NetBSD/bin/cp/cp.c

592 lines
15 KiB
C

/*
* Copyright (c) 1988 The Regents of the University of California.
* All rights reserved.
*
* This code is derived from software contributed to Berkeley by
* David Hitz of Auspex Systems Inc.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by the University of
* California, Berkeley and its contributors.
* 4. Neither the name of the University nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#ifndef lint
char copyright[] =
"@(#) Copyright (c) 1988 The Regents of the University of California.\n\
All rights reserved.\n";
#endif /* not lint */
#ifndef lint
/*static char sccsid[] = "from: @(#)cp.c 5.26 (Berkeley) 10/27/91";*/
static char rcsid[] = "$Id: cp.c,v 1.8 1994/02/25 00:43:25 jtc Exp $";
#endif /* not lint */
/*
* cp copies source files to target files.
*
* The global PATH_T structures "to" and "from" always contain paths to the
* current source and target files, respectively. Since cp does not change
* directories, these paths can be either absolute or dot-relative.
*
* The basic algorithm is to initialize "to" and "from", and then call the
* recursive copy() function to do the actual work. If "from" is a file,
* copy copies the data. If "from" is a directory, copy creates the
* corresponding "to" directory, and calls itself recursively on all of
* the entries in the "from" directory.
*/
#include <sys/param.h>
#include <sys/stat.h>
#include <sys/mman.h>
#include <sys/time.h>
#include <dirent.h>
#include <fcntl.h>
#include <errno.h>
#include <unistd.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "extern.h"
static void copy __P((void));
static void copy_dir __P((void));
static void copy_fifo __P((struct stat *, int));
static void copy_file __P((struct stat *, int));
static void copy_link __P((int));
static void copy_special __P((struct stat *, int));
static void setfile __P((struct stat *, int));
static void usage __P((void));
PATH_T from = { from.p_path, "" };
PATH_T to = { to.p_path, "" };
uid_t myuid;
int exit_val, myumask;
int iflag, pflag, orflag, rflag;
int (*statfcn)();
char *progname;
main(argc, argv)
int argc;
char **argv;
{
extern int optind;
struct stat to_stat;
register int c, r;
int symfollow, lstat(), stat();
char *old_to, *p;
/*
* The utility cp(1) is used by mv(1) -- except for usage statements,
* print the "called as" program name.
*/
progname = (p = rindex(*argv,'/')) ? ++p : *argv;
symfollow = 0;
while ((c = getopt(argc, argv, "Rfhipr")) != EOF) {
switch ((char)c) {
case 'f':
iflag = 0;
break;
case 'h':
symfollow = 1;
break;
case 'i':
iflag = isatty(fileno(stdin));
break;
case 'p':
pflag = 1;
break;
case 'R':
rflag = 1;
break;
case 'r':
orflag = 1;
break;
case '?':
default:
usage();
break;
}
}
argc -= optind;
argv += optind;
if (argc < 2)
usage();
if (rflag && orflag) {
(void)fprintf(stderr,
"cp: the -R and -r options are mutually exclusive.\n");
exit(1);
}
myuid = getuid();
/* copy the umask for explicit mode setting */
myumask = umask(0);
(void)umask(myumask);
/* consume last argument first. */
if (!path_set(&to, argv[--argc]))
exit(1);
statfcn = symfollow || !rflag ? stat : lstat;
/*
* Cp has two distinct cases:
*
* % cp [-rip] source target
* % cp [-rip] source1 ... directory
*
* In both cases, source can be either a file or a directory.
*
* In (1), the target becomes a copy of the source. That is, if the
* source is a file, the target will be a file, and likewise for
* directories.
*
* In (2), the real target is not directory, but "directory/source".
*/
r = stat(to.p_path, &to_stat);
if (r == -1 && errno != ENOENT) {
err("%s: %s", to.p_path, strerror(errno));
exit(1);
}
if (r == -1 || !S_ISDIR(to_stat.st_mode)) {
/*
* Case (1). Target is not a directory.
*/
if (argc > 1) {
usage();
exit(1);
}
if (!path_set(&from, *argv))
exit(1);
copy();
}
else {
/*
* Case (2). Target is a directory.
*/
for (;; ++argv) {
if (!path_set(&from, *argv))
continue;
if (!(old_to =
path_append(&to, path_basename(&from), -1)))
continue;
copy();
if (!--argc)
break;
path_restore(&to, old_to);
}
}
exit(exit_val);
}
/* copy file or directory at "from" to "to". */
static void
copy()
{
struct stat from_stat, to_stat;
int dne, statval;
statval = statfcn(from.p_path, &from_stat);
if (statval == -1) {
err("%s: %s", from.p_path, strerror(errno));
return;
}
/* not an error, but need to remember it happened */
if (stat(to.p_path, &to_stat) == -1)
dne = 1;
else {
if (to_stat.st_dev == from_stat.st_dev &&
to_stat.st_ino == from_stat.st_ino) {
(void)fprintf(stderr,
"%s: %s and %s are identical (not copied).\n",
progname, to.p_path, from.p_path);
exit_val = 1;
return;
}
if (!S_ISDIR(from_stat.st_mode) && S_ISDIR(to_stat.st_mode)) {
(void)fprintf(stderr,
"%s: %s: cannot overwrite directory with non-directory.\n",
progname, to.p_path);
exit_val = 1;
return;
}
dne = 0;
}
switch(from_stat.st_mode & S_IFMT) {
case S_IFLNK:
copy_link(!dne);
return;
case S_IFDIR:
if (!rflag && !orflag) {
(void)fprintf(stderr,
"%s: %s is a directory (not copied).\n",
progname, from.p_path);
exit_val = 1;
return;
}
if (dne) {
/*
* If the directory doesn't exist, create the new
* one with the from file mode plus owner RWX bits,
* modified by the umask. Trade-off between being
* able to write the directory (if from directory is
* 555) and not causing a permissions race. If the
* umask blocks owner writes cp fails.
*/
if (mkdir(to.p_path, from_stat.st_mode|S_IRWXU) < 0) {
err("%s: %s", to.p_path, strerror(errno));
return;
}
}
else if (!S_ISDIR(to_stat.st_mode)) {
(void)fprintf(stderr, "%s: %s: not a directory.\n",
progname, to.p_path);
return;
}
copy_dir();
/*
* If not -p and directory didn't exist, set it to be the
* same as the from directory, umodified by the umask;
* arguably wrong, but it's been that way forever.
*/
if (pflag)
setfile(&from_stat, 0);
else if (dne)
(void)chmod(to.p_path, from_stat.st_mode);
return;
case S_IFCHR:
case S_IFBLK:
if (rflag) {
copy_special(&from_stat, !dne);
return;
}
break;
case S_IFIFO:
if (rflag) {
copy_fifo(&from_stat, !dne);
return;
}
break;
}
copy_file(&from_stat, dne);
}
static void
copy_file(fs, dne)
struct stat *fs;
int dne;
{
static char buf[MAXBSIZE];
register int from_fd, to_fd, rcount, wcount;
struct stat to_stat;
char *p;
if ((from_fd = open(from.p_path, O_RDONLY, 0)) == -1) {
err("%s: %s", from.p_path, strerror(errno));
return;
}
/*
* If the file exists and we're interactive, verify with the user.
* If the file DNE, set the mode to be the from file, minus setuid
* bits, modified by the umask; arguably wrong, but it makes copying
* executables work right and it's been that way forever. (The
* other choice is 666 or'ed with the execute bits on the from file
* modified by the umask.)
*/
if (!dne) {
if (iflag) {
int checkch, ch;
(void)fprintf(stderr, "overwrite %s? ", to.p_path);
checkch = ch = getchar();
while (ch != '\n' && ch != EOF)
ch = getchar();
if (checkch != 'y' && checkch != 'Y') {
(void)close(from_fd);
return;
}
}
to_fd = open(to.p_path, O_WRONLY|O_TRUNC, 0);
} else
to_fd = open(to.p_path, O_WRONLY|O_CREAT|O_TRUNC,
fs->st_mode & ~(S_ISUID|S_ISGID));
if (to_fd == -1) {
err("%s: %s", to.p_path, strerror(errno));
(void)close(from_fd);
return;
}
/*
* Mmap and write if less than 8M (the limit is so we don't totally
* trash memory on big files. This is really a minor hack, but it
* wins some CPU back.
*/
if (fs->st_size <= 8 * 1048576) {
if ((p = mmap(NULL, fs->st_size, PROT_READ,
MAP_FILE, from_fd, (off_t)0)) == (char *)-1)
err("%s: %s", from.p_path, strerror(errno));
if (write(to_fd, p, fs->st_size) != fs->st_size)
err("%s: %s", to.p_path, strerror(errno));
if (munmap(p, fs->st_size) < 0)
err("%s: %s", from.p_path, strerror(errno));
} else {
while ((rcount = read(from_fd, buf, MAXBSIZE)) > 0) {
wcount = write(to_fd, buf, rcount);
if (rcount != wcount || wcount == -1) {
err("%s: %s", to.p_path, strerror(errno));
break;
}
}
if (rcount < 0)
err("%s: %s", from.p_path, strerror(errno));
}
if (pflag)
setfile(fs, to_fd);
/*
* If the source was setuid or setgid, lose the bits unless the
* copy is owned by the same user and group.
*/
else if (fs->st_mode & (S_ISUID|S_ISGID) && fs->st_uid == myuid)
if (fstat(to_fd, &to_stat))
err("%s: %s", to.p_path, strerror(errno));
#define RETAINBITS (S_ISUID|S_ISGID|S_ISVTX|S_IRWXU|S_IRWXG|S_IRWXO)
else if (fs->st_gid == to_stat.st_gid && fchmod(to_fd,
fs->st_mode & RETAINBITS & ~myumask))
err("%s: %s", to.p_path, strerror(errno));
(void)close(from_fd);
if (close(to_fd))
err("%s: %s", to.p_path, strerror(errno));
}
static void
copy_dir()
{
struct stat from_stat;
struct dirent *dp, **dir_list;
register int dir_cnt, i;
char *old_from, *old_to;
dir_cnt = scandir(from.p_path, &dir_list, NULL, NULL);
if (dir_cnt == -1) {
(void)fprintf(stderr, "%s: can't read directory %s.\n",
progname, from.p_path);
exit_val = 1;
}
/*
* Instead of handling directory entries in the order they appear
* on disk, do non-directory files before directory files.
* There are two reasons to do directories last. The first is
* efficiency. Files tend to be in the same cylinder group as
* their parent, whereas directories tend not to be. Copying files
* all at once reduces seeking. Second, deeply nested tree's
* could use up all the file descriptors if we didn't close one
* directory before recursivly starting on the next.
*/
/* copy files */
for (i = 0; i < dir_cnt; ++i) {
dp = dir_list[i];
if (dp->d_namlen <= 2 && dp->d_name[0] == '.'
&& (dp->d_name[1] == NULL || dp->d_name[1] == '.'))
goto done;
if (!(old_from =
path_append(&from, dp->d_name, (int)dp->d_namlen)))
goto done;
if (statfcn(from.p_path, &from_stat) < 0) {
err("%s: %s", dp->d_name, strerror(errno));
path_restore(&from, old_from);
goto done;
}
if (S_ISDIR(from_stat.st_mode)) {
path_restore(&from, old_from);
continue;
}
if (old_to = path_append(&to, dp->d_name, (int)dp->d_namlen)) {
copy();
path_restore(&to, old_to);
}
path_restore(&from, old_from);
done: dir_list[i] = NULL;
free(dp);
}
/* copy directories */
for (i = 0; i < dir_cnt; ++i) {
dp = dir_list[i];
if (!dp)
continue;
if (!(old_from =
path_append(&from, dp->d_name, (int)dp->d_namlen))) {
free(dp);
continue;
}
if (!(old_to =
path_append(&to, dp->d_name, (int)dp->d_namlen))) {
free(dp);
path_restore(&from, old_from);
continue;
}
copy();
free(dp);
path_restore(&from, old_from);
path_restore(&to, old_to);
}
free(dir_list);
}
static void
copy_link(exists)
int exists;
{
int len;
char link[MAXPATHLEN];
if ((len = readlink(from.p_path, link, sizeof(link))) == -1) {
err("readlink: %s: %s", from.p_path, strerror(errno));
return;
}
link[len] = '\0';
if (exists && unlink(to.p_path)) {
err("unlink: %s: %s", to.p_path, strerror(errno));
return;
}
if (symlink(link, to.p_path)) {
err("symlink: %s: %s", link, strerror(errno));
return;
}
}
static void
copy_fifo(from_stat, exists)
struct stat *from_stat;
int exists;
{
if (exists && unlink(to.p_path)) {
err("unlink: %s: %s", to.p_path, strerror(errno));
return;
}
if (mkfifo(to.p_path, from_stat->st_mode)) {
err("mkfifo: %s: %s", to.p_path, strerror(errno));
return;
}
if (pflag)
setfile(from_stat, 0);
}
static void
copy_special(from_stat, exists)
struct stat *from_stat;
int exists;
{
if (exists && unlink(to.p_path)) {
err("unlink: %s: %s", to.p_path, strerror(errno));
return;
}
if (mknod(to.p_path, from_stat->st_mode, from_stat->st_rdev)) {
err("mknod: %s: %s", to.p_path, strerror(errno));
return;
}
if (pflag)
setfile(from_stat, 0);
}
static void
setfile(fs, fd)
register struct stat *fs;
int fd;
{
static struct timeval tv[2];
fs->st_mode &= S_ISUID|S_ISGID|S_IRWXU|S_IRWXG|S_IRWXO;
tv[0].tv_sec = fs->st_atime;
tv[1].tv_sec = fs->st_mtime;
if (utimes(to.p_path, tv))
err("utimes: %s: %s", to.p_path, strerror(errno));
/*
* Changing the ownership probably won't succeed, unless we're root
* or POSIX_CHOWN_RESTRICTED is not set. Set uid/gid before setting
* the mode; current BSD behavior is to remove all setuid bits on
* chown. If chown fails, lose setuid/setgid bits.
*/
if (fd ? fchown(fd, fs->st_uid, fs->st_gid) :
chown(to.p_path, fs->st_uid, fs->st_gid)) {
if (errno != EPERM)
err("chown: %s: %s", to.p_path, strerror(errno));
fs->st_mode &= ~(S_ISUID|S_ISGID);
}
if (fd ? fchmod(fd, fs->st_mode) : chmod(to.p_path, fs->st_mode))
err("chown: %s: %s", to.p_path, strerror(errno));
}
static void
usage()
{
(void)fprintf(stderr,
"usage: cp [-Rfhip] src target;\n cp [-Rfhip] src1 ... srcN directory\n");
exit(1);
}
#if __STDC__
#include <stdarg.h>
#else
#include <varargs.h>
#endif
void
#if __STDC__
err(const char *fmt, ...)
#else
err(fmt, va_alist)
char *fmt;
va_dcl
#endif
{
va_list ap;
#if __STDC__
va_start(ap, fmt);
#else
va_start(ap);
#endif
(void)fprintf(stderr, "%s: ", progname);
(void)vfprintf(stderr, fmt, ap);
va_end(ap);
(void)fprintf(stderr, "\n");
exit_val = 1;
}