[PATCH] fix saa7146 compilation
[linux-2.6/history.git] / lib / inflate.c
blob6a9cac881978b813fe2230d0c75e7c565a405b7d
1 #define DEBG(x)
2 #define DEBG1(x)
3 /* inflate.c -- Not copyrighted 1992 by Mark Adler
4 version c10p1, 10 January 1993 */
6 /*
7 * Adapted for booting Linux by Hannu Savolainen 1993
8 * based on gzip-1.0.3
10 * Nicolas Pitre <nico@cam.org>, 1999/04/14 :
11 * Little mods for all variable to reside either into rodata or bss segments
12 * by marking constant variables with 'const' and initializing all the others
13 * at run-time only. This allows for the kernel uncompressor to run
14 * directly from Flash or ROM memory on embedded systems.
18 Inflate deflated (PKZIP's method 8 compressed) data. The compression
19 method searches for as much of the current string of bytes (up to a
20 length of 258) in the previous 32 K bytes. If it doesn't find any
21 matches (of at least length 3), it codes the next byte. Otherwise, it
22 codes the length of the matched string and its distance backwards from
23 the current position. There is a single Huffman code that codes both
24 single bytes (called "literals") and match lengths. A second Huffman
25 code codes the distance information, which follows a length code. Each
26 length or distance code actually represents a base value and a number
27 of "extra" (sometimes zero) bits to get to add to the base value. At
28 the end of each deflated block is a special end-of-block (EOB) literal/
29 length code. The decoding process is basically: get a literal/length
30 code; if EOB then done; if a literal, emit the decoded byte; if a
31 length then get the distance and emit the referred-to bytes from the
32 sliding window of previously emitted data.
34 There are (currently) three kinds of inflate blocks: stored, fixed, and
35 dynamic. The compressor deals with some chunk of data at a time, and
36 decides which method to use on a chunk-by-chunk basis. A chunk might
37 typically be 32 K or 64 K. If the chunk is incompressible, then the
38 "stored" method is used. In this case, the bytes are simply stored as
39 is, eight bits per byte, with none of the above coding. The bytes are
40 preceded by a count, since there is no longer an EOB code.
42 If the data is compressible, then either the fixed or dynamic methods
43 are used. In the dynamic method, the compressed data is preceded by
44 an encoding of the literal/length and distance Huffman codes that are
45 to be used to decode this block. The representation is itself Huffman
46 coded, and so is preceded by a description of that code. These code
47 descriptions take up a little space, and so for small blocks, there is
48 a predefined set of codes, called the fixed codes. The fixed method is
49 used if the block codes up smaller that way (usually for quite small
50 chunks), otherwise the dynamic method is used. In the latter case, the
51 codes are customized to the probabilities in the current block, and so
52 can code it much better than the pre-determined fixed codes.
54 The Huffman codes themselves are decoded using a multi-level table
55 lookup, in order to maximize the speed of decoding plus the speed of
56 building the decoding tables. See the comments below that precede the
57 lbits and dbits tuning parameters.
62 Notes beyond the 1.93a appnote.txt:
64 1. Distance pointers never point before the beginning of the output
65 stream.
66 2. Distance pointers can point back across blocks, up to 32k away.
67 3. There is an implied maximum of 7 bits for the bit length table and
68 15 bits for the actual data.
69 4. If only one code exists, then it is encoded using one bit. (Zero
70 would be more efficient, but perhaps a little confusing.) If two
71 codes exist, they are coded using one bit each (0 and 1).
72 5. There is no way of sending zero distance codes--a dummy must be
73 sent if there are none. (History: a pre 2.0 version of PKZIP would
74 store blocks with no distance codes, but this was discovered to be
75 too harsh a criterion.) Valid only for 1.93a. 2.04c does allow
76 zero distance codes, which is sent as one code of zero bits in
77 length.
78 6. There are up to 286 literal/length codes. Code 256 represents the
79 end-of-block. Note however that the static length tree defines
80 288 codes just to fill out the Huffman codes. Codes 286 and 287
81 cannot be used though, since there is no length base or extra bits
82 defined for them. Similarly, there are up to 30 distance codes.
83 However, static trees define 32 codes (all 5 bits) to fill out the
84 Huffman codes, but the last two had better not show up in the data.
85 7. Unzip can check dynamic Huffman blocks for complete code sets.
86 The exception is that a single code would not be complete (see #4).
87 8. The five bits following the block type is really the number of
88 literal codes sent minus 257.
89 9. Length codes 8,16,16 are interpreted as 13 length codes of 8 bits
90 (1+6+6). Therefore, to output three times the length, you output
91 three codes (1+1+1), whereas to output four times the same length,
92 you only need two codes (1+3). Hmm.
93 10. In the tree reconstruction algorithm, Code = Code + Increment
94 only if BitLength(i) is not zero. (Pretty obvious.)
95 11. Correction: 4 Bits: # of Bit Length codes - 4 (4 - 19)
96 12. Note: length code 284 can represent 227-258, but length code 285
97 really is 258. The last length deserves its own, short code
98 since it gets used a lot in very redundant files. The length
99 258 is special since 258 - 3 (the min match length) is 255.
100 13. The literal/length and distance code bit lengths are read as a
101 single stream of lengths. It is possible (and advantageous) for
102 a repeat code (16, 17, or 18) to go across the boundary between
103 the two sets of lengths.
105 #include <linux/compiler.h>
107 #ifdef RCSID
108 static char rcsid[] = "#Id: inflate.c,v 0.14 1993/06/10 13:27:04 jloup Exp #";
109 #endif
111 #ifndef STATIC
113 #if defined(STDC_HEADERS) || defined(HAVE_STDLIB_H)
114 # include <sys/types.h>
115 # include <stdlib.h>
116 #endif
118 #include "gzip.h"
119 #define STATIC
120 #endif /* !STATIC */
122 #define slide window
124 /* Huffman code lookup table entry--this entry is four bytes for machines
125 that have 16-bit pointers (e.g. PC's in the small or medium model).
126 Valid extra bits are 0..13. e == 15 is EOB (end of block), e == 16
127 means that v is a literal, 16 < e < 32 means that v is a pointer to
128 the next table, which codes e - 16 bits, and lastly e == 99 indicates
129 an unused code. If a code with e == 99 is looked up, this implies an
130 error in the data. */
131 struct huft {
132 uch e; /* number of extra bits or operation */
133 uch b; /* number of bits in this code or subcode */
134 union {
135 ush n; /* literal, length base, or distance base */
136 struct huft *t; /* pointer to next level of table */
137 } v;
141 /* Function prototypes */
142 STATIC int huft_build OF((unsigned *, unsigned, unsigned,
143 const ush *, const ush *, struct huft **, int *));
144 STATIC int huft_free OF((struct huft *));
145 STATIC int inflate_codes OF((struct huft *, struct huft *, int, int));
146 STATIC int inflate_stored OF((void));
147 STATIC int inflate_fixed OF((void));
148 STATIC int inflate_dynamic OF((void));
149 STATIC int inflate_block OF((int *));
150 STATIC int inflate OF((void));
153 /* The inflate algorithm uses a sliding 32 K byte window on the uncompressed
154 stream to find repeated byte strings. This is implemented here as a
155 circular buffer. The index is updated simply by incrementing and then
156 ANDing with 0x7fff (32K-1). */
157 /* It is left to other modules to supply the 32 K area. It is assumed
158 to be usable as if it were declared "uch slide[32768];" or as just
159 "uch *slide;" and then malloc'ed in the latter case. The definition
160 must be in unzip.h, included above. */
161 /* unsigned wp; current position in slide */
162 #define wp outcnt
163 #define flush_output(w) (wp=(w),flush_window())
165 /* Tables for deflate from PKZIP's appnote.txt. */
166 static const unsigned border[] = { /* Order of the bit length code lengths */
167 16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15};
168 static const ush cplens[] = { /* Copy lengths for literal codes 257..285 */
169 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 15, 17, 19, 23, 27, 31,
170 35, 43, 51, 59, 67, 83, 99, 115, 131, 163, 195, 227, 258, 0, 0};
171 /* note: see note #13 above about the 258 in this list. */
172 static const ush cplext[] = { /* Extra bits for literal codes 257..285 */
173 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2,
174 3, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5, 0, 99, 99}; /* 99==invalid */
175 static const ush cpdist[] = { /* Copy offsets for distance codes 0..29 */
176 1, 2, 3, 4, 5, 7, 9, 13, 17, 25, 33, 49, 65, 97, 129, 193,
177 257, 385, 513, 769, 1025, 1537, 2049, 3073, 4097, 6145,
178 8193, 12289, 16385, 24577};
179 static const ush cpdext[] = { /* Extra bits for distance codes */
180 0, 0, 0, 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6,
181 7, 7, 8, 8, 9, 9, 10, 10, 11, 11,
182 12, 12, 13, 13};
186 /* Macros for inflate() bit peeking and grabbing.
187 The usage is:
189 NEEDBITS(j)
190 x = b & mask_bits[j];
191 DUMPBITS(j)
193 where NEEDBITS makes sure that b has at least j bits in it, and
194 DUMPBITS removes the bits from b. The macros use the variable k
195 for the number of bits in b. Normally, b and k are register
196 variables for speed, and are initialized at the beginning of a
197 routine that uses these macros from a global bit buffer and count.
199 If we assume that EOB will be the longest code, then we will never
200 ask for bits with NEEDBITS that are beyond the end of the stream.
201 So, NEEDBITS should not read any more bytes than are needed to
202 meet the request. Then no bytes need to be "returned" to the buffer
203 at the end of the last block.
205 However, this assumption is not true for fixed blocks--the EOB code
206 is 7 bits, but the other literal/length codes can be 8 or 9 bits.
207 (The EOB code is shorter than other codes because fixed blocks are
208 generally short. So, while a block always has an EOB, many other
209 literal/length codes have a significantly lower probability of
210 showing up at all.) However, by making the first table have a
211 lookup of seven bits, the EOB code will be found in that first
212 lookup, and so will not require that too many bits be pulled from
213 the stream.
216 STATIC ulg bb; /* bit buffer */
217 STATIC unsigned bk; /* bits in bit buffer */
219 STATIC const ush mask_bits[] = {
220 0x0000,
221 0x0001, 0x0003, 0x0007, 0x000f, 0x001f, 0x003f, 0x007f, 0x00ff,
222 0x01ff, 0x03ff, 0x07ff, 0x0fff, 0x1fff, 0x3fff, 0x7fff, 0xffff
225 #define NEXTBYTE() ({ int v = get_byte(); if (v < 0) goto underrun; (uch)v; })
226 #define NEEDBITS(n) {while(k<(n)){b|=((ulg)NEXTBYTE())<<k;k+=8;}}
227 #define DUMPBITS(n) {b>>=(n);k-=(n);}
231 Huffman code decoding is performed using a multi-level table lookup.
232 The fastest way to decode is to simply build a lookup table whose
233 size is determined by the longest code. However, the time it takes
234 to build this table can also be a factor if the data being decoded
235 is not very long. The most common codes are necessarily the
236 shortest codes, so those codes dominate the decoding time, and hence
237 the speed. The idea is you can have a shorter table that decodes the
238 shorter, more probable codes, and then point to subsidiary tables for
239 the longer codes. The time it costs to decode the longer codes is
240 then traded against the time it takes to make longer tables.
242 This results of this trade are in the variables lbits and dbits
243 below. lbits is the number of bits the first level table for literal/
244 length codes can decode in one step, and dbits is the same thing for
245 the distance codes. Subsequent tables are also less than or equal to
246 those sizes. These values may be adjusted either when all of the
247 codes are shorter than that, in which case the longest code length in
248 bits is used, or when the shortest code is *longer* than the requested
249 table size, in which case the length of the shortest code in bits is
250 used.
252 There are two different values for the two tables, since they code a
253 different number of possibilities each. The literal/length table
254 codes 286 possible values, or in a flat code, a little over eight
255 bits. The distance table codes 30 possible values, or a little less
256 than five bits, flat. The optimum values for speed end up being
257 about one bit more than those, so lbits is 8+1 and dbits is 5+1.
258 The optimum values may differ though from machine to machine, and
259 possibly even between compilers. Your mileage may vary.
263 STATIC const int lbits = 9; /* bits in base literal/length lookup table */
264 STATIC const int dbits = 6; /* bits in base distance lookup table */
267 /* If BMAX needs to be larger than 16, then h and x[] should be ulg. */
268 #define BMAX 16 /* maximum bit length of any code (16 for explode) */
269 #define N_MAX 288 /* maximum number of codes in any set */
272 STATIC unsigned hufts; /* track memory usage */
275 STATIC int huft_build(
276 unsigned *b, /* code lengths in bits (all assumed <= BMAX) */
277 unsigned n, /* number of codes (assumed <= N_MAX) */
278 unsigned s, /* number of simple-valued codes (0..s-1) */
279 const ush *d, /* list of base values for non-simple codes */
280 const ush *e, /* list of extra bits for non-simple codes */
281 struct huft **t, /* result: starting table */
282 int *m /* maximum lookup bits, returns actual */
284 /* Given a list of code lengths and a maximum table size, make a set of
285 tables to decode that set of codes. Return zero on success, one if
286 the given code set is incomplete (the tables are still built in this
287 case), two if the input is invalid (all zero length codes or an
288 oversubscribed set of lengths), and three if not enough memory. */
290 unsigned a; /* counter for codes of length k */
291 unsigned c[BMAX+1]; /* bit length count table */
292 unsigned f; /* i repeats in table every f entries */
293 int g; /* maximum code length */
294 int h; /* table level */
295 register unsigned i; /* counter, current code */
296 register unsigned j; /* counter */
297 register int k; /* number of bits in current code */
298 int l; /* bits per table (returned in m) */
299 register unsigned *p; /* pointer into c[], b[], or v[] */
300 register struct huft *q; /* points to current table */
301 struct huft r; /* table entry for structure assignment */
302 struct huft *u[BMAX]; /* table stack */
303 unsigned v[N_MAX]; /* values in order of bit length */
304 register int w; /* bits before this table == (l * h) */
305 unsigned x[BMAX+1]; /* bit offsets, then code stack */
306 unsigned *xp; /* pointer into x */
307 int y; /* number of dummy codes added */
308 unsigned z; /* number of entries in current table */
310 DEBG("huft1 ");
312 /* Generate counts for each bit length */
313 memzero(c, sizeof(c));
314 p = b; i = n;
315 do {
316 Tracecv(*p, (stderr, (n-i >= ' ' && n-i <= '~' ? "%c %d\n" : "0x%x %d\n"),
317 n-i, *p));
318 c[*p]++; /* assume all entries <= BMAX */
319 p++; /* Can't combine with above line (Solaris bug) */
320 } while (--i);
321 if (c[0] == n) /* null input--all zero length codes */
323 *t = (struct huft *)NULL;
324 *m = 0;
325 return 0;
328 DEBG("huft2 ");
330 /* Find minimum and maximum length, bound *m by those */
331 l = *m;
332 for (j = 1; j <= BMAX; j++)
333 if (c[j])
334 break;
335 k = j; /* minimum code length */
336 if ((unsigned)l < j)
337 l = j;
338 for (i = BMAX; i; i--)
339 if (c[i])
340 break;
341 g = i; /* maximum code length */
342 if ((unsigned)l > i)
343 l = i;
344 *m = l;
346 DEBG("huft3 ");
348 /* Adjust last length count to fill out codes, if needed */
349 for (y = 1 << j; j < i; j++, y <<= 1)
350 if ((y -= c[j]) < 0)
351 return 2; /* bad input: more codes than bits */
352 if ((y -= c[i]) < 0)
353 return 2;
354 c[i] += y;
356 DEBG("huft4 ");
358 /* Generate starting offsets into the value table for each length */
359 x[1] = j = 0;
360 p = c + 1; xp = x + 2;
361 while (--i) { /* note that i == g from above */
362 *xp++ = (j += *p++);
365 DEBG("huft5 ");
367 /* Make a table of values in order of bit lengths */
368 p = b; i = 0;
369 do {
370 if ((j = *p++) != 0)
371 v[x[j]++] = i;
372 } while (++i < n);
374 DEBG("h6 ");
376 /* Generate the Huffman codes and for each, make the table entries */
377 x[0] = i = 0; /* first Huffman code is zero */
378 p = v; /* grab values in bit order */
379 h = -1; /* no tables yet--level -1 */
380 w = -l; /* bits decoded == (l * h) */
381 u[0] = (struct huft *)NULL; /* just to keep compilers happy */
382 q = (struct huft *)NULL; /* ditto */
383 z = 0; /* ditto */
384 DEBG("h6a ");
386 /* go through the bit lengths (k already is bits in shortest code) */
387 for (; k <= g; k++)
389 DEBG("h6b ");
390 a = c[k];
391 while (a--)
393 DEBG("h6b1 ");
394 /* here i is the Huffman code of length k bits for value *p */
395 /* make tables up to required level */
396 while (k > w + l)
398 DEBG1("1 ");
399 h++;
400 w += l; /* previous table always l bits */
402 /* compute minimum size table less than or equal to l bits */
403 z = (z = g - w) > (unsigned)l ? l : z; /* upper limit on table size */
404 if ((f = 1 << (j = k - w)) > a + 1) /* try a k-w bit table */
405 { /* too few codes for k-w bit table */
406 DEBG1("2 ");
407 f -= a + 1; /* deduct codes from patterns left */
408 xp = c + k;
409 while (++j < z) /* try smaller tables up to z bits */
411 if ((f <<= 1) <= *++xp)
412 break; /* enough codes to use up j bits */
413 f -= *xp; /* else deduct codes from patterns */
416 DEBG1("3 ");
417 z = 1 << j; /* table entries for j-bit table */
419 /* allocate and link in new table */
420 if ((q = (struct huft *)malloc((z + 1)*sizeof(struct huft))) ==
421 (struct huft *)NULL)
423 if (h)
424 huft_free(u[0]);
425 return 3; /* not enough memory */
427 DEBG1("4 ");
428 hufts += z + 1; /* track memory usage */
429 *t = q + 1; /* link to list for huft_free() */
430 *(t = &(q->v.t)) = (struct huft *)NULL;
431 u[h] = ++q; /* table starts after link */
433 DEBG1("5 ");
434 /* connect to last table, if there is one */
435 if (h)
437 x[h] = i; /* save pattern for backing up */
438 r.b = (uch)l; /* bits to dump before this table */
439 r.e = (uch)(16 + j); /* bits in this table */
440 r.v.t = q; /* pointer to this table */
441 j = i >> (w - l); /* (get around Turbo C bug) */
442 u[h-1][j] = r; /* connect to last table */
444 DEBG1("6 ");
446 DEBG("h6c ");
448 /* set up table entry in r */
449 r.b = (uch)(k - w);
450 if (p >= v + n)
451 r.e = 99; /* out of values--invalid code */
452 else if (*p < s)
454 r.e = (uch)(*p < 256 ? 16 : 15); /* 256 is end-of-block code */
455 r.v.n = (ush)(*p); /* simple code is just the value */
456 p++; /* one compiler does not like *p++ */
458 else
460 r.e = (uch)e[*p - s]; /* non-simple--look up in lists */
461 r.v.n = d[*p++ - s];
463 DEBG("h6d ");
465 /* fill code-like entries with r */
466 f = 1 << (k - w);
467 for (j = i >> w; j < z; j += f)
468 q[j] = r;
470 /* backwards increment the k-bit code i */
471 for (j = 1 << (k - 1); i & j; j >>= 1)
472 i ^= j;
473 i ^= j;
475 /* backup over finished tables */
476 while ((i & ((1 << w) - 1)) != x[h])
478 h--; /* don't need to update q */
479 w -= l;
481 DEBG("h6e ");
483 DEBG("h6f ");
486 DEBG("huft7 ");
488 /* Return true (1) if we were given an incomplete table */
489 return y != 0 && g != 1;
494 STATIC int huft_free(
495 struct huft *t /* table to free */
497 /* Free the malloc'ed tables built by huft_build(), which makes a linked
498 list of the tables it made, with the links in a dummy first entry of
499 each table. */
501 register struct huft *p, *q;
504 /* Go through linked list, freeing from the malloced (t[-1]) address. */
505 p = t;
506 while (p != (struct huft *)NULL)
508 q = (--p)->v.t;
509 free((char*)p);
510 p = q;
512 return 0;
516 STATIC int inflate_codes(
517 struct huft *tl, /* literal/length decoder tables */
518 struct huft *td, /* distance decoder tables */
519 int bl, /* number of bits decoded by tl[] */
520 int bd /* number of bits decoded by td[] */
522 /* inflate (decompress) the codes in a deflated (compressed) block.
523 Return an error code or zero if it all goes ok. */
525 register unsigned e; /* table entry flag/number of extra bits */
526 unsigned n, d; /* length and index for copy */
527 unsigned w; /* current window position */
528 struct huft *t; /* pointer to table entry */
529 unsigned ml, md; /* masks for bl and bd bits */
530 register ulg b; /* bit buffer */
531 register unsigned k; /* number of bits in bit buffer */
534 /* make local copies of globals */
535 b = bb; /* initialize bit buffer */
536 k = bk;
537 w = wp; /* initialize window position */
539 /* inflate the coded data */
540 ml = mask_bits[bl]; /* precompute masks for speed */
541 md = mask_bits[bd];
542 for (;;) /* do until end of block */
544 NEEDBITS((unsigned)bl)
545 if ((e = (t = tl + ((unsigned)b & ml))->e) > 16)
546 do {
547 if (e == 99)
548 return 1;
549 DUMPBITS(t->b)
550 e -= 16;
551 NEEDBITS(e)
552 } while ((e = (t = t->v.t + ((unsigned)b & mask_bits[e]))->e) > 16);
553 DUMPBITS(t->b)
554 if (e == 16) /* then it's a literal */
556 slide[w++] = (uch)t->v.n;
557 Tracevv((stderr, "%c", slide[w-1]));
558 if (w == WSIZE)
560 flush_output(w);
561 w = 0;
564 else /* it's an EOB or a length */
566 /* exit if end of block */
567 if (e == 15)
568 break;
570 /* get length of block to copy */
571 NEEDBITS(e)
572 n = t->v.n + ((unsigned)b & mask_bits[e]);
573 DUMPBITS(e);
575 /* decode distance of block to copy */
576 NEEDBITS((unsigned)bd)
577 if ((e = (t = td + ((unsigned)b & md))->e) > 16)
578 do {
579 if (e == 99)
580 return 1;
581 DUMPBITS(t->b)
582 e -= 16;
583 NEEDBITS(e)
584 } while ((e = (t = t->v.t + ((unsigned)b & mask_bits[e]))->e) > 16);
585 DUMPBITS(t->b)
586 NEEDBITS(e)
587 d = w - t->v.n - ((unsigned)b & mask_bits[e]);
588 DUMPBITS(e)
589 Tracevv((stderr,"\\[%d,%d]", w-d, n));
591 /* do the copy */
592 do {
593 n -= (e = (e = WSIZE - ((d &= WSIZE-1) > w ? d : w)) > n ? n : e);
594 #if !defined(NOMEMCPY) && !defined(DEBUG)
595 if (w - d >= e) /* (this test assumes unsigned comparison) */
597 memcpy(slide + w, slide + d, e);
598 w += e;
599 d += e;
601 else /* do it slow to avoid memcpy() overlap */
602 #endif /* !NOMEMCPY */
603 do {
604 slide[w++] = slide[d++];
605 Tracevv((stderr, "%c", slide[w-1]));
606 } while (--e);
607 if (w == WSIZE)
609 flush_output(w);
610 w = 0;
612 } while (n);
617 /* restore the globals from the locals */
618 wp = w; /* restore global window pointer */
619 bb = b; /* restore global bit buffer */
620 bk = k;
622 /* done */
623 return 0;
625 underrun:
626 return 4; /* Input underrun */
631 STATIC int inflate_stored(void)
632 /* "decompress" an inflated type 0 (stored) block. */
634 unsigned n; /* number of bytes in block */
635 unsigned w; /* current window position */
636 register ulg b; /* bit buffer */
637 register unsigned k; /* number of bits in bit buffer */
639 DEBG("<stor");
641 /* make local copies of globals */
642 b = bb; /* initialize bit buffer */
643 k = bk;
644 w = wp; /* initialize window position */
647 /* go to byte boundary */
648 n = k & 7;
649 DUMPBITS(n);
652 /* get the length and its complement */
653 NEEDBITS(16)
654 n = ((unsigned)b & 0xffff);
655 DUMPBITS(16)
656 NEEDBITS(16)
657 if (n != (unsigned)((~b) & 0xffff))
658 return 1; /* error in compressed data */
659 DUMPBITS(16)
662 /* read and output the compressed data */
663 while (n--)
665 NEEDBITS(8)
666 slide[w++] = (uch)b;
667 if (w == WSIZE)
669 flush_output(w);
670 w = 0;
672 DUMPBITS(8)
676 /* restore the globals from the locals */
677 wp = w; /* restore global window pointer */
678 bb = b; /* restore global bit buffer */
679 bk = k;
681 DEBG(">");
682 return 0;
684 underrun:
685 return 4; /* Input underrun */
690 * We use `noinline' here to prevent gcc-3.5 from using too much stack space
692 STATIC int noinline inflate_fixed(void)
693 /* decompress an inflated type 1 (fixed Huffman codes) block. We should
694 either replace this with a custom decoder, or at least precompute the
695 Huffman tables. */
697 int i; /* temporary variable */
698 struct huft *tl; /* literal/length code table */
699 struct huft *td; /* distance code table */
700 int bl; /* lookup bits for tl */
701 int bd; /* lookup bits for td */
702 unsigned l[288]; /* length list for huft_build */
704 DEBG("<fix");
706 /* set up literal table */
707 for (i = 0; i < 144; i++)
708 l[i] = 8;
709 for (; i < 256; i++)
710 l[i] = 9;
711 for (; i < 280; i++)
712 l[i] = 7;
713 for (; i < 288; i++) /* make a complete, but wrong code set */
714 l[i] = 8;
715 bl = 7;
716 if ((i = huft_build(l, 288, 257, cplens, cplext, &tl, &bl)) != 0)
717 return i;
720 /* set up distance table */
721 for (i = 0; i < 30; i++) /* make an incomplete code set */
722 l[i] = 5;
723 bd = 5;
724 if ((i = huft_build(l, 30, 0, cpdist, cpdext, &td, &bd)) > 1)
726 huft_free(tl);
728 DEBG(">");
729 return i;
733 /* decompress until an end-of-block code */
734 if (inflate_codes(tl, td, bl, bd))
735 return 1;
738 /* free the decoding tables, return */
739 huft_free(tl);
740 huft_free(td);
741 return 0;
746 * We use `noinline' here to prevent gcc-3.5 from using too much stack space
748 STATIC int noinline inflate_dynamic(void)
749 /* decompress an inflated type 2 (dynamic Huffman codes) block. */
751 int i; /* temporary variables */
752 unsigned j;
753 unsigned l; /* last length */
754 unsigned m; /* mask for bit lengths table */
755 unsigned n; /* number of lengths to get */
756 struct huft *tl; /* literal/length code table */
757 struct huft *td; /* distance code table */
758 int bl; /* lookup bits for tl */
759 int bd; /* lookup bits for td */
760 unsigned nb; /* number of bit length codes */
761 unsigned nl; /* number of literal/length codes */
762 unsigned nd; /* number of distance codes */
763 #ifdef PKZIP_BUG_WORKAROUND
764 unsigned ll[288+32]; /* literal/length and distance code lengths */
765 #else
766 unsigned ll[286+30]; /* literal/length and distance code lengths */
767 #endif
768 register ulg b; /* bit buffer */
769 register unsigned k; /* number of bits in bit buffer */
771 DEBG("<dyn");
773 /* make local bit buffer */
774 b = bb;
775 k = bk;
778 /* read in table lengths */
779 NEEDBITS(5)
780 nl = 257 + ((unsigned)b & 0x1f); /* number of literal/length codes */
781 DUMPBITS(5)
782 NEEDBITS(5)
783 nd = 1 + ((unsigned)b & 0x1f); /* number of distance codes */
784 DUMPBITS(5)
785 NEEDBITS(4)
786 nb = 4 + ((unsigned)b & 0xf); /* number of bit length codes */
787 DUMPBITS(4)
788 #ifdef PKZIP_BUG_WORKAROUND
789 if (nl > 288 || nd > 32)
790 #else
791 if (nl > 286 || nd > 30)
792 #endif
793 return 1; /* bad lengths */
795 DEBG("dyn1 ");
797 /* read in bit-length-code lengths */
798 for (j = 0; j < nb; j++)
800 NEEDBITS(3)
801 ll[border[j]] = (unsigned)b & 7;
802 DUMPBITS(3)
804 for (; j < 19; j++)
805 ll[border[j]] = 0;
807 DEBG("dyn2 ");
809 /* build decoding table for trees--single level, 7 bit lookup */
810 bl = 7;
811 if ((i = huft_build(ll, 19, 19, NULL, NULL, &tl, &bl)) != 0)
813 if (i == 1)
814 huft_free(tl);
815 return i; /* incomplete code set */
818 DEBG("dyn3 ");
820 /* read in literal and distance code lengths */
821 n = nl + nd;
822 m = mask_bits[bl];
823 i = l = 0;
824 while ((unsigned)i < n)
826 NEEDBITS((unsigned)bl)
827 j = (td = tl + ((unsigned)b & m))->b;
828 DUMPBITS(j)
829 j = td->v.n;
830 if (j < 16) /* length of code in bits (0..15) */
831 ll[i++] = l = j; /* save last length in l */
832 else if (j == 16) /* repeat last length 3 to 6 times */
834 NEEDBITS(2)
835 j = 3 + ((unsigned)b & 3);
836 DUMPBITS(2)
837 if ((unsigned)i + j > n)
838 return 1;
839 while (j--)
840 ll[i++] = l;
842 else if (j == 17) /* 3 to 10 zero length codes */
844 NEEDBITS(3)
845 j = 3 + ((unsigned)b & 7);
846 DUMPBITS(3)
847 if ((unsigned)i + j > n)
848 return 1;
849 while (j--)
850 ll[i++] = 0;
851 l = 0;
853 else /* j == 18: 11 to 138 zero length codes */
855 NEEDBITS(7)
856 j = 11 + ((unsigned)b & 0x7f);
857 DUMPBITS(7)
858 if ((unsigned)i + j > n)
859 return 1;
860 while (j--)
861 ll[i++] = 0;
862 l = 0;
866 DEBG("dyn4 ");
868 /* free decoding table for trees */
869 huft_free(tl);
871 DEBG("dyn5 ");
873 /* restore the global bit buffer */
874 bb = b;
875 bk = k;
877 DEBG("dyn5a ");
879 /* build the decoding tables for literal/length and distance codes */
880 bl = lbits;
881 if ((i = huft_build(ll, nl, 257, cplens, cplext, &tl, &bl)) != 0)
883 DEBG("dyn5b ");
884 if (i == 1) {
885 error("incomplete literal tree");
886 huft_free(tl);
888 return i; /* incomplete code set */
890 DEBG("dyn5c ");
891 bd = dbits;
892 if ((i = huft_build(ll + nl, nd, 0, cpdist, cpdext, &td, &bd)) != 0)
894 DEBG("dyn5d ");
895 if (i == 1) {
896 error("incomplete distance tree");
897 #ifdef PKZIP_BUG_WORKAROUND
898 i = 0;
900 #else
901 huft_free(td);
903 huft_free(tl);
904 return i; /* incomplete code set */
905 #endif
908 DEBG("dyn6 ");
910 /* decompress until an end-of-block code */
911 if (inflate_codes(tl, td, bl, bd))
912 return 1;
914 DEBG("dyn7 ");
916 /* free the decoding tables, return */
917 huft_free(tl);
918 huft_free(td);
920 DEBG(">");
921 return 0;
923 underrun:
924 return 4; /* Input underrun */
929 STATIC int inflate_block(
930 int *e /* last block flag */
932 /* decompress an inflated block */
934 unsigned t; /* block type */
935 register ulg b; /* bit buffer */
936 register unsigned k; /* number of bits in bit buffer */
938 DEBG("<blk");
940 /* make local bit buffer */
941 b = bb;
942 k = bk;
945 /* read in last block bit */
946 NEEDBITS(1)
947 *e = (int)b & 1;
948 DUMPBITS(1)
951 /* read in block type */
952 NEEDBITS(2)
953 t = (unsigned)b & 3;
954 DUMPBITS(2)
957 /* restore the global bit buffer */
958 bb = b;
959 bk = k;
961 /* inflate that block type */
962 if (t == 2)
963 return inflate_dynamic();
964 if (t == 0)
965 return inflate_stored();
966 if (t == 1)
967 return inflate_fixed();
969 DEBG(">");
971 /* bad block type */
972 return 2;
974 underrun:
975 return 4; /* Input underrun */
980 STATIC int inflate(void)
981 /* decompress an inflated entry */
983 int e; /* last block flag */
984 int r; /* result code */
985 unsigned h; /* maximum struct huft's malloc'ed */
986 void *ptr;
988 /* initialize window, bit buffer */
989 wp = 0;
990 bk = 0;
991 bb = 0;
994 /* decompress until the last block */
995 h = 0;
996 do {
997 hufts = 0;
998 gzip_mark(&ptr);
999 if ((r = inflate_block(&e)) != 0) {
1000 gzip_release(&ptr);
1001 return r;
1003 gzip_release(&ptr);
1004 if (hufts > h)
1005 h = hufts;
1006 } while (!e);
1008 /* Undo too much lookahead. The next read will be byte aligned so we
1009 * can discard unused bits in the last meaningful byte.
1011 while (bk >= 8) {
1012 bk -= 8;
1013 inptr--;
1016 /* flush out slide */
1017 flush_output(wp);
1020 /* return success */
1021 #ifdef DEBUG
1022 fprintf(stderr, "<%u> ", h);
1023 #endif /* DEBUG */
1024 return 0;
1027 /**********************************************************************
1029 * The following are support routines for inflate.c
1031 **********************************************************************/
1033 static ulg crc_32_tab[256];
1034 static ulg crc; /* initialized in makecrc() so it'll reside in bss */
1035 #define CRC_VALUE (crc ^ 0xffffffffUL)
1038 * Code to compute the CRC-32 table. Borrowed from
1039 * gzip-1.0.3/makecrc.c.
1042 static void
1043 makecrc(void)
1045 /* Not copyrighted 1990 Mark Adler */
1047 unsigned long c; /* crc shift register */
1048 unsigned long e; /* polynomial exclusive-or pattern */
1049 int i; /* counter for all possible eight bit values */
1050 int k; /* byte being shifted into crc apparatus */
1052 /* terms of polynomial defining this crc (except x^32): */
1053 static const int p[] = {0,1,2,4,5,7,8,10,11,12,16,22,23,26};
1055 /* Make exclusive-or pattern from polynomial */
1056 e = 0;
1057 for (i = 0; i < sizeof(p)/sizeof(int); i++)
1058 e |= 1L << (31 - p[i]);
1060 crc_32_tab[0] = 0;
1062 for (i = 1; i < 256; i++)
1064 c = 0;
1065 for (k = i | 256; k != 1; k >>= 1)
1067 c = c & 1 ? (c >> 1) ^ e : c >> 1;
1068 if (k & 1)
1069 c ^= e;
1071 crc_32_tab[i] = c;
1074 /* this is initialized here so this code could reside in ROM */
1075 crc = (ulg)0xffffffffUL; /* shift register contents */
1078 /* gzip flag byte */
1079 #define ASCII_FLAG 0x01 /* bit 0 set: file probably ASCII text */
1080 #define CONTINUATION 0x02 /* bit 1 set: continuation of multi-part gzip file */
1081 #define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */
1082 #define ORIG_NAME 0x08 /* bit 3 set: original file name present */
1083 #define COMMENT 0x10 /* bit 4 set: file comment present */
1084 #define ENCRYPTED 0x20 /* bit 5 set: file is encrypted */
1085 #define RESERVED 0xC0 /* bit 6,7: reserved */
1088 * Do the uncompression!
1090 static int gunzip(void)
1092 uch flags;
1093 unsigned char magic[2]; /* magic header */
1094 char method;
1095 ulg orig_crc = 0; /* original crc */
1096 ulg orig_len = 0; /* original uncompressed length */
1097 int res;
1099 magic[0] = NEXTBYTE();
1100 magic[1] = NEXTBYTE();
1101 method = NEXTBYTE();
1103 if (magic[0] != 037 ||
1104 ((magic[1] != 0213) && (magic[1] != 0236))) {
1105 error("bad gzip magic numbers");
1106 return -1;
1109 /* We only support method #8, DEFLATED */
1110 if (method != 8) {
1111 error("internal error, invalid method");
1112 return -1;
1115 flags = (uch)get_byte();
1116 if ((flags & ENCRYPTED) != 0) {
1117 error("Input is encrypted");
1118 return -1;
1120 if ((flags & CONTINUATION) != 0) {
1121 error("Multi part input");
1122 return -1;
1124 if ((flags & RESERVED) != 0) {
1125 error("Input has invalid flags");
1126 return -1;
1128 NEXTBYTE(); /* Get timestamp */
1129 NEXTBYTE();
1130 NEXTBYTE();
1131 NEXTBYTE();
1133 (void)NEXTBYTE(); /* Ignore extra flags for the moment */
1134 (void)NEXTBYTE(); /* Ignore OS type for the moment */
1136 if ((flags & EXTRA_FIELD) != 0) {
1137 unsigned len = (unsigned)NEXTBYTE();
1138 len |= ((unsigned)NEXTBYTE())<<8;
1139 while (len--) (void)NEXTBYTE();
1142 /* Get original file name if it was truncated */
1143 if ((flags & ORIG_NAME) != 0) {
1144 /* Discard the old name */
1145 while (NEXTBYTE() != 0) /* null */ ;
1148 /* Discard file comment if any */
1149 if ((flags & COMMENT) != 0) {
1150 while (NEXTBYTE() != 0) /* null */ ;
1153 /* Decompress */
1154 if ((res = inflate())) {
1155 switch (res) {
1156 case 0:
1157 break;
1158 case 1:
1159 error("invalid compressed format (err=1)");
1160 break;
1161 case 2:
1162 error("invalid compressed format (err=2)");
1163 break;
1164 case 3:
1165 error("out of memory");
1166 break;
1167 case 4:
1168 error("out of input data");
1169 break;
1170 default:
1171 error("invalid compressed format (other)");
1173 return -1;
1176 /* Get the crc and original length */
1177 /* crc32 (see algorithm.doc)
1178 * uncompressed input size modulo 2^32
1180 orig_crc = (ulg) NEXTBYTE();
1181 orig_crc |= (ulg) NEXTBYTE() << 8;
1182 orig_crc |= (ulg) NEXTBYTE() << 16;
1183 orig_crc |= (ulg) NEXTBYTE() << 24;
1185 orig_len = (ulg) NEXTBYTE();
1186 orig_len |= (ulg) NEXTBYTE() << 8;
1187 orig_len |= (ulg) NEXTBYTE() << 16;
1188 orig_len |= (ulg) NEXTBYTE() << 24;
1190 /* Validate decompression */
1191 if (orig_crc != CRC_VALUE) {
1192 error("crc error");
1193 return -1;
1195 if (orig_len != bytes_out) {
1196 error("length error");
1197 return -1;
1199 return 0;
1201 underrun: /* NEXTBYTE() goto's here if needed */
1202 error("out of input data");
1203 return -1;