pax/pax.c

   1 /*      $OpenBSD: pax.c,v 1.28 2005/08/04 10:02:44 mpf Exp $    */
   2 /*      $NetBSD: pax.c,v 1.5 1996/03/26 23:54:20 mrg Exp $      */
   3
   4 /*-
   5  * Copyright (c) 1992 Keith Muller.
   6  * Copyright (c) 1992, 1993
   7  *      The Regents of the University of California.  All rights reserved.
   8  *
   9  * This code is derived from software contributed to Berkeley by
  10  * Keith Muller of the University of California, San Diego.
  11  *
  12  * Redistribution and use in source and binary forms, with or without
  13  * modification, are permitted provided that the following conditions
  14  * are met:
  15  * 1. Redistributions of source code must retain the above copyright
  16  *    notice, this list of conditions and the following disclaimer.
  17  * 2. Redistributions in binary form must reproduce the above copyright
  18  *    notice, this list of conditions and the following disclaimer in the
  19  *    documentation and/or other materials provided with the distribution.
  20  * 3. 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 #include <sys/cdefs.h>
  38 #ifndef lint
  39 __used static const char copyright[] =
  40 "@(#) Copyright (c) 1992, 1993\n\
  41         The Regents of the University of California.  All rights reserved.\n";
  42 #endif /* not lint */
  43
  44 #ifndef lint
  45 #if 0
  46 static const char sccsid[] = "@(#)pax.c 8.2 (Berkeley) 4/18/94";
  47 #else
  48 __used static const char rcsid[] = "$OpenBSD: pax.c,v 1.28 2005/08/04 10:02:44 mpf Exp $";
  49 #endif
  50 #endif /* not lint */
  51
  52 #include <stdio.h>
  53 #include <sys/types.h>
  54 #include <sys/param.h>
  55 #include <sys/stat.h>
  56 #include <sys/time.h>
  57 #include <sys/resource.h>
  58 #include <signal.h>
  59 #include <unistd.h>
  60 #include <stdlib.h>
  61 #include <string.h>
  62 #include <errno.h>
  63 #include <err.h>
  64 #include <fcntl.h>
  65 #include <paths.h>
  66 #include <locale.h>
  67 #include "pax.h"
  68 #include "extern.h"
  69 static int gen_init(void);
  70
  71 /*
  72  * PAX main routines, general globals and some simple start up routines
  73  */
  74
  75 /*
  76  * Variables that can be accessed by any routine within pax
  77  */
  78 int     act = DEFOP;            /* read/write/append/copy */
  79 const FSUB      *frmt = NULL;           /* archive format type */
  80 int     cflag;                  /* match all EXCEPT pattern/file */
  81 int     cwdfd;                  /* starting cwd */
  82 int     dflag;                  /* directory member match only  */
  83 int     iflag;                  /* interactive file/archive rename */
  84 int     kflag;                  /* do not overwrite existing files */
  85 int     lflag;                  /* use hard links when possible */
  86 int     nflag;                  /* select first archive member match */
  87 int     tflag;                  /* restore access time after read */
  88 int     uflag;                  /* ignore older modification time files */
  89 int     vflag;                  /* produce verbose output */
  90 int     Dflag;                  /* same as uflag except inode change time */
  91 int     Hflag;                  /* follow command line symlinks (write only) */
  92 int     Lflag;                  /* follow symlinks when writing */
  93 int     Xflag;                  /* archive files with same device id only */
  94 int     Yflag;                  /* same as Dflag except after name mode */
  95 int     Zflag;                  /* same as uflag except after name mode */
  96 int     zeroflag;               /* use \0 as pathname terminator */
  97 int     vfpart;                 /* is partial verbose output in progress */
  98 int     patime = 1;             /* preserve file access time */
  99 int     pmtime = 1;             /* preserve file modification times */
 100 int     nodirs;                 /* do not create directories as needed */
 101 int     pmode;                  /* preserve file mode bits */
 102 int     pids;                   /* preserve file uid/gid */
 103 int     rmleadslash = 0;        /* remove leading '/' from pathnames */
 104 int     secure = 1;             /* don't extract names that contain .. */
 105 int     exit_val;               /* exit value */
 106 int     docrc;                  /* check/create file crc */
 107 char    *dirptr;                /* destination dir in a copy */
 108 char    *ltmfrmt;               /* -v locale time format (if any) */
 109 char    *argv0;                 /* root of argv[0] */
 110 sigset_t s_mask;                /* signal mask for cleanup critical sect */
 111 FILE    *listf;                 /* file pointer to print file list to */
 112 char    *tempfile;              /* tempfile to use for mkstemp(3) */
 113 char    *tempbase;              /* basename of tempfile to use for mkstemp(3) */
 114
 115 /*
 116  *      PAX - Portable Archive Interchange
 117  *
 118  *      A utility to read, write, and write lists of the members of archive
 119  *      files and copy directory hierarchies. A variety of archive formats
 120  *      are supported (some are described in POSIX 1003.1 10.1):
 121  *
 122  *              ustar - 10.1.1 extended tar interchange format
 123  *              cpio  - 10.1.2 extended cpio interchange format
 124  *              tar - old BSD 4.3 tar format
 125  *              binary cpio - old cpio with binary header format
 126  *              sysVR4 cpio -  with and without CRC
 127  *
 128  * This version is a superset of IEEE Std 1003.2b-d3
 129  *
 130  * Summary of Extensions to the IEEE Standard:
 131  *
 132  * 1    READ ENHANCEMENTS
 133  * 1.1  Operations which read archives will continue to operate even when
 134  *      processing archives which may be damaged, truncated, or fail to meet
 135  *      format specs in several different ways. Damaged sections of archives
 136  *      are detected and avoided if possible. Attempts will be made to resync
 137  *      archive read operations even with badly damaged media.
 138  * 1.2  Blocksize requirements are not strictly enforced on archive read.
 139  *      Tapes which have variable sized records can be read without errors.
 140  * 1.3  The user can specify via the non-standard option flag -E if error
 141  *      resync operation should stop on a media error, try a specified number
 142  *      of times to correct, or try to correct forever.
 143  * 1.4  Sparse files (lseek holes) stored on the archive (but stored with blocks
 144  *      of all zeros will be restored with holes appropriate for the target
 145  *      filesystem
 146  * 1.5  The user is notified whenever something is found during archive
 147  *      read operations which violates spec (but the read will continue).
 148  * 1.6  Multiple archive volumes can be read and may span over different
 149  *      archive devices
 150  * 1.7  Rigidly restores all file attributes exactly as they are stored on the
 151  *      archive.
 152  * 1.8  Modification change time ranges can be specified via multiple -T
 153  *      options. These allow a user to select files whose modification time
 154  *      lies within a specific time range.
 155  * 1.9  Files can be selected based on owner (user name or uid) via one or more
 156  *      -U options.
 157  * 1.10 Files can be selected based on group (group name or gid) via one o
 158  *      more -G options.
 159  * 1.11 File modification time can be checked against existing file after
 160  *      name modification (-Z)
 161  *
 162  * 2    WRITE ENHANCEMENTS
 163  * 2.1  Write operation will stop instead of allowing a user to create a flawed
 164  *      flawed archive (due to any problem).
 165  * 2.2  Archives written by pax are forced to strictly conform to both the
 166  *      archive and pax the specific format specifications.
 167  * 2.3  Blocking size and format is rigidly enforced on writes.
 168  * 2.4  Formats which may exhibit header overflow problems (they have fields
 169  *      too small for large file systems, such as inode number storage), use
 170  *      routines designed to repair this problem. These techniques still
 171  *      conform to both pax and format specifications, but no longer truncate
 172  *      these fields. This removes any restrictions on using these archive
 173  *      formats on large file systems.
 174  * 2.5  Multiple archive volumes can be written and may span over different
 175  *      archive devices
 176  * 2.6  A archive volume record limit allows the user to specify the number
 177  *      of bytes stored on an archive volume. When reached the user is
 178  *      prompted for the next archive volume. This is specified with the
 179  *      non-standard -B flag. The limit is rounded up to the next blocksize.
 180  * 2.7  All archive padding during write use zero filled sections. This makes
 181  *      it much easier to pull data out of flawed archive during read
 182  *      operations.
 183  * 2.8  Access time reset with the -t applies to all file nodes (including
 184  *      directories).
 185  * 2.9  Symbolic links can be followed with -L (optional in the spec).
 186  * 2.10 Modification or inode change time ranges can be specified via
 187  *      multiple -T options. These allow a user to select files whose
 188  *      modification or inode change time lies within a specific time range.
 189  * 2.11 Files can be selected based on owner (user name or uid) via one or more
 190  *      -U options.
 191  * 2.12 Files can be selected based on group (group name or gid) via one o
 192  *      more -G options.
 193  * 2.13 Symlinks which appear on the command line can be followed (without
 194  *      following other symlinks; -H flag)
 195  *
 196  * 3    COPY ENHANCEMENTS
 197  * 3.1  Sparse files (lseek holes) can be copied without expanding the holes
 198  *      into zero filled blocks. The file copy is created with holes which are
 199  *      appropriate for the target filesystem
 200  * 3.2  Access time as well as modification time on copied file trees can be
 201  *      preserved with the appropriate -p options.
 202  * 3.3  Access time reset with the -t applies to all file nodes (including
 203  *      directories).
 204  * 3.4  Symbolic links can be followed with -L (optional in the spec).
 205  * 3.5  Modification or inode change time ranges can be specified via
 206  *      multiple -T options. These allow a user to select files whose
 207  *      modification or inode change time lies within a specific time range.
 208  * 3.6  Files can be selected based on owner (user name or uid) via one or more
 209  *      -U options.
 210  * 3.7  Files can be selected based on group (group name or gid) via one o
 211  *      more -G options.
 212  * 3.8  Symlinks which appear on the command line can be followed (without
 213  *      following other symlinks; -H flag)
 214  * 3.9  File inode change time can be checked against existing file before
 215  *      name modification (-D)
 216  * 3.10 File inode change time can be checked against existing file after
 217  *      name modification (-Y)
 218  * 3.11 File modification time can be checked against existing file after
 219  *      name modification (-Z)
 220  *
 221  * 4    GENERAL ENHANCEMENTS
 222  * 4.1  Internal structure is designed to isolate format dependent and
 223  *      independent functions. Formats are selected via a format driver table.
 224  *      This encourages the addition of new archive formats by only having to
 225  *      write those routines which id, read and write the archive header.
 226  */
 227
 228 /*
 229  * main()
 230  *      parse options, set up and operate as specified by the user.
 231  *      any operational flaw will set exit_val to non-zero
 232  * Return: 0 if ok, 1 otherwise
 233  */
 234
 235 int
 236 main(int argc, char **argv)
 237 {
 238         char *tmpdir;
 239         size_t tdlen;
 240 #ifdef _HAVE_REGCOMP_
 241         setlocale(LC_CTYPE, "");
 242         setlocale(LC_COLLATE, "");
 243 #endif
 244
 245         listf = stderr;
 246         /*
 247          * Keep a reference to cwd, so we can always come back home.
 248          */
 249         cwdfd = open(".", O_RDONLY);
 250         if (cwdfd < 0) {
 251                 syswarn(1, errno, "Can't open current working directory.");
 252                 return(exit_val);
 253         }
 254
 255         if (updatepath() == -1)
 256                 return exit_val;
 257         /*
 258          * Where should we put temporary files?
 259          */
 260         if ((tmpdir = getenv("TMPDIR")) == NULL || *tmpdir == '\0')
 261                 tmpdir = _PATH_TMP;
 262         tdlen = strlen(tmpdir);
 263         while (tdlen > 0 && tmpdir[tdlen - 1] == '/')
 264                 tdlen--;
 265         tempfile = malloc(tdlen + 1 + sizeof(_TFILE_BASE));
 266         if (tempfile == NULL) {
 267                 paxwarn(1, "Cannot allocate memory for temp file name.");
 268                 return(exit_val);
 269         }
 270         if (tdlen)
 271                 memcpy(tempfile, tmpdir, tdlen);
 272         tempbase = tempfile + tdlen;
 273         *tempbase++ = '/';
 274
 275         /*
 276          * parse options, determine operational mode, general init
 277          */
 278         options(argc, argv);
 279         if ((gen_init() < 0) || (tty_init() < 0))
 280                 return(exit_val);
 281
 282         /*
 283          * select a primary operation mode
 284          */
 285         switch (act) {
 286         case EXTRACT:
 287                 extract();
 288                 break;
 289         case ARCHIVE:
 290                 archive();
 291                 break;
 292         case APPND:
 293                 if (gzip_program != NULL)
 294                         errx(1, "can not gzip while appending");
 295                 append();
 296                 break;
 297         case COPY:
 298                 copy();
 299                 break;
 300         default:
 301         case LIST:
 302                 list();
 303                 break;
 304         }
 305         return(exit_val);
 306 }
 307
 308 /*
 309  * sig_cleanup()
 310  *      when interrupted we try to do whatever delayed processing we can.
 311  *      This is not critical, but we really ought to limit our damage when we
 312  *      are aborted by the user.
 313  * Return:
 314  *      never....
 315  */
 316
 317 void
 318 sig_cleanup(int which_sig)
 319 {
 320         /* XXX signal races */
 321
 322         /*
 323          * restore modes and times for any dirs we may have created
 324          * or any dirs we may have read. Set vflag and vfpart so the user
 325          * will clearly see the message on a line by itself.
 326          */
 327         vflag = vfpart = 1;
 328         if (which_sig == SIGXCPU)
 329                 paxwarn(0, "Cpu time limit reached, cleaning up.");
 330         else
 331                 paxwarn(0, "Signal caught, cleaning up.");
 332
 333         ar_close();
 334         proc_dir();
 335         if (tflag)
 336                 atdir_end();
 337         exit(1);
 338 }
 339
 340 /*
 341  * gen_init()
 342  *      general setup routines. Not all are required, but they really help
 343  *      when dealing with a medium to large sized archives.
 344  */
 345
 346 static int
 347 gen_init(void)
 348 {
 349         struct rlimit reslimit;
 350         struct sigaction n_hand;
 351         struct sigaction o_hand;
 352
 353         /*
 354          * Really needed to handle large archives. We can run out of memory for
 355          * internal tables really fast when we have a whole lot of files...
 356          */
 357         if (getrlimit(RLIMIT_DATA , &reslimit) == 0){
 358                 reslimit.rlim_cur = reslimit.rlim_max;
 359                 (void)setrlimit(RLIMIT_DATA , &reslimit);
 360         }
 361
 362         /*
 363          * should file size limits be waived? if the os limits us, this is
 364          * needed if we want to write a large archive
 365          */
 366         if (getrlimit(RLIMIT_FSIZE , &reslimit) == 0){
 367                 reslimit.rlim_cur = reslimit.rlim_max;
 368                 (void)setrlimit(RLIMIT_FSIZE , &reslimit);
 369         }
 370
 371         /*
 372          * increase the size the stack can grow to
 373          */
 374         if (getrlimit(RLIMIT_STACK , &reslimit) == 0){
 375                 reslimit.rlim_cur = reslimit.rlim_max;
 376                 (void)setrlimit(RLIMIT_STACK , &reslimit);
 377         }
 378
 379         /*
 380          * not really needed, but doesn't hurt
 381          */
 382         if (getrlimit(RLIMIT_RSS , &reslimit) == 0){
 383                 reslimit.rlim_cur = reslimit.rlim_max;
 384                 (void)setrlimit(RLIMIT_RSS , &reslimit);
 385         }
 386
 387         /*
 388          * Handle posix locale
 389          *
 390          * set user defines time printing format for -v option
 391          */
 392         ltmfrmt = getenv("LC_TIME");
 393
 394         /*
 395          * signal handling to reset stored directory times and modes. Since
 396          * we deal with broken pipes via failed writes we ignore it. We also
 397          * deal with any file size limit through failed writes. Cpu time
 398          * limits are caught and a cleanup is forced.
 399          */
 400         if ((sigemptyset(&s_mask) < 0) || (sigaddset(&s_mask, SIGTERM) < 0) ||
 401             (sigaddset(&s_mask,SIGINT) < 0)||(sigaddset(&s_mask,SIGHUP) < 0) ||
 402             (sigaddset(&s_mask,SIGPIPE) < 0)||(sigaddset(&s_mask,SIGQUIT)<0) ||
 403             (sigaddset(&s_mask,SIGXCPU) < 0)||(sigaddset(&s_mask,SIGXFSZ)<0)) {
 404                 paxwarn(1, "Unable to set up signal mask");
 405                 return(-1);
 406         }
 407         memset(&n_hand, 0, sizeof n_hand);
 408         n_hand.sa_mask = s_mask;
 409         n_hand.sa_flags = 0;
 410         n_hand.sa_handler = sig_cleanup;
 411
 412         if ((sigaction(SIGHUP, &n_hand, &o_hand) < 0) &&
 413             (o_hand.sa_handler == SIG_IGN) &&
 414             (sigaction(SIGHUP, &o_hand, &o_hand) < 0))
 415                 goto out;
 416
 417         if ((sigaction(SIGTERM, &n_hand, &o_hand) < 0) &&
 418             (o_hand.sa_handler == SIG_IGN) &&
 419             (sigaction(SIGTERM, &o_hand, &o_hand) < 0))
 420                 goto out;
 421
 422         if ((sigaction(SIGINT, &n_hand, &o_hand) < 0) &&
 423             (o_hand.sa_handler == SIG_IGN) &&
 424             (sigaction(SIGINT, &o_hand, &o_hand) < 0))
 425                 goto out;
 426
 427         if ((sigaction(SIGQUIT, &n_hand, &o_hand) < 0) &&
 428             (o_hand.sa_handler == SIG_IGN) &&
 429             (sigaction(SIGQUIT, &o_hand, &o_hand) < 0))
 430                 goto out;
 431
 432         if ((sigaction(SIGXCPU, &n_hand, &o_hand) < 0) &&
 433             (o_hand.sa_handler == SIG_IGN) &&
 434             (sigaction(SIGXCPU, &o_hand, &o_hand) < 0))
 435                 goto out;
 436
 437         n_hand.sa_handler = SIG_IGN;
 438         if ((sigaction(SIGPIPE, &n_hand, &o_hand) < 0) ||
 439             (sigaction(SIGXFSZ, &n_hand, &o_hand) < 0))
 440                 goto out;
 441         return(0);
 442
 443     out:
 444         syswarn(1, errno, "Unable to set up signal handler");
 445         return(-1);
 446 }