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 1.10 2008/07/13 03:37:43 dillon 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 typedef long long       du_number_t;
 103
 104 int
 105 main(int argc, char **argv)
 106 {
 107         FTS             *fts;
 108         FTSENT          *p;
 109         long            blocksize;
 110         du_number_t     savednumber = 0;
 111         int             ftsoptions;
 112         int             listall;
 113         int             depth;
 114         int             Hflag, Lflag, Pflag, aflag, sflag, dflag, cflag, hflag, ch, notused, rval;
 115         char            **save;
 116
 117         Hflag = Lflag = Pflag = aflag = sflag = dflag = cflag = hflag = 0;
 118
 119         save = argv;
 120         ftsoptions = 0;
 121         depth = INT_MAX;
 122         SLIST_INIT(&ignores);
 123
 124         while ((ch = getopt(argc, argv, "HI:LPasd:chkrx")) != -1)
 125                 switch (ch) {
 126                         case 'H':
 127                                 Hflag = 1;
 128                                 break;
 129                         case 'I':
 130                                 ignoreadd(optarg);
 131                                 break;
 132                         case 'L':
 133                                 if (Pflag)
 134                                         usage();
 135                                 Lflag = 1;
 136                                 break;
 137                         case 'P':
 138                                 if (Lflag)
 139                                         usage();
 140                                 Pflag = 1;
 141                                 break;
 142                         case 'a':
 143                                 aflag = 1;
 144                                 break;
 145                         case 's':
 146                                 sflag = 1;
 147                                 break;
 148                         case 'd':
 149                                 dflag = 1;
 150                                 errno = 0;
 151                                 depth = atoi(optarg);
 152                                 if (errno == ERANGE || depth < 0) {
 153                                         warnx("invalid argument to option d: %s", optarg);
 154                                         usage();
 155                                 }
 156                                 break;
 157                         case 'c':
 158                                 cflag = 1;
 159                                 break;
 160                         case 'h':
 161                                 if (putenv("BLOCKSIZE=512") == -1)
 162                                         warn("putenv: cannot set BLOCKSIZE=512");
 163                                 hflag = 1;
 164                                 valp = vals_base2;
 165                                 break;
 166                         case 'k':
 167                                 hflag = 0;
 168                                 if (putenv("BLOCKSIZE=1024") == -1)
 169                                         warn("putenv: cannot set BLOCKSIZE=1024");
 170                                 break;
 171                         case 'r':                /* Compatibility. */
 172                                 break;
 173                         case 'x':
 174                                 ftsoptions |= FTS_XDEV;
 175                                 break;
 176                         case '?':
 177                         default:
 178                                 usage();
 179                 }
 180
 181         argc -= optind;
 182         argv += optind;
 183
 184         /*
 185          * XXX
 186          * Because of the way that fts(3) works, logical walks will not count
 187          * the blocks actually used by symbolic links.  We rationalize this by
 188          * noting that users computing logical sizes are likely to do logical
 189          * copies, so not counting the links is correct.  The real reason is
 190          * that we'd have to re-implement the kernel's symbolic link traversing
 191          * algorithm to get this right.  If, for example, you have relative
 192          * symbolic links referencing other relative symbolic links, it gets
 193          * very nasty, very fast.  The bottom line is that it's documented in
 194          * the man page, so it's a feature.
 195          */
 196
 197         if (Hflag + Lflag + Pflag > 1)
 198                 usage();
 199
 200         if (Hflag + Lflag + Pflag == 0)
 201                 Pflag = 1;                      /* -P (physical) is default */
 202
 203         if (Hflag)
 204                 ftsoptions |= FTS_COMFOLLOW;
 205
 206         if (Lflag)
 207                 ftsoptions |= FTS_LOGICAL;
 208
 209         if (Pflag)
 210                 ftsoptions |= FTS_PHYSICAL;
 211
 212         listall = 0;
 213
 214         if (aflag) {
 215                 if (sflag || dflag)
 216                         usage();
 217                 listall = 1;
 218         } else if (sflag) {
 219                 if (dflag)
 220                         usage();
 221                 depth = 0;
 222         }
 223
 224         if (!*argv) {
 225                 argv = save;
 226                 argv[0] = period;
 227                 argv[1] = NULL;
 228         }
 229
 230         (void) getbsize(&notused, &blocksize);
 231         blocksize /= 512;
 232
 233         rval = 0;
 234
 235         if ((fts = fts_open(argv, ftsoptions, NULL)) == NULL)
 236                 err(1, "fts_open");
 237
 238         while ((p = fts_read(fts)) != NULL) {
 239                 switch (p->fts_info) {
 240                         case FTS_D:                     /* Ignore. */
 241                                 if (ignorep(p))
 242                                         fts_set(fts, p, FTS_SKIP);
 243                                 break;
 244                         case FTS_DP:
 245                                 if (ignorep(p))
 246                                         break;
 247
 248                                 if (p->fts_pointer == NULL) {
 249                                         p->fts_pointer = malloc(sizeof(du_number_t));
 250                                         *(du_number_t *)p->fts_pointer = 0;
 251                                 }
 252                                 *(du_number_t *)p->fts_pointer += p->fts_statp->st_blocks;
 253
 254                                 if (p->fts_parent->fts_pointer == NULL) {
 255                                         p->fts_parent->fts_pointer = malloc(sizeof(du_number_t));
 256                                         *(du_number_t *)p->fts_parent->fts_pointer = 0;
 257                                 }
 258                                 *(du_number_t *)p->fts_parent->fts_pointer += *(du_number_t *)p->fts_pointer += p->fts_statp->st_blocks;
 259
 260                                 if (p->fts_level <= depth) {
 261                                         if (hflag) {
 262                                                 (void) prthumanval(howmany(*(du_number_t *)p->fts_pointer, blocksize));
 263                                                 (void) printf("\t%s\n", p->fts_path);
 264                                         } else {
 265                                         (void) printf("%lld\t%s\n",
 266                                             howmany(*(du_number_t *)p->fts_pointer, blocksize),
 267                                             p->fts_path);
 268                                         }
 269                                 }
 270                                 if (p->fts_pointer) {
 271                                         free(p->fts_pointer);
 272                                         p->fts_pointer = NULL;
 273                                 }
 274                                 break;
 275                         case FTS_DC:                    /* Ignore. */
 276                                 break;
 277                         case FTS_DNR:                   /* Warn, continue. */
 278                         case FTS_ERR:
 279                         case FTS_NS:
 280                                 warnx("%s: %s", p->fts_path, strerror(p->fts_errno));
 281                                 rval = 1;
 282                                 break;
 283                         default:
 284                                 if (ignorep(p))
 285                                         break;
 286
 287                                 if (p->fts_statp->st_nlink > 1 && linkchk(p))
 288                                         break;
 289
 290                                 if (listall || p->fts_level == 0) {
 291                                         if (hflag) {
 292                                                 (void) prthumanval(howmany(p->fts_statp->st_blocks,
 293                                                         blocksize));
 294                                                 (void) printf("\t%s\n", p->fts_path);
 295                                         } else {
 296                                                 (void) printf("%lld\t%s\n",
 297                                                         howmany((long long)p->fts_statp->st_blocks, blocksize),
 298                                                         p->fts_path);
 299                                         }
 300                                 }
 301                                 if (p->fts_parent->fts_pointer == NULL) {
 302                                         p->fts_parent->fts_pointer = malloc(sizeof(du_number_t));
 303                                         *(du_number_t *)p->fts_parent->fts_pointer = 0;
 304                                 }
 305                                 *(du_number_t *)p->fts_parent->fts_pointer += p->fts_statp->st_blocks;
 306                 }
 307                 if (p->fts_parent->fts_pointer)
 308                         savednumber = *(du_number_t *)p->fts_parent->fts_pointer;
 309         }
 310
 311         if (errno)
 312                 err(1, "fts_read");
 313
 314         if (cflag) {
 315                 if (hflag) {
 316                         (void) prthumanval(howmany(savednumber, blocksize));
 317                         (void) printf("\ttotal\n");
 318                 } else {
 319                         (void) printf("%lld\ttotal\n", howmany(savednumber, blocksize));
 320                 }
 321         }
 322
 323         ignoreclean();
 324         exit(rval);
 325 }
 326
 327 static int
 328 linkchk(FTSENT *p)
 329 {
 330         struct links_entry {
 331                 struct links_entry *next;
 332                 struct links_entry *previous;
 333                 int             links;
 334                 dev_t           dev;
 335                 ino_t           ino;
 336         };
 337
 338         static const size_t links_hash_initial_size = 8192;
 339         static struct links_entry **buckets;
 340         static struct links_entry *free_list;
 341         static size_t number_buckets;
 342         static unsigned long number_entries;
 343         static char stop_allocating;
 344         struct links_entry *le, **new_buckets;
 345         struct stat *st;
 346         size_t i, new_size;
 347         int hash;
 348
 349         st = p->fts_statp;
 350
 351         /* If necessary, initialize the hash table. */
 352         if (buckets == NULL) {
 353                 number_buckets = links_hash_initial_size;
 354                 buckets = malloc(number_buckets * sizeof(buckets[0]));
 355                 if (buckets == NULL)
 356                         errx(1, "No memory for hardlink detection");
 357                 for (i = 0; i < number_buckets; i++)
 358                         buckets[i] = NULL;
 359         }
 360
 361         /* If the hash table is getting too full, enlarge it. */
 362         if (number_entries > number_buckets * 10 && !stop_allocating) {
 363                 new_size = number_buckets * 2;
 364                 new_buckets = malloc(new_size * sizeof(struct links_entry *));
 365
 366                 /* Try releasing the free list to see if that helps. */
 367                 if (new_buckets == NULL && free_list != NULL) {
 368                         while (free_list != NULL) {
 369                                 le = free_list;
 370                                 free_list = le->next;
 371                                 free(le);
 372                         }
 373                         new_buckets = malloc(new_size * sizeof(new_buckets[0]));
 374                 }
 375
 376                 if (new_buckets == NULL) {
 377                         stop_allocating = 1;
 378                         warnx("No more memory for tracking hard links");
 379                 } else {
 380                         memset(new_buckets, 0, new_size * sizeof(struct links_entry *));
 381                         for (i = 0; i < number_buckets; i++) {
 382                                 while (buckets[i] != NULL) {
 383                                         /* Remove entry from old bucket. */
 384                                         le = buckets[i];
 385                                         buckets[i] = le->next;
 386
 387                                         /* Add entry to new bucket. */
 388                                         hash = (le->dev ^ le->ino) % new_size;
 389
 390                                         if (new_buckets[hash] != NULL)
 391                                                 new_buckets[hash]->previous = le;
 392                                         le->next = new_buckets[hash];
 393                                         le->previous = NULL;
 394                                         new_buckets[hash] = le;
 395                                 }
 396                         }
 397                         free(buckets);
 398                         buckets = new_buckets;
 399                         number_buckets = new_size;
 400                 }
 401         }
 402
 403         /* Try to locate this entry in the hash table. */
 404         hash = ( st->st_dev ^ st->st_ino ) % number_buckets;
 405         for (le = buckets[hash]; le != NULL; le = le->next) {
 406                 if (le->dev == st->st_dev && le->ino == st->st_ino) {
 407                         /*
 408                          * Save memory by releasing an entry when we've seen
 409                          * all of it's links.
 410                          */
 411                         if (--le->links <= 0) {
 412                                 if (le->previous != NULL)
 413                                         le->previous->next = le->next;
 414                                 if (le->next != NULL)
 415                                         le->next->previous = le->previous;
 416                                 if (buckets[hash] == le)
 417                                         buckets[hash] = le->next;
 418                                 number_entries--;
 419                                 /* Recycle this node through the free list */
 420                                 if (stop_allocating) {
 421                                         free(le);
 422                                 } else {
 423                                         le->next = free_list;
 424                                         free_list = le;
 425                                 }
 426                         }
 427                         return (1);
 428                 }
 429         }
 430
 431         if (stop_allocating)
 432                 return (0);
 433
 434         /* Add this entry to the links cache. */
 435         if (free_list != NULL) {
 436                 /* Pull a node from the free list if we can. */
 437                 le = free_list;
 438                 free_list = le->next;
 439         } else
 440                 /* Malloc one if we have to. */
 441                 le = malloc(sizeof(struct links_entry));
 442         if (le == NULL) {
 443                 stop_allocating = 1;
 444                 warnx("No more memory for tracking hard links");
 445                 return (0);
 446         }
 447         le->dev = st->st_dev;
 448         le->ino = st->st_ino;
 449         le->links = st->st_nlink - 1;
 450         number_entries++;
 451         le->next = buckets[hash];
 452         le->previous = NULL;
 453         if (buckets[hash] != NULL)
 454                 buckets[hash]->previous = le;
 455         buckets[hash] = le;
 456         return (0);
 457 }
 458
 459 /*
 460  * Output in "human-readable" format.  Uses 3 digits max and puts
 461  * unit suffixes at the end.  Makes output compact and easy to read,
 462  * especially on huge disks.
 463  *
 464  */
 465 unit_t
 466 unit_adjust(double *val)
 467 {
 468         double abval;
 469         unit_t unit;
 470         unsigned int unit_sz;
 471
 472         abval = fabs(*val);
 473
 474         unit_sz = abval ? ilogb(abval) / 10 : 0;
 475
 476         if (unit_sz >= UNIT_MAX) {
 477                 unit = NONE;
 478         } else {
 479                 unit = unitp[unit_sz];
 480                 *val /= (double)valp[unit_sz];
 481         }
 482
 483         return (unit);
 484 }
 485
 486 void
 487 prthumanval(double bytes)
 488 {
 489         unit_t unit;
 490
 491         bytes *= 512;
 492         unit = unit_adjust(&bytes);
 493
 494         if (bytes == 0)
 495                 (void)printf("  0B");
 496         else if (bytes > 10)
 497                 (void)printf("%3.0f%c", bytes, "BKMGTPE"[unit]);
 498         else
 499                 (void)printf("%3.1f%c", bytes, "BKMGTPE"[unit]);
 500 }
 501
 502 static void
 503 usage(void)
 504 {
 505         (void)fprintf(stderr,
 506                 "usage: du [-H | -L | -P] [-a | -s | -d depth] [-c] [-h | -k] [-x] [-I mask] [file ...]\n");
 507         exit(EX_USAGE);
 508 }
 509
 510 void
 511 ignoreadd(const char *mask)
 512 {
 513         struct ignentry *ign;
 514
 515         ign = calloc(1, sizeof(*ign));
 516         if (ign == NULL)
 517                 errx(1, "cannot allocate memory");
 518         ign->mask = strdup(mask);
 519         if (ign->mask == NULL)
 520                 errx(1, "cannot allocate memory");
 521         SLIST_INSERT_HEAD(&ignores, ign, next);
 522 }
 523
 524 void
 525 ignoreclean(void)
 526 {
 527         struct ignentry *ign;
 528
 529         while (!SLIST_EMPTY(&ignores)) {
 530                 ign = SLIST_FIRST(&ignores);
 531                 SLIST_REMOVE_HEAD(&ignores, next);
 532                 free(ign->mask);
 533                 free(ign);
 534         }
 535 }
 536
 537 int
 538 ignorep(FTSENT *ent)
 539 {
 540         struct ignentry *ign;
 541
 542         SLIST_FOREACH(ign, &ignores, next)
 543                 if (fnmatch(ign->mask, ent->fts_name, 0) != FNM_NOMATCH)
 544                         return 1;
 545         return 0;
 546 }