[3320] | 1 | /*
|
---|
| 2 | * XZ decompressor
|
---|
| 3 | *
|
---|
| 4 | * Authors: Lasse Collin <lasse.collin@tukaani.org>
|
---|
| 5 | * Igor Pavlov <http://7-zip.org/>
|
---|
| 6 | *
|
---|
| 7 | * This file has been put into the public domain.
|
---|
| 8 | * You can do whatever you want with this file.
|
---|
| 9 | */
|
---|
| 10 |
|
---|
| 11 | #ifndef XZ_H
|
---|
| 12 | #define XZ_H
|
---|
| 13 |
|
---|
| 14 | #ifdef __KERNEL__
|
---|
| 15 | # include <linux/stddef.h>
|
---|
| 16 | # include <linux/types.h>
|
---|
| 17 | #else
|
---|
| 18 | # include <stddef.h>
|
---|
| 19 | # include <stdint.h>
|
---|
| 20 | #endif
|
---|
| 21 |
|
---|
| 22 | /* In Linux, this is used to make extern functions static when needed. */
|
---|
| 23 | #ifndef XZ_EXTERN
|
---|
| 24 | # define XZ_EXTERN extern
|
---|
| 25 | #endif
|
---|
| 26 |
|
---|
| 27 | /* In Linux, this is used to mark the functions with __init when needed. */
|
---|
| 28 | #ifndef XZ_FUNC
|
---|
| 29 | # define XZ_FUNC
|
---|
| 30 | #endif
|
---|
| 31 |
|
---|
| 32 | /**
|
---|
| 33 | * enum xz_mode - Operation mode
|
---|
| 34 | *
|
---|
| 35 | * @XZ_SINGLE: Single-call mode. This uses less RAM than
|
---|
| 36 | * than multi-call modes, because the LZMA2
|
---|
| 37 | * dictionary doesn't need to be allocated as
|
---|
| 38 | * part of the decoder state. All required data
|
---|
| 39 | * structures are allocated at initialization,
|
---|
| 40 | * so xz_dec_run() cannot return XZ_MEM_ERROR.
|
---|
| 41 | * @XZ_PREALLOC: Multi-call mode with preallocated LZMA2
|
---|
| 42 | * dictionary buffer. All data structures are
|
---|
| 43 | * allocated at initialization, so xz_dec_run()
|
---|
| 44 | * cannot return XZ_MEM_ERROR.
|
---|
| 45 | * @XZ_DYNALLOC: Multi-call mode. The LZMA2 dictionary is
|
---|
| 46 | * allocated once the required size has been
|
---|
| 47 | * parsed from the stream headers. If the
|
---|
| 48 | * allocation fails, xz_dec_run() will return
|
---|
| 49 | * XZ_MEM_ERROR.
|
---|
| 50 | *
|
---|
| 51 | * It is possible to enable support only for a subset of the above
|
---|
| 52 | * modes at compile time by defining XZ_DEC_SINGLE, XZ_DEC_PREALLOC,
|
---|
| 53 | * or XZ_DEC_DYNALLOC. The xz_dec kernel module is always compiled
|
---|
| 54 | * with support for all operation modes, but the preboot code may
|
---|
| 55 | * be built with fewer features to minimize code size.
|
---|
| 56 | */
|
---|
| 57 | enum xz_mode {
|
---|
| 58 | XZ_SINGLE,
|
---|
| 59 | XZ_PREALLOC,
|
---|
| 60 | XZ_DYNALLOC
|
---|
| 61 | };
|
---|
| 62 |
|
---|
| 63 | /**
|
---|
| 64 | * enum xz_ret - Return codes
|
---|
| 65 | * @XZ_OK: Everything is OK so far. More input or more
|
---|
| 66 | * output space is required to continue. This
|
---|
| 67 | * return code is possible only in multi-call mode
|
---|
| 68 | * (XZ_PREALLOC or XZ_DYNALLOC).
|
---|
| 69 | * @XZ_STREAM_END: Operation finished successfully.
|
---|
| 70 | * @XZ_UNSUPPORTED_CHECK: Integrity check type is not supported. Decoding
|
---|
| 71 | * is still possible in multi-call mode by simply
|
---|
| 72 | * calling xz_dec_run() again.
|
---|
| 73 | * NOTE: This return value is used only if
|
---|
| 74 | * XZ_DEC_ANY_CHECK was defined at build time,
|
---|
| 75 | * which is not used in the kernel. Unsupported
|
---|
| 76 | * check types return XZ_OPTIONS_ERROR if
|
---|
| 77 | * XZ_DEC_ANY_CHECK was not defined at build time.
|
---|
| 78 | * @XZ_MEM_ERROR: Allocating memory failed. This return code is
|
---|
| 79 | * possible only if the decoder was initialized
|
---|
| 80 | * with XZ_DYNALLOC. The amount of memory that was
|
---|
| 81 | * tried to be allocated was no more than the
|
---|
| 82 | * dict_max argument given to xz_dec_init().
|
---|
| 83 | * @XZ_MEMLIMIT_ERROR: A bigger LZMA2 dictionary would be needed than
|
---|
| 84 | * allowed by the dict_max argument given to
|
---|
| 85 | * xz_dec_init(). This return value is possible
|
---|
| 86 | * only in multi-call mode (XZ_PREALLOC or
|
---|
| 87 | * XZ_DYNALLOC); the single-call mode (XZ_SINGLE)
|
---|
| 88 | * ignores the dict_max argument.
|
---|
| 89 | * @XZ_FORMAT_ERROR: File format was not recognized (wrong magic
|
---|
| 90 | * bytes).
|
---|
| 91 | * @XZ_OPTIONS_ERROR: This implementation doesn't support the requested
|
---|
| 92 | * compression options. In the decoder this means
|
---|
| 93 | * that the header CRC32 matches, but the header
|
---|
| 94 | * itself specifies something that we don't support.
|
---|
| 95 | * @XZ_DATA_ERROR: Compressed data is corrupt.
|
---|
| 96 | * @XZ_BUF_ERROR: Cannot make any progress. Details are slightly
|
---|
| 97 | * different between multi-call and single-call
|
---|
| 98 | * mode; more information below.
|
---|
| 99 | *
|
---|
| 100 | * In multi-call mode, XZ_BUF_ERROR is returned when two consecutive calls
|
---|
| 101 | * to XZ code cannot consume any input and cannot produce any new output.
|
---|
| 102 | * This happens when there is no new input available, or the output buffer
|
---|
| 103 | * is full while at least one output byte is still pending. Assuming your
|
---|
| 104 | * code is not buggy, you can get this error only when decoding a compressed
|
---|
| 105 | * stream that is truncated or otherwise corrupt.
|
---|
| 106 | *
|
---|
| 107 | * In single-call mode, XZ_BUF_ERROR is returned only when the output buffer
|
---|
| 108 | * is too small, or the compressed input is corrupt in a way that makes the
|
---|
| 109 | * decoder produce more output than the caller expected. When it is
|
---|
| 110 | * (relatively) clear that the compressed input is truncated, XZ_DATA_ERROR
|
---|
| 111 | * is used instead of XZ_BUF_ERROR.
|
---|
| 112 | */
|
---|
| 113 | enum xz_ret {
|
---|
| 114 | XZ_OK,
|
---|
| 115 | XZ_STREAM_END,
|
---|
| 116 | XZ_UNSUPPORTED_CHECK,
|
---|
| 117 | XZ_MEM_ERROR,
|
---|
| 118 | XZ_MEMLIMIT_ERROR,
|
---|
| 119 | XZ_FORMAT_ERROR,
|
---|
| 120 | XZ_OPTIONS_ERROR,
|
---|
| 121 | XZ_DATA_ERROR,
|
---|
| 122 | XZ_BUF_ERROR
|
---|
| 123 | };
|
---|
| 124 |
|
---|
| 125 | /**
|
---|
| 126 | * struct xz_buf - Passing input and output buffers to XZ code
|
---|
| 127 | * @in: Beginning of the input buffer. This may be NULL if and only
|
---|
| 128 | * if in_pos is equal to in_size.
|
---|
| 129 | * @in_pos: Current position in the input buffer. This must not exceed
|
---|
| 130 | * in_size.
|
---|
| 131 | * @in_size: Size of the input buffer
|
---|
| 132 | * @out: Beginning of the output buffer. This may be NULL if and only
|
---|
| 133 | * if out_pos is equal to out_size.
|
---|
| 134 | * @out_pos: Current position in the output buffer. This must not exceed
|
---|
| 135 | * out_size.
|
---|
| 136 | * @out_size: Size of the output buffer
|
---|
| 137 | *
|
---|
| 138 | * Only the contents of the output buffer from out[out_pos] onward, and
|
---|
| 139 | * the variables in_pos and out_pos are modified by the XZ code.
|
---|
| 140 | */
|
---|
| 141 | struct xz_buf {
|
---|
| 142 | const uint8_t *in;
|
---|
| 143 | size_t in_pos;
|
---|
| 144 | size_t in_size;
|
---|
| 145 |
|
---|
| 146 | uint8_t *out;
|
---|
| 147 | size_t out_pos;
|
---|
| 148 | size_t out_size;
|
---|
| 149 | };
|
---|
| 150 |
|
---|
| 151 | /**
|
---|
| 152 | * struct xz_dec - Opaque type to hold the XZ decoder state
|
---|
| 153 | */
|
---|
| 154 | struct xz_dec;
|
---|
| 155 |
|
---|
| 156 | /**
|
---|
| 157 | * xz_dec_init() - Allocate and initialize a XZ decoder state
|
---|
| 158 | * @mode: Operation mode
|
---|
| 159 | * @dict_max: Maximum size of the LZMA2 dictionary (history buffer) for
|
---|
| 160 | * multi-call decoding. This is ignored in single-call mode
|
---|
| 161 | * (mode == XZ_SINGLE). LZMA2 dictionary is always 2^n bytes
|
---|
| 162 | * or 2^n + 2^(n-1) bytes (the latter sizes are less common
|
---|
| 163 | * in practice), so other values for dict_max don't make sense.
|
---|
| 164 | * In the kernel, dictionary sizes of 64 KiB, 128 KiB, 256 KiB,
|
---|
| 165 | * 512 KiB, and 1 MiB are probably the only reasonable values,
|
---|
| 166 | * except for kernel and initramfs images where a bigger
|
---|
| 167 | * dictionary can be fine and useful.
|
---|
| 168 | *
|
---|
| 169 | * Single-call mode (XZ_SINGLE): xz_dec_run() decodes the whole stream at
|
---|
| 170 | * once. The caller must provide enough output space or the decoding will
|
---|
| 171 | * fail. The output space is used as the dictionary buffer, which is why
|
---|
| 172 | * there is no need to allocate the dictionary as part of the decoder's
|
---|
| 173 | * internal state.
|
---|
| 174 | *
|
---|
| 175 | * Because the output buffer is used as the workspace, streams encoded using
|
---|
| 176 | * a big dictionary are not a problem in single-call mode. It is enough that
|
---|
| 177 | * the output buffer is big enough to hold the actual uncompressed data; it
|
---|
| 178 | * can be smaller than the dictionary size stored in the stream headers.
|
---|
| 179 | *
|
---|
| 180 | * Multi-call mode with preallocated dictionary (XZ_PREALLOC): dict_max bytes
|
---|
| 181 | * of memory is preallocated for the LZMA2 dictionary. This way there is no
|
---|
| 182 | * risk that xz_dec_run() could run out of memory, since xz_dec_run() will
|
---|
| 183 | * never allocate any memory. Instead, if the preallocated dictionary is too
|
---|
| 184 | * small for decoding the given input stream, xz_dec_run() will return
|
---|
| 185 | * XZ_MEMLIMIT_ERROR. Thus, it is important to know what kind of data will be
|
---|
| 186 | * decoded to avoid allocating excessive amount of memory for the dictionary.
|
---|
| 187 | *
|
---|
| 188 | * Multi-call mode with dynamically allocated dictionary (XZ_DYNALLOC):
|
---|
| 189 | * dict_max specifies the maximum allowed dictionary size that xz_dec_run()
|
---|
| 190 | * may allocate once it has parsed the dictionary size from the stream
|
---|
| 191 | * headers. This way excessive allocations can be avoided while still
|
---|
| 192 | * limiting the maximum memory usage to a sane value to prevent running the
|
---|
| 193 | * system out of memory when decompressing streams from untrusted sources.
|
---|
| 194 | *
|
---|
| 195 | * On success, xz_dec_init() returns a pointer to struct xz_dec, which is
|
---|
| 196 | * ready to be used with xz_dec_run(). If memory allocation fails,
|
---|
| 197 | * xz_dec_init() returns NULL.
|
---|
| 198 | */
|
---|
| 199 | XZ_EXTERN struct xz_dec * XZ_FUNC xz_dec_init(
|
---|
| 200 | enum xz_mode mode, uint32_t dict_max);
|
---|
| 201 |
|
---|
| 202 | /**
|
---|
| 203 | * xz_dec_run() - Run the XZ decoder
|
---|
| 204 | * @s: Decoder state allocated using xz_dec_init()
|
---|
| 205 | * @b: Input and output buffers
|
---|
| 206 | *
|
---|
| 207 | * The possible return values depend on build options and operation mode.
|
---|
| 208 | * See enum xz_ret for details.
|
---|
| 209 | *
|
---|
| 210 | * NOTE: If an error occurs in single-call mode (return value is not
|
---|
| 211 | * XZ_STREAM_END), b->in_pos and b->out_pos are not modified, and the
|
---|
| 212 | * contents of the output buffer from b->out[b->out_pos] onward are
|
---|
| 213 | * undefined. This is true even after XZ_BUF_ERROR, because with some filter
|
---|
| 214 | * chains, there may be a second pass over the output buffer, and this pass
|
---|
| 215 | * cannot be properly done if the output buffer is truncated. Thus, you
|
---|
| 216 | * cannot give the single-call decoder a too small buffer and then expect to
|
---|
| 217 | * get that amount valid data from the beginning of the stream. You must use
|
---|
| 218 | * the multi-call decoder if you don't want to uncompress the whole stream.
|
---|
| 219 | */
|
---|
| 220 | XZ_EXTERN enum xz_ret XZ_FUNC xz_dec_run(struct xz_dec *s, struct xz_buf *b);
|
---|
| 221 |
|
---|
| 222 | /**
|
---|
| 223 | * xz_dec_reset() - Reset an already allocated decoder state
|
---|
| 224 | * @s: Decoder state allocated using xz_dec_init()
|
---|
| 225 | *
|
---|
| 226 | * This function can be used to reset the multi-call decoder state without
|
---|
| 227 | * freeing and reallocating memory with xz_dec_end() and xz_dec_init().
|
---|
| 228 | *
|
---|
| 229 | * In single-call mode, xz_dec_reset() is always called in the beginning of
|
---|
| 230 | * xz_dec_run(). Thus, explicit call to xz_dec_reset() is useful only in
|
---|
| 231 | * multi-call mode.
|
---|
| 232 | */
|
---|
| 233 | XZ_EXTERN void XZ_FUNC xz_dec_reset(struct xz_dec *s);
|
---|
| 234 |
|
---|
| 235 | /**
|
---|
| 236 | * xz_dec_end() - Free the memory allocated for the decoder state
|
---|
| 237 | * @s: Decoder state allocated using xz_dec_init(). If s is NULL,
|
---|
| 238 | * this function does nothing.
|
---|
| 239 | */
|
---|
| 240 | XZ_EXTERN void XZ_FUNC xz_dec_end(struct xz_dec *s);
|
---|
| 241 |
|
---|
| 242 | /*
|
---|
| 243 | * Standalone build (userspace build or in-kernel build for boot time use)
|
---|
| 244 | * needs a CRC32 implementation. For normal in-kernel use, kernel's own
|
---|
| 245 | * CRC32 module is used instead, and users of this module don't need to
|
---|
| 246 | * care about the functions below.
|
---|
| 247 | */
|
---|
| 248 | #ifndef XZ_INTERNAL_CRC32
|
---|
| 249 | # ifdef __KERNEL__
|
---|
| 250 | # define XZ_INTERNAL_CRC32 0
|
---|
| 251 | # else
|
---|
| 252 | # define XZ_INTERNAL_CRC32 1
|
---|
| 253 | # endif
|
---|
| 254 | #endif
|
---|
| 255 |
|
---|
| 256 | #if XZ_INTERNAL_CRC32
|
---|
| 257 | /*
|
---|
| 258 | * This must be called before any other xz_* function to initialize
|
---|
| 259 | * the CRC32 lookup table.
|
---|
| 260 | */
|
---|
| 261 | XZ_EXTERN void XZ_FUNC xz_crc32_init(void);
|
---|
| 262 |
|
---|
| 263 | /*
|
---|
| 264 | * Update CRC32 value using the polynomial from IEEE-802.3. To start a new
|
---|
| 265 | * calculation, the third argument must be zero. To continue the calculation,
|
---|
| 266 | * the previously returned value is passed as the third argument.
|
---|
| 267 | */
|
---|
| 268 | XZ_EXTERN uint32_t XZ_FUNC xz_crc32(
|
---|
| 269 | const uint8_t *buf, size_t size, uint32_t crc);
|
---|
| 270 | #endif
|
---|
| 271 | #endif
|
---|