요약·해설과 원문, 전문 번역을 서로 분리했습니다. API 이름, symbol, source path는 원문 표기를 사용합니다.
1. 요약·해설
원문의 핵심 논리와 kernel programming 관점의 보충 설명입니다. 아래의 전문 번역과는 별도로 작성했습니다.
2. 영어 원문 전체
번역 기준이 된 Linux v6.18.37 원문입니다. 줄 번호는 이 버전의 파일 좌표입니다.
원문 전체 펼치기
.. SPDX-License-Identifier: GPL-2.0
Index Nodes
-----------
In a regular UNIX filesystem, the inode stores all the metadata
pertaining to the file (time stamps, block maps, extended attributes,
etc), not the directory entry. To find the information associated with a
file, one must traverse the directory files to find the directory entry
associated with a file, then load the inode to find the metadata for
that file. ext4 appears to cheat (for performance reasons) a little bit
by storing a copy of the file type (normally stored in the inode) in the
directory entry. (Compare all this to FAT, which stores all the file
information directly in the directory entry, but does not support hard
links and is in general more seek-happy than ext4 due to its simpler
block allocator and extensive use of linked lists.)
The inode table is a linear array of ``struct ext4_inode``. The table is
sized to have enough blocks to store at least
``sb.s_inode_size * sb.s_inodes_per_group`` bytes. The number of the
block group containing an inode can be calculated as
``(inode_number - 1) / sb.s_inodes_per_group``, and the offset into the
group's table is ``(inode_number - 1) % sb.s_inodes_per_group``. There
is no inode 0.
The inode checksum is calculated against the FS UUID, the inode number,
and the inode structure itself.
The inode table entry is laid out in ``struct ext4_inode``.
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
:class: longtable
* - Offset
- Size
- Name
- Description
* - 0x0
- __le16
- i_mode
- File mode. See the table i_mode_ below.
* - 0x2
- __le16
- i_uid
- Lower 16-bits of Owner UID.
* - 0x4
- __le32
- i_size_lo
- Lower 32-bits of size in bytes.
* - 0x8
- __le32
- i_atime
- Last access time, in seconds since the epoch. However, if the EA_INODE
inode flag is set, this inode stores an extended attribute value and
this field contains the checksum of the value.
* - 0xC
- __le32
- i_ctime
- Last inode change time, in seconds since the epoch. However, if the
EA_INODE inode flag is set, this inode stores an extended attribute
value and this field contains the lower 32 bits of the attribute value's
reference count.
* - 0x10
- __le32
- i_mtime
- Last data modification time, in seconds since the epoch. However, if the
EA_INODE inode flag is set, this inode stores an extended attribute
value and this field contains the number of the inode that owns the
extended attribute.
* - 0x14
- __le32
- i_dtime
- Deletion Time, in seconds since the epoch.
* - 0x18
- __le16
- i_gid
- Lower 16-bits of GID.
* - 0x1A
- __le16
- i_links_count
- Hard link count. Normally, ext4 does not permit an inode to have more
than 65,000 hard links. This applies to files as well as directories,
which means that there cannot be more than 64,998 subdirectories in a
directory (each subdirectory's '..' entry counts as a hard link, as does
the '.' entry in the directory itself). With the DIR_NLINK feature
enabled, ext4 supports more than 64,998 subdirectories by setting this
field to 1 to indicate that the number of hard links is not known.
* - 0x1C
- __le32
- i_blocks_lo
- Lower 32-bits of “block” count. If the huge_file feature flag is not
set on the filesystem, the file consumes ``i_blocks_lo`` 512-byte blocks
on disk. If huge_file is set and EXT4_HUGE_FILE_FL is NOT set in
``inode.i_flags``, then the file consumes ``i_blocks_lo + (i_blocks_hi
<< 32)`` 512-byte blocks on disk. If huge_file is set and
EXT4_HUGE_FILE_FL IS set in ``inode.i_flags``, then this file
consumes (``i_blocks_lo + i_blocks_hi`` << 32) filesystem blocks on
disk.
* - 0x20
- __le32
- i_flags
- Inode flags. See the table i_flags_ below.
* - 0x24
- 4 bytes
- i_osd1
- See the table i_osd1_ for more details.
* - 0x28
- 60 bytes
- i_block[EXT4_N_BLOCKS=15]
- Block map or extent tree. See the section “The Contents of inode.i_block”.
* - 0x64
- __le32
- i_generation
- File version (for NFS).
* - 0x68
- __le32
- i_file_acl_lo
- Lower 32-bits of extended attribute block. ACLs are of course one of
many possible extended attributes; I think the name of this field is a
result of the first use of extended attributes being for ACLs.
* - 0x6C
- __le32
- i_size_high / i_dir_acl
- Upper 32-bits of file/directory size. In ext2/3 this field was named
i_dir_acl, though it was usually set to zero and never used.
* - 0x70
- __le32
- i_obso_faddr
- (Obsolete) fragment address.
* - 0x74
- 12 bytes
- i_osd2
- See the table i_osd2_ for more details.
* - 0x80
- __le16
- i_extra_isize
- Size of this inode - 128. Alternately, the size of the extended inode
fields beyond the original ext2 inode, including this field.
* - 0x82
- __le16
- i_checksum_hi
- Upper 16-bits of the inode checksum.
* - 0x84
- __le32
- i_ctime_extra
- Extra change time bits. This provides sub-second precision. See Inode
Timestamps section.
* - 0x88
- __le32
- i_mtime_extra
- Extra modification time bits. This provides sub-second precision.
* - 0x8C
- __le32
- i_atime_extra
- Extra access time bits. This provides sub-second precision.
* - 0x90
- __le32
- i_crtime
- File creation time, in seconds since the epoch.
* - 0x94
- __le32
- i_crtime_extra
- Extra file creation time bits. This provides sub-second precision.
* - 0x98
- __le32
- i_version_hi
- Upper 32-bits for version number.
* - 0x9C
- __le32
- i_projid
- Project ID.
.. _i_mode:
The ``i_mode`` value is a combination of the following flags:
.. list-table::
:widths: 16 64
:header-rows: 1
* - Value
- Description
* - 0x1
- S_IXOTH (Others may execute)
* - 0x2
- S_IWOTH (Others may write)
* - 0x4
- S_IROTH (Others may read)
* - 0x8
- S_IXGRP (Group members may execute)
* - 0x10
- S_IWGRP (Group members may write)
* - 0x20
- S_IRGRP (Group members may read)
* - 0x40
- S_IXUSR (Owner may execute)
* - 0x80
- S_IWUSR (Owner may write)
* - 0x100
- S_IRUSR (Owner may read)
* - 0x200
- S_ISVTX (Sticky bit)
* - 0x400
- S_ISGID (Set GID)
* - 0x800
- S_ISUID (Set UID)
* -
- These are mutually-exclusive file types:
* - 0x1000
- S_IFIFO (FIFO)
* - 0x2000
- S_IFCHR (Character device)
* - 0x4000
- S_IFDIR (Directory)
* - 0x6000
- S_IFBLK (Block device)
* - 0x8000
- S_IFREG (Regular file)
* - 0xA000
- S_IFLNK (Symbolic link)
* - 0xC000
- S_IFSOCK (Socket)
.. _i_flags:
The ``i_flags`` field is a combination of these values:
.. list-table::
:widths: 16 64
:header-rows: 1
* - Value
- Description
* - 0x1
- This file requires secure deletion (EXT4_SECRM_FL). (not implemented)
* - 0x2
- This file should be preserved, should undeletion be desired
(EXT4_UNRM_FL). (not implemented)
* - 0x4
- File is compressed (EXT4_COMPR_FL). (not really implemented)
* - 0x8
- All writes to the file must be synchronous (EXT4_SYNC_FL).
* - 0x10
- File is immutable (EXT4_IMMUTABLE_FL).
* - 0x20
- File can only be appended (EXT4_APPEND_FL).
* - 0x40
- The dump(1) utility should not dump this file (EXT4_NODUMP_FL).
* - 0x80
- Do not update access time (EXT4_NOATIME_FL).
* - 0x100
- Dirty compressed file (EXT4_DIRTY_FL). (not used)
* - 0x200
- File has one or more compressed clusters (EXT4_COMPRBLK_FL). (not used)
* - 0x400
- Do not compress file (EXT4_NOCOMPR_FL). (not used)
* - 0x800
- Encrypted inode (EXT4_ENCRYPT_FL). This bit value previously was
EXT4_ECOMPR_FL (compression error), which was never used.
* - 0x1000
- Directory has hashed indexes (EXT4_INDEX_FL).
* - 0x2000
- AFS magic directory (EXT4_IMAGIC_FL).
* - 0x4000
- File data must always be written through the journal
(EXT4_JOURNAL_DATA_FL).
* - 0x8000
- File tail should not be merged (EXT4_NOTAIL_FL). (not used by ext4)
* - 0x10000
- All directory entry data should be written synchronously (see
``dirsync``) (EXT4_DIRSYNC_FL).
* - 0x20000
- Top of directory hierarchy (EXT4_TOPDIR_FL).
* - 0x40000
- This is a huge file (EXT4_HUGE_FILE_FL).
* - 0x80000
- Inode uses extents (EXT4_EXTENTS_FL).
* - 0x100000
- Verity protected file (EXT4_VERITY_FL).
* - 0x200000
- Inode stores a large extended attribute value in its data blocks
(EXT4_EA_INODE_FL).
* - 0x400000
- This file has blocks allocated past EOF (EXT4_EOFBLOCKS_FL).
(deprecated)
* - 0x01000000
- Inode is a snapshot (``EXT4_SNAPFILE_FL``). (not in mainline)
* - 0x04000000
- Snapshot is being deleted (``EXT4_SNAPFILE_DELETED_FL``). (not in
mainline)
* - 0x08000000
- Snapshot shrink has completed (``EXT4_SNAPFILE_SHRUNK_FL``). (not in
mainline)
* - 0x10000000
- Inode has inline data (EXT4_INLINE_DATA_FL).
* - 0x20000000
- Create children with the same project ID (EXT4_PROJINHERIT_FL).
* - 0x80000000
- Reserved for ext4 library (EXT4_RESERVED_FL).
* -
- Aggregate flags:
* - 0x705BDFFF
- User-visible flags.
* - 0x604BC0FF
- User-modifiable flags. Note that while EXT4_JOURNAL_DATA_FL and
EXT4_EXTENTS_FL can be set with setattr, they are not in the kernel's
EXT4_FL_USER_MODIFIABLE mask, since it needs to handle the setting of
these flags in a special manner and they are masked out of the set of
flags that are saved directly to i_flags.
.. _i_osd1:
The ``osd1`` field has multiple meanings depending on the creator:
Linux:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le32
- l_i_version
- Inode version. However, if the EA_INODE inode flag is set, this inode
stores an extended attribute value and this field contains the upper 32
bits of the attribute value's reference count.
Hurd:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le32
- h_i_translator
- ??
Masix:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le32
- m_i_reserved
- ??
.. _i_osd2:
The ``osd2`` field has multiple meanings depending on the filesystem creator:
Linux:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le16
- l_i_blocks_high
- Upper 16-bits of the block count. Please see the note attached to
i_blocks_lo.
* - 0x2
- __le16
- l_i_file_acl_high
- Upper 16-bits of the extended attribute block (historically, the file
ACL location). See the Extended Attributes section below.
* - 0x4
- __le16
- l_i_uid_high
- Upper 16-bits of the Owner UID.
* - 0x6
- __le16
- l_i_gid_high
- Upper 16-bits of the GID.
* - 0x8
- __le16
- l_i_checksum_lo
- Lower 16-bits of the inode checksum.
* - 0xA
- __le16
- l_i_reserved
- Unused.
Hurd:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le16
- h_i_reserved1
- ??
* - 0x2
- __u16
- h_i_mode_high
- Upper 16-bits of the file mode.
* - 0x4
- __le16
- h_i_uid_high
- Upper 16-bits of the Owner UID.
* - 0x6
- __le16
- h_i_gid_high
- Upper 16-bits of the GID.
* - 0x8
- __u32
- h_i_author
- Author code?
Masix:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le16
- h_i_reserved1
- ??
* - 0x2
- __u16
- m_i_file_acl_high
- Upper 16-bits of the extended attribute block (historically, the file
ACL location).
* - 0x4
- __u32
- m_i_reserved2[2]
- ??
Inode Size
~~~~~~~~~~
In ext2 and ext3, the inode structure size was fixed at 128 bytes
(``EXT2_GOOD_OLD_INODE_SIZE``) and each inode had a disk record size of
128 bytes. Starting with ext4, it is possible to allocate a larger
on-disk inode at format time for all inodes in the filesystem to provide
space beyond the end of the original ext2 inode. The on-disk inode
record size is recorded in the superblock as ``s_inode_size``. The
number of bytes actually used by struct ext4_inode beyond the original
128-byte ext2 inode is recorded in the ``i_extra_isize`` field for each
inode, which allows struct ext4_inode to grow for a new kernel without
having to upgrade all of the on-disk inodes. Access to fields beyond
EXT2_GOOD_OLD_INODE_SIZE should be verified to be within
``i_extra_isize``. By default, ext4 inode records are 256 bytes, and (as
of August 2019) the inode structure is 160 bytes
(``i_extra_isize = 32``). The extra space between the end of the inode
structure and the end of the inode record can be used to store extended
attributes. Each inode record can be as large as the filesystem block
size, though this is not terribly efficient.
Finding an Inode
~~~~~~~~~~~~~~~~
Each block group contains ``sb->s_inodes_per_group`` inodes. Because
inode 0 is defined not to exist, this formula can be used to find the
block group that an inode lives in:
``bg = (inode_num - 1) / sb->s_inodes_per_group``. The particular inode
can be found within the block group's inode table at
``index = (inode_num - 1) % sb->s_inodes_per_group``. To get the byte
address within the inode table, use
``offset = index * sb->s_inode_size``.
Inode Timestamps
~~~~~~~~~~~~~~~~
Four timestamps are recorded in the lower 128 bytes of the inode
structure -- inode change time (ctime), access time (atime), data
modification time (mtime), and deletion time (dtime). The four fields
are 32-bit signed integers that represent seconds since the Unix epoch
(1970-01-01 00:00:00 GMT), which means that the fields will overflow in
January 2038. If the filesystem does not have orphan_file feature, inodes
that are not linked from any directory but are still open (orphan inodes) have
the dtime field overloaded for use with the orphan list. The superblock field
``s_last_orphan`` points to the first inode in the orphan list; dtime is then
the number of the next orphaned inode, or zero if there are no more orphans.
If the inode structure size ``sb->s_inode_size`` is larger than 128
bytes and the ``i_inode_extra`` field is large enough to encompass the
respective ``i_[cma]time_extra`` field, the ctime, atime, and mtime
inode fields are widened to 64 bits. Within this “extra” 32-bit field,
the lower two bits are used to extend the 32-bit seconds field to be 34
bit wide; the upper 30 bits are used to provide nanosecond timestamp
accuracy. Therefore, timestamps should not overflow until May 2446.
dtime was not widened. There is also a fifth timestamp to record inode
creation time (crtime); this field is 64-bits wide and decoded in the
same manner as 64-bit [cma]time. Neither crtime nor dtime are accessible
through the regular stat() interface, though debugfs will report them.
We use the 32-bit signed time value plus (2^32 * (extra epoch bits)).
In other words:
.. list-table::
:widths: 20 20 20 20 20
:header-rows: 1
* - Extra epoch bits
- MSB of 32-bit time
- Adjustment for signed 32-bit to 64-bit tv_sec
- Decoded 64-bit tv_sec
- valid time range
* - 0 0
- 1
- 0
- ``-0x80000000 - -0x00000001``
- 1901-12-13 to 1969-12-31
* - 0 0
- 0
- 0
- ``0x000000000 - 0x07fffffff``
- 1970-01-01 to 2038-01-19
* - 0 1
- 1
- 0x100000000
- ``0x080000000 - 0x0ffffffff``
- 2038-01-19 to 2106-02-07
* - 0 1
- 0
- 0x100000000
- ``0x100000000 - 0x17fffffff``
- 2106-02-07 to 2174-02-25
* - 1 0
- 1
- 0x200000000
- ``0x180000000 - 0x1ffffffff``
- 2174-02-25 to 2242-03-16
* - 1 0
- 0
- 0x200000000
- ``0x200000000 - 0x27fffffff``
- 2242-03-16 to 2310-04-04
* - 1 1
- 1
- 0x300000000
- ``0x280000000 - 0x2ffffffff``
- 2310-04-04 to 2378-04-22
* - 1 1
- 0
- 0x300000000
- ``0x300000000 - 0x37fffffff``
- 2378-04-22 to 2446-05-10
This is a somewhat odd encoding since there are effectively seven times
as many positive values as negative values. There have also been
long-standing bugs decoding and encoding dates beyond 2038, which don't
seem to be fixed as of kernel 3.12 and e2fsprogs 1.42.8. 64-bit kernels
incorrectly use the extra epoch bits 1,1 for dates between 1901 and
1970. At some point the kernel will be fixed and e2fsck will fix this
situation, assuming that it is run before 2310.
3. 한국어 전문 번역
영어 원문의 문단 순서와 의미를 유지한 전체 번역입니다. 코드, 함수명, symbol과 URL은 원문 표기를 유지합니다.
inode와 inode table
1-29일반적인 UNIX 파일시스템에서는 디렉터리 엔트리가 아니라 inode가 timestamp, block map, extended attribute 등 파일의 모든 메타데이터를 저장합니다. 파일 정보를 찾으려면 디렉터리 파일을 순회해 해당 파일의 directory entry를 찾고, 그 뒤 inode를 읽어 메타데이터를 가져와야 합니다.
ext4는 성능을 위해 원래 inode에 있는 file type의 사본을 directory entry에도 저장합니다. FAT는 파일 정보를 directory entry에 직접 저장하지만 hard link를 지원하지 않으며, 단순한 block allocator와 linked list를 많이 사용하므로 ext4보다 seek가 더 많이 발생하는 편입니다.
inode table은 `struct ext4_inode`의 선형 배열입니다. table은 최소한 `sb.s_inode_size * sb.s_inodes_per_group`바이트를 저장할 수 있는 블록 수로 구성됩니다.
inode가 속한 block group은 `(inode_number - 1) / sb.s_inodes_per_group`으로 계산하고, 해당 그룹 table 안의 index는 `(inode_number - 1) % sb.s_inodes_per_group`입니다. inode 0은 존재하지 않습니다.
inode checksum은 FS UUID, inode number, inode structure 자체를 입력으로 계산합니다. inode table entry의 디스크 형식은 `struct ext4_inode`입니다.
directory entry와 inode table을 거쳐 파일 메타데이터를 찾는 경로입니다.
.. SPDX-License-Identifier: GPL-2.0
Index Nodes
-----------
In a regular UNIX filesystem, the inode stores all the metadata
pertaining to the file (time stamps, block maps, extended attributes,
etc), not the directory entry. To find the information associated with a
file, one must traverse the directory files to find the directory entry
associated with a file, then load the inode to find the metadata for
that file. ext4 appears to cheat (for performance reasons) a little bit
by storing a copy of the file type (normally stored in the inode) in the
directory entry. (Compare all this to FAT, which stores all the file
information directly in the directory entry, but does not support hard
links and is in general more seek-happy than ext4 due to its simpler
block allocator and extensive use of linked lists.)
The inode table is a linear array of ``struct ext4_inode``. The table is
sized to have enough blocks to store at least
``sb.s_inode_size * sb.s_inodes_per_group`` bytes. The number of the
block group containing an inode can be calculated as
``(inode_number - 1) / sb.s_inodes_per_group``, and the offset into the
group's table is ``(inode_number - 1) % sb.s_inodes_per_group``. There
is no inode 0.
The inode checksum is calculated against the FS UUID, the inode number,
and the inode structure itself.
The inode table entry is laid out in ``struct ext4_inode``.
`struct ext4_inode` 필드
30-174`struct ext4_inode`의 앞 128바이트는 ext2 시절부터 이어진 기본 inode 형식이고, `0x80` 이후는 ext4의 확장 inode 필드입니다.
`i_mode`는 파일 종류와 권한을 함께 저장합니다. `i_uid`와 `i_gid`는 소유자 ID의 하위 16비트이며 상위 비트는 OS별 `i_osd2`에 있습니다. `i_size_lo`와 `i_size_high`를 결합해 파일 또는 디렉터리 크기를 표현합니다.
`i_atime`, `i_ctime`, `i_mtime`, `i_dtime`은 기본 timestamp입니다. `EXT4_EA_INODE_FL`이 설정된 EA inode에서는 각각 xattr checksum, 참조 횟수 하위 32비트, 소유 inode number로 의미가 바뀝니다.
`i_links_count`는 hard link 수입니다. ext4의 일반 한도는 65,000이며 디렉터리의 `.`과 하위 디렉터리의 `..`도 포함하므로 하위 디렉터리는 최대 64,998개입니다. `DIR_NLINK` 기능을 켜면 이 필드를 1로 설정해 실제 link 수를 모른다는 뜻으로 사용하여 한도를 넘길 수 있습니다.
`i_blocks_lo`는 디스크 사용량을 나타냅니다. `huge_file` 기능과 inode의 `EXT4_HUGE_FILE_FL` 조합에 따라 512-byte block 수 또는 filesystem block 수로 해석하며 `i_blocks_hi`와 함께 사용합니다.
`i_block[EXT4_N_BLOCKS=15]` 60바이트에는 block map 또는 extent tree가 들어갑니다. `i_file_acl_lo`는 extended attribute block 위치의 하위 32비트이고, `i_size_high / i_dir_acl`은 크기의 상위 32비트입니다. `i_obso_faddr`는 더 이상 쓰지 않는 fragment address입니다.
확장 영역의 `i_extra_isize`는 inode 전체 크기에서 128을 뺀 값, 즉 이 필드 자체를 포함해 원래 ext2 inode 뒤에 실제로 사용하는 확장 필드의 크기입니다. `i_checksum_hi`는 inode checksum의 상위 16비트입니다.
`i_ctime_extra`, `i_mtime_extra`, `i_atime_extra`, `i_crtime_extra`는 초 단위 필드를 확장하고 sub-second 정밀도를 제공합니다. `i_crtime`은 파일 생성 시각, `i_version_hi`는 version number의 상위 32비트, `i_projid`는 project ID입니다.
기본 128바이트와 ext4 확장 필드를 offset 순서대로 정리한 표입니다.
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
:class: longtable
* - Offset
- Size
- Name
- Description
* - 0x0
- __le16
- i_mode
- File mode. See the table i_mode_ below.
* - 0x2
- __le16
- i_uid
- Lower 16-bits of Owner UID.
* - 0x4
- __le32
- i_size_lo
- Lower 32-bits of size in bytes.
* - 0x8
- __le32
- i_atime
- Last access time, in seconds since the epoch. However, if the EA_INODE
inode flag is set, this inode stores an extended attribute value and
this field contains the checksum of the value.
* - 0xC
- __le32
- i_ctime
- Last inode change time, in seconds since the epoch. However, if the
EA_INODE inode flag is set, this inode stores an extended attribute
value and this field contains the lower 32 bits of the attribute value's
reference count.
* - 0x10
- __le32
- i_mtime
- Last data modification time, in seconds since the epoch. However, if the
EA_INODE inode flag is set, this inode stores an extended attribute
value and this field contains the number of the inode that owns the
extended attribute.
* - 0x14
- __le32
- i_dtime
- Deletion Time, in seconds since the epoch.
* - 0x18
- __le16
- i_gid
- Lower 16-bits of GID.
* - 0x1A
- __le16
- i_links_count
- Hard link count. Normally, ext4 does not permit an inode to have more
than 65,000 hard links. This applies to files as well as directories,
which means that there cannot be more than 64,998 subdirectories in a
directory (each subdirectory's '..' entry counts as a hard link, as does
the '.' entry in the directory itself). With the DIR_NLINK feature
enabled, ext4 supports more than 64,998 subdirectories by setting this
field to 1 to indicate that the number of hard links is not known.
* - 0x1C
- __le32
- i_blocks_lo
- Lower 32-bits of “block” count. If the huge_file feature flag is not
set on the filesystem, the file consumes ``i_blocks_lo`` 512-byte blocks
on disk. If huge_file is set and EXT4_HUGE_FILE_FL is NOT set in
``inode.i_flags``, then the file consumes ``i_blocks_lo + (i_blocks_hi
<< 32)`` 512-byte blocks on disk. If huge_file is set and
EXT4_HUGE_FILE_FL IS set in ``inode.i_flags``, then this file
consumes (``i_blocks_lo + i_blocks_hi`` << 32) filesystem blocks on
disk.
* - 0x20
- __le32
- i_flags
- Inode flags. See the table i_flags_ below.
* - 0x24
- 4 bytes
- i_osd1
- See the table i_osd1_ for more details.
* - 0x28
- 60 bytes
- i_block[EXT4_N_BLOCKS=15]
- Block map or extent tree. See the section “The Contents of inode.i_block”.
* - 0x64
- __le32
- i_generation
- File version (for NFS).
* - 0x68
- __le32
- i_file_acl_lo
- Lower 32-bits of extended attribute block. ACLs are of course one of
many possible extended attributes; I think the name of this field is a
result of the first use of extended attributes being for ACLs.
* - 0x6C
- __le32
- i_size_high / i_dir_acl
- Upper 32-bits of file/directory size. In ext2/3 this field was named
i_dir_acl, though it was usually set to zero and never used.
* - 0x70
- __le32
- i_obso_faddr
- (Obsolete) fragment address.
* - 0x74
- 12 bytes
- i_osd2
- See the table i_osd2_ for more details.
* - 0x80
- __le16
- i_extra_isize
- Size of this inode - 128. Alternately, the size of the extended inode
fields beyond the original ext2 inode, including this field.
* - 0x82
- __le16
- i_checksum_hi
- Upper 16-bits of the inode checksum.
* - 0x84
- __le32
- i_ctime_extra
- Extra change time bits. This provides sub-second precision. See Inode
Timestamps section.
* - 0x88
- __le32
- i_mtime_extra
- Extra modification time bits. This provides sub-second precision.
* - 0x8C
- __le32
- i_atime_extra
- Extra access time bits. This provides sub-second precision.
* - 0x90
- __le32
- i_crtime
- File creation time, in seconds since the epoch.
* - 0x94
- __le32
- i_crtime_extra
- Extra file creation time bits. This provides sub-second precision.
* - 0x98
- __le32
- i_version_hi
- Upper 32-bits for version number.
* - 0x9C
- __le32
- i_projid
- Project ID.
`i_mode` 권한과 파일 종류
175-225`i_mode`는 접근 권한, sticky/set-ID bit, 파일 종류를 조합한 값입니다. 권한 비트는 owner, group, others 각각의 read, write, execute 권한을 표현합니다.
파일 종류 비트는 서로 배타적이며 FIFO, character device, directory, block device, regular file, symbolic link, socket 가운데 하나를 선택합니다.
권한·특수 비트와 서로 배타적인 파일 종류입니다.
.. _i_mode:
The ``i_mode`` value is a combination of the following flags:
.. list-table::
:widths: 16 64
:header-rows: 1
* - Value
- Description
* - 0x1
- S_IXOTH (Others may execute)
* - 0x2
- S_IWOTH (Others may write)
* - 0x4
- S_IROTH (Others may read)
* - 0x8
- S_IXGRP (Group members may execute)
* - 0x10
- S_IWGRP (Group members may write)
* - 0x20
- S_IRGRP (Group members may read)
* - 0x40
- S_IXUSR (Owner may execute)
* - 0x80
- S_IWUSR (Owner may write)
* - 0x100
- S_IRUSR (Owner may read)
* - 0x200
- S_ISVTX (Sticky bit)
* - 0x400
- S_ISGID (Set GID)
* - 0x800
- S_ISUID (Set UID)
* -
- These are mutually-exclusive file types:
* - 0x1000
- S_IFIFO (FIFO)
* - 0x2000
- S_IFCHR (Character device)
* - 0x4000
- S_IFDIR (Directory)
* - 0x6000
- S_IFBLK (Block device)
* - 0x8000
- S_IFREG (Regular file)
* - 0xA000
- S_IFLNK (Symbolic link)
* - 0xC000
- S_IFSOCK (Socket)
`i_flags` inode 상태
226-312`i_flags`는 파일의 쓰기 방식, 변경 가능성, 압축·암호화·인덱싱·extent·verity·inline data 같은 동작과 형식을 제어하는 비트 조합입니다.
일부 오래된 압축, 삭제 복구, snapshot 플래그는 구현되지 않았거나 mainline에 없습니다. `EXT4_ECOMPR_FL`로 쓰인 적이 없는 `0x800` 비트는 현재 `EXT4_ENCRYPT_FL`입니다.
user-visible mask는 `0x705BDFFF`, user-modifiable mask는 `0x604BC0FF`입니다. `EXT4_JOURNAL_DATA_FL`과 `EXT4_EXTENTS_FL`은 `setattr`로 설정할 수 있지만 커널이 특별히 처리해야 하므로 `EXT4_FL_USER_MODIFIABLE` mask에 포함되지 않고, `i_flags`에 직접 저장하는 flag 집합에서도 가려집니다.
inode의 형식과 동작을 나타내는 개별 및 집계 flag입니다.
.. _i_flags:
The ``i_flags`` field is a combination of these values:
.. list-table::
:widths: 16 64
:header-rows: 1
* - Value
- Description
* - 0x1
- This file requires secure deletion (EXT4_SECRM_FL). (not implemented)
* - 0x2
- This file should be preserved, should undeletion be desired
(EXT4_UNRM_FL). (not implemented)
* - 0x4
- File is compressed (EXT4_COMPR_FL). (not really implemented)
* - 0x8
- All writes to the file must be synchronous (EXT4_SYNC_FL).
* - 0x10
- File is immutable (EXT4_IMMUTABLE_FL).
* - 0x20
- File can only be appended (EXT4_APPEND_FL).
* - 0x40
- The dump(1) utility should not dump this file (EXT4_NODUMP_FL).
* - 0x80
- Do not update access time (EXT4_NOATIME_FL).
* - 0x100
- Dirty compressed file (EXT4_DIRTY_FL). (not used)
* - 0x200
- File has one or more compressed clusters (EXT4_COMPRBLK_FL). (not used)
* - 0x400
- Do not compress file (EXT4_NOCOMPR_FL). (not used)
* - 0x800
- Encrypted inode (EXT4_ENCRYPT_FL). This bit value previously was
EXT4_ECOMPR_FL (compression error), which was never used.
* - 0x1000
- Directory has hashed indexes (EXT4_INDEX_FL).
* - 0x2000
- AFS magic directory (EXT4_IMAGIC_FL).
* - 0x4000
- File data must always be written through the journal
(EXT4_JOURNAL_DATA_FL).
* - 0x8000
- File tail should not be merged (EXT4_NOTAIL_FL). (not used by ext4)
* - 0x10000
- All directory entry data should be written synchronously (see
``dirsync``) (EXT4_DIRSYNC_FL).
* - 0x20000
- Top of directory hierarchy (EXT4_TOPDIR_FL).
* - 0x40000
- This is a huge file (EXT4_HUGE_FILE_FL).
* - 0x80000
- Inode uses extents (EXT4_EXTENTS_FL).
* - 0x100000
- Verity protected file (EXT4_VERITY_FL).
* - 0x200000
- Inode stores a large extended attribute value in its data blocks
(EXT4_EA_INODE_FL).
* - 0x400000
- This file has blocks allocated past EOF (EXT4_EOFBLOCKS_FL).
(deprecated)
* - 0x01000000
- Inode is a snapshot (``EXT4_SNAPFILE_FL``). (not in mainline)
* - 0x04000000
- Snapshot is being deleted (``EXT4_SNAPFILE_DELETED_FL``). (not in
mainline)
* - 0x08000000
- Snapshot shrink has completed (``EXT4_SNAPFILE_SHRUNK_FL``). (not in
mainline)
* - 0x10000000
- Inode has inline data (EXT4_INLINE_DATA_FL).
* - 0x20000000
- Create children with the same project ID (EXT4_PROJINHERIT_FL).
* - 0x80000000
- Reserved for ext4 library (EXT4_RESERVED_FL).
* -
- Aggregate flags:
* - 0x705BDFFF
- User-visible flags.
* - 0x604BC0FF
- User-modifiable flags. Note that while EXT4_JOURNAL_DATA_FL and
EXT4_EXTENTS_FL can be set with setattr, they are not in the kernel's
EXT4_FL_USER_MODIFIABLE mask, since it needs to handle the setting of
these flags in a special manner and they are masked out of the set of
flags that are saved directly to i_flags.
`i_osd1` creator OS별 의미
313-3634바이트 `i_osd1` 필드는 파일시스템을 만든 운영체제에 따라 의미가 다릅니다.
Linux의 `l_i_version`은 inode version입니다. `EXT4_EA_INODE_FL`이 설정된 inode에서는 큰 확장 속성 값의 참조 횟수 상위 32비트를 저장합니다.
Hurd는 `h_i_translator`, Masix는 `m_i_reserved`로 정의하지만 문서에는 두 필드의 의미가 확정되지 않은 `??`로 표시되어 있습니다.
creator별 4바이트 필드 해석입니다.
.. _i_osd1:
The ``osd1`` field has multiple meanings depending on the creator:
Linux:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le32
- l_i_version
- Inode version. However, if the EA_INODE inode flag is set, this inode
stores an extended attribute value and this field contains the upper 32
bits of the attribute value's reference count.
Hurd:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le32
- h_i_translator
- ??
Masix:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le32
- m_i_reserved
- ??
`i_osd2` creator OS별 의미
364-45912바이트 `i_osd2`도 filesystem creator에 따라 배치가 달라집니다.
Linux 배치는 block count, extended attribute block, UID, GID, inode checksum의 상위 또는 하위 비트를 보충하고 마지막 2바이트를 예약합니다.
Linux가 사용하는 12바이트 확장 필드입니다.
Hurd 배치는 file mode, UID, GID의 상위 비트와 author code를 보관합니다. `h_i_reserved1`의 의미는 문서에서 확정되지 않았습니다.
Hurd가 사용하는 12바이트 확장 필드입니다.
Masix 배치는 extended attribute block의 상위 16비트를 두고 나머지를 예약합니다. 원문 표는 offset `0x0`의 이름을 `h_i_reserved1`로 기록하고 있습니다.
Masix가 사용하는 12바이트 확장 필드입니다.
.. _i_osd2:
The ``osd2`` field has multiple meanings depending on the filesystem creator:
Linux:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le16
- l_i_blocks_high
- Upper 16-bits of the block count. Please see the note attached to
i_blocks_lo.
* - 0x2
- __le16
- l_i_file_acl_high
- Upper 16-bits of the extended attribute block (historically, the file
ACL location). See the Extended Attributes section below.
* - 0x4
- __le16
- l_i_uid_high
- Upper 16-bits of the Owner UID.
* - 0x6
- __le16
- l_i_gid_high
- Upper 16-bits of the GID.
* - 0x8
- __le16
- l_i_checksum_lo
- Lower 16-bits of the inode checksum.
* - 0xA
- __le16
- l_i_reserved
- Unused.
Hurd:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le16
- h_i_reserved1
- ??
* - 0x2
- __u16
- h_i_mode_high
- Upper 16-bits of the file mode.
* - 0x4
- __le16
- h_i_uid_high
- Upper 16-bits of the Owner UID.
* - 0x6
- __le16
- h_i_gid_high
- Upper 16-bits of the GID.
* - 0x8
- __u32
- h_i_author
- Author code?
Masix:
.. list-table::
:widths: 8 8 24 40
:header-rows: 1
* - Offset
- Size
- Name
- Description
* - 0x0
- __le16
- h_i_reserved1
- ??
* - 0x2
- __u16
- m_i_file_acl_high
- Upper 16-bits of the extended attribute block (historically, the file
ACL location).
* - 0x4
- __u32
- m_i_reserved2[2]
- ??
inode record 크기
460-480ext2와 ext3의 inode structure와 디스크 record는 `EXT2_GOOD_OLD_INODE_SIZE`인 128바이트로 고정되어 있었습니다.
ext4부터는 format 시 파일시스템의 모든 inode에 대해 더 큰 on-disk inode를 할당할 수 있습니다. record 전체 크기는 superblock의 `s_inode_size`에 기록합니다.
각 inode의 `i_extra_isize`는 원래 128바이트 ext2 inode 뒤에서 `struct ext4_inode`가 실제로 사용하는 바이트 수입니다. 새 커널이 structure를 확장해도 모든 기존 on-disk inode를 즉시 upgrade하지 않도록 하며, 128바이트 뒤 필드에 접근하기 전에 해당 필드가 `i_extra_isize` 범위에 있는지 확인해야 합니다.
기본 ext4 inode record는 256바이트입니다. 2019년 8월 기준 inode structure는 160바이트이고 `i_extra_isize = 32`입니다. structure 끝과 record 끝 사이의 여분에는 extended attribute를 저장할 수 있습니다.
inode record는 filesystem block size만큼 클 수 있지만 효율적이지는 않습니다.
128바이트 호환 영역 뒤의 record를 안전하게 해석하는 순서입니다.
Inode Size
~~~~~~~~~~
In ext2 and ext3, the inode structure size was fixed at 128 bytes
(``EXT2_GOOD_OLD_INODE_SIZE``) and each inode had a disk record size of
128 bytes. Starting with ext4, it is possible to allocate a larger
on-disk inode at format time for all inodes in the filesystem to provide
space beyond the end of the original ext2 inode. The on-disk inode
record size is recorded in the superblock as ``s_inode_size``. The
number of bytes actually used by struct ext4_inode beyond the original
128-byte ext2 inode is recorded in the ``i_extra_isize`` field for each
inode, which allows struct ext4_inode to grow for a new kernel without
having to upgrade all of the on-disk inodes. Access to fields beyond
EXT2_GOOD_OLD_INODE_SIZE should be verified to be within
``i_extra_isize``. By default, ext4 inode records are 256 bytes, and (as
of August 2019) the inode structure is 160 bytes
(``i_extra_isize = 32``). The extra space between the end of the inode
structure and the end of the inode record can be used to store extended
attributes. Each inode record can be as large as the filesystem block
size, though this is not terribly efficient.
inode 위치 계산
481-492각 block group에는 `sb->s_inodes_per_group`개의 inode가 있습니다. inode 0은 존재하지 않으므로 inode number에서 1을 뺀 뒤 group과 index를 계산합니다.
block group은 `bg = (inode_num - 1) / sb->s_inodes_per_group`, group의 inode table 안 index는 `index = (inode_num - 1) % sb->s_inodes_per_group`입니다. inode table 안 byte offset은 `offset = index * sb->s_inode_size`입니다.
inode number를 block group과 byte offset으로 바꾸는 세 단계입니다.
Finding an Inode
~~~~~~~~~~~~~~~~
Each block group contains ``sb->s_inodes_per_group`` inodes. Because
inode 0 is defined not to exist, this formula can be used to find the
block group that an inode lives in:
``bg = (inode_num - 1) / sb->s_inodes_per_group``. The particular inode
can be found within the block group's inode table at
``index = (inode_num - 1) % sb->s_inodes_per_group``. To get the byte
address within the inode table, use
``offset = index * sb->s_inode_size``.
inode timestamp 구성
493-518inode structure의 하위 128바이트에는 inode change time `ctime`, access time `atime`, data modification time `mtime`, deletion time `dtime` 네 timestamp가 기록됩니다.
네 필드는 Unix epoch인 `1970-01-01 00:00:00 GMT` 이후의 초를 나타내는 signed 32-bit integer이므로 2038년 1월에 overflow합니다.
filesystem에 `orphan_file` 기능이 없으면 어느 directory에서도 link되지 않았지만 여전히 열려 있는 orphan inode의 `dtime`을 orphan list 연결에 재사용합니다. superblock의 `s_last_orphan`이 첫 inode를 가리키고, 각 `dtime`은 다음 orphan inode number 또는 목록 끝의 0입니다.
`sb->s_inode_size > 128`이고 `i_inode_extra`가 해당 `i_[cma]time_extra` 필드를 포함할 만큼 크면 ctime, atime, mtime은 64비트로 확장됩니다. extra 32비트의 하위 2비트는 seconds field를 34비트로 넓히고, 상위 30비트는 nanosecond 정밀도를 제공합니다. 이 형식은 2446년 5월까지 overflow하지 않습니다.
`dtime`은 확장되지 않았습니다. 다섯 번째 timestamp인 creation time `crtime`은 64비트이며 64비트 `[cma]time`과 같은 방식으로 decode합니다. `crtime`과 `dtime`은 일반 `stat()` interface에서 볼 수 없지만 `debugfs`가 보고합니다.
기본 필드와 extra 필드, 접근 경로를 정리합니다.
Inode Timestamps
~~~~~~~~~~~~~~~~
Four timestamps are recorded in the lower 128 bytes of the inode
structure -- inode change time (ctime), access time (atime), data
modification time (mtime), and deletion time (dtime). The four fields
are 32-bit signed integers that represent seconds since the Unix epoch
(1970-01-01 00:00:00 GMT), which means that the fields will overflow in
January 2038. If the filesystem does not have orphan_file feature, inodes
that are not linked from any directory but are still open (orphan inodes) have
the dtime field overloaded for use with the orphan list. The superblock field
``s_last_orphan`` points to the first inode in the orphan list; dtime is then
the number of the next orphaned inode, or zero if there are no more orphans.
If the inode structure size ``sb->s_inode_size`` is larger than 128
bytes and the ``i_inode_extra`` field is large enough to encompass the
respective ``i_[cma]time_extra`` field, the ctime, atime, and mtime
inode fields are widened to 64 bits. Within this “extra” 32-bit field,
the lower two bits are used to extend the 32-bit seconds field to be 34
bit wide; the upper 30 bits are used to provide nanosecond timestamp
accuracy. Therefore, timestamps should not overflow until May 2446.
dtime was not widened. There is also a fifth timestamp to record inode
creation time (crtime); this field is 64-bits wide and decoded in the
same manner as 64-bit [cma]time. Neither crtime nor dtime are accessible
through the regular stat() interface, though debugfs will report them.
64비트 timestamp 해석
519-57864비트 seconds 값은 signed 32-bit time 값에 `(2^32 * extra epoch bits)`를 더해 계산합니다.
extra epoch bits는 두 비트지만 signed 32-bit time의 최상위 비트와 함께 해석하여 1901년부터 2446년까지의 구간을 표현합니다. 결과적으로 양수 시간 값이 음수 값보다 사실상 일곱 배 많아지는 다소 특이한 인코딩입니다.
extra epoch bits와 32비트 time의 MSB 조합별 64비트 범위입니다.
2038년 이후 날짜의 encode/decode에는 오랫동안 bug가 있었습니다. 문서 기준 kernel 3.12와 e2fsprogs 1.42.8에서도 수정되지 않은 것으로 보이며, 64-bit kernel은 1901년부터 1970년 사이 날짜에 extra epoch bits `1,1`을 잘못 사용합니다.
향후 kernel이 수정되면 e2fsck가 이 상태를 바로잡을 예정이며, 2310년 이전에 e2fsck를 실행한다는 전제가 붙습니다.
We use the 32-bit signed time value plus (2^32 * (extra epoch bits)).
In other words:
.. list-table::
:widths: 20 20 20 20 20
:header-rows: 1
* - Extra epoch bits
- MSB of 32-bit time
- Adjustment for signed 32-bit to 64-bit tv_sec
- Decoded 64-bit tv_sec
- valid time range
* - 0 0
- 1
- 0
- ``-0x80000000 - -0x00000001``
- 1901-12-13 to 1969-12-31
* - 0 0
- 0
- 0
- ``0x000000000 - 0x07fffffff``
- 1970-01-01 to 2038-01-19
* - 0 1
- 1
- 0x100000000
- ``0x080000000 - 0x0ffffffff``
- 2038-01-19 to 2106-02-07
* - 0 1
- 0
- 0x100000000
- ``0x100000000 - 0x17fffffff``
- 2106-02-07 to 2174-02-25
* - 1 0
- 1
- 0x200000000
- ``0x180000000 - 0x1ffffffff``
- 2174-02-25 to 2242-03-16
* - 1 0
- 0
- 0x200000000
- ``0x200000000 - 0x27fffffff``
- 2242-03-16 to 2310-04-04
* - 1 1
- 1
- 0x300000000
- ``0x280000000 - 0x2ffffffff``
- 2310-04-04 to 2378-04-22
* - 1 1
- 0
- 0x300000000
- ``0x300000000 - 0x37fffffff``
- 2378-04-22 to 2446-05-10
This is a somewhat odd encoding since there are effectively seven times
as many positive values as negative values. There have also been
long-standing bugs decoding and encoding dates beyond 2038, which don't
seem to be fixed as of kernel 3.12 and e2fsprogs 1.42.8. 64-bit kernels
incorrectly use the extra epoch bits 1,1 for dates between 1901 and
1970. At some point the kernel will be fixed and e2fsck will fix this
situation, assuming that it is run before 2310.
요약·해설
inodes.rst:1-578ext4 inode는 파일의 mode, 소유권, 크기, block mapping, xattr 위치, flag, checksum과 다섯 종류의 timestamp를 보관합니다. inode number에서 block group과 table index를 계산하며 inode 0은 존재하지 않습니다.
기본 128바이트 형식 뒤의 확장 필드는 `s_inode_size`와 `i_extra_isize`로 유효 범위를 판단합니다. creator OS에 따라 `i_osd1`과 `i_osd2`의 의미가 다르고, Linux 배치는 UID/GID·block count·xattr block·checksum의 상위 또는 하위 비트를 보충합니다.
ctime, atime, mtime은 extra 필드로 34비트 seconds와 30비트 nanoseconds를 표현해 2446년까지 확장되지만, dtime은 32비트이며 orphan list 연결에 재사용될 수 있습니다.
on-disk inode를 안전하게 읽고 검증하는 전체 흐름입니다.