du.c revision 132201 
1/*
2 * Copyright (c) 1989, 1993, 1994
3 *	The Regents of the University of California.  All rights reserved.
4 *
5 * This code is derived from software contributed to Berkeley by
6 * Chris Newcomb.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 * 1. Redistributions of source code must retain the above copyright
12 *    notice, this list of conditions and the following disclaimer.
13 * 2. Redistributions in binary form must reproduce the above copyright
14 *    notice, this list of conditions and the following disclaimer in the
15 *    documentation and/or other materials provided with the distribution.
16 * 3. All advertising materials mentioning features or use of this software
17 *    must display the following acknowledgement:
18 *	This product includes software developed by the University of
19 *	California, Berkeley and its contributors.
20 * 4. Neither the name of the University nor the names of its contributors
21 *    may be used to endorse or promote products derived from this software
22 *    without specific prior written permission.
23 *
24 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
25 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
26 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
27 * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
28 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
29 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
30 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
31 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
33 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
34 * SUCH DAMAGE.
35 */
36
37#ifndef lint
38static const char copyright[] =
39"@(#) Copyright (c) 1989, 1993, 1994\n\
40	The Regents of the University of California.  All rights reserved.\n";
41#endif /* not lint */
42
43#ifndef lint
44#if 0
45static const char sccsid[] = "@(#)du.c	8.5 (Berkeley) 5/4/95";
46#endif
47#endif /* not lint */
48#include <sys/cdefs.h>
49__FBSDID("$FreeBSD: head/usr.bin/du/du.c 132201 2004-07-15 08:54:41Z tjr $");
50
51#include <sys/param.h>
52#include <sys/queue.h>
53#include <sys/stat.h>
54
55#include <err.h>
56#include <errno.h>
57#include <fnmatch.h>
58#include <fts.h>
59#include <libutil.h>
60#include <locale.h>
61#include <stdio.h>
62#include <stdlib.h>
63#include <string.h>
64#include <sysexits.h>
65#include <unistd.h>
66
67SLIST_HEAD(ignhead, ignentry) ignores;
68struct ignentry {
69	char			*mask;
70	SLIST_ENTRY(ignentry)	next;
71};
72
73static int	linkchk(FTSENT *);
74static void	usage(void);
75void		prthumanval(int64_t);
76void		ignoreadd(const char *);
77void		ignoreclean(void);
78int		ignorep(FTSENT *);
79
80int
81main(int argc, char *argv[])
82{
83	FTS		*fts;
84	FTSENT		*p;
85	long		blocksize, savednumber = 0;
86	int		ftsoptions;
87	int		listall;
88	int		depth;
89	int		Hflag, Lflag, Pflag, aflag, sflag, dflag, cflag, hflag, ch, notused, rval;
90	char 		**save;
91	static char	dot[] = ".";
92
93	setlocale(LC_ALL, "");
94
95	Hflag = Lflag = Pflag = aflag = sflag = dflag = cflag = hflag = 0;
96
97	save = argv;
98	ftsoptions = 0;
99	depth = INT_MAX;
100	SLIST_INIT(&ignores);
101
102	while ((ch = getopt(argc, argv, "HI:LPasd:chkmrx")) != -1)
103		switch (ch) {
104			case 'H':
105				Hflag = 1;
106				break;
107			case 'I':
108				ignoreadd(optarg);
109				break;
110			case 'L':
111				if (Pflag)
112					usage();
113				Lflag = 1;
114				break;
115			case 'P':
116				if (Lflag)
117					usage();
118				Pflag = 1;
119				break;
120			case 'a':
121				aflag = 1;
122				break;
123			case 's':
124				sflag = 1;
125				break;
126			case 'd':
127				dflag = 1;
128				errno = 0;
129				depth = atoi(optarg);
130				if (errno == ERANGE || depth < 0) {
131					warnx("invalid argument to option d: %s", optarg);
132					usage();
133				}
134				break;
135			case 'c':
136				cflag = 1;
137				break;
138			case 'h':
139				putenv("BLOCKSIZE=512");
140				hflag = 1;
141				break;
142			case 'k':
143				hflag = 0;
144				putenv("BLOCKSIZE=1024");
145				break;
146			case 'm':
147				hflag = 0;
148				putenv("BLOCKSIZE=1048576");
149				break;
150			case 'r':		 /* Compatibility. */
151				break;
152			case 'x':
153				ftsoptions |= FTS_XDEV;
154				break;
155			case '?':
156			default:
157				usage();
158		}
159
160	argc -= optind;
161	argv += optind;
162
163	/*
164	 * XXX
165	 * Because of the way that fts(3) works, logical walks will not count
166	 * the blocks actually used by symbolic links.  We rationalize this by
167	 * noting that users computing logical sizes are likely to do logical
168	 * copies, so not counting the links is correct.  The real reason is
169	 * that we'd have to re-implement the kernel's symbolic link traversing
170	 * algorithm to get this right.  If, for example, you have relative
171	 * symbolic links referencing other relative symbolic links, it gets
172	 * very nasty, very fast.  The bottom line is that it's documented in
173	 * the man page, so it's a feature.
174	 */
175
176	if (Hflag + Lflag + Pflag > 1)
177		usage();
178
179	if (Hflag + Lflag + Pflag == 0)
180		Pflag = 1;			/* -P (physical) is default */
181
182	if (Hflag)
183		ftsoptions |= FTS_COMFOLLOW;
184
185	if (Lflag)
186		ftsoptions |= FTS_LOGICAL;
187
188	if (Pflag)
189		ftsoptions |= FTS_PHYSICAL;
190
191	listall = 0;
192
193	if (aflag) {
194		if (sflag || dflag)
195			usage();
196		listall = 1;
197	} else if (sflag) {
198		if (dflag)
199			usage();
200		depth = 0;
201	}
202
203	if (!*argv) {
204		argv = save;
205		argv[0] = dot;
206		argv[1] = NULL;
207	}
208
209	(void) getbsize(&notused, &blocksize);
210	blocksize /= 512;
211
212	rval = 0;
213
214	if ((fts = fts_open(argv, ftsoptions, NULL)) == NULL)
215		err(1, "fts_open");
216
217	while ((p = fts_read(fts)) != NULL) {
218		switch (p->fts_info) {
219			case FTS_D:			/* Ignore. */
220				if (ignorep(p))
221					fts_set(fts, p, FTS_SKIP);
222				break;
223			case FTS_DP:
224				if (ignorep(p))
225					break;
226
227				p->fts_parent->fts_number +=
228				    p->fts_number += p->fts_statp->st_blocks;
229
230				if (p->fts_level <= depth) {
231					if (hflag) {
232						(void) prthumanval(howmany(p->fts_number, blocksize));
233						(void) printf("\t%s\n", p->fts_path);
234					} else {
235					(void) printf("%ld\t%s\n",
236					    howmany(p->fts_number, blocksize),
237					    p->fts_path);
238					}
239				}
240				break;
241			case FTS_DC:			/* Ignore. */
242				break;
243			case FTS_DNR:			/* Warn, continue. */
244			case FTS_ERR:
245			case FTS_NS:
246				warnx("%s: %s", p->fts_path, strerror(p->fts_errno));
247				rval = 1;
248				break;
249			default:
250				if (ignorep(p))
251					break;
252
253				if (p->fts_statp->st_nlink > 1 && linkchk(p))
254					break;
255
256				if (listall || p->fts_level == 0) {
257					if (hflag) {
258						(void) prthumanval(howmany(p->fts_statp->st_blocks,
259							blocksize));
260						(void) printf("\t%s\n", p->fts_path);
261					} else {
262						(void) printf("%qd\t%s\n",
263							(long long)howmany(p->fts_statp->st_blocks, blocksize),
264							p->fts_path);
265					}
266				}
267
268				p->fts_parent->fts_number += p->fts_statp->st_blocks;
269		}
270		savednumber = p->fts_parent->fts_number;
271	}
272
273	if (errno)
274		err(1, "fts_read");
275
276	if (cflag) {
277		if (hflag) {
278			(void) prthumanval(howmany(savednumber, blocksize));
279			(void) printf("\ttotal\n");
280		} else {
281			(void) printf("%ld\ttotal\n", howmany(savednumber, blocksize));
282		}
283	}
284
285	ignoreclean();
286	exit(rval);
287}
288
289static int
290linkchk(FTSENT *p)
291{
292	struct links_entry {
293		struct links_entry *next;
294		struct links_entry *previous;
295		int	 links;
296		dev_t	 dev;
297		ino_t	 ino;
298	};
299	static const size_t links_hash_initial_size = 8192;
300	static struct links_entry **buckets;
301	static struct links_entry *free_list;
302	static size_t number_buckets;
303	static unsigned long number_entries;
304	static char stop_allocating;
305	struct links_entry *le, **new_buckets;
306	struct stat *st;
307	size_t i, new_size;
308	int count, hash;
309
310	st = p->fts_statp;
311
312	/* If necessary, initialize the hash table. */
313	if (buckets == NULL) {
314		number_buckets = links_hash_initial_size;
315		buckets = malloc(number_buckets * sizeof(buckets[0]));
316		if (buckets == NULL)
317			errx(1, "No memory for hardlink detection");
318		for (i = 0; i < number_buckets; i++)
319			buckets[i] = NULL;
320	}
321
322	/* If the hash table is getting too full, enlarge it. */
323	if (number_entries > number_buckets * 10 && !stop_allocating) {
324		new_size = number_buckets * 2;
325		new_buckets = malloc(new_size * sizeof(struct links_entry *));
326		count = 0;
327
328		/* Try releasing the free list to see if that helps. */
329		if (new_buckets == NULL && free_list != NULL) {
330			while (free_list != NULL) {
331				le = free_list;
332				free_list = le->next;
333				free(le);
334			}
335			new_buckets = malloc(new_size * sizeof(new_buckets[0]));
336		}
337
338		if (new_buckets == NULL) {
339			stop_allocating = 1;
340			warnx("No more memory for tracking hard links");
341		} else {
342			memset(new_buckets, 0,
343			    new_size * sizeof(struct links_entry *));
344			for (i = 0; i < number_buckets; i++) {
345				while (buckets[i] != NULL) {
346					/* Remove entry from old bucket. */
347					le = buckets[i];
348					buckets[i] = le->next;
349
350					/* Add entry to new bucket. */
351					hash = (le->dev ^ le->ino) % new_size;
352
353					if (new_buckets[hash] != NULL)
354						new_buckets[hash]->previous =
355						    le;
356					le->next = new_buckets[hash];
357					le->previous = NULL;
358					new_buckets[hash] = le;
359				}
360			}
361			free(buckets);
362			buckets = new_buckets;
363			number_buckets = new_size;
364		}
365	}
366
367	/* Try to locate this entry in the hash table. */
368	hash = ( st->st_dev ^ st->st_ino ) % number_buckets;
369	for (le = buckets[hash]; le != NULL; le = le->next) {
370		if (le->dev == st->st_dev && le->ino == st->st_ino) {
371			/*
372			 * Save memory by releasing an entry when we've seen
373			 * all of it's links.
374			 */
375			if (--le->links <= 0) {
376				if (le->previous != NULL)
377					le->previous->next = le->next;
378				if (le->next != NULL)
379					le->next->previous = le->previous;
380				if (buckets[hash] == le)
381					buckets[hash] = le->next;
382				number_entries--;
383				/* Recycle this node through the free list */
384				if (stop_allocating) {
385					free(le);
386				} else {
387					le->next = free_list;
388					free_list = le;
389				}
390			}
391			return (1);
392		}
393	}
394
395	if (stop_allocating)
396		return (0);
397
398	/* Add this entry to the links cache. */
399	if (free_list != NULL) {
400		/* Pull a node from the free list if we can. */
401		le = free_list;
402		free_list = le->next;
403	} else
404		/* Malloc one if we have to. */
405		le = malloc(sizeof(struct links_entry));
406	if (le == NULL) {
407		stop_allocating = 1;
408		warnx("No more memory for tracking hard links");
409		return (0);
410	}
411	le->dev = st->st_dev;
412	le->ino = st->st_ino;
413	le->links = st->st_nlink - 1;
414	number_entries++;
415	le->next = buckets[hash];
416	le->previous = NULL;
417	if (buckets[hash] != NULL)
418		buckets[hash]->previous = le;
419	buckets[hash] = le;
420	return (0);
421}
422
423void
424prthumanval(int64_t bytes)
425{
426	char buf[5];
427
428	bytes *= DEV_BSIZE;
429
430	humanize_number(buf, sizeof(buf), bytes, "", HN_AUTOSCALE,
431	    HN_B | HN_NOSPACE | HN_DECIMAL);
432
433	(void)printf("%4s", buf);
434}
435
436static void
437usage(void)
438{
439	(void)fprintf(stderr,
440		"usage: du [-H | -L | -P] [-a | -s | -d depth] [-c] [-h | -k | -m] [-x] [-I mask] [file ...]\n");
441	exit(EX_USAGE);
442}
443
444void
445ignoreadd(const char *mask)
446{
447	struct ignentry *ign;
448
449	ign = calloc(1, sizeof(*ign));
450	if (ign == NULL)
451		errx(1, "cannot allocate memory");
452	ign->mask = strdup(mask);
453	if (ign->mask == NULL)
454		errx(1, "cannot allocate memory");
455	SLIST_INSERT_HEAD(&ignores, ign, next);
456}
457
458void
459ignoreclean(void)
460{
461	struct ignentry *ign;
462
463	while (!SLIST_EMPTY(&ignores)) {
464		ign = SLIST_FIRST(&ignores);
465		SLIST_REMOVE_HEAD(&ignores, next);
466		free(ign->mask);
467		free(ign);
468	}
469}
470
471int
472ignorep(FTSENT *ent)
473{
474	struct ignentry *ign;
475
476	SLIST_FOREACH(ign, &ignores, next)
477		if (fnmatch(ign->mask, ent->fts_name, 0) != FNM_NOMATCH)
478			return 1;
479	return 0;
480}
481