pax/pax.c

   1 /*      $OpenBSD: pax.c,v 1.11 1997/09/01 18:29:58 deraadt 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. All advertising materials mentioning features or use of this software
  21  *    must display the following acknowledgement:
  22  *      This product includes software developed by the University of
  23  *      California, Berkeley and its contributors.
  24  * 4. Neither the name of the University nor the names of its contributors
  25  *    may be used to endorse or promote products derived from this software
  26  *    without specific prior written permission.
  27  *
  28  * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
  29  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
  30  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
  31  * ARE DISCLAIMED.  IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
  32  * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
  33  * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
  34  * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
  35  * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
  36  * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
  37  * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
  38  * SUCH DAMAGE.
  39  */
  40
  41 #ifndef lint
  42 static char copyright[] __attribute__((__unused__)) =
  43 "@(#) Copyright (c) 1992, 1993\n\
  44         The Regents of the University of California.  All rights reserved.\n";
  45 #endif /* not lint */
  46
  47 #ifndef lint
  48 #if 0
  49 static char sccsid[] = "@(#)pax.c       8.2 (Berkeley) 4/18/94";
  50 #else
  51 static char rcsid[] __attribute__((__unused__)) = "$OpenBSD: pax.c,v 1.11 1997/09/01 18:29:58 deraadt Exp $";
  52 #endif
  53 #endif /* not lint */
  54
  55 #include <stdio.h>
  56 #include <sys/types.h>
  57 #include <sys/param.h>
  58 #include <sys/stat.h>
  59 #include <sys/time.h>
  60 #include <sys/resource.h>
  61 #include <signal.h>
  62 #include <unistd.h>
  63 #include <stdlib.h>
  64 #include <errno.h>
  65 #include <fcntl.h>
  66 #include "pax.h"
  67 #include "extern.h"
  68 static int gen_init __P((void));
  69
  70 /*
  71  * PAX main routines, general globals and some simple start up routines
  72  */
  73
  74 /*
  75  * Variables that can be accessed by any routine within pax
  76  */
  77 int     act = DEFOP;            /* read/write/append/copy */
  78 FSUB    *frmt = NULL;           /* archive format type */
  79 int     cflag;                  /* match all EXCEPT pattern/file */
  80 int     cwdfd;                  /* starting cwd */
  81 int     dflag;                  /* directory member match only  */
  82 int     iflag;                  /* interactive file/archive rename */
  83 int     kflag;                  /* do not overwrite existing files */
  84 int     lflag;                  /* use hard links when possible */
  85 int     nflag;                  /* select first archive member match */
  86 int     tflag;                  /* restore access time after read */
  87 int     uflag;                  /* ignore older modification time files */
  88 int     vflag;                  /* produce verbose output */
  89 int     zflag;                  /* use gzip */
  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 Dflg except after name mode */
  95 int     Zflag;                  /* same as uflg except after name mode */
  96 int     vfpart;                 /* is partial verbose output in progress */
  97 int     patime = 1;             /* preserve file access time */
  98 int     pmtime = 1;             /* preserve file modification times */
  99 int     nodirs;                 /* do not create directories as needed */
 100 int     pmode;                  /* preserve file mode bits */
 101 int     pids;                   /* preserve file uid/gid */
 102 int     rmleadslash = 0;        /* remove leading '/' from pathnames */
 103 int     exit_val;               /* exit value */
 104 int     docrc;                  /* check/create file crc */
 105 char    *dirptr;                /* destination dir in a copy */
 106 char    *ltmfrmt;               /* -v locale time format (if any) */
 107 char    *argv0;                 /* root of argv[0] */
 108 sigset_t s_mask;                /* signal mask for cleanup critical sect */
 109
 110 /*
 111  *      PAX - Portable Archive Interchange
 112  *
 113  *      A utility to read, write, and write lists of the members of archive
 114  *      files and copy directory hierarchies. A variety of archive formats
 115  *      are supported (some are described in POSIX 1003.1 10.1):
 116  *
 117  *              ustar - 10.1.1 extended tar interchange format
 118  *              cpio  - 10.1.2 extended cpio interchange format
 119  *              tar - old BSD 4.3 tar format
 120  *              binary cpio - old cpio with binary header format
 121  *              sysVR4 cpio -  with and without CRC
 122  *
 123  * This version is a superset of IEEE Std 1003.2b-d3
 124  *
 125  * Summary of Extensions to the IEEE Standard:
 126  *
 127  * 1    READ ENHANCEMENTS
 128  * 1.1  Operations which read archives will continue to operate even when
 129  *      processing archives which may be damaged, truncated, or fail to meet
 130  *      format specs in several different ways. Damaged sections of archives
 131  *      are detected and avoided if possible. Attempts will be made to resync
 132  *      archive read operations even with badly damaged media.
 133  * 1.2  Blocksize requirements are not strictly enforced on archive read.
 134  *      Tapes which have variable sized records can be read without errors.
 135  * 1.3  The user can specify via the non-standard option flag -E if error
 136  *      resync operation should stop on a media error, try a specified number
 137  *      of times to correct, or try to correct forever.
 138  * 1.4  Sparse files (lseek holes) stored on the archive (but stored with blocks
 139  *      of all zeros will be restored with holes appropriate for the target
 140  *      filesystem
 141  * 1.5  The user is notified whenever something is found during archive
 142  *      read operations which violates spec (but the read will continue).
 143  * 1.6  Multiple archive volumes can be read and may span over different
 144  *      archive devices
 145  * 1.7  Rigidly restores all file attributes exactly as they are stored on the
 146  *      archive.
 147  * 1.8  Modification change time ranges can be specified via multiple -T
 148  *      options. These allow a user to select files whose modification time
 149  *      lies within a specific time range.
 150  * 1.9  Files can be selected based on owner (user name or uid) via one or more
 151  *      -U options.
 152  * 1.10 Files can be selected based on group (group name or gid) via one o
 153  *      more -G options.
 154  * 1.11 File modification time can be checked against exisiting file after
 155  *      name modification (-Z)
 156  *
 157  * 2    WRITE ENHANCEMENTS
 158  * 2.1  Write operation will stop instead of allowing a user to create a flawed
 159  *      flawed archive (due to any problem).
 160  * 2.2  Archives writtens by pax are forced to strictly conform to both the
 161  *      archive and pax the spceific format specifications.
 162  * 2.3  Blocking size and format is rigidly enforced on writes.
 163  * 2.4  Formats which may exhibit header overflow problems (they have fields
 164  *      too small for large file systems, such as inode number storage), use
 165  *      routines designed to repair this problem. These techniques still
 166  *      conform to both pax and format specifications, but no longer truncate
 167  *      these fields. This removes any restrictions on using these archive
 168  *      formats on large file systems.
 169  * 2.5  Multiple archive volumes can be written and may span over different
 170  *      archive devices
 171  * 2.6  A archive volume record limit allows the user to specify the number
 172  *      of bytes stored on an archive volume. When reached the user is
 173  *      prompted for the next archive volume. This is specified with the
 174  *      non-standard -B flag. THe limit is rounded up to the next blocksize.
 175  * 2.7  All archive padding during write use zero filled sections. This makes
 176  *      it much easier to pull data out of flawed archive during read
 177  *      operations.
 178  * 2.8  Access time reset with the -t applies to all file nodes (including
 179  *      directories).
 180  * 2.9  Symbolic links can be followed with -L (optional in the spec).
 181  * 2.10 Modification or inode change time ranges can be specified via
 182  *      multiple -T options. These allow a user to select files whose
 183  *      modification or inode change time lies within a specific time range.
 184  * 2.11 Files can be selected based on owner (user name or uid) via one or more
 185  *      -U options.
 186  * 2.12 Files can be selected based on group (group name or gid) via one o
 187  *      more -G options.
 188  * 2.13 Symlinks which appear on the command line can be followed (without
 189  *      following other symlinks; -H flag)
 190  *
 191  * 3    COPY ENHANCEMENTS
 192  * 3.1  Sparse files (lseek holes) can be copied without expanding the holes
 193  *      into zero filled blocks. The file copy is created with holes which are
 194  *      appropriate for the target filesystem
 195  * 3.2  Access time as well as modification time on copied file trees can be
 196  *      preserved with the appropriate -p options.
 197  * 3.3  Access time reset with the -t applies to all file nodes (including
 198  *      directories).
 199  * 3.4  Symbolic links can be followed with -L (optional in the spec).
 200  * 3.5  Modification or inode change time ranges can be specified via
 201  *      multiple -T options. These allow a user to select files whose
 202  *      modification or inode change time lies within a specific time range.
 203  * 3.6  Files can be selected based on owner (user name or uid) via one or more
 204  *      -U options.
 205  * 3.7  Files can be selected based on group (group name or gid) via one o
 206  *      more -G options.
 207  * 3.8  Symlinks which appear on the command line can be followed (without
 208  *      following other symlinks; -H flag)
 209  * 3.9  File inode change time can be checked against exisiting file before
 210  *      name modification (-D)
 211  * 3.10 File inode change time can be checked against exisiting file after
 212  *      name modification (-Y)
 213  * 3.11 File modification time can be checked against exisiting file after
 214  *      name modification (-Z)
 215  *
 216  * 4    GENERAL ENHANCEMENTS
 217  * 4.1  Internal structure is designed to isolate format dependent and
 218  *      independent functions. Formats are selected via a format driver table.
 219  *      This encourages the addition of new archive formats by only having to
 220  *      write those routines which id, read and write the archive header.
 221  */
 222
 223 /*
 224  * main()
 225  *      parse options, set up and operate as specified by the user.
 226  *      any operational flaw will set exit_val to non-zero
 227  * Return: 0 if ok, 1 otherwise
 228  */
 229
 230 #ifdef __STDC__
 231 int
 232 main(int argc, char **argv)
 233 #else
 234 int
 235 main(argc, argv)
 236         int argc;
 237         char **argv;
 238 #endif
 239 {
 240         /*
 241          * Keep a reference to cwd, so we can always come back home.
 242          */
 243         cwdfd = open(".", O_RDONLY);
 244         if (cwdfd < 0) {
 245                 syswarn(0, errno, "Can't open current working directory.");
 246                 return(exit_val);
 247         }
 248
 249         /*
 250          * parse options, determine operational mode, general init
 251          */
 252         options(argc, argv);
 253         if ((gen_init() < 0) || (tty_init() < 0))
 254                 return(exit_val);
 255
 256         /*
 257          * select a primary operation mode
 258          */
 259         switch(act) {
 260         case EXTRACT:
 261                 extract();
 262                 break;
 263         case ARCHIVE:
 264                 archive();
 265                 break;
 266         case APPND:
 267                 append();
 268                 break;
 269         case COPY:
 270                 copy();
 271                 break;
 272         default:
 273         case LIST:
 274                 list();
 275                 break;
 276         }
 277         return(exit_val);
 278 }
 279
 280 /*
 281  * sig_cleanup()
 282  *      when interrupted we try to do whatever delayed processing we can.
 283  *      This is not critical, but we really ought to limit our damage when we
 284  *      are aborted by the user.
 285  * Return:
 286  *      never....
 287  */
 288
 289 #ifdef __STDC__
 290 void
 291 sig_cleanup(int which_sig)
 292 #else
 293 void
 294 sig_cleanup(which_sig)
 295         int which_sig;
 296 #endif
 297 {
 298         /*
 299          * restore modes and times for any dirs we may have created
 300          * or any dirs we may have read. Set vflag and vfpart so the user
 301          * will clearly see the message on a line by itself.
 302          */
 303         vflag = vfpart = 1;
 304         if (which_sig == SIGXCPU)
 305                 paxwarn(0, "Cpu time limit reached, cleaning up.");
 306         else
 307                 paxwarn(0, "Signal caught, cleaning up.");
 308
 309         ar_close();
 310         proc_dir();
 311         if (tflag)
 312                 atdir_end();
 313         exit(1);
 314 }
 315
 316 /*
 317  * gen_init()
 318  *      general setup routines. Not all are required, but they really help
 319  *      when dealing with a medium to large sized archives.
 320  */
 321
 322 #ifdef __STDC__
 323 static int
 324 gen_init(void)
 325 #else
 326 static int
 327 gen_init()
 328 #endif
 329 {
 330         struct rlimit reslimit;
 331         struct sigaction n_hand;
 332         struct sigaction o_hand;
 333
 334         /*
 335          * Really needed to handle large archives. We can run out of memory for
 336          * internal tables really fast when we have a whole lot of files...
 337          */
 338         if (getrlimit(RLIMIT_DATA , &reslimit) == 0){
 339                 reslimit.rlim_cur = reslimit.rlim_max;
 340                 (void)setrlimit(RLIMIT_DATA , &reslimit);
 341         }
 342
 343         /*
 344          * should file size limits be waived? if the os limits us, this is
 345          * needed if we want to write a large archive
 346          */
 347         if (getrlimit(RLIMIT_FSIZE , &reslimit) == 0){
 348                 reslimit.rlim_cur = reslimit.rlim_max;
 349                 (void)setrlimit(RLIMIT_FSIZE , &reslimit);
 350         }
 351
 352         /*
 353          * increase the size the stack can grow to
 354          */
 355         if (getrlimit(RLIMIT_STACK , &reslimit) == 0){
 356                 reslimit.rlim_cur = reslimit.rlim_max;
 357                 (void)setrlimit(RLIMIT_STACK , &reslimit);
 358         }
 359
 360         /*
 361          * not really needed, but doesn't hurt
 362          */
 363         if (getrlimit(RLIMIT_RSS , &reslimit) == 0){
 364                 reslimit.rlim_cur = reslimit.rlim_max;
 365                 (void)setrlimit(RLIMIT_RSS , &reslimit);
 366         }
 367
 368         /*
 369          * Handle posix locale
 370          *
 371          * set user defines time printing format for -v option
 372          */
 373         ltmfrmt = getenv("LC_TIME");
 374
 375         /*
 376          * signal handling to reset stored directory times and modes. Since
 377          * we deal with broken pipes via failed writes we ignore it. We also
 378          * deal with any file size limit thorugh failed writes. Cpu time
 379          * limits are caught and a cleanup is forced.
 380          */
 381         if ((sigemptyset(&s_mask) < 0) || (sigaddset(&s_mask, SIGTERM) < 0) ||
 382             (sigaddset(&s_mask,SIGINT) < 0)||(sigaddset(&s_mask,SIGHUP) < 0) ||
 383             (sigaddset(&s_mask,SIGPIPE) < 0)||(sigaddset(&s_mask,SIGQUIT)<0) ||
 384             (sigaddset(&s_mask,SIGXCPU) < 0)||(sigaddset(&s_mask,SIGXFSZ)<0)) {
 385                 paxwarn(1, "Unable to set up signal mask");
 386                 return(-1);
 387         }
 388         n_hand.sa_mask = s_mask;
 389         n_hand.sa_flags = 0;
 390         n_hand.sa_handler = sig_cleanup;
 391
 392         if ((sigaction(SIGHUP, &n_hand, &o_hand) < 0) &&
 393             (o_hand.sa_handler == SIG_IGN) &&
 394             (sigaction(SIGHUP, &o_hand, &o_hand) < 0))
 395                 goto out;
 396
 397         if ((sigaction(SIGTERM, &n_hand, &o_hand) < 0) &&
 398             (o_hand.sa_handler == SIG_IGN) &&
 399             (sigaction(SIGTERM, &o_hand, &o_hand) < 0))
 400                 goto out;
 401
 402         if ((sigaction(SIGINT, &n_hand, &o_hand) < 0) &&
 403             (o_hand.sa_handler == SIG_IGN) &&
 404             (sigaction(SIGINT, &o_hand, &o_hand) < 0))
 405                 goto out;
 406
 407         if ((sigaction(SIGQUIT, &n_hand, &o_hand) < 0) &&
 408             (o_hand.sa_handler == SIG_IGN) &&
 409             (sigaction(SIGQUIT, &o_hand, &o_hand) < 0))
 410                 goto out;
 411
 412         if ((sigaction(SIGXCPU, &n_hand, &o_hand) < 0) &&
 413             (o_hand.sa_handler == SIG_IGN) &&
 414             (sigaction(SIGXCPU, &o_hand, &o_hand) < 0))
 415                 goto out;
 416
 417         n_hand.sa_handler = SIG_IGN;
 418         if ((sigaction(SIGPIPE, &n_hand, &o_hand) < 0) ||
 419             (sigaction(SIGXFSZ, &n_hand, &o_hand) < 0))
 420                 goto out;
 421         return(0);
 422
 423     out:
 424         syswarn(1, errno, "Unable to set up signal handler");
 425         return(-1);
 426 }