Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1 | #define DEBG(x) |
| 2 | #define DEBG1(x) |
| 3 | /* inflate.c -- Not copyrighted 1992 by Mark Adler |
| 4 | version c10p1, 10 January 1993 */ |
| 5 | |
| 6 | /* |
| 7 | * Adapted for booting Linux by Hannu Savolainen 1993 |
| 8 | * based on gzip-1.0.3 |
| 9 | * |
Nicolas Pitre | 2f82af0 | 2009-09-14 03:25:28 -0400 | [diff] [blame] | 10 | * Nicolas Pitre <nico@fluxnic.net>, 1999/04/14 : |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 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. |
| 15 | */ |
| 16 | |
| 17 | /* |
| 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. |
| 33 | |
| 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. |
| 41 | |
| 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. |
| 53 | |
| 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. |
| 58 | */ |
| 59 | |
| 60 | |
| 61 | /* |
| 62 | Notes beyond the 1.93a appnote.txt: |
| 63 | |
| 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. |
| 104 | */ |
| 105 | #include <linux/compiler.h> |
Tejun Heo | 5a0e3ad | 2010-03-24 17:04:11 +0900 | [diff] [blame] | 106 | #include <linux/slab.h> |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 107 | |
| 108 | #ifdef RCSID |
| 109 | static char rcsid[] = "#Id: inflate.c,v 0.14 1993/06/10 13:27:04 jloup Exp #"; |
| 110 | #endif |
| 111 | |
| 112 | #ifndef STATIC |
| 113 | |
| 114 | #if defined(STDC_HEADERS) || defined(HAVE_STDLIB_H) |
| 115 | # include <sys/types.h> |
| 116 | # include <stdlib.h> |
| 117 | #endif |
| 118 | |
| 119 | #include "gzip.h" |
| 120 | #define STATIC |
| 121 | #endif /* !STATIC */ |
| 122 | |
| 123 | #ifndef INIT |
| 124 | #define INIT |
| 125 | #endif |
| 126 | |
| 127 | #define slide window |
| 128 | |
| 129 | /* Huffman code lookup table entry--this entry is four bytes for machines |
| 130 | that have 16-bit pointers (e.g. PC's in the small or medium model). |
| 131 | Valid extra bits are 0..13. e == 15 is EOB (end of block), e == 16 |
| 132 | means that v is a literal, 16 < e < 32 means that v is a pointer to |
| 133 | the next table, which codes e - 16 bits, and lastly e == 99 indicates |
| 134 | an unused code. If a code with e == 99 is looked up, this implies an |
| 135 | error in the data. */ |
| 136 | struct huft { |
| 137 | uch e; /* number of extra bits or operation */ |
| 138 | uch b; /* number of bits in this code or subcode */ |
| 139 | union { |
| 140 | ush n; /* literal, length base, or distance base */ |
| 141 | struct huft *t; /* pointer to next level of table */ |
| 142 | } v; |
| 143 | }; |
| 144 | |
| 145 | |
| 146 | /* Function prototypes */ |
| 147 | STATIC int INIT huft_build OF((unsigned *, unsigned, unsigned, |
| 148 | const ush *, const ush *, struct huft **, int *)); |
| 149 | STATIC int INIT huft_free OF((struct huft *)); |
| 150 | STATIC int INIT inflate_codes OF((struct huft *, struct huft *, int, int)); |
| 151 | STATIC int INIT inflate_stored OF((void)); |
| 152 | STATIC int INIT inflate_fixed OF((void)); |
| 153 | STATIC int INIT inflate_dynamic OF((void)); |
| 154 | STATIC int INIT inflate_block OF((int *)); |
| 155 | STATIC int INIT inflate OF((void)); |
| 156 | |
| 157 | |
| 158 | /* The inflate algorithm uses a sliding 32 K byte window on the uncompressed |
| 159 | stream to find repeated byte strings. This is implemented here as a |
| 160 | circular buffer. The index is updated simply by incrementing and then |
| 161 | ANDing with 0x7fff (32K-1). */ |
| 162 | /* It is left to other modules to supply the 32 K area. It is assumed |
| 163 | to be usable as if it were declared "uch slide[32768];" or as just |
| 164 | "uch *slide;" and then malloc'ed in the latter case. The definition |
| 165 | must be in unzip.h, included above. */ |
| 166 | /* unsigned wp; current position in slide */ |
| 167 | #define wp outcnt |
| 168 | #define flush_output(w) (wp=(w),flush_window()) |
| 169 | |
| 170 | /* Tables for deflate from PKZIP's appnote.txt. */ |
| 171 | static const unsigned border[] = { /* Order of the bit length code lengths */ |
| 172 | 16, 17, 18, 0, 8, 7, 9, 6, 10, 5, 11, 4, 12, 3, 13, 2, 14, 1, 15}; |
| 173 | static const ush cplens[] = { /* Copy lengths for literal codes 257..285 */ |
| 174 | 3, 4, 5, 6, 7, 8, 9, 10, 11, 13, 15, 17, 19, 23, 27, 31, |
| 175 | 35, 43, 51, 59, 67, 83, 99, 115, 131, 163, 195, 227, 258, 0, 0}; |
| 176 | /* note: see note #13 above about the 258 in this list. */ |
| 177 | static const ush cplext[] = { /* Extra bits for literal codes 257..285 */ |
| 178 | 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 2, 2, 2, 2, |
| 179 | 3, 3, 3, 3, 4, 4, 4, 4, 5, 5, 5, 5, 0, 99, 99}; /* 99==invalid */ |
| 180 | static const ush cpdist[] = { /* Copy offsets for distance codes 0..29 */ |
| 181 | 1, 2, 3, 4, 5, 7, 9, 13, 17, 25, 33, 49, 65, 97, 129, 193, |
| 182 | 257, 385, 513, 769, 1025, 1537, 2049, 3073, 4097, 6145, |
| 183 | 8193, 12289, 16385, 24577}; |
| 184 | static const ush cpdext[] = { /* Extra bits for distance codes */ |
| 185 | 0, 0, 0, 0, 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 6, 6, |
| 186 | 7, 7, 8, 8, 9, 9, 10, 10, 11, 11, |
| 187 | 12, 12, 13, 13}; |
| 188 | |
| 189 | |
| 190 | |
| 191 | /* Macros for inflate() bit peeking and grabbing. |
| 192 | The usage is: |
| 193 | |
| 194 | NEEDBITS(j) |
| 195 | x = b & mask_bits[j]; |
| 196 | DUMPBITS(j) |
| 197 | |
| 198 | where NEEDBITS makes sure that b has at least j bits in it, and |
| 199 | DUMPBITS removes the bits from b. The macros use the variable k |
| 200 | for the number of bits in b. Normally, b and k are register |
| 201 | variables for speed, and are initialized at the beginning of a |
| 202 | routine that uses these macros from a global bit buffer and count. |
| 203 | |
| 204 | If we assume that EOB will be the longest code, then we will never |
| 205 | ask for bits with NEEDBITS that are beyond the end of the stream. |
| 206 | So, NEEDBITS should not read any more bytes than are needed to |
| 207 | meet the request. Then no bytes need to be "returned" to the buffer |
| 208 | at the end of the last block. |
| 209 | |
| 210 | However, this assumption is not true for fixed blocks--the EOB code |
| 211 | is 7 bits, but the other literal/length codes can be 8 or 9 bits. |
| 212 | (The EOB code is shorter than other codes because fixed blocks are |
| 213 | generally short. So, while a block always has an EOB, many other |
| 214 | literal/length codes have a significantly lower probability of |
| 215 | showing up at all.) However, by making the first table have a |
| 216 | lookup of seven bits, the EOB code will be found in that first |
| 217 | lookup, and so will not require that too many bits be pulled from |
| 218 | the stream. |
| 219 | */ |
| 220 | |
| 221 | STATIC ulg bb; /* bit buffer */ |
| 222 | STATIC unsigned bk; /* bits in bit buffer */ |
| 223 | |
| 224 | STATIC const ush mask_bits[] = { |
| 225 | 0x0000, |
| 226 | 0x0001, 0x0003, 0x0007, 0x000f, 0x001f, 0x003f, 0x007f, 0x00ff, |
| 227 | 0x01ff, 0x03ff, 0x07ff, 0x0fff, 0x1fff, 0x3fff, 0x7fff, 0xffff |
| 228 | }; |
| 229 | |
| 230 | #define NEXTBYTE() ({ int v = get_byte(); if (v < 0) goto underrun; (uch)v; }) |
| 231 | #define NEEDBITS(n) {while(k<(n)){b|=((ulg)NEXTBYTE())<<k;k+=8;}} |
| 232 | #define DUMPBITS(n) {b>>=(n);k-=(n);} |
| 233 | |
Thomas Petazzoni | 2d6ffcc | 2008-07-25 01:45:44 -0700 | [diff] [blame] | 234 | #ifndef NO_INFLATE_MALLOC |
| 235 | /* A trivial malloc implementation, adapted from |
| 236 | * malloc by Hannu Savolainen 1993 and Matthias Urlichs 1994 |
| 237 | */ |
| 238 | |
| 239 | static unsigned long malloc_ptr; |
| 240 | static int malloc_count; |
| 241 | |
| 242 | static void *malloc(int size) |
| 243 | { |
| 244 | void *p; |
| 245 | |
| 246 | if (size < 0) |
| 247 | error("Malloc error"); |
| 248 | if (!malloc_ptr) |
| 249 | malloc_ptr = free_mem_ptr; |
| 250 | |
| 251 | malloc_ptr = (malloc_ptr + 3) & ~3; /* Align */ |
| 252 | |
| 253 | p = (void *)malloc_ptr; |
| 254 | malloc_ptr += size; |
| 255 | |
| 256 | if (free_mem_end_ptr && malloc_ptr >= free_mem_end_ptr) |
| 257 | error("Out of memory"); |
| 258 | |
| 259 | malloc_count++; |
| 260 | return p; |
| 261 | } |
| 262 | |
| 263 | static void free(void *where) |
| 264 | { |
| 265 | malloc_count--; |
| 266 | if (!malloc_count) |
| 267 | malloc_ptr = free_mem_ptr; |
| 268 | } |
| 269 | #else |
| 270 | #define malloc(a) kmalloc(a, GFP_KERNEL) |
| 271 | #define free(a) kfree(a) |
| 272 | #endif |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 273 | |
| 274 | /* |
| 275 | Huffman code decoding is performed using a multi-level table lookup. |
| 276 | The fastest way to decode is to simply build a lookup table whose |
| 277 | size is determined by the longest code. However, the time it takes |
| 278 | to build this table can also be a factor if the data being decoded |
| 279 | is not very long. The most common codes are necessarily the |
| 280 | shortest codes, so those codes dominate the decoding time, and hence |
| 281 | the speed. The idea is you can have a shorter table that decodes the |
| 282 | shorter, more probable codes, and then point to subsidiary tables for |
| 283 | the longer codes. The time it costs to decode the longer codes is |
| 284 | then traded against the time it takes to make longer tables. |
| 285 | |
| 286 | This results of this trade are in the variables lbits and dbits |
| 287 | below. lbits is the number of bits the first level table for literal/ |
| 288 | length codes can decode in one step, and dbits is the same thing for |
| 289 | the distance codes. Subsequent tables are also less than or equal to |
| 290 | those sizes. These values may be adjusted either when all of the |
| 291 | codes are shorter than that, in which case the longest code length in |
| 292 | bits is used, or when the shortest code is *longer* than the requested |
| 293 | table size, in which case the length of the shortest code in bits is |
| 294 | used. |
| 295 | |
| 296 | There are two different values for the two tables, since they code a |
| 297 | different number of possibilities each. The literal/length table |
| 298 | codes 286 possible values, or in a flat code, a little over eight |
| 299 | bits. The distance table codes 30 possible values, or a little less |
| 300 | than five bits, flat. The optimum values for speed end up being |
| 301 | about one bit more than those, so lbits is 8+1 and dbits is 5+1. |
| 302 | The optimum values may differ though from machine to machine, and |
| 303 | possibly even between compilers. Your mileage may vary. |
| 304 | */ |
| 305 | |
| 306 | |
| 307 | STATIC const int lbits = 9; /* bits in base literal/length lookup table */ |
| 308 | STATIC const int dbits = 6; /* bits in base distance lookup table */ |
| 309 | |
| 310 | |
| 311 | /* If BMAX needs to be larger than 16, then h and x[] should be ulg. */ |
| 312 | #define BMAX 16 /* maximum bit length of any code (16 for explode) */ |
| 313 | #define N_MAX 288 /* maximum number of codes in any set */ |
| 314 | |
| 315 | |
| 316 | STATIC unsigned hufts; /* track memory usage */ |
| 317 | |
| 318 | |
| 319 | STATIC int INIT huft_build( |
| 320 | unsigned *b, /* code lengths in bits (all assumed <= BMAX) */ |
| 321 | unsigned n, /* number of codes (assumed <= N_MAX) */ |
| 322 | unsigned s, /* number of simple-valued codes (0..s-1) */ |
| 323 | const ush *d, /* list of base values for non-simple codes */ |
| 324 | const ush *e, /* list of extra bits for non-simple codes */ |
| 325 | struct huft **t, /* result: starting table */ |
| 326 | int *m /* maximum lookup bits, returns actual */ |
| 327 | ) |
| 328 | /* Given a list of code lengths and a maximum table size, make a set of |
| 329 | tables to decode that set of codes. Return zero on success, one if |
| 330 | the given code set is incomplete (the tables are still built in this |
| 331 | case), two if the input is invalid (all zero length codes or an |
| 332 | oversubscribed set of lengths), and three if not enough memory. */ |
| 333 | { |
| 334 | unsigned a; /* counter for codes of length k */ |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 335 | unsigned f; /* i repeats in table every f entries */ |
| 336 | int g; /* maximum code length */ |
| 337 | int h; /* table level */ |
| 338 | register unsigned i; /* counter, current code */ |
| 339 | register unsigned j; /* counter */ |
| 340 | register int k; /* number of bits in current code */ |
| 341 | int l; /* bits per table (returned in m) */ |
| 342 | register unsigned *p; /* pointer into c[], b[], or v[] */ |
| 343 | register struct huft *q; /* points to current table */ |
| 344 | struct huft r; /* table entry for structure assignment */ |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 345 | register int w; /* bits before this table == (l * h) */ |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 346 | unsigned *xp; /* pointer into x */ |
| 347 | int y; /* number of dummy codes added */ |
| 348 | unsigned z; /* number of entries in current table */ |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 349 | struct { |
| 350 | unsigned c[BMAX+1]; /* bit length count table */ |
| 351 | struct huft *u[BMAX]; /* table stack */ |
| 352 | unsigned v[N_MAX]; /* values in order of bit length */ |
| 353 | unsigned x[BMAX+1]; /* bit offsets, then code stack */ |
| 354 | } *stk; |
| 355 | unsigned *c, *v, *x; |
| 356 | struct huft **u; |
| 357 | int ret; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 358 | |
| 359 | DEBG("huft1 "); |
| 360 | |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 361 | stk = malloc(sizeof(*stk)); |
| 362 | if (stk == NULL) |
| 363 | return 3; /* out of memory */ |
| 364 | |
| 365 | c = stk->c; |
| 366 | v = stk->v; |
| 367 | x = stk->x; |
| 368 | u = stk->u; |
| 369 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 370 | /* Generate counts for each bit length */ |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 371 | memzero(stk->c, sizeof(stk->c)); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 372 | p = b; i = n; |
| 373 | do { |
| 374 | Tracecv(*p, (stderr, (n-i >= ' ' && n-i <= '~' ? "%c %d\n" : "0x%x %d\n"), |
| 375 | n-i, *p)); |
| 376 | c[*p]++; /* assume all entries <= BMAX */ |
| 377 | p++; /* Can't combine with above line (Solaris bug) */ |
| 378 | } while (--i); |
| 379 | if (c[0] == n) /* null input--all zero length codes */ |
| 380 | { |
| 381 | *t = (struct huft *)NULL; |
| 382 | *m = 0; |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 383 | ret = 2; |
| 384 | goto out; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 385 | } |
| 386 | |
| 387 | DEBG("huft2 "); |
| 388 | |
| 389 | /* Find minimum and maximum length, bound *m by those */ |
| 390 | l = *m; |
| 391 | for (j = 1; j <= BMAX; j++) |
| 392 | if (c[j]) |
| 393 | break; |
| 394 | k = j; /* minimum code length */ |
| 395 | if ((unsigned)l < j) |
| 396 | l = j; |
| 397 | for (i = BMAX; i; i--) |
| 398 | if (c[i]) |
| 399 | break; |
| 400 | g = i; /* maximum code length */ |
| 401 | if ((unsigned)l > i) |
| 402 | l = i; |
| 403 | *m = l; |
| 404 | |
| 405 | DEBG("huft3 "); |
| 406 | |
| 407 | /* Adjust last length count to fill out codes, if needed */ |
| 408 | for (y = 1 << j; j < i; j++, y <<= 1) |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 409 | if ((y -= c[j]) < 0) { |
| 410 | ret = 2; /* bad input: more codes than bits */ |
| 411 | goto out; |
| 412 | } |
| 413 | if ((y -= c[i]) < 0) { |
| 414 | ret = 2; |
| 415 | goto out; |
| 416 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 417 | c[i] += y; |
| 418 | |
| 419 | DEBG("huft4 "); |
| 420 | |
| 421 | /* Generate starting offsets into the value table for each length */ |
| 422 | x[1] = j = 0; |
| 423 | p = c + 1; xp = x + 2; |
| 424 | while (--i) { /* note that i == g from above */ |
| 425 | *xp++ = (j += *p++); |
| 426 | } |
| 427 | |
| 428 | DEBG("huft5 "); |
| 429 | |
| 430 | /* Make a table of values in order of bit lengths */ |
| 431 | p = b; i = 0; |
| 432 | do { |
| 433 | if ((j = *p++) != 0) |
| 434 | v[x[j]++] = i; |
| 435 | } while (++i < n); |
Tim Yamin | 4aad724 | 2005-07-25 23:16:13 +0100 | [diff] [blame] | 436 | n = x[g]; /* set n to length of v */ |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 437 | |
| 438 | DEBG("h6 "); |
| 439 | |
| 440 | /* Generate the Huffman codes and for each, make the table entries */ |
| 441 | x[0] = i = 0; /* first Huffman code is zero */ |
| 442 | p = v; /* grab values in bit order */ |
| 443 | h = -1; /* no tables yet--level -1 */ |
| 444 | w = -l; /* bits decoded == (l * h) */ |
| 445 | u[0] = (struct huft *)NULL; /* just to keep compilers happy */ |
| 446 | q = (struct huft *)NULL; /* ditto */ |
| 447 | z = 0; /* ditto */ |
| 448 | DEBG("h6a "); |
| 449 | |
| 450 | /* go through the bit lengths (k already is bits in shortest code) */ |
| 451 | for (; k <= g; k++) |
| 452 | { |
| 453 | DEBG("h6b "); |
| 454 | a = c[k]; |
| 455 | while (a--) |
| 456 | { |
| 457 | DEBG("h6b1 "); |
| 458 | /* here i is the Huffman code of length k bits for value *p */ |
| 459 | /* make tables up to required level */ |
| 460 | while (k > w + l) |
| 461 | { |
| 462 | DEBG1("1 "); |
| 463 | h++; |
| 464 | w += l; /* previous table always l bits */ |
| 465 | |
| 466 | /* compute minimum size table less than or equal to l bits */ |
| 467 | z = (z = g - w) > (unsigned)l ? l : z; /* upper limit on table size */ |
| 468 | if ((f = 1 << (j = k - w)) > a + 1) /* try a k-w bit table */ |
| 469 | { /* too few codes for k-w bit table */ |
| 470 | DEBG1("2 "); |
| 471 | f -= a + 1; /* deduct codes from patterns left */ |
| 472 | xp = c + k; |
Tim Yamin | 4aad724 | 2005-07-25 23:16:13 +0100 | [diff] [blame] | 473 | if (j < z) |
| 474 | while (++j < z) /* try smaller tables up to z bits */ |
| 475 | { |
| 476 | if ((f <<= 1) <= *++xp) |
| 477 | break; /* enough codes to use up j bits */ |
| 478 | f -= *xp; /* else deduct codes from patterns */ |
| 479 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 480 | } |
| 481 | DEBG1("3 "); |
| 482 | z = 1 << j; /* table entries for j-bit table */ |
| 483 | |
| 484 | /* allocate and link in new table */ |
| 485 | if ((q = (struct huft *)malloc((z + 1)*sizeof(struct huft))) == |
| 486 | (struct huft *)NULL) |
| 487 | { |
| 488 | if (h) |
| 489 | huft_free(u[0]); |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 490 | ret = 3; /* not enough memory */ |
| 491 | goto out; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 492 | } |
| 493 | DEBG1("4 "); |
| 494 | hufts += z + 1; /* track memory usage */ |
| 495 | *t = q + 1; /* link to list for huft_free() */ |
| 496 | *(t = &(q->v.t)) = (struct huft *)NULL; |
| 497 | u[h] = ++q; /* table starts after link */ |
| 498 | |
| 499 | DEBG1("5 "); |
| 500 | /* connect to last table, if there is one */ |
| 501 | if (h) |
| 502 | { |
| 503 | x[h] = i; /* save pattern for backing up */ |
| 504 | r.b = (uch)l; /* bits to dump before this table */ |
| 505 | r.e = (uch)(16 + j); /* bits in this table */ |
| 506 | r.v.t = q; /* pointer to this table */ |
| 507 | j = i >> (w - l); /* (get around Turbo C bug) */ |
| 508 | u[h-1][j] = r; /* connect to last table */ |
| 509 | } |
| 510 | DEBG1("6 "); |
| 511 | } |
| 512 | DEBG("h6c "); |
| 513 | |
| 514 | /* set up table entry in r */ |
| 515 | r.b = (uch)(k - w); |
| 516 | if (p >= v + n) |
| 517 | r.e = 99; /* out of values--invalid code */ |
| 518 | else if (*p < s) |
| 519 | { |
| 520 | r.e = (uch)(*p < 256 ? 16 : 15); /* 256 is end-of-block code */ |
| 521 | r.v.n = (ush)(*p); /* simple code is just the value */ |
| 522 | p++; /* one compiler does not like *p++ */ |
| 523 | } |
| 524 | else |
| 525 | { |
| 526 | r.e = (uch)e[*p - s]; /* non-simple--look up in lists */ |
| 527 | r.v.n = d[*p++ - s]; |
| 528 | } |
| 529 | DEBG("h6d "); |
| 530 | |
| 531 | /* fill code-like entries with r */ |
| 532 | f = 1 << (k - w); |
| 533 | for (j = i >> w; j < z; j += f) |
| 534 | q[j] = r; |
| 535 | |
| 536 | /* backwards increment the k-bit code i */ |
| 537 | for (j = 1 << (k - 1); i & j; j >>= 1) |
| 538 | i ^= j; |
| 539 | i ^= j; |
| 540 | |
| 541 | /* backup over finished tables */ |
| 542 | while ((i & ((1 << w) - 1)) != x[h]) |
| 543 | { |
| 544 | h--; /* don't need to update q */ |
| 545 | w -= l; |
| 546 | } |
| 547 | DEBG("h6e "); |
| 548 | } |
| 549 | DEBG("h6f "); |
| 550 | } |
| 551 | |
| 552 | DEBG("huft7 "); |
| 553 | |
| 554 | /* Return true (1) if we were given an incomplete table */ |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 555 | ret = y != 0 && g != 1; |
| 556 | |
| 557 | out: |
| 558 | free(stk); |
| 559 | return ret; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 560 | } |
| 561 | |
| 562 | |
| 563 | |
| 564 | STATIC int INIT huft_free( |
| 565 | struct huft *t /* table to free */ |
| 566 | ) |
| 567 | /* Free the malloc'ed tables built by huft_build(), which makes a linked |
| 568 | list of the tables it made, with the links in a dummy first entry of |
| 569 | each table. */ |
| 570 | { |
| 571 | register struct huft *p, *q; |
| 572 | |
| 573 | |
| 574 | /* Go through linked list, freeing from the malloced (t[-1]) address. */ |
| 575 | p = t; |
| 576 | while (p != (struct huft *)NULL) |
| 577 | { |
| 578 | q = (--p)->v.t; |
| 579 | free((char*)p); |
| 580 | p = q; |
| 581 | } |
| 582 | return 0; |
| 583 | } |
| 584 | |
| 585 | |
| 586 | STATIC int INIT inflate_codes( |
| 587 | struct huft *tl, /* literal/length decoder tables */ |
| 588 | struct huft *td, /* distance decoder tables */ |
| 589 | int bl, /* number of bits decoded by tl[] */ |
| 590 | int bd /* number of bits decoded by td[] */ |
| 591 | ) |
| 592 | /* inflate (decompress) the codes in a deflated (compressed) block. |
| 593 | Return an error code or zero if it all goes ok. */ |
| 594 | { |
| 595 | register unsigned e; /* table entry flag/number of extra bits */ |
| 596 | unsigned n, d; /* length and index for copy */ |
| 597 | unsigned w; /* current window position */ |
| 598 | struct huft *t; /* pointer to table entry */ |
| 599 | unsigned ml, md; /* masks for bl and bd bits */ |
| 600 | register ulg b; /* bit buffer */ |
| 601 | register unsigned k; /* number of bits in bit buffer */ |
| 602 | |
| 603 | |
| 604 | /* make local copies of globals */ |
| 605 | b = bb; /* initialize bit buffer */ |
| 606 | k = bk; |
| 607 | w = wp; /* initialize window position */ |
| 608 | |
| 609 | /* inflate the coded data */ |
| 610 | ml = mask_bits[bl]; /* precompute masks for speed */ |
| 611 | md = mask_bits[bd]; |
| 612 | for (;;) /* do until end of block */ |
| 613 | { |
| 614 | NEEDBITS((unsigned)bl) |
| 615 | if ((e = (t = tl + ((unsigned)b & ml))->e) > 16) |
| 616 | do { |
| 617 | if (e == 99) |
| 618 | return 1; |
| 619 | DUMPBITS(t->b) |
| 620 | e -= 16; |
| 621 | NEEDBITS(e) |
| 622 | } while ((e = (t = t->v.t + ((unsigned)b & mask_bits[e]))->e) > 16); |
| 623 | DUMPBITS(t->b) |
| 624 | if (e == 16) /* then it's a literal */ |
| 625 | { |
| 626 | slide[w++] = (uch)t->v.n; |
| 627 | Tracevv((stderr, "%c", slide[w-1])); |
| 628 | if (w == WSIZE) |
| 629 | { |
| 630 | flush_output(w); |
| 631 | w = 0; |
| 632 | } |
| 633 | } |
| 634 | else /* it's an EOB or a length */ |
| 635 | { |
| 636 | /* exit if end of block */ |
| 637 | if (e == 15) |
| 638 | break; |
| 639 | |
| 640 | /* get length of block to copy */ |
| 641 | NEEDBITS(e) |
| 642 | n = t->v.n + ((unsigned)b & mask_bits[e]); |
| 643 | DUMPBITS(e); |
| 644 | |
| 645 | /* decode distance of block to copy */ |
| 646 | NEEDBITS((unsigned)bd) |
| 647 | if ((e = (t = td + ((unsigned)b & md))->e) > 16) |
| 648 | do { |
| 649 | if (e == 99) |
| 650 | return 1; |
| 651 | DUMPBITS(t->b) |
| 652 | e -= 16; |
| 653 | NEEDBITS(e) |
| 654 | } while ((e = (t = t->v.t + ((unsigned)b & mask_bits[e]))->e) > 16); |
| 655 | DUMPBITS(t->b) |
| 656 | NEEDBITS(e) |
| 657 | d = w - t->v.n - ((unsigned)b & mask_bits[e]); |
| 658 | DUMPBITS(e) |
| 659 | Tracevv((stderr,"\\[%d,%d]", w-d, n)); |
| 660 | |
| 661 | /* do the copy */ |
| 662 | do { |
| 663 | n -= (e = (e = WSIZE - ((d &= WSIZE-1) > w ? d : w)) > n ? n : e); |
| 664 | #if !defined(NOMEMCPY) && !defined(DEBUG) |
| 665 | if (w - d >= e) /* (this test assumes unsigned comparison) */ |
| 666 | { |
| 667 | memcpy(slide + w, slide + d, e); |
| 668 | w += e; |
| 669 | d += e; |
| 670 | } |
| 671 | else /* do it slow to avoid memcpy() overlap */ |
| 672 | #endif /* !NOMEMCPY */ |
| 673 | do { |
| 674 | slide[w++] = slide[d++]; |
| 675 | Tracevv((stderr, "%c", slide[w-1])); |
| 676 | } while (--e); |
| 677 | if (w == WSIZE) |
| 678 | { |
| 679 | flush_output(w); |
| 680 | w = 0; |
| 681 | } |
| 682 | } while (n); |
| 683 | } |
| 684 | } |
| 685 | |
| 686 | |
| 687 | /* restore the globals from the locals */ |
| 688 | wp = w; /* restore global window pointer */ |
| 689 | bb = b; /* restore global bit buffer */ |
| 690 | bk = k; |
| 691 | |
| 692 | /* done */ |
| 693 | return 0; |
| 694 | |
| 695 | underrun: |
| 696 | return 4; /* Input underrun */ |
| 697 | } |
| 698 | |
| 699 | |
| 700 | |
| 701 | STATIC int INIT inflate_stored(void) |
| 702 | /* "decompress" an inflated type 0 (stored) block. */ |
| 703 | { |
| 704 | unsigned n; /* number of bytes in block */ |
| 705 | unsigned w; /* current window position */ |
| 706 | register ulg b; /* bit buffer */ |
| 707 | register unsigned k; /* number of bits in bit buffer */ |
| 708 | |
| 709 | DEBG("<stor"); |
| 710 | |
| 711 | /* make local copies of globals */ |
| 712 | b = bb; /* initialize bit buffer */ |
| 713 | k = bk; |
| 714 | w = wp; /* initialize window position */ |
| 715 | |
| 716 | |
| 717 | /* go to byte boundary */ |
| 718 | n = k & 7; |
| 719 | DUMPBITS(n); |
| 720 | |
| 721 | |
| 722 | /* get the length and its complement */ |
| 723 | NEEDBITS(16) |
| 724 | n = ((unsigned)b & 0xffff); |
| 725 | DUMPBITS(16) |
| 726 | NEEDBITS(16) |
| 727 | if (n != (unsigned)((~b) & 0xffff)) |
| 728 | return 1; /* error in compressed data */ |
| 729 | DUMPBITS(16) |
| 730 | |
| 731 | |
| 732 | /* read and output the compressed data */ |
| 733 | while (n--) |
| 734 | { |
| 735 | NEEDBITS(8) |
| 736 | slide[w++] = (uch)b; |
| 737 | if (w == WSIZE) |
| 738 | { |
| 739 | flush_output(w); |
| 740 | w = 0; |
| 741 | } |
| 742 | DUMPBITS(8) |
| 743 | } |
| 744 | |
| 745 | |
| 746 | /* restore the globals from the locals */ |
| 747 | wp = w; /* restore global window pointer */ |
| 748 | bb = b; /* restore global bit buffer */ |
| 749 | bk = k; |
| 750 | |
| 751 | DEBG(">"); |
| 752 | return 0; |
| 753 | |
| 754 | underrun: |
| 755 | return 4; /* Input underrun */ |
| 756 | } |
| 757 | |
| 758 | |
| 759 | /* |
| 760 | * We use `noinline' here to prevent gcc-3.5 from using too much stack space |
| 761 | */ |
| 762 | STATIC int noinline INIT inflate_fixed(void) |
| 763 | /* decompress an inflated type 1 (fixed Huffman codes) block. We should |
| 764 | either replace this with a custom decoder, or at least precompute the |
| 765 | Huffman tables. */ |
| 766 | { |
| 767 | int i; /* temporary variable */ |
| 768 | struct huft *tl; /* literal/length code table */ |
| 769 | struct huft *td; /* distance code table */ |
| 770 | int bl; /* lookup bits for tl */ |
| 771 | int bd; /* lookup bits for td */ |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 772 | unsigned *l; /* length list for huft_build */ |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 773 | |
| 774 | DEBG("<fix"); |
| 775 | |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 776 | l = malloc(sizeof(*l) * 288); |
| 777 | if (l == NULL) |
| 778 | return 3; /* out of memory */ |
| 779 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 780 | /* set up literal table */ |
| 781 | for (i = 0; i < 144; i++) |
| 782 | l[i] = 8; |
| 783 | for (; i < 256; i++) |
| 784 | l[i] = 9; |
| 785 | for (; i < 280; i++) |
| 786 | l[i] = 7; |
| 787 | for (; i < 288; i++) /* make a complete, but wrong code set */ |
| 788 | l[i] = 8; |
| 789 | bl = 7; |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 790 | if ((i = huft_build(l, 288, 257, cplens, cplext, &tl, &bl)) != 0) { |
| 791 | free(l); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 792 | return i; |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 793 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 794 | |
| 795 | /* set up distance table */ |
| 796 | for (i = 0; i < 30; i++) /* make an incomplete code set */ |
| 797 | l[i] = 5; |
| 798 | bd = 5; |
| 799 | if ((i = huft_build(l, 30, 0, cpdist, cpdext, &td, &bd)) > 1) |
| 800 | { |
| 801 | huft_free(tl); |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 802 | free(l); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 803 | |
| 804 | DEBG(">"); |
| 805 | return i; |
| 806 | } |
| 807 | |
| 808 | |
| 809 | /* decompress until an end-of-block code */ |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 810 | if (inflate_codes(tl, td, bl, bd)) { |
| 811 | free(l); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 812 | return 1; |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 813 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 814 | |
| 815 | /* free the decoding tables, return */ |
Jeremy Fitzhardinge | 35c7422 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 816 | free(l); |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 817 | huft_free(tl); |
| 818 | huft_free(td); |
| 819 | return 0; |
| 820 | } |
| 821 | |
| 822 | |
| 823 | /* |
| 824 | * We use `noinline' here to prevent gcc-3.5 from using too much stack space |
| 825 | */ |
| 826 | STATIC int noinline INIT inflate_dynamic(void) |
| 827 | /* decompress an inflated type 2 (dynamic Huffman codes) block. */ |
| 828 | { |
| 829 | int i; /* temporary variables */ |
| 830 | unsigned j; |
| 831 | unsigned l; /* last length */ |
| 832 | unsigned m; /* mask for bit lengths table */ |
| 833 | unsigned n; /* number of lengths to get */ |
| 834 | struct huft *tl; /* literal/length code table */ |
| 835 | struct huft *td; /* distance code table */ |
| 836 | int bl; /* lookup bits for tl */ |
| 837 | int bd; /* lookup bits for td */ |
| 838 | unsigned nb; /* number of bit length codes */ |
| 839 | unsigned nl; /* number of literal/length codes */ |
| 840 | unsigned nd; /* number of distance codes */ |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 841 | unsigned *ll; /* literal/length and distance code lengths */ |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 842 | register ulg b; /* bit buffer */ |
| 843 | register unsigned k; /* number of bits in bit buffer */ |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 844 | int ret; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 845 | |
| 846 | DEBG("<dyn"); |
| 847 | |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 848 | #ifdef PKZIP_BUG_WORKAROUND |
| 849 | ll = malloc(sizeof(*ll) * (288+32)); /* literal/length and distance code lengths */ |
| 850 | #else |
| 851 | ll = malloc(sizeof(*ll) * (286+30)); /* literal/length and distance code lengths */ |
| 852 | #endif |
| 853 | |
Jim Meyering | 22caa04 | 2008-04-29 00:59:09 -0700 | [diff] [blame] | 854 | if (ll == NULL) |
| 855 | return 1; |
| 856 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 857 | /* make local bit buffer */ |
| 858 | b = bb; |
| 859 | k = bk; |
| 860 | |
| 861 | |
| 862 | /* read in table lengths */ |
| 863 | NEEDBITS(5) |
| 864 | nl = 257 + ((unsigned)b & 0x1f); /* number of literal/length codes */ |
| 865 | DUMPBITS(5) |
| 866 | NEEDBITS(5) |
| 867 | nd = 1 + ((unsigned)b & 0x1f); /* number of distance codes */ |
| 868 | DUMPBITS(5) |
| 869 | NEEDBITS(4) |
| 870 | nb = 4 + ((unsigned)b & 0xf); /* number of bit length codes */ |
| 871 | DUMPBITS(4) |
| 872 | #ifdef PKZIP_BUG_WORKAROUND |
| 873 | if (nl > 288 || nd > 32) |
| 874 | #else |
| 875 | if (nl > 286 || nd > 30) |
| 876 | #endif |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 877 | { |
| 878 | ret = 1; /* bad lengths */ |
| 879 | goto out; |
| 880 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 881 | |
| 882 | DEBG("dyn1 "); |
| 883 | |
| 884 | /* read in bit-length-code lengths */ |
| 885 | for (j = 0; j < nb; j++) |
| 886 | { |
| 887 | NEEDBITS(3) |
| 888 | ll[border[j]] = (unsigned)b & 7; |
| 889 | DUMPBITS(3) |
| 890 | } |
| 891 | for (; j < 19; j++) |
| 892 | ll[border[j]] = 0; |
| 893 | |
| 894 | DEBG("dyn2 "); |
| 895 | |
| 896 | /* build decoding table for trees--single level, 7 bit lookup */ |
| 897 | bl = 7; |
| 898 | if ((i = huft_build(ll, 19, 19, NULL, NULL, &tl, &bl)) != 0) |
| 899 | { |
| 900 | if (i == 1) |
| 901 | huft_free(tl); |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 902 | ret = i; /* incomplete code set */ |
| 903 | goto out; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 904 | } |
| 905 | |
| 906 | DEBG("dyn3 "); |
| 907 | |
| 908 | /* read in literal and distance code lengths */ |
| 909 | n = nl + nd; |
| 910 | m = mask_bits[bl]; |
| 911 | i = l = 0; |
| 912 | while ((unsigned)i < n) |
| 913 | { |
| 914 | NEEDBITS((unsigned)bl) |
| 915 | j = (td = tl + ((unsigned)b & m))->b; |
| 916 | DUMPBITS(j) |
| 917 | j = td->v.n; |
| 918 | if (j < 16) /* length of code in bits (0..15) */ |
| 919 | ll[i++] = l = j; /* save last length in l */ |
| 920 | else if (j == 16) /* repeat last length 3 to 6 times */ |
| 921 | { |
| 922 | NEEDBITS(2) |
| 923 | j = 3 + ((unsigned)b & 3); |
| 924 | DUMPBITS(2) |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 925 | if ((unsigned)i + j > n) { |
| 926 | ret = 1; |
| 927 | goto out; |
| 928 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 929 | while (j--) |
| 930 | ll[i++] = l; |
| 931 | } |
| 932 | else if (j == 17) /* 3 to 10 zero length codes */ |
| 933 | { |
| 934 | NEEDBITS(3) |
| 935 | j = 3 + ((unsigned)b & 7); |
| 936 | DUMPBITS(3) |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 937 | if ((unsigned)i + j > n) { |
| 938 | ret = 1; |
| 939 | goto out; |
| 940 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 941 | while (j--) |
| 942 | ll[i++] = 0; |
| 943 | l = 0; |
| 944 | } |
| 945 | else /* j == 18: 11 to 138 zero length codes */ |
| 946 | { |
| 947 | NEEDBITS(7) |
| 948 | j = 11 + ((unsigned)b & 0x7f); |
| 949 | DUMPBITS(7) |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 950 | if ((unsigned)i + j > n) { |
| 951 | ret = 1; |
| 952 | goto out; |
| 953 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 954 | while (j--) |
| 955 | ll[i++] = 0; |
| 956 | l = 0; |
| 957 | } |
| 958 | } |
| 959 | |
| 960 | DEBG("dyn4 "); |
| 961 | |
| 962 | /* free decoding table for trees */ |
| 963 | huft_free(tl); |
| 964 | |
| 965 | DEBG("dyn5 "); |
| 966 | |
| 967 | /* restore the global bit buffer */ |
| 968 | bb = b; |
| 969 | bk = k; |
| 970 | |
| 971 | DEBG("dyn5a "); |
| 972 | |
| 973 | /* build the decoding tables for literal/length and distance codes */ |
| 974 | bl = lbits; |
| 975 | if ((i = huft_build(ll, nl, 257, cplens, cplext, &tl, &bl)) != 0) |
| 976 | { |
| 977 | DEBG("dyn5b "); |
| 978 | if (i == 1) { |
| 979 | error("incomplete literal tree"); |
| 980 | huft_free(tl); |
| 981 | } |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 982 | ret = i; /* incomplete code set */ |
| 983 | goto out; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 984 | } |
| 985 | DEBG("dyn5c "); |
| 986 | bd = dbits; |
| 987 | if ((i = huft_build(ll + nl, nd, 0, cpdist, cpdext, &td, &bd)) != 0) |
| 988 | { |
| 989 | DEBG("dyn5d "); |
| 990 | if (i == 1) { |
| 991 | error("incomplete distance tree"); |
| 992 | #ifdef PKZIP_BUG_WORKAROUND |
| 993 | i = 0; |
| 994 | } |
| 995 | #else |
| 996 | huft_free(td); |
| 997 | } |
| 998 | huft_free(tl); |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 999 | ret = i; /* incomplete code set */ |
| 1000 | goto out; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1001 | #endif |
| 1002 | } |
| 1003 | |
| 1004 | DEBG("dyn6 "); |
| 1005 | |
| 1006 | /* decompress until an end-of-block code */ |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 1007 | if (inflate_codes(tl, td, bl, bd)) { |
| 1008 | ret = 1; |
| 1009 | goto out; |
| 1010 | } |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1011 | |
| 1012 | DEBG("dyn7 "); |
| 1013 | |
| 1014 | /* free the decoding tables, return */ |
| 1015 | huft_free(tl); |
| 1016 | huft_free(td); |
| 1017 | |
| 1018 | DEBG(">"); |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 1019 | ret = 0; |
| 1020 | out: |
| 1021 | free(ll); |
| 1022 | return ret; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1023 | |
Jeremy Fitzhardinge | 39b7ee0 | 2007-05-02 19:27:15 +0200 | [diff] [blame] | 1024 | underrun: |
| 1025 | ret = 4; /* Input underrun */ |
| 1026 | goto out; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1027 | } |
| 1028 | |
| 1029 | |
| 1030 | |
| 1031 | STATIC int INIT inflate_block( |
| 1032 | int *e /* last block flag */ |
| 1033 | ) |
| 1034 | /* decompress an inflated block */ |
| 1035 | { |
| 1036 | unsigned t; /* block type */ |
| 1037 | register ulg b; /* bit buffer */ |
| 1038 | register unsigned k; /* number of bits in bit buffer */ |
| 1039 | |
| 1040 | DEBG("<blk"); |
| 1041 | |
| 1042 | /* make local bit buffer */ |
| 1043 | b = bb; |
| 1044 | k = bk; |
| 1045 | |
| 1046 | |
| 1047 | /* read in last block bit */ |
| 1048 | NEEDBITS(1) |
| 1049 | *e = (int)b & 1; |
| 1050 | DUMPBITS(1) |
| 1051 | |
| 1052 | |
| 1053 | /* read in block type */ |
| 1054 | NEEDBITS(2) |
| 1055 | t = (unsigned)b & 3; |
| 1056 | DUMPBITS(2) |
| 1057 | |
| 1058 | |
| 1059 | /* restore the global bit buffer */ |
| 1060 | bb = b; |
| 1061 | bk = k; |
| 1062 | |
| 1063 | /* inflate that block type */ |
| 1064 | if (t == 2) |
| 1065 | return inflate_dynamic(); |
| 1066 | if (t == 0) |
| 1067 | return inflate_stored(); |
| 1068 | if (t == 1) |
| 1069 | return inflate_fixed(); |
| 1070 | |
| 1071 | DEBG(">"); |
| 1072 | |
| 1073 | /* bad block type */ |
| 1074 | return 2; |
| 1075 | |
| 1076 | underrun: |
| 1077 | return 4; /* Input underrun */ |
| 1078 | } |
| 1079 | |
| 1080 | |
| 1081 | |
| 1082 | STATIC int INIT inflate(void) |
| 1083 | /* decompress an inflated entry */ |
| 1084 | { |
| 1085 | int e; /* last block flag */ |
| 1086 | int r; /* result code */ |
| 1087 | unsigned h; /* maximum struct huft's malloc'ed */ |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1088 | |
| 1089 | /* initialize window, bit buffer */ |
| 1090 | wp = 0; |
| 1091 | bk = 0; |
| 1092 | bb = 0; |
| 1093 | |
| 1094 | |
| 1095 | /* decompress until the last block */ |
| 1096 | h = 0; |
| 1097 | do { |
| 1098 | hufts = 0; |
Thomas Petazzoni | 2d6ffcc | 2008-07-25 01:45:44 -0700 | [diff] [blame] | 1099 | #ifdef ARCH_HAS_DECOMP_WDOG |
| 1100 | arch_decomp_wdog(); |
| 1101 | #endif |
| 1102 | r = inflate_block(&e); |
| 1103 | if (r) |
| 1104 | return r; |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1105 | if (hufts > h) |
| 1106 | h = hufts; |
| 1107 | } while (!e); |
| 1108 | |
| 1109 | /* Undo too much lookahead. The next read will be byte aligned so we |
| 1110 | * can discard unused bits in the last meaningful byte. |
| 1111 | */ |
| 1112 | while (bk >= 8) { |
| 1113 | bk -= 8; |
| 1114 | inptr--; |
| 1115 | } |
| 1116 | |
| 1117 | /* flush out slide */ |
| 1118 | flush_output(wp); |
| 1119 | |
| 1120 | |
| 1121 | /* return success */ |
| 1122 | #ifdef DEBUG |
| 1123 | fprintf(stderr, "<%u> ", h); |
| 1124 | #endif /* DEBUG */ |
| 1125 | return 0; |
| 1126 | } |
| 1127 | |
| 1128 | /********************************************************************** |
| 1129 | * |
| 1130 | * The following are support routines for inflate.c |
| 1131 | * |
| 1132 | **********************************************************************/ |
| 1133 | |
| 1134 | static ulg crc_32_tab[256]; |
| 1135 | static ulg crc; /* initialized in makecrc() so it'll reside in bss */ |
| 1136 | #define CRC_VALUE (crc ^ 0xffffffffUL) |
| 1137 | |
| 1138 | /* |
| 1139 | * Code to compute the CRC-32 table. Borrowed from |
| 1140 | * gzip-1.0.3/makecrc.c. |
| 1141 | */ |
| 1142 | |
| 1143 | static void INIT |
| 1144 | makecrc(void) |
| 1145 | { |
| 1146 | /* Not copyrighted 1990 Mark Adler */ |
| 1147 | |
| 1148 | unsigned long c; /* crc shift register */ |
| 1149 | unsigned long e; /* polynomial exclusive-or pattern */ |
| 1150 | int i; /* counter for all possible eight bit values */ |
| 1151 | int k; /* byte being shifted into crc apparatus */ |
| 1152 | |
| 1153 | /* terms of polynomial defining this crc (except x^32): */ |
| 1154 | static const int p[] = {0,1,2,4,5,7,8,10,11,12,16,22,23,26}; |
| 1155 | |
| 1156 | /* Make exclusive-or pattern from polynomial */ |
| 1157 | e = 0; |
| 1158 | for (i = 0; i < sizeof(p)/sizeof(int); i++) |
| 1159 | e |= 1L << (31 - p[i]); |
| 1160 | |
| 1161 | crc_32_tab[0] = 0; |
| 1162 | |
| 1163 | for (i = 1; i < 256; i++) |
| 1164 | { |
| 1165 | c = 0; |
| 1166 | for (k = i | 256; k != 1; k >>= 1) |
| 1167 | { |
| 1168 | c = c & 1 ? (c >> 1) ^ e : c >> 1; |
| 1169 | if (k & 1) |
| 1170 | c ^= e; |
| 1171 | } |
| 1172 | crc_32_tab[i] = c; |
| 1173 | } |
| 1174 | |
| 1175 | /* this is initialized here so this code could reside in ROM */ |
| 1176 | crc = (ulg)0xffffffffUL; /* shift register contents */ |
| 1177 | } |
| 1178 | |
| 1179 | /* gzip flag byte */ |
| 1180 | #define ASCII_FLAG 0x01 /* bit 0 set: file probably ASCII text */ |
| 1181 | #define CONTINUATION 0x02 /* bit 1 set: continuation of multi-part gzip file */ |
| 1182 | #define EXTRA_FIELD 0x04 /* bit 2 set: extra field present */ |
| 1183 | #define ORIG_NAME 0x08 /* bit 3 set: original file name present */ |
| 1184 | #define COMMENT 0x10 /* bit 4 set: file comment present */ |
| 1185 | #define ENCRYPTED 0x20 /* bit 5 set: file is encrypted */ |
| 1186 | #define RESERVED 0xC0 /* bit 6,7: reserved */ |
| 1187 | |
| 1188 | /* |
| 1189 | * Do the uncompression! |
| 1190 | */ |
| 1191 | static int INIT gunzip(void) |
| 1192 | { |
| 1193 | uch flags; |
| 1194 | unsigned char magic[2]; /* magic header */ |
| 1195 | char method; |
| 1196 | ulg orig_crc = 0; /* original crc */ |
| 1197 | ulg orig_len = 0; /* original uncompressed length */ |
| 1198 | int res; |
| 1199 | |
| 1200 | magic[0] = NEXTBYTE(); |
| 1201 | magic[1] = NEXTBYTE(); |
| 1202 | method = NEXTBYTE(); |
| 1203 | |
| 1204 | if (magic[0] != 037 || |
| 1205 | ((magic[1] != 0213) && (magic[1] != 0236))) { |
| 1206 | error("bad gzip magic numbers"); |
| 1207 | return -1; |
| 1208 | } |
| 1209 | |
| 1210 | /* We only support method #8, DEFLATED */ |
| 1211 | if (method != 8) { |
| 1212 | error("internal error, invalid method"); |
| 1213 | return -1; |
| 1214 | } |
| 1215 | |
| 1216 | flags = (uch)get_byte(); |
| 1217 | if ((flags & ENCRYPTED) != 0) { |
| 1218 | error("Input is encrypted"); |
| 1219 | return -1; |
| 1220 | } |
| 1221 | if ((flags & CONTINUATION) != 0) { |
| 1222 | error("Multi part input"); |
| 1223 | return -1; |
| 1224 | } |
| 1225 | if ((flags & RESERVED) != 0) { |
| 1226 | error("Input has invalid flags"); |
| 1227 | return -1; |
| 1228 | } |
| 1229 | NEXTBYTE(); /* Get timestamp */ |
| 1230 | NEXTBYTE(); |
| 1231 | NEXTBYTE(); |
| 1232 | NEXTBYTE(); |
| 1233 | |
| 1234 | (void)NEXTBYTE(); /* Ignore extra flags for the moment */ |
| 1235 | (void)NEXTBYTE(); /* Ignore OS type for the moment */ |
| 1236 | |
| 1237 | if ((flags & EXTRA_FIELD) != 0) { |
| 1238 | unsigned len = (unsigned)NEXTBYTE(); |
| 1239 | len |= ((unsigned)NEXTBYTE())<<8; |
| 1240 | while (len--) (void)NEXTBYTE(); |
| 1241 | } |
| 1242 | |
| 1243 | /* Get original file name if it was truncated */ |
| 1244 | if ((flags & ORIG_NAME) != 0) { |
| 1245 | /* Discard the old name */ |
| 1246 | while (NEXTBYTE() != 0) /* null */ ; |
| 1247 | } |
| 1248 | |
| 1249 | /* Discard file comment if any */ |
| 1250 | if ((flags & COMMENT) != 0) { |
| 1251 | while (NEXTBYTE() != 0) /* null */ ; |
| 1252 | } |
| 1253 | |
| 1254 | /* Decompress */ |
| 1255 | if ((res = inflate())) { |
| 1256 | switch (res) { |
| 1257 | case 0: |
| 1258 | break; |
| 1259 | case 1: |
| 1260 | error("invalid compressed format (err=1)"); |
| 1261 | break; |
| 1262 | case 2: |
| 1263 | error("invalid compressed format (err=2)"); |
| 1264 | break; |
| 1265 | case 3: |
| 1266 | error("out of memory"); |
| 1267 | break; |
| 1268 | case 4: |
| 1269 | error("out of input data"); |
| 1270 | break; |
| 1271 | default: |
| 1272 | error("invalid compressed format (other)"); |
| 1273 | } |
| 1274 | return -1; |
| 1275 | } |
| 1276 | |
| 1277 | /* Get the crc and original length */ |
| 1278 | /* crc32 (see algorithm.doc) |
| 1279 | * uncompressed input size modulo 2^32 |
| 1280 | */ |
| 1281 | orig_crc = (ulg) NEXTBYTE(); |
| 1282 | orig_crc |= (ulg) NEXTBYTE() << 8; |
| 1283 | orig_crc |= (ulg) NEXTBYTE() << 16; |
| 1284 | orig_crc |= (ulg) NEXTBYTE() << 24; |
| 1285 | |
| 1286 | orig_len = (ulg) NEXTBYTE(); |
| 1287 | orig_len |= (ulg) NEXTBYTE() << 8; |
| 1288 | orig_len |= (ulg) NEXTBYTE() << 16; |
| 1289 | orig_len |= (ulg) NEXTBYTE() << 24; |
| 1290 | |
| 1291 | /* Validate decompression */ |
| 1292 | if (orig_crc != CRC_VALUE) { |
| 1293 | error("crc error"); |
| 1294 | return -1; |
| 1295 | } |
| 1296 | if (orig_len != bytes_out) { |
| 1297 | error("length error"); |
| 1298 | return -1; |
| 1299 | } |
| 1300 | return 0; |
| 1301 | |
| 1302 | underrun: /* NEXTBYTE() goto's here if needed */ |
| 1303 | error("out of input data"); |
| 1304 | return -1; |
| 1305 | } |
| 1306 | |
| 1307 | |