1PATH_RESOLUTION(7) Linux Programmer's Manual PATH_RESOLUTION(7)
2
6 path_resolution - how a pathname is resolved to a file
7
9 Some UNIX/Linux system calls have as parameter one or more filenames.
10 A filename (or pathname) is resolved as follows.
11 Step 1: start of the resolution process
13 If the pathname starts with the '/' character, the starting lookup
14 directory is the root directory of the calling process. (A process
15 inherits its root directory from its parent. Usually this will be the
16 root directory of the file hierarchy. A process may get a different
17 root directory by use of the chroot(2) system call. A process may get
18 an entirely private mount namespace in case it—or one of its ancestors—
19 was started by an invocation of the clone(2) system call that had the
20 CLONE_NEWNS flag set.) This handles the '/' part of the pathname.
21 If the pathname does not start with the '/' character, the starting
23 lookup directory of the resolution process is the current working
24 directory of the process. (This is also inherited from the parent. It
25 can be changed by use of the chdir(2) system call.)
26 Pathnames starting with a '/' character are called absolute pathnames.
28 Pathnames not starting with a '/' are called relative pathnames.
29 Step 2: walk along the path
31 Set the current lookup directory to the starting lookup directory.
32 Now, for each nonfinal component of the pathname, where a component is
33 a substring delimited by '/' characters, this component is looked up in
34 the current lookup directory.
35 If the process does not have search permission on the current lookup
37 directory, an EACCES error is returned ("Permission denied").
38 If the component is not found, an ENOENT error is returned ("No such
40 file or directory").
41 If the component is found, but is neither a directory nor a symbolic
43 link, an ENOTDIR error is returned ("Not a directory").
44 If the component is found and is a directory, we set the current lookup
46 directory to that directory, and go to the next component.
47 If the component is found and is a symbolic link (symlink), we first
49 resolve this symbolic link (with the current lookup directory as start‐
50 ing lookup directory). Upon error, that error is returned. If the
51 result is not a directory, an ENOTDIR error is returned. If the reso‐
52 lution of the symlink is successful and returns a directory, we set the
53 current lookup directory to that directory, and go to the next compo‐
54 nent. Note that the resolution process here involves recursion. In
55 order to protect the kernel against stack overflow, and also to protect
56 against denial of service, there are limits on the maximum recursion
57 depth, and on the maximum number of symbolic links followed. An ELOOP
58 error is returned when the maximum is exceeded ("Too many levels of
59 symbolic links").
60 Step 3: find the final entry
62 The lookup of the final component of the pathname goes just like that
63 of all other components, as described in the previous step, with two
64 differences: (i) the final component need not be a directory (at least
65 as far as the path resolution process is concerned—it may have to be a
66 directory, or a nondirectory, because of the requirements of the spe‐
67 cific system call), and (ii) it is not necessarily an error if the com‐
68 ponent is not found—maybe we are just creating it. The details on the
69 treatment of the final entry are described in the manual pages of the
70 specific system calls.
71 . and ..
73 By convention, every directory has the entries "." and "..", which
74 refer to the directory itself and to its parent directory, respec‐
75 tively.
76 The path resolution process will assume that these entries have their
78 conventional meanings, regardless of whether they are actually present
79 in the physical file system.
80 One cannot walk down past the root: "/.." is the same as "/".
82 Mount points
84 After a "mount dev path" command, the pathname "path" refers to the
85 root of the file system hierarchy on the device "dev", and no longer to
86 whatever it referred to earlier.
87 One can walk out of a mounted file system: "path/.." refers to the par‐
89 ent directory of "path", outside of the file system hierarchy on "dev".
90 Trailing slashes
92 If a pathname ends in a '/', that forces resolution of the preceding
93 component as in Step 2: it has to exist and resolve to a directory.
94 Otherwise a trailing '/' is ignored. (Or, equivalently, a pathname
95 with a trailing '/' is equivalent to the pathname obtained by appending
96 '.' to it.)
97 Final symlink
99 If the last component of a pathname is a symbolic link, then it depends
100 on the system call whether the file referred to will be the symbolic
101 link or the result of path resolution on its contents. For example,
102 the system call lstat(2) will operate on the symlink, while stat(2)
103 operates on the file pointed to by the symlink.
104 Length limit
106 There is a maximum length for pathnames. If the pathname (or some
107 intermediate pathname obtained while resolving symbolic links) is too
108 long, an ENAMETOOLONG error is returned ("Filename too long").
109 Empty pathname
111 In the original UNIX, the empty pathname referred to the current direc‐
112 tory. Nowadays POSIX decrees that an empty pathname must not be
113 resolved successfully. Linux returns ENOENT in this case.
114 Permissions
116 The permission bits of a file consist of three groups of three bits,
117 cf. chmod(1) and stat(2). The first group of three is used when the
118 effective user ID of the calling process equals the owner ID of the
119 file. The second group of three is used when the group ID of the file
120 either equals the effective group ID of the calling process, or is one
121 of the supplementary group IDs of the calling process (as set by set‐
122 groups(2)). When neither holds, the third group is used.
123 Of the three bits used, the first bit determines read permission, the
125 second write permission, and the last execute permission in case of
126 ordinary files, or search permission in case of directories.
127 Linux uses the fsuid instead of the effective user ID in permission
129 checks. Ordinarily the fsuid will equal the effective user ID, but the
130 fsuid can be changed by the system call setfsuid(2).
131 (Here "fsuid" stands for something like "file system user ID". The
133 concept was required for the implementation of a user space NFS server
134 at a time when processes could send a signal to a process with the same
135 effective user ID. It is obsolete now. Nobody should use setf‐
136 suid(2).)
137 Similarly, Linux uses the fsgid ("file system group ID") instead of the
139 effective group ID. See setfsgid(2).
140 Bypassing permission checks: superuser and capabilities
142 On a traditional UNIX system, the superuser (root, user ID 0) is all-
143 powerful, and bypasses all permissions restrictions when accessing
144 files.
145 On Linux, superuser privileges are divided into capabilities (see capa‐
147 bilities(7)). Two capabilities are relevant for file permissions
148 checks: CAP_DAC_OVERRIDE and CAP_DAC_READ_SEARCH. (A process has these
149 capabilities if its fsuid is 0.)
150 The CAP_DAC_OVERRIDE capability overrides all permission checking, but
152 grants execute permission only when at least one of the file's three
153 execute permission bits is set.
154 The CAP_DAC_READ_SEARCH capability grants read and search permission on
156 directories, and read permission on ordinary files.
157
159 readlink(2), capabilities(7), credentials(7), symlink(7)
160
162 This page is part of release 3.53 of the Linux man-pages project. A
163 description of the project, and information about reporting bugs, can
164 be found at http://www.kernel.org/doc/man-pages/.
165Linux 2009-12-05 PATH_RESOLUTION(7)