1 /* Copyright (C) 1991-2023 Free Software Foundation, Inc.
2 This file is part of the GNU C Library.
3
4 The GNU C Library is free software; you can redistribute it and/or
5 modify it under the terms of the GNU Lesser General Public
6 License as published by the Free Software Foundation; either
7 version 2.1 of the License, or (at your option) any later version.
8
9 The GNU C Library is distributed in the hope that it will be useful,
10 but WITHOUT ANY WARRANTY; without even the implied warranty of
11 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
12 Lesser General Public License for more details.
13
14 You should have received a copy of the GNU Lesser General Public
15 License along with the GNU C Library; if not, see
16 <https://www.gnu.org/licenses/>. */
17
18 #if !_LIBC
19 # include <config.h>
20 # include <unistd.h>
21 # include "pathmax.h"
22 #else
23 # define HAVE_OPENAT 1
24 # define D_INO_IN_DIRENT 1
25 # define HAVE_MSVC_INVALID_PARAMETER_HANDLER 0
26 # define HAVE_MINIMALLY_WORKING_GETCWD 0
27 #endif
28
29 #include <errno.h>
30 #include <sys/types.h>
31 #include <sys/stat.h>
32 #include <stdbool.h>
33 #include <stddef.h>
34
35 #include <fcntl.h> /* For AT_FDCWD on Solaris 9. */
36
37 /* If this host provides the openat function or if we're using the
38 gnulib replacement function with a native fdopendir, then enable
39 code below to make getcwd more efficient and robust. */
40 #if defined HAVE_OPENAT || (defined GNULIB_OPENAT && defined HAVE_FDOPENDIR)
41 # define HAVE_OPENAT_SUPPORT 1
42 #else
43 # define HAVE_OPENAT_SUPPORT 0
44 #endif
45
46 #ifndef __set_errno
47 # define __set_errno(val) (errno = (val))
48 #endif
49
50 #include <dirent.h>
51 #ifndef _D_EXACT_NAMLEN
52 # define _D_EXACT_NAMLEN(d) strlen ((d)->d_name)
53 #endif
54 #ifndef _D_ALLOC_NAMLEN
55 # define _D_ALLOC_NAMLEN(d) (_D_EXACT_NAMLEN (d) + 1)
56 #endif
57
58 #include <unistd.h>
59 #include <stdlib.h>
60 #include <string.h>
61
62 #if _LIBC
63 # ifndef mempcpy
64 # define mempcpy __mempcpy
65 # endif
66 #endif
67
68 #ifndef MAX
69 # define MAX(a, b) ((a) < (b) ? (b) : (a))
70 #endif
71 #ifndef MIN
72 # define MIN(a, b) ((a) < (b) ? (a) : (b))
73 #endif
74
75 /* In this file, PATH_MAX only serves as a threshold for choosing among two
76 algorithms. */
77 #ifndef PATH_MAX
78 # define PATH_MAX 8192
79 #endif
80
81 #if D_INO_IN_DIRENT
82 # define MATCHING_INO(dp, ino) ((dp)->d_ino == (ino))
83 #else
84 # define MATCHING_INO(dp, ino) true
85 #endif
86
87 #if HAVE_MSVC_INVALID_PARAMETER_HANDLER
88 # include "msvc-inval.h"
89 #endif
90
91 #if !_LIBC
92 # define __close_nocancel_nostatus close
93 # define __getcwd_generic rpl_getcwd
94 # define stat64 stat
95 # define __fstat64 fstat
96 # define __fstatat64 fstatat
97 # define __lstat64 lstat
98 # define __closedir closedir
99 # define __opendir opendir
100 # define __readdir64 readdir
101 # define __fdopendir fdopendir
102 # define __openat openat
103 # define __rewinddir rewinddir
104 # define __openat64 openat
105 # define dirent64 dirent
106 #else
107 # include <not-cancel.h>
108 #endif
109
110 /* The results of opendir() in this file are not used with dirfd and fchdir,
111 and we do not leak fds to any single-threaded code that could use stdio,
112 therefore save some unnecessary recursion in fchdir.c.
113 FIXME - if the kernel ever adds support for multi-thread safety for
114 avoiding standard fds, then we should use opendir_safer and
115 openat_safer. */
116 #ifdef GNULIB_defined_opendir
117 # undef opendir
118 #endif
119 #ifdef GNULIB_defined_closedir
120 # undef closedir
121 #endif
122
123 #if defined _WIN32 && !defined __CYGWIN__
124 # if HAVE_MSVC_INVALID_PARAMETER_HANDLER
125 static char *
126 getcwd_nothrow (char *buf, size_t size)
127 {
128 char *result;
129
130 TRY_MSVC_INVAL
131 {
132 result = _getcwd (buf, size);
133 }
134 CATCH_MSVC_INVAL
135 {
136 result = NULL;
137 errno = ERANGE;
138 }
139 DONE_MSVC_INVAL;
140
141 return result;
142 }
143 # else
144 # define getcwd_nothrow _getcwd
145 # endif
146 # define getcwd_system getcwd_nothrow
147 #else
148 # define getcwd_system getcwd
149 #endif
150
151 /* Get the name of the current working directory, and put it in SIZE
152 bytes of BUF. Returns NULL with errno set if the directory couldn't be
153 determined or SIZE was too small. If successful, returns BUF. In GNU,
154 if BUF is NULL, an array is allocated with 'malloc'; the array is SIZE
155 bytes long, unless SIZE == 0, in which case it is as big as necessary. */
156
157 GETCWD_RETURN_TYPE
158 __getcwd_generic (char *buf, size_t size)
159 {
160 /* Lengths of big file name components and entire file names, and a
161 deep level of file name nesting. These numbers are not upper
162 bounds; they are merely large values suitable for initial
163 allocations, designed to be large enough for most real-world
164 uses. */
165 enum
166 {
167 BIG_FILE_NAME_COMPONENT_LENGTH = 255,
168 BIG_FILE_NAME_LENGTH = MIN (4095, PATH_MAX - 1),
169 DEEP_NESTING = 100
170 };
171
172 #if HAVE_OPENAT_SUPPORT
173 int fd = AT_FDCWD;
174 bool fd_needs_closing = false;
175 #else
176 char dots[DEEP_NESTING * sizeof ".." + BIG_FILE_NAME_COMPONENT_LENGTH + 1];
177 char *dotlist = dots;
178 size_t dotsize = sizeof dots;
179 size_t dotlen = 0;
180 #endif
181 DIR *dirstream = NULL;
182 dev_t rootdev, thisdev;
183 ino_t rootino, thisino;
184 char *dir;
185 register char *dirp;
186 struct __stat64_t64 st;
187 size_t allocated = size;
188 size_t used;
189
190 /* A size of 1 byte is never useful. */
191 if (allocated == 1)
192 {
193 __set_errno (ERANGE);
194 return NULL;
195 }
196
197 #if HAVE_MINIMALLY_WORKING_GETCWD
198 /* If AT_FDCWD is not defined, the algorithm below is O(N**2) and
199 this is much slower than the system getcwd (at least on
200 GNU/Linux). So trust the system getcwd's results unless they
201 look suspicious.
202
203 Use the system getcwd even if we have openat support, since the
204 system getcwd works even when a parent is unreadable, while the
205 openat-based approach does not.
206
207 But on AIX 5.1..7.1, the system getcwd is not even minimally
208 working: If the current directory name is slightly longer than
209 PATH_MAX, it omits the first directory component and returns
210 this wrong result with errno = 0. */
211
212 # undef getcwd
213 dir = getcwd_system (buf, size);
214 if (dir || (size && errno == ERANGE))
215 return dir;
216
217 /* Solaris getcwd (NULL, 0) fails with errno == EINVAL, but it has
218 internal magic that lets it work even if an ancestor directory is
219 inaccessible, which is better in many cases. So in this case try
220 again with a buffer that's almost always big enough. */
221 if (errno == EINVAL && buf == NULL && size == 0)
222 {
223 char big_buffer[BIG_FILE_NAME_LENGTH + 1];
224 dir = getcwd_system (big_buffer, sizeof big_buffer);
225 if (dir)
226 return strdup (dir);
227 }
228
229 # if HAVE_PARTLY_WORKING_GETCWD
230 /* The system getcwd works, except it sometimes fails when it
231 shouldn't, setting errno to ERANGE, ENAMETOOLONG, or ENOENT. */
232 if (errno != ERANGE && errno != ENAMETOOLONG && errno != ENOENT)
233 return NULL;
234 # endif
235 #endif
236 if (size == 0)
237 {
238 if (buf != NULL)
239 {
240 __set_errno (EINVAL);
241 return NULL;
242 }
243
244 allocated = BIG_FILE_NAME_LENGTH + 1;
245 }
246
247 if (buf == NULL)
248 {
249 dir = malloc (allocated);
250 if (dir == NULL)
251 return NULL;
252 }
253 else
254 dir = buf;
255
256 dirp = dir + allocated;
257 *--dirp = '\0';
258
259 if (__lstat64_time64 (".", &st) < 0)
260 goto lose;
261 thisdev = st.st_dev;
262 thisino = st.st_ino;
263
264 if (__lstat64_time64 ("/", &st) < 0)
265 goto lose;
266 rootdev = st.st_dev;
267 rootino = st.st_ino;
268
269 while (!(thisdev == rootdev && thisino == rootino))
270 {
271 struct dirent64 *d;
272 dev_t dotdev;
273 ino_t dotino;
274 bool mount_point;
275 int parent_status;
276 size_t dirroom;
277 size_t namlen;
278 bool use_d_ino = true;
279
280 /* Look at the parent directory. */
281 #if HAVE_OPENAT_SUPPORT
282 fd = __openat64 (fd, "..", O_RDONLY);
283 if (fd < 0)
284 goto lose;
285 fd_needs_closing = true;
286 parent_status = __fstat64_time64 (fd, &st);
287 #else
288 dotlist[dotlen++] = '.';
289 dotlist[dotlen++] = '.';
290 dotlist[dotlen] = '\0';
291 parent_status = __lstat64_time64 (dotlist, &st);
292 #endif
293 if (parent_status != 0)
294 goto lose;
295
296 if (dirstream && __closedir (dirstream) != 0)
297 {
298 dirstream = NULL;
299 goto lose;
300 }
301
302 /* Figure out if this directory is a mount point. */
303 dotdev = st.st_dev;
304 dotino = st.st_ino;
305 mount_point = dotdev != thisdev;
306
307 /* Search for the last directory. */
308 #if HAVE_OPENAT_SUPPORT
309 dirstream = __fdopendir (fd);
310 if (dirstream == NULL)
311 goto lose;
312 fd_needs_closing = false;
313 #else
314 dirstream = __opendir (dotlist);
315 if (dirstream == NULL)
316 goto lose;
317 dotlist[dotlen++] = '/';
318 #endif
319 for (;;)
320 {
321 /* Clear errno to distinguish EOF from error if readdir returns
322 NULL. */
323 __set_errno (0);
324 d = __readdir64 (dirstream);
325
326 /* When we've iterated through all directory entries without finding
327 one with a matching d_ino, rewind the stream and consider each
328 name again, but this time, using lstat. This is necessary in a
329 chroot on at least one system (glibc-2.3.6 + linux 2.6.12), where
330 .., ../.., ../../.., etc. all had the same device number, yet the
331 d_ino values for entries in / did not match those obtained
332 via lstat. */
333 if (d == NULL && errno == 0 && use_d_ino)
334 {
335 use_d_ino = false;
336 __rewinddir (dirstream);
337 d = __readdir64 (dirstream);
338 }
339
340 if (d == NULL)
341 {
342 if (errno == 0)
343 /* EOF on dirstream, which can mean e.g., that the current
344 directory has been removed. */
345 __set_errno (ENOENT);
346 goto lose;
347 }
348 if (d->d_name[0] == '.' &&
349 (d->d_name[1] == '\0' ||
350 (d->d_name[1] == '.' && d->d_name[2] == '\0')))
351 continue;
352
353 if (use_d_ino)
354 {
355 bool match = (MATCHING_INO (d, thisino) || mount_point);
356 if (! match)
357 continue;
358 }
359
360 {
361 int entry_status;
362 #if HAVE_OPENAT_SUPPORT
363 entry_status = __fstatat64_time64 (fd, d->d_name, &st,
364 AT_SYMLINK_NOFOLLOW);
365 #else
366 /* Compute size needed for this file name, or for the file
367 name ".." in the same directory, whichever is larger.
368 Room for ".." might be needed the next time through
369 the outer loop. */
370 size_t name_alloc = _D_ALLOC_NAMLEN (d);
371 size_t filesize = dotlen + MAX (sizeof "..", name_alloc);
372
373 if (filesize < dotlen)
374 goto memory_exhausted;
375
376 if (dotsize < filesize)
377 {
378 /* My, what a deep directory tree you have, Grandma. */
379 size_t newsize = MAX (filesize, dotsize * 2);
380 size_t i;
381 if (newsize < dotsize)
382 goto memory_exhausted;
383 if (dotlist != dots)
384 free (dotlist);
385 dotlist = malloc (newsize);
386 if (dotlist == NULL)
387 goto lose;
388 dotsize = newsize;
389
390 i = 0;
391 do
392 {
393 dotlist[i++] = '.';
394 dotlist[i++] = '.';
395 dotlist[i++] = '/';
396 }
397 while (i < dotlen);
398 }
399
400 memcpy (dotlist + dotlen, d->d_name, _D_ALLOC_NAMLEN (d));
401 entry_status = __lstat64_time64 (dotlist, &st);
402 #endif
403 /* We don't fail here if we cannot stat() a directory entry.
404 This can happen when (network) file systems fail. If this
405 entry is in fact the one we are looking for we will find
406 out soon as we reach the end of the directory without
407 having found anything. */
408 if (entry_status == 0 && S_ISDIR (st.st_mode)
409 && st.st_dev == thisdev && st.st_ino == thisino)
410 break;
411 }
412 }
413
414 dirroom = dirp - dir;
415 namlen = _D_EXACT_NAMLEN (d);
416
417 if (dirroom <= namlen)
418 {
419 if (size != 0)
420 {
421 __set_errno (ERANGE);
422 goto lose;
423 }
424 else
425 {
426 char *tmp;
427 size_t oldsize = allocated;
428
429 allocated += MAX (allocated, namlen);
430 if (allocated < oldsize
431 || ! (tmp = realloc (dir, allocated)))
432 goto memory_exhausted;
433
434 /* Move current contents up to the end of the buffer.
435 This is guaranteed to be non-overlapping. */
436 dirp = memcpy (tmp + allocated - (oldsize - dirroom),
437 tmp + dirroom,
438 oldsize - dirroom);
439 dir = tmp;
440 }
441 }
442 dirp -= namlen;
443 memcpy (dirp, d->d_name, namlen);
444 *--dirp = '/';
445
446 thisdev = dotdev;
447 thisino = dotino;
448 }
449
450 if (dirstream && __closedir (dirstream) != 0)
451 {
452 dirstream = NULL;
453 goto lose;
454 }
455
456 if (dirp == &dir[allocated - 1])
457 *--dirp = '/';
458
459 #if ! HAVE_OPENAT_SUPPORT
460 if (dotlist != dots)
461 free (dotlist);
462 #endif
463
464 used = dir + allocated - dirp;
465 memmove (dir, dirp, used);
466
467 if (size == 0)
468 /* Ensure that the buffer is only as large as necessary. */
469 buf = (used < allocated ? realloc (dir, used) : dir);
470
471 if (buf == NULL)
472 /* Either buf was NULL all along, or 'realloc' failed but
473 we still have the original string. */
474 buf = dir;
475
476 return buf;
477
478 memory_exhausted:
479 __set_errno (ENOMEM);
480 lose:
481 {
482 int save = errno;
483 if (dirstream)
484 __closedir (dirstream);
485 #if HAVE_OPENAT_SUPPORT
486 if (fd_needs_closing)
487 __close_nocancel_nostatus (fd);
488 #else
489 if (dotlist != dots)
490 free (dotlist);
491 #endif
492 if (buf == NULL)
493 free (dir);
494 __set_errno (save);
495 }
496 return NULL;
497 }
498
499 #if defined _LIBC && !defined GETCWD_RETURN_TYPE
500 libc_hidden_def (__getcwd)
501 weak_alias (__getcwd, getcwd)
502 #endif