usr.bin/du/du.c

   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  * @(#) Copyright (c) 1989, 1993, 1994 The Regents of the University of California.  All rights reserved.
  37  * @(#)du.c     8.5 (Berkeley) 5/4/95
  38  * $FreeBSD: src/usr.bin/du/du.c,v 1.17.2.4 2002/12/12 16:29:39 trhodes Exp $
  39  * $DragonFly: src/usr.bin/du/du.c,v 1.9 2006/01/12 13:43:10 corecode Exp $
  40  */
  41
  42 #include <sys/param.h>
  43 #include <sys/queue.h>
  44 #include <sys/stat.h>
  45
  46 #include <err.h>
  47 #include <errno.h>
  48 #include <fnmatch.h>
  49 #include <fts.h>
  50 #include <math.h>
  51 #include <stdio.h>
  52 #include <stdlib.h>
  53 #include <string.h>
  54 #include <sysexits.h>
  55 #include <unistd.h>
  56
  57 #define KILO_SZ(n) (n)
  58 #define MEGA_SZ(n) ((n) * (n))
  59 #define GIGA_SZ(n) ((n) * (n) * (n))
  60 #define TERA_SZ(n) ((n) * (n) * (n) * (n))
  61 #define PETA_SZ(n) ((n) * (n) * (n) * (n) * (n))
  62
  63 #define KILO_2_SZ (KILO_SZ(1024ULL))
  64 #define MEGA_2_SZ (MEGA_SZ(1024ULL))
  65 #define GIGA_2_SZ (GIGA_SZ(1024ULL))
  66 #define TERA_2_SZ (TERA_SZ(1024ULL))
  67 #define PETA_2_SZ (PETA_SZ(1024ULL))
  68
  69 #define KILO_SI_SZ (KILO_SZ(1000ULL))
  70 #define MEGA_SI_SZ (MEGA_SZ(1000ULL))
  71 #define GIGA_SI_SZ (GIGA_SZ(1000ULL))
  72 #define TERA_SI_SZ (TERA_SZ(1000ULL))
  73 #define PETA_SI_SZ (PETA_SZ(1000ULL))
  74
  75 #define HASHSIZE        256             /* power of 2 only */
  76 #define HASHMASK        (HASHSIZE - 1)
  77
  78 unsigned long long vals_si [] = {1, KILO_SI_SZ, MEGA_SI_SZ, GIGA_SI_SZ, TERA_SI_SZ, PETA_SI_SZ};
  79 unsigned long long vals_base2[] = {1, KILO_2_SZ, MEGA_2_SZ, GIGA_2_SZ, TERA_2_SZ, PETA_2_SZ};
  80 unsigned long long *valp;
  81
  82 typedef enum { NONE, KILO, MEGA, GIGA, TERA, PETA, UNIT_MAX } unit_t;
  83
  84 int unitp [] = { NONE, KILO, MEGA, GIGA, TERA, PETA };
  85
  86 SLIST_HEAD(ignhead, ignentry) ignores;
  87 struct ignentry {
  88         char                    *mask;
  89         SLIST_ENTRY(ignentry)   next;
  90 };
  91
  92 static int      linkchk(FTSENT *);
  93 static void     usage(void);
  94 void            prthumanval(double);
  95 unit_t          unit_adjust(double *);
  96 void            ignoreadd(const char *);
  97 void            ignoreclean(void);
  98 int             ignorep(FTSENT *);
  99
 100 static char period[] = ".";
 101
 102 int
 103 main(int argc, char **argv)
 104 {
 105         FTS             *fts;
 106         FTSENT          *p;
 107         long            blocksize, savednumber = 0;
 108         int             ftsoptions;
 109         int             listall;
 110         int             depth;
 111         int             Hflag, Lflag, Pflag, aflag, sflag, dflag, cflag, hflag, ch, notused, rval;
 112         char            **save;
 113
 114         Hflag = Lflag = Pflag = aflag = sflag = dflag = cflag = hflag = 0;
 115
 116         save = argv;
 117         ftsoptions = 0;
 118         depth = INT_MAX;
 119         SLIST_INIT(&ignores);
 120
 121         while ((ch = getopt(argc, argv, "HI:LPasd:chkrx")) != -1)
 122                 switch (ch) {
 123                         case 'H':
 124                                 Hflag = 1;
 125                                 break;
 126                         case 'I':
 127                                 ignoreadd(optarg);
 128                                 break;
 129                         case 'L':
 130                                 if (Pflag)
 131                                         usage();
 132                                 Lflag = 1;
 133                                 break;
 134                         case 'P':
 135                                 if (Lflag)
 136                                         usage();
 137                                 Pflag = 1;
 138                                 break;
 139                         case 'a':
 140                                 aflag = 1;
 141                                 break;
 142                         case 's':
 143                                 sflag = 1;
 144                                 break;
 145                         case 'd':
 146                                 dflag = 1;
 147                                 errno = 0;
 148                                 depth = atoi(optarg);
 149                                 if (errno == ERANGE || depth < 0) {
 150                                         warnx("invalid argument to option d: %s", optarg);
 151                                         usage();
 152                                 }
 153                                 break;
 154                         case 'c':
 155                                 cflag = 1;
 156                                 break;
 157                         case 'h':
 158                                 if (putenv("BLOCKSIZE=512") == -1)
 159                                         warn("putenv: cannot set BLOCKSIZE=512");
 160                                 hflag = 1;
 161                                 valp = vals_base2;
 162                                 break;
 163                         case 'k':
 164                                 hflag = 0;
 165                                 if (putenv("BLOCKSIZE=1024") == -1)
 166                                         warn("putenv: cannot set BLOCKSIZE=1024");
 167                                 break;
 168                         case 'r':                /* Compatibility. */
 169                                 break;
 170                         case 'x':
 171                                 ftsoptions |= FTS_XDEV;
 172                                 break;
 173                         case '?':
 174                         default:
 175                                 usage();
 176                 }
 177
 178         argc -= optind;
 179         argv += optind;
 180
 181         /*
 182          * XXX
 183          * Because of the way that fts(3) works, logical walks will not count
 184          * the blocks actually used by symbolic links.  We rationalize this by
 185          * noting that users computing logical sizes are likely to do logical
 186          * copies, so not counting the links is correct.  The real reason is
 187          * that we'd have to re-implement the kernel's symbolic link traversing
 188          * algorithm to get this right.  If, for example, you have relative
 189          * symbolic links referencing other relative symbolic links, it gets
 190          * very nasty, very fast.  The bottom line is that it's documented in
 191          * the man page, so it's a feature.
 192          */
 193
 194         if (Hflag + Lflag + Pflag > 1)
 195                 usage();
 196
 197         if (Hflag + Lflag + Pflag == 0)
 198                 Pflag = 1;                      /* -P (physical) is default */
 199
 200         if (Hflag)
 201                 ftsoptions |= FTS_COMFOLLOW;
 202
 203         if (Lflag)
 204                 ftsoptions |= FTS_LOGICAL;
 205
 206         if (Pflag)
 207                 ftsoptions |= FTS_PHYSICAL;
 208
 209         listall = 0;
 210
 211         if (aflag) {
 212                 if (sflag || dflag)
 213                         usage();
 214                 listall = 1;
 215         } else if (sflag) {
 216                 if (dflag)
 217                         usage();
 218                 depth = 0;
 219         }
 220
 221         if (!*argv) {
 222                 argv = save;
 223                 argv[0] = period;
 224                 argv[1] = NULL;
 225         }
 226
 227         (void) getbsize(&notused, &blocksize);
 228         blocksize /= 512;
 229
 230         rval = 0;
 231
 232         if ((fts = fts_open(argv, ftsoptions, NULL)) == NULL)
 233                 err(1, "fts_open");
 234
 235         while ((p = fts_read(fts)) != NULL) {
 236                 switch (p->fts_info) {
 237                         case FTS_D:                     /* Ignore. */
 238                                 if (ignorep(p))
 239                                         fts_set(fts, p, FTS_SKIP);
 240                                 break;
 241                         case FTS_DP:
 242                                 if (ignorep(p))
 243                                         break;
 244
 245                                 p->fts_parent->fts_number +=
 246                                     p->fts_number += p->fts_statp->st_blocks;
 247
 248                                 if (p->fts_level <= depth) {
 249                                         if (hflag) {
 250                                                 (void) prthumanval(howmany(p->fts_number, blocksize));
 251                                                 (void) printf("\t%s\n", p->fts_path);
 252                                         } else {
 253                                         (void) printf("%ld\t%s\n",
 254                                             howmany(p->fts_number, blocksize),
 255                                             p->fts_path);
 256                                         }
 257                                 }
 258                                 break;
 259                         case FTS_DC:                    /* Ignore. */
 260                                 break;
 261                         case FTS_DNR:                   /* Warn, continue. */
 262                         case FTS_ERR:
 263                         case FTS_NS:
 264                                 warnx("%s: %s", p->fts_path, strerror(p->fts_errno));
 265                                 rval = 1;
 266                                 break;
 267                         default:
 268                                 if (ignorep(p))
 269                                         break;
 270
 271                                 if (p->fts_statp->st_nlink > 1 && linkchk(p))
 272                                         break;
 273
 274                                 if (listall || p->fts_level == 0) {
 275                                         if (hflag) {
 276                                                 (void) prthumanval(howmany(p->fts_statp->st_blocks,
 277                                                         blocksize));
 278                                                 (void) printf("\t%s\n", p->fts_path);
 279                                         } else {
 280                                                 (void) printf("%qd\t%s\n",
 281                                                         howmany(p->fts_statp->st_blocks, blocksize),
 282                                                         p->fts_path);
 283                                         }
 284                                 }
 285
 286                                 p->fts_parent->fts_number += p->fts_statp->st_blocks;
 287                 }
 288                 savednumber = p->fts_parent->fts_number;
 289         }
 290
 291         if (errno)
 292                 err(1, "fts_read");
 293
 294         if (cflag) {
 295                 if (hflag) {
 296                         (void) prthumanval(howmany(savednumber, blocksize));
 297                         (void) printf("\ttotal\n");
 298                 } else {
 299                         (void) printf("%ld\ttotal\n", howmany(savednumber, blocksize));
 300                 }
 301         }
 302
 303         ignoreclean();
 304         exit(rval);
 305 }
 306
 307 static int
 308 linkchk(FTSENT *p)
 309 {
 310         struct links_entry {
 311                 struct links_entry *next;
 312                 struct links_entry *previous;
 313                 int             links;
 314                 dev_t           dev;
 315                 ino_t           ino;
 316         };
 317
 318         static const size_t links_hash_initial_size = 8192;
 319         static struct links_entry **buckets;
 320         static struct links_entry *free_list;
 321         static size_t number_buckets;
 322         static unsigned long number_entries;
 323         static char stop_allocating;
 324         struct links_entry *le, **new_buckets;
 325         struct stat *st;
 326         size_t i, new_size;
 327         int hash;
 328
 329         st = p->fts_statp;
 330
 331         /* If necessary, initialize the hash table. */
 332         if (buckets == NULL) {
 333                 number_buckets = links_hash_initial_size;
 334                 buckets = malloc(number_buckets * sizeof(buckets[0]));
 335                 if (buckets == NULL)
 336                         errx(1, "No memory for hardlink detection");
 337                 for (i = 0; i < number_buckets; i++)
 338                         buckets[i] = NULL;
 339         }
 340
 341         /* If the hash table is getting too full, enlarge it. */
 342         if (number_entries > number_buckets * 10 && !stop_allocating) {
 343                 new_size = number_buckets * 2;
 344                 new_buckets = malloc(new_size * sizeof(struct links_entry *));
 345
 346                 /* Try releasing the free list to see if that helps. */
 347                 if (new_buckets == NULL && free_list != NULL) {
 348                         while (free_list != NULL) {
 349                                 le = free_list;
 350                                 free_list = le->next;
 351                                 free(le);
 352                         }
 353                         new_buckets = malloc(new_size * sizeof(new_buckets[0]));
 354                 }
 355
 356                 if (new_buckets == NULL) {
 357                         stop_allocating = 1;
 358                         warnx("No more memory for tracking hard links");
 359                 } else {
 360                         memset(new_buckets, 0, new_size * sizeof(struct links_entry *));
 361                         for (i = 0; i < number_buckets; i++) {
 362                                 while (buckets[i] != NULL) {
 363                                         /* Remove entry from old bucket. */
 364                                         le = buckets[i];
 365                                         buckets[i] = le->next;
 366
 367                                         /* Add entry to new bucket. */
 368                                         hash = (le->dev ^ le->ino) % new_size;
 369
 370                                         if (new_buckets[hash] != NULL)
 371                                                 new_buckets[hash]->previous = le;
 372                                         le->next = new_buckets[hash];
 373                                         le->previous = NULL;
 374                                         new_buckets[hash] = le;
 375                                 }
 376                         }
 377                         free(buckets);
 378                         buckets = new_buckets;
 379                         number_buckets = new_size;
 380                 }
 381         }
 382
 383         /* Try to locate this entry in the hash table. */
 384         hash = ( st->st_dev ^ st->st_ino ) % number_buckets;
 385         for (le = buckets[hash]; le != NULL; le = le->next) {
 386                 if (le->dev == st->st_dev && le->ino == st->st_ino) {
 387                         /*
 388                          * Save memory by releasing an entry when we've seen
 389                          * all of it's links.
 390                          */
 391                         if (--le->links <= 0) {
 392                                 if (le->previous != NULL)
 393                                         le->previous->next = le->next;
 394                                 if (le->next != NULL)
 395                                         le->next->previous = le->previous;
 396                                 if (buckets[hash] == le)
 397                                         buckets[hash] = le->next;
 398                                 number_entries--;
 399                                 /* Recycle this node through the free list */
 400                                 if (stop_allocating) {
 401                                         free(le);
 402                                 } else {
 403                                         le->next = free_list;
 404                                         free_list = le;
 405                                 }
 406                         }
 407                         return (1);
 408                 }
 409         }
 410
 411         if (stop_allocating)
 412                 return (0);
 413
 414         /* Add this entry to the links cache. */
 415         if (free_list != NULL) {
 416                 /* Pull a node from the free list if we can. */
 417                 le = free_list;
 418                 free_list = le->next;
 419         } else
 420                 /* Malloc one if we have to. */
 421                 le = malloc(sizeof(struct links_entry));
 422         if (le == NULL) {
 423                 stop_allocating = 1;
 424                 warnx("No more memory for tracking hard links");
 425                 return (0);
 426         }
 427         le->dev = st->st_dev;
 428         le->ino = st->st_ino;
 429         le->links = st->st_nlink - 1;
 430         number_entries++;
 431         le->next = buckets[hash];
 432         le->previous = NULL;
 433         if (buckets[hash] != NULL)
 434                 buckets[hash]->previous = le;
 435         buckets[hash] = le;
 436         return (0);
 437 }
 438
 439 /*
 440  * Output in "human-readable" format.  Uses 3 digits max and puts
 441  * unit suffixes at the end.  Makes output compact and easy to read,
 442  * especially on huge disks.
 443  *
 444  */
 445 unit_t
 446 unit_adjust(double *val)
 447 {
 448         double abval;
 449         unit_t unit;
 450         unsigned int unit_sz;
 451
 452         abval = fabs(*val);
 453
 454         unit_sz = abval ? ilogb(abval) / 10 : 0;
 455
 456         if (unit_sz >= UNIT_MAX) {
 457                 unit = NONE;
 458         } else {
 459                 unit = unitp[unit_sz];
 460                 *val /= (double)valp[unit_sz];
 461         }
 462
 463         return (unit);
 464 }
 465
 466 void
 467 prthumanval(double bytes)
 468 {
 469         unit_t unit;
 470
 471         bytes *= 512;
 472         unit = unit_adjust(&bytes);
 473
 474         if (bytes == 0)
 475                 (void)printf("  0B");
 476         else if (bytes > 10)
 477                 (void)printf("%3.0f%c", bytes, "BKMGTPE"[unit]);
 478         else
 479                 (void)printf("%3.1f%c", bytes, "BKMGTPE"[unit]);
 480 }
 481
 482 static void
 483 usage(void)
 484 {
 485         (void)fprintf(stderr,
 486                 "usage: du [-H | -L | -P] [-a | -s | -d depth] [-c] [-h | -k] [-x] [-I mask] [file ...]\n");
 487         exit(EX_USAGE);
 488 }
 489
 490 void
 491 ignoreadd(const char *mask)
 492 {
 493         struct ignentry *ign;
 494
 495         ign = calloc(1, sizeof(*ign));
 496         if (ign == NULL)
 497                 errx(1, "cannot allocate memory");
 498         ign->mask = strdup(mask);
 499         if (ign->mask == NULL)
 500                 errx(1, "cannot allocate memory");
 501         SLIST_INSERT_HEAD(&ignores, ign, next);
 502 }
 503
 504 void
 505 ignoreclean(void)
 506 {
 507         struct ignentry *ign;
 508
 509         while (!SLIST_EMPTY(&ignores)) {
 510                 ign = SLIST_FIRST(&ignores);
 511                 SLIST_REMOVE_HEAD(&ignores, next);
 512                 free(ign->mask);
 513                 free(ign);
 514         }
 515 }
 516
 517 int
 518 ignorep(FTSENT *ent)
 519 {
 520         struct ignentry *ign;
 521
 522         SLIST_FOREACH(ign, &ignores, next)
 523                 if (fnmatch(ign->mask, ent->fts_name, 0) != FNM_NOMATCH)
 524                         return 1;
 525         return 0;
 526 }