pax.c revision 81601
1/*- 2 * Copyright (c) 1992 Keith Muller. 3 * Copyright (c) 1992, 1993 4 * The Regents of the University of California. All rights reserved. 5 * 6 * This code is derived from software contributed to Berkeley by 7 * Keith Muller of the University of California, San Diego. 8 * 9 * Redistribution and use in source and binary forms, with or without 10 * modification, are permitted provided that the following conditions 11 * are met: 12 * 1. Redistributions of source code must retain the above copyright 13 * notice, this list of conditions and the following disclaimer. 14 * 2. Redistributions in binary form must reproduce the above copyright 15 * notice, this list of conditions and the following disclaimer in the 16 * documentation and/or other materials provided with the distribution. 17 * 3. All advertising materials mentioning features or use of this software 18 * must display the following acknowledgement: 19 * This product includes software developed by the University of 20 * California, Berkeley and its contributors. 21 * 4. Neither the name of the University nor the names of its contributors 22 * may be used to endorse or promote products derived from this software 23 * without specific prior written permission. 24 * 25 * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND 26 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE 27 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE 28 * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE 29 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL 30 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS 31 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) 32 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT 33 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY 34 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF 35 * SUCH DAMAGE. 36 */ 37 38#ifndef lint 39static char const 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 46static char sccsid[] = "@(#)pax.c 8.2 (Berkeley) 4/18/94"; 47#endif 48static const char rcsid[] = 49 "$FreeBSD: head/bin/pax/pax.c 81601 2001-08-13 21:54:47Z peter $"; 50#endif /* not lint */ 51 52#include <sys/types.h> 53#include <sys/stat.h> 54#include <sys/time.h> 55#include <sys/resource.h> 56#include <err.h> 57#include <errno.h> 58#include <fcntl.h> 59#include <locale.h> 60#include <paths.h> 61#include <signal.h> 62#include <stdio.h> 63#include <stdlib.h> 64#include <string.h> 65#include <unistd.h> 66#include "pax.h" 67#include "extern.h" 68static 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 */ 77int act = DEFOP; /* read/write/append/copy */ 78FSUB *frmt = NULL; /* archive format type */ 79int cflag; /* match all EXCEPT pattern/file */ 80int cwdfd; /* starting cwd */ 81int dflag; /* directory member match only */ 82int iflag; /* interactive file/archive rename */ 83int kflag; /* do not overwrite existing files */ 84int lflag; /* use hard links when possible */ 85int nflag; /* select first archive member match */ 86int tflag; /* restore access time after read */ 87int uflag; /* ignore older modification time files */ 88int vflag; /* produce verbose output */ 89int Dflag; /* same as uflag except inode change time */ 90int Hflag; /* follow command line symlinks (write only) */ 91int Lflag; /* follow symlinks when writing */ 92int Xflag; /* archive files with same device id only */ 93int Yflag; /* same as Dflg except after name mode */ 94int Zflag; /* same as uflg except after name mode */ 95int vfpart; /* is partial verbose output in progress */ 96int patime = 1; /* preserve file access time */ 97int pmtime = 1; /* preserve file modification times */ 98int nodirs; /* do not create directories as needed */ 99int pmode; /* preserve file mode bits */ 100int pids; /* preserve file uid/gid */ 101int rmleadslash = 0; /* remove leading '/' from pathnames */ 102int exit_val; /* exit value */ 103int docrc; /* check/create file crc */ 104char *dirptr; /* destination dir in a copy */ 105char *argv0; /* root of argv[0] */ 106sigset_t s_mask; /* signal mask for cleanup critical sect */ 107FILE *listf; /* file pointer to print file list to */ 108char *tempfile; /* tempfile to use for mkstemp(3) */ 109char *tempbase; /* basename of tempfile to use for mkstemp(3) */ 110 111/* 112 * PAX - Portable Archive Interchange 113 * 114 * A utility to read, write, and write lists of the members of archive 115 * files and copy directory hierarchies. A variety of archive formats 116 * are supported (some are described in POSIX 1003.1 10.1): 117 * 118 * ustar - 10.1.1 extended tar interchange format 119 * cpio - 10.1.2 extended cpio interchange format 120 * tar - old BSD 4.3 tar format 121 * binary cpio - old cpio with binary header format 122 * sysVR4 cpio - with and without CRC 123 * 124 * This version is a superset of IEEE Std 1003.2b-d3 125 * 126 * Summary of Extensions to the IEEE Standard: 127 * 128 * 1 READ ENHANCEMENTS 129 * 1.1 Operations which read archives will continue to operate even when 130 * processing archives which may be damaged, truncated, or fail to meet 131 * format specs in several different ways. Damaged sections of archives 132 * are detected and avoided if possible. Attempts will be made to resync 133 * archive read operations even with badly damaged media. 134 * 1.2 Blocksize requirements are not strictly enforced on archive read. 135 * Tapes which have variable sized records can be read without errors. 136 * 1.3 The user can specify via the non-standard option flag -E if error 137 * resync operation should stop on a media error, try a specified number 138 * of times to correct, or try to correct forever. 139 * 1.4 Sparse files (lseek holes) stored on the archive (but stored with blocks 140 * of all zeros will be restored with holes appropriate for the target 141 * filesystem 142 * 1.5 The user is notified whenever something is found during archive 143 * read operations which violates spec (but the read will continue). 144 * 1.6 Multiple archive volumes can be read and may span over different 145 * archive devices 146 * 1.7 Rigidly restores all file attributes exactly as they are stored on the 147 * archive. 148 * 1.8 Modification change time ranges can be specified via multiple -T 149 * options. These allow a user to select files whose modification time 150 * lies within a specific time range. 151 * 1.9 Files can be selected based on owner (user name or uid) via one or more 152 * -U options. 153 * 1.10 Files can be selected based on group (group name or gid) via one o 154 * more -G options. 155 * 1.11 File modification time can be checked against existing file after 156 * name modification (-Z) 157 * 158 * 2 WRITE ENHANCEMENTS 159 * 2.1 Write operation will stop instead of allowing a user to create a flawed 160 * flawed archive (due to any problem). 161 * 2.2 Archives written by pax are forced to strictly conform to both the 162 * archive and pax the specific format specifications. 163 * 2.3 Blocking size and format is rigidly enforced on writes. 164 * 2.4 Formats which may exhibit header overflow problems (they have fields 165 * too small for large file systems, such as inode number storage), use 166 * routines designed to repair this problem. These techniques still 167 * conform to both pax and format specifications, but no longer truncate 168 * these fields. This removes any restrictions on using these archive 169 * formats on large file systems. 170 * 2.5 Multiple archive volumes can be written and may span over different 171 * archive devices 172 * 2.6 A archive volume record limit allows the user to specify the number 173 * of bytes stored on an archive volume. When reached the user is 174 * prompted for the next archive volume. This is specified with the 175 * non-standard -B flag. The limit is rounded up to the next blocksize. 176 * 2.7 All archive padding during write use zero filled sections. This makes 177 * it much easier to pull data out of flawed archive during read 178 * operations. 179 * 2.8 Access time reset with the -t applies to all file nodes (including 180 * directories). 181 * 2.9 Symbolic links can be followed with -L (optional in the spec). 182 * 2.10 Modification or inode change time ranges can be specified via 183 * multiple -T options. These allow a user to select files whose 184 * modification or inode change time lies within a specific time range. 185 * 2.11 Files can be selected based on owner (user name or uid) via one or more 186 * -U options. 187 * 2.12 Files can be selected based on group (group name or gid) via one o 188 * more -G options. 189 * 2.13 Symlinks which appear on the command line can be followed (without 190 * following other symlinks; -H flag) 191 * 192 * 3 COPY ENHANCEMENTS 193 * 3.1 Sparse files (lseek holes) can be copied without expanding the holes 194 * into zero filled blocks. The file copy is created with holes which are 195 * appropriate for the target filesystem 196 * 3.2 Access time as well as modification time on copied file trees can be 197 * preserved with the appropriate -p options. 198 * 3.3 Access time reset with the -t applies to all file nodes (including 199 * directories). 200 * 3.4 Symbolic links can be followed with -L (optional in the spec). 201 * 3.5 Modification or inode change time ranges can be specified via 202 * multiple -T options. These allow a user to select files whose 203 * modification or inode change time lies within a specific time range. 204 * 3.6 Files can be selected based on owner (user name or uid) via one or more 205 * -U options. 206 * 3.7 Files can be selected based on group (group name or gid) via one o 207 * more -G options. 208 * 3.8 Symlinks which appear on the command line can be followed (without 209 * following other symlinks; -H flag) 210 * 3.9 File inode change time can be checked against existing file before 211 * name modification (-D) 212 * 3.10 File inode change time can be checked against existing file after 213 * name modification (-Y) 214 * 3.11 File modification time can be checked against existing file after 215 * name modification (-Z) 216 * 217 * 4 GENERAL ENHANCEMENTS 218 * 4.1 Internal structure is designed to isolate format dependent and 219 * independent functions. Formats are selected via a format driver table. 220 * This encourages the addition of new archive formats by only having to 221 * write those routines which id, read and write the archive header. 222 */ 223 224/* 225 * main() 226 * parse options, set up and operate as specified by the user. 227 * any operational flaw will set exit_val to non-zero 228 * Return: 0 if ok, 1 otherwise 229 */ 230 231#ifdef __STDC__ 232int 233main(int argc, char **argv) 234#else 235int 236main(argc, argv) 237 int argc; 238 char **argv; 239#endif 240{ 241 char *tmpdir; 242 size_t tdlen; 243 244 (void) setlocale(LC_ALL, ""); 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(0, errno, "Can't open current working directory."); 252 return(exit_val); 253 } 254 255 /* 256 * Where should we put temporary files? 257 */ 258 if ((tmpdir = getenv("TMPDIR")) == NULL || *tmpdir == '\0') 259 tmpdir = _PATH_TMP; 260 tdlen = strlen(tmpdir); 261 while(tdlen > 0 && tmpdir[tdlen - 1] == '/') 262 tdlen--; 263 tempfile = malloc(tdlen + 1 + sizeof(_TFILE_BASE)); 264 if (tempfile == NULL) { 265 paxwarn(1, "Cannot allocate memory for temp file name."); 266 return(exit_val); 267 } 268 if (tdlen) 269 memcpy(tempfile, tmpdir, tdlen); 270 tempbase = tempfile + tdlen; 271 *tempbase++ = '/'; 272 273 /* 274 * parse options, determine operational mode, general init 275 */ 276 options(argc, argv); 277 if ((gen_init() < 0) || (tty_init() < 0)) 278 return(exit_val); 279 280 /* 281 * select a primary operation mode 282 */ 283 switch(act) { 284 case EXTRACT: 285 extract(); 286 break; 287 case ARCHIVE: 288 archive(); 289 break; 290 case APPND: 291 if (gzip_program != NULL) 292 err(1, "can not gzip while appending"); 293 append(); 294 break; 295 case COPY: 296 copy(); 297 break; 298 default: 299 case LIST: 300 list(); 301 break; 302 } 303 return(exit_val); 304} 305 306/* 307 * sig_cleanup() 308 * when interrupted we try to do whatever delayed processing we can. 309 * This is not critical, but we really ought to limit our damage when we 310 * are aborted by the user. 311 * Return: 312 * never.... 313 */ 314 315#ifdef __STDC__ 316void 317sig_cleanup(int which_sig) 318#else 319void 320sig_cleanup(which_sig) 321 int which_sig; 322#endif 323{ 324 /* 325 * restore modes and times for any dirs we may have created 326 * or any dirs we may have read. Set vflag and vfpart so the user 327 * will clearly see the message on a line by itself. 328 */ 329 vflag = vfpart = 1; 330 if (which_sig == SIGXCPU) 331 paxwarn(0, "Cpu time limit reached, cleaning up."); 332 else 333 paxwarn(0, "Signal caught, cleaning up."); 334 335 ar_close(); 336 proc_dir(); 337 if (tflag) 338 atdir_end(); 339 exit(1); 340} 341 342/* 343 * gen_init() 344 * general setup routines. Not all are required, but they really help 345 * when dealing with a medium to large sized archives. 346 */ 347 348#ifdef __STDC__ 349static int 350gen_init(void) 351#else 352static int 353gen_init() 354#endif 355{ 356 struct rlimit reslimit; 357 struct sigaction n_hand; 358 struct sigaction o_hand; 359 360 /* 361 * Really needed to handle large archives. We can run out of memory for 362 * internal tables really fast when we have a whole lot of files... 363 */ 364 if (getrlimit(RLIMIT_DATA , &reslimit) == 0){ 365 reslimit.rlim_cur = reslimit.rlim_max; 366 (void)setrlimit(RLIMIT_DATA , &reslimit); 367 } 368 369 /* 370 * should file size limits be waived? if the os limits us, this is 371 * needed if we want to write a large archive 372 */ 373 if (getrlimit(RLIMIT_FSIZE , &reslimit) == 0){ 374 reslimit.rlim_cur = reslimit.rlim_max; 375 (void)setrlimit(RLIMIT_FSIZE , &reslimit); 376 } 377 378 /* 379 * increase the size the stack can grow to 380 */ 381 if (getrlimit(RLIMIT_STACK , &reslimit) == 0){ 382 reslimit.rlim_cur = reslimit.rlim_max; 383 (void)setrlimit(RLIMIT_STACK , &reslimit); 384 } 385 386 /* 387 * not really needed, but doesn't hurt 388 */ 389 if (getrlimit(RLIMIT_RSS , &reslimit) == 0){ 390 reslimit.rlim_cur = reslimit.rlim_max; 391 (void)setrlimit(RLIMIT_RSS , &reslimit); 392 } 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 thorugh 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} 447