MKNOD(2) Linux Programmer's Manual MKNOD(2) NAME mknod(1,2) - create a special or ordinary file(1,n) SYNOPSIS #include <sys/types.h> #include <sys/stat.h> #include <fcntl.h> #include <unistd.h> int mknod(1,2)(const char *pathname, mode_t mode, dev_t dev); DESCRIPTION The system call mknod(1,2) creates a filesystem node (file(1,n), device special file(1,n) or named(5,8) pipe(2,8)) named(5,8) pathname, with attributes specified by mode and dev. The mode argument specifies both the permissions to use and the type of node to be created. It should be a combination (using bitwise OR) of one of the file(1,n) types listed below and the permissions for the new node. The permissions are modified by the process's umask in(1,8) the usual way: the permissions of the created node are (mode & ~umask). The file(1,n) type must be one of S_IFREG, S_IFCHR, S_IFBLK, S_IFIFO or S_IFSOCK to specify a normal file(1,n) (which will be created empty), char- acter special file(1,n), block special file(1,n), FIFO (named(5,8) pipe(2,8)), or Unix domain socket(2,7,n), respectively. (Zero file(1,n) type is equivalent to type S_IFREG.) If the file(1,n) type is S_IFCHR or S_IFBLK then dev specifies the major and minor numbers of the newly created device special file(1,n); otherwise it is ignored. If pathname already exists, or is a symbolic link(1,2), this call fails with an EEXIST error. The newly created node will be owned by the effective user ID of the process. If the directory containing the node has the set(7,n,1 builtins) group ID bit set(7,n,1 builtins), or if(3,n) the filesystem is mounted with BSD group semantics, the new node will inherit the group ownership from its parent directory; other- wise it will be owned by the effective group ID of the process. RETURN VALUE mknod(1,2) returns zero on success, or -1 if(3,n) an error(8,n) occurred (in(1,8) which case, errno is set(7,n,1 builtins) appropriately). ERRORS EACCES The parent directory does not allow write(1,2) permission to the process, or one of the directories in(1,8) the path prefix of path- name did not allow search permission. (See also path_resolu- tion(2).) EEXIST pathname already exists. EFAULT pathname points outside your accessible address space. EINVAL mode requested creation of something other than a normal file(1,n), device special file(1,n), FIFO or socket. ELOOP Too many symbolic links were encountered in(1,8) resolving pathname. ENAMETOOLONG pathname was too long. ENOENT A directory component in(1,8) pathname does not exist or is a dan- gling symbolic link. ENOMEM Insufficient kernel memory was available. ENOSPC The device containing pathname has no room for the new node. ENOTDIR A component used as a directory in(1,8) pathname is not, in(1,8) fact, a directory. EPERM mode requested creation of something other than a regular file(1,n), FIFO (named(5,8) pipe(2,8)), or Unix domain socket(2,7,n), and the caller is not privileged (Linux: does not have the CAP_MKNOD capability); also returned if(3,n) the filesystem containing pathname does not support the type of node requested. EROFS pathname refers to a file(1,n) on a read-only filesystem. CONFORMING TO SVr4 (but the call requires privilege and is thus not in(1,8) POSIX), 4.4BSD. The Linux version(1,3,5) differs from the SVr4 version(1,3,5) in(1,8) that it does not require root permission to create pipes, also in(1,8) that no EMUL- TIHOP, ENOLINK, or EINTR error(8,n) is documented. NOTES POSIX 1003.1-2001 says: "The only portable use of mknod(1,2)() is to create a FIFO-special file. If mode is not S_IFIFO or dev is not 0, the behav- ior of mknod(1,2)() is unspecified." Under Linux, this call cannot be used to create directories. One should make directories with mkdir(1,2), and FIFOs with mkfifo(1,3). There are many infelicities in(1,8) the protocol underlying NFS. Some of these affect mknod(1,2). SEE ALSO fcntl(2), mkdir(1,2)(2), mount(2,8)(2), path_resolution(2), socket(2,7,n)(2), stat(1,2)(2), umask(2), unlink(1,2)(2), mkfifo(1,3)(3) Linux 2.6.7 2004-06-23 MKNOD(2)