du.c revision 1.27 1 /* $NetBSD: du.c,v 1.27 2006/04/30 19:24:50 liamjfoy Exp $ */
2
3 /*
4 * Copyright (c) 1989, 1993, 1994
5 * The Regents of the University of California. All rights reserved.
6 *
7 * This code is derived from software contributed to Berkeley by
8 * Chris Newcomb.
9 *
10 * Redistribution and use in source and binary forms, with or without
11 * modification, are permitted provided that the following conditions
12 * are met:
13 * 1. Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * 2. Redistributions in binary form must reproduce the above copyright
16 * notice, this list of conditions and the following disclaimer in the
17 * documentation and/or other materials provided with the distribution.
18 * 3. Neither the name of the University nor the names of its contributors
19 * may be used to endorse or promote products derived from this software
20 * without specific prior written permission.
21 *
22 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
23 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
24 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
25 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
26 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
27 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
28 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
29 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
30 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
31 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
32 * SUCH DAMAGE.
33 */
34
35 #include <sys/cdefs.h>
36 #ifndef lint
37 __COPYRIGHT("@(#) Copyright (c) 1989, 1993, 1994\n\
38 The Regents of the University of California. All rights reserved.\n");
39 #endif /* not lint */
40
41 #ifndef lint
42 #if 0
43 static char sccsid[] = "@(#)du.c 8.5 (Berkeley) 5/4/95";
44 #else
45 __RCSID("$NetBSD: du.c,v 1.27 2006/04/30 19:24:50 liamjfoy Exp $");
46 #endif
47 #endif /* not lint */
48
49 #include <sys/types.h>
50 #include <sys/stat.h>
51
52 #include <dirent.h>
53 #include <err.h>
54 #include <errno.h>
55 #include <fts.h>
56 #include <util.h>
57 #include <stdio.h>
58 #include <stdlib.h>
59 #include <string.h>
60 #include <unistd.h>
61 #include <limits.h>
62
63 int linkchk(dev_t, ino_t);
64 void prstat(const char *, int64_t);
65 void usage(void);
66
67 int hflag;
68 long blocksize;
69
70 int
71 main(int argc, char *argv[])
72 {
73 FTS *fts;
74 FTSENT *p;
75 int64_t totalblocks;
76 int ftsoptions, listdirs, listfiles;
77 int Hflag, Lflag, aflag, ch, cflag, gkmflag, nflag, rval, sflag;
78 const char *noargv[2];
79
80 Hflag = Lflag = aflag = cflag = gkmflag = nflag = sflag = 0;
81 totalblocks = 0;
82 ftsoptions = FTS_PHYSICAL;
83 while ((ch = getopt(argc, argv, "HLPacghkmnrsx")) != -1)
84 switch (ch) {
85 case 'H':
86 Hflag = 1;
87 Lflag = 0;
88 break;
89 case 'L':
90 Lflag = 1;
91 Hflag = 0;
92 break;
93 case 'P':
94 Hflag = Lflag = 0;
95 break;
96 case 'a':
97 aflag = 1;
98 break;
99 case 'c':
100 cflag = 1;
101 break;
102 case 'g':
103 blocksize = 1024 * 1024 * 1024;
104 gkmflag = 1;
105 break;
106 case 'h':
107 hflag = 1;
108 break;
109 case 'k':
110 blocksize = 1024;
111 gkmflag = 1;
112 break;
113 case 'm':
114 blocksize = 1024 * 1024;
115 gkmflag = 1;
116 break;
117 case 'n':
118 nflag = 1;
119 break;
120 case 'r':
121 break;
122 case 's':
123 sflag = 1;
124 break;
125 case 'x':
126 ftsoptions |= FTS_XDEV;
127 break;
128 case '?':
129 default:
130 usage();
131 }
132 argc -= optind;
133 argv += optind;
134
135 /*
136 * XXX
137 * Because of the way that fts(3) works, logical walks will not count
138 * the blocks actually used by symbolic links. We rationalize this by
139 * noting that users computing logical sizes are likely to do logical
140 * copies, so not counting the links is correct. The real reason is
141 * that we'd have to re-implement the kernel's symbolic link traversing
142 * algorithm to get this right. If, for example, you have relative
143 * symbolic links referencing other relative symbolic links, it gets
144 * very nasty, very fast. The bottom line is that it's documented in
145 * the man page, so it's a feature.
146 */
147 if (Hflag)
148 ftsoptions |= FTS_COMFOLLOW;
149 if (Lflag) {
150 ftsoptions &= ~FTS_PHYSICAL;
151 ftsoptions |= FTS_LOGICAL;
152 }
153
154 if (aflag) {
155 if (sflag)
156 usage();
157 listdirs = listfiles = 1;
158 } else if (sflag)
159 listdirs = listfiles = 0;
160 else {
161 listfiles = 0;
162 listdirs = 1;
163 }
164
165 if (!*argv) {
166 noargv[0] = ".";
167 noargv[1] = NULL;
168 (const char *)argv = noargv;
169 }
170
171 if (!gkmflag)
172 (void)getbsize(NULL, &blocksize);
173 blocksize /= 512;
174
175 if ((fts = fts_open(argv, ftsoptions, NULL)) == NULL)
176 err(1, "fts_open `%s'", *argv);
177
178 for (rval = 0; (p = fts_read(fts)) != NULL;) {
179 if (nflag) {
180 switch (p->fts_info) {
181 case FTS_NS:
182 case FTS_SLNONE:
183 /* nothing */
184 break;
185 default:
186 if (p->fts_statp->st_flags & UF_NODUMP) {
187 fts_set(fts, p, FTS_SKIP);
188 continue;
189 }
190 }
191 }
192 switch (p->fts_info) {
193 case FTS_D: /* Ignore. */
194 break;
195 case FTS_DP:
196 p->fts_parent->fts_number +=
197 p->fts_number += p->fts_statp->st_blocks;
198 if (cflag)
199 totalblocks += p->fts_statp->st_blocks;
200 /*
201 * If listing each directory, or not listing files
202 * or directories and this is post-order of the
203 * root of a traversal, display the total.
204 */
205 if (listdirs || (!listfiles && !p->fts_level))
206 prstat(p->fts_path, p->fts_number);
207 break;
208 case FTS_DC: /* Ignore. */
209 break;
210 case FTS_DNR: /* Warn, continue. */
211 case FTS_ERR:
212 case FTS_NS:
213 warnx("%s: %s", p->fts_path, strerror(p->fts_errno));
214 rval = 1;
215 break;
216 default:
217 if (p->fts_statp->st_nlink > 1 &&
218 linkchk(p->fts_statp->st_dev, p->fts_statp->st_ino))
219 break;
220 /*
221 * If listing each file, or a non-directory file was
222 * the root of a traversal, display the total.
223 */
224 if (listfiles || !p->fts_level)
225 prstat(p->fts_path, p->fts_statp->st_blocks);
226 p->fts_parent->fts_number += p->fts_statp->st_blocks;
227 if (cflag)
228 totalblocks += p->fts_statp->st_blocks;
229 }
230 }
231 if (errno)
232 err(1, "fts_read");
233 if (cflag)
234 prstat("total", totalblocks);
235 exit(rval);
236 }
237
238 void
239 prstat(const char *fname, int64_t blocks)
240 {
241 if (hflag) {
242 char buf[5];
243 int64_t sz = blocks * 512;
244
245 humanize_number(buf, sizeof(buf), sz, "", HN_AUTOSCALE,
246 HN_B | HN_NOSPACE | HN_DECIMAL);
247
248 (void)printf("%s\t%s\n", buf, fname);
249 } else
250 (void)printf("%lld\t%s\n",
251 (long long)howmany(blocks, (int64_t)blocksize),
252 fname);
253 }
254
255 int
256 linkchk(dev_t dev, ino_t ino)
257 {
258 static struct entry {
259 dev_t dev;
260 ino_t ino;
261 } *htable;
262 static int htshift; /* log(allocated size) */
263 static int htmask; /* allocated size - 1 */
264 static int htused; /* 2*number of insertions */
265 static int sawzero; /* Whether zero is in table or not */
266 int h, h2;
267 uint64_t tmp;
268 /* this constant is (1<<64)/((1+sqrt(5))/2)
269 * aka (word size)/(golden ratio)
270 */
271 const uint64_t HTCONST = 11400714819323198485ULL;
272 const int HTBITS = CHAR_BIT * sizeof(tmp);
273
274 /* Never store zero in hashtable */
275 if (dev == 0 && ino == 0) {
276 h = sawzero;
277 sawzero = 1;
278 return h;
279 }
280
281 /* Extend hash table if necessary, keep load under 0.5 */
282 if (htused<<1 >= htmask) {
283 struct entry *ohtable;
284
285 if (!htable)
286 htshift = 10; /* starting hashtable size */
287 else
288 htshift++; /* exponential hashtable growth */
289
290 htmask = (1 << htshift) - 1;
291 htused = 0;
292
293 ohtable = htable;
294 htable = calloc(htmask+1, sizeof(*htable));
295 if (!htable)
296 err(1, "calloc");
297
298 /* populate newly allocated hashtable */
299 if (ohtable) {
300 int i;
301 for (i = 0; i <= htmask>>1; i++)
302 if (ohtable[i].ino || ohtable[i].dev)
303 linkchk(ohtable[i].dev, ohtable[i].ino);
304 free(ohtable);
305 }
306 }
307
308 /* multiplicative hashing */
309 tmp = dev;
310 tmp <<= HTBITS>>1;
311 tmp |= ino;
312 tmp *= HTCONST;
313 h = tmp >> (HTBITS - htshift);
314 h2 = 1 | ( tmp >> (HTBITS - (htshift<<1) - 1)); /* must be odd */
315
316 /* open address hashtable search with double hash probing */
317 while (htable[h].ino || htable[h].dev) {
318 if ((htable[h].ino == ino) && (htable[h].dev == dev))
319 return 1;
320 h = (h + h2) & htmask;
321 }
322
323 /* Insert the current entry into hashtable */
324 htable[h].dev = dev;
325 htable[h].ino = ino;
326 htused++;
327 return 0;
328 }
329
330 void
331 usage(void)
332 {
333
334 (void)fprintf(stderr,
335 "usage: du [-H | -L | -P] [-a | -s] [-cghkmrx] [file ...]\n");
336 exit(1);
337 }
338