#include <unistd.h> int unlink(const char *path); #include <fcntl.h> int unlinkat(int fd, const char *path, int flag);
When the file's link count becomes 0 and no process has the file open, the space occupied by the file shall be freed and the file shall no longer be accessible. If one or more processes have the file open when the last link is removed, the link shall be removed before unlink() returns, but the removal of the file contents shall be postponed until all references to the file are closed.
The path argument shall not name a directory unless the process has appropriate privileges and the implementation supports using unlink() on directories.
Upon successful completion, unlink() shall mark for update the last data modification and last file status change timestamps of the parent directory. Also, if the file's link count is not 0, the last file status change timestamp of the file shall be marked for update.
The unlinkat() function shall be equivalent to the unlink() or rmdir() function except in the case where path specifies a relative path. In this case the directory entry to be removed is determined relative to the directory associated with the file descriptor fd instead of the current working directory. If the access mode of the open file description associated with the file descriptor is not O_SEARCH, the function shall check whether directory searches are permitted using the current permissions of the directory underlying the file descriptor. If the access mode is O_SEARCH, the function shall not perform the check.
Values for flag are constructed by a bitwise-inclusive OR of flags from the following list, defined in <fcntl.h>:
If unlinkat() is passed the special value AT_FDCWD in the fd parameter, the current working directory shall be used and the behavior shall be identical to a call to unlink() or rmdir() respectively, depending on whether or not the AT_REMOVEDIR bit is set in flag.
The unlinkat() function shall fail if:
These functions may fail and not unlink the file if:
The unlinkat() function may fail if:
The following sections are informative.
The following example shows how to remove a link to a file named /home/cnd/mod1 by removing the entry named /modules/pass1.
#include <unistd.h> char *path = "/modules/pass1"; int status; ... status = unlink(path);
The following example fragment creates a temporary password lock file named LOCKFILE, which is defined as /etc/ptmp, and gets a file descriptor for it. If the file cannot be opened for writing, unlink() is used to remove the link between the file descriptor and LOCKFILE.
#include <sys/types.h> #include <stdio.h> #include <fcntl.h> #include <errno.h> #include <unistd.h> #include <sys/stat.h> #define LOCKFILE "/etc/ptmp" int pfd; /* Integer for file descriptor returned by open call. */ FILE *fpfd; /* File pointer for use in putpwent(). */ ... /* Open password Lock file. If it exists, this is an error. */ if ((pfd = open(LOCKFILE, O_WRONLY| O_CREAT | O_EXCL, S_IRUSR | S_IWUSR | S_IRGRP | S_IROTH)) == -1) { fprintf(stderr, "Cannot open /etc/ptmp. Try again later.\n"); exit(1); } /* Lock file created; proceed with fdopen of lock file so that putpwent() can be used. */ if ((fpfd = fdopen(pfd, "w")) == NULL) { close(pfd); unlink(LOCKFILE); exit(1); }
The following example fragment uses unlink() to discard links to files, so that they can be replaced with new versions of the files. The first call removes the link to LOCKFILE if an error occurs. Successive calls remove the links to SAVEFILE and PASSWDFILE so that new links can be created, then removes the link to LOCKFILE when it is no longer needed.
#include <sys/types.h> #include <stdio.h> #include <fcntl.h> #include <errno.h> #include <unistd.h> #include <sys/stat.h> #define LOCKFILE "/etc/ptmp" #define PASSWDFILE "/etc/passwd" #define SAVEFILE "/etc/opasswd" ... /* If no change was made, assume error and leave passwd unchanged. */ if (!valid_change) { fprintf(stderr, "Could not change password for user %s\n", user); unlink(LOCKFILE); exit(1); } /* Change permissions on new password file. */ chmod(LOCKFILE, S_IRUSR | S_IRGRP | S_IROTH); /* Remove saved password file. */ unlink(SAVEFILE); /* Save current password file. */ link(PASSWDFILE, SAVEFILE); /* Remove current password file. */ unlink(PASSWDFILE); /* Save new password file as current password file. */ link(LOCKFILE,PASSWDFILE); /* Remove lock file. */ unlink(LOCKFILE); exit(0);
The meaning of [EBUSY] in historical implementations is ``mount point busy''. Since this volume of POSIX.1-2017 does not cover the system administration concepts of mounting and unmounting, the description of the error was changed to ``resource busy''. (This meaning is used by some device drivers when a second process tries to open an exclusive use device.) The wording is also intended to allow implementations to refuse to remove a directory if it is the root or current working directory of any process.
The standard developers reviewed TR 24715-2006 and noted that LSB-conforming implementations may return [EISDIR] instead of [EPERM] when unlinking a directory. A change to permit this behavior by changing the requirement for [EPERM] to [EPERM] or [EISDIR] was considered, but decided against since it would break existing strictly conforming and conforming applications. Applications written for portability to both POSIX.1-2008 and the LSB should be prepared to handle either error code.
The purpose of the unlinkat() function is to remove directory entries in directories other than the current working directory without exposure to race conditions. Any part of the path of a file could be changed in parallel to a call to unlink(), resulting in unspecified behavior. By opening a file descriptor for the target directory and using the unlinkat() function it can be guaranteed that the removed directory entry is located relative to the desired directory.
The Base Definitions volume of POSIX.1-2017, Section 4.3, Directory Protection, <fcntl.h>, <unistd.h>
Any typographical or formatting errors that appear in this page are most likely to have been introduced during the conversion of the source files to man page format. To report such errors, see https://www.kernel.org/doc/man-pages/reporting_bugs.html .