dissect.volume 3.12.dev2__tar.gz → 3.13.dev1__tar.gz

{dissect_volume-3.12.dev2 → dissect_volume-3.13.dev1}/.gitattributes

@@ -2,3 +2,4 @@ tests/data/dm/* filter=lfs diff=lfs merge=lfs -text
 tests/data/md/* filter=lfs diff=lfs merge=lfs -text
 tests/data/ddf/* filter=lfs diff=lfs merge=lfs -text
 tests/data/lvm/* filter=lfs diff=lfs merge=lfs -text
+tests/data/vinum/* filter=lfs diff=lfs merge=lfs -text

{dissect_volume-3.12.dev2 → dissect_volume-3.13.dev1}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: dissect.volume
-Version: 3.12.dev2
+Version: 3.13.dev1
 Summary: A Dissect module implementing a parser for different disk volume and partition systems, for example LVM2, GPT and MBR
 Author-email: Dissect Team <dissect@fox-it.com>
 License: Affero General Public License v3

{dissect_volume-3.12.dev2 → dissect_volume-3.13.dev1}/dissect/volume/ddf/ddf.py

@@ -1,43 +1,51 @@
 from __future__ import annotations
 
 import io
-from typing import BinaryIO, Union
+from typing import TYPE_CHECKING, BinaryIO
 
 from dissect.util import ts
 
 from dissect.volume.ddf.c_ddf import DEFAULT_SECTOR_SIZE, c_ddf
 from dissect.volume.exceptions import DDFError
-from dissect.volume.raid.raid import RAID, Configuration, PhysicalDisk, VirtualDisk
+from dissect.volume.raid.raid import (
+    RAID,
+    Configuration,
+    DiskMap,
+    PhysicalDisk,
+    VirtualDisk,
+)
 from dissect.volume.raid.stream import Layout, Level
 
+if TYPE_CHECKING:
+    DDFPhysicalDiskDescriptor = BinaryIO | "DDFPhysicalDisk"
+
 DECADE = 3600 * 24 * (365 * 10 + 2)
 
 
 class DDF(RAID):
-    def __init__(self, fh: list[Union[BinaryIO, DDFPhysicalDisk]], sector_size: int = DEFAULT_SECTOR_SIZE):
+    def __init__(
+        self,
+        fh: list[DDFPhysicalDiskDescriptor] | DDFPhysicalDiskDescriptor,
+        sector_size: int = DEFAULT_SECTOR_SIZE,
+    ):
         fhs = [fh] if not isinstance(fh, list) else fh
-        self.disks = [DDFPhysicalDisk(f, sector_size) if not isinstance(f, DDFPhysicalDisk) else f for f in fhs]
-        self.sector_size = sector_size
+        physical_disks = [DDFPhysicalDisk(f, sector_size) if not isinstance(f, DDFPhysicalDisk) else f for f in fhs]
 
         config_map = {}
-        for pd in self.disks:
+        for pd in physical_disks:
             config_map.setdefault(pd.anchor.DDF_Header_GUID, []).append(pd)
 
         super().__init__([DDFConfiguration(disks) for disks in config_map.values()])
 
 
 class DDFConfiguration(Configuration):
-    def __init__(self, fh: list[Union[BinaryIO, PhysicalDisk]], sector_size: int = DEFAULT_SECTOR_SIZE):
-        fhs = [fh] if not isinstance(fh, list) else fh
-        self.disks = [DDFPhysicalDisk(f, sector_size) if not isinstance(f, DDFPhysicalDisk) else f for f in fhs]
-        self.sector_size = sector_size
-
+    def __init__(self, physical_disks: list[DDFPhysicalDisk]):
         pd_map: dict[int, DDFPhysicalDisk] = {}
         vde_map: dict[bytes, VirtualDiskRecord] = {}
         vdcr_map: dict[bytes, VirtualDiskConfigurationRecord] = {}
         vdcr_uniq: dict[tuple[bytes, int], VirtualDiskConfigurationRecord] = {}
 
-        for pd in self.disks:
+        for pd in physical_disks:
             pd_map[pd.reference] = pd
             vde_map.update({vde.guid: vde for vde in pd.virtual_disk_records})
             vdcr_map.update({vdcr.guid: vdcr for vdcr in pd.virtual_disk_configuration_records})
@@ -61,7 +69,7 @@ class DDFConfiguration(Configuration):
                     i += 1
 
         virtual_disks = [DDFVirtualDisk(vdcr_map[guid], vde_map[guid], vd_map[guid]) for guid in vd_map.keys()]
-        super().__init__(self.disks, virtual_disks)
+        super().__init__(physical_disks, virtual_disks)
 
 
 class DDFVirtualDisk(VirtualDisk):
@@ -69,14 +77,14 @@ class DDFVirtualDisk(VirtualDisk):
         self,
         vdcr: VirtualDiskConfigurationRecord,
         vdr: VirtualDiskRecord,
-        disks: dict[int, tuple[int, DDFPhysicalDisk]],
+        disks: DiskMap,
     ):
         self.vdcr = vdcr
         self.vdr = vdr
         self.disks = disks
 
         if (block_size := self.vdcr.block_size) == 0xFFFF:
-            block_size = list(self.disks.values())[0][1].block_size
+            block_size = list(disks.values())[0][1].block_size
 
         level, layout, num_disks = _convert_raid_layout(
             vdcr.primary_raid_level,
@@ -218,7 +226,7 @@ class VirtualDiskRecord:
         self.state = self.header.VD_State
         self.init_state = self.header.Init_State
         name = self.header.VD_Name.split(b"\x00")[0]
-        self.name = name.decode("utf-8") if self.type & 0x02 else name.decode()
+        self.name = name.decode(encoding="utf-8") if self.type & 0x02 else name.decode(encoding="ascii")
 
     def __repr__(self) -> str:
         return f"<VirtualDiskRecord guid={self.guid} number={self.number} type={self.type:#x} name={self.name!r}>"

dissect_volume-3.13.dev1/dissect/volume/md/__init__.py

@@ -0,0 +1,6 @@
+from dissect.volume.md.md import MD, MDPhysicalDisk
+
+__all__ = [
+    "MD",
+    "MDPhysicalDisk",
+]

{dissect_volume-3.12.dev2 → dissect_volume-3.13.dev1}/dissect/volume/md/md.py

@@ -4,7 +4,7 @@ import datetime
 import io
 import operator
 import struct
-from typing import BinaryIO, Union
+from typing import TYPE_CHECKING, BinaryIO
 from uuid import UUID
 
 from dissect.util import ts
@@ -13,7 +13,8 @@ from dissect.volume.md.c_md import SECTOR_SIZE, c_md
 from dissect.volume.raid.raid import RAID, Configuration, PhysicalDisk, VirtualDisk
 from dissect.volume.raid.stream import Level
 
-DeviceDescriptor = Union[BinaryIO, "Device"]
+if TYPE_CHECKING:
+    MDPhysicalDiskDescriptor = BinaryIO | "MDPhysicalDisk"
 
 
 class MD(RAID):
@@ -22,62 +23,63 @@ class MD(RAID):
     Use this class to read from a RAID set.
 
     Args:
-        fh: A single file-like object or :class:`Device`, or a list of multiple belonging to the same RAID set.
+        fh: A single file-like object or :class:`MDPhysicalDisk`, or a list of multiple belonging to the same RAID set.
     """
 
-    def __init__(self, fh: Union[list[DeviceDescriptor], DeviceDescriptor]):
+    def __init__(self, fh: list[MDPhysicalDiskDescriptor] | MDPhysicalDiskDescriptor):
         fhs = [fh] if not isinstance(fh, list) else fh
-        self.devices = [Device(fh) if not isinstance(fh, Device) else fh for fh in fhs]
+        physical_disks = [MDPhysicalDisk(fh) if not isinstance(fh, MDPhysicalDisk) else fh for fh in fhs]
 
         config_map = {}
-        for dev in self.devices:
-            config_map.setdefault(dev.set_uuid, []).append(dev)
+        for disk in physical_disks:
+            config_map.setdefault(disk.set_uuid, []).append(disk)
 
-        super().__init__([MDConfiguration(devices) for devices in config_map.values()])
+        super().__init__([MDConfiguration(disks) for disks in config_map.values()])
 
 
 class MDConfiguration(Configuration):
-    def __init__(self, devices: list[DeviceDescriptor]):
-        devices = [Device(fh) if not isinstance(fh, Device) else fh for fh in devices]
+    def __init__(self, physical_disks: list[MDPhysicalDisk]):
+        physical_disks = sorted(physical_disks, key=operator.attrgetter("raid_disk"))
 
-        self.devices = sorted(devices, key=operator.attrgetter("raid_disk"))
-        if len({dev.set_uuid for dev in self.devices}) != 1:
-            raise ValueError("Multiple MD sets detected, supply only the devices of a single set")
+        if len({disk.set_uuid for disk in physical_disks}) != 1:
+            raise ValueError("Multiple MD sets detected, supply only the disks of a single set")
 
-        virtual_disk = MDDisk(self)
-        super().__init__(self.devices, [virtual_disk])
+        virtual_disks = [MDVirtualDisk(physical_disks)]
+        super().__init__(physical_disks, virtual_disks)
 
 
-class MDDisk(VirtualDisk):
-    def __init__(self, configuration: MDConfiguration):
-        self.configuration = configuration
-        reference_dev = sorted(configuration.devices, key=operator.attrgetter("events"), reverse=True)[0]
-        disks = {dev.raid_disk: (0, dev) for dev in self.configuration.devices if dev.raid_disk is not None}
+class MDVirtualDisk(VirtualDisk):
+    def __init__(self, physical_disks: list[MDPhysicalDisk]):
+        reference_disk = sorted(physical_disks, key=operator.attrgetter("events"), reverse=True)[0]
+        disk_map = {disk.raid_disk: (0, disk) for disk in physical_disks if disk.raid_disk is not None}
 
-        if reference_dev.level == Level.LINEAR:
-            size = sum(disk.size for _, disk in disks.values())
-        elif reference_dev.level == Level.RAID0:
+        if reference_disk.level == Level.LINEAR:
+            size = sum(disk.size for _, disk in disk_map.values())
+        elif reference_disk.level == Level.RAID0:
             size = 0
-            for _, disk in disks.values():
-                size += disk.size & ~(reference_dev.chunk_size - 1)
-        elif reference_dev.level in (Level.RAID1, Level.RAID4, Level.RAID5, Level.RAID6, Level.RAID10):
-            size = reference_dev.sb.size * SECTOR_SIZE
+            for _, disk in disk_map.values():
+                size += disk.size & ~(reference_disk.chunk_size - 1)
+        elif reference_disk.level in (Level.RAID1, Level.RAID4, Level.RAID5, Level.RAID6, Level.RAID10):
+            size = reference_disk.sb.size * SECTOR_SIZE
         else:
-            raise ValueError("Invalid MD RAID configuration: No valid RAID level found for the reference disk")
+            raise ValueError(
+                "Invalid MD RAID configuration: No valid RAID level found for the reference disk, found: %d",
+                reference_disk.level,
+            )
 
         super().__init__(
-            reference_dev.set_name,
-            reference_dev.set_uuid,
+            reference_disk.set_name,
+            reference_disk.set_uuid,
             size,
-            reference_dev.level,
-            reference_dev.layout,
-            reference_dev.chunk_size,
-            reference_dev.raid_disks,
-            disks,
+            reference_disk.level,
+            reference_disk.layout,
+            reference_disk.chunk_size,
+            reference_disk.raid_disks,
+            disk_map,
         )
 
 
-class Device(PhysicalDisk):
+class MDPhysicalDisk(PhysicalDisk):
     """Parse metadata from an MD device.
 
     Supports 0.90 and 1.x metadata.

{dissect_volume-3.12.dev2 → dissect_volume-3.13.dev1}/dissect/volume/raid/raid.py

@@ -12,6 +12,9 @@ from dissect.volume.raid.stream import (
     RAID456Stream,
 )
 
+DiskMap = dict[int, tuple[int, "PhysicalDisk"]]
+"""A dict of the structure: {disk_idx: {data_offset: PhysicalDisk}}"""
+
 
 class RAID:
     def __init__(self, configurations: list[Configuration]):
@@ -45,7 +48,7 @@ class VirtualDisk:
         layout: int,
         stripe_size: int,
         num_disks: int,
-        physical_disks: dict[int, tuple[int, PhysicalDisk]],
+        disk_map: DiskMap,
     ):
         self.name = name
         self.uuid = uuid
@@ -54,7 +57,7 @@ class VirtualDisk:
         self.layout = layout
         self.stripe_size = stripe_size
         self.num_disks = num_disks
-        self.physical_disks = physical_disks
+        self.disk_map = disk_map
 
     def open(self) -> BinaryIO:
         """Return a file-like object of the RAID volume in this set."""
@@ -64,7 +67,7 @@ class VirtualDisk:
             return RAID0Stream(self)
         elif self.level == Level.RAID1:
             # Don't really care which mirror to read from, so just open the first disk
-            return self.physical_disks[0][1].open()
+            return self.disk_map[0][1].open()
         elif self.level in (Level.RAID4, Level.RAID5, Level.RAID6):
             return RAID456Stream(self)
         elif self.level == Level.RAID10:

{dissect_volume-3.12.dev2 → dissect_volume-3.13.dev1}/dissect/volume/raid/stream.py

@@ -8,7 +8,7 @@ from dissect.util.stream import AlignedStream, MappingStream
 from dissect.volume.exceptions import RAIDError
 
 if TYPE_CHECKING:
-    from dissect.volume.raid.raid import PhysicalDisk, VirtualDisk
+    from dissect.volume.raid.raid import DiskMap, PhysicalDisk, VirtualDisk
 
 
 class Level(IntEnum):
@@ -54,13 +54,13 @@ class LinearStream(MappingStream):
     def __init__(self, virtual_disk: VirtualDisk):
         super().__init__()
         self.virtual_disk = virtual_disk
+        self.disk_map: DiskMap = dict(sorted(virtual_disk.disk_map.items()))
 
-        physical_disks: dict[int, tuple[int, PhysicalDisk]] = dict(sorted(virtual_disk.physical_disks.items()))
-        if len(physical_disks) != virtual_disk.num_disks:
+        if len(self.disk_map) != virtual_disk.num_disks:
             raise RAIDError(f"Missing disks in linear RAID set {virtual_disk.uuid} ({virtual_disk.name})")
 
         offset = 0
-        for disk_offset, disk in physical_disks.values():
+        for disk_offset, disk in self.disk_map.values():
             self.add(offset, disk.size, disk.open(), disk_offset)
             offset += disk.size
 
@@ -68,7 +68,7 @@ class LinearStream(MappingStream):
 class Zone(NamedTuple):
     zone_end: int
     dev_start: int
-    devices: list[tuple[int, PhysicalDisk]]
+    devices: DiskMap
 
 
 class RAID0Stream(AlignedStream):
@@ -76,27 +76,25 @@ class RAID0Stream(AlignedStream):
 
     def __init__(self, virtual_disk: VirtualDisk):
         self.virtual_disk = virtual_disk
+        self.disk_map = dict(sorted(self.virtual_disk.disk_map.items()))
 
-        disks = self.virtual_disk.physical_disks
-        if len(disks) != virtual_disk.num_disks:
+        if len(self.disk_map) != virtual_disk.num_disks:
             raise RAIDError(f"Missing disks in RAID0 set {virtual_disk.uuid} ({virtual_disk.name})")
 
         # Determine how many strip zones we need to construct
         # If a RAID0 set consists of devices with different sizes, additional strip zones
         # may exist on the larger devices but not on the smaller ones
         # Reference: create_strip_zones
-        disks: dict[int, tuple[int, PhysicalDisk]] = dict(sorted(disks.items()))
-        rounded_sizes = {}
-
+        rounded_sizes: dict[PhysicalDisk, int] = {}
         stripe_size = virtual_disk.stripe_size
         num_strip_zones = 0
-        for idx1, (_, dev1) in disks.items():
+        for idx1, (_, dev1) in self.disk_map.items():
             rounded_sizes[dev1] = (dev1.size // stripe_size) * stripe_size
 
             has_same_size = False
             # Check if dev1 is unequal in size to the sizes of any of the previous devices
             # If so, this means an extra strip zone is present
-            for idx2, (_, dev2) in disks.items():
+            for idx2, (_, dev2) in self.disk_map.items():
                 if idx1 == idx2:
                     break
 
@@ -109,25 +107,25 @@ class RAID0Stream(AlignedStream):
 
         # Determine the smallest device
         smallest = None
-        for _, dev in disks.values():
+        for _, dev in self.disk_map.values():
             if not smallest or rounded_sizes[dev] < rounded_sizes[smallest]:
                 smallest = dev
 
         # Construct the strip zones
-        zones = [Zone(rounded_sizes[smallest] * len(disks), 0, disks)]
+        zones = [Zone(rounded_sizes[smallest] * len(self.disk_map), 0, self.disk_map)]
 
         cur_zone_end = zones[0].zone_end
         for _ in range(1, num_strip_zones):
-            zone_devices = []
+            zone_devices = {}
             dev_start = rounded_sizes[smallest]
             smallest = None
 
             # Look for the next smallest device, that is: the smallest device that is larger than the "dev_start" device
-            for _, dev in disks.values():
+            for disk_idx, (data_offset, dev) in self.disk_map.items():
                 if rounded_sizes[dev] <= dev_start:
                     continue
 
-                zone_devices.append(dev)
+                zone_devices[disk_idx] = (data_offset, dev)
                 if not smallest or rounded_sizes[dev] < rounded_sizes[smallest]:
                     smallest = dev
 
@@ -192,8 +190,8 @@ class RAID456Stream(AlignedStream):
         self.algorithm = self.virtual_disk.layout
         self.max_degraded = 2 if self.level == 6 else 1
 
-        self.disks = self.virtual_disk.physical_disks
-        if len(self.disks) < self.virtual_disk.num_disks - self.max_degraded:
+        self.disk_map = self.virtual_disk.disk_map
+        if len(self.disk_map) < self.virtual_disk.num_disks - self.max_degraded:
             raise RAIDError(f"Missing disks in RAID{self.level} set {virtual_disk.uuid} ({virtual_disk.name})")
 
         super().__init__(self.virtual_disk.size, self.virtual_disk.stripe_size)
@@ -362,7 +360,7 @@ class RAID456Stream(AlignedStream):
         while length:
             stripe, offset_in_stripe, dd_idx, pd_idx, qd_idx, ddf_layout = self._get_stripe_read_info(offset)
             offset_in_device = stripe * stripe_size + offset_in_stripe
-            dd_start, dd_dev = self.disks[dd_idx]
+            dd_start, dd_dev = self.disk_map[dd_idx]
 
             stripe_remaining = stripe_size - offset_in_stripe
             read_length = min(length, stripe_remaining)
@@ -383,7 +381,7 @@ class RAID10Stream(AlignedStream):
     def __init__(self, virtual_disk: VirtualDisk):
         self.virtual_disk = virtual_disk
         self.raid_disks = self.virtual_disk.num_disks
-        self.devices = virtual_disk.physical_disks
+        self.disk_map = virtual_disk.disk_map
 
         # Reference: setup_geo
         layout = virtual_disk.layout
@@ -420,7 +418,7 @@ class RAID10Stream(AlignedStream):
 
             if self.far_offset:
                 stripe *= self.far_copies
-            device_start, device = self.devices[dev]
+            device_start, device = self.disk_map[dev]
 
             stripe_remaining = stripe_size - offset_in_stripe
             read_length = min(length, stripe_remaining)

dissect_volume-3.13.dev1/dissect/volume/vinum/c_vinum.py

@@ -0,0 +1,83 @@
+from dissect.cstruct import cstruct
+
+# Structures are copied from:
+# https://github.com/freebsd/freebsd-src/blob/f21a6a6a8fc59393173d9a537ed8cebbdbd6343c/sys/geom/vinum/geom_vinum_var.h
+
+vinum_def = """
+struct timeval {
+    uint64 sec;
+    uint64 usec;
+};
+
+typedef uint64 off_t;
+
+/*
+ * Slice header
+ *
+ * Vinum drives start with this structure:
+ *
+ *\                                            Sector
+ * |--------------------------------------|
+ * |   PDP-11 memorial boot block         |      0
+ * |--------------------------------------|
+ * |   Disk label, maybe                  |      1
+ * |--------------------------------------|
+ * |   Slice definition  (vinum_hdr)      |      8
+ * |--------------------------------------|
+ * |                                      |
+ * |   Configuration info, first copy     |      9
+ * |                                      |
+ * |--------------------------------------|
+ * |                                      |
+ * |   Configuration info, second copy    |      9 + size of config
+ * |                                      |
+ * |--------------------------------------|
+ */
+
+/* Sizes and offsets of our information. */
+#define GV_HDR_OFFSET       4096    /* Offset of vinum header. */
+#define GV_HDR_LEN          512     /* Size of vinum header. */
+#define GV_CFG_OFFSET       4608    /* Offset of first config copy. */
+#define GV_CFG_LEN          65536   /* Size of config copy. */
+
+/* This is where the actual data starts. */
+#define GV_DATA_START       (GV_CFG_LEN * 2 + GV_CFG_OFFSET)
+/* #define GV_DATA_START       (GV_CFG_LEN * 2 + GV_HDR_LEN) */
+
+#define GV_MAXDRIVENAME     32      /* Maximum length of a device name. */
+
+/*
+ * hostname is 256 bytes long, but we don't need to shlep multiple copies in
+ * vinum.  We use the host name just to identify this system, and 32 bytes
+ * should be ample for that purpose.
+ */
+
+#define GV_HOSTNAME_LEN     32
+struct gv_label {
+    char            sysname[GV_HOSTNAME_LEN];   /* System name at creation time. */
+    char            name[GV_MAXDRIVENAME];      /* Our name of the drive. */
+    struct timeval  date_of_birth;              /* The time it was created ... */
+    struct timeval  last_update;                /* ... and the time of last update. */
+    off_t           drive_size;                 /* Total size incl. headers. */
+};
+
+#define GV_OLD_MAGIC        0x494E2056494E4F00LL
+#define GV_OLD_NOMAGIC      0x4E4F2056494E4F00LL
+#define GV_MAGIC            0x56494E554D2D3100LL
+#define GV_NOMAGIC          0x56494E554D2D2D00LL
+
+/* The 'header' of each valid vinum drive. */
+struct gv_hdr {
+    uint64_t        magic;
+    uint64_t        config_length;
+    struct gv_label label;
+} header;
+"""  # noqa W605
+
+c_vinum = cstruct(endian=">").load(vinum_def)
+
+# Not really needed as this size is hardcoded in the various GV_*_OFFSET and related values
+SECTOR_SIZE = 512
+
+MAGIC_ACTIVE = {c_vinum.GV_OLD_MAGIC, c_vinum.GV_MAGIC}
+MAGIC_INACTIVE = {c_vinum.GV_OLD_NOMAGIC, c_vinum.GV_NOMAGIC}