Debian Codegraph @

upstream/debian-unstable/man7/xattr.7

raw
1 .\" Copyright 2000-2007, Andreas Grünbacher <[email protected]>
2 .\" Copyright 2001-2007, Silicon Graphics, Inc.
3 .\" Copyright, the authors of the Linux man-pages project
4 .\"
5 .\" SPDX-License-Identifier: GPL-2.0-or-later
6 .\"
7 .TH xattr 7 2026-02-08 "Linux man-pages 6.18"
8 .SH NAME
9 xattr \- Extended attributes
10 .SH DESCRIPTION
11 Extended attributes are name:value pairs associated permanently with
12 files and directories, similar to the environment strings associated
13 with a process.
14 An attribute may be defined or undefined.
15 If it is defined, its value may be empty or non-empty.
16 .P
17 Extended attributes are extensions to the normal attributes which are
18 associated with all inodes in the system (i.e., the
19 .BR stat (2)
20 data).
21 They are often used to provide additional functionality
22 to a filesystem\[em]for example, additional security features such as
23 Access Control Lists (ACLs) may be implemented using extended attributes.
24 .P
25 Users with search access to a file or directory may use
26 .BR listxattr (2)
27 to retrieve a list of attribute names defined for that file or directory.
28 .P
29 Extended attributes are accessed as atomic objects.
30 Reading
31 .RB ( getxattr (2))
32 retrieves the whole value of an attribute and stores it in a buffer.
33 Writing
34 .RB ( setxattr (2))
35 replaces any previous value with the new value.
36 .P
37 Space consumed for extended attributes may be counted towards the disk quotas
38 of the file owner and file group.
39 .SS Extended attribute namespaces
40 Attribute names are null-terminated strings.
41 The attribute name is always specified in the fully qualified
42 .I namespace.attribute
43 form, for example,
44 .IR user.mime_type ,
45 .IR trusted.md5sum ,
46 .IR system.posix_acl_access ,
47 or
48 .IR security.selinux .
49 .P
50 The namespace mechanism is used to define different classes of extended
51 attributes.
52 These different classes exist for several reasons;
53 for example, the permissions
54 and capabilities required for manipulating extended attributes of one
55 namespace may differ to another.
56 .P
57 Currently, the
58 .IR security ,
59 .IR system ,
60 .IR trusted ,
61 and
62 .I user
63 extended attribute classes are defined as described below.
64 Additional classes may be added in the future.
65 .SS Extended security attributes
66 The security attribute namespace is used by kernel security modules,
67 such as Security Enhanced Linux, and also to implement file capabilities (see
68 .BR capabilities (7)).
69 Read and write access permissions to security attributes depend on the
70 policy implemented for each security attribute by the security module.
71 When no security module is loaded, all processes have read access to
72 extended security attributes, and write access is limited to processes
73 that have the
74 .B CAP_SYS_ADMIN
75 capability.
76 .SS System extended attributes
77 System extended attributes are used by the kernel to store system
78 objects such as Access Control Lists.
79 Read and write
80 access permissions to system attributes depend on the policy implemented
81 for each system attribute implemented by filesystems in the kernel.
82 .SS Trusted extended attributes
83 Trusted extended attributes are visible and accessible only to processes that
84 have the
85 .B CAP_SYS_ADMIN
86 capability.
87 Attributes in this class are used to implement mechanisms in user
88 space (i.e., outside the kernel) which keep information in extended attributes
89 to which ordinary processes should not have access.
90 .SS User extended attributes
91 User extended attributes may be assigned to files and directories for
92 storing arbitrary additional information such as the mime type,
93 character set or encoding of a file.
94 The access permissions for user
95 attributes are defined by the file permission bits:
96 read permission is required to retrieve the attribute value,
97 and writer permission is required to change it.
98 .P
99 The file permission bits of regular files and directories are
100 interpreted differently from the file permission bits of special files
101 and symbolic links.
102 For regular files and directories the file
103 permission bits define access to the file's contents, while for device special
104 files they define access to the device described by the special file.
105 The file permissions of symbolic links are not used in access checks.
106 These differences would allow users to consume filesystem resources in
107 a way not controllable by disk quotas for group or world writable
108 special files and directories.
109 .P
110 For this reason,
111 user extended attributes are allowed only for regular files and directories,
112 and access to user extended attributes is restricted to the
113 owner and to users with appropriate capabilities for directories with the
114 sticky bit set (see the
115 .BR chmod (1)
116 manual page for an explanation of the sticky bit).
117 .SS Filesystem differences
118 The kernel and the filesystem may place limits on the maximum number
119 and size of extended attributes that can be associated with a file.
120 The VFS-imposed limits on attribute names and values are 255 bytes
121 and 64\ kB, respectively.
122 The list of attribute names that
123 can be returned is also limited to 64\ kB
124 (see BUGS in
125 .BR listxattr (2)).
126 .P
127 Some filesystems, such as Reiserfs (and, historically, ext2 and ext3),
128 require the filesystem to be mounted with the
129 .B user_xattr
130 mount option in order for user extended attributes to be used.
131 .P
132 In the current ext2, ext3, and ext4 filesystem implementations,
133 the total bytes used by the names and values of all of a file's
134 extended attributes must fit in a single filesystem block (1024, 2048
135 or 4096 bytes, depending on the block size specified when the
136 filesystem was created).
137 .P
138 In the Btrfs, XFS, and Reiserfs filesystem implementations, there is no
139 practical limit on the number of extended attributes
140 associated with a file, and the algorithms used to store extended
141 attribute information on disk are scalable.
142 .P
143 In the JFS, XFS, and Reiserfs filesystem implementations,
144 the limit on bytes used in an EA value is the ceiling imposed by the VFS.
145 .P
146 In the Btrfs filesystem implementation,
147 the total bytes used for the name, value, and implementation overhead bytes
148 is limited to the filesystem
149 .I nodesize
150 value (16\ kB by default).
151 .SH STANDARDS
152 Extended attributes are not specified in POSIX.1, but some other systems
153 (e.g., the BSDs and Solaris) provide a similar feature.
154 .SH NOTES
155 Since the filesystems on which extended attributes are stored might also
156 be used on architectures with a different byte order and machine word
157 size, care should be taken to store attribute values in an
158 architecture-independent format.
159 .P
160 This page was formerly named
161 .BR attr (5).
162 .\" .SH AUTHORS
163 .\" Andreas Gruenbacher,
164 .\" .RI < [email protected] >
165 .\" and the SGI XFS development team,
166 .\" .RI < [email protected] >.
167 .SH SEE ALSO
168 .BR attr (1),
169 .BR getfattr (1),
170 .BR setfattr (1),
171 .BR getxattr (2),
172 .BR FS_IOC_GETFLAGS (2const),
173 .BR FS_IOC_SETFLAGS (2const),
174 .BR listxattr (2),
175 .BR removexattr (2),
176 .BR setxattr (2),
177 .BR acl (5),
178 .BR capabilities (7),
179 .BR selinux (8)