Merge branch 'develop' into linux_fix_mnt_namespace_issue_1187

.github/workflows/test.yaml

@@ -46,7 +46,7 @@ jobs:
 
     - name: Clean up post-test
       run: |
-        rm -rf *.lime
+        rm -rf *.bin
         rm -rf *.img
         cd volatility3/symbols
         rm -rf linux

volatility3/framework/constants/_version.py

@@ -1,7 +1,7 @@
 # We use the SemVer 2.0.0 versioning scheme
 VERSION_MAJOR = 2  # Number of releases of the library with a breaking change
 VERSION_MINOR = 8  # Number of changes that only add to the interface
-VERSION_PATCH = 0  # Number of changes that do not change the interface
+VERSION_PATCH = 1  # Number of changes that do not change the interface
 VERSION_SUFFIX = ""
 
 PACKAGE_VERSION = (

volatility3/framework/plugins/linux/ebpf.py

@@ -0,0 +1,73 @@
+# 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
+
+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
+
+vollog = logging.getLogger(__name__)
+
+
+class EBPF(plugins.PluginInterface):
+    """Enumerate eBPF programs"""
+
+    _required_framework_version = (2, 0, 0)
+
+    _version = (1, 0, 0)
+
+    @classmethod
+    def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]:
+        return [
+            requirements.ModuleRequirement(
+                name="kernel",
+                description="Linux kernel",
+                architectures=["Intel32", "Intel64"],
+            ),
+        ]
+
+    def get_ebpf_programs(
+        self,
+        context: interfaces.context.ContextInterface,
+        vmlinux_module_name: str,
+    ) -> interfaces.objects.ObjectInterface:
+        """Enumerate eBPF programs walking its IDR.
+
+        Args:
+            context: The context to retrieve required elements (layers, symbol tables) from
+            vmlinux_module_name: The name of the kernel module on which to operate
+        Yields:
+            eBPF program objects
+        """
+        vmlinux = context.modules[vmlinux_module_name]
+
+        if not vmlinux.has_symbol("prog_idr"):
+            raise exceptions.VolatilityException(
+                "Cannot find the eBPF prog idr. Unsupported kernel"
+            )
+
+        prog_idr = vmlinux.object_from_symbol("prog_idr")
+        for page_addr in prog_idr.get_entries():
+            bpf_prog = vmlinux.object("bpf_prog", offset=page_addr, absolute=True)
+            yield bpf_prog
+
+    def _generator(self):
+        for prog in self.get_ebpf_programs(self.context, self.config["kernel"]):
+            prog_addr = prog.vol.offset
+            prog_type = prog.get_type() or renderers.NotAvailableValue()
+            prog_tag = prog.get_tag() or renderers.NotAvailableValue()
+            prog_name = prog.get_name() or renderers.NotAvailableValue()
+            fields = (format_hints.Hex(prog_addr), prog_name, prog_tag, prog_type)
+            yield (0, fields)
+
+    def run(self):
+        headers = [
+            ("Address", format_hints.Hex),
+            ("Name", str),
+            ("Tag", str),
+            ("Type", str),
+        ]
+        return renderers.TreeGrid(headers, self._generator())

volatility3/framework/plugins/linux/lsof.py

@@ -1,27 +1,27 @@
-# This file is Copyright 2019 Volatility Foundation and licensed under the Volatility Software License 1.0
+# 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
 #
 """A module containing a collection of plugins that produce data typically
 found in Linux's /proc file system."""
-import logging
+import logging, datetime
 from typing import List, Callable
 
-from volatility3.framework import renderers, interfaces, constants
+from volatility3.framework import renderers, interfaces, constants, exceptions
 from volatility3.framework.configuration import requirements
 from volatility3.framework.interfaces import plugins
 from volatility3.framework.objects import utility
 from volatility3.framework.symbols import linux
 from volatility3.plugins.linux import pslist
+from volatility3.plugins import timeliner
 
 vollog = logging.getLogger(__name__)
 
 
-class Lsof(plugins.PluginInterface):
-    """Lists all memory maps for all processes."""
+class Lsof(plugins.PluginInterface, timeliner.TimeLinerInterface):
+    """Lists open files for each processes."""
 
     _required_framework_version = (2, 0, 0)
-
-    _version = (1, 1, 0)
+    _version = (1, 2, 0)
 
     @classmethod
     def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]:
@@ -45,6 +45,29 @@ class Lsof(plugins.PluginInterface):
             ),
         ]
 
+    @classmethod
+    def get_inode_metadata(cls, filp: interfaces.objects.ObjectInterface):
+        try:
+            dentry = filp.get_dentry()
+            if dentry:
+                inode_object = dentry.d_inode
+                if inode_object and inode_object.is_valid():
+                    itype = (
+                        inode_object.get_inode_type() or renderers.NotAvailableValue()
+                    )
+                    return (
+                        inode_object.i_ino,
+                        itype,
+                        inode_object.i_size,
+                        inode_object.get_file_mode(),
+                        inode_object.get_change_time(),
+                        inode_object.get_modification_time(),
+                        inode_object.get_access_time(),
+                    )
+        except (exceptions.InvalidAddressException, AttributeError) as e:
+            vollog.warning(f"Can't get inode metadata: {e}")
+        return None
+
     @classmethod
     def list_fds(
         cls,
@@ -52,7 +75,7 @@ class Lsof(plugins.PluginInterface):
         symbol_table: str,
         filter_func: Callable[[int], bool] = lambda _: False,
     ):
-        linuxutils_symbol_table = None  # type: ignore
+        linuxutils_symbol_table = None
         for task in pslist.PsList.list_tasks(context, symbol_table, filter_func):
             if linuxutils_symbol_table is None:
                 if constants.BANG not in task.vol.type_name:
@@ -69,21 +92,79 @@ class Lsof(plugins.PluginInterface):
             for fd_fields in fd_generator:
                 yield pid, task_comm, task, fd_fields
 
+    @classmethod
+    def list_fds_and_inodes(
+        cls,
+        context: interfaces.context.ContextInterface,
+        symbol_table: str,
+        filter_func: Callable[[int], bool] = lambda _: False,
+    ):
+        for pid, task_comm, task, (fd_num, filp, full_path) in cls.list_fds(
+            context, symbol_table, filter_func
+        ):
+            inode_metadata = cls.get_inode_metadata(filp)
+            if inode_metadata is None:
+                inode_metadata = tuple(
+                    interfaces.renderers.BaseAbsentValue() for _ in range(7)
+                )
+            yield pid, task_comm, task, fd_num, filp, full_path, inode_metadata
+
     def _generator(self, pids, symbol_table):
         filter_func = pslist.PsList.create_pid_filter(pids)
-        fds_generator = self.list_fds(
+        fds_generator = self.list_fds_and_inodes(
             self.context, symbol_table, filter_func=filter_func
         )
 
-        for pid, task_comm, _task, fd_fields in fds_generator:
-            fd_num, _filp, full_path = fd_fields
-
-            fields = (pid, task_comm, fd_num, full_path)
+        for (
+            pid,
+            task_comm,
+            task,
+            fd_num,
+            filp,
+            full_path,
+            inode_metadata,
+        ) in fds_generator:
+            inode_num, itype, file_size, imode, ctime, mtime, atime = inode_metadata
+            fields = (
+                pid,
+                task_comm,
+                fd_num,
+                full_path,
+                inode_num,
+                itype,
+                imode,
+                ctime,
+                mtime,
+                atime,
+                file_size,
+            )
             yield (0, fields)
 
     def run(self):
         pids = self.config.get("pid", None)
         symbol_table = self.config["kernel"]
 
-        tree_grid_args = [("PID", int), ("Process", str), ("FD", int), ("Path", str)]
+        tree_grid_args = [
+            ("PID", int),
+            ("Process", str),
+            ("FD", int),
+            ("Path", str),
+            ("Inode", int),
+            ("Type", str),
+            ("Mode", str),
+            ("Changed", datetime.datetime),
+            ("Modified", datetime.datetime),
+            ("Accessed", datetime.datetime),
+            ("Size", int),
+        ]
         return renderers.TreeGrid(tree_grid_args, self._generator(pids, symbol_table))
+
+    def generate_timeline(self):
+        pids = self.config.get("pid", None)
+        symbol_table = self.config["kernel"]
+        for row in self._generator(pids, symbol_table):
+            _depth, row_data = row
+            description = f'Process {row_data[1]} ({row_data[0]}) Open "{row_data[3]}"'
+            yield description, timeliner.TimeLinerType.CHANGED, row_data[7]
+            yield description, timeliner.TimeLinerType.MODIFIED, row_data[8]
+            yield description, timeliner.TimeLinerType.ACCESSED, row_data[9]

volatility3/framework/plugins/linux/mountinfo.py

@@ -37,7 +37,7 @@ class MountInfo(plugins.PluginInterface):
 
     _required_framework_version = (2, 2, 0)
 
-    _version = (1, 0, 0)
+    _version = (1, 2, 0)
 
     @classmethod
     def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]:
@@ -143,10 +143,10 @@ class MountInfo(plugins.PluginInterface):
             sb_opts,
         )
 
+    @staticmethod
     def _get_tasks_mountpoints(
-        self,
         tasks: Iterable[interfaces.objects.ObjectInterface],
-        filtered_by_pids: bool,
+        filtered_by_pids: bool = False,
     ):
         seen_mountpoints = set()
         for task in tasks:
@@ -184,8 +184,8 @@ class MountInfo(plugins.PluginInterface):
         self,
         tasks: Iterable[interfaces.objects.ObjectInterface],
         mnt_ns_ids: List[int],
-        mount_format: bool,
-        filtered_by_pids: bool,
+        mount_format: bool = False,
+        filtered_by_pids: bool = False,
     ) -> Iterable[Tuple[int, Tuple]]:
         show_filter_warning = False
         for task, mnt, mnt_ns_id in self._get_tasks_mountpoints(
@@ -247,6 +247,37 @@ class MountInfo(plugins.PluginInterface):
                 "Could not filter by mount namespace id. This field is not available in this kernel."
             )
 
+    @classmethod
+    def get_superblocks(
+        cls,
+        context: interfaces.context.ContextInterface,
+        vmlinux_module_name: str,
+    ) -> Iterable[interfaces.objects.ObjectInterface]:
+        """Yield file system superblocks based on the task's mounted filesystems.
+
+        Args:
+            context: The context to retrieve required elements (layers, symbol tables) from
+            vmlinux_module_name: The name of the kernel module on which to operate
+
+        Yields:
+            super_block: Kernel's struct super_block object
+        """
+        # No filter so that we get all the mount namespaces from all tasks
+        tasks = pslist.PsList.list_tasks(context, vmlinux_module_name)
+
+        seen_sb_ptr = set()
+        for task, mnt, _mnt_ns_id in cls._get_tasks_mountpoints(tasks):
+            path_root = linux.LinuxUtilities.get_path_mnt(task, mnt)
+            if not path_root:
+                continue
+
+            sb_ptr = mnt.get_mnt_sb()
+            if not sb_ptr or sb_ptr in seen_sb_ptr:
+                continue
+            seen_sb_ptr.add(sb_ptr)
+
+            yield sb_ptr.dereference(), path_root
+
     def run(self):
         pids = self.config.get("pids")
         mount_ns_ids = self.config.get("mntns")

volatility3/framework/plugins/linux/pagecache.py

@@ -0,0 +1,531 @@
+# 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 math
+import logging
+import datetime
+from dataclasses import dataclass, astuple
+from typing import List, Set, Type, Iterable
+
+from volatility3.framework import renderers, interfaces
+from volatility3.framework.renderers import format_hints
+from volatility3.framework.interfaces import plugins
+from volatility3.framework.configuration import requirements
+from volatility3.plugins import timeliner
+from volatility3.plugins.linux import mountinfo
+
+vollog = logging.getLogger(__name__)
+
+
+@dataclass
+class InodeUser:
+    """Inode user representation, featuring augmented information and formatted fields.
+    This is the data the plugin will eventually display.
+    """
+
+    superblock_addr: int
+    mountpoint: str
+    device: str
+    inode_num: int
+    inode_addr: int
+    type: str
+    inode_pages: int
+    cached_pages: int
+    file_mode: str
+    access_time: str
+    modification_time: str
+    change_time: str
+    path: str
+
+
+@dataclass
+class InodeInternal:
+    """Inode internal representation containing only the core objects
+
+    Fields:
+        superblock: 'super_block' struct
+        mountpoint: Superblock mountpoint path
+        inode: 'inode' struct
+        path: Dentry full path
+    """
+
+    superblock: interfaces.objects.ObjectInterface
+    mountpoint: str
+    inode: interfaces.objects.ObjectInterface
+    path: str
+
+    def to_user(
+        self, kernel_layer: interfaces.layers.TranslationLayerInterface
+    ) -> InodeUser:
+        """Augment the inode information to be presented to the user
+
+        Args:
+            kernel_layer: The kernel layer to obtain the page size
+
+        Returns:
+            An InodeUser dataclass
+        """
+        # Ensure all types are atomic immutable. Otherwise, astuple() will take a long
+        # time doing a deepcopy of the Volatility objects.
+        superblock_addr = self.superblock.vol.offset
+        device = f"{self.superblock.major}:{self.superblock.minor}"
+        inode_num = int(self.inode.i_ino)
+        inode_addr = self.inode.vol.offset
+        inode_type = self.inode.get_inode_type() or renderers.UnparsableValue()
+        # Round up the number of pages to fit the inode's size
+        inode_pages = int(math.ceil(self.inode.i_size / float(kernel_layer.page_size)))
+        cached_pages = int(self.inode.i_mapping.nrpages)
+        file_mode = self.inode.get_file_mode()
+        access_time_dt = self.inode.get_access_time()
+        modification_time_str = self.inode.get_modification_time()
+        change_time_str = self.inode.get_change_time()
+
+        inode_user = InodeUser(
+            superblock_addr=superblock_addr,
+            mountpoint=self.mountpoint,
+            device=device,
+            inode_num=inode_num,
+            inode_addr=inode_addr,
+            type=inode_type,
+            inode_pages=inode_pages,
+            cached_pages=cached_pages,
+            file_mode=file_mode,
+            access_time=access_time_dt,
+            modification_time=modification_time_str,
+            change_time=change_time_str,
+            path=self.path,
+        )
+        return inode_user
+
+
+class Files(plugins.PluginInterface, timeliner.TimeLinerInterface):
+    """Lists files from memory"""
+
+    _required_framework_version = (2, 0, 0)
+
+    _version = (1, 0, 0)
+
+    @classmethod
+    def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]:
+        return [
+            requirements.ModuleRequirement(
+                name="kernel",
+                description="Linux kernel",
+                architectures=["Intel32", "Intel64"],
+            ),
+            requirements.PluginRequirement(
+                name="mountinfo", plugin=mountinfo.MountInfo, version=(1, 2, 0)
+            ),
+            requirements.ListRequirement(
+                name="type",
+                description="List of space-separated file type filters i.e. --type REG DIR",
+                element_type=str,
+                optional=True,
+            ),
+            requirements.StringRequirement(
+                name="find",
+                description="Filename (full path) to find",
+                optional=True,
+            ),
+        ]
+
+    @staticmethod
+    def _follow_symlink(
+        inode: interfaces.objects.ObjectInterface,
+        symlink_path: str,
+    ) -> str:
+        """Follows (fast) symlinks (kernels >= 4.2.x).
+        Fast symlinks are filesystem agnostic.
+
+        Args:
+            inode: The inode (or pointer) to dump
+            symlink_path: The symlink name
+
+        Returns:
+            If it can resolve the symlink, it returns a string "symlink_path -> target_path"
+            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:
+            i_link_str = inode.i_link.dereference().cast(
+                "string", max_length=255, encoding="utf-8", errors="replace"
+            )
+            symlink_path = f"{symlink_path} -> {i_link_str}"
+
+        return symlink_path
+
+    @classmethod
+    def _walk_dentry(
+        cls,
+        seen_dentries: Set[int],
+        root_dentry: interfaces.objects.ObjectInterface,
+        parent_dir: str,
+    ):
+        """Walks dentries recursively
+
+        Args:
+            seen_dentries: A set to ensure each dentry is processed only once
+            root_dentry: Root dentry object
+            parent_dir: Parent directory path
+
+        Yields:
+           file_path: Filename including path
+           dentry: Dentry object
+        """
+
+        for dentry in root_dentry.get_subdirs():
+            dentry_addr = dentry.vol.offset
+
+            # corruption
+            if dentry_addr == root_dentry.vol.offset:
+                continue
+
+            if dentry_addr in seen_dentries:
+                continue
+
+            seen_dentries.add(dentry_addr)
+
+            inode = dentry.d_inode
+            if not (inode and inode.is_valid()):
+                continue
+
+            # This allows us to have consistent paths
+            if dentry.d_name.name:
+                basename = dentry.d_name.name_as_str()
+                # Do NOT use os.path.join() below
+                file_path = parent_dir + "/" + basename
+            else:
+                continue
+
+            yield file_path, dentry
+
+            if inode.is_dir:
+                yield from cls._walk_dentry(seen_dentries, dentry, parent_dir=file_path)
+
+    @classmethod
+    def get_inodes(
+        cls,
+        context: interfaces.context.ContextInterface,
+        vmlinux_module_name: str,
+    ) -> Iterable[InodeInternal]:
+        """Retrieves the inodes from the superblocks
+
+        Args:
+            context: The context that the plugin will operate within
+            vmlinux_module_name: The name of the kernel module on which to operate
+
+        Yields:
+            An InodeInternal object
+        """
+
+        superblocks_iter = mountinfo.MountInfo.get_superblocks(
+            context=context,
+            vmlinux_module_name=vmlinux_module_name,
+        )
+
+        seen_inodes = set()
+        seen_dentries = set()
+        for superblock, mountpoint in superblocks_iter:
+            parent_dir = "" if mountpoint == "/" else mountpoint
+
+            # Superblock root dentry
+            root_dentry_ptr = superblock.s_root
+            if not root_dentry_ptr:
+                continue
+
+            root_dentry = root_dentry_ptr.dereference()
+
+            # Dentry sanity check
+            if not root_dentry.is_root():
+                continue
+
+            # More dentry/inode sanity checks
+            root_inode_ptr = root_dentry.d_inode
+            if not root_inode_ptr:
+                continue
+            root_inode = root_inode_ptr.dereference()
+            if not root_inode.is_valid():
+                continue
+
+            # Inode already processed?
+            if root_inode_ptr in seen_inodes:
+                continue
+            seen_inodes.add(root_inode_ptr)
+
+            root_path = mountpoint
+
+            inode_in = InodeInternal(
+                superblock=superblock,
+                mountpoint=mountpoint,
+                inode=root_inode,
+                path=root_path,
+            )
+            yield inode_in
+
+            # Children
+            for file_path, file_dentry in cls._walk_dentry(
+                seen_dentries, root_dentry, parent_dir
+            ):
+                if not file_dentry:
+                    continue
+                # Dentry/inode sanity checks
+                file_inode_ptr = file_dentry.d_inode
+                if not file_inode_ptr:
+                    continue
+                file_inode = file_inode_ptr.dereference()
+                if not file_inode.is_valid():
+                    continue
+
+                # Inode already processed?
+                if file_inode_ptr in seen_inodes:
+                    continue
+                seen_inodes.add(file_inode_ptr)
+
+                file_path = cls._follow_symlink(file_inode_ptr, file_path)
+                inode_in = InodeInternal(
+                    superblock=superblock,
+                    mountpoint=mountpoint,
+                    inode=file_inode,
+                    path=file_path,
+                )
+                yield inode_in
+
+    def _generator(self):
+        vmlinux_module_name = self.config["kernel"]
+        vmlinux = self.context.modules[vmlinux_module_name]
+        vmlinux_layer = self.context.layers[vmlinux.layer_name]
+
+        inodes_iter = self.get_inodes(
+            context=self.context,
+            vmlinux_module_name=vmlinux_module_name,
+        )
+
+        types_filter = self.config["type"]
+        for inode_in in inodes_iter:
+            if types_filter and inode_in.inode.get_inode_type() not in types_filter:
+                continue
+
+            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):
+        """Generates tuples of (description, timestamp_type, timestamp)
+
+        These need not be generated in any particular order, sorting
+        will be done later
+        """
+        vmlinux_module_name = self.config["kernel"]
+        vmlinux = self.context.modules[vmlinux_module_name]
+        vmlinux_layer = self.context.layers[vmlinux.layer_name]
+
+        inodes_iter = self.get_inodes(
+            context=self.context,
+            vmlinux_module_name=vmlinux_module_name,
+        )
+
+        for inode_in in inodes_iter:
+            inode_out = inode_in.to_user(vmlinux_layer)
+            description = f"Cached Inode for {inode_out.path}"
+            yield description, timeliner.TimeLinerType.ACCESSED, inode_out.access_time
+            yield description, timeliner.TimeLinerType.MODIFIED, inode_out.modification_time
+            yield description, timeliner.TimeLinerType.CHANGE, inode_out.change_time
+
+    @staticmethod
+    def format_fields_with_headers(headers, generator):
+        """Uses the headers type to cast the fields obtained from the generator"""
+        for level, fields in generator:
+            formatted_fields = []
+            for header, field in zip(headers, fields):
+                header_type = header[1]
+
+                if isinstance(
+                    field, (header_type, interfaces.renderers.BaseAbsentValue)
+                ):
+                    formatted_field = field
+                else:
+                    formatted_field = header_type(field)
+
+                formatted_fields.append(formatted_field)
+            yield level, formatted_fields
+
+    def run(self):
+        headers = [
+            ("SuperblockAddr", format_hints.Hex),
+            ("MountPoint", str),
+            ("Device", str),
+            ("InodeNum", int),
+            ("InodeAddr", format_hints.Hex),
+            ("FileType", str),
+            ("InodePages", int),
+            ("CachedPages", int),
+            ("FileMode", str),
+            ("AccessTime", datetime.datetime),
+            ("ModificationTime", datetime.datetime),
+            ("ChangeTime", datetime.datetime),
+            ("FilePath", str),
+        ]
+
+        return renderers.TreeGrid(
+            headers, self.format_fields_with_headers(headers, self._generator())
+        )
+
+
+class InodePages(plugins.PluginInterface):
+    """Lists and recovers cached inode pages"""
+
+    _required_framework_version = (2, 0, 0)
+
+    _version = (1, 0, 0)
+
+    @classmethod
+    def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]:
+        return [
+            requirements.ModuleRequirement(
+                name="kernel",
+                description="Linux kernel",
+                architectures=["Intel32", "Intel64"],
+            ),
+            requirements.PluginRequirement(
+                name="files", plugin=Files, version=(1, 0, 0)
+            ),
+            requirements.StringRequirement(
+                name="find",
+                description="Filename (full path) to find ",
+                optional=True,
+            ),
+            requirements.IntRequirement(
+                name="inode",
+                description="Inode address",
+                optional=True,
+            ),
+            requirements.StringRequirement(
+                name="dump",
+                description="Output file path",
+                optional=True,
+            ),
+        ]
+
+    @staticmethod
+    def write_inode_content_to_file(
+        inode: interfaces.objects.ObjectInterface,
+        filename: str,
+        open_method: Type[interfaces.plugins.FileHandlerInterface],
+        vmlinux_layer: interfaces.layers.TranslationLayerInterface,
+    ) -> None:
+        """Extracts the inode's contents from the page cache and saves them to a file
+
+        Args:
+            inode: The inode to dump
+            filename: Filename for writing the inode content
+            open_method: class for constructing output files
+            vmlinux_layer: The kernel layer to obtain the page size
+        """
+        if not inode.is_reg:
+            vollog.error("The inode is not a regular file")
+            return
+
+        # By using truncate/seek, provided the filesystem supports it, a sparse file will be
+        # 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.
+        try:
+            with open_method(filename) as f:
+                inode_size = inode.i_size
+                f.truncate(inode_size)
+
+                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:
+                        vollog.error(
+                            "Page out of file bounds: inode 0x%x, inode size %d, page index %d",
+                            inode.vol.object,
+                            inode_size,
+                            page_idx,
+                        )
+                    f.seek(current_fp)
+                    f.write(page_bytes)
+
+        except IOError as e:
+            vollog.error("Unable to write to file (%s): %s", filename, e)
+
+    def _generator(self):
+        vmlinux_module_name = self.config["kernel"]
+        vmlinux = self.context.modules[vmlinux_module_name]
+        vmlinux_layer = self.context.layers[vmlinux.layer_name]
+
+        if self.config["inode"] and self.config["find"]:
+            vollog.error("Cannot use --inode and --find simultaneously")
+            return
+
+        if self.config["find"]:
+            inodes_iter = Files.get_inodes(
+                context=self.context,
+                vmlinux_module_name=vmlinux_module_name,
+            )
+            for inode_in in inodes_iter:
+                if inode_in.path == self.config["find"]:
+                    inode = inode_in.inode
+                    break  # Only the first match
+
+        elif self.config["inode"]:
+            inode = vmlinux.object("inode", self.config["inode"], absolute=True)
+        else:
+            vollog.error("You must use either --inode or --find")
+            return
+
+        if not inode.is_reg:
+            vollog.error("The inode is not a regular file")
+            return
+
+        inode_size = inode.i_size
+        if not inode.is_valid():
+            vollog.error("Invalid inode at 0x%x", self.config["inode"])
+            return
+
+        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"]:
+            filename = self.config["dump"]
+            vollog.info("[*] Writing inode at 0x%x to '%s'", inode.vol.offset, filename)
+            self.write_inode_content_to_file(inode, filename, self.open, vmlinux_layer)
+
+    def run(self):
+        headers = [
+            ("PageVAddr", format_hints.Hex),
+            ("PagePAddr", format_hints.Hex),
+            ("MappingAddr", format_hints.Hex),
+            ("Index", int),
+            ("DumpSafe", bool),
+            ("Flags", str),
+        ]
+
+        return renderers.TreeGrid(
+            headers, Files.format_fields_with_headers(headers, self._generator())
+        )

volatility3/framework/plugins/linux/pidhashtable.py

@@ -0,0 +1,258 @@
+# 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
+
+from volatility3.framework import renderers, interfaces, constants
+from volatility3.framework.symbols import linux
+from volatility3.framework.renderers import format_hints
+from volatility3.framework.interfaces import plugins
+from volatility3.framework.configuration import requirements
+from volatility3.plugins.linux import pslist
+
+vollog = logging.getLogger(__name__)
+
+
+class PIDHashTable(plugins.PluginInterface):
+    """Enumerates processes through the PID hash table"""
+
+    _required_framework_version = (2, 0, 0)
+
+    _version = (1, 0, 0)
+
+    @classmethod
+    def get_requirements(cls) -> List[interfaces.configuration.RequirementInterface]:
+        return [
+            requirements.ModuleRequirement(
+                name="kernel",
+                description="Linux kernel",
+                architectures=["Intel32", "Intel64"],
+            ),
+            requirements.PluginRequirement(
+                name="pslist", plugin=pslist.PsList, version=(2, 0, 0)
+            ),
+            requirements.VersionRequirement(
+                name="linuxutils", component=linux.LinuxUtilities, version=(2, 1, 0)
+            ),
+            requirements.BooleanRequirement(
+                name="decorate_comm",
+                description="Show `user threads` comm in curly brackets, and `kernel threads` comm in square brackets",
+                optional=True,
+                default=False,
+            ),
+        ]
+
+    def _is_valid_task(self, task) -> bool:
+        vmlinux = self.context.modules[self.config["kernel"]]
+        vmlinux_layer = self.context.layers[vmlinux.layer_name]
+        return bool(task and task.pid > 0 and vmlinux_layer.is_valid(task.parent))
+
+    def _get_pidtype_pid(self):
+        vmlinux = self.context.modules[self.config["kernel"]]
+
+        # The pid_type enumeration is present since 2.5.37, just in case
+        pid_type_enum = vmlinux.get_enumeration("pid_type")
+        if not pid_type_enum:
+            vollog.error("Cannot find pid_type enum. Unsupported kernel")
+            return None
+
+        pidtype_pid = pid_type_enum.choices.get("PIDTYPE_PID")
+        if pidtype_pid is None:
+            vollog.error("Cannot find PIDTYPE_PID. Unsupported kernel")
+            return None
+
+        # Typically PIDTYPE_PID = 0
+        return pidtype_pid
+
+    def _get_pidhash_array(self):
+        vmlinux = self.context.modules[self.config["kernel"]]
+
+        pidhash_shift = vmlinux.object_from_symbol("pidhash_shift")
+        pidhash_size = 1 << pidhash_shift
+
+        array_type_name = vmlinux.symbol_table_name + constants.BANG + "array"
+
+        pidhash_ptr = vmlinux.object_from_symbol("pid_hash")
+        # pidhash is an array of hlist_heads
+        pidhash = self._context.object(
+            array_type_name,
+            offset=pidhash_ptr,
+            subtype=vmlinux.get_type("hlist_head"),
+            count=pidhash_size,
+            layer_name=vmlinux.layer_name,
+        )
+
+        return pidhash
+
+    def _walk_upid(self, seen_upids, upid):
+        vmlinux = self.context.modules[self.config["kernel"]]
+        vmlinux_layer = self.context.layers[vmlinux.layer_name]
+
+        while upid and vmlinux_layer.is_valid(upid.vol.offset):
+            if upid.vol.offset in seen_upids:
+                break
+            seen_upids.add(upid.vol.offset)
+
+            pid_chain = upid.pid_chain
+            if not (pid_chain and vmlinux_layer.is_valid(pid_chain.vol.offset)):
+                break
+
+            upid = linux.LinuxUtilities.container_of(
+                pid_chain.next, "upid", "pid_chain", vmlinux
+            )
+
+    def _get_upids(self):
+        vmlinux = self.context.modules[self.config["kernel"]]
+        vmlinux_layer = self.context.layers[vmlinux.layer_name]
+
+        # 2.6.24 <= kernels < 4.15
+        pidhash = self._get_pidhash_array()
+
+        seen_upids = set()
+        for hlist in pidhash:
+            # each entry in the hlist is a upid which is wrapped in a pid
+            ent = hlist.first
+
+            while ent and vmlinux_layer.is_valid(ent.vol.offset):
+                # upid->pid_chain exists 2.6.24 <= kernel < 4.15
+                upid = linux.LinuxUtilities.container_of(
+                    ent.vol.offset, "upid", "pid_chain", vmlinux
+                )
+
+                if upid.vol.offset in seen_upids:
+                    break
+
+                self._walk_upid(seen_upids, upid)
+
+                ent = ent.next
+
+        return seen_upids
+
+    def _pid_hash_implementation(self):
+        vmlinux = self.context.modules[self.config["kernel"]]
+
+        # 2.6.24 <= kernels < 4.15
+        task_pids_off = vmlinux.get_type("task_struct").relative_child_offset("pids")
+        pidtype_pid = self._get_pidtype_pid()
+
+        for upid in self._get_upids():
+            pid = linux.LinuxUtilities.container_of(upid, "pid", "numbers", vmlinux)
+            if not pid:
+                continue
+
+            pid_tasks_0 = pid.tasks[pidtype_pid].first
+            if not pid_tasks_0:
+                continue
+
+            task = vmlinux.object(
+                "task_struct", offset=pid_tasks_0 - task_pids_off, absolute=True
+            )
+            if self._is_valid_task(task):
+                yield task
+
+    def _task_for_radix_pid_node(self, nodep):
+        vmlinux = self.context.modules[self.config["kernel"]]
+
+        # kernels >= 4.15
+        pid = vmlinux.object("pid", offset=nodep, absolute=True)
+        pidtype_pid = self._get_pidtype_pid()
+
+        pid_tasks_0 = pid.tasks[pidtype_pid].first
+        if not pid_tasks_0:
+            return None
+
+        task_struct_type = vmlinux.get_type("task_struct")
+        if task_struct_type.has_member("pids"):
+            member = "pids"
+        elif task_struct_type.has_member("pid_links"):
+            member = "pid_links"
+        else:
+            return None
+
+        task_pids_off = task_struct_type.relative_child_offset(member)
+        task = vmlinux.object(
+            "task_struct", offset=pid_tasks_0 - task_pids_off, absolute=True
+        )
+        return task
+
+    def _pid_namespace_idr(self):
+        vmlinux = self.context.modules[self.config["kernel"]]
+
+        # kernels >= 4.15
+        ns_addr = vmlinux.get_symbol("init_pid_ns").address
+        ns = vmlinux.object("pid_namespace", offset=ns_addr)
+
+        for page_addr in ns.idr.get_entries():
+            task = self._task_for_radix_pid_node(page_addr)
+            if self._is_valid_task(task):
+                yield task
+
+    def _determine_pid_func(self):
+        vmlinux = self.context.modules[self.config["kernel"]]
+
+        pid_hash = vmlinux.has_symbol("pid_hash") and vmlinux.has_symbol(
+            "pidhash_shift"
+        )  # 2.5.55 <= kernels < 4.15
+
+        has_pid_numbers = vmlinux.has_type("pid") and vmlinux.get_type(
+            "pid"
+        ).has_member(
+            "numbers"
+        )  # kernels >= 2.6.24
+
+        has_pid_chain = vmlinux.has_type("upid") and vmlinux.get_type(
+            "upid"
+        ).has_member(
+            "pid_chain"
+        )  # 2.6.24 <= kernels < 4.15
+
+        # kernels >= 4.15
+        pid_idr = vmlinux.has_type("pid_namespace") and vmlinux.get_type(
+            "pid_namespace"
+        ).has_member("idr")
+
+        if pid_idr:
+            # kernels >= 4.15
+            return self._pid_namespace_idr
+        elif pid_hash and has_pid_numbers and has_pid_numbers and has_pid_chain:
+            # 2.6.24 <= kernels < 4.15
+            return self._pid_hash_implementation
+
+        return None
+
+    def get_tasks(self) -> interfaces.objects.ObjectInterface:
+        """Enumerates processes through the PID hash table
+
+        Yields:
+            task_struct objects
+        """
+        pid_func = self._determine_pid_func()
+        if not pid_func:
+            vollog.error("Cannot determine which PID hash table this kernel is using")
+            return
+
+        yield from sorted(pid_func(), key=lambda t: (t.tgid, t.pid))
+
+    def _generator(
+        self, decorate_comm: bool = False
+    ) -> interfaces.objects.ObjectInterface:
+        for task in self.get_tasks():
+            offset, pid, tid, ppid, name = pslist.PsList.get_task_fields(
+                task, decorate_comm
+            )
+            fields = format_hints.Hex(offset), pid, tid, ppid, name
+            yield 0, fields
+
+    def run(self):
+        decorate_comm = self.config.get("decorate_comm")
+
+        headers = [
+            ("OFFSET", format_hints.Hex),
+            ("PID", int),
+            ("TID", int),
+            ("PPID", int),
+            ("COMM", str),
+        ]
+        return renderers.TreeGrid(headers, self._generator(decorate_comm=decorate_comm))

volatility3/framework/plugins/linux/sockstat.py

@@ -151,17 +151,15 @@ class SockHandlers(interfaces.configuration.VersionableInterface):
 
         bpfprog = sock_filter.prog
 
-        BPF_PROG_TYPE_UNSPEC = 0  # cBPF filter
-        try:
-            bpfprog_type = bpfprog.get_type()
-            if bpfprog_type == BPF_PROG_TYPE_UNSPEC:
-                return  # cBPF filter
-        except AttributeError:
+        bpfprog_type = bpfprog.get_type()
+        if not bpfprog_type:
             # kernel < 3.18.140, it's a cBPF filter
             return None
 
-        BPF_PROG_TYPE_SOCKET_FILTER = 1  # eBPF filter
-        if bpfprog_type != BPF_PROG_TYPE_SOCKET_FILTER:
+        if bpfprog_type == "BPF_PROG_TYPE_UNSPEC":
+            return None  # cBPF filter
+
+        if bpfprog_type != "BPF_PROG_TYPE_SOCKET_FILTER":
             socket_filter["bpf_filter_type"] = f"UNK({bpfprog_type})"
             vollog.warning(f"Unexpected BPF type {bpfprog_type} for a socket")
             return None

volatility3/framework/renderers/conversion.py

@@ -11,6 +11,7 @@ from typing import Union
 from volatility3.framework import interfaces, renderers
 
 
+# FIXME: Move wintime_to_datetime() and unixtime_to_datetime() out of renderers, possibly framework.objects.utility
 def wintime_to_datetime(
     wintime: int,
 ) -> Union[interfaces.renderers.BaseAbsentValue, datetime.datetime]:

volatility3/framework/symbols/linux/__init__.py

@@ -1,6 +1,9 @@
 # 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
 #
+import math
+import contextlib
+from abc import ABC, abstractmethod
 from typing import Iterator, List, Tuple, Optional, Union
 
 from volatility3 import framework
@@ -30,9 +33,13 @@ class LinuxKernelIntermedSymbols(intermed.IntermediateSymbolTable):
         self.set_type_class("kobject", extensions.kobject)
         self.set_type_class("cred", extensions.cred)
         self.set_type_class("inode", extensions.inode)
+        self.set_type_class("idr", extensions.IDR)
+        self.set_type_class("address_space", extensions.address_space)
+        self.set_type_class("page", extensions.page)
         # Might not exist in the current symbols
         self.optional_set_type_class("module", extensions.module)
         self.optional_set_type_class("bpf_prog", extensions.bpf_prog)
+        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)
 
@@ -68,7 +75,7 @@ class LinuxKernelIntermedSymbols(intermed.IntermediateSymbolTable):
 class LinuxUtilities(interfaces.configuration.VersionableInterface):
     """Class with multiple useful linux functions."""
 
-    _version = (2, 1, 0)
+    _version = (2, 1, 1)
     _required_framework_version = (2, 0, 0)
 
     framework.require_interface_version(*_required_framework_version)
@@ -413,9 +420,7 @@ class LinuxUtilities(interfaces.configuration.VersionableInterface):
         Returns:
             A kernel object (vmlinux)
         """
-        symbol_table_arr = volobj.vol.type_name.split("!", 1)
-        symbol_table = symbol_table_arr[0] if len(symbol_table_arr) == 2 else None
-
+        symbol_table = volobj.get_symbol_table_name()
         module_names = context.modules.get_modules_by_symbol_tables(symbol_table)
         module_names = list(module_names)
 
@@ -426,3 +431,353 @@ class LinuxUtilities(interfaces.configuration.VersionableInterface):
         kernel = context.modules[kernel_module_name]
 
         return kernel
+
+
+class IDStorage(ABC):
+    """Abstraction to support both XArray and RadixTree"""
+
+    # Dynamic values, these will be initialized later
+    CHUNK_SHIFT = None
+    CHUNK_SIZE = None
+    CHUNK_MASK = None
+
+    def __init__(
+        self,
+        context: interfaces.context.ContextInterface,
+        kernel_module_name: str,
+    ):
+        self.vmlinux = context.modules[kernel_module_name]
+        self.vmlinux_layer = self.vmlinux.context.layers[self.vmlinux.layer_name]
+
+        self.pointer_size = self.vmlinux.get_type("pointer").size
+        # Dynamically work out the (XA_CHUNK|RADIX_TREE_MAP)_SHIFT values based on
+        # the node.slots[] array size
+        node_type = self.vmlinux.get_type(self.node_type_name)
+        slots_array_size = node_type.child_template("slots").count
+
+        # Calculate the LSB index - 1
+        self.CHUNK_SHIFT = slots_array_size.bit_length() - 1
+        self.CHUNK_SIZE = 1 << self.CHUNK_SHIFT
+        self.CHUNK_MASK = self.CHUNK_SIZE - 1
+
+    @classmethod
+    def choose_id_storage(
+        cls,
+        context: interfaces.context.ContextInterface,
+        kernel_module_name: str,
+    ) -> "IDStorage":
+        """Returns the appropriate ID storage data structure instance for the current kernel implementation.
+        This is used by the IDR and the PageCache to choose between the XArray and RadixTree.
+
+        Args:
+            context: The context to retrieve required elements (layers, symbol tables) from
+            kernel_module_name: The name of the kernel module on which to operate
+
+        Returns:
+            The appropriate ID storage instance for the current kernel
+        """
+        vmlinux = context.modules[kernel_module_name]
+        address_space_type = vmlinux.get_type("address_space")
+        address_space_has_i_pages = address_space_type.has_member("i_pages")
+        i_pages_type_name = (
+            address_space_type.child_template("i_pages").vol.type_name
+            if address_space_has_i_pages
+            else ""
+        )
+        i_pages_is_xarray = i_pages_type_name.endswith(constants.BANG + "xarray")
+        i_pages_is_radix_tree_root = i_pages_type_name.endswith(
+            constants.BANG + "radix_tree_root"
+        ) and vmlinux.get_type("radix_tree_root").has_member("xa_head")
+
+        if i_pages_is_xarray or i_pages_is_radix_tree_root:
+            return XArray(context, kernel_module_name)
+        else:
+            return RadixTree(context, kernel_module_name)
+
+    @property
+    @abstractmethod
+    def node_type_name(self) -> str:
+        """Returns the Tree implementation node type name
+
+        Returns:
+            A string with the node type name
+        """
+        raise NotImplementedError()
+
+    @property
+    def tag_internal_value(self) -> int:
+        """Returns the internal node flag for the tree"""
+        raise NotImplementedError()
+
+    @abstractmethod
+    def node_is_internal(self, nodep) -> bool:
+        """Checks if the node is internal"""
+        raise NotImplementedError
+
+    @abstractmethod
+    def is_node_tagged(self, nodep) -> bool:
+        """Checks if the node pointer is tagged"""
+        raise NotImplementedError
+
+    @abstractmethod
+    def untag_node(self, nodep) -> int:
+        """Untags a node pointer"""
+        raise NotImplementedError
+
+    @abstractmethod
+    def get_tree_height(self, treep) -> int:
+        """Returns the tree height"""
+        raise NotImplementedError
+
+    @abstractmethod
+    def get_node_height(self, nodep) -> int:
+        """Returns the node height"""
+        raise NotImplementedError
+
+    @abstractmethod
+    def get_head_node(self, tree) -> int:
+        """Returns a pointer to the tree's head"""
+        raise NotImplementedError
+
+    @abstractmethod
+    def is_valid_node(self, nodep) -> bool:
+        """Validates a node pointer"""
+        raise NotImplementedError
+
+    def nodep_to_node(self, nodep) -> interfaces.objects.ObjectInterface:
+        """Instanciates a tree node from its pointer
+
+        Args:
+            nodep: Pointer to the XArray/RadixTree node
+
+        Returns:
+            A XArray/RadixTree node instance
+        """
+        node = self.vmlinux.object(self.node_type_name, offset=nodep, absolute=True)
+        return node
+
+    def _slot_to_nodep(self, slot) -> int:
+        if self.node_is_internal(slot):
+            nodep = slot & ~self.tag_internal_value
+        else:
+            nodep = slot
+
+        return nodep
+
+    def _iter_node(self, nodep, height) -> int:
+        node = self.nodep_to_node(nodep)
+        node_slots = node.slots
+        for off in range(self.CHUNK_SIZE):
+            slot = node_slots[off]
+            if slot == 0:
+                continue
+
+            nodep = self._slot_to_nodep(slot)
+
+            if height == 1:
+                if self.is_valid_node(nodep):
+                    yield nodep
+            else:
+                for child_node in self._iter_node(nodep, height - 1):
+                    yield child_node
+
+    def get_entries(self, root: interfaces.objects.ObjectInterface) -> int:
+        """Walks the tree data structure
+
+        Args:
+            root: The tree root object
+
+        Yields:
+            A tree node pointer
+        """
+        height = self.get_tree_height(root.vol.offset)
+
+        nodep = self.get_head_node(root)
+        if not nodep:
+            return
+
+        # Keep the internal flag before untagging it
+        is_internal = self.node_is_internal(nodep)
+        if self.is_node_tagged(nodep):
+            nodep = self.untag_node(nodep)
+
+        if is_internal:
+            height = self.get_node_height(nodep)
+
+        if height == 0:
+            if self.is_valid_node(nodep):
+                yield nodep
+        else:
+            for child_node in self._iter_node(nodep, height):
+                yield child_node
+
+
+class XArray(IDStorage):
+    XARRAY_TAG_MASK = 3
+    XARRAY_TAG_INTERNAL = 2
+
+    def get_tree_height(self, treep) -> int:
+        return 0
+
+    @property
+    def node_type_name(self) -> str:
+        return "xa_node"
+
+    @property
+    def tag_internal_value(self) -> int:
+        return self.XARRAY_TAG_INTERNAL
+
+    def get_node_height(self, nodep) -> int:
+        node = self.nodep_to_node(nodep)
+        return (node.shift / self.CHUNK_SHIFT) + 1
+
+    def get_head_node(self, tree) -> int:
+        return tree.xa_head
+
+    def node_is_internal(self, nodep) -> bool:
+        return (nodep & self.XARRAY_TAG_MASK) == self.XARRAY_TAG_INTERNAL
+
+    def is_node_tagged(self, nodep) -> bool:
+        return (nodep & self.XARRAY_TAG_MASK) != 0
+
+    def untag_node(self, nodep) -> int:
+        return nodep & (~self.XARRAY_TAG_MASK)
+
+    def is_valid_node(self, nodep) -> bool:
+        # It should have the tag mask clear
+        return not self.is_node_tagged(nodep)
+
+
+class RadixTree(IDStorage):
+    RADIX_TREE_INTERNAL_NODE = 1
+    RADIX_TREE_EXCEPTIONAL_ENTRY = 2
+    RADIX_TREE_ENTRY_MASK = 3
+
+    # Dynamic values. These will be initialized later
+    RADIX_TREE_INDEX_BITS = None
+    RADIX_TREE_MAX_PATH = None
+    RADIX_TREE_HEIGHT_SHIFT = None
+    RADIX_TREE_HEIGHT_MASK = None
+
+    def __init__(self, *args, **kwargs) -> None:
+        super().__init__(*args, **kwargs)
+
+        char_bits = 8
+        self.RADIX_TREE_INDEX_BITS = char_bits * self.pointer_size
+        self.RADIX_TREE_MAX_PATH = int(
+            math.ceil(self.RADIX_TREE_INDEX_BITS / float(self.CHUNK_SHIFT))
+        )
+        self.RADIX_TREE_HEIGHT_SHIFT = self.RADIX_TREE_MAX_PATH + 1
+        self.RADIX_TREE_HEIGHT_MASK = (1 << self.RADIX_TREE_HEIGHT_SHIFT) - 1
+
+        if not self.vmlinux.has_type("radix_tree_root"):
+            # In kernels 4.20, RADIX_TREE_INTERNAL_NODE flag took RADIX_TREE_EXCEPTIONAL_ENTRY's
+            # value. RADIX_TREE_EXCEPTIONAL_ENTRY was removed but that's managed in is_valid_node()
+            # Note that the Radix Tree is still in use for IDR, even after kernels 4.20 when XArray
+            # mostly replace it
+            self.RADIX_TREE_INTERNAL_NODE = 2
+
+    @property
+    def node_type_name(self) -> str:
+        return "radix_tree_node"
+
+    @property
+    def tag_internal_value(self) -> int:
+        return self.RADIX_TREE_INTERNAL_NODE
+
+    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
+                radix_tree_root = self.vmlinux.object(
+                    "radix_tree_root", offset=treep, absolute=True
+                )
+                return radix_tree_root.height
+
+        # kernels >= 4.7.10
+        return 0
+
+    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
+        else:
+            # Kernels < 4.7
+            height_to_maxindex_array = self.vmlinux.object_from_symbol(
+                "height_to_maxindex"
+            )
+            maxindex = height_to_maxindex_array[height]
+            return maxindex
+
+    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
+        elif hasattr(node, "path"):
+            # 3.15 <= Kernels < 4.7
+            return node.path & self.RADIX_TREE_HEIGHT_MASK
+        elif hasattr(node, "height"):
+            # Kernels < 3.15
+            return node.height
+        else:
+            raise exceptions.VolatilityException("Cannot find radix-tree node height")
+
+    def get_head_node(self, tree) -> int:
+        return tree.rnode
+
+    def node_is_internal(self, nodep) -> bool:
+        return (nodep & self.RADIX_TREE_INTERNAL_NODE) != 0
+
+    def is_node_tagged(self, nodep) -> bool:
+        return self.node_is_internal(nodep)
+
+    def untag_node(self, nodep) -> int:
+        return nodep & (~self.RADIX_TREE_ENTRY_MASK)
+
+    def is_valid_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 True
+
+
+class PageCache(object):
+    """Linux Page Cache abstraction"""
+
+    def __init__(
+        self,
+        context: interfaces.context.ContextInterface,
+        kernel_module_name: str,
+        page_cache: interfaces.objects.ObjectInterface,
+    ):
+        """
+        Args:
+            context: interfaces.context.ContextInterface,
+            kernel_module_name: The name of the kernel module on which to operate
+            page_cache: Page cache address space
+        """
+        self.vmlinux = context.modules[kernel_module_name]
+
+        self._page_cache = page_cache
+        self._idstorage = IDStorage.choose_id_storage(context, kernel_module_name)
+
+    def get_cached_pages(self) -> interfaces.objects.ObjectInterface:
+        """Returns all page cache contents
+
+        Yields:
+            Page objects
+        """
+
+        for page_addr in self._idstorage.get_entries(self._page_cache.i_pages):
+            if not page_addr:
+                continue
+
+            page = self.vmlinux.object("page", offset=page_addr, absolute=True)
+            if page:
+                yield page

volatility3/framework/symbols/linux/extensions/__init__.py

@@ -4,13 +4,16 @@
 
 import collections.abc
 import logging
+import functools
+import binascii
 import stat
 from datetime import datetime
 import socket as socket_module
-from typing import Generator, Iterable, Iterator, Optional, Tuple, List, Union
+from typing import Generator, Iterable, Iterator, Optional, Tuple, List, Union, Dict
 
 from volatility3.framework import constants, exceptions, objects, interfaces, symbols
 from volatility3.framework.renderers import conversion
+from volatility3.framework.configuration import requirements
 from volatility3.framework.constants.linux import SOCK_TYPES, SOCK_FAMILY
 from volatility3.framework.constants.linux import IP_PROTOCOLS, IPV6_PROTOCOLS
 from volatility3.framework.constants.linux import TCP_STATES, NETLINK_PROTOCOLS
@@ -47,7 +50,7 @@ class module(generic.GenericIntelProcess):
                 ).choices
             except exceptions.SymbolError:
                 vollog.debug(
-                    f"Unable to find mod_mem_type enum. This message can be ignored for kernels < 6.4"
+                    "Unable to find mod_mem_type enum. This message can be ignored for kernels < 6.4"
                 )
                 # set to empty dict to show that the enum was not found, and so shouldn't be searched for again
                 self._mod_mem_type = {}
@@ -820,6 +823,26 @@ class dentry(objects.StructType):
             current_dentry = current_dentry.d_parent
         return None
 
+    def get_subdirs(self) -> interfaces.objects.ObjectInterface:
+        """Walks dentry subdirs
+
+        Yields:
+            A dentry object
+        """
+        if self.has_member("d_sib") and self.has_member("d_children"):
+            # kernels >= 6.8
+            walk_member = "d_sib"
+            list_head_member = self.d_children.first
+        elif self.has_member("d_child") and self.has_member("d_subdirs"):
+            # 2.5.0 <= kernels < 6.8
+            walk_member = "d_child"
+            list_head_member = self.d_subdirs
+        else:
+            raise exceptions.VolatilityException("Unsupported dentry type")
+
+        dentry_type_name = self.get_symbol_table_name() + constants.BANG + "dentry"
+        yield from list_head_member.to_list(dentry_type_name, walk_member)
+
 
 class struct_file(objects.StructType):
     def get_dentry(self) -> interfaces.objects.ObjectInterface:
@@ -936,7 +959,8 @@ class mount(objects.StructType):
         MNT_RELATIME: "relatime",
     }
 
-    def get_mnt_sb(self):
+    def get_mnt_sb(self) -> int:
+        """Returns a pointer to the super_block"""
         if self.has_member("mnt"):
             return self.mnt.mnt_sb
         elif self.has_member("mnt_sb"):
@@ -1251,6 +1275,7 @@ class vfsmount(objects.StructType):
             return self._get_real_mnt().has_parent()
 
     def get_mnt_sb(self):
+        """Returns a pointer to the super_block"""
         return self.mnt_sb
 
     def get_flags_access(self) -> str:
@@ -1612,20 +1637,72 @@ class xdp_sock(objects.StructType):
 
 
 class bpf_prog(objects.StructType):
-    def get_type(self):
+    def _get_vmlinux(self):
+        linuxutils_required_version = (2, 1, 1)
+        linuxutils_current_version = linux.LinuxUtilities._version
+        if not requirements.VersionRequirement.matches_required(
+            linuxutils_required_version, linuxutils_current_version
+        ):
+            raise exceptions.PluginRequirementException(
+                f"linux.LinuxUtilities version not suitable: required {linuxutils_required_version} found {linuxutils_current_version}"
+            )
+
+        vmlinux = linux.LinuxUtilities.get_module_from_volobj_type(self._context, self)
+        return vmlinux
+
+    def get_type(self) -> Union[str, None]:
+        """Returns a string with the eBPF program type"""
+
         # The program type was in `bpf_prog_aux::prog_type` from 3.18.140 to
         # 4.1.52 before it was moved to `bpf_prog::type`
         if self.has_member("type"):
             # kernel >= 4.1.52
-            return self.type
+            return self.type.description
 
         if self.has_member("aux") and self.aux:
             if self.aux.has_member("prog_type"):
                 # 3.18.140 <= kernel < 4.1.52
-                return self.aux.prog_type
+                return self.aux.prog_type.description
 
         # kernel < 3.18.140
-        raise AttributeError("Unable to find the BPF type")
+        return None
+
+    def get_tag(self) -> Union[str, None]:
+        """Returns a string with the eBPF program tag"""
+        # 'tag' was added in kernels 4.10
+        if not self.has_member("tag"):
+            return None
+
+        vmlinux = self._get_vmlinux()
+        vmlinux_layer = vmlinux.context.layers[vmlinux.layer_name]
+
+        prog_tag_addr = self.tag.vol.offset
+        prog_tag_size = self.tag.count
+        prog_tag_bytes = vmlinux_layer.read(prog_tag_addr, prog_tag_size)
+
+        prog_tag = binascii.hexlify(prog_tag_bytes).decode()
+        return prog_tag
+
+    def get_name(self) -> Union[str, None]:
+        """Returns a string with the eBPF program name"""
+        if not self.has_member("aux"):
+            # 'prog_aux' was added in kernels 3.18
+            return None
+
+        return self.aux.get_name()
+
+
+class bpf_prog_aux(objects.StructType):
+    def get_name(self) -> Union[str, None]:
+        """Returns a string with the eBPF program name"""
+        if not self.has_member("name"):
+            # 'name' was added in kernels 4.15
+            return None
+
+        if not self.name:
+            return None
+
+        return utility.array_to_string(self.name)
 
 
 class cred(objects.StructType):
@@ -1926,6 +2003,279 @@ class inode(objects.StructType):
         """
         return stat.filemode(self.i_mode)
 
+    def get_pages(self) -> interfaces.objects.ObjectInterface:
+        """Gets the inode's cached pages
+
+        Yields:
+            The inode's cached pages
+        """
+        if not self.i_size:
+            return
+        elif not (self.i_mapping and self.i_mapping.nrpages > 0):
+            return
+
+        page_cache = linux.PageCache(
+            context=self._context,
+            kernel_module_name="kernel",
+            page_cache=self.i_mapping.dereference(),
+        )
+        yield from page_cache.get_cached_pages()
+
+    def get_contents(self):
+        """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
+        """
+        for page_obj in self.get_pages():
+            page_index = int(page_obj.index)
+            page_content = page_obj.get_content()
+            yield page_index, page_content
+
+
+class address_space(objects.StructType):
+    @property
+    def i_pages(self):
+        """Returns the appropriate member containing the page cache tree"""
+        if self.has_member("i_pages"):
+            # Kernel >= 4.17
+            return self.member("i_pages")
+        elif self.has_member("page_tree"):
+            # Kernel < 4.17
+            return self.member("page_tree")
+
+        raise exceptions.VolatilityException("Unsupported page cache tree")
+
+
+class page(objects.StructType):
+    @property
+    @functools.lru_cache()
+    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"
+            ).choices
+        except exceptions.SymbolError:
+            vollog.debug(
+                "Unable to find pageflags enum. This can happen in kernels < 2.6.26 or wrong ISF"
+            )
+            # set to empty dict to show that the enum was not found, and so shouldn't be searched for again
+            pageflags_enum = {}
+
+        return pageflags_enum
+
+    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
+
+    def _get_vmlinux(self):
+        linuxutils_required_version = (2, 1, 1)
+        linuxutils_current_version = linux.LinuxUtilities._version
+        if not requirements.VersionRequirement.matches_required(
+            linuxutils_required_version, linuxutils_current_version
+        ):
+            raise exceptions.PluginRequirementException(
+                f"linux.LinuxUtilities version not suitable: required {linuxutils_required_version} found {linuxutils_current_version}"
+            )
+
+        vmlinux = linux.LinuxUtilities.get_module_from_volobj_type(self._context, self)
+        return vmlinux
+
+    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
+        """
+        vmlinux = self._get_vmlinux()
+        vmlinux_layer = vmlinux.context.layers[vmlinux.layer_name]
+
+        vmemmap_start = None
+        if vmlinux.has_symbol("mem_section"):
+            # SPARSEMEM_VMEMMAP physical memory model: memmap is virtually contiguous
+            if vmlinux.has_symbol("vmemmap_base"):
+                # CONFIG_DYNAMIC_MEMORY_LAYOUT - KASLR kernels >= 4.9
+                vmemmap_start = vmlinux.object_from_symbol("vmemmap_base")
+            else:
+                # !CONFIG_DYNAMIC_MEMORY_LAYOUT
+                if vmlinux_layer._maxvirtaddr < 57:
+                    # 4-Level paging -> VMEMMAP_START = __VMEMMAP_BASE_L4
+                    vmemmap_base_l4 = 0xFFFFEA0000000000
+                    vmemmap_start = vmemmap_base_l4
+                else:
+                    # 5-Level paging -> VMEMMAP_START = __VMEMMAP_BASE_L5
+                    # FIXME: Once 5-level paging is supported, uncomment the following lines and remove the exception
+                    # vmemmap_base_l5 = 0xFFD4000000000000
+                    # vmemmap_start = vmemmap_base_l5
+                    raise exceptions.VolatilityException(
+                        "5-level paging is not yet supported"
+                    )
+
+        elif vmlinux.has_symbol("mem_map"):
+            # FLATMEM physical memory model, typically 32bit
+            vmemmap_start = vmlinux.object_from_symbol("mem_map")
+
+        elif vmlinux.has_symbol("node_data"):
+            raise exceptions.VolatilityException("NUMA systems are not yet supported")
+        else:
+            raise exceptions.VolatilityException("Unsupported Linux memory model")
+
+        if not vmemmap_start:
+            raise exceptions.VolatilityException(
+                "Something went wrong, we shouldn't be here"
+            )
+
+        page_type_size = vmlinux.get_type("page").size
+        pagec = vmlinux_layer.canonicalize(self.vol.offset)
+        pfn = (pagec - vmemmap_start) // page_type_size
+        page_paddr = pfn * vmlinux_layer.page_size
+
+        return page_paddr
+
+    def get_content(self) -> Union[str, None]:
+        """Returns the page content
+
+        Returns:
+            The page content
+        """
+        vmlinux = self._get_vmlinux()
+        vmlinux_layer = vmlinux.context.layers[vmlinux.layer_name]
+        physical_layer = vmlinux.context.layers["memory_layer"]
+        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
+
+
+class IDR(objects.StructType):
+    IDR_BITS = 8
+    IDR_MASK = (1 << IDR_BITS) - 1
+    INT_SIZE = 4
+    MAX_IDR_SHIFT = INT_SIZE * 8 - 1
+    MAX_IDR_BIT = 1 << MAX_IDR_SHIFT
+
+    def _get_vmlinux(self):
+        linuxutils_required_version = (2, 1, 1)
+        linuxutils_current_version = linux.LinuxUtilities._version
+        if not requirements.VersionRequirement.matches_required(
+            linuxutils_required_version, linuxutils_current_version
+        ):
+            raise exceptions.PluginRequirementException(
+                f"linux.LinuxUtilities version not suitable: required {linuxutils_required_version} found {linuxutils_current_version}"
+            )
+
+        vmlinux = linux.LinuxUtilities.get_module_from_volobj_type(self._context, self)
+        return vmlinux
+
+    def idr_max(self, num_layers: int) -> int:
+        """Returns the maximum ID which can be allocated given idr::layers
+
+        Args:
+            num_layers: Number of layers
+
+        Returns:
+            Maximum ID for a given number of layers
+        """
+        # Kernel < 4.17
+        bits = min([self.INT_SIZE, num_layers * self.IDR_BITS, self.MAX_IDR_SHIFT])
+
+        return (1 << bits) - 1
+
+    def idr_find(self, idr_id: int) -> int:
+        """Finds an ID within the IDR data structure.
+        Based on idr_find_slowpath(), 3.9 <= Kernel < 4.11
+        Args:
+            idr_id: The IDR lookup ID
+
+        Returns:
+            A pointer to the given ID element
+        """
+        vmlinux = self._get_vmlinux()
+        if not vmlinux.get_type("idr_layer").has_member("layer"):
+            vollog.info(
+                "Unsupported IDR implementation, it should be a very very old kernel, probabably < 2.6"
+            )
+            return None
+
+        if idr_id < 0:
+            return None
+
+        idr_layer = self.top
+        if not idr_layer:
+            return None
+
+        n = (idr_layer.layer + 1) * self.IDR_BITS
+
+        if idr_id > self.idr_max(idr_layer.layer + 1):
+            return None
+
+        assert n != 0
+
+        while n > 0 and idr_layer:
+            n -= self.IDR_BITS
+            assert n == idr_layer.layer * self.IDR_BITS
+            idr_layer = idr_layer.ary[(idr_id >> n) & self.IDR_MASK]
+
+        return idr_layer
+
+    def _old_kernel_get_entries(self) -> int:
+        # Kernels < 4.11
+        cur = self.cur
+        total = next_id = 0
+        while next_id < cur:
+            entry = self.idr_find(next_id)
+            if entry:
+                yield entry
+                total += 1
+
+            next_id += 1
+
+    def _new_kernel_get_entries(self) -> int:
+        # Kernels >= 4.11
+        id_storage = linux.IDStorage.choose_id_storage(
+            self._context, kernel_module_name="kernel"
+        )
+        for page_addr in id_storage.get_entries(root=self.idr_rt):
+            yield page_addr
+
+    def get_entries(self) -> int:
+        """Walks the IDR and yield a pointer associated with each element.
+
+        Args:
+            in_use (int, optional): _description_. Defaults to 0.
+
+        Yields:
+            A pointer associated with each element.
+        """
+        if self.has_member("idr_rt"):
+            # Kernels >= 4.11
+            get_entries_func = self._new_kernel_get_entries
+        else:
+            # Kernels < 4.11
+            get_entries_func = self._old_kernel_get_entries
+
+        for page_addr in get_entries_func():
+            yield page_addr
+
 
 class rb_root(objects.StructType):
     def _walk_nodes(self, root_node) -> Iterator[int]: