2 * Copyright (c) 2003-2007 Tim Kientzle
5 * Redistribution and use in source and binary forms, with or without
6 * modification, are permitted provided that the following conditions
8 * 1. Redistributions of source code must retain the above copyright
9 * notice, this list of conditions and the following disclaimer.
10 * 2. Redistributions in binary form must reproduce the above copyright
11 * notice, this list of conditions and the following disclaimer in the
12 * documentation and/or other materials provided with the distribution.
14 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR(S) ``AS IS'' AND ANY EXPRESS OR
15 * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES
16 * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED.
17 * IN NO EVENT SHALL THE AUTHOR(S) BE LIABLE FOR ANY DIRECT, INDIRECT,
18 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
19 * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
20 * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
21 * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
22 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
23 * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
26 #include "archive_platform.h"
27 __FBSDID("$FreeBSD: src/lib/libarchive/archive_read_support_format_cpio.c,v 1.23 2007/04/13 16:07:25 kientzle Exp $");
32 /* #include <stdint.h> */ /* See archive_platform.h */
41 #include "archive_entry.h"
42 #include "archive_private.h"
43 #include "archive_read_private.h"
45 struct cpio_bin_header
{
46 unsigned char c_magic
[2];
47 unsigned char c_dev
[2];
48 unsigned char c_ino
[2];
49 unsigned char c_mode
[2];
50 unsigned char c_uid
[2];
51 unsigned char c_gid
[2];
52 unsigned char c_nlink
[2];
53 unsigned char c_rdev
[2];
54 unsigned char c_mtime
[4];
55 unsigned char c_namesize
[2];
56 unsigned char c_filesize
[4];
59 struct cpio_odc_header
{
73 struct cpio_newc_header
{
91 struct links_entry
*next
;
92 struct links_entry
*previous
;
99 #define CPIO_MAGIC 0x13141516
102 int (*read_header
)(struct archive_read
*, struct cpio
*,
103 struct archive_entry
*, size_t *, size_t *);
104 struct links_entry
*links_head
;
105 struct archive_string entry_name
;
106 struct archive_string entry_linkname
;
107 off_t entry_bytes_remaining
;
112 static int64_t atol16(const char *, unsigned);
113 static int64_t atol8(const char *, unsigned);
114 static int archive_read_format_cpio_bid(struct archive_read
*);
115 static int archive_read_format_cpio_cleanup(struct archive_read
*);
116 static int archive_read_format_cpio_read_data(struct archive_read
*,
117 const void **, size_t *, off_t
*);
118 static int archive_read_format_cpio_read_header(struct archive_read
*,
119 struct archive_entry
*);
120 static int be4(const unsigned char *);
121 static int header_bin_be(struct archive_read
*, struct cpio
*,
122 struct archive_entry
*, size_t *, size_t *);
123 static int header_bin_le(struct archive_read
*, struct cpio
*,
124 struct archive_entry
*, size_t *, size_t *);
125 static int header_newc(struct archive_read
*, struct cpio
*,
126 struct archive_entry
*, size_t *, size_t *);
127 static int header_odc(struct archive_read
*, struct cpio
*,
128 struct archive_entry
*, size_t *, size_t *);
129 static int le4(const unsigned char *);
130 static void record_hardlink(struct cpio
*cpio
, struct archive_entry
*entry
);
133 archive_read_support_format_cpio(struct archive
*_a
)
135 struct archive_read
*a
= (struct archive_read
*)_a
;
139 cpio
= (struct cpio
*)malloc(sizeof(*cpio
));
141 archive_set_error(&a
->archive
, ENOMEM
, "Can't allocate cpio data");
142 return (ARCHIVE_FATAL
);
144 memset(cpio
, 0, sizeof(*cpio
));
145 cpio
->magic
= CPIO_MAGIC
;
147 r
= __archive_read_register_format(a
,
149 archive_read_format_cpio_bid
,
150 archive_read_format_cpio_read_header
,
151 archive_read_format_cpio_read_data
,
153 archive_read_format_cpio_cleanup
);
162 archive_read_format_cpio_bid(struct archive_read
*a
)
166 const unsigned char *p
;
169 cpio
= (struct cpio
*)(a
->format
->data
);
171 bytes_read
= (a
->decompressor
->read_ahead
)(a
, &h
, 6);
172 /* Convert error code into error return. */
174 return ((int)bytes_read
);
178 p
= (const unsigned char *)h
;
179 if (memcmp(p
, "070707", 6) == 0) {
180 /* ASCII cpio archive (odc, POSIX.1) */
181 cpio
->read_header
= header_odc
;
184 * XXX TODO: More verification; Could check that only octal
185 * digits appear in appropriate header locations. XXX
187 } else if (memcmp(p
, "070701", 6) == 0) {
188 /* ASCII cpio archive (SVR4 without CRC) */
189 cpio
->read_header
= header_newc
;
192 * XXX TODO: More verification; Could check that only hex
193 * digits appear in appropriate header locations. XXX
195 } else if (memcmp(p
, "070702", 6) == 0) {
196 /* ASCII cpio archive (SVR4 with CRC) */
197 /* XXX TODO: Flag that we should check the CRC. XXX */
198 cpio
->read_header
= header_newc
;
201 * XXX TODO: More verification; Could check that only hex
202 * digits appear in appropriate header locations. XXX
204 } else if (p
[0] * 256 + p
[1] == 070707) {
205 /* big-endian binary cpio archives */
206 cpio
->read_header
= header_bin_be
;
208 /* Is more verification possible here? */
209 } else if (p
[0] + p
[1] * 256 == 070707) {
210 /* little-endian binary cpio archives */
211 cpio
->read_header
= header_bin_le
;
213 /* Is more verification possible here? */
215 return (ARCHIVE_WARN
);
221 archive_read_format_cpio_read_header(struct archive_read
*a
,
222 struct archive_entry
*entry
)
231 cpio
= (struct cpio
*)(a
->format
->data
);
232 r
= (cpio
->read_header(a
, cpio
, entry
, &namelength
, &name_pad
));
237 /* Read name from buffer. */
238 bytes
= (a
->decompressor
->read_ahead
)(a
, &h
, namelength
+ name_pad
);
239 if (bytes
< namelength
+ name_pad
)
240 return (ARCHIVE_FATAL
);
241 (a
->decompressor
->consume
)(a
, namelength
+ name_pad
);
242 archive_strncpy(&cpio
->entry_name
, (const char *)h
, namelength
);
243 archive_entry_set_pathname(entry
, cpio
->entry_name
.s
);
244 cpio
->entry_offset
= 0;
246 /* If this is a symlink, read the link contents. */
247 if (archive_entry_filetype(entry
) == AE_IFLNK
) {
248 bytes
= (a
->decompressor
->read_ahead
)(a
, &h
,
249 cpio
->entry_bytes_remaining
);
250 if ((off_t
)bytes
< cpio
->entry_bytes_remaining
)
251 return (ARCHIVE_FATAL
);
252 (a
->decompressor
->consume
)(a
, cpio
->entry_bytes_remaining
);
253 archive_strncpy(&cpio
->entry_linkname
, (const char *)h
,
254 cpio
->entry_bytes_remaining
);
255 archive_entry_set_symlink(entry
, cpio
->entry_linkname
.s
);
256 cpio
->entry_bytes_remaining
= 0;
259 /* Compare name to "TRAILER!!!" to test for end-of-archive. */
260 if (namelength
== 11 && strcmp((const char *)h
, "TRAILER!!!") == 0) {
261 /* TODO: Store file location of start of block. */
262 archive_set_error(&a
->archive
, 0, NULL
);
263 return (ARCHIVE_EOF
);
266 /* Detect and record hardlinks to previously-extracted entries. */
267 record_hardlink(cpio
, entry
);
273 archive_read_format_cpio_read_data(struct archive_read
*a
,
274 const void **buff
, size_t *size
, off_t
*offset
)
279 cpio
= (struct cpio
*)(a
->format
->data
);
280 if (cpio
->entry_bytes_remaining
> 0) {
281 bytes_read
= (a
->decompressor
->read_ahead
)(a
, buff
, 1);
283 return (ARCHIVE_FATAL
);
284 if (bytes_read
> cpio
->entry_bytes_remaining
)
285 bytes_read
= cpio
->entry_bytes_remaining
;
287 *offset
= cpio
->entry_offset
;
288 cpio
->entry_offset
+= bytes_read
;
289 cpio
->entry_bytes_remaining
-= bytes_read
;
290 (a
->decompressor
->consume
)(a
, bytes_read
);
293 while (cpio
->entry_padding
> 0) {
294 bytes_read
= (a
->decompressor
->read_ahead
)(a
, buff
, 1);
296 return (ARCHIVE_FATAL
);
297 if (bytes_read
> cpio
->entry_padding
)
298 bytes_read
= cpio
->entry_padding
;
299 (a
->decompressor
->consume
)(a
, bytes_read
);
300 cpio
->entry_padding
-= bytes_read
;
304 *offset
= cpio
->entry_offset
;
305 return (ARCHIVE_EOF
);
310 header_newc(struct archive_read
*a
, struct cpio
*cpio
,
311 struct archive_entry
*entry
, size_t *namelength
, size_t *name_pad
)
314 const struct cpio_newc_header
*header
;
317 /* Read fixed-size portion of header. */
318 bytes
= (a
->decompressor
->read_ahead
)(a
, &h
, sizeof(struct cpio_newc_header
));
319 if (bytes
< sizeof(struct cpio_newc_header
))
320 return (ARCHIVE_FATAL
);
321 (a
->decompressor
->consume
)(a
, sizeof(struct cpio_newc_header
));
323 /* Parse out hex fields. */
324 header
= (const struct cpio_newc_header
*)h
;
326 if (memcmp(header
->c_magic
, "070701", 6) == 0) {
327 a
->archive
.archive_format
= ARCHIVE_FORMAT_CPIO_SVR4_NOCRC
;
328 a
->archive
.archive_format_name
= "ASCII cpio (SVR4 with no CRC)";
329 } else if (memcmp(header
->c_magic
, "070702", 6) == 0) {
330 a
->archive
.archive_format
= ARCHIVE_FORMAT_CPIO_SVR4_CRC
;
331 a
->archive
.archive_format_name
= "ASCII cpio (SVR4 with CRC)";
333 /* TODO: Abort here? */
336 archive_entry_set_devmajor(entry
, atol16(header
->c_devmajor
, sizeof(header
->c_devmajor
)));
337 archive_entry_set_devminor(entry
, atol16(header
->c_devminor
, sizeof(header
->c_devminor
)));
338 archive_entry_set_ino(entry
, atol16(header
->c_ino
, sizeof(header
->c_ino
)));
339 archive_entry_set_mode(entry
, atol16(header
->c_mode
, sizeof(header
->c_mode
)));
340 archive_entry_set_uid(entry
, atol16(header
->c_uid
, sizeof(header
->c_uid
)));
341 archive_entry_set_gid(entry
, atol16(header
->c_gid
, sizeof(header
->c_gid
)));
342 archive_entry_set_nlink(entry
, atol16(header
->c_nlink
, sizeof(header
->c_nlink
)));
343 archive_entry_set_rdevmajor(entry
, atol16(header
->c_rdevmajor
, sizeof(header
->c_rdevmajor
)));
344 archive_entry_set_rdevminor(entry
, atol16(header
->c_rdevminor
, sizeof(header
->c_rdevminor
)));
345 archive_entry_set_mtime(entry
, atol16(header
->c_mtime
, sizeof(header
->c_mtime
)), 0);
346 *namelength
= atol16(header
->c_namesize
, sizeof(header
->c_namesize
));
347 /* Pad name to 2 more than a multiple of 4. */
348 *name_pad
= (2 - *namelength
) & 3;
351 * Note: entry_bytes_remaining is at least 64 bits and
352 * therefore guaranteed to be big enough for a 33-bit file
355 cpio
->entry_bytes_remaining
=
356 atol16(header
->c_filesize
, sizeof(header
->c_filesize
));
357 archive_entry_set_size(entry
, cpio
->entry_bytes_remaining
);
358 /* Pad file contents to a multiple of 4. */
359 cpio
->entry_padding
= 3 & -cpio
->entry_bytes_remaining
;
364 header_odc(struct archive_read
*a
, struct cpio
*cpio
,
365 struct archive_entry
*entry
, size_t *namelength
, size_t *name_pad
)
368 const struct cpio_odc_header
*header
;
371 a
->archive
.archive_format
= ARCHIVE_FORMAT_CPIO_POSIX
;
372 a
->archive
.archive_format_name
= "POSIX octet-oriented cpio";
374 /* Read fixed-size portion of header. */
375 bytes
= (a
->decompressor
->read_ahead
)(a
, &h
, sizeof(struct cpio_odc_header
));
376 if (bytes
< sizeof(struct cpio_odc_header
))
377 return (ARCHIVE_FATAL
);
378 (a
->decompressor
->consume
)(a
, sizeof(struct cpio_odc_header
));
380 /* Parse out octal fields. */
381 header
= (const struct cpio_odc_header
*)h
;
383 archive_entry_set_dev(entry
, atol8(header
->c_dev
, sizeof(header
->c_dev
)));
384 archive_entry_set_ino(entry
, atol8(header
->c_ino
, sizeof(header
->c_ino
)));
385 archive_entry_set_mode(entry
, atol8(header
->c_mode
, sizeof(header
->c_mode
)));
386 archive_entry_set_uid(entry
, atol8(header
->c_uid
, sizeof(header
->c_uid
)));
387 archive_entry_set_gid(entry
, atol8(header
->c_gid
, sizeof(header
->c_gid
)));
388 archive_entry_set_nlink(entry
, atol8(header
->c_nlink
, sizeof(header
->c_nlink
)));
389 archive_entry_set_rdev(entry
, atol8(header
->c_rdev
, sizeof(header
->c_rdev
)));
390 archive_entry_set_mtime(entry
, atol8(header
->c_mtime
, sizeof(header
->c_mtime
)), 0);
391 *namelength
= atol8(header
->c_namesize
, sizeof(header
->c_namesize
));
392 *name_pad
= 0; /* No padding of filename. */
395 * Note: entry_bytes_remaining is at least 64 bits and
396 * therefore guaranteed to be big enough for a 33-bit file
399 cpio
->entry_bytes_remaining
=
400 atol8(header
->c_filesize
, sizeof(header
->c_filesize
));
401 archive_entry_set_size(entry
, cpio
->entry_bytes_remaining
);
402 cpio
->entry_padding
= 0;
407 header_bin_le(struct archive_read
*a
, struct cpio
*cpio
,
408 struct archive_entry
*entry
, size_t *namelength
, size_t *name_pad
)
411 const struct cpio_bin_header
*header
;
414 a
->archive
.archive_format
= ARCHIVE_FORMAT_CPIO_BIN_LE
;
415 a
->archive
.archive_format_name
= "cpio (little-endian binary)";
417 /* Read fixed-size portion of header. */
418 bytes
= (a
->decompressor
->read_ahead
)(a
, &h
, sizeof(struct cpio_bin_header
));
419 if (bytes
< sizeof(struct cpio_bin_header
))
420 return (ARCHIVE_FATAL
);
421 (a
->decompressor
->consume
)(a
, sizeof(struct cpio_bin_header
));
423 /* Parse out binary fields. */
424 header
= (const struct cpio_bin_header
*)h
;
426 archive_entry_set_dev(entry
, header
->c_dev
[0] + header
->c_dev
[1] * 256);
427 archive_entry_set_ino(entry
, header
->c_ino
[0] + header
->c_ino
[1] * 256);
428 archive_entry_set_mode(entry
, header
->c_mode
[0] + header
->c_mode
[1] * 256);
429 archive_entry_set_uid(entry
, header
->c_uid
[0] + header
->c_uid
[1] * 256);
430 archive_entry_set_gid(entry
, header
->c_gid
[0] + header
->c_gid
[1] * 256);
431 archive_entry_set_nlink(entry
, header
->c_nlink
[0] + header
->c_nlink
[1] * 256);
432 archive_entry_set_rdev(entry
, header
->c_rdev
[0] + header
->c_rdev
[1] * 256);
433 archive_entry_set_mtime(entry
, le4(header
->c_mtime
), 0);
434 *namelength
= header
->c_namesize
[0] + header
->c_namesize
[1] * 256;
435 *name_pad
= *namelength
& 1; /* Pad to even. */
437 cpio
->entry_bytes_remaining
= le4(header
->c_filesize
);
438 archive_entry_set_size(entry
, cpio
->entry_bytes_remaining
);
439 cpio
->entry_padding
= cpio
->entry_bytes_remaining
& 1; /* Pad to even. */
444 header_bin_be(struct archive_read
*a
, struct cpio
*cpio
,
445 struct archive_entry
*entry
, size_t *namelength
, size_t *name_pad
)
448 const struct cpio_bin_header
*header
;
451 a
->archive
.archive_format
= ARCHIVE_FORMAT_CPIO_BIN_BE
;
452 a
->archive
.archive_format_name
= "cpio (big-endian binary)";
454 /* Read fixed-size portion of header. */
455 bytes
= (a
->decompressor
->read_ahead
)(a
, &h
,
456 sizeof(struct cpio_bin_header
));
457 if (bytes
< sizeof(struct cpio_bin_header
))
458 return (ARCHIVE_FATAL
);
459 (a
->decompressor
->consume
)(a
, sizeof(struct cpio_bin_header
));
461 /* Parse out binary fields. */
462 header
= (const struct cpio_bin_header
*)h
;
463 archive_entry_set_dev(entry
, header
->c_dev
[0] * 256 + header
->c_dev
[1]);
464 archive_entry_set_ino(entry
, header
->c_ino
[0] * 256 + header
->c_ino
[1]);
465 archive_entry_set_mode(entry
, header
->c_mode
[0] * 256 + header
->c_mode
[1]);
466 archive_entry_set_uid(entry
, header
->c_uid
[0] * 256 + header
->c_uid
[1]);
467 archive_entry_set_gid(entry
, header
->c_gid
[0] * 256 + header
->c_gid
[1]);
468 archive_entry_set_nlink(entry
, header
->c_nlink
[0] * 256 + header
->c_nlink
[1]);
469 archive_entry_set_rdev(entry
, header
->c_rdev
[0] * 256 + header
->c_rdev
[1]);
470 archive_entry_set_mtime(entry
, be4(header
->c_mtime
), 0);
471 *namelength
= header
->c_namesize
[0] * 256 + header
->c_namesize
[1];
472 *name_pad
= *namelength
& 1; /* Pad to even. */
474 cpio
->entry_bytes_remaining
= be4(header
->c_filesize
);
475 archive_entry_set_size(entry
, cpio
->entry_bytes_remaining
);
476 cpio
->entry_padding
= cpio
->entry_bytes_remaining
& 1; /* Pad to even. */
481 archive_read_format_cpio_cleanup(struct archive_read
*a
)
485 cpio
= (struct cpio
*)(a
->format
->data
);
486 /* Free inode->name map */
487 while (cpio
->links_head
!= NULL
) {
488 struct links_entry
*lp
= cpio
->links_head
->next
;
490 if (cpio
->links_head
->name
)
491 free(cpio
->links_head
->name
);
492 free(cpio
->links_head
);
493 cpio
->links_head
= lp
;
495 archive_string_free(&cpio
->entry_name
);
497 (a
->format
->data
) = NULL
;
502 le4(const unsigned char *p
)
504 return ((p
[0]<<16) + (p
[1]<<24) + (p
[2]<<0) + (p
[3]<<8));
509 be4(const unsigned char *p
)
511 return (p
[0] + (p
[1]<<8) + (p
[2]<<16) + (p
[3]<<24));
515 * Note that this implementation does not (and should not!) obey
516 * locale settings; you cannot simply substitute strtol here, since
517 * it does obey locale.
520 atol8(const char *p
, unsigned char_cnt
)
526 while (char_cnt
-- > 0) {
527 if (*p
>= '0' && *p
<= '7')
539 atol16(const char *p
, unsigned char_cnt
)
545 while (char_cnt
-- > 0) {
546 if (*p
>= 'a' && *p
<= 'f')
547 digit
= *p
- 'a' + 10;
548 else if (*p
>= 'A' && *p
<= 'F')
549 digit
= *p
- 'A' + 10;
550 else if (*p
>= '0' && *p
<= '9')
562 record_hardlink(struct cpio
*cpio
, struct archive_entry
*entry
)
564 struct links_entry
*le
;
568 dev
= archive_entry_dev(entry
);
569 ino
= archive_entry_ino(entry
);
572 * First look in the list of multiply-linked files. If we've
573 * already dumped it, convert this entry to a hard link entry.
575 for (le
= cpio
->links_head
; le
; le
= le
->next
) {
576 if (le
->dev
== dev
&& le
->ino
== ino
) {
577 archive_entry_set_hardlink(entry
, le
->name
);
579 if (--le
->links
<= 0) {
580 if (le
->previous
!= NULL
)
581 le
->previous
->next
= le
->next
;
582 if (le
->next
!= NULL
)
583 le
->next
->previous
= le
->previous
;
584 if (cpio
->links_head
== le
)
585 cpio
->links_head
= le
->next
;
593 le
= (struct links_entry
*)malloc(sizeof(struct links_entry
));
595 __archive_errx(1, "Out of memory adding file to list");
596 if (cpio
->links_head
!= NULL
)
597 cpio
->links_head
->previous
= le
;
598 le
->next
= cpio
->links_head
;
600 cpio
->links_head
= le
;
603 le
->links
= archive_entry_nlink(entry
) - 1;
604 le
->name
= strdup(archive_entry_pathname(entry
));
605 if (le
->name
== NULL
)
606 __archive_errx(1, "Out of memory adding file to list");