:     bool describing whether the host filesystem should be consulted when opening files.
+        :param 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.
+        :param kwargs:          Any additional keyword args will be passed to the SimState constructor.
+        :return:                The blank state.
+        :rtype:                 SimState
+        """
+        return self.project.simos.state_blank(**kwargs)
+
+    def entry_state(self, **kwargs) -> SimState:
+        """
+        Returns a state object representing the program at its entry point. All parameters are optional.
+
+        :param addr:            The address the state should start at instead of the entry point.
+        :param initial_prefix:  If this is provided, all symbolic registers will hold symbolic values with names
+                                prefixed by this string.
+        :param fs:              a dictionary of file names with associated preset SimFile objects.
+        :param concrete_fs:     boolean describing whether the host filesystem should be consulted when opening files.
+        :param 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.
+        :param 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.
+        :param args:            a list of values to use as the program's argv. May be mixed strings and bitvectors.
+        :param env:             a dictionary to use as the environment for the program. Both keys and values may be
+                                mixed strings and bitvectors.
+        :return:                The entry state.
+        :rtype:                 SimState
+        """
+
+        return self.project.simos.state_entry(**kwargs)
+
+    def full_init_state(self, **kwargs):
+        """
+        Very much like :meth:`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``.
+        """
+        return self.project.simos.state_full_init(**kwargs)
+
+    def call_state(self, addr, *args, **kwargs):
+        """
+        Returns a state object initialized to the start of a given function, as if it were called with given parameters.
+
+        :param addr:            The address the state should start at instead of the entry point.
+        :param args:            Any additional positional arguments will be used as arguments to the function call.
+
+        The following parameters are optional.
+
+        :param base_state:      Use this SimState as the base for the new state instead of a blank state.
+        :param cc:              Optionally provide a SimCC object to use a specific calling convention.
+        :param ret_addr:        Use this address as the function's return target.
+        :param stack_base:      An optional pointer to use as the top of the stack, circa the function entry point
+        :param alloc_base:      An optional pointer to use as the place to put excess argument data
+        :param grow_like_stack: When allocating data at alloc_base, whether to allocate at decreasing addresses
+        :param toc:             The address of the table of contents for ppc64
+        :param initial_prefix:  If this is provided, all symbolic registers will hold symbolic values with names
+                                prefixed by this string.
+        :param fs:              A dictionary of file names with associated preset SimFile objects.
+        :param concrete_fs:     bool describing whether the host filesystem should be consulted when opening files.
+        :param 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.
+        :param kwargs:          Any additional keyword args will be passed to the SimState constructor.
+        :return:                The state at the beginning of the function.
+        :rtype:                 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.
+        """
+        return self.project.simos.state_call(addr, *args, **kwargs)
+
+    def simulation_manager(self, thing: list[SimState] | SimState | None = None, **kwargs) -> "SimulationManager":
+        """
+        Constructs a new simulation manager.
+
+        :param thing:           What to put in the new SimulationManager's active stash (either a SimState or a list of
+                                SimStates).
+        :param kwargs:          Any additional keyword arguments will be passed to the SimulationManager constructor
+        :returns:               The new SimulationManager
+        :rtype:                 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. :meth:`entry_state()`.
+        * If a :class:`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.
+        """
+        if thing is None:
+            thing = [self.entry_state()]
+        elif isinstance(thing, (list, tuple)):
+            if any(not isinstance(val, SimState) for val in thing):
+                raise AngrError("Bad type to initialize SimulationManager")
+        elif isinstance(thing, SimState):
+            thing = [thing]
+        else:
+            raise AngrError("BadType to initialze SimulationManager: %s" % repr(thing))
+
+        return SimulationManager(self.project, active_states=thing, **kwargs)
+
+    def simgr(self, *args, **kwargs):
+        """
+        Alias for `simulation_manager` to save our poor fingers
+        """
+        return self.simulation_manager(*args, **kwargs)
+
+    def callable(
+        self,
+        addr,
+        prototype=None,
+        concrete_only=False,
+        perform_merge=True,
+        base_state=None,
+        toc=None,
+        cc=None,
+        add_options=None,
+        remove_options=None,
+    ):
+        """
+        A Callable is a representation of a function in the binary that can be interacted with like a native python
+        function.
+
+        :param addr:            The address of the function to use
+        :param prototype:         The prototype of the call to use, as a string or a SimTypeFunction
+        :param concrete_only:   Throw an exception if the execution splits into multiple states
+        :param perform_merge:   Merge all result states into one at the end (only relevant if concrete_only=False)
+        :param base_state:      The state from which to do these runs
+        :param toc:             The address of the table of contents for ppc64
+        :param 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.
+        :rtype:                 angr.callable.Callable
+        """
+        return Callable(
+            self.project,
+            addr=addr,
+            prototype=prototype,
+            concrete_only=concrete_only,
+            perform_merge=perform_merge,
+            base_state=base_state,
+            toc=toc,
+            cc=cc,
+            add_options=add_options,
+            remove_options=remove_options,
+        )
+
+    def cc(self):
+        """
+        Return a SimCC (calling convention) parameterized 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.
+        """
+
+        return self._default_cc(arch=self.project.arch)
+
+    def function_prototype(self):
+        """
+        Return a default function prototype parameterized for this project and SimOS.
+        """
+        return SimTypeFunction(tuple(), SimTypeInt()).with_arch(self.project.arch)
+
+    # pylint: disable=unused-argument, no-self-use, function-redefined
+    @overload
+    def block(
+        self,
+        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,  # backward compatibility
+        strict_block_end=None,
+        collect_data_refs=False,
+        cross_insn_opt=True,
+        load_from_ro_regions=False,
+        initial_regs=None,
+        skip_stmts=False,
+    ) -> "Block": ...
+
+    # pylint: disable=unused-argument, no-self-use, function-redefined
+    @overload
+    def block(
+        self,
+        addr: 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,  # backward compatibility
+        strict_block_end=None,
+        collect_data_refs=False,
+        cross_insn_opt=True,
+        skip_stmts=False,
+    ) -> "SootBlock": ...
+
+    def block(
+        self,
+        addr,
+        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,  # backward compatibility
+        strict_block_end=None,
+        collect_data_refs=False,
+        cross_insn_opt=True,
+        load_from_ro_regions=False,
+        initial_regs=None,
+        skip_stmts=False,
+    ):
+        if isinstance(self.project.arch, ArchSoot) and isinstance(addr, SootAddressDescriptor):
+            return SootBlock(addr, arch=self.project.arch, project=self.project)
+
+        if insn_bytes is not None and insn_text is not None:
+            raise AngrError("You cannot provide both 'insn_bytes' and 'insn_text'!")
+
+        if insn_bytes is not None:
+            byte_string = insn_bytes
+
+        if insn_text is not None:
+            byte_string = self.project.arch.asm(insn_text, addr=addr, as_bytes=True, thumb=thumb)
+            if byte_string is None:
+                # assembly failed
+                raise AngrAssemblyError(
+                    "Assembling failed. Please make sure keystone is installed, and the assembly string is correct."
+                )
+
+        return Block(
+            addr,
+            project=self.project,
+            size=size,
+            max_size=max_size,
+            byte_string=byte_string,
+            vex=vex,
+            extra_stop_points=extra_stop_points,
+            thumb=thumb,
+            backup_state=backup_state,
+            opt_level=opt_level,
+            num_inst=num_inst,
+            traceflags=traceflags,
+            strict_block_end=strict_block_end,
+            collect_data_refs=collect_data_refs,
+            cross_insn_opt=cross_insn_opt,
+            load_from_ro_regions=load_from_ro_regions,
+            initial_regs=initial_regs,
+            skip_stmts=skip_stmts,
+        )
+
+    def fresh_block(self, addr, size, backup_state=None):
+        return Block(addr, project=self.project, size=size, backup_state=backup_state)
+
+    cc.SimRegArg = SimRegArg
+    cc.SimStackArg = SimStackArg
+    callable.PointerWrapper = PointerWrapper
+    call_state.PointerWrapper = PointerWrapper

angr/flirt/__init__.py

@@ -0,0 +1,126 @@
+# This submodule stores and manages FLIRT signatures
+
+from typing import Set, List, Dict, Optional
+import os
+import json
+from collections import defaultdict
+import logging
+
+import nampa
+
+_l = logging.getLogger(__name__)
+
+
+class FlirtSignature:
+    """
+    This class describes a FLIRT signature.
+    """
+
+    def __init__(
+        self,
+        arch: str,
+        platform: str,
+        sig_name: str,
+        sig_path: str,
+        unique_strings: set[str] | None = None,
+        compiler: str | None = None,
+        compiler_version: str | None = None,
+        os_name: str | None = None,
+        os_version: str | None = None,
+    ):
+        self.arch = arch
+        self.platform = platform
+        self.sig_name = sig_name
+        self.sig_path = sig_path
+        self.unique_strings = unique_strings
+        self.compiler = compiler
+        self.compiler_version = compiler_version
+        self.os_name = os_name
+        self.os_version = os_version
+
+    def __repr__(self):
+        if self.os_name:
+            if self.os_version:
+                return f"<{self.sig_name}@{self.arch}-{self.os_name}-{self.os_version}>"
+            return f"<{self.sig_name}@{self.arch}-{self.os_name}>"
+        return f"<{self.sig_name}@{self.arch}-{self.platform}>"
+
+
+FS = FlirtSignature
+
+# A dict from architecture names to FLIRT signatures under that architecture. Arch names are always in lower case.
+FLIRT_SIGNATURES_BY_ARCH: dict[str, list[FlirtSignature]] = defaultdict(list)
+LIBRARY_TO_SIGNATURES: dict[str, list[FlirtSignature]] = defaultdict(list)
+STRING_TO_LIBRARIES: dict[str, set[str]] = defaultdict(set)
+
+
+def load_signatures(path: str) -> None:
+    """
+    Recursively load all FLIRT signatures under a specific path.
+
+    :param path:    Location of FLIRT signatures.
+    """
+
+    FLIRT_SIGNATURES_BY_ARCH.clear()
+    LIBRARY_TO_SIGNATURES.clear()
+    STRING_TO_LIBRARIES.clear()
+
+    for root, _, filenames in os.walk(path):
+        for filename in filenames:
+            if filename.endswith(".sig"):
+                # parse it
+                sig_path = os.path.join(root, filename)
+                try:
+                    with open(sig_path, "rb") as f:
+                        flirt_header = nampa.flirt.parse_header(f)
+                except nampa.flirt.FlirtException:
+                    _l.warning("Failed to load FLIRT signature file %s.", sig_path)
+                    continue
+
+                # is there a meta data file?
+                meta_path = os.path.join(root, filename[:-4] + ".meta")
+                if os.path.isfile(meta_path):
+                    # yes!
+                    with open(meta_path) as f:
+                        meta = json.load(f)
+
+                    arch = meta.get("arch", None)
+                    platform = meta.get("platform", None)
+                    os_name = meta.get("os", None)
+                    os_version = meta.get("os_version", None)
+                    compiler = meta.get("compiler", None)
+                    compiler_version = meta.get("compiler_version", None)
+                    unique_strings = meta.get("unique_strings", None)
+
+                else:
+                    # nope... we need to extract information from the signature file
+                    # TODO: Convert them to angr-specific strings
+                    arch = flirt_header.arch
+                    platform = flirt_header.os_types
+                    os_name = None
+                    os_version = None
+                    unique_strings = None
+                    compiler = None
+                    compiler_version = None
+
+                signature = FlirtSignature(
+                    arch,
+                    platform,
+                    flirt_header.library_name.decode("utf-8"),
+                    sig_path,
+                    unique_strings=unique_strings,
+                    compiler=compiler,
+                    compiler_version=compiler_version,
+                    os_name=os_name,
+                    os_version=os_version,
+                )
+
+                FLIRT_SIGNATURES_BY_ARCH[arch].append(signature)
+
+    # fill in LIBRARY_TO_SIGNATURES and STRING_TO_LIBRARIES
+    for sigs in FLIRT_SIGNATURES_BY_ARCH.values():
+        for sig in sigs:
+            LIBRARY_TO_SIGNATURES[sig.sig_name].append(sig)
+            if sig.unique_strings:
+                for us in sig.unique_strings:
+                    STRING_TO_LIBRARIES[us].add(sig.sig_name)