hush/archival/libunarchive/decompress_uncompress.c

308 lines
7.7 KiB
C
Raw Normal View History

/* vi: set sw=4 ts=4: */
/* uncompress for busybox -- (c) 2002 Robert Griebl
*
* based on the original compress42.c source
* (see disclaimer below)
*/
/* (N)compress42.c - File compression ala IEEE Computer, Mar 1992.
*
* Authors:
* Spencer W. Thomas (decvax!harpo!utah-cs!utah-gr!thomas)
* Jim McKie (decvax!mcvax!jim)
* Steve Davies (decvax!vax135!petsd!peora!srd)
* Ken Turkowski (decvax!decwrl!turtlevax!ken)
* James A. Woods (decvax!ihnp4!ames!jaw)
* Joe Orost (decvax!vax135!petsd!joe)
* Dave Mack (csu@alembic.acs.com)
* Peter Jannesen, Network Communication Systems
* (peter@ncs.nl)
*
* marc@suse.de : a small security fix for a buffer overflow
*
* [... History snipped ...]
*
*/
#include "libbb.h"
#include "unarchive.h"
/* Default input buffer size */
2002-11-28 09:09:47 +00:00
#define IBUFSIZ 2048
/* Default output buffer size */
#define OBUFSIZ 2048
/* Defines for third byte of header */
#define BIT_MASK 0x1f /* Mask for 'number of compresssion bits' */
/* Masks 0x20 and 0x40 are free. */
/* I think 0x20 should mean that there is */
/* a fourth header byte (for expansion). */
#define BLOCK_MODE 0x80 /* Block compression if table is full and */
/* compression rate is dropping flush tables */
/* the next two codes should not be changed lightly, as they must not */
/* lie within the contiguous general code space. */
#define FIRST 257 /* first free entry */
#define CLEAR 256 /* table clear output code */
2002-11-28 09:09:47 +00:00
#define INIT_BITS 9 /* initial number of bits/code */
/* machine variants which require cc -Dmachine: pdp11, z8000, DOS */
#define HBITS 17 /* 50% occupancy */
#define HSIZE (1<<HBITS)
#define HMASK (HSIZE-1) /* unused */
#define HPRIME 9941 /* unused */
#define BITS 16
#define BITS_STR "16"
#undef MAXSEG_64K /* unused */
#define MAXCODE(n) (1L << (n))
#define htabof(i) htab[i]
#define codetabof(i) codetab[i]
#define tab_prefixof(i) codetabof(i)
#define tab_suffixof(i) ((unsigned char *)(htab))[i]
#define de_stack ((unsigned char *)&(htab[HSIZE-1]))
#define clear_tab_prefixof() memset(codetab, 0, 256)
/*
* Decompress stdin to stdout. This routine adapts to the codes in the
* file building the "string" table on-the-fly; requiring no table to
* be stored in the compressed file.
*/
IF_DESKTOP(long long) int FAST_FUNC
unpack_Z_stream(int fd_in, int fd_out)
{
IF_DESKTOP(long long total_written = 0;)
IF_DESKTOP(long long) int retval = -1;
2002-11-28 09:09:47 +00:00
unsigned char *stackp;
2007-04-13 23:22:00 +00:00
long code;
2002-11-28 09:09:47 +00:00
int finchar;
2007-04-13 23:22:00 +00:00
long oldcode;
long incode;
2002-11-28 09:09:47 +00:00
int inbits;
int posbits;
int outpos;
int insize;
int bitmask;
2007-04-13 23:22:00 +00:00
long free_ent;
long maxcode;
long maxmaxcode;
2002-11-28 09:09:47 +00:00
int n_bits;
int rsize = 0;
unsigned char *inbuf; /* were eating insane amounts of stack - */
unsigned char *outbuf; /* bad for some embedded targets */
unsigned char *htab;
unsigned short *codetab;
/* Hmm, these were statics - why?! */
/* user settable max # bits/code */
int maxbits; /* = BITS; */
/* block compress mode -C compatible with 2.0 */
int block_mode; /* = BLOCK_MODE; */
inbuf = xzalloc(IBUFSIZ + 64);
outbuf = xzalloc(OBUFSIZ + 2048);
htab = xzalloc(HSIZE); /* wsn't zeroed out before, maybe can xmalloc? */
codetab = xzalloc(HSIZE * sizeof(codetab[0]));
insize = 0;
/* xread isn't good here, we have to return - caller may want
* to do some cleanup (e.g. delete incomplete unpacked file etc) */
if (full_read(fd_in, inbuf, 1) != 1) {
bb_error_msg("short read");
goto err;
}
maxbits = inbuf[0] & BIT_MASK;
block_mode = inbuf[0] & BLOCK_MODE;
maxmaxcode = MAXCODE(maxbits);
2002-11-28 09:09:47 +00:00
if (maxbits > BITS) {
bb_error_msg("compressed with %d bits, can only handle "
BITS_STR" bits", maxbits);
goto err;
}
n_bits = INIT_BITS;
maxcode = MAXCODE(INIT_BITS) - 1;
bitmask = (1 << INIT_BITS) - 1;
oldcode = -1;
finchar = 0;
outpos = 0;
2002-11-28 09:09:47 +00:00
posbits = 0 << 3;
2002-11-28 09:09:47 +00:00
free_ent = ((block_mode) ? FIRST : 256);
2002-11-28 09:09:47 +00:00
/* As above, initialize the first 256 entries in the table. */
/*clear_tab_prefixof(); - done by xzalloc */
2002-11-28 09:09:47 +00:00
for (code = 255; code >= 0; --code) {
tab_suffixof(code) = (unsigned char) code;
}
2002-11-28 09:09:47 +00:00
do {
resetbuf:
{
2002-11-28 09:09:47 +00:00
int i;
int e;
int o;
o = posbits >> 3;
e = insize - o;
2002-11-28 09:09:47 +00:00
for (i = 0; i < e; ++i)
inbuf[i] = inbuf[i + o];
insize = e;
posbits = 0;
}
if (insize < (int) (IBUFSIZ + 64) - IBUFSIZ) {
2003-11-21 22:17:28 +00:00
rsize = safe_read(fd_in, inbuf + insize, IBUFSIZ);
//error check??
insize += rsize;
}
2002-11-28 09:09:47 +00:00
inbits = ((rsize > 0) ? (insize - insize % n_bits) << 3 :
(insize << 3) - (n_bits - 1));
2002-11-28 09:09:47 +00:00
while (inbits > posbits) {
if (free_ent > maxcode) {
posbits =
((posbits - 1) +
((n_bits << 3) -
(posbits - 1 + (n_bits << 3)) % (n_bits << 3)));
++n_bits;
2002-11-28 09:09:47 +00:00
if (n_bits == maxbits) {
maxcode = maxmaxcode;
2002-11-28 09:09:47 +00:00
} else {
maxcode = MAXCODE(n_bits) - 1;
}
bitmask = (1 << n_bits) - 1;
goto resetbuf;
}
2002-11-28 09:09:47 +00:00
{
unsigned char *p = &inbuf[posbits >> 3];
code = ((((long) (p[0])) | ((long) (p[1]) << 8) |
((long) (p[2]) << 16)) >> (posbits & 0x7)) & bitmask;
}
posbits += n_bits;
2002-11-28 09:09:47 +00:00
if (oldcode == -1) {
oldcode = code;
finchar = (int) oldcode;
outbuf[outpos++] = (unsigned char) finchar;
continue;
}
2002-11-28 09:09:47 +00:00
if (code == CLEAR && block_mode) {
clear_tab_prefixof();
2002-11-28 09:09:47 +00:00
free_ent = FIRST - 1;
posbits =
((posbits - 1) +
((n_bits << 3) -
(posbits - 1 + (n_bits << 3)) % (n_bits << 3)));
n_bits = INIT_BITS;
maxcode = MAXCODE(INIT_BITS) - 1;
bitmask = (1 << INIT_BITS) - 1;
goto resetbuf;
}
incode = code;
2002-11-28 09:09:47 +00:00
stackp = de_stack;
2002-11-28 09:09:47 +00:00
/* Special case for KwKwK string. */
if (code >= free_ent) {
if (code > free_ent) {
unsigned char *p;
posbits -= n_bits;
2002-11-28 09:09:47 +00:00
p = &inbuf[posbits >> 3];
2003-03-19 09:13:01 +00:00
bb_error_msg
2002-11-28 09:09:47 +00:00
("insize:%d posbits:%d inbuf:%02X %02X %02X %02X %02X (%d)",
insize, posbits, p[-1], p[0], p[1], p[2], p[3],
(posbits & 07));
bb_error_msg("corrupted data");
goto err;
}
2002-11-28 09:09:47 +00:00
*--stackp = (unsigned char) finchar;
code = oldcode;
}
2002-11-28 09:09:47 +00:00
/* Generate output characters in reverse order */
2007-04-13 23:22:00 +00:00
while ((long) code >= (long) 256) {
2002-11-28 09:09:47 +00:00
*--stackp = tab_suffixof(code);
code = tab_prefixof(code);
}
finchar = tab_suffixof(code);
*--stackp = (unsigned char) finchar;
2002-11-28 09:09:47 +00:00
/* And put them out in forward order */
{
2002-11-28 09:09:47 +00:00
int i;
i = de_stack - stackp;
if (outpos + i >= OBUFSIZ) {
2002-11-28 09:09:47 +00:00
do {
if (i > OBUFSIZ - outpos) {
i = OBUFSIZ - outpos;
}
2002-11-28 09:09:47 +00:00
if (i > 0) {
memcpy(outbuf + outpos, stackp, i);
outpos += i;
}
2002-11-28 09:09:47 +00:00
if (outpos >= OBUFSIZ) {
full_write(fd_out, outbuf, outpos);
//error check??
IF_DESKTOP(total_written += outpos;)
outpos = 0;
}
2002-11-28 09:09:47 +00:00
stackp += i;
i = de_stack - stackp;
} while (i > 0);
2002-11-28 09:09:47 +00:00
} else {
memcpy(outbuf + outpos, stackp, i);
outpos += i;
}
}
2002-11-28 09:09:47 +00:00
/* Generate the new entry. */
code = free_ent;
if (code < maxmaxcode) {
2002-11-28 09:09:47 +00:00
tab_prefixof(code) = (unsigned short) oldcode;
tab_suffixof(code) = (unsigned char) finchar;
free_ent = code + 1;
}
2002-11-28 09:09:47 +00:00
/* Remember previous code. */
oldcode = incode;
}
2002-11-28 09:09:47 +00:00
} while (rsize > 0);
if (outpos > 0) {
full_write(fd_out, outbuf, outpos);
//error check??
IF_DESKTOP(total_written += outpos;)
}
retval = IF_DESKTOP(total_written) + 0;
err:
free(inbuf);
free(outbuf);
free(htab);
free(codetab);
return retval;
}