pax.c revision 1.29
1 /* $NetBSD: pax.c,v 1.29 2003/08/07 09:05:22 agc Exp $ */ 2 3 /*- 4 * Copyright (c) 1992, 1993 5 * The Regents of the University of California. All rights reserved. 6 * 7 * This code is derived from software contributed to Berkeley by 8 * Keith Muller of the University of California, San Diego. 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 /*- 36 * Copyright (c) 1992 Keith Muller. 37 * 38 * This code is derived from software contributed to Berkeley by 39 * Keith Muller of the University of California, San Diego. 40 * 41 * Redistribution and use in source and binary forms, with or without 42 * modification, are permitted provided that the following conditions 43 * are met: 44 * 1. Redistributions of source code must retain the above copyright 45 * notice, this list of conditions and the following disclaimer. 46 * 2. Redistributions in binary form must reproduce the above copyright 47 * notice, this list of conditions and the following disclaimer in the 48 * documentation and/or other materials provided with the distribution. 49 * 3. All advertising materials mentioning features or use of this software 50 * must display the following acknowledgement: 51 * This product includes software developed by the University of 52 * California, Berkeley and its contributors. 53 * 4. Neither the name of the University nor the names of its contributors 54 * may be used to endorse or promote products derived from this software 55 * without specific prior written permission. 56 * 57 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 58 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 59 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 60 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 61 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 62 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 63 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 64 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 65 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 66 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 67 * SUCH DAMAGE. 68 */ 69 70 #include <sys/cdefs.h> 71 #if defined(__COPYRIGHT) && !defined(lint) 72 __COPYRIGHT("@(#) Copyright (c) 1992, 1993\n\ 73 The Regents of the University of California. All rights reserved.\n"); 74 #endif /* not lint */ 75 76 #if defined(__RCSID) && !defined(lint) 77 #if 0 78 static char sccsid[] = "@(#)pax.c 8.2 (Berkeley) 4/18/94"; 79 #else 80 __RCSID("$NetBSD: pax.c,v 1.29 2003/08/07 09:05:22 agc Exp $"); 81 #endif 82 #endif /* not lint */ 83 84 #include <sys/types.h> 85 #include <sys/param.h> 86 #include <sys/stat.h> 87 #include <sys/time.h> 88 #include <sys/resource.h> 89 #include <stdio.h> 90 #include <fcntl.h> 91 #include <signal.h> 92 #include <unistd.h> 93 #include <stdlib.h> 94 #include <string.h> 95 #include <errno.h> 96 #include <paths.h> 97 #include "pax.h" 98 #include "extern.h" 99 static int gen_init(void); 100 101 /* 102 * PAX main routines, general globals and some simple start up routines 103 */ 104 105 /* 106 * Variables that can be accessed by any routine within pax 107 */ 108 int act = ERROR; /* read/write/append/copy */ 109 FSUB *frmt = NULL; /* archive format type */ 110 int cflag; /* match all EXCEPT pattern/file */ 111 int cwdfd; /* starting cwd */ 112 int dflag; /* directory member match only */ 113 int iflag; /* interactive file/archive rename */ 114 int jflag; /* pass through bzip2 */ 115 int kflag; /* do not overwrite existing files */ 116 int lflag; /* use hard links when possible */ 117 int nflag; /* select first archive member match */ 118 int tflag; /* restore access time after read */ 119 int uflag; /* ignore older modification time files */ 120 int vflag; /* produce verbose output */ 121 int zflag; /* use gzip */ 122 int Aflag; /* honor absolute path */ 123 int Dflag; /* same as uflag except inode change time */ 124 int Hflag; /* follow command line symlinks (write only) */ 125 int Lflag; /* follow symlinks when writing */ 126 int Mflag; /* treat stdin as an mtree(8) specfile */ 127 int Xflag; /* archive files with same device id only */ 128 int Yflag; /* same as Dflg except after name mode */ 129 int Zflag; /* same as uflg except after name mode */ 130 int vfpart; /* is partial verbose output in progress */ 131 int patime = 1; /* preserve file access time */ 132 int pmtime = 1; /* preserve file modification times */ 133 int nodirs; /* do not create directories as needed */ 134 int pfflags = 1; /* preserve file flags */ 135 int pmode; /* preserve file mode bits */ 136 int pids; /* preserve file uid/gid */ 137 int rmleadslash = 0; /* remove leading '/' from pathnames */ 138 int exit_val; /* exit value */ 139 int docrc; /* check/create file crc */ 140 int to_stdout; /* extract to stdout */ 141 char *dirptr; /* destination dir in a copy */ 142 char *ltmfrmt; /* -v locale time format (if any) */ 143 char *argv0; /* root of argv[0] */ 144 sigset_t s_mask; /* signal mask for cleanup critical sect */ 145 FILE *listf; /* file pointer to print file list to */ 146 char *tempfile; /* tempfile to use for mkstemp(3) */ 147 char *tempbase; /* basename of tempfile to use for mkstemp(3) */ 148 int forcelocal; /* force local operation even if the name 149 * contains a : 150 */ 151 int secure = 1; /* don't extract names that contain .. */ 152 153 /* 154 * PAX - Portable Archive Interchange 155 * 156 * A utility to read, write, and write lists of the members of archive 157 * files and copy directory hierarchies. A variety of archive formats 158 * are supported (some are described in POSIX 1003.1 10.1): 159 * 160 * ustar - 10.1.1 extended tar interchange format 161 * cpio - 10.1.2 extended cpio interchange format 162 * tar - old BSD 4.3 tar format 163 * binary cpio - old cpio with binary header format 164 * sysVR4 cpio - with and without CRC 165 * 166 * This version is a superset of IEEE Std 1003.2b-d3 167 * 168 * Summary of Extensions to the IEEE Standard: 169 * 170 * 1 READ ENHANCEMENTS 171 * 1.1 Operations which read archives will continue to operate even when 172 * processing archives which may be damaged, truncated, or fail to meet 173 * format specs in several different ways. Damaged sections of archives 174 * are detected and avoided if possible. Attempts will be made to resync 175 * archive read operations even with badly damaged media. 176 * 1.2 Blocksize requirements are not strictly enforced on archive read. 177 * Tapes which have variable sized records can be read without errors. 178 * 1.3 The user can specify via the non-standard option flag -E if error 179 * resync operation should stop on a media error, try a specified number 180 * of times to correct, or try to correct forever. 181 * 1.4 Sparse files (lseek holes) stored on the archive (but stored with blocks 182 * of all zeros will be restored with holes appropriate for the target 183 * filesystem 184 * 1.5 The user is notified whenever something is found during archive 185 * read operations which violates spec (but the read will continue). 186 * 1.6 Multiple archive volumes can be read and may span over different 187 * archive devices 188 * 1.7 Rigidly restores all file attributes exactly as they are stored on the 189 * archive. 190 * 1.8 Modification change time ranges can be specified via multiple -T 191 * options. These allow a user to select files whose modification time 192 * lies within a specific time range. 193 * 1.9 Files can be selected based on owner (user name or uid) via one or more 194 * -U options. 195 * 1.10 Files can be selected based on group (group name or gid) via one o 196 * more -G options. 197 * 1.11 File modification time can be checked against existing file after 198 * name modification (-Z) 199 * 200 * 2 WRITE ENHANCEMENTS 201 * 2.1 Write operation will stop instead of allowing a user to create a flawed 202 * flawed archive (due to any problem). 203 * 2.2 Archives written by pax are forced to strictly conform to both the 204 * archive and pax the spceific format specifications. 205 * 2.3 Blocking size and format is rigidly enforced on writes. 206 * 2.4 Formats which may exhibit header overflow problems (they have fields 207 * too small for large file systems, such as inode number storage), use 208 * routines designed to repair this problem. These techniques still 209 * conform to both pax and format specifications, but no longer truncate 210 * these fields. This removes any restrictions on using these archive 211 * formats on large file systems. 212 * 2.5 Multiple archive volumes can be written and may span over different 213 * archive devices 214 * 2.6 A archive volume record limit allows the user to specify the number 215 * of bytes stored on an archive volume. When reached the user is 216 * prompted for the next archive volume. This is specified with the 217 * non-standard -B flag. The limit is rounded up to the next blocksize. 218 * 2.7 All archive padding during write use zero filled sections. This makes 219 * it much easier to pull data out of flawed archive during read 220 * operations. 221 * 2.8 Access time reset with the -t applies to all file nodes (including 222 * directories). 223 * 2.9 Symbolic links can be followed with -L (optional in the spec). 224 * 2.10 Modification or inode change time ranges can be specified via 225 * multiple -T options. These allow a user to select files whose 226 * modification or inode change time lies within a specific time range. 227 * 2.11 Files can be selected based on owner (user name or uid) via one or more 228 * -U options. 229 * 2.12 Files can be selected based on group (group name or gid) via one o 230 * more -G options. 231 * 2.13 Symlinks which appear on the command line can be followed (without 232 * following other symlinks; -H flag) 233 * 234 * 3 COPY ENHANCEMENTS 235 * 3.1 Sparse files (lseek holes) can be copied without expanding the holes 236 * into zero filled blocks. The file copy is created with holes which are 237 * appropriate for the target filesystem 238 * 3.2 Access time as well as modification time on copied file trees can be 239 * preserved with the appropriate -p options. 240 * 3.3 Access time reset with the -t applies to all file nodes (including 241 * directories). 242 * 3.4 Symbolic links can be followed with -L (optional in the spec). 243 * 3.5 Modification or inode change time ranges can be specified via 244 * multiple -T options. These allow a user to select files whose 245 * modification or inode change time lies within a specific time range. 246 * 3.6 Files can be selected based on owner (user name or uid) via one or more 247 * -U options. 248 * 3.7 Files can be selected based on group (group name or gid) via one o 249 * more -G options. 250 * 3.8 Symlinks which appear on the command line can be followed (without 251 * following other symlinks; -H flag) 252 * 3.9 File inode change time can be checked against existing file before 253 * name modification (-D) 254 * 3.10 File inode change time can be checked against existing file after 255 * name modification (-Y) 256 * 3.11 File modification time can be checked against existing file after 257 * name modification (-Z) 258 * 259 * 4 GENERAL ENHANCEMENTS 260 * 4.1 Internal structure is designed to isolate format dependent and 261 * independent functions. Formats are selected via a format driver table. 262 * This encourages the addition of new archive formats by only having to 263 * write those routines which id, read and write the archive header. 264 */ 265 266 /* 267 * main() 268 * parse options, set up and operate as specified by the user. 269 * any operational flaw will set exit_val to non-zero 270 * Return: 0 if ok, 1 otherwise 271 */ 272 273 int 274 main(int argc, char **argv) 275 { 276 char *tmpdir; 277 size_t tdlen; 278 279 setprogname(argv[0]); 280 281 listf = stderr; 282 283 /* 284 * Keep a reference to cwd, so we can always come back home. 285 */ 286 cwdfd = open(".", O_RDONLY); 287 if (cwdfd < 0) { 288 syswarn(0, errno, "Can't open current working directory."); 289 return(exit_val); 290 } 291 292 /* 293 * Where should we put temporary files? 294 */ 295 if ((tmpdir = getenv("TMPDIR")) == NULL || *tmpdir == '\0') 296 tmpdir = _PATH_TMP; 297 tdlen = strlen(tmpdir); 298 while(tdlen > 0 && tmpdir[tdlen - 1] == '/') 299 tdlen--; 300 tempfile = malloc(tdlen + 1 + sizeof(_TFILE_BASE)); 301 if (tempfile == NULL) { 302 tty_warn(1, "Cannot allocate memory for temp file name."); 303 return(exit_val); 304 } 305 if (tdlen) 306 memcpy(tempfile, tmpdir, tdlen); 307 tempbase = tempfile + tdlen; 308 *tempbase++ = '/'; 309 310 /* 311 * parse options, determine operational mode, general init 312 */ 313 options(argc, argv); 314 if ((gen_init() < 0) || (tty_init() < 0)) 315 return(exit_val); 316 317 (void)time(&starttime); 318 #ifdef SIGINFO 319 (void)signal(SIGINFO, ar_summary); 320 #endif 321 /* 322 * select a primary operation mode 323 */ 324 switch(act) { 325 case EXTRACT: 326 extract(); 327 break; 328 case ARCHIVE: 329 archive(); 330 break; 331 case APPND: 332 if (gzip_program != NULL) 333 err(1, "cannot gzip while appending"); 334 append(); 335 /* 336 * Check if we tried to append on an empty file and 337 * turned into ARCHIVE mode. 338 */ 339 if (act == -ARCHIVE) { 340 act = ARCHIVE; 341 archive(); 342 } 343 break; 344 case COPY: 345 copy(); 346 break; 347 default: 348 case LIST: 349 list(); 350 break; 351 } 352 return(exit_val); 353 } 354 355 /* 356 * sig_cleanup() 357 * when interrupted we try to do whatever delayed processing we can. 358 * This is not critical, but we really ought to limit our damage when we 359 * are aborted by the user. 360 * Return: 361 * never.... 362 */ 363 364 void 365 sig_cleanup(int which_sig) 366 { 367 /* 368 * restore modes and times for any dirs we may have created 369 * or any dirs we may have read. Set vflag and vfpart so the user 370 * will clearly see the message on a line by itself. 371 */ 372 vflag = vfpart = 1; 373 if (which_sig == SIGXCPU) 374 tty_warn(0, "Cpu time limit reached, cleaning up."); 375 else 376 tty_warn(0, "Signal caught, cleaning up."); 377 378 ar_close(); 379 proc_dir(); 380 if (tflag) 381 atdir_end(); 382 exit(1); 383 } 384 385 /* 386 * gen_init() 387 * general setup routines. Not all are required, but they really help 388 * when dealing with a medium to large sized archives. 389 */ 390 391 static int 392 gen_init(void) 393 { 394 struct rlimit reslimit; 395 struct sigaction n_hand; 396 struct sigaction o_hand; 397 398 /* 399 * Really needed to handle large archives. We can run out of memory for 400 * internal tables really fast when we have a whole lot of files... 401 */ 402 if (getrlimit(RLIMIT_DATA , &reslimit) == 0){ 403 reslimit.rlim_cur = reslimit.rlim_max; 404 (void)setrlimit(RLIMIT_DATA , &reslimit); 405 } 406 407 /* 408 * should file size limits be waived? if the os limits us, this is 409 * needed if we want to write a large archive 410 */ 411 if (getrlimit(RLIMIT_FSIZE , &reslimit) == 0){ 412 reslimit.rlim_cur = reslimit.rlim_max; 413 (void)setrlimit(RLIMIT_FSIZE , &reslimit); 414 } 415 416 /* 417 * increase the size the stack can grow to 418 */ 419 if (getrlimit(RLIMIT_STACK , &reslimit) == 0){ 420 reslimit.rlim_cur = reslimit.rlim_max; 421 (void)setrlimit(RLIMIT_STACK , &reslimit); 422 } 423 424 #ifdef RLIMIT_RSS 425 /* 426 * not really needed, but doesn't hurt 427 */ 428 if (getrlimit(RLIMIT_RSS , &reslimit) == 0){ 429 reslimit.rlim_cur = reslimit.rlim_max; 430 (void)setrlimit(RLIMIT_RSS , &reslimit); 431 } 432 #endif 433 434 /* 435 * Handle posix locale 436 * 437 * set user defines time printing format for -v option 438 */ 439 ltmfrmt = getenv("LC_TIME"); 440 441 /* 442 * signal handling to reset stored directory times and modes. Since 443 * we deal with broken pipes via failed writes we ignore it. We also 444 * deal with any file size limit thorugh failed writes. Cpu time 445 * limits are caught and a cleanup is forced. 446 */ 447 if ((sigemptyset(&s_mask) < 0) || (sigaddset(&s_mask, SIGTERM) < 0) || 448 (sigaddset(&s_mask,SIGINT) < 0)||(sigaddset(&s_mask,SIGHUP) < 0) || 449 (sigaddset(&s_mask,SIGPIPE) < 0)||(sigaddset(&s_mask,SIGQUIT)<0) || 450 (sigaddset(&s_mask,SIGXCPU) < 0)||(sigaddset(&s_mask,SIGXFSZ)<0)) { 451 tty_warn(1, "Unable to set up signal mask"); 452 return(-1); 453 } 454 memset(&n_hand, 0, sizeof n_hand); 455 n_hand.sa_mask = s_mask; 456 n_hand.sa_flags = 0; 457 n_hand.sa_handler = sig_cleanup; 458 459 if ((sigaction(SIGHUP, &n_hand, &o_hand) < 0) && 460 (o_hand.sa_handler == SIG_IGN) && 461 (sigaction(SIGHUP, &o_hand, &o_hand) < 0)) 462 goto out; 463 464 if ((sigaction(SIGTERM, &n_hand, &o_hand) < 0) && 465 (o_hand.sa_handler == SIG_IGN) && 466 (sigaction(SIGTERM, &o_hand, &o_hand) < 0)) 467 goto out; 468 469 if ((sigaction(SIGINT, &n_hand, &o_hand) < 0) && 470 (o_hand.sa_handler == SIG_IGN) && 471 (sigaction(SIGINT, &o_hand, &o_hand) < 0)) 472 goto out; 473 474 if ((sigaction(SIGQUIT, &n_hand, &o_hand) < 0) && 475 (o_hand.sa_handler == SIG_IGN) && 476 (sigaction(SIGQUIT, &o_hand, &o_hand) < 0)) 477 goto out; 478 479 if ((sigaction(SIGXCPU, &n_hand, &o_hand) < 0) && 480 (o_hand.sa_handler == SIG_IGN) && 481 (sigaction(SIGXCPU, &o_hand, &o_hand) < 0)) 482 goto out; 483 484 n_hand.sa_handler = SIG_IGN; 485 if ((sigaction(SIGPIPE, &n_hand, &o_hand) < 0) || 486 (sigaction(SIGXFSZ, &n_hand, &o_hand) < 0)) 487 goto out; 488 return(0); 489 490 out: 491 syswarn(1, errno, "Unable to set up signal handler"); 492 return(-1); 493 } 494
Indexes created Fri Sep 26 20:09:58 GMT 2025