9 /* $Id: compat_fts.c,v 1.8 2015/02/07 07:53:01 schwarze Exp $ */
10 /* $OpenBSD: fts.c,v 1.50 2015/01/16 16:48:51 deraadt Exp $ */
13 * Copyright (c) 1990, 1993, 1994
14 * The Regents of the University of California. All rights reserved.
16 * Redistribution and use in source and binary forms, with or without
17 * modification, are permitted provided that the following conditions
19 * 1. Redistributions of source code must retain the above copyright
20 * notice, this list of conditions and the following disclaimer.
21 * 2. Redistributions in binary form must reproduce the above copyright
22 * notice, this list of conditions and the following disclaimer in the
23 * documentation and/or other materials provided with the distribution.
24 * 3. Neither the name of the University nor the names of its contributors
25 * may be used to endorse or promote products derived from this software
26 * without specific prior written permission.
28 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
29 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
30 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
31 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
32 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
33 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
34 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
35 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
36 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
37 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
42 #include <sys/types.h>
51 #include "compat_fts.h"
53 #define MAXIMUM(a, b) (((a) > (b)) ? (a) : (b))
55 static FTSENT *fts_alloc(FTS *, const char *, size_t);
56 static FTSENT *fts_build(FTS *);
57 static void fts_lfree(FTSENT *);
58 static void fts_load(FTS *, FTSENT *);
59 static size_t fts_maxarglen(char * const *);
60 static void fts_padjust(FTS *, FTSENT *);
61 static int fts_palloc(FTS *, size_t);
62 static unsigned short fts_stat(FTS *, FTSENT *);
63 static int fts_safe_changedir(FTS *, FTSENT *, int, const char *);
65 #define ISDOT(a) (a[0] == '.' && (!a[1] || (a[1] == '.' && !a[2])))
73 #define CLR(opt) (sp->fts_options &= ~(opt))
74 #define ISSET(opt) (sp->fts_options & (opt))
75 #define SET(opt) (sp->fts_options |= (opt))
77 #define FCHDIR(sp, fd) (!ISSET(FTS_NOCHDIR) && fchdir(fd))
80 fts_open(char * const *argv, int options, void *dummy)
89 if (options & ~FTS_OPTIONMASK) {
94 /* Allocate/initialize the stream */
95 if ((sp = calloc(1, sizeof(FTS))) == NULL)
97 sp->fts_options = options;
100 * Start out with 1K of path space, and enough, in any case,
101 * to hold the user's paths.
103 if (fts_palloc(sp, MAXIMUM(fts_maxarglen(argv), PATH_MAX)))
106 /* Allocate/initialize root's parent. */
107 if ((parent = fts_alloc(sp, "", 0)) == NULL)
109 parent->fts_level = FTS_ROOTPARENTLEVEL;
111 /* Allocate/initialize root(s). */
112 for (root = NULL, nitems = 0; *argv; ++argv, ++nitems) {
113 /* Don't allow zero-length paths. */
114 if ((len = strlen(*argv)) == 0) {
119 if ((p = fts_alloc(sp, *argv, len)) == NULL)
121 p->fts_level = FTS_ROOTLEVEL;
122 p->fts_parent = parent;
123 p->fts_accpath = p->fts_name;
124 p->fts_info = fts_stat(sp, p);
126 /* Command-line "." and ".." are real directories. */
127 if (p->fts_info == FTS_DOT)
140 * Allocate a dummy pointer and make fts_read think that we've just
141 * finished the node before the root(s); set p->fts_info to FTS_INIT
142 * so that everything about the "current" node is ignored.
144 if ((sp->fts_cur = fts_alloc(sp, "", 0)) == NULL)
146 sp->fts_cur->fts_link = root;
147 sp->fts_cur->fts_info = FTS_INIT;
150 * If using chdir(2), grab a file descriptor pointing to dot to ensure
151 * that we can get back here; this could be avoided for some paths,
152 * but almost certainly not worth the effort. Slashes, symbolic links,
153 * and ".." are all fairly nasty problems. Note, if we can't get the
154 * descriptor we run anyway, just more slowly.
156 if (!ISSET(FTS_NOCHDIR) &&
157 (sp->fts_rfd = open(".", O_RDONLY | O_CLOEXEC)) < 0)
165 mem3: fts_lfree(root);
167 mem2: free(sp->fts_path);
173 fts_load(FTS *sp, FTSENT *p)
179 * Load the stream structure for the next traversal. Since we don't
180 * actually enter the directory until after the preorder visit, set
181 * the fts_accpath field specially so the chdir gets done to the right
182 * place and the user can access the first node. From fts_open it's
183 * known that the path will fit.
185 len = p->fts_pathlen = p->fts_namelen;
186 memmove(sp->fts_path, p->fts_name, len + 1);
187 if ((cp = strrchr(p->fts_name, '/')) && (cp != p->fts_name || cp[1])) {
189 memmove(p->fts_name, cp, len + 1);
190 p->fts_namelen = len;
192 p->fts_accpath = p->fts_path = sp->fts_path;
193 sp->fts_dev = p->fts_dev;
203 * This still works if we haven't read anything -- the dummy structure
204 * points to the root list, so we step through to the end of the root
205 * list which has a valid parent pointer.
208 for (p = sp->fts_cur; p->fts_level >= FTS_ROOTLEVEL;) {
210 p = p->fts_link ? p->fts_link : p->fts_parent;
216 /* Stash the original directory fd if needed. */
217 rfd = ISSET(FTS_NOCHDIR) ? -1 : sp->fts_rfd;
219 /* Free up child linked list, sort array, path buffer, stream ptr.*/
221 fts_lfree(sp->fts_child);
225 /* Return to original directory, checking for error. */
238 * Special case of "/" at the end of the path so that slashes aren't
239 * appended which would cause paths to be written as "....//foo".
242 (p->fts_path[p->fts_pathlen - 1] == '/' \
243 ? p->fts_pathlen - 1 : p->fts_pathlen)
252 /* If finished or unrecoverable error, return NULL. */
253 if (sp->fts_cur == NULL || ISSET(FTS_STOP))
256 /* Set current node pointer. */
259 /* Save and zero out user instructions. */
260 instr = p->fts_instr;
261 p->fts_instr = FTS_NOINSTR;
263 /* Directory in pre-order. */
264 if (p->fts_info == FTS_D) {
265 /* If skipped or crossed mount point, do post-order visit. */
266 if (instr == FTS_SKIP ||
267 (ISSET(FTS_XDEV) && p->fts_dev != sp->fts_dev)) {
269 fts_lfree(sp->fts_child);
270 sp->fts_child = NULL;
272 p->fts_info = FTS_DP;
277 * Cd to the subdirectory.
279 * If have already read and now fail to chdir, whack the list
280 * to make the names come out right, and set the parent errno
281 * so the application will eventually get an error condition.
282 * Set the FTS_DONTCHDIR flag so that when we logically change
283 * directories back to the parent we don't do a chdir.
285 * If haven't read do so. If the read fails, fts_build sets
286 * FTS_STOP or the fts_info field of the node.
289 if (fts_safe_changedir(sp, p, -1, p->fts_accpath)) {
290 p->fts_errno = errno;
291 p->fts_flags |= FTS_DONTCHDIR;
292 for (p = sp->fts_child; p; p = p->fts_link)
294 p->fts_parent->fts_accpath;
296 } else if ((sp->fts_child = fts_build(sp)) == NULL) {
302 sp->fts_child = NULL;
306 /* Move to the next node on this level. */
308 if ((p = p->fts_link)) {
312 * If reached the top, return to the original directory (or
313 * the root of the tree), and load the paths for the next root.
315 if (p->fts_level == FTS_ROOTLEVEL) {
316 if (FCHDIR(sp, sp->fts_rfd)) {
321 return (sp->fts_cur = p);
325 * User may have called fts_set on the node. If skipped,
326 * ignore. If followed, get a file descriptor so we can
327 * get back if necessary.
329 if (p->fts_instr == FTS_SKIP)
332 name: t = sp->fts_path + NAPPEND(p->fts_parent);
334 memmove(t, p->fts_name, p->fts_namelen + 1);
335 return (sp->fts_cur = p);
338 /* Move up to the parent node. */
342 if (p->fts_level == FTS_ROOTPARENTLEVEL) {
344 * Done; free everything up and set errno to 0 so the user
345 * can distinguish between error and EOF.
349 return (sp->fts_cur = NULL);
352 /* NUL terminate the pathname. */
353 sp->fts_path[p->fts_pathlen] = '\0';
356 * Return to the parent directory. If at a root node or came through
357 * a symlink, go back through the file descriptor. Otherwise, cd up
360 if (p->fts_level == FTS_ROOTLEVEL) {
361 if (FCHDIR(sp, sp->fts_rfd)) {
366 } else if (!(p->fts_flags & FTS_DONTCHDIR) &&
367 fts_safe_changedir(sp, p->fts_parent, -1, "..")) {
372 p->fts_info = p->fts_errno ? FTS_ERR : FTS_DP;
373 return (sp->fts_cur = p);
377 * Fts_set takes the stream as an argument although it's not used in this
378 * implementation; it would be necessary if anyone wanted to add global
379 * semantics to fts using fts_set. An error return is allowed for similar
384 fts_set(FTS *sp, FTSENT *p, int instr)
386 if (instr && instr != FTS_NOINSTR && instr != FTS_SKIP) {
390 p->fts_instr = instr;
395 * This is the tricky part -- do not casually change *anything* in here. The
396 * idea is to build the linked list of entries that are used by fts_children
397 * and fts_read. There are lots of special cases.
399 * The real slowdown in walking the tree is the stat calls. If FTS_NOSTAT is
400 * set and it's a physical walk (so that symbolic links can't be directories),
401 * we can do things quickly. First, if it's a 4.4BSD file system, the type
402 * of the file is in the directory entry. Otherwise, we assume that the number
403 * of subdirectories in a node is equal to the number of links to the parent.
404 * The former skips all stat calls. The latter skips stat calls in any leaf
405 * directories and for any files after the subdirectories in the directory have
406 * been found, cutting the stat calls by about 2/3.
416 size_t dlen, len, maxlen;
417 int nitems, cderrno, descend, level, doadjust;
421 /* Set current node pointer. */
425 * Open the directory for reading. If this fails, we're done.
426 * If being called from fts_read, set the fts_info field.
428 if ((dirp = opendir(cur->fts_accpath)) == NULL) {
429 cur->fts_info = FTS_DNR;
430 cur->fts_errno = errno;
435 * If we're going to need to stat anything or we want to descend
436 * and stay in the directory, chdir. If this fails we keep going,
437 * but set a flag so we don't chdir after the post-order visit.
438 * We won't be able to stat anything, but we can still return the
439 * names themselves. Note, that since fts_read won't be able to
440 * chdir into the directory, it will have to return different path
441 * names than before, i.e. "a/b" instead of "b". Since the node
442 * has already been visited in pre-order, have to wait until the
443 * post-order visit to return the error. There is a special case
444 * here, if there was nothing to stat then it's not an error to
445 * not be able to stat. This is all fairly nasty. If a program
446 * needed sorted entries or stat information, they had better be
447 * checking FTS_NS on the returned nodes.
450 if (fts_safe_changedir(sp, cur, dirfd(dirp), NULL)) {
451 cur->fts_errno = errno;
452 cur->fts_flags |= FTS_DONTCHDIR;
455 (void)closedir(dirp);
461 * Figure out the max file name length that can be stored in the
462 * current path -- the inner loop allocates more path as necessary.
463 * We really wouldn't have to do the maxlen calculations here, we
464 * could do them in fts_read before returning the path, but it's a
465 * lot easier here since the length is part of the dirent structure.
467 * If not changing directories set a pointer so that can just append
468 * each new name into the path.
471 if (ISSET(FTS_NOCHDIR)) {
472 cp = sp->fts_path + len;
476 maxlen = sp->fts_pathlen - len;
479 * fts_level is signed so we must prevent it from wrapping
480 * around to FTS_ROOTLEVEL and FTS_ROOTPARENTLEVEL.
482 level = cur->fts_level;
483 if (level < FTS_MAXLEVEL)
486 /* Read the directory, attaching each entry to the `link' pointer. */
488 for (head = tail = NULL, nitems = 0; dirp && (dp = readdir(dirp));) {
489 if (ISDOT(dp->d_name))
492 #if HAVE_DIRENT_NAMLEN
495 dlen = strlen(dp->d_name);
498 if (!(p = fts_alloc(sp, dp->d_name, dlen)))
500 if (dlen >= maxlen) { /* include space for NUL */
501 oldaddr = sp->fts_path;
502 if (fts_palloc(sp, dlen + len + 1)) {
504 * No more memory for path or structures. Save
505 * errno, free up the current structure and the
506 * structures already allocated.
508 mem1: saved_errno = errno;
512 (void)closedir(dirp);
513 cur->fts_info = FTS_ERR;
518 /* Did realloc() change the pointer? */
519 if (oldaddr != sp->fts_path) {
521 if (ISSET(FTS_NOCHDIR))
522 cp = sp->fts_path + len;
524 maxlen = sp->fts_pathlen - len;
527 p->fts_level = level;
528 p->fts_parent = sp->fts_cur;
529 p->fts_pathlen = len + dlen;
530 if (p->fts_pathlen < len) {
532 * If we wrap, free up the current structure and
533 * the structures already allocated, then error
534 * out with ENAMETOOLONG.
538 (void)closedir(dirp);
539 cur->fts_info = FTS_ERR;
541 errno = ENAMETOOLONG;
546 p->fts_info = FTS_NS;
547 p->fts_errno = cderrno;
548 p->fts_accpath = cur->fts_accpath;
550 /* Build a file name for fts_stat to stat. */
551 if (ISSET(FTS_NOCHDIR)) {
552 p->fts_accpath = p->fts_path;
553 memmove(cp, p->fts_name, p->fts_namelen + 1);
555 p->fts_accpath = p->fts_name;
557 p->fts_info = fts_stat(sp, p);
560 /* We walk in directory order so "ls -f" doesn't get upset. */
571 (void)closedir(dirp);
574 * If realloc() changed the address of the path, adjust the
575 * addresses for the rest of the tree and the dir list.
578 fts_padjust(sp, head);
581 * If not changing directories, reset the path back to original
584 if (ISSET(FTS_NOCHDIR)) {
585 if (len == sp->fts_pathlen || nitems == 0)
591 * If descended after called from fts_children or after called from
592 * fts_read and nothing found, get back. At the root level we use
593 * the saved fd; if one of fts_open()'s arguments is a relative path
594 * to an empty directory, we wind up here with no other way back. If
595 * can't get back, we're done.
597 if (descend && !nitems &&
598 (cur->fts_level == FTS_ROOTLEVEL ? FCHDIR(sp, sp->fts_rfd) :
599 fts_safe_changedir(sp, cur->fts_parent, -1, ".."))) {
600 cur->fts_info = FTS_ERR;
605 /* If didn't find anything, return NULL. */
607 cur->fts_info = FTS_DP;
613 static unsigned short
614 fts_stat(FTS *sp, FTSENT *p)
621 /* If user needs stat info, stat buffer already allocated. */
624 if (lstat(p->fts_accpath, sbp)) {
625 p->fts_errno = errno;
626 memset(sbp, 0, sizeof(struct stat));
630 if (S_ISDIR(sbp->st_mode)) {
632 * Set the device/inode. Used to find cycles and check for
633 * crossing mount points. Also remember the link count, used
634 * in fts_build to limit the number of stat calls. It is
635 * understood that these fields are only referenced if fts_info
638 dev = p->fts_dev = sbp->st_dev;
639 ino = p->fts_ino = sbp->st_ino;
640 p->fts_nlink = sbp->st_nlink;
642 if (ISDOT(p->fts_name))
646 * Cycle detection is done by brute force when the directory
647 * is first encountered. If the tree gets deep enough or the
648 * number of symbolic links to directories is high enough,
649 * something faster might be worthwhile.
651 for (t = p->fts_parent;
652 t->fts_level >= FTS_ROOTLEVEL; t = t->fts_parent)
653 if (ino == t->fts_ino && dev == t->fts_dev) {
659 if (S_ISLNK(sbp->st_mode))
661 if (S_ISREG(sbp->st_mode))
663 return (FTS_DEFAULT);
667 fts_alloc(FTS *sp, const char *name, size_t namelen)
672 len = sizeof(FTSENT) + namelen;
673 if ((p = calloc(1, len)) == NULL)
676 p->fts_path = sp->fts_path;
677 p->fts_namelen = namelen;
678 p->fts_instr = FTS_NOINSTR;
679 p->fts_statp = malloc(sizeof(struct stat));
680 if (p->fts_statp == NULL) {
684 memcpy(p->fts_name, name, namelen);
690 fts_lfree(FTSENT *head)
694 /* Free a linked list of structures. */
696 head = head->fts_link;
702 * Allow essentially unlimited paths; find, rm, ls should all work on any tree.
703 * Most systems will allow creation of paths much longer than PATH_MAX, even
704 * though the kernel won't resolve them. Add the size (not just what's needed)
705 * plus 256 bytes so don't realloc the path 2 bytes at a time.
708 fts_palloc(FTS *sp, size_t more)
713 * Check for possible wraparound.
716 if (sp->fts_pathlen + more < sp->fts_pathlen) {
720 errno = ENAMETOOLONG;
723 sp->fts_pathlen += more;
724 p = realloc(sp->fts_path, sp->fts_pathlen);
736 * When the path is realloc'd, have to fix all of the pointers in structures
740 fts_padjust(FTS *sp, FTSENT *head)
743 char *addr = sp->fts_path;
745 #define ADJUST(p) { \
746 if ((p)->fts_accpath != (p)->fts_name) { \
748 (char *)addr + ((p)->fts_accpath - (p)->fts_path); \
750 (p)->fts_path = addr; \
752 /* Adjust the current set of children. */
753 for (p = sp->fts_child; p; p = p->fts_link)
756 /* Adjust the rest of the tree, including the current level. */
757 for (p = head; p->fts_level >= FTS_ROOTLEVEL;) {
759 p = p->fts_link ? p->fts_link : p->fts_parent;
764 fts_maxarglen(char * const *argv)
768 for (max = 0; *argv; ++argv)
769 if ((len = strlen(*argv)) > max)
775 * Change to dir specified by fd or p->fts_accpath without getting
776 * tricked by someone changing the world out from underneath us.
777 * Assumes p->fts_dev and p->fts_ino are filled in.
780 fts_safe_changedir(FTS *sp, FTSENT *p, int fd, const char *path)
782 int ret, oerrno, newfd;
786 if (ISSET(FTS_NOCHDIR))
788 if (fd < 0 && (newfd = open(path, O_RDONLY|O_DIRECTORY|O_CLOEXEC)) < 0)
790 if (fstat(newfd, &sb)) {
794 if (p->fts_dev != sb.st_dev || p->fts_ino != sb.st_ino) {
795 errno = ENOENT; /* disinformation */