diff --git a/.github/workflows/build-pyinstaller.yml b/.github/workflows/build-pyinstaller.yml new file mode 100644 index 000000000..bcba95403 --- /dev/null +++ b/.github/workflows/build-pyinstaller.yml @@ -0,0 +1,50 @@ +name: build-pyinstaller +on: + push: + branches: + - stable + - develop + - 'release/**' + pull_request: + branches: + - stable + - 'release/**' + +jobs: + + exe: + runs-on: windows-latest + strategy: + matrix: + python-version: ["3.11"] + steps: + - uses: actions/checkout@v3 + - name: Set up Python ${{ matrix.python-version }} + uses: actions/setup-python@v4 + with: + python-version: ${{ matrix.python-version }} + + - name: Install dependencies + run: | + python -m pip install --upgrade pip + pip install pyinstaller + + - name: Pyinstall executable + run: | + pyinstaller --clean -y vol.spec + pyinstaller --clean -y volshell.spec + + - name: Move files + run: | + mv dist/vol.exe vol.exe + mv dist/volshell.exe volshell.exe + + - name: Archive + uses: actions/upload-artifact@v4 + with: + name: volatility3-pyinstaller + path: | + vol.exe + volshell.exe + README.md + LICENSE.txt diff --git a/README.md b/README.md index cc33d3cc4..b74bdab0b 100644 --- a/README.md +++ b/README.md @@ -88,7 +88,7 @@ The latest generated copy of the documentation can be found at: =2024.8.26", -] +dependencies = ["pefile>=2024.8.26"] [project.optional-dependencies] full = [ @@ -26,15 +32,12 @@ full = [ "pillow>=10.0.0,<11.0.0", ] -cloud = [ - "gcsfs>=2024.10.0", - "s3fs>=2024.10.0", -] +cloud = ["gcsfs>=2024.10.0", "s3fs>=2024.10.0"] dev = [ "volatility3[full,cloud]", "jsonschema>=4.23.0,<5", - "pyinstaller>=6.11.0,<7", + "pyinstaller>=6.5.0,<7", "pyinstaller-hooks-contrib>=2024.9", "types-jsonschema>=4.23.0,<5", ] @@ -48,8 +51,8 @@ test = [ docs = [ "volatility3[dev]", - "sphinx>=8.0.0,<7", - "sphinx-autodoc-typehints>=2.5.0,<3", + "sphinx>=4.0.0,<9", + "sphinx-autodoc-typehints>=2.0.0,<3", "sphinx-rtd-theme>=3.0.1,<4", ] @@ -79,16 +82,16 @@ target-version = "py38" [tool.ruff.lint] select = [ - "F", # pyflakes - "E", # pycodestyle errors - "W", # pycodestyle warnings - "G", # flake8-logging-format - "PIE", # flake8-pie - "UP", # pyupgrade + "F", # pyflakes + "E", # pycodestyle errors + "W", # pycodestyle warnings + "G", # flake8-logging-format + "PIE", # flake8-pie + "UP", # pyupgrade ] ignore = [ - "E501", # ignore due to conflict with formatter + "E501", # ignore due to conflict with formatter ] [build-system] diff --git a/test/test_volatility.py b/test/test_volatility.py index bb7c9a851..8676d1f3e 100644 --- a/test/test_volatility.py +++ b/test/test_volatility.py @@ -708,6 +708,24 @@ def test_linux_page_cache_inodepages(image, volatility, python): inode_address = hex(0x88001AB5C270) inode_dump_filename = f"inode_{inode_address}.dmp" + + rc, out, _err = runvol_plugin( + "linux.pagecache.InodePages", + image, + volatility, + python, + pluginargs=["--inode", inode_address], + ) + + assert rc == 0 + assert out.count(b"\n") > 4 + + # PageVAddr PagePAddr MappingAddr .. DumpSafe + assert re.search( + rb"0xea000054c5f8\s0x18389000\s0x88001ab5c3b0.*?True", + out, + ) + try: rc, out, _err = runvol_plugin( "linux.pagecache.InodePages", @@ -718,13 +736,8 @@ def test_linux_page_cache_inodepages(image, volatility, python): ) assert rc == 0 - assert out.count(b"\n") > 4 + assert out.count(b"\n") >= 4 - # PageVAddr PagePAddr MappingAddr .. DumpSafe - assert re.search( - rb"0xea000054c5f8\s0x18389000\s0x88001ab5c3b0.*?True", - out, - ) assert os.path.exists(inode_dump_filename) with open(inode_dump_filename, "rb") as fp: inode_contents = fp.read() diff --git a/volatility3/cli/__init__.py b/volatility3/cli/__init__.py index 6172a17f3..82a2a4205 100644 --- a/volatility3/cli/__init__.py +++ b/volatility3/cli/__init__.py @@ -367,6 +367,7 @@ class CommandLine: plugin, help=plugin_list[plugin].__doc__, description=plugin_list[plugin].__doc__, + epilog=plugin_list[plugin].additional_description, ) self.populate_requirements_argparse(plugin_parser, plugin_list[plugin]) diff --git a/volatility3/cli/volshell/linux.py b/volatility3/cli/volshell/linux.py index cc58fa1c2..b3689c3ae 100644 --- a/volatility3/cli/volshell/linux.py +++ b/volatility3/cli/volshell/linux.py @@ -3,6 +3,7 @@ # from typing import Any, List, Optional, Tuple, Union +from enum import Enum from volatility3.cli.volshell import generic from volatility3.framework import constants, interfaces @@ -10,6 +11,16 @@ from volatility3.framework.configuration import requirements from volatility3.plugins.linux import pslist +# Could import the enum from psscan.py to avoid code duplication +class DescExitStateEnum(Enum): + """Enum for linux task exit_state as defined in include/linux/sched.h""" + + TASK_RUNNING = 0x00000000 + EXIT_DEAD = 0x00000010 + EXIT_ZOMBIE = 0x00000020 + EXIT_TRACE = EXIT_ZOMBIE | EXIT_DEAD + + class Volshell(generic.Volshell): """Shell environment to directly interact with a linux memory image.""" @@ -40,6 +51,71 @@ class Volshell(generic.Volshell): return None print(f"No task with task ID {pid} found") + def get_process(self, pid=None, virtaddr=None, physaddr=None): + """Return the task_struct object that matches the pid. If a physical or a virtual address is provided, construct the task_struct object at said address. Only one parameter is allowed. + + Args: + pid (int, optional): PID to search for + virtaddr (int, optional): Virtual address to construct object at + physaddr (int, optional): Physical address to construct object at + + Returns: + ObjectInterface: task_struct Object + """ + + if sum(1 if x is not None else 0 for x in [pid, virtaddr, physaddr]) != 1: + print("Only one parameter is accepted") + return None + + vmlinux_module_name = self.config["kernel"] + vmlinux = self.context.modules[vmlinux_module_name] + + kernel_layer_name = vmlinux.layer_name + kernel_layer = self.context.layers[kernel_layer_name] + + memory_layer_name = kernel_layer.dependencies[0] + + task_struct_symbol = vmlinux.symbol_table_name + constants.BANG + "task_struct" + + if virtaddr is not None: + task = self.context.object( + task_struct_symbol, + layer_name=kernel_layer_name, + offset=virtaddr, + ) + + if physaddr is not None: + task = self.context.object( + task_struct_symbol, + layer_name=memory_layer_name, + offset=physaddr, + native_layer_name=kernel_layer_name, + ) + + if physaddr is not None or virtaddr is not None: + try: + DescExitStateEnum(task.exit_state) + except ValueError: + print( + f"task_struct @ {hex(task.vol.offset)} as exit_state {task.exit_state} is likely not valid" + ) + + if not (0 < task.pid < 65535): + print( + f"task_struct @ {hex(task.vol.offset)} as pid {task.pid} is likely not valid" + ) + + return task + + if pid is not None: + tasks = self.list_tasks() + for task in tasks: + if task.pid == pid: + return task + print(f"No task with task ID {pid} found") + + return None + def list_tasks(self): """Returns a list of task objects from the primary layer""" # We always use the main kernel memory and associated symbols @@ -50,6 +126,7 @@ class Volshell(generic.Volshell): result += [ (["ct", "change_task", "cp"], self.change_task), (["lt", "list_tasks", "ps"], self.list_tasks), + (["gp", "get_process", "get_task"], self.get_process), (["symbols"], self.context.symbol_space[self.current_symbol_table]), ] if self.config.get("pid", None) is not None: diff --git a/volatility3/cli/volshell/windows.py b/volatility3/cli/volshell/windows.py index 303d4d5c3..9b89a8b81 100644 --- a/volatility3/cli/volshell/windows.py +++ b/volatility3/cli/volshell/windows.py @@ -44,11 +44,67 @@ class Volshell(generic.Volshell): ) ) + def get_process(self, pid=None, virtaddr=None, physaddr=None): + """Returns the _EPROCESS object that matches the pid. If a physical or a virtual address is provided, construct the _EPROCESS object at said address. Only one parameter is allowed. + + Args: + pid (int, optional): PID / UniqueProcessId to search for. + virtaddr (int, optional): Virtual address to construct object at + physaddr (int, optional): Physical address to construct object at + + Returns: + ObjectInterface: _EPROCESS Object + """ + + if sum(1 if x is not None else 0 for x in [pid, virtaddr, physaddr]) != 1: + print("Only one parameter is accepted") + return None + + kernel_name = self.config["kernel"] + kernel = self.context.modules[kernel_name] + + kernel_layer_name = kernel.layer_name + + kernel_layer = self.context.layers[kernel_layer_name] + memory_layer_name = kernel_layer.dependencies[0] + + eprocess_symbol = kernel.symbol_table_name + constants.BANG + "_EPROCESS" + + if virtaddr is not None: + eproc = self.context.object( + eprocess_symbol, + layer_name=kernel_layer_name, + offset=virtaddr, + ) + + return eproc + + if physaddr is not None: + eproc = self.context.object( + eprocess_symbol, + layer_name=memory_layer_name, + offset=physaddr, + native_layer_name=kernel_layer_name, + ) + + return eproc + + if pid is not None: + processes = self.list_processes() + for process in processes: + if process.UniqueProcessId == pid: + return process + print(f"No process with process ID {pid} found") + return None + + return None + def construct_locals(self) -> List[Tuple[List[str], Any]]: result = super().construct_locals() result += [ (["cp", "change_process"], self.change_process), (["lp", "list_processes", "ps"], self.list_processes), + (["gp", "get_process"], self.get_process), (["symbols"], self.context.symbol_space[self.current_symbol_table]), ] if self.config.get("pid", None) is not None: diff --git a/volatility3/framework/__init__.py b/volatility3/framework/__init__.py index a1925faef..754939460 100644 --- a/volatility3/framework/__init__.py +++ b/volatility3/framework/__init__.py @@ -5,6 +5,7 @@ # Check the python version to ensure it's suitable import glob import sys +from volatility3.framework import check_python_version as check_python_version import zipfile import importlib import inspect @@ -57,7 +58,7 @@ class NonInheritable: self.default_value = value self.cls = cls - def __get__(self, obj: Any, get_type: Optional[Type] = None) -> Any: + def __get__(self, obj: Any, get_type: Type = Optional[None]) -> Any: if type is self.cls: if hasattr(self.default_value, "__get__"): return self.default_value.__get__(obj, get_type) diff --git a/volatility3/framework/automagic/linux.py b/volatility3/framework/automagic/linux.py index f22cae012..cb4f3cc64 100644 --- a/volatility3/framework/automagic/linux.py +++ b/volatility3/framework/automagic/linux.py @@ -71,6 +71,11 @@ class LinuxIntelStacker(interfaces.automagic.StackerLayerInterface): elif "init_level4_pgt" in table.symbols: layer_class = intel.LinuxIntel32e dtb_symbol_name = "init_level4_pgt" + elif "pkmap_count" in table.symbols and table.get_symbol( + "pkmap_count" + ).type.count in (512, 2048): + layer_class = intel.LinuxIntelPAE + dtb_symbol_name = "swapper_pg_dir" else: layer_class = intel.LinuxIntel dtb_symbol_name = "swapper_pg_dir" diff --git a/volatility3/framework/automagic/symbol_cache.py b/volatility3/framework/automagic/symbol_cache.py index 065eb6d43..cd1a348a4 100644 --- a/volatility3/framework/automagic/symbol_cache.py +++ b/volatility3/framework/automagic/symbol_cache.py @@ -301,6 +301,13 @@ class SqliteCache(CacheManagerInterface): This also updates remote locations based on a cache timeout. """ + if progress_callback is None: + + def dummy_progress(*args, **kargs) -> None: + return None + + progress_callback = dummy_progress + on_disk_locations = set( [ filename diff --git a/volatility3/framework/constants/_version.py b/volatility3/framework/constants/_version.py index 694375538..689e39664 100644 --- a/volatility3/framework/constants/_version.py +++ b/volatility3/framework/constants/_version.py @@ -1,6 +1,6 @@ # We use the SemVer 2.0.0 versioning scheme VERSION_MAJOR = 2 # Number of releases of the library with a breaking change -VERSION_MINOR = 14 # Number of changes that only add to the interface +VERSION_MINOR = 18 # Number of changes that only add to the interface VERSION_PATCH = 1 # Number of changes that do not change the interface VERSION_SUFFIX = "" diff --git a/volatility3/framework/constants/linux/__init__.py b/volatility3/framework/constants/linux/__init__.py index 6e49e6f37..ee3a6c748 100644 --- a/volatility3/framework/constants/linux/__init__.py +++ b/volatility3/framework/constants/linux/__init__.py @@ -6,6 +6,7 @@ Linux-specific values that aren't found in debug symbols """ from enum import IntEnum, Flag +from dataclasses import dataclass KERNEL_NAME = "__kernel__" @@ -352,3 +353,57 @@ NSEC_PER_SEC = 1e9 MODULE_MAXIMUM_CORE_SIZE = 20000000 MODULE_MAXIMUM_CORE_TEXT_SIZE = 20000000 MODULE_MINIMUM_SIZE = 4096 + + +@dataclass +class TaintFlag: + shift: int + desc: str + when_present: bool + module: bool + + +TAINT_FLAGS = { + "P": TaintFlag( + shift=1 << 0, desc="PROPRIETARY_MODULE", when_present=True, module=True + ), + "G": TaintFlag( + shift=1 << 0, desc="PROPRIETARY_MODULE", when_present=False, module=True + ), + "F": TaintFlag(shift=1 << 1, desc="FORCED_MODULE", when_present=True, module=False), + "S": TaintFlag( + shift=1 << 2, desc="CPU_OUT_OF_SPEC", when_present=True, module=False + ), + "R": TaintFlag(shift=1 << 3, desc="FORCED_RMMOD", when_present=True, module=False), + "M": TaintFlag(shift=1 << 4, desc="MACHINE_CHECK", when_present=True, module=False), + "B": TaintFlag(shift=1 << 5, desc="BAD_PAGE", when_present=True, module=False), + "U": TaintFlag(shift=1 << 6, desc="USER", when_present=True, module=False), + "D": TaintFlag(shift=1 << 7, desc="DIE", when_present=True, module=False), + "A": TaintFlag( + shift=1 << 8, desc="OVERRIDDEN_ACPI_TABLE", when_present=True, module=False + ), + "W": TaintFlag(shift=1 << 9, desc="WARN", when_present=True, module=False), + "C": TaintFlag(shift=1 << 10, desc="CRAP", when_present=True, module=True), + "I": TaintFlag( + shift=1 << 11, desc="FIRMWARE_WORKAROUND", when_present=True, module=False + ), + "O": TaintFlag(shift=1 << 12, desc="OOT_MODULE", when_present=True, module=True), + "E": TaintFlag( + shift=1 << 13, desc="UNSIGNED_MODULE", when_present=True, module=True + ), + "L": TaintFlag(shift=1 << 14, desc="SOFTLOCKUP", when_present=True, module=False), + "K": TaintFlag(shift=1 << 15, desc="LIVEPATCH", when_present=True, module=True), + "X": TaintFlag(shift=1 << 16, desc="AUX", when_present=True, module=True), + "T": TaintFlag(shift=1 << 17, desc="RANDSTRUCT", when_present=True, module=True), + "N": TaintFlag(shift=1 << 18, desc="TEST", when_present=True, module=True), +} +"""Flags used to taint kernel and modules, for debugging purposes. + +Map based on 6.12-rc5. + +Documentation : + - https://www.kernel.org/doc/Documentation/admin-guide/sysctl/kernel.rst#:~:text=guide/sysrq.rst.-,tainted,-%3D%3D%3D%3D%3D%3D%3D%0A%0ANon%2Dzero%20if + - https://www.kernel.org/doc/Documentation/admin-guide/tainted-kernels.rst#:~:text=More%20detailed%20explanation%20for%20tainting + - taint_flag kernel struct + - taint_flags kernel constant +""" diff --git a/volatility3/framework/contexts/__init__.py b/volatility3/framework/contexts/__init__.py index e1fb56d94..e7c423a10 100644 --- a/volatility3/framework/contexts/__init__.py +++ b/volatility3/framework/contexts/__init__.py @@ -338,7 +338,7 @@ class Module(interfaces.context.ModuleInterface): ) @property - def symbols(self): + def symbols(self) -> Iterable[str]: return self.context.symbol_space[self.symbol_table_name].symbols get_symbol = get_module_wrapper("get_symbol") diff --git a/volatility3/framework/exceptions.py b/volatility3/framework/exceptions.py index c44fb4f2e..41c67b88d 100644 --- a/volatility3/framework/exceptions.py +++ b/volatility3/framework/exceptions.py @@ -130,3 +130,7 @@ class OfflineException(VolatilityException): class RenderException(VolatilityException): """Thrown if there is an error during rendering""" + + +class LinuxPageCacheException(VolatilityException): + """Thrown if there is an error during Linux Page Cache processing""" diff --git a/volatility3/framework/interfaces/context.py b/volatility3/framework/interfaces/context.py index 48c066e96..723f2fd46 100644 --- a/volatility3/framework/interfaces/context.py +++ b/volatility3/framework/interfaces/context.py @@ -304,8 +304,8 @@ class ModuleInterface(interfaces.configuration.ConfigurableInterface): @property @abstractmethod - def symbols(self) -> List: - """Lists the symbols contained in the symbol table for this module""" + def symbols(self) -> Iterable[str]: + """Returns an iterable of the symbols contained in the symbol table for this module""" raise NotImplementedError("Symbols property has not been implemented.") @abstractmethod diff --git a/volatility3/framework/interfaces/plugins.py b/volatility3/framework/interfaces/plugins.py index f763815a6..7ad78d0ba 100644 --- a/volatility3/framework/interfaces/plugins.py +++ b/volatility3/framework/interfaces/plugins.py @@ -112,6 +112,8 @@ class PluginInterface( # Be careful with inheritance around this (We default to requiring a version which doesn't exist, so it must be set) _required_framework_version: Tuple[int, int, int] = (0, 0, 0) """The _version variable is a quick way for plugins to define their current interface, it should follow SemVer rules""" + additional_description: str = None + """Display additional description of the plugin after the description of the arguments. See: https://docs.python.org/3/library/argparse.html#epilog""" def __init__( self, diff --git a/volatility3/framework/interfaces/symbols.py b/volatility3/framework/interfaces/symbols.py index b8712e38d..2d142de9a 100644 --- a/volatility3/framework/interfaces/symbols.py +++ b/volatility3/framework/interfaces/symbols.py @@ -122,7 +122,7 @@ class BaseSymbolTableInterface: @property def symbols(self) -> Iterable[str]: - """Returns an iterator of the Symbol names.""" + """Returns an iterable of the available symbol names.""" raise NotImplementedError( "Abstract property symbols not implemented by subclass." ) @@ -131,7 +131,7 @@ class BaseSymbolTableInterface: @property def types(self) -> Iterable[str]: - """Returns an iterator of the Symbol type names.""" + """Returns an iterable of the available symbol type names.""" raise NotImplementedError( "Abstract property types not implemented by subclass." ) @@ -149,7 +149,7 @@ class BaseSymbolTableInterface: @property def enumerations(self) -> Iterable[Any]: - """Returns an iterator of the Enumeration names.""" + """Returns an iterable of the available enumerations.""" raise NotImplementedError( "Abstract property enumerations not implemented by subclass." ) @@ -366,6 +366,7 @@ class NativeTableInterface(BaseSymbolTableInterface): @property def symbols(self) -> Iterable[str]: + """Returns an iterable of the available symbol names.""" return [] def get_enumeration(self, name: str) -> objects.Template: @@ -374,7 +375,13 @@ class NativeTableInterface(BaseSymbolTableInterface): ) @property - def enumerations(self) -> Iterable[str]: + def enumerations(self) -> Iterable[Any]: + """Returns an iterable of the available enumerations.""" + return [] + + @property + def types(self) -> Iterable[str]: + """Returns an iterable of the available symbol type names.""" return [] diff --git a/volatility3/framework/layers/leechcore.py b/volatility3/framework/layers/leechcore.py index eeede1673..06c359203 100644 --- a/volatility3/framework/layers/leechcore.py +++ b/volatility3/framework/layers/leechcore.py @@ -129,6 +129,8 @@ if HAS_LEECHCORE: def readline(self, __size: Optional[int] = ...) -> bytes: data = b"" + if not __size: + __size = 0 while __size > self._chunk_size or __size < 0: data += self.read(self._chunk_size) index = data.find(b"\n") diff --git a/volatility3/framework/layers/registry.py b/volatility3/framework/layers/registry.py index c684ccd40..21e1a938e 100644 --- a/volatility3/framework/layers/registry.py +++ b/volatility3/framework/layers/registry.py @@ -192,9 +192,9 @@ class RegistryHive(linear.LinearlyMappedLayer): while key_array and node_key: subkeys = node_key[-1].get_subkeys() for subkey in subkeys: - # registry keys are not case sensitive so compare lowercase - # https://msdn.microsoft.com/en-us/library/windows/desktop/ms724946(v=vs.85).aspx - if subkey.get_name().lower() == key_array[0].lower(): + # registry keys are not case sensitive so compare likewise + # https://learn.microsoft.com/en-us/windows/win32/sysinfo/structure-of-the-registry + if subkey.get_name().casefold() == key_array[0].casefold(): node_key = node_key + [subkey] found_key, key_array = found_key + [key_array[0]], key_array[1:] break diff --git a/volatility3/framework/plugins/linux/bash.py b/volatility3/framework/plugins/linux/bash.py index 056e3cd51..8acfeb848 100644 --- a/volatility3/framework/plugins/linux/bash.py +++ b/volatility3/framework/plugins/linux/bash.py @@ -1,8 +1,8 @@ # This file is Copyright 2019 Volatility Foundation and licensed under the Volatility Software License 1.0 # which is available at https://www.volatilityfoundation.org/license/vsl-v1.0 # -"""A module containing a collection of plugins that produce data typically -found in Linux's /proc file system.""" +"""A module containing a plugin that recovers bash command history +from bash process memory.""" import datetime import struct diff --git a/volatility3/framework/plugins/linux/check_afinfo.py b/volatility3/framework/plugins/linux/check_afinfo.py index 201a443f7..7aa3cbdd2 100644 --- a/volatility3/framework/plugins/linux/check_afinfo.py +++ b/volatility3/framework/plugins/linux/check_afinfo.py @@ -1,8 +1,8 @@ # This file is Copyright 2019 Volatility Foundation and licensed under the Volatility Software License 1.0 # which is available at https://www.volatilityfoundation.org/license/vsl-v1.0 # -"""A module containing a collection of plugins that produce data typically -found in Linux's /proc file system.""" +"""A module containing a plugin that verifies the operation function +pointers of network protocols.""" import logging from typing import List diff --git a/volatility3/framework/plugins/linux/check_modules.py b/volatility3/framework/plugins/linux/check_modules.py index 9b3594c5e..0ed638d9c 100644 --- a/volatility3/framework/plugins/linux/check_modules.py +++ b/volatility3/framework/plugins/linux/check_modules.py @@ -18,6 +18,7 @@ vollog = logging.getLogger(__name__) class Check_modules(plugins.PluginInterface): """Compares module list to sysfs info, if available""" + _version = (1, 0, 0) _required_framework_version = (2, 0, 0) @classmethod diff --git a/volatility3/framework/plugins/linux/check_syscall.py b/volatility3/framework/plugins/linux/check_syscall.py index 3537a9fa1..13d312f2f 100644 --- a/volatility3/framework/plugins/linux/check_syscall.py +++ b/volatility3/framework/plugins/linux/check_syscall.py @@ -1,8 +1,7 @@ # This file is Copyright 2019 Volatility Foundation and licensed under the Volatility Software License 1.0 # which is available at https://www.volatilityfoundation.org/license/vsl-v1.0 # -"""A module containing a collection of plugins that produce data typically -found in Linux's /proc file system.""" +"""A module containing a plugin that checks the system call table for hooks.""" import contextlib import logging from typing import List diff --git a/volatility3/framework/plugins/linux/elfs.py b/volatility3/framework/plugins/linux/elfs.py index 2fd740941..0d1c9c2dd 100644 --- a/volatility3/framework/plugins/linux/elfs.py +++ b/volatility3/framework/plugins/linux/elfs.py @@ -1,8 +1,8 @@ # This file is Copyright 2019 Volatility Foundation and licensed under the Volatility Software License 1.0 # which is available at https://www.volatilityfoundation.org/license/vsl-v1.0 # -"""A module containing a collection of plugins that produce data typically -found in Linux's /proc file system.""" +"""A module containing a plugin for enumerating memory-mapped +ELF files across all processes.""" import logging from typing import List, Optional, Type diff --git a/volatility3/framework/plugins/linux/kmsg.py b/volatility3/framework/plugins/linux/kmsg.py index d66e3b9ca..638c2ccf2 100644 --- a/volatility3/framework/plugins/linux/kmsg.py +++ b/volatility3/framework/plugins/linux/kmsg.py @@ -149,7 +149,7 @@ class ABCKmsg(ABC): # This might seem insignificant but it could cause some issues # when compared with userland tool results or when used in # timelines. - return f"{nsec / 1000000000:lu}.{(nsec % 1000000000) / 1000:06lu}" + return f"{nsec // 1000000000}.{(nsec % 1000000000) // 1000:06}" def get_timestamp_in_sec_str(self, obj) -> str: # obj could be log, printk_log or printk_info @@ -166,7 +166,7 @@ class ABCKmsg(ABC): def get_caller_text(self, caller_id): caller_name = "CPU" if caller_id & 0x80000000 else "Task" - caller = f"{caller_name}({caller_id & ~0x80000000:u})" + caller = f"{caller_name}({caller_id & ~0x80000000})" return caller def get_prefix(self, obj) -> Tuple[int, int, str, str]: @@ -317,23 +317,27 @@ class Kmsg_3_5_to_3_11(ABCKmsg): while cur_idx < end_idx: msg_offset = log_buf_ptr + cur_idx # type: ignore msg = self.vmlinux.object(object_type=log_struct_name, offset=msg_offset) - if msg.len == 0: - # As per kernel/printk.c: - # A length == 0 for the next message indicates a wrap-around to - # the beginning of the buffer. - cur_idx = 0 - end_idx = log_next_idx - else: - facility, level, timestamp, caller = self.get_prefix(msg) - level_txt = self.get_level_text(level) - facility_txt = self.get_facility_text(facility) + try: + if msg.len == 0: + # As per kernel/printk.c: + # A length == 0 for the next message indicates a wrap-around to + # the beginning of the buffer. + cur_idx = 0 + end_idx = log_next_idx + else: + facility, level, timestamp, caller = self.get_prefix(msg) + level_txt = self.get_level_text(level) + facility_txt = self.get_facility_text(facility) - for line in self.get_log_lines(msg): - yield facility_txt, level_txt, timestamp, caller, line - for line in self.get_dict_lines(msg): - yield facility_txt, level_txt, timestamp, caller, line + for line in self.get_log_lines(msg): + yield facility_txt, level_txt, timestamp, caller, line + for line in self.get_dict_lines(msg): + yield facility_txt, level_txt, timestamp, caller, line - cur_idx += msg.len + cur_idx += msg.len + except exceptions.InvalidAddressException: + vollog.warning("Kmsg buffer msg length could not be read") + return class Kmsg_3_11_to_5_10(Kmsg_3_5_to_3_11): diff --git a/volatility3/framework/plugins/linux/lsmod.py b/volatility3/framework/plugins/linux/lsmod.py index 49e990e93..e9a2a7137 100644 --- a/volatility3/framework/plugins/linux/lsmod.py +++ b/volatility3/framework/plugins/linux/lsmod.py @@ -1,8 +1,7 @@ # This file is Copyright 2019 Volatility Foundation and licensed under the Volatility Software License 1.0 # which is available at https://www.volatilityfoundation.org/license/vsl-v1.0 # -"""A module containing a collection of plugins that produce data typically -found in Linux's /proc file system.""" +"""A module containing a plugin that lists loaded kernel modules.""" import logging from typing import List, Iterable diff --git a/volatility3/framework/plugins/linux/modxview.py b/volatility3/framework/plugins/linux/modxview.py new file mode 100644 index 000000000..c74bf28e8 --- /dev/null +++ b/volatility3/framework/plugins/linux/modxview.py @@ -0,0 +1,189 @@ +# This file is Copyright 2024 Volatility Foundation and licensed under the Volatility Software License 1.0 +# which is available at https://www.volatilityfoundation.org/license/vsl-v1.0 +# +import logging +from typing import List, Dict, Iterator +from volatility3.plugins.linux import lsmod, check_modules, hidden_modules +from volatility3.framework import interfaces +from volatility3.framework.configuration import requirements +from volatility3.framework.renderers import format_hints, TreeGrid, NotAvailableValue +from volatility3.framework.symbols.linux import extensions +from volatility3.framework.constants import architectures +from volatility3.framework.symbols.linux.utilities import tainting + +vollog = logging.getLogger(__name__) + + +class Modxview(interfaces.plugins.PluginInterface): + """Centralize lsmod, check_modules and hidden_modules results to efficiently + spot modules presence and taints.""" + + _version = (1, 0, 0) + _required_framework_version = (2, 17, 0) + + @classmethod + def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]: + return [ + requirements.ModuleRequirement( + name="kernel", + description="Linux kernel", + architectures=architectures.LINUX_ARCHS, + ), + requirements.VersionRequirement( + name="linux-tainting", component=tainting.Tainting, version=(1, 0, 0) + ), + requirements.PluginRequirement( + name="lsmod", plugin=lsmod.Lsmod, version=(2, 0, 0) + ), + requirements.PluginRequirement( + name="check_modules", + plugin=check_modules.Check_modules, + version=(1, 0, 0), + ), + requirements.PluginRequirement( + name="hidden_modules", + plugin=hidden_modules.Hidden_modules, + version=(1, 0, 0), + ), + requirements.BooleanRequirement( + name="plain_taints", + description="Display the plain taints string for each module.", + optional=True, + default=False, + ), + ] + + @classmethod + def flatten_run_modules_results( + cls, run_results: Dict[str, List[extensions.module]], deduplicate: bool = True + ) -> Iterator[extensions.module]: + """Flatten a dictionary mapping plugin names and modules list, to a single merged list. + This is useful to get a generic lookup list of all the detected modules. + + Args: + run_results: dictionary of plugin names mapping a list of detected modules + deduplicate: remove duplicate modules, based on their offsets + + Returns: + Iterator of modules objects + """ + seen_addresses = set() + for modules in run_results.values(): + for module in modules: + if deduplicate and module.vol.offset in seen_addresses: + continue + seen_addresses.add(module.vol.offset) + yield module + + @classmethod + def run_modules_scanners( + cls, + context: interfaces.context.ContextInterface, + kernel_name: str, + run_hidden_modules: bool = True, + ) -> Dict[str, List[extensions.module]]: + """Run module scanning plugins and aggregate the results. It is designed + to not operate any inter-plugin results triage. + + Args: + run_hidden_modules: specify if the hidden_modules plugin should be run + Returns: + Dictionary mapping each plugin to its corresponding result + """ + + kernel = context.modules[kernel_name] + run_results = {} + # lsmod + run_results["lsmod"] = list(lsmod.Lsmod.list_modules(context, kernel_name)) + # check_modules + sysfs_modules: dict = check_modules.Check_modules.get_kset_modules( + context, kernel_name + ) + ## Convert get_kset_modules() offsets back to module objects + run_results["check_modules"] = [ + kernel.object(object_type="module", offset=m_offset, absolute=True) + for m_offset in sysfs_modules.values() + ] + # hidden_modules + if run_hidden_modules: + known_modules_addresses = set( + context.layers[kernel.layer_name].canonicalize(module.vol.offset) + for module in run_results["lsmod"] + run_results["check_modules"] + ) + modules_memory_boundaries = ( + hidden_modules.Hidden_modules.get_modules_memory_boundaries( + context, kernel_name + ) + ) + run_results["hidden_modules"] = list( + hidden_modules.Hidden_modules.get_hidden_modules( + context, + kernel_name, + known_modules_addresses, + modules_memory_boundaries, + ) + ) + + return run_results + + def _generator(self): + kernel_name = self.config["kernel"] + run_results = self.run_modules_scanners(self.context, kernel_name) + aggregated_modules = {} + # We want to be explicit on the plugins results we are interested in + for plugin_name in ["lsmod", "check_modules", "hidden_modules"]: + # Iterate over each recovered module + for module in run_results[plugin_name]: + # Use offsets as unique keys, whether a module + # appears in many plugin runs or not + if aggregated_modules.get(module.vol.offset, None) is not None: + # Append the plugin to the list of originating plugins + aggregated_modules[module.vol.offset][1].append(plugin_name) + else: + aggregated_modules[module.vol.offset] = (module, [plugin_name]) + + for module_offset, (module, originating_plugins) in aggregated_modules.items(): + # Tainting parsing capabilities applied to the module + if self.config.get("plain_taints"): + taints = tainting.Tainting.get_taints_as_plain_string( + self.context, + kernel_name, + module.taints, + True, + ) + else: + taints = ",".join( + tainting.Tainting.get_taints_parsed( + self.context, + kernel_name, + module.taints, + True, + ) + ) + + yield ( + 0, + ( + module.get_name() or NotAvailableValue(), + format_hints.Hex(module_offset), + "lsmod" in originating_plugins, + "check_modules" in originating_plugins, + "hidden_modules" in originating_plugins, + taints or NotAvailableValue(), + ), + ) + + def run(self): + columns = [ + ("Name", str), + ("Address", format_hints.Hex), + ("In procfs", bool), + ("In sysfs", bool), + ("Hidden", bool), + ("Taints", str), + ] + + return TreeGrid( + columns, + self._generator(), + ) diff --git a/volatility3/framework/plugins/linux/pagecache.py b/volatility3/framework/plugins/linux/pagecache.py index 382268515..32b176b72 100644 --- a/volatility3/framework/plugins/linux/pagecache.py +++ b/volatility3/framework/plugins/linux/pagecache.py @@ -6,9 +6,9 @@ import math import logging import datetime from dataclasses import dataclass, astuple -from typing import List, Set, Type, Iterable +from typing import List, Set, Type, Iterable, Tuple -from volatility3.framework import renderers, interfaces +from volatility3.framework import renderers, interfaces, exceptions from volatility3.framework.renderers import format_hints from volatility3.framework.interfaces import plugins from volatility3.framework.configuration import requirements @@ -104,7 +104,7 @@ class Files(plugins.PluginInterface, timeliner.TimeLinerInterface): _required_framework_version = (2, 0, 0) - _version = (1, 0, 1) + _version = (1, 0, 2) @classmethod def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]: @@ -147,7 +147,13 @@ class Files(plugins.PluginInterface, timeliner.TimeLinerInterface): Otherwise, it returns the same symlink_path """ # i_link (fast symlinks) were introduced in 4.2 - if inode and inode.is_link and inode.has_member("i_link") and inode.i_link: + if ( + inode + and inode.is_link + and inode.has_member("i_link") + and inode.i_link + and inode.i_link.is_readable() + ): i_link_str = inode.i_link.dereference().cast( "string", max_length=255, encoding="utf-8", errors="replace" ) @@ -253,6 +259,10 @@ class Files(plugins.PluginInterface, timeliner.TimeLinerInterface): if not root_inode.is_valid(): continue + if not (root_inode.i_mapping and root_inode.i_mapping.is_readable()): + # Retrieving data from the page cache requires a valid address space + continue + # Inode already processed? if root_inode_ptr in seen_inodes: continue @@ -284,6 +294,10 @@ class Files(plugins.PluginInterface, timeliner.TimeLinerInterface): if not file_inode.is_valid(): continue + if not (file_inode.i_mapping and file_inode.i_mapping.is_readable()): + # Retrieving data from the page cache requires a valid address space + continue + # Inode already processed? if file_inode_ptr in seen_inodes: continue @@ -316,10 +330,12 @@ class Files(plugins.PluginInterface, timeliner.TimeLinerInterface): if self.config["find"]: if inode_in.path == self.config["find"]: inode_out = inode_in.to_user(vmlinux_layer) + yield (0, astuple(inode_out)) break # Only the first match else: inode_out = inode_in.to_user(vmlinux_layer) + yield (0, astuple(inode_out)) def generate_timeline(self): @@ -389,7 +405,7 @@ class InodePages(plugins.PluginInterface): _required_framework_version = (2, 0, 0) - _version = (2, 0, 0) + _version = (2, 0, 1) @classmethod def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]: @@ -443,28 +459,80 @@ class InodePages(plugins.PluginInterface): # created, saving both disk space and I/O time. # Additionally, using the page index will guarantee that each page is written at the # appropriate file position. + inode_size = inode.i_size try: - with open_method(filename) as f: - inode_size = inode.i_size - f.truncate(inode_size) - + file_initialized = False + with open_method(filename) as file_obj: for page_idx, page_content in inode.get_contents(): current_fp = page_idx * vmlinux_layer.page_size max_length = inode_size - current_fp - page_bytes = page_content[:max_length] - if current_fp + len(page_bytes) > inode_size: + page_bytes_len = min(max_length, len(page_content)) + if ( + current_fp >= inode_size + or current_fp + page_bytes_len > inode_size + ): vollog.error( "Page out of file bounds: inode 0x%x, inode size %d, page index %d", inode.vol.offset, inode_size, page_idx, ) - f.seek(current_fp) - f.write(page_bytes) + continue + page_bytes = page_content[:page_bytes_len] + if not file_initialized: + # Lazy initialization to avoid truncating the file until we are + # certain there is something to write + file_obj.truncate(inode_size) + file_initialized = True + + file_obj.seek(current_fp) + file_obj.write(page_bytes) + except exceptions.LinuxPageCacheException: + vollog.error( + f"Error dumping cached pages for inode at {inode.vol.offset:#x}" + ) except OSError as e: vollog.error("Unable to write to file (%s): %s", filename, e) + def _generate_inode_fields( + self, + inode: interfaces.objects.ObjectInterface, + vmlinux_layer: interfaces.layers.TranslationLayerInterface, + ) -> Iterable[Tuple[int, int, int, int, bool, str]]: + inode_size = inode.i_size + try: + for page_obj in inode.get_pages(): + if page_obj.mapping != inode.i_mapping: + vollog.warning( + f"Cached page at {page_obj.vol.offset:#x} has a mismatched address space with the inode. Skipping page" + ) + continue + page_vaddr = page_obj.vol.offset + page_paddr = page_obj.to_paddr() + page_mapping_addr = page_obj.mapping + page_index = page_obj.index + page_file_offset = page_index * vmlinux_layer.page_size + dump_safe = ( + page_file_offset < inode_size + and page_mapping_addr + and page_mapping_addr.is_readable() + ) + page_flags_list = page_obj.get_flags_list() + page_flags = ",".join([x.replace("PG_", "") for x in page_flags_list]) + fields = ( + page_vaddr, + page_paddr, + page_mapping_addr, + page_index, + dump_safe, + page_flags, + ) + + yield 0, fields + except exceptions.LinuxPageCacheException: + vollog.warning(f"Page cache for inode at {inode.vol.offset:#x} is corrupt") + def _generator(self): vmlinux_module_name = self.config["kernel"] vmlinux = self.context.modules[vmlinux_module_name] @@ -486,7 +554,6 @@ class InodePages(plugins.PluginInterface): else: vollog.error("Unable to find inode with path %s", self.config["find"]) return None - elif self.config["inode"]: inode = vmlinux.object("inode", self.config["inode"], absolute=True) else: @@ -501,27 +568,6 @@ class InodePages(plugins.PluginInterface): vollog.error("The inode is not a regular file") return None - inode_size = inode.i_size - for page_obj in inode.get_pages(): - page_vaddr = page_obj.vol.offset - page_paddr = page_obj.to_paddr() - page_mapping_addr = page_obj.mapping - page_index = int(page_obj.index) - page_file_offset = page_index * vmlinux_layer.page_size - dump_safe = page_file_offset < inode_size - page_flags_list = page_obj.get_flags_list() - page_flags = ",".join([x.replace("PG_", "") for x in page_flags_list]) - fields = ( - page_vaddr, - page_paddr, - page_mapping_addr, - page_index, - dump_safe, - page_flags, - ) - - yield 0, fields - if self.config["dump"]: open_method = self.open inode_address = inode.vol.offset @@ -530,6 +576,8 @@ class InodePages(plugins.PluginInterface): self.write_inode_content_to_file( inode, filename, open_method, vmlinux_layer ) + else: + yield from self._generate_inode_fields(inode, vmlinux_layer) def run(self): headers = [ diff --git a/volatility3/framework/plugins/linux/pslist.py b/volatility3/framework/plugins/linux/pslist.py index 37cf000fc..b8fac4a8b 100644 --- a/volatility3/framework/plugins/linux/pslist.py +++ b/volatility3/framework/plugins/linux/pslist.py @@ -34,7 +34,7 @@ class PsList(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterface): """Lists the processes present in a particular linux memory image.""" _required_framework_version = (2, 13, 0) - _version = (4, 0, 0) + _version = (4, 1, 0) @classmethod def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]: @@ -179,6 +179,10 @@ class PsList(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterface): file_output = "VMA start matching task start_code not found" return file_output + @staticmethod + def _format_cred(cred): + return renderers.NotAvailableValue() if cred is None else cred + def _generator( self, pid_filter: Callable[[Any], bool], @@ -212,16 +216,21 @@ class PsList(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterface): task_fields = self.get_task_fields(task, decorate_comm) + task_uid = self._format_cred(task_fields.uid) + task_gid = self._format_cred(task_fields.gid) + task_euid = self._format_cred(task_fields.euid) + task_egid = self._format_cred(task_fields.egid) + yield 0, ( format_hints.Hex(task_fields.offset), task_fields.user_pid, task_fields.user_tid, task_fields.user_ppid, task_fields.name, - task_fields.uid or renderers.NotAvailableValue(), - task_fields.gid or renderers.NotAvailableValue(), - task_fields.euid or renderers.NotAvailableValue(), - task_fields.egid or renderers.NotAvailableValue(), + task_uid, + task_gid, + task_euid, + task_egid, task_fields.creation_time or renderers.NotAvailableValue(), file_output, ) @@ -250,6 +259,9 @@ class PsList(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterface): # Note that the init_task itself is not yielded, since "ps" also never shows it. for task in init_task.tasks: + if not task.is_valid(): + continue + if filter_func(task): continue diff --git a/volatility3/framework/plugins/windows/amcache.py b/volatility3/framework/plugins/windows/amcache.py index 1e918d61c..46a742233 100644 --- a/volatility3/framework/plugins/windows/amcache.py +++ b/volatility3/framework/plugins/windows/amcache.py @@ -543,7 +543,7 @@ class Amcache(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterface): amcache.get_key("Root\\InventoryDriverBinary") # type: ignore ) ) - except KeyError: + except (KeyError, registry.RegistryFormatException): # Registry key not found pass @@ -554,7 +554,7 @@ class Amcache(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterface): amcache.get_key("Root\\Programs") ) # type: ignore } - except KeyError: + except (KeyError, registry.RegistryFormatException): programs = {} try: @@ -564,7 +564,7 @@ class Amcache(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterface): ), key=_entry_sort_key, ) - except KeyError: + except (KeyError, registry.RegistryFormatException): files = [] for program_id, file_entries in itertools.groupby( @@ -593,7 +593,7 @@ class Amcache(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterface): amcache.get_key("Root\\InventoryApplication") # type: ignore ) ) - except KeyError: + except (KeyError, registry.RegistryFormatException): programs = {} try: @@ -603,7 +603,7 @@ class Amcache(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterface): ), key=_entry_sort_key, ) - except KeyError: + except (KeyError, registry.RegistryFormatException): files = [] for program_id, file_entries in itertools.groupby( diff --git a/volatility3/framework/plugins/windows/cachedump.py b/volatility3/framework/plugins/windows/cachedump.py index 6e667984a..6c730e6ae 100644 --- a/volatility3/framework/plugins/windows/cachedump.py +++ b/volatility3/framework/plugins/windows/cachedump.py @@ -8,7 +8,7 @@ from typing import Tuple from Crypto.Cipher import ARC4, AES from Crypto.Hash import HMAC -from volatility3.framework import interfaces, renderers +from volatility3.framework import interfaces, renderers, exceptions from volatility3.framework.configuration import requirements from volatility3.framework.layers import registry from volatility3.framework.symbols.windows import versions @@ -140,9 +140,14 @@ class Cachedump(interfaces.plugins.PluginInterface): if cache_item.Name == "NL$Control": continue - data = sechive.read(cache_item.Data + 4, cache_item.DataLength) - if data is None: + try: + data = sechive.read(cache_item.Data + 4, cache_item.DataLength) + except exceptions.InvalidAddressException: continue + + if not data: + continue + ( uname_len, domain_len, diff --git a/volatility3/framework/plugins/windows/cmdscan.py b/volatility3/framework/plugins/windows/cmdscan.py index 9645ee507..0cd0addb2 100644 --- a/volatility3/framework/plugins/windows/cmdscan.py +++ b/volatility3/framework/plugins/windows/cmdscan.py @@ -67,6 +67,7 @@ class CmdScan(interfaces.plugins.PluginInterface): Args: conhost_proc: the process object for conhost.exe + size_filter: size above which vads will not be returned Returns: A list of tuples of: @@ -99,8 +100,8 @@ class CmdScan(interfaces.plugins.PluginInterface): kernel_layer_name: The name of the layer on which to operate kernel_symbol_table_name: The name of the table containing the kernel symbols config_path: The config path where to find symbol files - procs: list of process objects - max_history: an initial set of CommandHistorySize values + procs: List of process objects + max_history: An initial set of CommandHistorySize values Returns: The conhost process object, the command history structure, a dictionary of properties for @@ -227,7 +228,6 @@ class CmdScan(interfaces.plugins.PluginInterface): "data": command_history.CommandCountMax, } ) - command_history_properties.append( { "level": 1, @@ -236,6 +236,7 @@ class CmdScan(interfaces.plugins.PluginInterface): "data": "", } ) + for ( cmd_index, bucket_cmd, @@ -352,7 +353,7 @@ class CmdScan(interfaces.plugins.PluginInterface): def _conhost_proc_filter(self, proc: interfaces.objects.ObjectInterface): """ - Used to filter to only conhost.exe processes + Used to filter only conhost.exe processes """ process_name = utility.array_to_string(proc.ImageFileName) diff --git a/volatility3/framework/plugins/windows/driverscan.py b/volatility3/framework/plugins/windows/driverscan.py index 24d81c3d5..d388ffbb7 100644 --- a/volatility3/framework/plugins/windows/driverscan.py +++ b/volatility3/framework/plugins/windows/driverscan.py @@ -64,7 +64,7 @@ class DriverScan(interfaces.plugins.PluginInterface): names associated with a driver Args: - driver: A Eriver object + driver: A Driver object Returns: A tuple of strings of (driver name, service key, driver alt. name) diff --git a/volatility3/framework/plugins/windows/envars.py b/volatility3/framework/plugins/windows/envars.py index cac4ecf40..48e1ef671 100644 --- a/volatility3/framework/plugins/windows/envars.py +++ b/volatility3/framework/plugins/windows/envars.py @@ -76,14 +76,14 @@ class Envars(interfaces.plugins.PluginInterface): "CurrentControlSet\\Control\\Session Manager\\Environment" ) sys = True - except KeyError: - with contextlib.suppress(KeyError): + except (KeyError, registry.RegistryFormatException): + with contextlib.suppress(KeyError, registry.RegistryFormatException): key = hive.get_key( "ControlSet001\\Control\\Session Manager\\Environment" ) sys = True if sys: - with contextlib.suppress(KeyError): + with contextlib.suppress(KeyError, registry.RegistryFormatException): for node in key.get_values(): try: value_node_name = node.get_name() @@ -100,11 +100,11 @@ class Envars(interfaces.plugins.PluginInterface): continue ## The user-specific variables - with contextlib.suppress(KeyError): + with contextlib.suppress(KeyError, registry.RegistryFormatException): key = hive.get_key("Environment") ntuser = True if ntuser: - with contextlib.suppress(KeyError): + with contextlib.suppress(KeyError, registry.RegistryFormatException): for node in key.get_values(): try: value_node_name = node.get_name() @@ -123,7 +123,7 @@ class Envars(interfaces.plugins.PluginInterface): ## The volatile user variables try: key = hive.get_key("Volatile Environment") - except KeyError: + except (KeyError, registry.RegistryFormatException): continue try: for node in key.get_values(): diff --git a/volatility3/framework/plugins/windows/getservicesids.py b/volatility3/framework/plugins/windows/getservicesids.py index eece7fb6c..207d0e2ad 100644 --- a/volatility3/framework/plugins/windows/getservicesids.py +++ b/volatility3/framework/plugins/windows/getservicesids.py @@ -10,6 +10,7 @@ from typing import List from volatility3.framework import renderers, interfaces, constants, exceptions from volatility3.framework.configuration import requirements +from volatility3.framework.layers import registry from volatility3.plugins.windows.registry import hivelist vollog = logging.getLogger(__name__) @@ -86,10 +87,18 @@ class GetServiceSIDs(interfaces.plugins.PluginInterface): # Get ControlSet\Services. try: services = hive.get_key(r"CurrentControlSet\Services") - except (KeyError, exceptions.InvalidAddressException): + except ( + KeyError, + exceptions.InvalidAddressException, + registry.RegistryFormatException, + ): try: services = hive.get_key(r"ControlSet001\Services") - except (KeyError, exceptions.InvalidAddressException): + except ( + KeyError, + exceptions.InvalidAddressException, + registry.RegistryFormatException, + ): continue if services: diff --git a/volatility3/framework/plugins/windows/getsids.py b/volatility3/framework/plugins/windows/getsids.py index df0c7a835..a75bbe7ea 100644 --- a/volatility3/framework/plugins/windows/getsids.py +++ b/volatility3/framework/plugins/windows/getsids.py @@ -158,7 +158,11 @@ class GetSIDs(interfaces.plugins.PluginInterface): layers.registry.RegistryFormatException, ): continue - except (KeyError, exceptions.InvalidAddressException): + except ( + KeyError, + exceptions.InvalidAddressException, + layers.registry.RegistryFormatException, + ): continue return sids diff --git a/volatility3/framework/plugins/windows/handles.py b/volatility3/framework/plugins/windows/handles.py index 38ccfbfbc..6a391fe35 100644 --- a/volatility3/framework/plugins/windows/handles.py +++ b/volatility3/framework/plugins/windows/handles.py @@ -341,7 +341,7 @@ class Handles(interfaces.plugins.PluginInterface): try: obj_name = entry.NameInfo.Name.String except (ValueError, exceptions.InvalidAddressException): - obj_name = "" + obj_name = None except exceptions.InvalidAddressException: vollog.log( @@ -359,7 +359,7 @@ class Handles(interfaces.plugins.PluginInterface): format_hints.Hex(entry.HandleValue), obj_type, format_hints.Hex(entry.GrantedAccess), - obj_name, + obj_name or renderers.NotAvailableValue(), ), ) diff --git a/volatility3/framework/plugins/windows/hashdump.py b/volatility3/framework/plugins/windows/hashdump.py index 0c98ab8ca..621b0ae53 100644 --- a/volatility3/framework/plugins/windows/hashdump.py +++ b/volatility3/framework/plugins/windows/hashdump.py @@ -332,7 +332,7 @@ class Hashdump(interfaces.plugins.PluginInterface): try: if hive: result = hive.get_key(key) - except KeyError: + except (KeyError, registry.RegistryFormatException): vollog.info( f"Unable to load the required registry key {hive.get_name()}\\{key} from this memory image" ) diff --git a/volatility3/framework/plugins/windows/lsadump.py b/volatility3/framework/plugins/windows/lsadump.py index da8dee325..f3925f2a2 100644 --- a/volatility3/framework/plugins/windows/lsadump.py +++ b/volatility3/framework/plugins/windows/lsadump.py @@ -8,7 +8,7 @@ from typing import Optional from Crypto.Cipher import ARC4, DES, AES from Crypto.Hash import MD5, SHA256 -from volatility3.framework import interfaces, renderers +from volatility3.framework import interfaces, renderers, exceptions from volatility3.framework.configuration import requirements from volatility3.framework.layers import registry from volatility3.framework.symbols.windows import versions @@ -81,7 +81,10 @@ class Lsadump(interfaces.plugins.PluginInterface): if not enc_reg_value: return None - obf_lsa_key = sechive.read(enc_reg_value.Data + 4, enc_reg_value.DataLength) + try: + obf_lsa_key = sechive.read(enc_reg_value.Data + 4, enc_reg_value.DataLength) + except exceptions.InvalidAddressException: + return None if not obf_lsa_key: return None diff --git a/volatility3/framework/plugins/windows/registry/userassist.py b/volatility3/framework/plugins/windows/registry/userassist.py index 932ee9d6f..87016553a 100644 --- a/volatility3/framework/plugins/windows/registry/userassist.py +++ b/volatility3/framework/plugins/windows/registry/userassist.py @@ -13,7 +13,7 @@ from typing import Any, Generator, List, Tuple from volatility3.framework import constants, exceptions, interfaces, renderers from volatility3.framework.configuration import requirements from volatility3.framework.layers.physical import BufferDataLayer -from volatility3.framework.layers.registry import RegistryHive +from volatility3.framework.layers.registry import RegistryHive, RegistryFormatException from volatility3.framework.renderers import conversion, format_hints from volatility3.framework.symbols import intermed from volatility3.plugins.windows.registry import hivelist @@ -167,10 +167,21 @@ class UserAssist(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterfac self._determine_userassist_type() - userassist_node_path = hive.get_key( - "software\\microsoft\\windows\\currentversion\\explorer\\userassist", - return_list=True, - ) + try: + userassist_node_path = hive.get_key( + "software\\microsoft\\windows\\currentversion\\explorer\\userassist", + return_list=True, + ) + except RegistryFormatException as e: + vollog.warning( + f"Error accessing UserAssist key in {hive_name} at {hive.hive_offset:#x}: {e}" + ) + return None + except KeyError: + vollog.warning( + f"UserAssist key not found in {hive_name} at {hive.hive_offset:#x}" + ) + return None if not userassist_node_path: vollog.warning("list_userassist did not find a valid node_path (or None)") diff --git a/volatility3/framework/plugins/windows/shimcachemem.py b/volatility3/framework/plugins/windows/shimcachemem.py index b8e9b5bd7..9d968c30a 100644 --- a/volatility3/framework/plugins/windows/shimcachemem.py +++ b/volatility3/framework/plugins/windows/shimcachemem.py @@ -305,14 +305,14 @@ class ShimcacheMem(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterf If a number of validity checks are passed, this method will return the `SHIM_CACHE_HEAD` object. Otherwise, `None` is returned. """ - # print("checking RTL_AVL_TABLE at offset %s" % hex(offset)) + # Check RTL_AVL_TABLE at offset rtl_avl_table = context.object( symbol_table + constants.BANG + "_RTL_AVL_TABLE", layer_name, offset ) if not rtl_avl_table.is_valid(mod_page_start, mod_page_end): return None - vollog.debug(f"Candidate RTL_AVL_TABLE found at offset {hex(offset)}") + vollog.debug(f"Candidate RTL_AVL_TABLE found at offset {offset:#x}") ersrc_size = context.symbol_space.get_type( kernel_symbol_table + constants.BANG + "_ERESOURCE" @@ -324,13 +324,13 @@ class ShimcacheMem(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterf # 0x20 if context.symbol_space.get_type("pointer").size == 8 else 0x10 ) vollog.debug( - f"ERESOURCE size: {hex(ersrc_size)}, ERESOURCE alignment: {hex(ersrc_alignment)}" + f"ERESOURCE size: {ersrc_size:#x}, ERESOURCE alignment: {ersrc_alignment:#x}" ) eresource_rel_off = ersrc_size + ((offset - ersrc_size) % ersrc_alignment) eresource_offset = offset - eresource_rel_off - vollog.debug(f"Constructing ERESOURCE at {hex(eresource_offset)}") + vollog.debug(f"Constructing ERESOURCE at {eresource_offset:#x}") eresource = context.object( kernel_symbol_table + constants.BANG + "_ERESOURCE", layer_name, @@ -408,8 +408,8 @@ class ShimcacheMem(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterf # iterate over ahcache kernel module's .data section in search of *two* SHIM handles shim_heads = [] - vollog.debug(f"PAGE offset: {hex(mod_page_offset)}") - vollog.debug(f".data offset: {hex(data_sec_offset)}") + vollog.debug(f"PAGE offset: {mod_page_offset:#x}") + vollog.debug(f".data offset: {data_sec_offset:#x}") handle_type = context.symbol_space.get_type( shimcache_symbol_table + constants.BANG + "SHIM_CACHE_HANDLE" @@ -419,7 +419,7 @@ class ShimcacheMem(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterf data_sec_offset + data_sec_size, 8 if symbols.symbol_table_is_64bit(context, nt_symbol_table) else 4, ): - vollog.debug(f"Building shim handle pointer at {hex(offset)}") + vollog.debug(f"Building shim handle pointer at {offset:#x}") shim_handle = context.object( object_type=shimcache_symbol_table + constants.BANG + "pointer", layer_name=kernel_layer_name, @@ -430,7 +430,7 @@ class ShimcacheMem(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterf if shim_handle.is_valid(mod_page_offset, mod_page_offset + mod_page_size): if shim_handle.head is not None: vollog.debug( - f"Found valid shim handle @ {hex(shim_handle.vol.offset)}" + f"Found valid shim handle @ {shim_handle.vol.offset:#x}" ) shim_heads.append(shim_handle.head) if len(shim_heads) == 2: @@ -440,7 +440,7 @@ class ShimcacheMem(interfaces.plugins.PluginInterface, timeliner.TimeLinerInterf vollog.debug("Failed to identify two valid SHIM_CACHE_HANDLE structures") return - # On Windows 8 x64, the frist cache contains the shim cache + # On Windows 8 x64, the first cache contains the shim cache. # On Windows 8 x86, 8.1 x86/x64, and 10, the second cache contains the shim cache. if ( not symbols.symbol_table_is_64bit(context, nt_symbol_table) diff --git a/volatility3/framework/plugins/windows/svcscan.py b/volatility3/framework/plugins/windows/svcscan.py index bd477ba27..17baac5b0 100644 --- a/volatility3/framework/plugins/windows/svcscan.py +++ b/volatility3/framework/plugins/windows/svcscan.py @@ -15,7 +15,7 @@ from volatility3.framework import ( symbols, ) from volatility3.framework.configuration import requirements -from volatility3.framework.layers import scanners +from volatility3.framework.layers import scanners, registry from volatility3.framework.renderers import format_hints from volatility3.framework.symbols import intermed from volatility3.framework.symbols.windows import versions @@ -159,12 +159,20 @@ class SvcScan(interfaces.plugins.PluginInterface): return cast( objects.StructType, hive.get_key(r"CurrentControlSet\Services") ) - except (KeyError, exceptions.InvalidAddressException): + except ( + KeyError, + exceptions.InvalidAddressException, + registry.RegistryFormatException, + ): try: return cast( objects.StructType, hive.get_key(r"ControlSet001\Services") ) - except (KeyError, exceptions.InvalidAddressException): + except ( + KeyError, + exceptions.InvalidAddressException, + registry.RegistryFormatException, + ): vollog.log( constants.LOGLEVEL_VVVV, "Could not retrieve any control set from SYSTEM hive", diff --git a/volatility3/framework/renderers/conversion.py b/volatility3/framework/renderers/conversion.py index e48684b31..f848b2dad 100644 --- a/volatility3/framework/renderers/conversion.py +++ b/volatility3/framework/renderers/conversion.py @@ -18,7 +18,7 @@ def wintime_to_datetime( unix_time = wintime // 10000000 if unix_time == 0: return renderers.NotApplicableValue() - unix_time = unix_time - 11644473600 + unix_time -= 11644473600 try: return datetime.datetime.fromtimestamp(unix_time, datetime.timezone.utc) # Windows sometimes throws OSErrors rather than ValueError/OverflowError when it can't convert a value @@ -71,7 +71,7 @@ def round(addr: int, align: int, up: bool = False) -> int: Args: addr: the address align: the alignment value - up: Whether to round up or not + up: whether to round up or not Returns: The aligned address @@ -122,11 +122,12 @@ def convert_port(port_as_integer): def convert_network_four_tuple(family, four_tuple): - """Converts the connection four_tuple: (source ip, source port, dest ip, - dest port) + """Converts the connection four_tuple: + + (source ip, source port, dest ip, dest port) into their string equivalents. IP addresses are expected as a tuple - of unsigned shorts Ports are converted to proper endianness as well + of unsigned shorts. Ports are converted to proper endianness as well. """ if family == socket.AF_INET: diff --git a/volatility3/framework/symbols/intermed.py b/volatility3/framework/symbols/intermed.py index 6802af7d6..cb0b67969 100644 --- a/volatility3/framework/symbols/intermed.py +++ b/volatility3/framework/symbols/intermed.py @@ -411,18 +411,27 @@ class Version1Format(ISFormatTable): @property def symbols(self) -> Iterable[str]: - """Returns an iterator of the symbol names.""" - return list(self._json_object.get("symbols", {})) + """Returns an iterable (KeysView) of the available symbol names.""" + return self._json_object.get("symbols", {}).keys() @property - def enumerations(self) -> Iterable[str]: - """Returns an iterator of the available enumerations.""" - return list(self._json_object.get("enums", {})) + def enumerations(self) -> Iterable[Any]: + """Returns an iterable (KeysView) of the available enumerations.""" + return self._json_object.get("enums", {}).keys() @property def types(self) -> Iterable[str]: - """Returns an iterator of the symbol type names.""" - return list(self._json_object.get("user_types", {})) + list(self.natives.types) + """Returns an iterable (KeysView) of the available symbol type names.""" + # We use ** instead of + # `set(self._json_object.get("user_types", {}).keys()).union(self.natives.types)` + # because converting user_types dict to a set is costly. + # It is more efficient to convert the (very small) self.natives.types set to a dict. + # FIXME: On Python3.8 support drop, merge the two dicts using the merge operator: + # (self._json_object.get("user_types", {}) | dict.fromkeys(self.natives.types)).keys() + return { + **self._json_object.get("user_types", {}), + **dict.fromkeys(self.natives.types), + }.keys() def get_type_class(self, name: str) -> Type[interfaces.objects.ObjectInterface]: return self._overrides.get(name, objects.AggregateType) diff --git a/volatility3/framework/symbols/linux/__init__.py b/volatility3/framework/symbols/linux/__init__.py index 5aa27b964..423284b03 100644 --- a/volatility3/framework/symbols/linux/__init__.py +++ b/volatility3/framework/symbols/linux/__init__.py @@ -3,6 +3,8 @@ # import math import contextlib +import functools +import logging from abc import ABC, abstractmethod from typing import Iterator, List, Tuple, Optional, Union @@ -12,6 +14,8 @@ from volatility3.framework.objects import utility from volatility3.framework.symbols import intermed from volatility3.framework.symbols.linux import extensions +vollog = logging.getLogger(__name__) + class LinuxKernelIntermedSymbols(intermed.IntermediateSymbolTable): provides = {"type": "interface"} @@ -43,6 +47,7 @@ class LinuxKernelIntermedSymbols(intermed.IntermediateSymbolTable): self.optional_set_type_class("bpf_prog_aux", extensions.bpf_prog_aux) self.optional_set_type_class("kernel_cap_struct", extensions.kernel_cap_struct) self.optional_set_type_class("kernel_cap_t", extensions.kernel_cap_t) + self.optional_set_type_class("scatterlist", extensions.scatterlist) # kernels >= 4.18 self.optional_set_type_class("timespec64", extensions.timespec64) @@ -612,7 +617,7 @@ class IDStorage(ABC): raise NotImplementedError def nodep_to_node(self, nodep) -> interfaces.objects.ObjectInterface: - """Instanciates a tree node from its pointer + """Instantiates a tree node from its pointer Args: nodep: Pointer to the XArray/RadixTree node @@ -659,7 +664,7 @@ class IDStorage(ABC): height = self.get_tree_height(root.vol.offset) nodep = self.get_head_node(root) - if not nodep: + if not (nodep and nodep.is_readable()): return # Keep the internal flag before untagging it @@ -694,7 +699,7 @@ class XArray(IDStorage): def get_node_height(self, nodep) -> int: node = self.nodep_to_node(nodep) - return (node.shift / self.CHUNK_SHIFT) + 1 + return (node.shift // self.CHUNK_SHIFT) + 1 def get_head_node(self, tree) -> int: return tree.xa_head @@ -717,6 +722,7 @@ class RadixTree(IDStorage): RADIX_TREE_INTERNAL_NODE = 1 RADIX_TREE_EXCEPTIONAL_ENTRY = 2 RADIX_TREE_ENTRY_MASK = 3 + RADIX_TREE_MAP_SHIFT = 6 # CONFIG_BASE_FULL # Dynamic values. These will be initialized later RADIX_TREE_INDEX_BITS = None @@ -753,43 +759,57 @@ class RadixTree(IDStorage): def get_tree_height(self, treep) -> int: with contextlib.suppress(exceptions.SymbolError): if self.vmlinux.get_type("radix_tree_root").has_member("height"): - # kernels < 4.7.10 + # kernels < 4.7 d0891265bbc988dc91ed8580b38eb3dac128581b radix_tree_root = self.vmlinux.object( "radix_tree_root", offset=treep, absolute=True ) return radix_tree_root.height - # kernels >= 4.7.10 + # kernels >= 4.7 return 0 + @functools.cached_property + def _max_height_array(self): + if self.vmlinux.has_symbol("height_to_maxindex"): + # 2.6.24 26fb1589cb0aaec3a0b4418c54f30c1a2b1781f6 <= Kernels < 4.7 d0891265bbc988dc91ed8580b38eb3dac128581b + return self.vmlinux.object_from_symbol("height_to_maxindex") + elif self.vmlinux.has_symbol("height_to_maxnodes"): + # 4.8 c78c66d1ddfdbd2353f3fcfeba0268524537b096 <= kernels < 4.20 8cf2f98411e3a0865026a1061af637161b16d32b + return self.vmlinux.object_from_symbol("height_to_maxnodes") + + return None + def _radix_tree_maxindex(self, node, height) -> int: """Return the maximum key which can be store into a radix tree with this height.""" - if not self.vmlinux.has_symbol("height_to_maxindex"): - # Kernels >= 4.7 - return (self.CHUNK_SIZE << node.shift) - 1 + if self._max_height_array: + # 2.6.24 <= kernels <= 4.20 See _max_height_array() + return self._max_height_array[height] else: - # Kernels < 4.7 - height_to_maxindex_array = self.vmlinux.object_from_symbol( - "height_to_maxindex" - ) - maxindex = height_to_maxindex_array[height] - return maxindex + # Kernels >= 4.20 + return (self.CHUNK_SIZE << node.shift) - 1 def get_node_height(self, nodep) -> int: node = self.nodep_to_node(nodep) if hasattr(node, "shift"): # 4.7 <= Kernels < 4.20 - return (node.shift / self.CHUNK_SHIFT) + 1 + height = (node.shift // self.CHUNK_SHIFT) + 1 elif hasattr(node, "path"): # 3.15 <= Kernels < 4.7 - return node.path & self.RADIX_TREE_HEIGHT_MASK + height = node.path & self.RADIX_TREE_HEIGHT_MASK elif hasattr(node, "height"): # Kernels < 3.15 - return node.height + height = node.height else: raise exceptions.VolatilityException("Cannot find radix-tree node height") + if self._max_height_array and not (0 <= height < self._max_height_array.count): + error_msg = f"Radix Tree node {node.vol.offset:#x} height {height} exceeds max height of {self._max_height_array.count}" + vollog.error(error_msg) + raise exceptions.LinuxPageCacheException(error_msg) + + return height + def get_head_node(self, tree) -> int: return tree.rnode @@ -802,14 +822,16 @@ class RadixTree(IDStorage): def untag_node(self, nodep) -> int: return nodep & (~self.RADIX_TREE_ENTRY_MASK) - def is_valid_node(self, nodep) -> bool: + def _is_exceptional_node(self, nodep) -> bool: # In kernels 4.20, exceptional nodes were removed and internal entries took their bitmask - if self.vmlinux.has_type("radix_tree_root"): - return ( - nodep & self.RADIX_TREE_ENTRY_MASK - ) != self.RADIX_TREE_EXCEPTIONAL_ENTRY + return ( + self.vmlinux.has_type("radix_tree_root") + and (nodep & self.RADIX_TREE_ENTRY_MASK) + == self.RADIX_TREE_EXCEPTIONAL_ENTRY + ) - return True + def is_valid_node(self, nodep) -> bool: + return not self._is_exceptional_node(nodep) class PageCache: @@ -838,11 +860,17 @@ class PageCache: Yields: Page objects """ - + layer = self.vmlinux.context.layers[self.vmlinux.layer_name] for page_addr in self._idstorage.get_entries(self._page_cache.i_pages): - if not page_addr: - continue + if not layer.is_valid(page_addr): + error_msg = f"Invalid cached page address at {page_addr:#x}, aborting" + vollog.error(error_msg) + raise exceptions.LinuxPageCacheException(error_msg) page = self.vmlinux.object("page", offset=page_addr, absolute=True) - if page: - yield page + if not page.is_valid(): + error_msg = f"Invalid cached page at {page_addr:#x}, aborting" + vollog.error(error_msg) + raise exceptions.LinuxPageCacheException(error_msg) + + yield page diff --git a/volatility3/framework/symbols/linux/extensions/__init__.py b/volatility3/framework/symbols/linux/extensions/__init__.py index 34d0fcba9..90a3cbc6e 100644 --- a/volatility3/framework/symbols/linux/extensions/__init__.py +++ b/volatility3/framework/symbols/linux/extensions/__init__.py @@ -15,12 +15,11 @@ from typing import Generator, Iterable, Iterator, Optional, Tuple, List, Union, from volatility3.framework import constants, exceptions, objects, interfaces, symbols from volatility3.framework.renderers import conversion from volatility3.framework.constants import linux as linux_constants -from volatility3.framework.layers import linear +from volatility3.framework.layers import linear, intel from volatility3.framework.objects import utility from volatility3.framework.symbols import generic, linux, intermed from volatility3.framework.symbols.linux.extensions import elf - vollog = logging.getLogger(__name__) # Keep these in a basic module, to prevent import cycles when symbol providers require them @@ -307,6 +306,46 @@ class module(generic.GenericIntelProcess): class task_struct(generic.GenericIntelProcess): + def is_valid(self) -> bool: + layer = self._context.layers[self.vol.layer_name] + # Make sure the entire task content is readable + if not layer.is_valid(self.vol.offset, self.vol.size): + return False + + if self.pid < 0 or self.tgid < 0: + return False + + if self.has_member("signal") and not ( + self.signal and self.signal.is_readable() + ): + return False + + if self.has_member("nsproxy") and not ( + self.nsproxy and self.nsproxy.is_readable() + ): + return False + + if self.has_member("real_parent") and not ( + self.real_parent and self.real_parent.is_readable() + ): + return False + + if ( + self.has_member("active_mm") + and self.active_mm + and not self.active_mm.is_readable() + ): + return False + + if self.mm: + if not self.mm.is_readable(): + return False + + if self.mm != self.active_mm: + return False + + return True + def add_process_layer( self, config_prefix: Optional[str] = None, preferred_name: Optional[str] = None ) -> Optional[str]: @@ -324,9 +363,11 @@ class task_struct(generic.GenericIntelProcess): raise TypeError( "Parent layer is not a translation layer, unable to construct process layer" ) - dtb, layer_name = parent_layer.translate(pgd) - if not dtb: + try: + dtb, layer_name = parent_layer.translate(pgd) + except exceptions.InvalidAddressException: return None + if preferred_name is None: preferred_name = self.vol.layer_name + f"_Process{self.pid}" # Add the constructed layer and return the name @@ -399,6 +440,8 @@ class task_struct(generic.GenericIntelProcess): tasks_iterable = self._get_tasks_iterable() threads_seen = set([self.vol.offset]) for task in tasks_iterable: + if not task.is_valid(): + continue if task.vol.offset not in threads_seen: threads_seen.add(task.vol.offset) yield task @@ -809,23 +852,30 @@ class mm_struct(objects.StructType): def _get_mmap_iter(self) -> Iterable[interfaces.objects.ObjectInterface]: """Returns an iterator for the mmap list member of an mm_struct. Use this only if required, get_vma_iter() will choose the correct _get_maple_tree_iter() or - _get_mmap_iter() automatically as required.""" + _get_mmap_iter() automatically as required. + + Yields: + vm_area_struct objects + """ if not self.has_member("mmap"): raise AttributeError( "_get_mmap_iter called on mm_struct where no mmap member exists." ) - if not self.mmap: + vma_pointer = self.mmap + if not (vma_pointer and vma_pointer.is_readable()): return None - yield self.mmap + vma_object = vma_pointer.dereference() + yield vma_object - seen = {self.mmap.vol.offset} - link = self.mmap.vm_next + seen = {vma_pointer} + vma_pointer = vma_pointer.vm_next - while link != 0 and link.vol.offset not in seen: - yield link - seen.add(link.vol.offset) - link = link.vm_next + while vma_pointer and vma_pointer.is_readable() and vma_pointer not in seen: + vma_object = vma_pointer.dereference() + yield vma_object + seen.add(vma_pointer) + vma_pointer = vma_pointer.vm_next # TODO: As of version 3.0.0 this method should be removed def get_maple_tree_iter(self) -> Iterable[interfaces.objects.ObjectInterface]: @@ -840,7 +890,11 @@ class mm_struct(objects.StructType): def _get_maple_tree_iter(self) -> Iterable[interfaces.objects.ObjectInterface]: """Returns an iterator for the mm_mt member of an mm_struct. Use this only if required, get_vma_iter() will choose the correct _get_maple_tree_iter() or - get_mmap_iter() automatically as required.""" + get_mmap_iter() automatically as required. + + Yields: + vm_area_struct objects + """ if not self.has_member("mm_mt"): raise AttributeError( @@ -848,20 +902,27 @@ class mm_struct(objects.StructType): ) symbol_table_name = self.get_symbol_table_name() for vma_pointer in self.mm_mt.get_slot_iter(): - # convert pointer to vm_area_struct and yield - vma = self._context.object( + # Convert pointer to vm_area_struct and yield + vma_object = self._context.object( symbol_table_name + constants.BANG + "vm_area_struct", layer_name=self.vol.native_layer_name, offset=vma_pointer, ) - yield vma + yield vma_object def get_vma_iter(self) -> Iterable[interfaces.objects.ObjectInterface]: - """Returns an iterator for the VMAs in an mm_struct. Automatically choosing the mmap or mm_mt as required.""" + """Returns an iterator for the VMAs in an mm_struct. + Automatically choosing the mmap or mm_mt as required. + + Yields: + vm_area_struct objects + """ if self.has_member("mmap"): + # kernels < 6.1 yield from self._get_mmap_iter() elif self.has_member("mm_mt"): + # kernels >= 6.1 d4af56c5c7c6781ca6ca8075e2cf5bc119ed33d1 yield from self._get_maple_tree_iter() else: raise AttributeError("Unable to find mmap or mm_mt in mm_struct") @@ -1207,35 +1268,43 @@ class list_head(objects.StructType, collections.abc.Iterable): Objects of the type specified via the "symbol_type" argument. """ - layer = layer or self.vol.layer_name + layer_name = layer or self.vol.layer_name + + trans_layer = self._context.layers[layer_name] + if not trans_layer.is_valid(self.vol.offset): + return None relative_offset = self._context.symbol_space.get_type( symbol_type ).relative_child_offset(member) - direction = "prev" - if forward: - direction = "next" - try: - link = getattr(self, direction).dereference() - except exceptions.InvalidAddressException: + direction = "next" if forward else "prev" + + link_ptr = getattr(self, direction) + if not (link_ptr and link_ptr.is_readable()): return None + link = link_ptr.dereference() + if not sentinel: - yield self._context.object( - symbol_type, layer, offset=self.vol.offset - relative_offset - ) + obj_offset = self.vol.offset - relative_offset + if not trans_layer.is_valid(obj_offset): + return None + + yield self._context.object(symbol_type, layer_name, offset=obj_offset) + seen = {self.vol.offset} while link.vol.offset not in seen: - obj = self._context.object( - symbol_type, layer, offset=link.vol.offset - relative_offset - ) - yield obj + obj_offset = link.vol.offset - relative_offset + if not trans_layer.is_valid(obj_offset): + return None + + yield self._context.object(symbol_type, layer_name, offset=obj_offset) seen.add(link.vol.offset) - try: - link = getattr(link, direction).dereference() - except exceptions.InvalidAddressException: + link_ptr = getattr(link, direction) + if not (link_ptr and link_ptr.is_readable()): break + link = link_ptr.dereference() def __iter__(self) -> Iterator[interfaces.objects.ObjectInterface]: return self.to_list(self.vol.parent.vol.type_name, self.vol.member_name) @@ -2487,7 +2556,12 @@ class inode(objects.StructType): """ if not self.i_size: return - elif not (self.i_mapping and self.i_mapping.nrpages > 0): + + if not ( + self.i_mapping + and self.i_mapping.is_readable() + and self.i_mapping.nrpages > 0 + ): return page_cache = linux.PageCache( @@ -2495,19 +2569,26 @@ class inode(objects.StructType): kernel_module_name="kernel", page_cache=self.i_mapping.dereference(), ) + yield from page_cache.get_cached_pages() - def get_contents(self): + def get_contents(self) -> Iterable[Tuple[int, bytes]]: """Get the inode cached pages from the page cache Yields: page_index (int): The page index in the Tree. File offset is page_index * PAGE_SIZE. - page_content (str): The page content + page_content (bytes): The page content """ for page_obj in self.get_pages(): + if page_obj.mapping != self.i_mapping: + vollog.warning( + f"Cached page at {page_obj.vol.offset:#x} has a mismatched address space with the inode. Skipping page" + ) + continue page_index = int(page_obj.index) page_content = page_obj.get_content() - yield page_index, page_content + if page_content: + yield page_index, page_content class address_space(objects.StructType): @@ -2515,7 +2596,7 @@ class address_space(objects.StructType): def i_pages(self): """Returns the appropriate member containing the page cache tree""" if self.has_member("i_pages"): - # Kernel >= 4.17 + # Kernel >= 4.17 b93b016313b3ba8003c3b8bb71f569af91f19fc7 return self.member("i_pages") elif self.has_member("page_tree"): # Kernel < 4.17 @@ -2525,16 +2606,22 @@ class address_space(objects.StructType): class page(objects.StructType): - @property - @functools.lru_cache + def is_valid(self) -> bool: + if self.mapping and not self.mapping.is_readable(): + return False + + if self.to_paddr() < 0: + return False + + return True + + @functools.cached_property def pageflags_enum(self) -> Dict: """Returns 'pageflags' enumeration key/values Returns: A dictionary with the pageflags enumeration key/values """ - # FIXME: It would be even better to use @functools.cached_property instead, - # however, this requires Python +3.8 try: pageflags_enum = self._context.symbol_space.get_enumeration( self.get_symbol_table_name() + constants.BANG + "pageflags" @@ -2548,24 +2635,12 @@ class page(objects.StructType): return pageflags_enum - def get_flags_list(self) -> List[str]: - """Returns a list of page flags + @functools.cached_property + def _intel_vmemmap_start(self) -> int: + """Determine the start of the struct page array, for Intel systems. Returns: - List of page flags - """ - flags = [] - for name, value in self.pageflags_enum.items(): - if self.flags & (1 << value) != 0: - flags.append(name) - - return flags - - def to_paddr(self) -> int: - """Converts a page's virtual address to its physical address using the current physical memory model. - - Returns: - int: page physical address + int: vmemmap_start address """ vmlinux = linux.LinuxUtilities.get_module_from_volobj_type(self._context, self) vmlinux_layer = vmlinux.context.layers[vmlinux.layer_name] @@ -2605,14 +2680,40 @@ class page(objects.StructType): "Something went wrong, we shouldn't be here" ) - page_type_size = vmlinux.get_type("page").size + return vmemmap_start + + def _intel_to_paddr(self) -> int: + """Converts a page's virtual address to its physical address using the current Intel memory model. + + Returns: + int: page physical address + """ + vmlinux = linux.LinuxUtilities.get_module_from_volobj_type(self._context, self) + vmlinux_layer = vmlinux.context.layers[vmlinux.layer_name] pagec = vmlinux_layer.canonicalize(self.vol.offset) - pfn = (pagec - vmemmap_start) // page_type_size + pfn = (pagec - self._intel_vmemmap_start) // vmlinux.get_type("page").size page_paddr = pfn * vmlinux_layer.page_size return page_paddr - def get_content(self) -> Union[str, None]: + def to_paddr(self) -> int: + """Converts a page's virtual address to its physical address using the current CPU memory model. + + Returns: + int: page physical address + """ + vmlinux = linux.LinuxUtilities.get_module_from_volobj_type(self._context, self) + vmlinux_layer = vmlinux.context.layers[vmlinux.layer_name] + if isinstance(vmlinux_layer, intel.Intel): + page_paddr = self._intel_to_paddr() + else: + raise exceptions.LayerException( + f"Architecture {type(vmlinux_layer)} vmemmap_start calculation isn't currently supported." + ) + + return page_paddr + + def get_content(self) -> Union[bytes, None]: """Returns the page content Returns: @@ -2620,13 +2721,34 @@ class page(objects.StructType): """ vmlinux = linux.LinuxUtilities.get_module_from_volobj_type(self._context, self) vmlinux_layer = vmlinux.context.layers[vmlinux.layer_name] - physical_layer = vmlinux.context.layers["memory_layer"] + physical_layer_name = self._context.layers[self.vol.layer_name].config.get( + "memory_layer", self.vol.layer_name + ) + physical_layer = self._context.layers[physical_layer_name] page_paddr = self.to_paddr() if not page_paddr: return None - page_data = physical_layer.read(page_paddr, vmlinux_layer.page_size) - return page_data + if not physical_layer.is_valid(page_paddr, length=vmlinux_layer.page_size): + vollog.debug( + "Unable to read page 0x%x content at 0x%x", self.vol.offset, page_paddr + ) + return None + + return physical_layer.read(page_paddr, vmlinux_layer.page_size) + + def get_flags_list(self) -> List[str]: + """Returns a list of page flags + + Returns: + List of page flags + """ + flags = [] + for name, value in self.pageflags_enum.items(): + if self.flags & (1 << value) != 0: + flags.append(name) + + return flags class IDR(objects.StructType): @@ -2726,17 +2848,17 @@ class IDR(objects.StructType): class rb_root(objects.StructType): - def _walk_nodes(self, root_node) -> Iterator[int]: + def _walk_nodes(self, root_node: int) -> Iterator[int]: """Traverses the Red-Black tree from the root node and yields a pointer to each node in this tree. Args: - root_node: A Red-Black tree node from which to start descending + root_node: A Red-Black tree node pointer from which to start descending Yields: A pointer to every node descending from the specified root node """ - if not root_node: + if not (root_node and root_node.is_readable()): return yield root_node @@ -2751,3 +2873,111 @@ class rb_root(objects.StructType): """ yield from self._walk_nodes(root_node=self.rb_node) + + +class scatterlist(objects.StructType): + SG_CHAIN = 0x01 + SG_END = 0x02 + SG_PAGE_LINK_MASK = SG_CHAIN | SG_END + + def _sg_flags(self) -> int: + return self.page_link & self.SG_PAGE_LINK_MASK + + def _sg_is_chain(self) -> int: + return self._sg_flags() & self.SG_CHAIN + + def _sg_is_last(self) -> int: + return self._sg_flags() & self.SG_END + + def _sg_chain_ptr(self) -> int: + """Clears the last two bits basically.""" + return self.page_link & ~self.SG_PAGE_LINK_MASK + + def _sg_dma_len(self) -> int: + # Depends on CONFIG_NEED_SG_DMA_LENGTH + if self.has_member("dma_length"): + return self.dma_length + return self.length + + def _get_sg_max_single_alloc(self) -> int: + """Based on kernel's SG_MAX_SINGLE_ALLOC. + + Doc. from kernel source : + * Maximum number of entries that will be allocated in one piece, if + * a list larger than this is required then chaining will be utilized. + """ + return self._context.layers[self.vol.layer_name].page_size // self.vol.size + + def _sg_next(self) -> Optional[interfaces.objects.ObjectInterface]: + """Get the next scatterlist struct from the list. + Based on kernel's sg_next. + + Doc. from kernel source : + * Notes on SG table design. + * + * We use the unsigned long page_link field in the scatterlist struct to place + * the page pointer AND encode information about the sg table as well. The two + * lower bits are reserved for this information. + * + * If bit 0 is set, then the page_link contains a pointer to the next sg + * table list. Otherwise the next entry is at sg + 1. + * + * If bit 1 is set, then this sg entry is the last element in a list. + """ + if self._sg_is_last(): + return None + + if self._sg_is_chain(): + next_address = self._sg_chain_ptr() + else: + next_address = self.vol.offset + self.vol.size + + sg = self._context.object( + self.get_symbol_table_name() + constants.BANG + "scatterlist", + self.vol.layer_name, + next_address, + ) + return sg + + def for_each_sg(self) -> Optional[Iterator[interfaces.objects.ObjectInterface]]: + """Iterate over each struct in the scatterlist.""" + sg = self + sg_max_single_alloc = self._get_sg_max_single_alloc() + + # Empty scatterlists protection + if sg.page_link == 0 and sg._sg_dma_len() == 0 and sg.dma_address == 0: + return None + else: + # Yield itself first + yield sg + + entries_count = 1 + # entries_count <= sg_max_single_alloc should always be true if the + # scatterlists were correctly chained. + while entries_count <= sg_max_single_alloc: + sg = sg._sg_next() + if sg is None: + break + # Points to a new scatterlist + elif sg._sg_is_chain(): + entries_count = 0 + else: + entries_count += 1 + yield sg + + def get_content( + self, + ) -> Optional[Iterator[bytes]]: + """Traverse a scatterlist to gather content located at each + dma_address position. + + Returns: + An iterator of bytes + """ + # Either "physical" is layer-1 because this is a module layer, or "physical" is the current layer + physical_layer_name = self._context.layers[self.vol.layer_name].config.get( + "memory_layer", self.vol.layer_name + ) + physical_layer = self._context.layers[physical_layer_name] + for sg in self.for_each_sg(): + yield from physical_layer.read(sg.dma_address, sg._sg_dma_len()) diff --git a/volatility3/framework/symbols/linux/utilities/__init__.py b/volatility3/framework/symbols/linux/utilities/__init__.py new file mode 100644 index 000000000..e69de29bb diff --git a/volatility3/framework/symbols/linux/utilities/tainting.py b/volatility3/framework/symbols/linux/utilities/tainting.py new file mode 100644 index 000000000..2360401d5 --- /dev/null +++ b/volatility3/framework/symbols/linux/utilities/tainting.py @@ -0,0 +1,161 @@ +import functools + +from volatility3 import framework +from volatility3.framework import interfaces +from volatility3.framework.constants import linux as linux_constants +from typing import List, Optional + + +class Tainting(interfaces.configuration.VersionableInterface): + """Tainted kernel and modules parsing capabilities. + + Relevant Linux kernel functions: + - modules: module_flags_taint + - kernel: print_tainted + """ + + _version = (1, 0, 0) + _required_framework_version = (2, 0, 0) + + framework.require_interface_version(*_required_framework_version) + + @classmethod + @functools.lru_cache + def _get_kernel_taint_flags_list( + cls, + context: interfaces.context.ContextInterface, + kernel_module_name: str, + ) -> Optional[List[interfaces.objects.ObjectInterface]]: + """Determine whether the kernel embeds taint flags definition + in-memory or not. + + Returns: + A list of "taint_flag" kernel objects if taint_flags symbol exists + """ + kernel = context.modules[kernel_module_name] + if kernel.has_symbol("taint_flags"): + return list(kernel.object_from_symbol("taint_flags")) + return None + + @classmethod + def _module_flags_taint_pre_4_10_rc1( + cls, + taints: int, + is_module: bool = False, + ) -> str: + """Convert the module's taints value to a 1-1 character mapping. + Relies on statically defined taints mappings in the framework. + + Args: + taints: The taints value, represented by an integer + is_module: Indicates if the taints value is associated with a built-in/LKM module + + Returns: + The raw taints string. + """ + taints_string = "" + for char, taint_flag in linux_constants.TAINT_FLAGS.items(): + if is_module and not taint_flag.module: + continue + + if taints & taint_flag.shift: + taints_string += char + + return taints_string + + @classmethod + def _module_flags_taint_post_4_10_rc1( + cls, + context: interfaces.context.ContextInterface, + kernel_module_name: str, + taints: int, + is_module: bool = False, + ) -> str: + """Convert the module's taints value to a 1-1 character mapping. + Relies on kernel symbol embedded taints definitions. + + struct taint_flag { + char c_true; /* character printed when tainted */ + char c_false; /* character printed when not tainted */ + bool module; /* also show as a per-module taint flag */ + }; + + Args: + taints: The taints value, represented by an integer + is_module: Indicates if the taints value is associated with a built-in/LKM module + + Returns: + The raw taints string. + """ + taints_string = "" + for taint_bit, taint_flag in enumerate( + cls._get_kernel_taint_flags_list(context, kernel_module_name) + ): + if is_module and not taint_flag.module: + continue + c_true = chr(taint_flag.c_true) + c_false = chr(taint_flag.c_false) + if taints & (1 << taint_bit): + taints_string += c_true + elif c_false != " ": + taints_string += c_false + + return taints_string + + @classmethod + def get_taints_as_plain_string( + cls, + context: interfaces.context.ContextInterface, + kernel_module_name: str, + taints: int, + is_module: bool = False, + ) -> str: + """Convert the taints value to a 1-1 character mapping. + + Args: + taints: The taints value, represented by an integer + is_module: Indicates if the taints value is associated with a built-in/LKM module + Returns: + The raw taints string. + + Documentation: + - module_flags_taint kernel function + """ + + if cls._get_kernel_taint_flags_list(context, kernel_module_name): + return cls._module_flags_taint_post_4_10_rc1( + context, kernel_module_name, taints, is_module + ) + return cls._module_flags_taint_pre_4_10_rc1(taints, is_module) + + @classmethod + def get_taints_parsed( + cls, + context: interfaces.context.ContextInterface, + kernel_module_name: str, + taints: int, + is_module: bool = False, + ) -> List[str]: + """Convert the taints string to a 1-1 descriptor mapping. + + Args: + taints: The taints value, represented by an integer + is_module: Indicates if the taints value is associated with a built-in/LKM module + + Returns: + A comprehensive (user-friendly) taint descriptor list. + + Documentation: + - module_flags_taint kernel function + """ + comprehensive_taints = [] + for character in cls.get_taints_as_plain_string( + context, kernel_module_name, taints, is_module + ): + taint_flag = linux_constants.TAINT_FLAGS.get(character) + if not taint_flag: + comprehensive_taints.append(f"") + elif taint_flag.when_present: + comprehensive_taints.append(taint_flag.desc) + + return comprehensive_taints diff --git a/volatility3/framework/symbols/native.py b/volatility3/framework/symbols/native.py index 7c3e1b312..61417532e 100644 --- a/volatility3/framework/symbols/native.py +++ b/volatility3/framework/symbols/native.py @@ -30,7 +30,7 @@ class NativeTable(interfaces.symbols.NativeTableInterface): @property def types(self) -> Iterable[str]: - """Returns an iterator of the symbol type names.""" + """Returns an iterable (set) of the available symbol type names.""" return self._types def get_type(self, type_name: str) -> interfaces.objects.Template: diff --git a/volatility3/framework/symbols/windows/extensions/__init__.py b/volatility3/framework/symbols/windows/extensions/__init__.py index f12fd3f5b..214002f49 100755 --- a/volatility3/framework/symbols/windows/extensions/__init__.py +++ b/volatility3/framework/symbols/windows/extensions/__init__.py @@ -962,56 +962,55 @@ class LIST_ENTRY(objects.StructType, collections.abc.Iterable): ) -> Iterator[interfaces.objects.ObjectInterface]: """Returns an iterator of the entries in the list.""" - layer = layer or self.vol.layer_name + layer_name = layer or self.vol.layer_name + native_layer_name = layer_name or self.vol.native_layer_name + + trans_layer = self._context.layers[layer_name] + if not trans_layer.is_valid(self.vol.offset): + return None relative_offset = self._context.symbol_space.get_type( symbol_type ).relative_child_offset(member) - direction = "Blink" - if forward: - direction = "Flink" + direction = "Flink" if forward else "Blink" - trans_layer = self._context.layers[layer] - - try: - is_valid = trans_layer.is_valid(self.vol.offset) - if not is_valid: - return None - - link = getattr(self, direction).dereference() - except exceptions.InvalidAddressException: + link_ptr = getattr(self, direction) + if not (link_ptr and link_ptr.is_readable()): return None + link = link_ptr.dereference() if not sentinel: + obj_offset = self.vol.offset - relative_offset + if not trans_layer.is_valid(obj_offset): + return None + yield self._context.object( symbol_type, - layer, - offset=self.vol.offset - relative_offset, - native_layer_name=layer or self.vol.native_layer_name, + layer_name, + offset=obj_offset, + native_layer_name=native_layer_name, ) seen = {self.vol.offset} while link.vol.offset not in seen: obj_offset = link.vol.offset - relative_offset - if not trans_layer.is_valid(obj_offset): return None - obj = self._context.object( + yield self._context.object( symbol_type, - layer, + layer_name, offset=obj_offset, - native_layer_name=layer or self.vol.native_layer_name, + native_layer_name=native_layer_name, ) - yield obj seen.add(link.vol.offset) - try: - link = getattr(link, direction).dereference() - except exceptions.InvalidAddressException: + link_ptr = getattr(link, direction) + if not (link_ptr and link_ptr.is_readable()): return None + link = link_ptr.dereference() def __iter__(self) -> Iterator[interfaces.objects.ObjectInterface]: return self.to_list(self.vol.parent.vol.type_name, self.vol.member_name) diff --git a/volatility3/framework/symbols/windows/extensions/registry.py b/volatility3/framework/symbols/windows/extensions/registry.py index 9e2f8df3b..a8cc7703c 100644 --- a/volatility3/framework/symbols/windows/extensions/registry.py +++ b/volatility3/framework/symbols/windows/extensions/registry.py @@ -133,8 +133,17 @@ class CM_KEY_BODY(objects.StructType): def get_full_key_name(self) -> str: output = [] + seen = set() + kcb = self.KeyControlBlock while kcb.ParentKcb: + if kcb.ParentKcb.vol.offset in seen: + return None + seen.add(kcb.ParentKcb.vol.offset) + + if len(output) > 128: + return None + if kcb.NameBlock.Name is None: break @@ -159,14 +168,20 @@ class CM_KEY_NODE(objects.StructType): """Extension to allow traversal of registry keys.""" def get_volatile(self) -> bool: + """ + Returns a bool indicating whether or not the key is volatile. + + Raises TypeError if the key was not instantiated on a RegistryHive layer + """ if not isinstance(self._context.layers[self.vol.layer_name], RegistryHive): - raise ValueError( - "Cannot determine volatility of registry key without an offset in a RegistryHive layer" - ) + raise TypeError("CM_KEY_NODE was not instantiated on a RegistryHive layer") return bool(self.vol.offset & 0x80000000) def get_subkeys(self) -> Iterator["CM_KEY_NODE"]: - """Returns a list of the key nodes.""" + """Returns a list of the key nodes. + + Raises TypeError if the key was not instantiated on a RegistryHive layer + """ hive = self._context.layers[self.vol.layer_name] if not isinstance(hive, RegistryHive): raise TypeError("CM_KEY_NODE was not instantiated on a RegistryHive layer") @@ -222,7 +237,10 @@ class CM_KEY_NODE(objects.StructType): yield from self._get_subkeys_recursive(hive, subnode) def get_values(self) -> Iterator["CM_KEY_VALUE"]: - """Returns a list of the Value nodes for a key.""" + """Returns a list of the Value nodes for a key. + + Raises TypeError if the key was not instantiated on a RegistryHive layer + """ hive = self._context.layers[self.vol.layer_name] if not isinstance(hive, RegistryHive): raise TypeError("CM_KEY_NODE was not instantiated on a RegistryHive layer") @@ -251,6 +269,11 @@ class CM_KEY_NODE(objects.StructType): return self.Name.cast("string", max_length=namelength, encoding="latin-1") def get_key_path(self) -> str: + """ + Returns the full path to this registry key. + + Raises TypeError if the key was not instantiated on a RegistryHive layer + """ reg = self._context.layers[self.vol.layer_name] if not isinstance(reg, RegistryHive): raise TypeError("Key was not instantiated on a RegistryHive layer") @@ -276,7 +299,16 @@ class CM_KEY_VALUE(objects.StructType): return RegValueTypes(self.Type) def decode_data(self) -> Union[int, bytes]: - """Properly decodes the data associated with the value node""" + """ + Properly decodes the data associated with the value node. + + If an InvalidAddressException occurs when reading data from the + underlying RegistryHive layer, the data will be padded with null bytes + of the same length. + + Raises ValueError if the data cannot be read + Raises TypeError if the class was not instantiated on a RegistryHive layer + """ # Determine if the data is stored inline datalen = self.DataLength data = b"" @@ -310,14 +342,26 @@ class CM_KEY_VALUE(objects.StructType): and block_offset < layer.maximum_address ): amount = min(BIG_DATA_MAXLEN, datalen) - data += layer.read( - offset=layer.get_cell(block_offset).vol.offset, length=amount - ) + try: + data += layer.read( + offset=layer.get_cell(block_offset).vol.offset, + length=amount, + ) + except exceptions.InvalidAddressException: + vollog.debug( + f"Failed to read {amount:x} bytes of data, padding with {amount:x}" + ) datalen -= amount else: # Suspect Data actually points to a Cell, # but the length at the start could be negative so just adding 4 to jump past it - data = layer.read(self.Data + 4, datalen) + try: + data = layer.read(self.Data + 4, datalen) + except exceptions.InvalidAddressException: + vollog.debug( + f"Failed to read {datalen:x} bytes of data, returning {datalen:x} null bytes" + ) + data = b"\x00" * datalen if self.get_type() == RegValueTypes.REG_DWORD: if len(data) != struct.calcsize(" Set[str]: """Loads up the list of successfully cached json objects, so we don't need @@ -93,6 +95,13 @@ def valid( return True try: import jsonschema + + schema_key = json.dumps(schema, sort_keys=True) + if schema_key not in validators: + validator_class = jsonschema.validators.validator_for(schema) + validator_class.check_schema(schema) + validator = validator_class(schema) + validators[schema_key] = validator except ImportError: vollog.info("Dependency for validation unavailable: jsonschema") vollog.debug("All validations will report success, even with malformed input") @@ -100,7 +109,7 @@ def valid( try: vollog.debug("Validating JSON against schema...") - jsonschema.validate(input, schema) + validators[schema_key].validate(input) cached_validations.add(input_hash) vollog.debug("JSON validated against schema (result cached)") except jsonschema.exceptions.SchemaError: