Linux Audio

Check our new training course

Embedded Linux Audio

Check our new training course
with Creative Commons CC-BY-SA
lecture materials

Bootlin logo

Elixir Cross Referencer

Loading...
  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
/* vi: set sw=4 ts=4: */
/* Licensed under GPLv2 or later, see file LICENSE in this source tree.
 *
 * FIXME:
 *    In privileged mode if uname and gname map to a uid and gid then use the
 *    mapped value instead of the uid/gid values in tar header
 *
 * References:
 *    GNU tar and star man pages,
 *    Opengroup's ustar interchange format,
 *    http://www.opengroup.org/onlinepubs/007904975/utilities/pax.html
 */

#include "libbb.h"
#include "archive.h"

typedef uint32_t aliased_uint32_t FIX_ALIASING;
typedef off_t    aliased_off_t    FIX_ALIASING;


const char* FAST_FUNC strip_unsafe_prefix(const char *str)
{
	const char *cp = str;
	while (1) {
		char *cp2;
		if (*cp == '/') {
			cp++;
			continue;
		}
		if (strncmp(cp, "/../"+1, 3) == 0) {
			cp += 3;
			continue;
		}
		cp2 = strstr(cp, "/../");
		if (!cp2)
			break;
		cp = cp2 + 4;
	}
	if (cp != str) {
		static smallint warned = 0;
		if (!warned) {
			warned = 1;
			bb_error_msg("removing leading '%.*s' from member names",
				(int)(cp - str), str);
		}
	}
	return cp;
}

/* NB: _DESTROYS_ str[len] character! */
static unsigned long long getOctal(char *str, int len)
{
	unsigned long long v;
	char *end;
	/* NB: leading spaces are allowed. Using strtoull to handle that.
	 * The downside is that we accept e.g. "-123" too :(
	 */
	str[len] = '\0';
	v = strtoull(str, &end, 8);
	/* std: "Each numeric field is terminated by one or more
	 * <space> or NUL characters". We must support ' '! */
	if (*end != '\0' && *end != ' ') {
		int8_t first = str[0];
		if (!(first & 0x80))
			bb_error_msg_and_die("corrupted octal value in tar header");
		/*
		 * GNU tar uses "base-256 encoding" for very large numbers.
		 * Encoding is binary, with highest bit always set as a marker
		 * and sign in next-highest bit:
		 * 80 00 .. 00 - zero
		 * bf ff .. ff - largest positive number
		 * ff ff .. ff - minus 1
		 * c0 00 .. 00 - smallest negative number
		 *
		 * Example of tar file with 8914993153 (0x213600001) byte file.
		 * Field starts at offset 7c:
		 * 00070  30 30 30 00 30 30 30 30  30 30 30 00 80 00 00 00  |000.0000000.....|
		 * 00080  00 00 00 02 13 60 00 01  31 31 31 32 30 33 33 36  |.....`..11120336|
		 *
		 * NB: tarballs with NEGATIVE unix times encoded that way were seen!
		 */
		v = first;
		/* Sign-extend using 6th bit: */
		v <<= sizeof(unsigned long long)*8 - 7;
		v = (long long)v >> (sizeof(unsigned long long)*8 - 7);
		while (--len != 0)
			v = (v << 8) + (unsigned char) *str++;
	}
	return v;
}
#define GET_OCTAL(a) getOctal((a), sizeof(a))

#if ENABLE_FEATURE_TAR_SELINUX
/* Scan a PAX header for SELinux contexts, via "RHT.security.selinux" keyword.
 * This is what Red Hat's patched version of tar uses.
 */
# define SELINUX_CONTEXT_KEYWORD "RHT.security.selinux"
static char *get_selinux_sctx_from_pax_hdr(archive_handle_t *archive_handle, unsigned sz)
{
	char *buf, *p;
	char *result;

	p = buf = xmalloc(sz + 1);
	/* prevent bb_strtou from running off the buffer */
	buf[sz] = '\0';
	xread(archive_handle->src_fd, buf, sz);
	archive_handle->offset += sz;

	result = NULL;
	while (sz != 0) {
		char *end, *value;
		unsigned len;

		/* Every record has this format: "LEN NAME=VALUE\n" */
		len = bb_strtou(p, &end, 10);
		/* expect errno to be EINVAL, because the character
		 * following the digits should be a space
		 */
		p += len;
		sz -= len;
		if ((int)sz < 0
		 || len == 0
		 || errno != EINVAL
		 || *end != ' '
		) {
			bb_error_msg("malformed extended header, skipped");
			// More verbose version:
			//bb_error_msg("malformed extended header at %"OFF_FMT"d, skipped",
			//		archive_handle->offset - (sz + len));
			break;
		}
		/* overwrite the terminating newline with NUL
		 * (we do not bother to check that it *was* a newline)
		 */
		p[-1] = '\0';
		/* Is it selinux security context? */
		value = end + 1;
		if (strncmp(value, SELINUX_CONTEXT_KEYWORD"=", sizeof(SELINUX_CONTEXT_KEYWORD"=") - 1) == 0) {
			value += sizeof(SELINUX_CONTEXT_KEYWORD"=") - 1;
			result = xstrdup(value);
			break;
		}
	}

	free(buf);
	return result;
}
#endif

char FAST_FUNC get_header_tar(archive_handle_t *archive_handle)
{
	file_header_t *file_header = archive_handle->file_header;
	struct tar_header_t tar;
	char *cp;
	int i, sum_u, sum;
#if ENABLE_FEATURE_TAR_OLDSUN_COMPATIBILITY
	int sum_s;
#endif
	int parse_names;

	/* Our "private data" */
#if ENABLE_FEATURE_TAR_GNU_EXTENSIONS
# define p_longname (archive_handle->tar__longname)
# define p_linkname (archive_handle->tar__linkname)
#else
# define p_longname 0
# define p_linkname 0
#endif

#if ENABLE_FEATURE_TAR_GNU_EXTENSIONS || ENABLE_FEATURE_TAR_SELINUX
 again:
#endif
	/* Align header */
	data_align(archive_handle, 512);

 again_after_align:

#if ENABLE_DESKTOP || ENABLE_FEATURE_TAR_AUTODETECT
	/* to prevent misdetection of bz2 sig */
	*(aliased_uint32_t*)&tar = 0;
	i = full_read(archive_handle->src_fd, &tar, 512);
	/* If GNU tar sees EOF in above read, it says:
	 * "tar: A lone zero block at N", where N = kilobyte
	 * where EOF was met (not EOF block, actual EOF!),
	 * and exits with EXIT_SUCCESS.
	 * We will mimic exit(EXIT_SUCCESS), although we will not mimic
	 * the message and we don't check whether we indeed
	 * saw zero block directly before this. */
	if (i == 0) {
		xfunc_error_retval = 0;
 short_read:
		bb_error_msg_and_die("short read");
	}
	if (i != 512) {
		IF_FEATURE_TAR_AUTODETECT(goto autodetect;)
		goto short_read;
	}

#else
	i = 512;
	xread(archive_handle->src_fd, &tar, i);
#endif
	archive_handle->offset += i;

	/* If there is no filename its an empty header */
	if (tar.name[0] == 0 && tar.prefix[0] == 0) {
		if (archive_handle->tar__end) {
			/* Second consecutive empty header - end of archive.
			 * Read until the end to empty the pipe from gz or bz2
			 */
			while (full_read(archive_handle->src_fd, &tar, 512) == 512)
				continue;
			return EXIT_FAILURE;
		}
		archive_handle->tar__end = 1;
		return EXIT_SUCCESS;
	}
	archive_handle->tar__end = 0;

	/* Check header has valid magic, "ustar" is for the proper tar,
	 * five NULs are for the old tar format  */
	if (strncmp(tar.magic, "ustar", 5) != 0
	 && (!ENABLE_FEATURE_TAR_OLDGNU_COMPATIBILITY
	     || memcmp(tar.magic, "\0\0\0\0", 5) != 0)
	) {
#if ENABLE_FEATURE_TAR_AUTODETECT
		char FAST_FUNC (*get_header_ptr)(archive_handle_t *);
		uint16_t magic2;

 autodetect:
		magic2 = *(bb__aliased_uint16_t*)tar.name;
		/* tar gz/bz autodetect: check for gz/bz2 magic.
		 * If we see the magic, and it is the very first block,
		 * we can switch to get_header_tar_gz/bz2/lzma().
		 * Needs seekable fd. I wish recv(MSG_PEEK) works
		 * on any fd... */
# if ENABLE_FEATURE_SEAMLESS_GZ
		if (magic2 == GZIP_MAGIC) {
			get_header_ptr = get_header_tar_gz;
		} else
# endif
# if ENABLE_FEATURE_SEAMLESS_BZ2
		if (magic2 == BZIP2_MAGIC
		 && tar.name[2] == 'h' && isdigit(tar.name[3])
		) { /* bzip2 */
			get_header_ptr = get_header_tar_bz2;
		} else
# endif
# if ENABLE_FEATURE_SEAMLESS_XZ
		//TODO: if (magic2 == XZ_MAGIC1)...
		//else
# endif
			goto err;
		/* Two different causes for lseek() != 0:
		 * unseekable fd (would like to support that too, but...),
		 * or not first block (false positive, it's not .gz/.bz2!) */
		if (lseek(archive_handle->src_fd, -i, SEEK_CUR) != 0)
			goto err;
		while (get_header_ptr(archive_handle) == EXIT_SUCCESS)
			continue;
		return EXIT_FAILURE;
 err:
#endif /* FEATURE_TAR_AUTODETECT */
		bb_error_msg_and_die("invalid tar magic");
	}

	/* Do checksum on headers.
	 * POSIX says that checksum is done on unsigned bytes, but
	 * Sun and HP-UX gets it wrong... more details in
	 * GNU tar source. */
#if ENABLE_FEATURE_TAR_OLDSUN_COMPATIBILITY
	sum_s = ' ' * sizeof(tar.chksum);
#endif
	sum_u = ' ' * sizeof(tar.chksum);
	for (i = 0; i < 148; i++) {
		sum_u += ((unsigned char*)&tar)[i];
#if ENABLE_FEATURE_TAR_OLDSUN_COMPATIBILITY
		sum_s += ((signed char*)&tar)[i];
#endif
	}
	for (i = 156; i < 512; i++) {
		sum_u += ((unsigned char*)&tar)[i];
#if ENABLE_FEATURE_TAR_OLDSUN_COMPATIBILITY
		sum_s += ((signed char*)&tar)[i];
#endif
	}
	/* This field does not need special treatment (getOctal) */
	{
		char *endp; /* gcc likes temp var for &endp */
		sum = strtoul(tar.chksum, &endp, 8);
		if ((*endp != '\0' && *endp != ' ')
		 || (sum_u != sum IF_FEATURE_TAR_OLDSUN_COMPATIBILITY(&& sum_s != sum))
		) {
			bb_error_msg_and_die("invalid tar header checksum");
		}
	}
	/* don't use xstrtoul, tar.chksum may have leading spaces */
	sum = strtoul(tar.chksum, NULL, 8);
	if (sum_u != sum IF_FEATURE_TAR_OLDSUN_COMPATIBILITY(&& sum_s != sum)) {
		bb_error_msg_and_die("invalid tar header checksum");
	}

	/* 0 is reserved for high perf file, treat as normal file */
	if (!tar.typeflag) tar.typeflag = '0';
	parse_names = (tar.typeflag >= '0' && tar.typeflag <= '7');

	/* getOctal trashes subsequent field, therefore we call it
	 * on fields in reverse order */
	if (tar.devmajor[0]) {
		char t = tar.prefix[0];
		/* we trash prefix[0] here, but we DO need it later! */
		unsigned minor = GET_OCTAL(tar.devminor);
		unsigned major = GET_OCTAL(tar.devmajor);
		file_header->device = makedev(major, minor);
		tar.prefix[0] = t;
	}
	file_header->link_target = NULL;
	if (!p_linkname && parse_names && tar.linkname[0]) {
		file_header->link_target = xstrndup(tar.linkname, sizeof(tar.linkname));
		/* FIXME: what if we have non-link object with link_target? */
		/* Will link_target be free()ed? */
	}
#if ENABLE_FEATURE_TAR_UNAME_GNAME
	file_header->tar__uname = tar.uname[0] ? xstrndup(tar.uname, sizeof(tar.uname)) : NULL;
	file_header->tar__gname = tar.gname[0] ? xstrndup(tar.gname, sizeof(tar.gname)) : NULL;
#endif
	file_header->mtime = GET_OCTAL(tar.mtime);
	file_header->size = GET_OCTAL(tar.size);
	file_header->gid = GET_OCTAL(tar.gid);
	file_header->uid = GET_OCTAL(tar.uid);
	/* Set bits 0-11 of the files mode */
	file_header->mode = 07777 & GET_OCTAL(tar.mode);

	file_header->name = NULL;
	if (!p_longname && parse_names) {
		/* we trash mode[0] here, it's ok */
		//tar.name[sizeof(tar.name)] = '\0'; - gcc 4.3.0 would complain
		tar.mode[0] = '\0';
		if (tar.prefix[0]) {
			/* and padding[0] */
			//tar.prefix[sizeof(tar.prefix)] = '\0'; - gcc 4.3.0 would complain
			tar.padding[0] = '\0';
			file_header->name = concat_path_file(tar.prefix, tar.name);
		} else
			file_header->name = xstrdup(tar.name);
	}

	/* Set bits 12-15 of the files mode */
	/* (typeflag was not trashed because chksum does not use getOctal) */
	switch (tar.typeflag) {
	case '1': /* hardlink */
		/* we mark hardlinks as regular files with zero size and a link name */
		file_header->mode |= S_IFREG;
		/* on size of link fields from star(4)
		 * ... For tar archives written by pre POSIX.1-1988
		 * implementations, the size field usually contains the size of
		 * the file and needs to be ignored as no data may follow this
		 * header type.  For POSIX.1- 1988 compliant archives, the size
		 * field needs to be 0.  For POSIX.1-2001 compliant archives,
		 * the size field may be non zero, indicating that file data is
		 * included in the archive.
		 * i.e; always assume this is zero for safety.
		 */
		goto size0;
	case '7':
	/* case 0: */
	case '0':
#if ENABLE_FEATURE_TAR_OLDGNU_COMPATIBILITY
		if (last_char_is(file_header->name, '/')) {
			goto set_dir;
		}
#endif
		file_header->mode |= S_IFREG;
		break;
	case '2':
		file_header->mode |= S_IFLNK;
		/* have seen tarballs with size field containing
		 * the size of the link target's name */
 size0:
		file_header->size = 0;
		break;
	case '3':
		file_header->mode |= S_IFCHR;
		goto size0; /* paranoia */
	case '4':
		file_header->mode |= S_IFBLK;
		goto size0;
	case '5':
 IF_FEATURE_TAR_OLDGNU_COMPATIBILITY(set_dir:)
		file_header->mode |= S_IFDIR;
		goto size0;
	case '6':
		file_header->mode |= S_IFIFO;
		goto size0;
#if ENABLE_FEATURE_TAR_GNU_EXTENSIONS
	case 'L':
		/* free: paranoia: tar with several consecutive longnames */
		free(p_longname);
		/* For paranoia reasons we allocate extra NUL char */
		p_longname = xzalloc(file_header->size + 1);
		/* We read ASCIZ string, including NUL */
		xread(archive_handle->src_fd, p_longname, file_header->size);
		archive_handle->offset += file_header->size;
		/* return get_header_tar(archive_handle); */
		/* gcc 4.1.1 didn't optimize it into jump */
		/* so we will do it ourself, this also saves stack */
		goto again;
	case 'K':
		free(p_linkname);
		p_linkname = xzalloc(file_header->size + 1);
		xread(archive_handle->src_fd, p_linkname, file_header->size);
		archive_handle->offset += file_header->size;
		/* return get_header_tar(archive_handle); */
		goto again;
	case 'D':	/* GNU dump dir */
	case 'M':	/* Continuation of multi volume archive */
	case 'N':	/* Old GNU for names > 100 characters */
	case 'S':	/* Sparse file */
	case 'V':	/* Volume header */
#endif
#if !ENABLE_FEATURE_TAR_SELINUX
	case 'g':	/* pax global header */
	case 'x':	/* pax extended header */
#else
 skip_ext_hdr:
#endif
	{
		off_t sz;
		bb_error_msg("warning: skipping header '%c'", tar.typeflag);
		sz = (file_header->size + 511) & ~(off_t)511;
		archive_handle->offset += sz;
		sz >>= 9; /* sz /= 512 but w/o contortions for signed div */
		while (sz--)
			xread(archive_handle->src_fd, &tar, 512);
		/* return get_header_tar(archive_handle); */
		goto again_after_align;
	}
#if ENABLE_FEATURE_TAR_SELINUX
	case 'g':	/* pax global header */
	case 'x': {	/* pax extended header */
		char **pp;
		if ((uoff_t)file_header->size > 0xfffff) /* paranoia */
			goto skip_ext_hdr;
		pp = (tar.typeflag == 'g') ? &archive_handle->tar__global_sctx : &archive_handle->tar__next_file_sctx;
		free(*pp);
		*pp = get_selinux_sctx_from_pax_hdr(archive_handle, file_header->size);
		goto again;
	}
#endif
	default:
		bb_error_msg_and_die("unknown typeflag: 0x%x", tar.typeflag);
	}

#if ENABLE_FEATURE_TAR_GNU_EXTENSIONS
	if (p_longname) {
		file_header->name = p_longname;
		p_longname = NULL;
	}
	if (p_linkname) {
		file_header->link_target = p_linkname;
		p_linkname = NULL;
	}
#endif

	/* Everything up to and including last ".." component is stripped */
	overlapping_strcpy(file_header->name, strip_unsafe_prefix(file_header->name));

	/* Strip trailing '/' in directories */
	/* Must be done after mode is set as '/' is used to check if it's a directory */
	cp = last_char_is(file_header->name, '/');

	if (archive_handle->filter(archive_handle) == EXIT_SUCCESS) {
		archive_handle->action_header(/*archive_handle->*/ file_header);
		/* Note that we kill the '/' only after action_header() */
		/* (like GNU tar 1.15.1: verbose mode outputs "dir/dir/") */
		if (cp)
			*cp = '\0';
		archive_handle->action_data(archive_handle);
		if (archive_handle->accept || archive_handle->reject)
			llist_add_to(&archive_handle->passed, file_header->name);
		else /* Caller isn't interested in list of unpacked files */
			free(file_header->name);
	} else {
		data_skip(archive_handle);
		free(file_header->name);
	}
	archive_handle->offset += file_header->size;

	free(file_header->link_target);
	/* Do not free(file_header->name)!
	 * It might be inserted in archive_handle->passed - see above */
#if ENABLE_FEATURE_TAR_UNAME_GNAME
	free(file_header->tar__uname);
	free(file_header->tar__gname);
#endif
	return EXIT_SUCCESS;
}