1 .. SPDX-License-Identifier: GPL-2.0 1 .. SPDX-License-Identifier: GPL-2.0
2 2
3 Extended Attributes 3 Extended Attributes
4 ------------------- 4 -------------------
5 5
6 Extended attributes (xattrs) are typically sto 6 Extended attributes (xattrs) are typically stored in a separate data
7 block on the disk and referenced from inodes v 7 block on the disk and referenced from inodes via ``inode.i_file_acl*``.
8 The first use of extended attributes seems to 8 The first use of extended attributes seems to have been for storing file
9 ACLs and other security data (selinux). With t 9 ACLs and other security data (selinux). With the ``user_xattr`` mount
10 option it is possible for users to store exten 10 option it is possible for users to store extended attributes so long as
11 all attribute names begin with “user”; thi 11 all attribute names begin with “user”; this restriction seems to have
12 disappeared as of Linux 3.0. 12 disappeared as of Linux 3.0.
13 13
14 There are two places where extended attributes 14 There are two places where extended attributes can be found. The first
15 place is between the end of each inode entry a 15 place is between the end of each inode entry and the beginning of the
16 next inode entry. For example, if inode.i_extr 16 next inode entry. For example, if inode.i_extra_isize = 28 and
17 sb.inode_size = 256, then there are 256 - (128 17 sb.inode_size = 256, then there are 256 - (128 + 28) = 100 bytes
18 available for in-inode extended attribute stor 18 available for in-inode extended attribute storage. The second place
19 where extended attributes can be found is in t 19 where extended attributes can be found is in the block pointed to by
20 ``inode.i_file_acl``. As of Linux 3.11, it is 20 ``inode.i_file_acl``. As of Linux 3.11, it is not possible for this
21 block to contain a pointer to a second extende 21 block to contain a pointer to a second extended attribute block (or even
22 the remaining blocks of a cluster). In theory 22 the remaining blocks of a cluster). In theory it is possible for each
23 attribute's value to be stored in a separate d 23 attribute's value to be stored in a separate data block, though as of
24 Linux 3.11 the code does not permit this. 24 Linux 3.11 the code does not permit this.
25 25
26 Keys are generally assumed to be ASCIIZ string 26 Keys are generally assumed to be ASCIIZ strings, whereas values can be
27 strings or binary data. 27 strings or binary data.
28 28
29 Extended attributes, when stored after the ino 29 Extended attributes, when stored after the inode, have a header
30 ``ext4_xattr_ibody_header`` that is 4 bytes lo 30 ``ext4_xattr_ibody_header`` that is 4 bytes long:
31 31
32 .. list-table:: 32 .. list-table::
33 :widths: 8 8 24 40 33 :widths: 8 8 24 40
34 :header-rows: 1 34 :header-rows: 1
35 35
36 * - Offset 36 * - Offset
37 - Type 37 - Type
38 - Name 38 - Name
39 - Description 39 - Description
40 * - 0x0 40 * - 0x0
41 - __le32 41 - __le32
42 - h_magic 42 - h_magic
43 - Magic number for identification, 0xEA02 43 - Magic number for identification, 0xEA020000. This value is set by the
44 Linux driver, though e2fsprogs doesn't 44 Linux driver, though e2fsprogs doesn't seem to check it(?)
45 45
46 The beginning of an extended attribute block i 46 The beginning of an extended attribute block is in
47 ``struct ext4_xattr_header``, which is 32 byte 47 ``struct ext4_xattr_header``, which is 32 bytes long:
48 48
49 .. list-table:: 49 .. list-table::
50 :widths: 8 8 24 40 50 :widths: 8 8 24 40
51 :header-rows: 1 51 :header-rows: 1
52 52
53 * - Offset 53 * - Offset
54 - Type 54 - Type
55 - Name 55 - Name
56 - Description 56 - Description
57 * - 0x0 57 * - 0x0
58 - __le32 58 - __le32
59 - h_magic 59 - h_magic
60 - Magic number for identification, 0xEA02 60 - Magic number for identification, 0xEA020000.
61 * - 0x4 61 * - 0x4
62 - __le32 62 - __le32
63 - h_refcount 63 - h_refcount
64 - Reference count. 64 - Reference count.
65 * - 0x8 65 * - 0x8
66 - __le32 66 - __le32
67 - h_blocks 67 - h_blocks
68 - Number of disk blocks used. 68 - Number of disk blocks used.
69 * - 0xC 69 * - 0xC
70 - __le32 70 - __le32
71 - h_hash 71 - h_hash
72 - Hash value of all attributes. 72 - Hash value of all attributes.
73 * - 0x10 73 * - 0x10
74 - __le32 74 - __le32
75 - h_checksum 75 - h_checksum
76 - Checksum of the extended attribute bloc 76 - Checksum of the extended attribute block.
77 * - 0x14 77 * - 0x14
78 - __u32 78 - __u32
79 - h_reserved[3] 79 - h_reserved[3]
80 - Zero. 80 - Zero.
81 81
82 The checksum is calculated against the FS UUID 82 The checksum is calculated against the FS UUID, the 64-bit block number
83 of the extended attribute block, and the entir 83 of the extended attribute block, and the entire block (header +
84 entries). 84 entries).
85 85
86 Following the ``struct ext4_xattr_header`` or 86 Following the ``struct ext4_xattr_header`` or
87 ``struct ext4_xattr_ibody_header`` is an array 87 ``struct ext4_xattr_ibody_header`` is an array of
88 ``struct ext4_xattr_entry``; each of these ent 88 ``struct ext4_xattr_entry``; each of these entries is at least 16 bytes
89 long. When stored in an external block, the `` 89 long. When stored in an external block, the ``struct ext4_xattr_entry``
90 entries must be stored in sorted order. The so 90 entries must be stored in sorted order. The sort order is
91 ``e_name_index``, then ``e_name_len``, and fin 91 ``e_name_index``, then ``e_name_len``, and finally ``e_name``.
92 Attributes stored inside an inode do not need 92 Attributes stored inside an inode do not need be stored in sorted order.
93 93
94 .. list-table:: 94 .. list-table::
95 :widths: 8 8 24 40 95 :widths: 8 8 24 40
96 :header-rows: 1 96 :header-rows: 1
97 97
98 * - Offset 98 * - Offset
99 - Type 99 - Type
100 - Name 100 - Name
101 - Description 101 - Description
102 * - 0x0 102 * - 0x0
103 - __u8 103 - __u8
104 - e_name_len 104 - e_name_len
105 - Length of name. 105 - Length of name.
106 * - 0x1 106 * - 0x1
107 - __u8 107 - __u8
108 - e_name_index 108 - e_name_index
109 - Attribute name index. There is a discus 109 - Attribute name index. There is a discussion of this below.
110 * - 0x2 110 * - 0x2
111 - __le16 111 - __le16
112 - e_value_offs 112 - e_value_offs
113 - Location of this attribute's value on t 113 - Location of this attribute's value on the disk block where it is stored.
114 Multiple attributes can share the same 114 Multiple attributes can share the same value. For an inode attribute
115 this value is relative to the start of 115 this value is relative to the start of the first entry; for a block this
116 value is relative to the start of the b 116 value is relative to the start of the block (i.e. the header).
117 * - 0x4 117 * - 0x4
118 - __le32 118 - __le32
119 - e_value_inum 119 - e_value_inum
120 - The inode where the value is stored. Ze 120 - The inode where the value is stored. Zero indicates the value is in the
121 same block as this entry. This field is 121 same block as this entry. This field is only used if the
122 INCOMPAT_EA_INODE feature is enabled. 122 INCOMPAT_EA_INODE feature is enabled.
123 * - 0x8 123 * - 0x8
124 - __le32 124 - __le32
125 - e_value_size 125 - e_value_size
126 - Length of attribute value. 126 - Length of attribute value.
127 * - 0xC 127 * - 0xC
128 - __le32 128 - __le32
129 - e_hash 129 - e_hash
130 - Hash value of attribute name and attrib 130 - Hash value of attribute name and attribute value. The kernel doesn't
131 update the hash for in-inode attributes 131 update the hash for in-inode attributes, so for that case this value
132 must be zero, because e2fsck validates 132 must be zero, because e2fsck validates any non-zero hash regardless of
133 where the xattr lives. 133 where the xattr lives.
134 * - 0x10 134 * - 0x10
135 - char 135 - char
136 - e_name[e_name_len] 136 - e_name[e_name_len]
137 - Attribute name. Does not include traili 137 - Attribute name. Does not include trailing NULL.
138 138
139 Attribute values can follow the end of the ent 139 Attribute values can follow the end of the entry table. There appears to
140 be a requirement that they be aligned to 4-byt 140 be a requirement that they be aligned to 4-byte boundaries. The values
141 are stored starting at the end of the block an 141 are stored starting at the end of the block and grow towards the
142 xattr_header/xattr_entry table. When the two c 142 xattr_header/xattr_entry table. When the two collide, the overflow is
143 put into a separate disk block. If the disk bl 143 put into a separate disk block. If the disk block fills up, the
144 filesystem returns -ENOSPC. 144 filesystem returns -ENOSPC.
145 145
146 The first four fields of the ``ext4_xattr_entr 146 The first four fields of the ``ext4_xattr_entry`` are set to zero to
147 mark the end of the key list. 147 mark the end of the key list.
148 148
149 Attribute Name Indices 149 Attribute Name Indices
150 ~~~~~~~~~~~~~~~~~~~~~~ 150 ~~~~~~~~~~~~~~~~~~~~~~
151 151
152 Logically speaking, extended attributes are a 152 Logically speaking, extended attributes are a series of key=value pairs.
153 The keys are assumed to be NULL-terminated str 153 The keys are assumed to be NULL-terminated strings. To reduce the amount
154 of on-disk space that the keys consume, the be 154 of on-disk space that the keys consume, the beginning of the key string
155 is matched against the attribute name index. I 155 is matched against the attribute name index. If a match is found, the
156 attribute name index field is set, and matchin 156 attribute name index field is set, and matching string is removed from
157 the key name. Here is a map of name index valu 157 the key name. Here is a map of name index values to key prefixes:
158 158
159 .. list-table:: 159 .. list-table::
160 :widths: 16 64 160 :widths: 16 64
161 :header-rows: 1 161 :header-rows: 1
162 162
163 * - Name Index 163 * - Name Index
164 - Key Prefix 164 - Key Prefix
165 * - 0 165 * - 0
166 - (no prefix) 166 - (no prefix)
167 * - 1 167 * - 1
168 - “user.” 168 - “user.”
169 * - 2 169 * - 2
170 - “system.posix_acl_access” 170 - “system.posix_acl_access”
171 * - 3 171 * - 3
172 - “system.posix_acl_default” 172 - “system.posix_acl_default”
173 * - 4 173 * - 4
174 - “trusted.” 174 - “trusted.”
175 * - 6 175 * - 6
176 - “security.” 176 - “security.”
177 * - 7 177 * - 7
178 - “system.” (inline_data only?) 178 - “system.” (inline_data only?)
179 * - 8 179 * - 8
180 - “system.richacl” (SuSE kernels only 180 - “system.richacl” (SuSE kernels only?)
181 181
182 For example, if the attribute key is “user.f 182 For example, if the attribute key is “user.fubar”, the attribute name
183 index is set to 1 and the “fubar” name is 183 index is set to 1 and the “fubar” name is recorded on disk.
184 184
185 POSIX ACLs 185 POSIX ACLs
186 ~~~~~~~~~~ 186 ~~~~~~~~~~
187 187
188 POSIX ACLs are stored in a reduced version of 188 POSIX ACLs are stored in a reduced version of the Linux kernel (and
189 libacl's) internal ACL format. The key differe 189 libacl's) internal ACL format. The key difference is that the version
190 number is different (1) and the ``e_id`` field 190 number is different (1) and the ``e_id`` field is only stored for named
191 user and group ACLs. 191 user and group ACLs.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.