diff --git a/CHANGELOG.md b/CHANGELOG.md --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,5 +1,9 @@ # lz4 +## 0.1.5.2 + + * Upgrade bundled sources to 1.9.4 + ## 0.1.5.1 * Upgrade to lz4 1.9.3 under the hood diff --git a/README.md b/README.md --- a/README.md +++ b/README.md @@ -1,7 +1,7 @@ # lz4-hs Haskell bindings to [lz4](https://github.com/lz4/lz4). Includes a bundled copy -of lz4 1.9.2 +of lz4 1.9.3 ## Comparison diff --git a/bench/Bench.hs b/bench/Bench.hs --- a/bench/Bench.hs +++ b/bench/Bench.hs @@ -8,19 +8,16 @@ decompressDump :: FilePath -> FilePath -> IO () decompressDump dir fp' = - BSL.writeFile (dir "dump.tar") =<< - (decompress <$> BSL.readFile fp') + BSL.writeFile (dir "dump.tar") . decompress =<< BSL.readFile fp' decompressFile :: FilePath -> IO () -decompressFile fp' = forceBSL =<< - (decompress <$> BSL.readFile fp') +decompressFile fp' = forceBSL . decompress =<< BSL.readFile fp' where forceBSL = (`seq` mempty) . last . BSL.toChunks compressDump :: FilePath -> FilePath -> IO () compressDump dir fp' = - BSL.writeFile (dir "dump.tar.lz4") =<< - (compress <$> BSL.readFile fp') + BSL.writeFile (dir "dump.tar.lz4") . compress =<< BSL.readFile fp' main :: IO () main = diff --git a/cbits/lz4.c b/cbits/lz4.c --- a/cbits/lz4.c +++ b/cbits/lz4.c @@ -1,6 +1,6 @@ /* LZ4 - Fast LZ compression algorithm - Copyright (C) 2011-present, Yann Collet. + Copyright (C) 2011-2020, Yann Collet. BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) @@ -124,6 +124,7 @@ #if defined(_MSC_VER) && (_MSC_VER >= 1400) /* Visual Studio 2005+ */ # include /* only present in VS2005+ */ # pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ +# pragma warning(disable : 6237) /* disable: C6237: conditional expression is always 0 */ #endif /* _MSC_VER */ #ifndef LZ4_FORCE_INLINE @@ -187,7 +188,27 @@ /*-************************************ * Memory routines **************************************/ -#ifdef LZ4_USER_MEMORY_FUNCTIONS + +/*! LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION : + * Disable relatively high-level LZ4/HC functions that use dynamic memory + * allocation functions (malloc(), calloc(), free()). + * + * Note that this is a compile-time switch. And since it disables + * public/stable LZ4 v1 API functions, we don't recommend using this + * symbol to generate a library for distribution. + * + * The following public functions are removed when this symbol is defined. + * - lz4 : LZ4_createStream, LZ4_freeStream, + * LZ4_createStreamDecode, LZ4_freeStreamDecode, LZ4_create (deprecated) + * - lz4hc : LZ4_createStreamHC, LZ4_freeStreamHC, + * LZ4_createHC (deprecated), LZ4_freeHC (deprecated) + * - lz4frame, lz4file : All LZ4F_* functions + */ +#if defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) +# define ALLOC(s) lz4_error_memory_allocation_is_disabled +# define ALLOC_AND_ZERO(s) lz4_error_memory_allocation_is_disabled +# define FREEMEM(p) lz4_error_memory_allocation_is_disabled +#elif defined(LZ4_USER_MEMORY_FUNCTIONS) /* memory management functions can be customized by user project. * Below functions must exist somewhere in the Project * and be available at link time */ @@ -204,8 +225,13 @@ # define FREEMEM(p) free(p) #endif -#include /* memset, memcpy */ -#define MEM_INIT(p,v,s) memset((p),(v),(s)) +#if ! LZ4_FREESTANDING +# include /* memset, memcpy */ +#endif +#if !defined(LZ4_memset) +# define LZ4_memset(p,v,s) memset((p),(v),(s)) +#endif +#define MEM_INIT(p,v,s) LZ4_memset((p),(v),(s)) /*-************************************ @@ -316,12 +342,22 @@ * memcpy() as if it were standard compliant, so it can inline it in freestanding * environments. This is needed when decompressing the Linux Kernel, for example. */ -#if defined(__GNUC__) && (__GNUC__ >= 4) -#define LZ4_memcpy(dst, src, size) __builtin_memcpy(dst, src, size) -#else -#define LZ4_memcpy(dst, src, size) memcpy(dst, src, size) +#if !defined(LZ4_memcpy) +# if defined(__GNUC__) && (__GNUC__ >= 4) +# define LZ4_memcpy(dst, src, size) __builtin_memcpy(dst, src, size) +# else +# define LZ4_memcpy(dst, src, size) memcpy(dst, src, size) +# endif #endif +#if !defined(LZ4_memmove) +# if defined(__GNUC__) && (__GNUC__ >= 4) +# define LZ4_memmove __builtin_memmove +# else +# define LZ4_memmove memmove +# endif +#endif + static unsigned LZ4_isLittleEndian(void) { const union { U32 u; BYTE c[4]; } one = { 1 }; /* don't use static : performance detrimental */ @@ -343,14 +379,14 @@ /* __pack instructions are safer, but compiler specific, hence potentially problematic for some compilers */ /* currently only defined for gcc and icc */ -typedef union { U16 u16; U32 u32; reg_t uArch; } __attribute__((packed)) unalign; +typedef union { U16 u16; U32 u32; reg_t uArch; } __attribute__((packed)) LZ4_unalign; -static U16 LZ4_read16(const void* ptr) { return ((const unalign*)ptr)->u16; } -static U32 LZ4_read32(const void* ptr) { return ((const unalign*)ptr)->u32; } -static reg_t LZ4_read_ARCH(const void* ptr) { return ((const unalign*)ptr)->uArch; } +static U16 LZ4_read16(const void* ptr) { return ((const LZ4_unalign*)ptr)->u16; } +static U32 LZ4_read32(const void* ptr) { return ((const LZ4_unalign*)ptr)->u32; } +static reg_t LZ4_read_ARCH(const void* ptr) { return ((const LZ4_unalign*)ptr)->uArch; } -static void LZ4_write16(void* memPtr, U16 value) { ((unalign*)memPtr)->u16 = value; } -static void LZ4_write32(void* memPtr, U32 value) { ((unalign*)memPtr)->u32 = value; } +static void LZ4_write16(void* memPtr, U16 value) { ((LZ4_unalign*)memPtr)->u16 = value; } +static void LZ4_write32(void* memPtr, U32 value) { ((LZ4_unalign*)memPtr)->u32 = value; } #else /* safe and portable access using memcpy() */ @@ -421,10 +457,12 @@ #ifndef LZ4_FAST_DEC_LOOP # if defined __i386__ || defined _M_IX86 || defined __x86_64__ || defined _M_X64 # define LZ4_FAST_DEC_LOOP 1 +# elif defined(__aarch64__) && defined(__APPLE__) +# define LZ4_FAST_DEC_LOOP 1 # elif defined(__aarch64__) && !defined(__clang__) - /* On aarch64, we disable this optimization for clang because on certain - * mobile chipsets, performance is reduced with clang. For information - * refer to https://github.com/lz4/lz4/pull/707 */ + /* On non-Apple aarch64, we disable this optimization for clang because + * on certain mobile chipsets, performance is reduced with clang. For + * more information refer to https://github.com/lz4/lz4/pull/707 */ # define LZ4_FAST_DEC_LOOP 1 # else # define LZ4_FAST_DEC_LOOP 0 @@ -486,7 +524,14 @@ case 2: LZ4_memcpy(v, srcPtr, 2); LZ4_memcpy(&v[2], srcPtr, 2); +#if defined(_MSC_VER) && (_MSC_VER <= 1933) /* MSVC 2022 ver 17.3 or earlier */ +# pragma warning(push) +# pragma warning(disable : 6385) /* warning C6385: Reading invalid data from 'v'. */ +#endif LZ4_memcpy(&v[4], v, 4); +#if defined(_MSC_VER) && (_MSC_VER <= 1933) /* MSVC 2022 ver 17.3 or earlier */ +# pragma warning(pop) +#endif break; case 4: LZ4_memcpy(v, srcPtr, 4); @@ -515,9 +560,20 @@ assert(val != 0); if (LZ4_isLittleEndian()) { if (sizeof(val) == 8) { -# if defined(_MSC_VER) && (_MSC_VER >= 1800) && defined(_M_AMD64) && !defined(LZ4_FORCE_SW_BITCOUNT) +# if defined(_MSC_VER) && (_MSC_VER >= 1800) && (defined(_M_AMD64) && !defined(_M_ARM64EC)) && !defined(LZ4_FORCE_SW_BITCOUNT) +/*-************************************************************************************************* +* ARM64EC is a Microsoft-designed ARM64 ABI compatible with AMD64 applications on ARM64 Windows 11. +* The ARM64EC ABI does not support AVX/AVX2/AVX512 instructions, nor their relevant intrinsics +* including _tzcnt_u64. Therefore, we need to neuter the _tzcnt_u64 code path for ARM64EC. +****************************************************************************************************/ +# if defined(__clang__) && (__clang_major__ < 10) + /* Avoid undefined clang-cl intrinsics issue. + * See https://github.com/lz4/lz4/pull/1017 for details. */ + return (unsigned)__builtin_ia32_tzcnt_u64(val) >> 3; +# else /* x64 CPUS without BMI support interpret `TZCNT` as `REP BSF` */ return (unsigned)_tzcnt_u64(val) >> 3; +# endif # elif defined(_MSC_VER) && defined(_WIN64) && !defined(LZ4_FORCE_SW_BITCOUNT) unsigned long r = 0; _BitScanForward64(&r, (U64)val); @@ -652,10 +708,10 @@ * - usingExtDict : Like withPrefix64k, but the preceding content is somewhere * else in memory, starting at ctx->dictionary with length * ctx->dictSize. - * - usingDictCtx : Like usingExtDict, but everything concerning the preceding - * content is in a separate context, pointed to by - * ctx->dictCtx. ctx->dictionary, ctx->dictSize, and table - * entries in the current context that refer to positions + * - usingDictCtx : Everything concerning the preceding content is + * in a separate context, pointed to by ctx->dictCtx. + * ctx->dictionary, ctx->dictSize, and table entries + * in the current context that refer to positions * preceding the beginning of the current compression are * ignored. Instead, ctx->dictCtx->dictionary and ctx->dictCtx * ->dictSize describe the location and size of the preceding @@ -672,12 +728,12 @@ int LZ4_versionNumber (void) { return LZ4_VERSION_NUMBER; } const char* LZ4_versionString(void) { return LZ4_VERSION_STRING; } int LZ4_compressBound(int isize) { return LZ4_COMPRESSBOUND(isize); } -int LZ4_sizeofState(void) { return LZ4_STREAMSIZE; } +int LZ4_sizeofState(void) { return sizeof(LZ4_stream_t); } -/*-************************************ -* Internal Definitions used in Tests -**************************************/ +/*-**************************************** +* Internal Definitions, used only in Tests +*******************************************/ #if defined (__cplusplus) extern "C" { #endif @@ -687,7 +743,9 @@ int LZ4_decompress_safe_forceExtDict(const char* source, char* dest, int compressedSize, int maxOutputSize, const void* dictStart, size_t dictSize); - +int LZ4_decompress_safe_partial_forceExtDict(const char* source, char* dest, + int compressedSize, int targetOutputSize, int dstCapacity, + const void* dictStart, size_t dictSize); #if defined (__cplusplus) } #endif @@ -827,9 +885,10 @@ } } - /* Adding a gap, so all previous entries are > LZ4_DISTANCE_MAX back, is faster - * than compressing without a gap. However, compressing with - * currentOffset == 0 is faster still, so we preserve that case. + /* Adding a gap, so all previous entries are > LZ4_DISTANCE_MAX back, + * is faster than compressing without a gap. + * However, compressing with currentOffset == 0 is faster still, + * so we preserve that case. */ if (cctx->currentOffset != 0 && tableType == byU32) { DEBUGLOG(5, "LZ4_prepareTable: adding 64KB to currentOffset"); @@ -853,7 +912,7 @@ const char* const source, char* const dest, const int inputSize, - int *inputConsumed, /* only written when outputDirective == fillOutput */ + int* inputConsumed, /* only written when outputDirective == fillOutput */ const int maxOutputSize, const limitedOutput_directive outputDirective, const tableType_t tableType, @@ -885,7 +944,8 @@ /* the dictCtx currentOffset is indexed on the start of the dictionary, * while a dictionary in the current context precedes the currentOffset */ - const BYTE* dictBase = !dictionary ? NULL : (dictDirective == usingDictCtx) ? + const BYTE* dictBase = (dictionary == NULL) ? NULL : + (dictDirective == usingDictCtx) ? dictionary + dictSize - dictCtx->currentOffset : dictionary + dictSize - startIndex; @@ -981,10 +1041,11 @@ match = base + matchIndex; lowLimit = (const BYTE*)source; } - } else if (dictDirective==usingExtDict) { + } else if (dictDirective == usingExtDict) { if (matchIndex < startIndex) { DEBUGLOG(7, "extDict candidate: matchIndex=%5u < startIndex=%5u", matchIndex, startIndex); assert(startIndex - matchIndex >= MINMATCH); + assert(dictBase); match = dictBase + matchIndex; lowLimit = dictionary; } else { @@ -1048,7 +1109,7 @@ _next_match: /* at this stage, the following variables must be correctly set : * - ip : at start of LZ operation - * - match : at start of previous pattern occurence; can be within current prefix, or within extDict + * - match : at start of previous pattern occurrence; can be within current prefix, or within extDict * - offset : if maybe_ext_memSegment==1 (constant) * - lowLimit : must be == dictionary to mean "match is within extDict"; must be == source otherwise * - token and *token : position to write 4-bits for match length; higher 4-bits for literal length supposed already written @@ -1173,6 +1234,7 @@ } } else if (dictDirective==usingExtDict) { if (matchIndex < startIndex) { + assert(dictBase); match = dictBase + matchIndex; lowLimit = dictionary; /* required for match length counter */ } else { @@ -1355,7 +1417,7 @@ { int result; #if (LZ4_HEAPMODE) - LZ4_stream_t* ctxPtr = ALLOC(sizeof(LZ4_stream_t)); /* malloc-calloc always properly aligned */ + LZ4_stream_t* ctxPtr = (LZ4_stream_t*)ALLOC(sizeof(LZ4_stream_t)); /* malloc-calloc always properly aligned */ if (ctxPtr == NULL) return 0; #else LZ4_stream_t ctx; @@ -1420,15 +1482,17 @@ * Streaming functions ********************************/ +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) LZ4_stream_t* LZ4_createStream(void) { LZ4_stream_t* const lz4s = (LZ4_stream_t*)ALLOC(sizeof(LZ4_stream_t)); - LZ4_STATIC_ASSERT(LZ4_STREAMSIZE >= sizeof(LZ4_stream_t_internal)); /* A compilation error here means LZ4_STREAMSIZE is not large enough */ + LZ4_STATIC_ASSERT(sizeof(LZ4_stream_t) >= sizeof(LZ4_stream_t_internal)); DEBUGLOG(4, "LZ4_createStream %p", lz4s); if (lz4s == NULL) return NULL; LZ4_initStream(lz4s, sizeof(*lz4s)); return lz4s; } +#endif static size_t LZ4_stream_t_alignment(void) { @@ -1462,6 +1526,7 @@ LZ4_prepareTable(&(ctx->internal_donotuse), 0, byU32); } +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) int LZ4_freeStream (LZ4_stream_t* LZ4_stream) { if (!LZ4_stream) return 0; /* support free on NULL */ @@ -1469,6 +1534,7 @@ FREEMEM(LZ4_stream); return (0); } +#endif #define HASH_UNIT sizeof(reg_t) @@ -1514,8 +1580,9 @@ return (int)dict->dictSize; } -void LZ4_attach_dictionary(LZ4_stream_t* workingStream, const LZ4_stream_t* dictionaryStream) { - const LZ4_stream_t_internal* dictCtx = dictionaryStream == NULL ? NULL : +void LZ4_attach_dictionary(LZ4_stream_t* workingStream, const LZ4_stream_t* dictionaryStream) +{ + const LZ4_stream_t_internal* dictCtx = (dictionaryStream == NULL) ? NULL : &(dictionaryStream->internal_donotuse); DEBUGLOG(4, "LZ4_attach_dictionary (%p, %p, size %u)", @@ -1568,36 +1635,40 @@ int acceleration) { const tableType_t tableType = byU32; - LZ4_stream_t_internal* streamPtr = &LZ4_stream->internal_donotuse; - const BYTE* dictEnd = streamPtr->dictionary + streamPtr->dictSize; + LZ4_stream_t_internal* const streamPtr = &LZ4_stream->internal_donotuse; + const char* dictEnd = streamPtr->dictSize ? (const char*)streamPtr->dictionary + streamPtr->dictSize : NULL; - DEBUGLOG(5, "LZ4_compress_fast_continue (inputSize=%i)", inputSize); + DEBUGLOG(5, "LZ4_compress_fast_continue (inputSize=%i, dictSize=%u)", inputSize, streamPtr->dictSize); - LZ4_renormDictT(streamPtr, inputSize); /* avoid index overflow */ + LZ4_renormDictT(streamPtr, inputSize); /* fix index overflow */ if (acceleration < 1) acceleration = LZ4_ACCELERATION_DEFAULT; if (acceleration > LZ4_ACCELERATION_MAX) acceleration = LZ4_ACCELERATION_MAX; /* invalidate tiny dictionaries */ - if ( (streamPtr->dictSize-1 < 4-1) /* intentional underflow */ - && (dictEnd != (const BYTE*)source) ) { + if ( (streamPtr->dictSize < 4) /* tiny dictionary : not enough for a hash */ + && (dictEnd != source) /* prefix mode */ + && (inputSize > 0) /* tolerance : don't lose history, in case next invocation would use prefix mode */ + && (streamPtr->dictCtx == NULL) /* usingDictCtx */ + ) { DEBUGLOG(5, "LZ4_compress_fast_continue: dictSize(%u) at addr:%p is too small", streamPtr->dictSize, streamPtr->dictionary); + /* remove dictionary existence from history, to employ faster prefix mode */ streamPtr->dictSize = 0; streamPtr->dictionary = (const BYTE*)source; - dictEnd = (const BYTE*)source; + dictEnd = source; } /* Check overlapping input/dictionary space */ - { const BYTE* sourceEnd = (const BYTE*) source + inputSize; - if ((sourceEnd > streamPtr->dictionary) && (sourceEnd < dictEnd)) { + { const char* const sourceEnd = source + inputSize; + if ((sourceEnd > (const char*)streamPtr->dictionary) && (sourceEnd < dictEnd)) { streamPtr->dictSize = (U32)(dictEnd - sourceEnd); if (streamPtr->dictSize > 64 KB) streamPtr->dictSize = 64 KB; if (streamPtr->dictSize < 4) streamPtr->dictSize = 0; - streamPtr->dictionary = dictEnd - streamPtr->dictSize; + streamPtr->dictionary = (const BYTE*)dictEnd - streamPtr->dictSize; } } /* prefix mode : source data follows dictionary */ - if (dictEnd == (const BYTE*)source) { + if (dictEnd == source) { if ((streamPtr->dictSize < 64 KB) && (streamPtr->dictSize < streamPtr->currentOffset)) return LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, withPrefix64k, dictSmall, acceleration); else @@ -1623,7 +1694,7 @@ } else { result = LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, usingDictCtx, noDictIssue, acceleration); } - } else { + } else { /* small data <= 4 KB */ if ((streamPtr->dictSize < 64 KB) && (streamPtr->dictSize < streamPtr->currentOffset)) { result = LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, usingExtDict, dictSmall, acceleration); } else { @@ -1661,21 +1732,25 @@ /*! LZ4_saveDict() : * If previously compressed data block is not guaranteed to remain available at its memory location, * save it into a safer place (char* safeBuffer). - * Note : you don't need to call LZ4_loadDict() afterwards, - * dictionary is immediately usable, you can therefore call LZ4_compress_fast_continue(). - * Return : saved dictionary size in bytes (necessarily <= dictSize), or 0 if error. + * Note : no need to call LZ4_loadDict() afterwards, dictionary is immediately usable, + * one can therefore call LZ4_compress_fast_continue() right after. + * @return : saved dictionary size in bytes (necessarily <= dictSize), or 0 if error. */ int LZ4_saveDict (LZ4_stream_t* LZ4_dict, char* safeBuffer, int dictSize) { LZ4_stream_t_internal* const dict = &LZ4_dict->internal_donotuse; - const BYTE* const previousDictEnd = dict->dictionary + dict->dictSize; + DEBUGLOG(5, "LZ4_saveDict : dictSize=%i, safeBuffer=%p", dictSize, safeBuffer); + if ((U32)dictSize > 64 KB) { dictSize = 64 KB; } /* useless to define a dictionary > 64 KB */ if ((U32)dictSize > dict->dictSize) { dictSize = (int)dict->dictSize; } if (safeBuffer == NULL) assert(dictSize == 0); - if (dictSize > 0) - memmove(safeBuffer, previousDictEnd - dictSize, dictSize); + if (dictSize > 0) { + const BYTE* const previousDictEnd = dict->dictionary + dict->dictSize; + assert(dict->dictionary); + LZ4_memmove(safeBuffer, previousDictEnd - dictSize, (size_t)dictSize); + } dict->dictionary = (const BYTE*)safeBuffer; dict->dictSize = (U32)dictSize; @@ -1689,40 +1764,164 @@ * Decompression functions ********************************/ -typedef enum { endOnOutputSize = 0, endOnInputSize = 1 } endCondition_directive; typedef enum { decode_full_block = 0, partial_decode = 1 } earlyEnd_directive; #undef MIN #define MIN(a,b) ( (a) < (b) ? (a) : (b) ) + +/* variant for decompress_unsafe() + * does not know end of input + * presumes input is well formed + * note : will consume at least one byte */ +size_t read_long_length_no_check(const BYTE** pp) +{ + size_t b, l = 0; + do { b = **pp; (*pp)++; l += b; } while (b==255); + DEBUGLOG(6, "read_long_length_no_check: +length=%zu using %zu input bytes", l, l/255 + 1) + return l; +} + +/* core decoder variant for LZ4_decompress_fast*() + * for legacy support only : these entry points are deprecated. + * - Presumes input is correctly formed (no defense vs malformed inputs) + * - Does not know input size (presume input buffer is "large enough") + * - Decompress a full block (only) + * @return : nb of bytes read from input. + * Note : this variant is not optimized for speed, just for maintenance. + * the goal is to remove support of decompress_fast*() variants by v2.0 +**/ +LZ4_FORCE_INLINE int +LZ4_decompress_unsafe_generic( + const BYTE* const istart, + BYTE* const ostart, + int decompressedSize, + + size_t prefixSize, + const BYTE* const dictStart, /* only if dict==usingExtDict */ + const size_t dictSize /* note: =0 if dictStart==NULL */ + ) +{ + const BYTE* ip = istart; + BYTE* op = (BYTE*)ostart; + BYTE* const oend = ostart + decompressedSize; + const BYTE* const prefixStart = ostart - prefixSize; + + DEBUGLOG(5, "LZ4_decompress_unsafe_generic"); + if (dictStart == NULL) assert(dictSize == 0); + + while (1) { + /* start new sequence */ + unsigned token = *ip++; + + /* literals */ + { size_t ll = token >> ML_BITS; + if (ll==15) { + /* long literal length */ + ll += read_long_length_no_check(&ip); + } + if ((size_t)(oend-op) < ll) return -1; /* output buffer overflow */ + LZ4_memmove(op, ip, ll); /* support in-place decompression */ + op += ll; + ip += ll; + if ((size_t)(oend-op) < MFLIMIT) { + if (op==oend) break; /* end of block */ + DEBUGLOG(5, "invalid: literals end at distance %zi from end of block", oend-op); + /* incorrect end of block : + * last match must start at least MFLIMIT==12 bytes before end of output block */ + return -1; + } } + + /* match */ + { size_t ml = token & 15; + size_t const offset = LZ4_readLE16(ip); + ip+=2; + + if (ml==15) { + /* long literal length */ + ml += read_long_length_no_check(&ip); + } + ml += MINMATCH; + + if ((size_t)(oend-op) < ml) return -1; /* output buffer overflow */ + + { const BYTE* match = op - offset; + + /* out of range */ + if (offset > (size_t)(op - prefixStart) + dictSize) { + DEBUGLOG(6, "offset out of range"); + return -1; + } + + /* check special case : extDict */ + if (offset > (size_t)(op - prefixStart)) { + /* extDict scenario */ + const BYTE* const dictEnd = dictStart + dictSize; + const BYTE* extMatch = dictEnd - (offset - (size_t)(op-prefixStart)); + size_t const extml = (size_t)(dictEnd - extMatch); + if (extml > ml) { + /* match entirely within extDict */ + LZ4_memmove(op, extMatch, ml); + op += ml; + ml = 0; + } else { + /* match split between extDict & prefix */ + LZ4_memmove(op, extMatch, extml); + op += extml; + ml -= extml; + } + match = prefixStart; + } + + /* match copy - slow variant, supporting overlap copy */ + { size_t u; + for (u=0; u= lencheck. - * loop_check - check ip >= lencheck in body of loop. Returns loop_error if so. - * initial_check - check ip >= lencheck before start of loop. Returns initial_error if so. - * error (output) - error code. Should be set to 0 before call. - */ -typedef enum { loop_error = -2, initial_error = -1, ok = 0 } variable_length_error; -LZ4_FORCE_INLINE unsigned -read_variable_length(const BYTE**ip, const BYTE* lencheck, - int loop_check, int initial_check, - variable_length_error* error) + * @ip : input pointer + * @ilimit : position after which if length is not decoded, the input is necessarily corrupted. + * @initial_check - check ip >= ipmax before start of loop. Returns initial_error if so. + * @error (output) - error code. Must be set to 0 before call. +**/ +typedef size_t Rvl_t; +static const Rvl_t rvl_error = (Rvl_t)(-1); +LZ4_FORCE_INLINE Rvl_t +read_variable_length(const BYTE** ip, const BYTE* ilimit, + int initial_check) { - U32 length = 0; - U32 s; - if (initial_check && unlikely((*ip) >= lencheck)) { /* overflow detection */ - *error = initial_error; - return length; + Rvl_t s, length = 0; + assert(ip != NULL); + assert(*ip != NULL); + assert(ilimit != NULL); + if (initial_check && unlikely((*ip) >= ilimit)) { /* read limit reached */ + return rvl_error; } do { s = **ip; (*ip)++; length += s; - if (loop_check && unlikely((*ip) >= lencheck)) { /* overflow detection */ - *error = loop_error; - return length; + if (unlikely((*ip) > ilimit)) { /* read limit reached */ + return rvl_error; } + /* accumulator overflow detection (32-bit mode only) */ + if ((sizeof(length)<8) && unlikely(length > ((Rvl_t)(-1)/2)) ) { + return rvl_error; + } } while (s==255); return length; @@ -1741,7 +1940,6 @@ int srcSize, int outputSize, /* If endOnInput==endOnInputSize, this value is `dstCapacity` */ - endCondition_directive endOnInput, /* endOnOutputSize, endOnInputSize */ earlyEnd_directive partialDecoding, /* full, partial */ dict_directive dict, /* noDict, withPrefix64k, usingExtDict */ const BYTE* const lowPrefix, /* always <= dst, == dst when no prefix */ @@ -1749,7 +1947,7 @@ const size_t dictSize /* note : = 0 if noDict */ ) { - if (src == NULL) { return -1; } + if ((src == NULL) || (outputSize < 0)) { return -1; } { const BYTE* ip = (const BYTE*) src; const BYTE* const iend = ip + srcSize; @@ -1760,13 +1958,12 @@ const BYTE* const dictEnd = (dictStart == NULL) ? NULL : dictStart + dictSize; - const int safeDecode = (endOnInput==endOnInputSize); - const int checkOffset = ((safeDecode) && (dictSize < (int)(64 KB))); + const int checkOffset = (dictSize < (int)(64 KB)); /* Set up the "end" pointers for the shortcut. */ - const BYTE* const shortiend = iend - (endOnInput ? 14 : 8) /*maxLL*/ - 2 /*offset*/; - const BYTE* const shortoend = oend - (endOnInput ? 14 : 8) /*maxLL*/ - 18 /*maxML*/; + const BYTE* const shortiend = iend - 14 /*maxLL*/ - 2 /*offset*/; + const BYTE* const shortoend = oend - 14 /*maxLL*/ - 18 /*maxML*/; const BYTE* match; size_t offset; @@ -1778,83 +1975,70 @@ /* Special cases */ assert(lowPrefix <= op); - if ((endOnInput) && (unlikely(outputSize==0))) { + if (unlikely(outputSize==0)) { /* Empty output buffer */ if (partialDecoding) return 0; return ((srcSize==1) && (*ip==0)) ? 0 : -1; } - if ((!endOnInput) && (unlikely(outputSize==0))) { return (*ip==0 ? 1 : -1); } - if ((endOnInput) && unlikely(srcSize==0)) { return -1; } + if (unlikely(srcSize==0)) { return -1; } - /* Currently the fast loop shows a regression on qualcomm arm chips. */ + /* LZ4_FAST_DEC_LOOP: + * designed for modern OoO performance cpus, + * where copying reliably 32-bytes is preferable to an unpredictable branch. + * note : fast loop may show a regression for some client arm chips. */ #if LZ4_FAST_DEC_LOOP if ((oend - op) < FASTLOOP_SAFE_DISTANCE) { DEBUGLOG(6, "skip fast decode loop"); goto safe_decode; } - /* Fast loop : decode sequences as long as output < iend-FASTLOOP_SAFE_DISTANCE */ + /* Fast loop : decode sequences as long as output < oend-FASTLOOP_SAFE_DISTANCE */ while (1) { /* Main fastloop assertion: We can always wildcopy FASTLOOP_SAFE_DISTANCE */ assert(oend - op >= FASTLOOP_SAFE_DISTANCE); - if (endOnInput) { assert(ip < iend); } + assert(ip < iend); token = *ip++; length = token >> ML_BITS; /* literal length */ - assert(!endOnInput || ip <= iend); /* ip < iend before the increment */ - /* decode literal length */ if (length == RUN_MASK) { - variable_length_error error = ok; - length += read_variable_length(&ip, iend-RUN_MASK, (int)endOnInput, (int)endOnInput, &error); - if (error == initial_error) { goto _output_error; } - if ((safeDecode) && unlikely((uptrval)(op)+length<(uptrval)(op))) { goto _output_error; } /* overflow detection */ - if ((safeDecode) && unlikely((uptrval)(ip)+length<(uptrval)(ip))) { goto _output_error; } /* overflow detection */ + size_t const addl = read_variable_length(&ip, iend-RUN_MASK, 1); + if (addl == rvl_error) { goto _output_error; } + length += addl; + if (unlikely((uptrval)(op)+length<(uptrval)(op))) { goto _output_error; } /* overflow detection */ + if (unlikely((uptrval)(ip)+length<(uptrval)(ip))) { goto _output_error; } /* overflow detection */ /* copy literals */ cpy = op+length; LZ4_STATIC_ASSERT(MFLIMIT >= WILDCOPYLENGTH); - if (endOnInput) { /* LZ4_decompress_safe() */ - if ((cpy>oend-32) || (ip+length>iend-32)) { goto safe_literal_copy; } - LZ4_wildCopy32(op, ip, cpy); - } else { /* LZ4_decompress_fast() */ - if (cpy>oend-8) { goto safe_literal_copy; } - LZ4_wildCopy8(op, ip, cpy); /* LZ4_decompress_fast() cannot copy more than 8 bytes at a time : - * it doesn't know input length, and only relies on end-of-block properties */ - } + if ((cpy>oend-32) || (ip+length>iend-32)) { goto safe_literal_copy; } + LZ4_wildCopy32(op, ip, cpy); ip += length; op = cpy; } else { cpy = op+length; - if (endOnInput) { /* LZ4_decompress_safe() */ - DEBUGLOG(7, "copy %u bytes in a 16-bytes stripe", (unsigned)length); - /* We don't need to check oend, since we check it once for each loop below */ - if (ip > iend-(16 + 1/*max lit + offset + nextToken*/)) { goto safe_literal_copy; } - /* Literals can only be 14, but hope compilers optimize if we copy by a register size */ - LZ4_memcpy(op, ip, 16); - } else { /* LZ4_decompress_fast() */ - /* LZ4_decompress_fast() cannot copy more than 8 bytes at a time : - * it doesn't know input length, and relies on end-of-block properties */ - LZ4_memcpy(op, ip, 8); - if (length > 8) { LZ4_memcpy(op+8, ip+8, 8); } - } + DEBUGLOG(7, "copy %u bytes in a 16-bytes stripe", (unsigned)length); + /* We don't need to check oend, since we check it once for each loop below */ + if (ip > iend-(16 + 1/*max lit + offset + nextToken*/)) { goto safe_literal_copy; } + /* Literals can only be <= 14, but hope compilers optimize better when copy by a register size */ + LZ4_memcpy(op, ip, 16); ip += length; op = cpy; } /* get offset */ offset = LZ4_readLE16(ip); ip+=2; match = op - offset; - assert(match <= op); + assert(match <= op); /* overflow check */ /* get matchlength */ length = token & ML_MASK; if (length == ML_MASK) { - variable_length_error error = ok; - if ((checkOffset) && (unlikely(match + dictSize < lowPrefix))) { goto _output_error; } /* Error : offset outside buffers */ - length += read_variable_length(&ip, iend - LASTLITERALS + 1, (int)endOnInput, 0, &error); - if (error != ok) { goto _output_error; } - if ((safeDecode) && unlikely((uptrval)(op)+length<(uptrval)op)) { goto _output_error; } /* overflow detection */ + size_t const addl = read_variable_length(&ip, iend - LASTLITERALS + 1, 0); + if (addl == rvl_error) { goto _output_error; } + length += addl; length += MINMATCH; + if (unlikely((uptrval)(op)+length<(uptrval)op)) { goto _output_error; } /* overflow detection */ + if ((checkOffset) && (unlikely(match + dictSize < lowPrefix))) { goto _output_error; } /* Error : offset outside buffers */ if (op + length >= oend - FASTLOOP_SAFE_DISTANCE) { goto safe_match_copy; } @@ -1864,7 +2048,7 @@ goto safe_match_copy; } - /* Fastpath check: Avoids a branch in LZ4_wildCopy32 if true */ + /* Fastpath check: skip LZ4_wildCopy32 when true */ if ((dict == withPrefix64k) || (match >= lowPrefix)) { if (offset >= 8) { assert(match >= lowPrefix); @@ -1881,6 +2065,7 @@ if (checkOffset && (unlikely(match + dictSize < lowPrefix))) { goto _output_error; } /* Error : offset outside buffers */ /* match starting within external dictionary */ if ((dict==usingExtDict) && (match < lowPrefix)) { + assert(dictEnd != NULL); if (unlikely(op+length > oend-LASTLITERALS)) { if (partialDecoding) { DEBUGLOG(7, "partialDecoding: dictionary match, close to dstEnd"); @@ -1891,7 +2076,7 @@ if (length <= (size_t)(lowPrefix-match)) { /* match fits entirely within external dictionary : just copy */ - memmove(op, dictEnd - (lowPrefix-match), length); + LZ4_memmove(op, dictEnd - (lowPrefix-match), length); op += length; } else { /* match stretches into both external dictionary and current block */ @@ -1927,11 +2112,10 @@ /* Main Loop : decode remaining sequences where output < FASTLOOP_SAFE_DISTANCE */ while (1) { + assert(ip < iend); token = *ip++; length = token >> ML_BITS; /* literal length */ - assert(!endOnInput || ip <= iend); /* ip < iend before the increment */ - /* A two-stage shortcut for the most common case: * 1) If the literal length is 0..14, and there is enough space, * enter the shortcut and copy 16 bytes on behalf of the literals @@ -1941,11 +2125,11 @@ * those 18 bytes earlier, upon entering the shortcut (in other words, * there is a combined check for both stages). */ - if ( (endOnInput ? length != RUN_MASK : length <= 8) + if ( (length != RUN_MASK) /* strictly "less than" on input, to re-enter the loop with at least one byte */ - && likely((endOnInput ? ip < shortiend : 1) & (op <= shortoend)) ) { + && likely((ip < shortiend) & (op <= shortoend)) ) { /* Copy the literals */ - LZ4_memcpy(op, ip, endOnInput ? 16 : 8); + LZ4_memcpy(op, ip, 16); op += length; ip += length; /* The second stage: prepare for match copying, decode full info. @@ -1975,11 +2159,11 @@ /* decode literal length */ if (length == RUN_MASK) { - variable_length_error error = ok; - length += read_variable_length(&ip, iend-RUN_MASK, (int)endOnInput, (int)endOnInput, &error); - if (error == initial_error) { goto _output_error; } - if ((safeDecode) && unlikely((uptrval)(op)+length<(uptrval)(op))) { goto _output_error; } /* overflow detection */ - if ((safeDecode) && unlikely((uptrval)(ip)+length<(uptrval)(ip))) { goto _output_error; } /* overflow detection */ + size_t const addl = read_variable_length(&ip, iend-RUN_MASK, 1); + if (addl == rvl_error) { goto _output_error; } + length += addl; + if (unlikely((uptrval)(op)+length<(uptrval)(op))) { goto _output_error; } /* overflow detection */ + if (unlikely((uptrval)(ip)+length<(uptrval)(ip))) { goto _output_error; } /* overflow detection */ } /* copy literals */ @@ -1988,9 +2172,7 @@ safe_literal_copy: #endif LZ4_STATIC_ASSERT(MFLIMIT >= WILDCOPYLENGTH); - if ( ((endOnInput) && ((cpy>oend-MFLIMIT) || (ip+length>iend-(2+1+LASTLITERALS))) ) - || ((!endOnInput) && (cpy>oend-WILDCOPYLENGTH)) ) - { + if ((cpy>oend-MFLIMIT) || (ip+length>iend-(2+1+LASTLITERALS))) { /* We've either hit the input parsing restriction or the output parsing restriction. * In the normal scenario, decoding a full block, it must be the last sequence, * otherwise it's an error (invalid input or dimensions). @@ -2000,7 +2182,6 @@ /* Since we are partial decoding we may be in this block because of the output parsing * restriction, which is not valid since the output buffer is allowed to be undersized. */ - assert(endOnInput); DEBUGLOG(7, "partialDecoding: copying literals, close to input or output end") DEBUGLOG(7, "partialDecoding: literal length = %u", (unsigned)length); DEBUGLOG(7, "partialDecoding: remaining space in dstBuffer : %i", (int)(oend - op)); @@ -2021,21 +2202,17 @@ length = (size_t)(oend-op); } } else { - /* We must be on the last sequence because of the parsing limitations so check - * that we exactly regenerate the original size (must be exact when !endOnInput). - */ - if ((!endOnInput) && (cpy != oend)) { goto _output_error; } /* We must be on the last sequence (or invalid) because of the parsing limitations * so check that we exactly consume the input and don't overrun the output buffer. */ - if ((endOnInput) && ((ip+length != iend) || (cpy > oend))) { + if ((ip+length != iend) || (cpy > oend)) { DEBUGLOG(6, "should have been last run of literals") DEBUGLOG(6, "ip(%p) + length(%i) = %p != iend (%p)", ip, (int)length, ip+length, iend); DEBUGLOG(6, "or cpy(%p) > oend(%p)", cpy, oend); goto _output_error; } } - memmove(op, ip, length); /* supports overlapping memory regions; only matters for in-place decompression scenarios */ + LZ4_memmove(op, ip, length); /* supports overlapping memory regions, for in-place decompression scenarios */ ip += length; op += length; /* Necessarily EOF when !partialDecoding. @@ -2047,7 +2224,7 @@ break; } } else { - LZ4_wildCopy8(op, ip, cpy); /* may overwrite up to WILDCOPYLENGTH beyond cpy */ + LZ4_wildCopy8(op, ip, cpy); /* can overwrite up to 8 bytes beyond cpy */ ip += length; op = cpy; } @@ -2060,10 +2237,10 @@ _copy_match: if (length == ML_MASK) { - variable_length_error error = ok; - length += read_variable_length(&ip, iend - LASTLITERALS + 1, (int)endOnInput, 0, &error); - if (error != ok) goto _output_error; - if ((safeDecode) && unlikely((uptrval)(op)+length<(uptrval)op)) goto _output_error; /* overflow detection */ + size_t const addl = read_variable_length(&ip, iend - LASTLITERALS + 1, 0); + if (addl == rvl_error) { goto _output_error; } + length += addl; + if (unlikely((uptrval)(op)+length<(uptrval)op)) goto _output_error; /* overflow detection */ } length += MINMATCH; @@ -2073,6 +2250,7 @@ if ((checkOffset) && (unlikely(match + dictSize < lowPrefix))) goto _output_error; /* Error : offset outside buffers */ /* match starting within external dictionary */ if ((dict==usingExtDict) && (match < lowPrefix)) { + assert(dictEnd != NULL); if (unlikely(op+length > oend-LASTLITERALS)) { if (partialDecoding) length = MIN(length, (size_t)(oend-op)); else goto _output_error; /* doesn't respect parsing restriction */ @@ -2080,7 +2258,7 @@ if (length <= (size_t)(lowPrefix-match)) { /* match fits entirely within external dictionary : just copy */ - memmove(op, dictEnd - (lowPrefix-match), length); + LZ4_memmove(op, dictEnd - (lowPrefix-match), length); op += length; } else { /* match stretches into both external dictionary and current block */ @@ -2151,12 +2329,8 @@ } /* end of decoding */ - if (endOnInput) { - DEBUGLOG(5, "decoded %i bytes", (int) (((char*)op)-dst)); - return (int) (((char*)op)-dst); /* Nb of output bytes decoded */ - } else { - return (int) (((const char*)ip)-src); /* Nb of input bytes read */ - } + DEBUGLOG(5, "decoded %i bytes", (int) (((char*)op)-dst)); + return (int) (((char*)op)-dst); /* Nb of output bytes decoded */ /* Overflow error detected */ _output_error: @@ -2171,7 +2345,7 @@ int LZ4_decompress_safe(const char* source, char* dest, int compressedSize, int maxDecompressedSize) { return LZ4_decompress_generic(source, dest, compressedSize, maxDecompressedSize, - endOnInputSize, decode_full_block, noDict, + decode_full_block, noDict, (BYTE*)dest, NULL, 0); } @@ -2180,16 +2354,17 @@ { dstCapacity = MIN(targetOutputSize, dstCapacity); return LZ4_decompress_generic(src, dst, compressedSize, dstCapacity, - endOnInputSize, partial_decode, + partial_decode, noDict, (BYTE*)dst, NULL, 0); } LZ4_FORCE_O2 int LZ4_decompress_fast(const char* source, char* dest, int originalSize) { - return LZ4_decompress_generic(source, dest, 0, originalSize, - endOnOutputSize, decode_full_block, withPrefix64k, - (BYTE*)dest - 64 KB, NULL, 0); + DEBUGLOG(5, "LZ4_decompress_fast"); + return LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + 0, NULL, 0); } /*===== Instantiate a few more decoding cases, used more than once. =====*/ @@ -2198,16 +2373,25 @@ int LZ4_decompress_safe_withPrefix64k(const char* source, char* dest, int compressedSize, int maxOutputSize) { return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, - endOnInputSize, decode_full_block, withPrefix64k, + decode_full_block, withPrefix64k, (BYTE*)dest - 64 KB, NULL, 0); } +LZ4_FORCE_O2 +static int LZ4_decompress_safe_partial_withPrefix64k(const char* source, char* dest, int compressedSize, int targetOutputSize, int dstCapacity) +{ + dstCapacity = MIN(targetOutputSize, dstCapacity); + return LZ4_decompress_generic(source, dest, compressedSize, dstCapacity, + partial_decode, withPrefix64k, + (BYTE*)dest - 64 KB, NULL, 0); +} + /* Another obsolete API function, paired with the previous one. */ int LZ4_decompress_fast_withPrefix64k(const char* source, char* dest, int originalSize) { - /* LZ4_decompress_fast doesn't validate match offsets, - * and thus serves well with any prefixed dictionary. */ - return LZ4_decompress_fast(source, dest, originalSize); + return LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + 64 KB, NULL, 0); } LZ4_FORCE_O2 @@ -2215,27 +2399,48 @@ size_t prefixSize) { return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, - endOnInputSize, decode_full_block, noDict, + decode_full_block, noDict, (BYTE*)dest-prefixSize, NULL, 0); } LZ4_FORCE_O2 +static int LZ4_decompress_safe_partial_withSmallPrefix(const char* source, char* dest, int compressedSize, int targetOutputSize, int dstCapacity, + size_t prefixSize) +{ + dstCapacity = MIN(targetOutputSize, dstCapacity); + return LZ4_decompress_generic(source, dest, compressedSize, dstCapacity, + partial_decode, noDict, + (BYTE*)dest-prefixSize, NULL, 0); +} + +LZ4_FORCE_O2 int LZ4_decompress_safe_forceExtDict(const char* source, char* dest, int compressedSize, int maxOutputSize, const void* dictStart, size_t dictSize) { return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, - endOnInputSize, decode_full_block, usingExtDict, + decode_full_block, usingExtDict, (BYTE*)dest, (const BYTE*)dictStart, dictSize); } LZ4_FORCE_O2 +int LZ4_decompress_safe_partial_forceExtDict(const char* source, char* dest, + int compressedSize, int targetOutputSize, int dstCapacity, + const void* dictStart, size_t dictSize) +{ + dstCapacity = MIN(targetOutputSize, dstCapacity); + return LZ4_decompress_generic(source, dest, compressedSize, dstCapacity, + partial_decode, usingExtDict, + (BYTE*)dest, (const BYTE*)dictStart, dictSize); +} + +LZ4_FORCE_O2 static int LZ4_decompress_fast_extDict(const char* source, char* dest, int originalSize, const void* dictStart, size_t dictSize) { - return LZ4_decompress_generic(source, dest, 0, originalSize, - endOnOutputSize, decode_full_block, usingExtDict, - (BYTE*)dest, (const BYTE*)dictStart, dictSize); + return LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + 0, (const BYTE*)dictStart, dictSize); } /* The "double dictionary" mode, for use with e.g. ring buffers: the first part @@ -2247,26 +2452,17 @@ size_t prefixSize, const void* dictStart, size_t dictSize) { return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, - endOnInputSize, decode_full_block, usingExtDict, - (BYTE*)dest-prefixSize, (const BYTE*)dictStart, dictSize); -} - -LZ4_FORCE_INLINE -int LZ4_decompress_fast_doubleDict(const char* source, char* dest, int originalSize, - size_t prefixSize, const void* dictStart, size_t dictSize) -{ - return LZ4_decompress_generic(source, dest, 0, originalSize, - endOnOutputSize, decode_full_block, usingExtDict, + decode_full_block, usingExtDict, (BYTE*)dest-prefixSize, (const BYTE*)dictStart, dictSize); } /*===== streaming decompression functions =====*/ +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) LZ4_streamDecode_t* LZ4_createStreamDecode(void) { - LZ4_streamDecode_t* lz4s = (LZ4_streamDecode_t*) ALLOC_AND_ZERO(sizeof(LZ4_streamDecode_t)); - LZ4_STATIC_ASSERT(LZ4_STREAMDECODESIZE >= sizeof(LZ4_streamDecode_t_internal)); /* A compilation error here means LZ4_STREAMDECODESIZE is not large enough */ - return lz4s; + LZ4_STATIC_ASSERT(sizeof(LZ4_streamDecode_t) >= sizeof(LZ4_streamDecode_t_internal)); + return (LZ4_streamDecode_t*) ALLOC_AND_ZERO(sizeof(LZ4_streamDecode_t)); } int LZ4_freeStreamDecode (LZ4_streamDecode_t* LZ4_stream) @@ -2275,6 +2471,7 @@ FREEMEM(LZ4_stream); return 0; } +#endif /*! LZ4_setStreamDecode() : * Use this function to instruct where to find the dictionary. @@ -2285,8 +2482,13 @@ int LZ4_setStreamDecode (LZ4_streamDecode_t* LZ4_streamDecode, const char* dictionary, int dictSize) { LZ4_streamDecode_t_internal* lz4sd = &LZ4_streamDecode->internal_donotuse; - lz4sd->prefixSize = (size_t) dictSize; - lz4sd->prefixEnd = (const BYTE*) dictionary + dictSize; + lz4sd->prefixSize = (size_t)dictSize; + if (dictSize) { + assert(dictionary != NULL); + lz4sd->prefixEnd = (const BYTE*) dictionary + dictSize; + } else { + lz4sd->prefixEnd = (const BYTE*) dictionary; + } lz4sd->externalDict = NULL; lz4sd->extDictSize = 0; return 1; @@ -2358,29 +2560,35 @@ return result; } -LZ4_FORCE_O2 -int LZ4_decompress_fast_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int originalSize) +LZ4_FORCE_O2 int +LZ4_decompress_fast_continue (LZ4_streamDecode_t* LZ4_streamDecode, + const char* source, char* dest, int originalSize) { - LZ4_streamDecode_t_internal* lz4sd = &LZ4_streamDecode->internal_donotuse; + LZ4_streamDecode_t_internal* const lz4sd = + (assert(LZ4_streamDecode!=NULL), &LZ4_streamDecode->internal_donotuse); int result; + + DEBUGLOG(5, "LZ4_decompress_fast_continue (toDecodeSize=%i)", originalSize); assert(originalSize >= 0); if (lz4sd->prefixSize == 0) { + DEBUGLOG(5, "first invocation : no prefix nor extDict"); assert(lz4sd->extDictSize == 0); result = LZ4_decompress_fast(source, dest, originalSize); if (result <= 0) return result; lz4sd->prefixSize = (size_t)originalSize; lz4sd->prefixEnd = (BYTE*)dest + originalSize; } else if (lz4sd->prefixEnd == (BYTE*)dest) { - if (lz4sd->prefixSize >= 64 KB - 1 || lz4sd->extDictSize == 0) - result = LZ4_decompress_fast(source, dest, originalSize); - else - result = LZ4_decompress_fast_doubleDict(source, dest, originalSize, - lz4sd->prefixSize, lz4sd->externalDict, lz4sd->extDictSize); + DEBUGLOG(5, "continue using existing prefix"); + result = LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + lz4sd->prefixSize, + lz4sd->externalDict, lz4sd->extDictSize); if (result <= 0) return result; lz4sd->prefixSize += (size_t)originalSize; lz4sd->prefixEnd += originalSize; } else { + DEBUGLOG(5, "prefix becomes extDict"); lz4sd->extDictSize = lz4sd->prefixSize; lz4sd->externalDict = lz4sd->prefixEnd - lz4sd->extDictSize; result = LZ4_decompress_fast_extDict(source, dest, originalSize, @@ -2416,10 +2624,27 @@ return LZ4_decompress_safe_forceExtDict(source, dest, compressedSize, maxOutputSize, dictStart, (size_t)dictSize); } +int LZ4_decompress_safe_partial_usingDict(const char* source, char* dest, int compressedSize, int targetOutputSize, int dstCapacity, const char* dictStart, int dictSize) +{ + if (dictSize==0) + return LZ4_decompress_safe_partial(source, dest, compressedSize, targetOutputSize, dstCapacity); + if (dictStart+dictSize == dest) { + if (dictSize >= 64 KB - 1) { + return LZ4_decompress_safe_partial_withPrefix64k(source, dest, compressedSize, targetOutputSize, dstCapacity); + } + assert(dictSize >= 0); + return LZ4_decompress_safe_partial_withSmallPrefix(source, dest, compressedSize, targetOutputSize, dstCapacity, (size_t)dictSize); + } + assert(dictSize >= 0); + return LZ4_decompress_safe_partial_forceExtDict(source, dest, compressedSize, targetOutputSize, dstCapacity, dictStart, (size_t)dictSize); +} + int LZ4_decompress_fast_usingDict(const char* source, char* dest, int originalSize, const char* dictStart, int dictSize) { if (dictSize==0 || dictStart+dictSize == dest) - return LZ4_decompress_fast(source, dest, originalSize); + return LZ4_decompress_unsafe_generic( + (const BYTE*)source, (BYTE*)dest, originalSize, + (size_t)dictSize, NULL, 0); assert(dictSize >= 0); return LZ4_decompress_fast_extDict(source, dest, originalSize, dictStart, (size_t)dictSize); } @@ -2471,7 +2696,7 @@ /* Obsolete Streaming functions */ -int LZ4_sizeofStreamState(void) { return LZ4_STREAMSIZE; } +int LZ4_sizeofStreamState(void) { return sizeof(LZ4_stream_t); } int LZ4_resetStreamState(void* state, char* inputBuffer) { @@ -2480,11 +2705,13 @@ return 0; } +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) void* LZ4_create (char* inputBuffer) { (void)inputBuffer; return LZ4_createStream(); } +#endif char* LZ4_slideInputBuffer (void* state) { diff --git a/cbits/lz4.h b/cbits/lz4.h --- a/cbits/lz4.h +++ b/cbits/lz4.h @@ -1,7 +1,7 @@ /* * LZ4 - Fast LZ compression algorithm * Header File - * Copyright (C) 2011-present, Yann Collet. + * Copyright (C) 2011-2020, Yann Collet. BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) @@ -97,37 +97,78 @@ # define LZ4LIB_API LZ4LIB_VISIBILITY #endif +/*! LZ4_FREESTANDING : + * When this macro is set to 1, it enables "freestanding mode" that is + * suitable for typical freestanding environment which doesn't support + * standard C library. + * + * - LZ4_FREESTANDING is a compile-time switch. + * - It requires the following macros to be defined: + * LZ4_memcpy, LZ4_memmove, LZ4_memset. + * - It only enables LZ4/HC functions which don't use heap. + * All LZ4F_* functions are not supported. + * - See tests/freestanding.c to check its basic setup. + */ +#if defined(LZ4_FREESTANDING) && (LZ4_FREESTANDING == 1) +# define LZ4_HEAPMODE 0 +# define LZ4HC_HEAPMODE 0 +# define LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION 1 +# if !defined(LZ4_memcpy) +# error "LZ4_FREESTANDING requires macro 'LZ4_memcpy'." +# endif +# if !defined(LZ4_memset) +# error "LZ4_FREESTANDING requires macro 'LZ4_memset'." +# endif +# if !defined(LZ4_memmove) +# error "LZ4_FREESTANDING requires macro 'LZ4_memmove'." +# endif +#elif ! defined(LZ4_FREESTANDING) +# define LZ4_FREESTANDING 0 +#endif + + /*------ Version ------*/ #define LZ4_VERSION_MAJOR 1 /* for breaking interface changes */ #define LZ4_VERSION_MINOR 9 /* for new (non-breaking) interface capabilities */ -#define LZ4_VERSION_RELEASE 3 /* for tweaks, bug-fixes, or development */ +#define LZ4_VERSION_RELEASE 4 /* for tweaks, bug-fixes, or development */ #define LZ4_VERSION_NUMBER (LZ4_VERSION_MAJOR *100*100 + LZ4_VERSION_MINOR *100 + LZ4_VERSION_RELEASE) #define LZ4_LIB_VERSION LZ4_VERSION_MAJOR.LZ4_VERSION_MINOR.LZ4_VERSION_RELEASE #define LZ4_QUOTE(str) #str #define LZ4_EXPAND_AND_QUOTE(str) LZ4_QUOTE(str) -#define LZ4_VERSION_STRING LZ4_EXPAND_AND_QUOTE(LZ4_LIB_VERSION) +#define LZ4_VERSION_STRING LZ4_EXPAND_AND_QUOTE(LZ4_LIB_VERSION) /* requires v1.7.3+ */ -LZ4LIB_API int LZ4_versionNumber (void); /**< library version number; useful to check dll version */ -LZ4LIB_API const char* LZ4_versionString (void); /**< library version string; useful to check dll version */ +LZ4LIB_API int LZ4_versionNumber (void); /**< library version number; useful to check dll version; requires v1.3.0+ */ +LZ4LIB_API const char* LZ4_versionString (void); /**< library version string; useful to check dll version; requires v1.7.5+ */ /*-************************************ * Tuning parameter **************************************/ +#define LZ4_MEMORY_USAGE_MIN 10 +#define LZ4_MEMORY_USAGE_DEFAULT 14 +#define LZ4_MEMORY_USAGE_MAX 20 + /*! * LZ4_MEMORY_USAGE : - * Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; etc.) - * Increasing memory usage improves compression ratio. - * Reduced memory usage may improve speed, thanks to better cache locality. + * Memory usage formula : N->2^N Bytes (examples : 10 -> 1KB; 12 -> 4KB ; 16 -> 64KB; 20 -> 1MB; ) + * Increasing memory usage improves compression ratio, at the cost of speed. + * Reduced memory usage may improve speed at the cost of ratio, thanks to better cache locality. * Default value is 14, for 16KB, which nicely fits into Intel x86 L1 cache */ #ifndef LZ4_MEMORY_USAGE -# define LZ4_MEMORY_USAGE 14 +# define LZ4_MEMORY_USAGE LZ4_MEMORY_USAGE_DEFAULT #endif +#if (LZ4_MEMORY_USAGE < LZ4_MEMORY_USAGE_MIN) +# error "LZ4_MEMORY_USAGE is too small !" +#endif +#if (LZ4_MEMORY_USAGE > LZ4_MEMORY_USAGE_MAX) +# error "LZ4_MEMORY_USAGE is too large !" +#endif + /*-************************************ * Simple Functions **************************************/ @@ -270,8 +311,25 @@ ***********************************************/ typedef union LZ4_stream_u LZ4_stream_t; /* incomplete type (defined later) */ +/** + Note about RC_INVOKED + + - RC_INVOKED is predefined symbol of rc.exe (the resource compiler which is part of MSVC/Visual Studio). + https://docs.microsoft.com/en-us/windows/win32/menurc/predefined-macros + + - Since rc.exe is a legacy compiler, it truncates long symbol (> 30 chars) + and reports warning "RC4011: identifier truncated". + + - To eliminate the warning, we surround long preprocessor symbol with + "#if !defined(RC_INVOKED) ... #endif" block that means + "skip this block when rc.exe is trying to read it". +*/ +#if !defined(RC_INVOKED) /* https://docs.microsoft.com/en-us/windows/win32/menurc/predefined-macros */ +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) LZ4LIB_API LZ4_stream_t* LZ4_createStream(void); LZ4LIB_API int LZ4_freeStream (LZ4_stream_t* streamPtr); +#endif /* !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) */ +#endif /*! LZ4_resetStream_fast() : v1.9.0+ * Use this to prepare an LZ4_stream_t for a new chain of dependent blocks @@ -355,8 +413,12 @@ * creation / destruction of streaming decompression tracking context. * A tracking context can be re-used multiple times. */ +#if !defined(RC_INVOKED) /* https://docs.microsoft.com/en-us/windows/win32/menurc/predefined-macros */ +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) LZ4LIB_API LZ4_streamDecode_t* LZ4_createStreamDecode(void); LZ4LIB_API int LZ4_freeStreamDecode (LZ4_streamDecode_t* LZ4_stream); +#endif /* !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) */ +#endif /*! LZ4_setStreamDecode() : * An LZ4_streamDecode_t context can be allocated once and re-used multiple times. @@ -406,7 +468,10 @@ * save the last 64KB of decoded data into a safe buffer where it can't be modified during decompression, * then indicate where this data is saved using LZ4_setStreamDecode(), before decompressing next block. */ -LZ4LIB_API int LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* src, char* dst, int srcSize, int dstCapacity); +LZ4LIB_API int +LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, + const char* src, char* dst, + int srcSize, int dstCapacity); /*! LZ4_decompress_*_usingDict() : @@ -417,8 +482,17 @@ * Performance tip : Decompression speed can be substantially increased * when dst == dictStart + dictSize. */ -LZ4LIB_API int LZ4_decompress_safe_usingDict (const char* src, char* dst, int srcSize, int dstCapcity, const char* dictStart, int dictSize); +LZ4LIB_API int +LZ4_decompress_safe_usingDict(const char* src, char* dst, + int srcSize, int dstCapacity, + const char* dictStart, int dictSize); +LZ4LIB_API int +LZ4_decompress_safe_partial_usingDict(const char* src, char* dst, + int compressedSize, + int targetOutputSize, int maxOutputSize, + const char* dictStart, int dictSize); + #endif /* LZ4_H_2983827168210 */ @@ -496,13 +570,15 @@ * stream (and source buffer) must remain in-place / accessible / unchanged * through the completion of the first compression call on the stream. */ -LZ4LIB_STATIC_API void LZ4_attach_dictionary(LZ4_stream_t* workingStream, const LZ4_stream_t* dictionaryStream); +LZ4LIB_STATIC_API void +LZ4_attach_dictionary(LZ4_stream_t* workingStream, + const LZ4_stream_t* dictionaryStream); /*! In-place compression and decompression * * It's possible to have input and output sharing the same buffer, - * for highly contrained memory environments. + * for highly constrained memory environments. * In both cases, it requires input to lay at the end of the buffer, * and decompression to start at beginning of the buffer. * Buffer size must feature some margin, hence be larger than final size. @@ -592,38 +668,26 @@ typedef unsigned int LZ4_u32; #endif +/*! LZ4_stream_t : + * Never ever use below internal definitions directly ! + * These definitions are not API/ABI safe, and may change in future versions. + * If you need static allocation, declare or allocate an LZ4_stream_t object. +**/ + typedef struct LZ4_stream_t_internal LZ4_stream_t_internal; struct LZ4_stream_t_internal { LZ4_u32 hashTable[LZ4_HASH_SIZE_U32]; - LZ4_u32 currentOffset; - LZ4_u32 tableType; const LZ4_byte* dictionary; const LZ4_stream_t_internal* dictCtx; + LZ4_u32 currentOffset; + LZ4_u32 tableType; LZ4_u32 dictSize; + /* Implicit padding to ensure structure is aligned */ }; -typedef struct { - const LZ4_byte* externalDict; - size_t extDictSize; - const LZ4_byte* prefixEnd; - size_t prefixSize; -} LZ4_streamDecode_t_internal; - - -/*! LZ4_stream_t : - * Do not use below internal definitions directly ! - * Declare or allocate an LZ4_stream_t instead. - * LZ4_stream_t can also be created using LZ4_createStream(), which is recommended. - * The structure definition can be convenient for static allocation - * (on stack, or as part of larger structure). - * Init this structure with LZ4_initStream() before first use. - * note : only use this definition in association with static linking ! - * this definition is not API/ABI safe, and may change in future versions. - */ -#define LZ4_STREAMSIZE 16416 /* static size, for inter-version compatibility */ -#define LZ4_STREAMSIZE_VOIDP (LZ4_STREAMSIZE / sizeof(void*)) +#define LZ4_STREAM_MINSIZE ((1UL << LZ4_MEMORY_USAGE) + 32) /* static size, for inter-version compatibility */ union LZ4_stream_u { - void* table[LZ4_STREAMSIZE_VOIDP]; + char minStateSize[LZ4_STREAM_MINSIZE]; LZ4_stream_t_internal internal_donotuse; }; /* previously typedef'd to LZ4_stream_t */ @@ -641,21 +705,25 @@ * In which case, the function will @return NULL. * Note2: An LZ4_stream_t structure guarantees correct alignment and size. * Note3: Before v1.9.0, use LZ4_resetStream() instead - */ +**/ LZ4LIB_API LZ4_stream_t* LZ4_initStream (void* buffer, size_t size); /*! LZ4_streamDecode_t : - * information structure to track an LZ4 stream during decompression. - * init this structure using LZ4_setStreamDecode() before first use. - * note : only use in association with static linking ! - * this definition is not API/ABI safe, - * and may change in a future version ! - */ -#define LZ4_STREAMDECODESIZE_U64 (4 + ((sizeof(void*)==16) ? 2 : 0) /*AS-400*/ ) -#define LZ4_STREAMDECODESIZE (LZ4_STREAMDECODESIZE_U64 * sizeof(unsigned long long)) + * Never ever use below internal definitions directly ! + * These definitions are not API/ABI safe, and may change in future versions. + * If you need static allocation, declare or allocate an LZ4_streamDecode_t object. +**/ +typedef struct { + const LZ4_byte* externalDict; + const LZ4_byte* prefixEnd; + size_t extDictSize; + size_t prefixSize; +} LZ4_streamDecode_t_internal; + +#define LZ4_STREAMDECODE_MINSIZE 32 union LZ4_streamDecode_u { - unsigned long long table[LZ4_STREAMDECODESIZE_U64]; + char minStateSize[LZ4_STREAMDECODE_MINSIZE]; LZ4_streamDecode_t_internal internal_donotuse; } ; /* previously typedef'd to LZ4_streamDecode_t */ diff --git a/cbits/lz4frame.c b/cbits/lz4frame.c --- a/cbits/lz4frame.c +++ b/cbits/lz4frame.c @@ -45,7 +45,7 @@ * Compiler Options **************************************/ #ifdef _MSC_VER /* Visual Studio */ -# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ +# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ #endif @@ -63,6 +63,19 @@ /*-************************************ +* Library declarations +**************************************/ +#define LZ4F_STATIC_LINKING_ONLY +#include "lz4frame.h" +#define LZ4_STATIC_LINKING_ONLY +#include "lz4.h" +#define LZ4_HC_STATIC_LINKING_ONLY +#include "lz4hc.h" +#define XXH_STATIC_LINKING_ONLY +#include "xxhash.h" + + +/*-************************************ * Memory routines **************************************/ /* @@ -70,7 +83,13 @@ * malloc(), calloc() and free() * towards another library or solution of their choice * by modifying below section. - */ +**/ + +#include /* memset, memcpy, memmove */ +#ifndef LZ4_SRC_INCLUDED /* avoid redefinition when sources are coalesced */ +# define MEM_INIT(p,v,s) memset((p),(v),(s)) +#endif + #ifndef LZ4_SRC_INCLUDED /* avoid redefinition when sources are coalesced */ # include /* malloc, calloc, free */ # define ALLOC(s) malloc(s) @@ -78,23 +97,42 @@ # define FREEMEM(p) free(p) #endif -#include /* memset, memcpy, memmove */ -#ifndef LZ4_SRC_INCLUDED /* avoid redefinition when sources are coalesced */ -# define MEM_INIT(p,v,s) memset((p),(v),(s)) -#endif +static void* LZ4F_calloc(size_t s, LZ4F_CustomMem cmem) +{ + /* custom calloc defined : use it */ + if (cmem.customCalloc != NULL) { + return cmem.customCalloc(cmem.opaqueState, s); + } + /* nothing defined : use default 's calloc() */ + if (cmem.customAlloc == NULL) { + return ALLOC_AND_ZERO(s); + } + /* only custom alloc defined : use it, and combine it with memset() */ + { void* const p = cmem.customAlloc(cmem.opaqueState, s); + if (p != NULL) MEM_INIT(p, 0, s); + return p; +} } +static void* LZ4F_malloc(size_t s, LZ4F_CustomMem cmem) +{ + /* custom malloc defined : use it */ + if (cmem.customAlloc != NULL) { + return cmem.customAlloc(cmem.opaqueState, s); + } + /* nothing defined : use default 's malloc() */ + return ALLOC(s); +} -/*-************************************ -* Library declarations -**************************************/ -#define LZ4F_STATIC_LINKING_ONLY -#include "lz4frame.h" -#define LZ4_STATIC_LINKING_ONLY -#include "lz4.h" -#define LZ4_HC_STATIC_LINKING_ONLY -#include "lz4hc.h" -#define XXH_STATIC_LINKING_ONLY -#include "xxhash.h" +static void LZ4F_free(void* p, LZ4F_CustomMem cmem) +{ + /* custom malloc defined : use it */ + if (cmem.customFree != NULL) { + cmem.customFree(cmem.opaqueState, p); + return; + } + /* nothing defined : use default 's free() */ + FREEMEM(p); +} /*-************************************ @@ -143,7 +181,7 @@ #endif -/* unoptimized version; solves endianess & alignment issues */ +/* unoptimized version; solves endianness & alignment issues */ static U32 LZ4F_readLE32 (const void* src) { const BYTE* const srcPtr = (const BYTE*)src; @@ -206,8 +244,6 @@ #define _4BITS 0x0F #define _8BITS 0xFF -#define LZ4F_MAGIC_SKIPPABLE_START 0x184D2A50U -#define LZ4F_MAGICNUMBER 0x184D2204U #define LZ4F_BLOCKUNCOMPRESSED_FLAG 0x80000000U #define LZ4F_BLOCKSIZEID_DEFAULT LZ4F_max64KB @@ -220,22 +256,27 @@ /*-************************************ * Structures and local types **************************************/ + +typedef enum { LZ4B_COMPRESSED, LZ4B_UNCOMPRESSED} LZ4F_blockCompression_t; + typedef struct LZ4F_cctx_s { + LZ4F_CustomMem cmem; LZ4F_preferences_t prefs; U32 version; U32 cStage; const LZ4F_CDict* cdict; size_t maxBlockSize; size_t maxBufferSize; - BYTE* tmpBuff; - BYTE* tmpIn; - size_t tmpInSize; + BYTE* tmpBuff; /* internal buffer, for streaming */ + BYTE* tmpIn; /* starting position of data compress within internal buffer (>= tmpBuff) */ + size_t tmpInSize; /* amount of data to compress after tmpIn */ U64 totalInSize; XXH32_state_t xxh; void* lz4CtxPtr; U16 lz4CtxAlloc; /* sized for: 0 = none, 1 = lz4 ctx, 2 = lz4hc ctx */ U16 lz4CtxState; /* in use as: 0 = none, 1 = lz4 ctx, 2 = lz4hc ctx */ + LZ4F_blockCompression_t blockCompression; } LZ4F_cctx_t; @@ -264,27 +305,33 @@ return (LZ4F_errorCodes)(-(ptrdiff_t)functionResult); } -static LZ4F_errorCode_t err0r(LZ4F_errorCodes code) +static LZ4F_errorCode_t LZ4F_returnErrorCode(LZ4F_errorCodes code) { /* A compilation error here means sizeof(ptrdiff_t) is not large enough */ LZ4F_STATIC_ASSERT(sizeof(ptrdiff_t) >= sizeof(size_t)); return (LZ4F_errorCode_t)-(ptrdiff_t)code; } +#define RETURN_ERROR(e) return LZ4F_returnErrorCode(LZ4F_ERROR_ ## e) + +#define RETURN_ERROR_IF(c,e) if (c) RETURN_ERROR(e) + +#define FORWARD_IF_ERROR(r) if (LZ4F_isError(r)) return (r) + unsigned LZ4F_getVersion(void) { return LZ4F_VERSION; } int LZ4F_compressionLevel_max(void) { return LZ4HC_CLEVEL_MAX; } -size_t LZ4F_getBlockSize(unsigned blockSizeID) +size_t LZ4F_getBlockSize(LZ4F_blockSizeID_t blockSizeID) { static const size_t blockSizes[4] = { 64 KB, 256 KB, 1 MB, 4 MB }; if (blockSizeID == 0) blockSizeID = LZ4F_BLOCKSIZEID_DEFAULT; if (blockSizeID < LZ4F_max64KB || blockSizeID > LZ4F_max4MB) - return err0r(LZ4F_ERROR_maxBlockSize_invalid); - blockSizeID -= LZ4F_max64KB; - return blockSizes[blockSizeID]; -} + RETURN_ERROR(maxBlockSize_invalid); + { int const blockSizeIdx = (int)blockSizeID - (int)LZ4F_max64KB; + return blockSizes[blockSizeIdx]; +} } /*-************************************ * Private functions @@ -397,21 +444,20 @@ MEM_INIT(&options, 0, sizeof(options)); options.stableSrc = 1; - if (dstCapacity < LZ4F_compressFrameBound(srcSize, &prefs)) /* condition to guarantee success */ - return err0r(LZ4F_ERROR_dstMaxSize_tooSmall); + RETURN_ERROR_IF(dstCapacity < LZ4F_compressFrameBound(srcSize, &prefs), dstMaxSize_tooSmall); { size_t const headerSize = LZ4F_compressBegin_usingCDict(cctx, dstBuffer, dstCapacity, cdict, &prefs); /* write header */ - if (LZ4F_isError(headerSize)) return headerSize; + FORWARD_IF_ERROR(headerSize); dstPtr += headerSize; /* header size */ } assert(dstEnd >= dstPtr); { size_t const cSize = LZ4F_compressUpdate(cctx, dstPtr, (size_t)(dstEnd-dstPtr), srcBuffer, srcSize, &options); - if (LZ4F_isError(cSize)) return cSize; + FORWARD_IF_ERROR(cSize); dstPtr += cSize; } assert(dstEnd >= dstPtr); { size_t const tailSize = LZ4F_compressEnd(cctx, dstPtr, (size_t)(dstEnd-dstPtr), &options); /* flush last block, and generate suffix */ - if (LZ4F_isError(tailSize)) return tailSize; + FORWARD_IF_ERROR(tailSize); dstPtr += tailSize; } assert(dstEnd >= dstStart); @@ -432,27 +478,26 @@ { size_t result; #if (LZ4F_HEAPMODE) - LZ4F_cctx_t *cctxPtr; + LZ4F_cctx_t* cctxPtr; result = LZ4F_createCompressionContext(&cctxPtr, LZ4F_VERSION); - if (LZ4F_isError(result)) return result; + FORWARD_IF_ERROR(result); #else LZ4F_cctx_t cctx; LZ4_stream_t lz4ctx; - LZ4F_cctx_t *cctxPtr = &cctx; + LZ4F_cctx_t* const cctxPtr = &cctx; - DEBUGLOG(4, "LZ4F_compressFrame"); MEM_INIT(&cctx, 0, sizeof(cctx)); cctx.version = LZ4F_VERSION; cctx.maxBufferSize = 5 MB; /* mess with real buffer size to prevent dynamic allocation; works only because autoflush==1 & stableSrc==1 */ - if (preferencesPtr == NULL || - preferencesPtr->compressionLevel < LZ4HC_CLEVEL_MIN) - { + if ( preferencesPtr == NULL + || preferencesPtr->compressionLevel < LZ4HC_CLEVEL_MIN ) { LZ4_initStream(&lz4ctx, sizeof(lz4ctx)); cctxPtr->lz4CtxPtr = &lz4ctx; cctxPtr->lz4CtxAlloc = 1; cctxPtr->lz4CtxState = 1; } #endif + DEBUGLOG(4, "LZ4F_compressFrame"); result = LZ4F_compressFrame_usingCDict(cctxPtr, dstBuffer, dstCapacity, srcBuffer, srcSize, @@ -461,10 +506,9 @@ #if (LZ4F_HEAPMODE) LZ4F_freeCompressionContext(cctxPtr); #else - if (preferencesPtr != NULL && - preferencesPtr->compressionLevel >= LZ4HC_CLEVEL_MIN) - { - FREEMEM(cctxPtr->lz4CtxPtr); + if ( preferencesPtr != NULL + && preferencesPtr->compressionLevel >= LZ4HC_CLEVEL_MIN ) { + LZ4F_free(cctxPtr->lz4CtxPtr, cctxPtr->cmem); } #endif return result; @@ -476,30 +520,31 @@ *****************************************************/ struct LZ4F_CDict_s { + LZ4F_CustomMem cmem; void* dictContent; LZ4_stream_t* fastCtx; LZ4_streamHC_t* HCCtx; }; /* typedef'd to LZ4F_CDict within lz4frame_static.h */ -/*! LZ4F_createCDict() : - * When compressing multiple messages / blocks with the same dictionary, it's recommended to load it just once. - * LZ4F_createCDict() will create a digested dictionary, ready to start future compression operations without startup delay. - * LZ4F_CDict can be created once and shared by multiple threads concurrently, since its usage is read-only. - * `dictBuffer` can be released after LZ4F_CDict creation, since its content is copied within CDict - * @return : digested dictionary for compression, or NULL if failed */ -LZ4F_CDict* LZ4F_createCDict(const void* dictBuffer, size_t dictSize) +LZ4F_CDict* +LZ4F_createCDict_advanced(LZ4F_CustomMem cmem, const void* dictBuffer, size_t dictSize) { const char* dictStart = (const char*)dictBuffer; - LZ4F_CDict* cdict = (LZ4F_CDict*) ALLOC(sizeof(*cdict)); - DEBUGLOG(4, "LZ4F_createCDict"); + LZ4F_CDict* const cdict = (LZ4F_CDict*)LZ4F_malloc(sizeof(*cdict), cmem); + DEBUGLOG(4, "LZ4F_createCDict_advanced"); if (!cdict) return NULL; + cdict->cmem = cmem; if (dictSize > 64 KB) { dictStart += dictSize - 64 KB; dictSize = 64 KB; } - cdict->dictContent = ALLOC(dictSize); - cdict->fastCtx = LZ4_createStream(); - cdict->HCCtx = LZ4_createStreamHC(); + cdict->dictContent = LZ4F_malloc(dictSize, cmem); + cdict->fastCtx = (LZ4_stream_t*)LZ4F_malloc(sizeof(LZ4_stream_t), cmem); + if (cdict->fastCtx) + LZ4_initStream(cdict->fastCtx, sizeof(LZ4_stream_t)); + cdict->HCCtx = (LZ4_streamHC_t*)LZ4F_malloc(sizeof(LZ4_streamHC_t), cmem); + if (cdict->HCCtx) + LZ4_initStream(cdict->HCCtx, sizeof(LZ4_streamHC_t)); if (!cdict->dictContent || !cdict->fastCtx || !cdict->HCCtx) { LZ4F_freeCDict(cdict); return NULL; @@ -511,13 +556,25 @@ return cdict; } +/*! LZ4F_createCDict() : + * When compressing multiple messages / blocks with the same dictionary, it's recommended to load it just once. + * LZ4F_createCDict() will create a digested dictionary, ready to start future compression operations without startup delay. + * LZ4F_CDict can be created once and shared by multiple threads concurrently, since its usage is read-only. + * @dictBuffer can be released after LZ4F_CDict creation, since its content is copied within CDict + * @return : digested dictionary for compression, or NULL if failed */ +LZ4F_CDict* LZ4F_createCDict(const void* dictBuffer, size_t dictSize) +{ + DEBUGLOG(4, "LZ4F_createCDict"); + return LZ4F_createCDict_advanced(LZ4F_defaultCMem, dictBuffer, dictSize); +} + void LZ4F_freeCDict(LZ4F_CDict* cdict) { if (cdict==NULL) return; /* support free on NULL */ - FREEMEM(cdict->dictContent); - LZ4_freeStream(cdict->fastCtx); - LZ4_freeStreamHC(cdict->HCCtx); - FREEMEM(cdict); + LZ4F_free(cdict->dictContent, cdict->cmem); + LZ4F_free(cdict->fastCtx, cdict->cmem); + LZ4F_free(cdict->HCCtx, cdict->cmem); + LZ4F_free(cdict, cdict->cmem); } @@ -525,6 +582,20 @@ * Advanced compression functions ***********************************/ +LZ4F_cctx* +LZ4F_createCompressionContext_advanced(LZ4F_CustomMem customMem, unsigned version) +{ + LZ4F_cctx* const cctxPtr = + (LZ4F_cctx*)LZ4F_calloc(sizeof(LZ4F_cctx), customMem); + if (cctxPtr==NULL) return NULL; + + cctxPtr->cmem = customMem; + cctxPtr->version = version; + cctxPtr->cStage = 0; /* Uninitialized. Next stage : init cctx */ + + return cctxPtr; +} + /*! LZ4F_createCompressionContext() : * The first thing to do is to create a compressionContext object, which will be used in all compression operations. * This is achieved using LZ4F_createCompressionContext(), which takes as argument a version and an LZ4F_preferences_t structure. @@ -532,17 +603,16 @@ * The function will provide a pointer to an allocated LZ4F_compressionContext_t object. * If the result LZ4F_errorCode_t is not OK_NoError, there was an error during context creation. * Object can release its memory using LZ4F_freeCompressionContext(); - */ -LZ4F_errorCode_t LZ4F_createCompressionContext(LZ4F_cctx** LZ4F_compressionContextPtr, unsigned version) +**/ +LZ4F_errorCode_t +LZ4F_createCompressionContext(LZ4F_cctx** LZ4F_compressionContextPtr, unsigned version) { - LZ4F_cctx_t* const cctxPtr = (LZ4F_cctx_t*)ALLOC_AND_ZERO(sizeof(LZ4F_cctx_t)); - if (cctxPtr==NULL) return err0r(LZ4F_ERROR_allocation_failed); - - cctxPtr->version = version; - cctxPtr->cStage = 0; /* Next stage : init stream */ - - *LZ4F_compressionContextPtr = cctxPtr; + assert(LZ4F_compressionContextPtr != NULL); /* considered a violation of narrow contract */ + /* in case it nonetheless happen in production */ + RETURN_ERROR_IF(LZ4F_compressionContextPtr == NULL, parameter_null); + *LZ4F_compressionContextPtr = LZ4F_createCompressionContext_advanced(LZ4F_defaultCMem, version); + RETURN_ERROR_IF(*LZ4F_compressionContextPtr==NULL, allocation_failed); return LZ4F_OK_NoError; } @@ -550,11 +620,10 @@ LZ4F_errorCode_t LZ4F_freeCompressionContext(LZ4F_cctx* cctxPtr) { if (cctxPtr != NULL) { /* support free on NULL */ - FREEMEM(cctxPtr->lz4CtxPtr); /* note: LZ4_streamHC_t and LZ4_stream_t are simple POD types */ - FREEMEM(cctxPtr->tmpBuff); - FREEMEM(cctxPtr); + LZ4F_free(cctxPtr->lz4CtxPtr, cctxPtr->cmem); /* note: LZ4_streamHC_t and LZ4_stream_t are simple POD types */ + LZ4F_free(cctxPtr->tmpBuff, cctxPtr->cmem); + LZ4F_free(cctxPtr, cctxPtr->cmem); } - return LZ4F_OK_NoError; } @@ -588,11 +657,21 @@ } } +static int ctxTypeID_to_size(int ctxTypeID) { + switch(ctxTypeID) { + case 1: + return LZ4_sizeofState(); + case 2: + return LZ4_sizeofStateHC(); + default: + return 0; + } +} /*! LZ4F_compressBegin_usingCDict() : - * init streaming compression and writes frame header into dstBuffer. - * dstBuffer must be >= LZ4F_HEADER_SIZE_MAX bytes. - * @return : number of bytes written into dstBuffer for the header + * init streaming compression AND writes frame header into @dstBuffer. + * @dstCapacity must be >= LZ4F_HEADER_SIZE_MAX bytes. + * @return : number of bytes written into @dstBuffer for the header * or an error code (can be tested using LZ4F_isError()) */ size_t LZ4F_compressBegin_usingCDict(LZ4F_cctx* cctxPtr, @@ -600,41 +679,46 @@ const LZ4F_CDict* cdict, const LZ4F_preferences_t* preferencesPtr) { - LZ4F_preferences_t prefNull; + LZ4F_preferences_t const prefNull = LZ4F_INIT_PREFERENCES; BYTE* const dstStart = (BYTE*)dstBuffer; BYTE* dstPtr = dstStart; - BYTE* headerStart; - if (dstCapacity < maxFHSize) return err0r(LZ4F_ERROR_dstMaxSize_tooSmall); - MEM_INIT(&prefNull, 0, sizeof(prefNull)); + RETURN_ERROR_IF(dstCapacity < maxFHSize, dstMaxSize_tooSmall); if (preferencesPtr == NULL) preferencesPtr = &prefNull; cctxPtr->prefs = *preferencesPtr; - /* Ctx Management */ + /* cctx Management */ { U16 const ctxTypeID = (cctxPtr->prefs.compressionLevel < LZ4HC_CLEVEL_MIN) ? 1 : 2; - if (cctxPtr->lz4CtxAlloc < ctxTypeID) { - FREEMEM(cctxPtr->lz4CtxPtr); + int requiredSize = ctxTypeID_to_size(ctxTypeID); + int allocatedSize = ctxTypeID_to_size(cctxPtr->lz4CtxAlloc); + if (allocatedSize < requiredSize) { + /* not enough space allocated */ + LZ4F_free(cctxPtr->lz4CtxPtr, cctxPtr->cmem); if (cctxPtr->prefs.compressionLevel < LZ4HC_CLEVEL_MIN) { - cctxPtr->lz4CtxPtr = LZ4_createStream(); + /* must take ownership of memory allocation, + * in order to respect custom allocator contract */ + cctxPtr->lz4CtxPtr = LZ4F_malloc(sizeof(LZ4_stream_t), cctxPtr->cmem); + if (cctxPtr->lz4CtxPtr) + LZ4_initStream(cctxPtr->lz4CtxPtr, sizeof(LZ4_stream_t)); } else { - cctxPtr->lz4CtxPtr = LZ4_createStreamHC(); + cctxPtr->lz4CtxPtr = LZ4F_malloc(sizeof(LZ4_streamHC_t), cctxPtr->cmem); + if (cctxPtr->lz4CtxPtr) + LZ4_initStreamHC(cctxPtr->lz4CtxPtr, sizeof(LZ4_streamHC_t)); } - if (cctxPtr->lz4CtxPtr == NULL) - return err0r(LZ4F_ERROR_allocation_failed); + RETURN_ERROR_IF(cctxPtr->lz4CtxPtr == NULL, allocation_failed); cctxPtr->lz4CtxAlloc = ctxTypeID; cctxPtr->lz4CtxState = ctxTypeID; } else if (cctxPtr->lz4CtxState != ctxTypeID) { - /* otherwise, a sufficient buffer is allocated, but we need to - * reset it to the correct context type */ + /* otherwise, a sufficient buffer is already allocated, + * but we need to reset it to the correct context type */ if (cctxPtr->prefs.compressionLevel < LZ4HC_CLEVEL_MIN) { - LZ4_initStream((LZ4_stream_t *) cctxPtr->lz4CtxPtr, sizeof (LZ4_stream_t)); + LZ4_initStream((LZ4_stream_t*)cctxPtr->lz4CtxPtr, sizeof(LZ4_stream_t)); } else { - LZ4_initStreamHC((LZ4_streamHC_t *) cctxPtr->lz4CtxPtr, sizeof(LZ4_streamHC_t)); - LZ4_setCompressionLevel((LZ4_streamHC_t *) cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel); + LZ4_initStreamHC((LZ4_streamHC_t*)cctxPtr->lz4CtxPtr, sizeof(LZ4_streamHC_t)); + LZ4_setCompressionLevel((LZ4_streamHC_t*)cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel); } cctxPtr->lz4CtxState = ctxTypeID; - } - } + } } /* Buffer Management */ if (cctxPtr->prefs.frameInfo.blockSizeID == 0) @@ -647,9 +731,9 @@ if (cctxPtr->maxBufferSize < requiredBuffSize) { cctxPtr->maxBufferSize = 0; - FREEMEM(cctxPtr->tmpBuff); - cctxPtr->tmpBuff = (BYTE*)ALLOC_AND_ZERO(requiredBuffSize); - if (cctxPtr->tmpBuff == NULL) return err0r(LZ4F_ERROR_allocation_failed); + LZ4F_free(cctxPtr->tmpBuff, cctxPtr->cmem); + cctxPtr->tmpBuff = (BYTE*)LZ4F_calloc(requiredBuffSize, cctxPtr->cmem); + RETURN_ERROR_IF(cctxPtr->tmpBuff == NULL, allocation_failed); cctxPtr->maxBufferSize = requiredBuffSize; } } cctxPtr->tmpIn = cctxPtr->tmpBuff; @@ -669,31 +753,32 @@ /* Magic Number */ LZ4F_writeLE32(dstPtr, LZ4F_MAGICNUMBER); dstPtr += 4; - headerStart = dstPtr; + { BYTE* const headerStart = dstPtr; - /* FLG Byte */ - *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ - + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) - + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) - + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) - + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) - + (cctxPtr->prefs.frameInfo.dictID > 0) ); - /* BD Byte */ - *dstPtr++ = (BYTE)((cctxPtr->prefs.frameInfo.blockSizeID & _3BITS) << 4); - /* Optional Frame content size field */ - if (cctxPtr->prefs.frameInfo.contentSize) { - LZ4F_writeLE64(dstPtr, cctxPtr->prefs.frameInfo.contentSize); - dstPtr += 8; - cctxPtr->totalInSize = 0; - } - /* Optional dictionary ID field */ - if (cctxPtr->prefs.frameInfo.dictID) { - LZ4F_writeLE32(dstPtr, cctxPtr->prefs.frameInfo.dictID); - dstPtr += 4; + /* FLG Byte */ + *dstPtr++ = (BYTE)(((1 & _2BITS) << 6) /* Version('01') */ + + ((cctxPtr->prefs.frameInfo.blockMode & _1BIT ) << 5) + + ((cctxPtr->prefs.frameInfo.blockChecksumFlag & _1BIT ) << 4) + + ((unsigned)(cctxPtr->prefs.frameInfo.contentSize > 0) << 3) + + ((cctxPtr->prefs.frameInfo.contentChecksumFlag & _1BIT ) << 2) + + (cctxPtr->prefs.frameInfo.dictID > 0) ); + /* BD Byte */ + *dstPtr++ = (BYTE)((cctxPtr->prefs.frameInfo.blockSizeID & _3BITS) << 4); + /* Optional Frame content size field */ + if (cctxPtr->prefs.frameInfo.contentSize) { + LZ4F_writeLE64(dstPtr, cctxPtr->prefs.frameInfo.contentSize); + dstPtr += 8; + cctxPtr->totalInSize = 0; + } + /* Optional dictionary ID field */ + if (cctxPtr->prefs.frameInfo.dictID) { + LZ4F_writeLE32(dstPtr, cctxPtr->prefs.frameInfo.dictID); + dstPtr += 4; + } + /* Header CRC Byte */ + *dstPtr = LZ4F_headerChecksum(headerStart, (size_t)(dstPtr - headerStart)); + dstPtr++; } - /* Header CRC Byte */ - *dstPtr = LZ4F_headerChecksum(headerStart, (size_t)(dstPtr - headerStart)); - dstPtr++; cctxPtr->cStage = 1; /* header written, now request input data block */ return (size_t)(dstPtr - dstStart); @@ -701,9 +786,9 @@ /*! LZ4F_compressBegin() : - * init streaming compression and writes frame header into dstBuffer. - * dstBuffer must be >= LZ4F_HEADER_SIZE_MAX bytes. - * preferencesPtr can be NULL, in which case default parameters are selected. + * init streaming compression AND writes frame header into @dstBuffer. + * @dstCapacity must be >= LZ4F_HEADER_SIZE_MAX bytes. + * @preferencesPtr can be NULL, in which case default parameters are selected. * @return : number of bytes written into dstBuffer for the header * or an error code (can be tested using LZ4F_isError()) */ @@ -744,11 +829,13 @@ LZ4F_blockChecksum_t crcFlag) { BYTE* const cSizePtr = (BYTE*)dst; - U32 cSize = (U32)compress(lz4ctx, (const char*)src, (char*)(cSizePtr+BHSize), - (int)(srcSize), (int)(srcSize-1), - level, cdict); - if (cSize == 0) { /* compression failed */ - DEBUGLOG(5, "LZ4F_makeBlock: compression failed, creating a raw block (size %u)", (U32)srcSize); + U32 cSize; + assert(compress != NULL); + cSize = (U32)compress(lz4ctx, (const char*)src, (char*)(cSizePtr+BHSize), + (int)(srcSize), (int)(srcSize-1), + level, cdict); + + if (cSize == 0 || cSize >= srcSize) { cSize = (U32)srcSize; LZ4F_writeLE32(cSizePtr, cSize | LZ4F_BLOCKUNCOMPRESSED_FLAG); memcpy(cSizePtr+BHSize, src, srcSize); @@ -766,6 +853,7 @@ static int LZ4F_compressBlock(void* ctx, const char* src, char* dst, int srcSize, int dstCapacity, int level, const LZ4F_CDict* cdict) { int const acceleration = (level < 0) ? -level + 1 : 1; + DEBUGLOG(5, "LZ4F_compressBlock (srcSize=%i)", srcSize); LZ4F_initStream(ctx, cdict, level, LZ4F_blockIndependent); if (cdict) { return LZ4_compress_fast_continue((LZ4_stream_t*)ctx, src, dst, srcSize, dstCapacity, acceleration); @@ -778,6 +866,7 @@ { int const acceleration = (level < 0) ? -level + 1 : 1; (void)cdict; /* init once at beginning of frame */ + DEBUGLOG(5, "LZ4F_compressBlock_continue (srcSize=%i)", srcSize); return LZ4_compress_fast_continue((LZ4_stream_t*)ctx, src, dst, srcSize, dstCapacity, acceleration); } @@ -796,8 +885,15 @@ return LZ4_compress_HC_continue((LZ4_streamHC_t*)ctx, src, dst, srcSize, dstCapacity); } -static compressFunc_t LZ4F_selectCompression(LZ4F_blockMode_t blockMode, int level) +static int LZ4F_doNotCompressBlock(void* ctx, const char* src, char* dst, int srcSize, int dstCapacity, int level, const LZ4F_CDict* cdict) { + (void)ctx; (void)src; (void)dst; (void)srcSize; (void)dstCapacity; (void)level; (void)cdict; + return 0; +} + +static compressFunc_t LZ4F_selectCompression(LZ4F_blockMode_t blockMode, int level, LZ4F_blockCompression_t compressMode) +{ + if (compressMode == LZ4B_UNCOMPRESSED) return LZ4F_doNotCompressBlock; if (level < LZ4HC_CLEVEL_MIN) { if (blockMode == LZ4F_blockIndependent) return LZ4F_compressBlock; return LZ4F_compressBlock_continue; @@ -806,6 +902,7 @@ return LZ4F_compressBlockHC_continue; } +/* Save history (up to 64KB) into @tmpBuff */ static int LZ4F_localSaveDict(LZ4F_cctx_t* cctxPtr) { if (cctxPtr->prefs.compressionLevel < LZ4HC_CLEVEL_MIN) @@ -815,38 +912,57 @@ typedef enum { notDone, fromTmpBuffer, fromSrcBuffer } LZ4F_lastBlockStatus; -/*! LZ4F_compressUpdate() : +static const LZ4F_compressOptions_t k_cOptionsNull = { 0, { 0, 0, 0 } }; + + + /*! LZ4F_compressUpdateImpl() : * LZ4F_compressUpdate() can be called repetitively to compress as much data as necessary. - * dstBuffer MUST be >= LZ4F_compressBound(srcSize, preferencesPtr). - * LZ4F_compressOptions_t structure is optional : you can provide NULL as argument. + * When successful, the function always entirely consumes @srcBuffer. + * src data is either buffered or compressed into @dstBuffer. + * If the block compression does not match the compression of the previous block, the old data is flushed + * and operations continue with the new compression mode. + * @dstCapacity MUST be >= LZ4F_compressBound(srcSize, preferencesPtr) when block compression is turned on. + * @compressOptionsPtr is optional : provide NULL to mean "default". * @return : the number of bytes written into dstBuffer. It can be zero, meaning input data was just buffered. * or an error code if it fails (which can be tested using LZ4F_isError()) + * After an error, the state is left in a UB state, and must be re-initialized. */ -size_t LZ4F_compressUpdate(LZ4F_cctx* cctxPtr, - void* dstBuffer, size_t dstCapacity, +static size_t LZ4F_compressUpdateImpl(LZ4F_cctx* cctxPtr, + void* dstBuffer, size_t dstCapacity, const void* srcBuffer, size_t srcSize, - const LZ4F_compressOptions_t* compressOptionsPtr) -{ - LZ4F_compressOptions_t cOptionsNull; + const LZ4F_compressOptions_t* compressOptionsPtr, + LZ4F_blockCompression_t blockCompression) + { size_t const blockSize = cctxPtr->maxBlockSize; const BYTE* srcPtr = (const BYTE*)srcBuffer; const BYTE* const srcEnd = srcPtr + srcSize; BYTE* const dstStart = (BYTE*)dstBuffer; BYTE* dstPtr = dstStart; LZ4F_lastBlockStatus lastBlockCompressed = notDone; - compressFunc_t const compress = LZ4F_selectCompression(cctxPtr->prefs.frameInfo.blockMode, cctxPtr->prefs.compressionLevel); - + compressFunc_t const compress = LZ4F_selectCompression(cctxPtr->prefs.frameInfo.blockMode, cctxPtr->prefs.compressionLevel, blockCompression); + size_t bytesWritten; DEBUGLOG(4, "LZ4F_compressUpdate (srcSize=%zu)", srcSize); - if (cctxPtr->cStage != 1) return err0r(LZ4F_ERROR_GENERIC); + RETURN_ERROR_IF(cctxPtr->cStage != 1, compressionState_uninitialized); /* state must be initialized and waiting for next block */ if (dstCapacity < LZ4F_compressBound_internal(srcSize, &(cctxPtr->prefs), cctxPtr->tmpInSize)) - return err0r(LZ4F_ERROR_dstMaxSize_tooSmall); - MEM_INIT(&cOptionsNull, 0, sizeof(cOptionsNull)); - if (compressOptionsPtr == NULL) compressOptionsPtr = &cOptionsNull; + RETURN_ERROR(dstMaxSize_tooSmall); + if (blockCompression == LZ4B_UNCOMPRESSED && dstCapacity < srcSize) + RETURN_ERROR(dstMaxSize_tooSmall); + + /* flush currently written block, to continue with new block compression */ + if (cctxPtr->blockCompression != blockCompression) { + bytesWritten = LZ4F_flush(cctxPtr, dstBuffer, dstCapacity, compressOptionsPtr); + dstPtr += bytesWritten; + cctxPtr->blockCompression = blockCompression; + } + + if (compressOptionsPtr == NULL) compressOptionsPtr = &k_cOptionsNull; + /* complete tmp buffer */ if (cctxPtr->tmpInSize > 0) { /* some data already within tmp buffer */ size_t const sizeToCopy = blockSize - cctxPtr->tmpInSize; + assert(blockSize > cctxPtr->tmpInSize); if (sizeToCopy > srcSize) { /* add src to tmpIn buffer */ memcpy(cctxPtr->tmpIn + cctxPtr->tmpInSize, srcBuffer, srcSize); @@ -864,11 +980,9 @@ compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel, cctxPtr->cdict, cctxPtr->prefs.frameInfo.blockChecksumFlag); - if (cctxPtr->prefs.frameInfo.blockMode==LZ4F_blockLinked) cctxPtr->tmpIn += blockSize; cctxPtr->tmpInSize = 0; - } - } + } } while ((size_t)(srcEnd - srcPtr) >= blockSize) { /* compress full blocks */ @@ -882,33 +996,38 @@ } if ((cctxPtr->prefs.autoFlush) && (srcPtr < srcEnd)) { - /* compress remaining input < blockSize */ + /* autoFlush : remaining input (< blockSize) is compressed */ lastBlockCompressed = fromSrcBuffer; dstPtr += LZ4F_makeBlock(dstPtr, srcPtr, (size_t)(srcEnd - srcPtr), compress, cctxPtr->lz4CtxPtr, cctxPtr->prefs.compressionLevel, cctxPtr->cdict, cctxPtr->prefs.frameInfo.blockChecksumFlag); - srcPtr = srcEnd; + srcPtr = srcEnd; } - /* preserve dictionary if necessary */ + /* preserve dictionary within @tmpBuff whenever necessary */ if ((cctxPtr->prefs.frameInfo.blockMode==LZ4F_blockLinked) && (lastBlockCompressed==fromSrcBuffer)) { + /* linked blocks are only supported in compressed mode, see LZ4F_uncompressedUpdate */ + assert(blockCompression == LZ4B_COMPRESSED); if (compressOptionsPtr->stableSrc) { - cctxPtr->tmpIn = cctxPtr->tmpBuff; + cctxPtr->tmpIn = cctxPtr->tmpBuff; /* src is stable : dictionary remains in src across invocations */ } else { int const realDictSize = LZ4F_localSaveDict(cctxPtr); - if (realDictSize==0) return err0r(LZ4F_ERROR_GENERIC); + assert(0 <= realDictSize && realDictSize <= 64 KB); cctxPtr->tmpIn = cctxPtr->tmpBuff + realDictSize; } } /* keep tmpIn within limits */ - if ((cctxPtr->tmpIn + blockSize) > (cctxPtr->tmpBuff + cctxPtr->maxBufferSize) /* necessarily LZ4F_blockLinked && lastBlockCompressed==fromTmpBuffer */ - && !(cctxPtr->prefs.autoFlush)) + if (!(cctxPtr->prefs.autoFlush) /* no autoflush : there may be some data left within internal buffer */ + && (cctxPtr->tmpIn + blockSize) > (cctxPtr->tmpBuff + cctxPtr->maxBufferSize) ) /* not enough room to store next block */ { + /* only preserve 64KB within internal buffer. Ensures there is enough room for next block. + * note: this situation necessarily implies lastBlockCompressed==fromTmpBuffer */ int const realDictSize = LZ4F_localSaveDict(cctxPtr); cctxPtr->tmpIn = cctxPtr->tmpBuff + realDictSize; + assert((cctxPtr->tmpIn + blockSize) <= (cctxPtr->tmpBuff + cctxPtr->maxBufferSize)); } /* some input data left, necessarily < blockSize */ @@ -926,7 +1045,54 @@ return (size_t)(dstPtr - dstStart); } +/*! LZ4F_compressUpdate() : + * LZ4F_compressUpdate() can be called repetitively to compress as much data as necessary. + * When successful, the function always entirely consumes @srcBuffer. + * src data is either buffered or compressed into @dstBuffer. + * If previously an uncompressed block was written, buffered data is flushed + * before appending compressed data is continued. + * @dstCapacity MUST be >= LZ4F_compressBound(srcSize, preferencesPtr). + * @compressOptionsPtr is optional : provide NULL to mean "default". + * @return : the number of bytes written into dstBuffer. It can be zero, meaning input data was just buffered. + * or an error code if it fails (which can be tested using LZ4F_isError()) + * After an error, the state is left in a UB state, and must be re-initialized. + */ +size_t LZ4F_compressUpdate(LZ4F_cctx* cctxPtr, + void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_compressOptions_t* compressOptionsPtr) +{ + return LZ4F_compressUpdateImpl(cctxPtr, + dstBuffer, dstCapacity, + srcBuffer, srcSize, + compressOptionsPtr, LZ4B_COMPRESSED); +} +/*! LZ4F_compressUpdate() : + * LZ4F_compressUpdate() can be called repetitively to compress as much data as necessary. + * When successful, the function always entirely consumes @srcBuffer. + * src data is either buffered or compressed into @dstBuffer. + * If previously an uncompressed block was written, buffered data is flushed + * before appending compressed data is continued. + * This is only supported when LZ4F_blockIndependent is used + * @dstCapacity MUST be >= LZ4F_compressBound(srcSize, preferencesPtr). + * @compressOptionsPtr is optional : provide NULL to mean "default". + * @return : the number of bytes written into dstBuffer. It can be zero, meaning input data was just buffered. + * or an error code if it fails (which can be tested using LZ4F_isError()) + * After an error, the state is left in a UB state, and must be re-initialized. + */ +size_t LZ4F_uncompressedUpdate(LZ4F_cctx* cctxPtr, + void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_compressOptions_t* compressOptionsPtr) { + RETURN_ERROR_IF(cctxPtr->prefs.frameInfo.blockMode != LZ4F_blockIndependent, blockMode_invalid); + return LZ4F_compressUpdateImpl(cctxPtr, + dstBuffer, dstCapacity, + srcBuffer, srcSize, + compressOptionsPtr, LZ4B_UNCOMPRESSED); +} + + /*! LZ4F_flush() : * When compressed data must be sent immediately, without waiting for a block to be filled, * invoke LZ4_flush(), which will immediately compress any remaining data stored within LZ4F_cctx. @@ -944,13 +1110,12 @@ compressFunc_t compress; if (cctxPtr->tmpInSize == 0) return 0; /* nothing to flush */ - if (cctxPtr->cStage != 1) return err0r(LZ4F_ERROR_GENERIC); - if (dstCapacity < (cctxPtr->tmpInSize + BHSize + BFSize)) - return err0r(LZ4F_ERROR_dstMaxSize_tooSmall); - (void)compressOptionsPtr; /* not yet useful */ + RETURN_ERROR_IF(cctxPtr->cStage != 1, compressionState_uninitialized); + RETURN_ERROR_IF(dstCapacity < (cctxPtr->tmpInSize + BHSize + BFSize), dstMaxSize_tooSmall); + (void)compressOptionsPtr; /* not useful (yet) */ /* select compression function */ - compress = LZ4F_selectCompression(cctxPtr->prefs.frameInfo.blockMode, cctxPtr->prefs.compressionLevel); + compress = LZ4F_selectCompression(cctxPtr->prefs.frameInfo.blockMode, cctxPtr->prefs.compressionLevel, cctxPtr->blockCompression); /* compress tmp buffer */ dstPtr += LZ4F_makeBlock(dstPtr, @@ -992,19 +1157,19 @@ size_t const flushSize = LZ4F_flush(cctxPtr, dstBuffer, dstCapacity, compressOptionsPtr); DEBUGLOG(5,"LZ4F_compressEnd: dstCapacity=%u", (unsigned)dstCapacity); - if (LZ4F_isError(flushSize)) return flushSize; + FORWARD_IF_ERROR(flushSize); dstPtr += flushSize; assert(flushSize <= dstCapacity); dstCapacity -= flushSize; - if (dstCapacity < 4) return err0r(LZ4F_ERROR_dstMaxSize_tooSmall); + RETURN_ERROR_IF(dstCapacity < 4, dstMaxSize_tooSmall); LZ4F_writeLE32(dstPtr, 0); dstPtr += 4; /* endMark */ if (cctxPtr->prefs.frameInfo.contentChecksumFlag == LZ4F_contentChecksumEnabled) { U32 const xxh = XXH32_digest(&(cctxPtr->xxh)); - if (dstCapacity < 8) return err0r(LZ4F_ERROR_dstMaxSize_tooSmall); + RETURN_ERROR_IF(dstCapacity < 8, dstMaxSize_tooSmall); DEBUGLOG(5,"Writing 32-bit content checksum"); LZ4F_writeLE32(dstPtr, xxh); dstPtr+=4; /* content Checksum */ @@ -1015,7 +1180,7 @@ if (cctxPtr->prefs.frameInfo.contentSize) { if (cctxPtr->prefs.frameInfo.contentSize != cctxPtr->totalInSize) - return err0r(LZ4F_ERROR_frameSize_wrong); + RETURN_ERROR(frameSize_wrong); } return (size_t)(dstPtr - dstStart); @@ -1039,6 +1204,7 @@ } dStage_t; struct LZ4F_dctx_s { + LZ4F_CustomMem cmem; LZ4F_frameInfo_t frameInfo; U32 version; dStage_t dStage; @@ -1056,26 +1222,37 @@ size_t tmpOutStart; XXH32_state_t xxh; XXH32_state_t blockChecksum; + int skipChecksum; BYTE header[LZ4F_HEADER_SIZE_MAX]; }; /* typedef'd to LZ4F_dctx in lz4frame.h */ +LZ4F_dctx* LZ4F_createDecompressionContext_advanced(LZ4F_CustomMem customMem, unsigned version) +{ + LZ4F_dctx* const dctx = (LZ4F_dctx*)LZ4F_calloc(sizeof(LZ4F_dctx), customMem); + if (dctx == NULL) return NULL; + + dctx->cmem = customMem; + dctx->version = version; + return dctx; +} + /*! LZ4F_createDecompressionContext() : * Create a decompressionContext object, which will track all decompression operations. * Provides a pointer to a fully allocated and initialized LZ4F_decompressionContext object. * Object can later be released using LZ4F_freeDecompressionContext(). * @return : if != 0, there was an error during context creation. */ -LZ4F_errorCode_t LZ4F_createDecompressionContext(LZ4F_dctx** LZ4F_decompressionContextPtr, unsigned versionNumber) +LZ4F_errorCode_t +LZ4F_createDecompressionContext(LZ4F_dctx** LZ4F_decompressionContextPtr, unsigned versionNumber) { - LZ4F_dctx* const dctx = (LZ4F_dctx*)ALLOC_AND_ZERO(sizeof(LZ4F_dctx)); - if (dctx == NULL) { /* failed allocation */ - *LZ4F_decompressionContextPtr = NULL; - return err0r(LZ4F_ERROR_allocation_failed); - } + assert(LZ4F_decompressionContextPtr != NULL); /* violation of narrow contract */ + RETURN_ERROR_IF(LZ4F_decompressionContextPtr == NULL, parameter_null); /* in case it nonetheless happen in production */ - dctx->version = versionNumber; - *LZ4F_decompressionContextPtr = dctx; + *LZ4F_decompressionContextPtr = LZ4F_createDecompressionContext_advanced(LZ4F_defaultCMem, versionNumber); + if (*LZ4F_decompressionContextPtr == NULL) { /* failed allocation */ + RETURN_ERROR(allocation_failed); + } return LZ4F_OK_NoError; } @@ -1084,9 +1261,9 @@ LZ4F_errorCode_t result = LZ4F_OK_NoError; if (dctx != NULL) { /* can accept NULL input, like free() */ result = (LZ4F_errorCode_t)dctx->dStage; - FREEMEM(dctx->tmpIn); - FREEMEM(dctx->tmpOutBuffer); - FREEMEM(dctx); + LZ4F_free(dctx->tmpIn, dctx->cmem); + LZ4F_free(dctx->tmpOutBuffer, dctx->cmem); + LZ4F_free(dctx, dctx->cmem); } return result; } @@ -1099,6 +1276,7 @@ dctx->dStage = dstage_getFrameHeader; dctx->dict = NULL; dctx->dictSize = 0; + dctx->skipChecksum = 0; } @@ -1118,7 +1296,7 @@ DEBUGLOG(5, "LZ4F_decodeHeader"); /* need to decode header to get frameInfo */ - if (srcSize < minFHSize) return err0r(LZ4F_ERROR_frameHeader_incomplete); /* minimal frame header size */ + RETURN_ERROR_IF(srcSize < minFHSize, frameHeader_incomplete); /* minimal frame header size */ MEM_INIT(&(dctx->frameInfo), 0, sizeof(dctx->frameInfo)); /* special case : skippable frames */ @@ -1132,14 +1310,13 @@ } else { dctx->dStage = dstage_getSFrameSize; return 4; - } - } + } } /* control magic number */ #ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION if (LZ4F_readLE32(srcPtr) != LZ4F_MAGICNUMBER) { DEBUGLOG(4, "frame header error : unknown magic number"); - return err0r(LZ4F_ERROR_frameType_unknown); + RETURN_ERROR(frameType_unknown); } #endif dctx->frameInfo.frameType = LZ4F_frame; @@ -1153,8 +1330,8 @@ contentChecksumFlag = (FLG>>2) & _1BIT; dictIDFlag = FLG & _1BIT; /* validate */ - if (((FLG>>1)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ - if (version != 1) return err0r(LZ4F_ERROR_headerVersion_wrong); /* Version Number, only supported value */ + if (((FLG>>1)&_1BIT) != 0) RETURN_ERROR(reservedFlag_set); /* Reserved bit */ + if (version != 1) RETURN_ERROR(headerVersion_wrong); /* Version Number, only supported value */ } /* Frame Header Size */ @@ -1173,17 +1350,16 @@ { U32 const BD = srcPtr[5]; blockSizeID = (BD>>4) & _3BITS; /* validate */ - if (((BD>>7)&_1BIT) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bit */ - if (blockSizeID < 4) return err0r(LZ4F_ERROR_maxBlockSize_invalid); /* 4-7 only supported values for the time being */ - if (((BD>>0)&_4BITS) != 0) return err0r(LZ4F_ERROR_reservedFlag_set); /* Reserved bits */ + if (((BD>>7)&_1BIT) != 0) RETURN_ERROR(reservedFlag_set); /* Reserved bit */ + if (blockSizeID < 4) RETURN_ERROR(maxBlockSize_invalid); /* 4-7 only supported values for the time being */ + if (((BD>>0)&_4BITS) != 0) RETURN_ERROR(reservedFlag_set); /* Reserved bits */ } /* check header */ assert(frameHeaderSize > 5); #ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION { BYTE const HC = LZ4F_headerChecksum(srcPtr+4, frameHeaderSize-5); - if (HC != srcPtr[frameHeaderSize-1]) - return err0r(LZ4F_ERROR_headerChecksum_invalid); + RETURN_ERROR_IF(HC != srcPtr[frameHeaderSize-1], headerChecksum_invalid); } #endif @@ -1192,10 +1368,9 @@ dctx->frameInfo.blockChecksumFlag = (LZ4F_blockChecksum_t)blockChecksumFlag; dctx->frameInfo.contentChecksumFlag = (LZ4F_contentChecksum_t)contentChecksumFlag; dctx->frameInfo.blockSizeID = (LZ4F_blockSizeID_t)blockSizeID; - dctx->maxBlockSize = LZ4F_getBlockSize(blockSizeID); + dctx->maxBlockSize = LZ4F_getBlockSize((LZ4F_blockSizeID_t)blockSizeID); if (contentSizeFlag) - dctx->frameRemainingSize = - dctx->frameInfo.contentSize = LZ4F_readLE64(srcPtr+6); + dctx->frameRemainingSize = dctx->frameInfo.contentSize = LZ4F_readLE64(srcPtr+6); if (dictIDFlag) dctx->frameInfo.dictID = LZ4F_readLE32(srcPtr + frameHeaderSize - 5); @@ -1211,11 +1386,11 @@ */ size_t LZ4F_headerSize(const void* src, size_t srcSize) { - if (src == NULL) return err0r(LZ4F_ERROR_srcPtr_wrong); + RETURN_ERROR_IF(src == NULL, srcPtr_wrong); /* minimal srcSize to determine header size */ if (srcSize < LZ4F_MIN_SIZE_TO_KNOW_HEADER_LENGTH) - return err0r(LZ4F_ERROR_frameHeader_incomplete); + RETURN_ERROR(frameHeader_incomplete); /* special case : skippable frames */ if ((LZ4F_readLE32(src) & 0xFFFFFFF0U) == LZ4F_MAGIC_SKIPPABLE_START) @@ -1224,7 +1399,7 @@ /* control magic number */ #ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION if (LZ4F_readLE32(src) != LZ4F_MAGICNUMBER) - return err0r(LZ4F_ERROR_frameType_unknown); + RETURN_ERROR(frameType_unknown); #endif /* Frame Header Size */ @@ -1266,13 +1441,13 @@ if (dctx->dStage == dstage_storeFrameHeader) { /* frame decoding already started, in the middle of header => automatic fail */ *srcSizePtr = 0; - return err0r(LZ4F_ERROR_frameDecoding_alreadyStarted); + RETURN_ERROR(frameDecoding_alreadyStarted); } else { size_t const hSize = LZ4F_headerSize(srcBuffer, *srcSizePtr); if (LZ4F_isError(hSize)) { *srcSizePtr=0; return hSize; } if (*srcSizePtr < hSize) { *srcSizePtr=0; - return err0r(LZ4F_ERROR_frameHeader_incomplete); + RETURN_ERROR(frameHeader_incomplete); } { size_t decodeResult = LZ4F_decodeHeader(dctx, srcBuffer, hSize); @@ -1290,16 +1465,14 @@ /* LZ4F_updateDict() : * only used for LZ4F_blockLinked mode - * Condition : dstPtr != NULL + * Condition : @dstPtr != NULL */ static void LZ4F_updateDict(LZ4F_dctx* dctx, const BYTE* dstPtr, size_t dstSize, const BYTE* dstBufferStart, unsigned withinTmp) { assert(dstPtr != NULL); - if (dctx->dictSize==0) { - dctx->dict = (const BYTE*)dstPtr; /* priority to prefix mode */ - } + if (dctx->dictSize==0) dctx->dict = (const BYTE*)dstPtr; /* will lead to prefix mode */ assert(dctx->dict != NULL); if (dctx->dict + dctx->dictSize == dstPtr) { /* prefix mode, everything within dstBuffer */ @@ -1362,7 +1535,6 @@ } - /*! LZ4F_decompress() : * Call this function repetitively to regenerate compressed data in srcBuffer. * The function will attempt to decode up to *srcSizePtr bytes from srcBuffer @@ -1406,6 +1578,7 @@ *srcSizePtr = 0; *dstSizePtr = 0; assert(dctx != NULL); + dctx->skipChecksum |= (decompressOptionsPtr->skipChecksums != 0); /* once set, disable for the remainder of the frame */ /* behaves as a state machine */ @@ -1418,7 +1591,7 @@ DEBUGLOG(6, "dstage_getFrameHeader"); if ((size_t)(srcEnd-srcPtr) >= maxFHSize) { /* enough to decode - shortcut */ size_t const hSize = LZ4F_decodeHeader(dctx, srcPtr, (size_t)(srcEnd-srcPtr)); /* will update dStage appropriately */ - if (LZ4F_isError(hSize)) return hSize; + FORWARD_IF_ERROR(hSize); srcPtr += hSize; break; } @@ -1440,9 +1613,7 @@ doAnotherStage = 0; /* not enough src data, ask for some more */ break; } - { size_t const hSize = LZ4F_decodeHeader(dctx, dctx->header, dctx->tmpInTarget); /* will update dStage appropriately */ - if (LZ4F_isError(hSize)) return hSize; - } + FORWARD_IF_ERROR( LZ4F_decodeHeader(dctx, dctx->header, dctx->tmpInTarget) ); /* will update dStage appropriately */ break; case dstage_init: @@ -1453,14 +1624,12 @@ + ((dctx->frameInfo.blockMode==LZ4F_blockLinked) ? 128 KB : 0); if (bufferNeeded > dctx->maxBufferSize) { /* tmp buffers too small */ dctx->maxBufferSize = 0; /* ensure allocation will be re-attempted on next entry*/ - FREEMEM(dctx->tmpIn); - dctx->tmpIn = (BYTE*)ALLOC(dctx->maxBlockSize + BFSize /* block checksum */); - if (dctx->tmpIn == NULL) - return err0r(LZ4F_ERROR_allocation_failed); - FREEMEM(dctx->tmpOutBuffer); - dctx->tmpOutBuffer= (BYTE*)ALLOC(bufferNeeded); - if (dctx->tmpOutBuffer== NULL) - return err0r(LZ4F_ERROR_allocation_failed); + LZ4F_free(dctx->tmpIn, dctx->cmem); + dctx->tmpIn = (BYTE*)LZ4F_malloc(dctx->maxBlockSize + BFSize /* block checksum */, dctx->cmem); + RETURN_ERROR_IF(dctx->tmpIn == NULL, allocation_failed); + LZ4F_free(dctx->tmpOutBuffer, dctx->cmem); + dctx->tmpOutBuffer= (BYTE*)LZ4F_malloc(bufferNeeded, dctx->cmem); + RETURN_ERROR_IF(dctx->tmpOutBuffer== NULL, allocation_failed); dctx->maxBufferSize = bufferNeeded; } } dctx->tmpInSize = 0; @@ -1509,7 +1678,7 @@ break; } if (nextCBlockSize > dctx->maxBlockSize) { - return err0r(LZ4F_ERROR_maxBlockSize_invalid); + RETURN_ERROR(maxBlockSize_invalid); } if (blockHeader & LZ4F_BLOCKUNCOMPRESSED_FLAG) { /* next block is uncompressed */ @@ -1540,11 +1709,13 @@ size_t const minBuffSize = MIN((size_t)(srcEnd-srcPtr), (size_t)(dstEnd-dstPtr)); sizeToCopy = MIN(dctx->tmpInTarget, minBuffSize); memcpy(dstPtr, srcPtr, sizeToCopy); - if (dctx->frameInfo.blockChecksumFlag) { - (void)XXH32_update(&dctx->blockChecksum, srcPtr, sizeToCopy); + if (!dctx->skipChecksum) { + if (dctx->frameInfo.blockChecksumFlag) { + (void)XXH32_update(&dctx->blockChecksum, srcPtr, sizeToCopy); + } + if (dctx->frameInfo.contentChecksumFlag) + (void)XXH32_update(&dctx->xxh, srcPtr, sizeToCopy); } - if (dctx->frameInfo.contentChecksumFlag) - (void)XXH32_update(&dctx->xxh, srcPtr, sizeToCopy); if (dctx->frameInfo.contentSize) dctx->frameRemainingSize -= sizeToCopy; @@ -1590,14 +1761,15 @@ } crcSrc = dctx->header; } - { U32 const readCRC = LZ4F_readLE32(crcSrc); + if (!dctx->skipChecksum) { + U32 const readCRC = LZ4F_readLE32(crcSrc); U32 const calcCRC = XXH32_digest(&dctx->blockChecksum); #ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION DEBUGLOG(6, "compare block checksum"); if (readCRC != calcCRC) { DEBUGLOG(4, "incorrect block checksum: %08X != %08X", readCRC, calcCRC); - return err0r(LZ4F_ERROR_blockChecksum_invalid); + RETURN_ERROR(blockChecksum_invalid); } #else (void)readCRC; @@ -1637,37 +1809,44 @@ } /* At this stage, input is large enough to decode a block */ + + /* First, decode and control block checksum if it exists */ if (dctx->frameInfo.blockChecksumFlag) { + assert(dctx->tmpInTarget >= 4); dctx->tmpInTarget -= 4; assert(selectedIn != NULL); /* selectedIn is defined at this stage (either srcPtr, or dctx->tmpIn) */ { U32 const readBlockCrc = LZ4F_readLE32(selectedIn + dctx->tmpInTarget); U32 const calcBlockCrc = XXH32(selectedIn, dctx->tmpInTarget, 0); #ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION - if (readBlockCrc != calcBlockCrc) - return err0r(LZ4F_ERROR_blockChecksum_invalid); + RETURN_ERROR_IF(readBlockCrc != calcBlockCrc, blockChecksum_invalid); #else (void)readBlockCrc; (void)calcBlockCrc; #endif } } - if ((size_t)(dstEnd-dstPtr) >= dctx->maxBlockSize) { + /* decode directly into destination buffer if there is enough room */ + if ( ((size_t)(dstEnd-dstPtr) >= dctx->maxBlockSize) + /* unless the dictionary is stored in tmpOut: + * in which case it's faster to decode within tmpOut + * to benefit from prefix speedup */ + && !(dctx->dict!= NULL && (const BYTE*)dctx->dict + dctx->dictSize == dctx->tmpOut) ) + { const char* dict = (const char*)dctx->dict; size_t dictSize = dctx->dictSize; int decodedSize; assert(dstPtr != NULL); if (dict && dictSize > 1 GB) { - /* the dictSize param is an int, avoid truncation / sign issues */ + /* overflow control : dctx->dictSize is an int, avoid truncation / sign issues */ dict += dictSize - 64 KB; dictSize = 64 KB; } - /* enough capacity in `dst` to decompress directly there */ decodedSize = LZ4_decompress_safe_usingDict( (const char*)selectedIn, (char*)dstPtr, (int)dctx->tmpInTarget, (int)dctx->maxBlockSize, dict, (int)dictSize); - if (decodedSize < 0) return err0r(LZ4F_ERROR_GENERIC); /* decompression failed */ - if (dctx->frameInfo.contentChecksumFlag) + RETURN_ERROR_IF(decodedSize < 0, decompressionFailed); + if ((dctx->frameInfo.contentChecksumFlag) && (!dctx->skipChecksum)) XXH32_update(&(dctx->xxh), dstPtr, (size_t)decodedSize); if (dctx->frameInfo.contentSize) dctx->frameRemainingSize -= (size_t)decodedSize; @@ -1678,25 +1857,27 @@ } dstPtr += decodedSize; - dctx->dStage = dstage_getBlockHeader; + dctx->dStage = dstage_getBlockHeader; /* end of block, let's get another one */ break; } /* not enough place into dst : decode into tmpOut */ - /* ensure enough place for tmpOut */ + + /* manage dictionary */ if (dctx->frameInfo.blockMode == LZ4F_blockLinked) { if (dctx->dict == dctx->tmpOutBuffer) { + /* truncate dictionary to 64 KB if too big */ if (dctx->dictSize > 128 KB) { memcpy(dctx->tmpOutBuffer, dctx->dict + dctx->dictSize - 64 KB, 64 KB); dctx->dictSize = 64 KB; } dctx->tmpOut = dctx->tmpOutBuffer + dctx->dictSize; - } else { /* dict not within tmp */ + } else { /* dict not within tmpOut */ size_t const reservedDictSpace = MIN(dctx->dictSize, 64 KB); dctx->tmpOut = dctx->tmpOutBuffer + reservedDictSpace; } } - /* Decode block */ + /* Decode block into tmpOut */ { const char* dict = (const char*)dctx->dict; size_t dictSize = dctx->dictSize; int decodedSize; @@ -1709,9 +1890,8 @@ (const char*)selectedIn, (char*)dctx->tmpOut, (int)dctx->tmpInTarget, (int)dctx->maxBlockSize, dict, (int)dictSize); - if (decodedSize < 0) /* decompression failed */ - return err0r(LZ4F_ERROR_decompressionFailed); - if (dctx->frameInfo.contentChecksumFlag) + RETURN_ERROR_IF(decodedSize < 0, decompressionFailed); + if (dctx->frameInfo.contentChecksumFlag && !dctx->skipChecksum) XXH32_update(&(dctx->xxh), dctx->tmpOut, (size_t)decodedSize); if (dctx->frameInfo.contentSize) dctx->frameRemainingSize -= (size_t)decodedSize; @@ -1744,8 +1924,7 @@ break; case dstage_getSuffix: - if (dctx->frameRemainingSize) - return err0r(LZ4F_ERROR_frameSize_wrong); /* incorrect frame size decoded */ + RETURN_ERROR_IF(dctx->frameRemainingSize, frameSize_wrong); /* incorrect frame size decoded */ if (!dctx->frameInfo.contentChecksumFlag) { /* no checksum, frame is completed */ nextSrcSizeHint = 0; LZ4F_resetDecompressionContext(dctx); @@ -1777,20 +1956,20 @@ } /* if (dctx->dStage == dstage_storeSuffix) */ /* case dstage_checkSuffix: */ /* no direct entry, avoid initialization risks */ - { U32 const readCRC = LZ4F_readLE32(selectedIn); + if (!dctx->skipChecksum) { + U32 const readCRC = LZ4F_readLE32(selectedIn); U32 const resultCRC = XXH32_digest(&(dctx->xxh)); #ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION - if (readCRC != resultCRC) - return err0r(LZ4F_ERROR_contentChecksum_invalid); + RETURN_ERROR_IF(readCRC != resultCRC, contentChecksum_invalid); #else (void)readCRC; (void)resultCRC; #endif - nextSrcSizeHint = 0; - LZ4F_resetDecompressionContext(dctx); - doAnotherStage = 0; - break; } + nextSrcSizeHint = 0; + LZ4F_resetDecompressionContext(dctx); + doAnotherStage = 0; + break; case dstage_getSFrameSize: if ((srcEnd - srcPtr) >= 4) { @@ -1841,7 +2020,7 @@ } /* switch (dctx->dStage) */ } /* while (doAnotherStage) */ - /* preserve history within tmp whenever necessary */ + /* preserve history within tmpOut whenever necessary */ LZ4F_STATIC_ASSERT((unsigned)dstage_init == 2); if ( (dctx->frameInfo.blockMode==LZ4F_blockLinked) /* next block will use up to 64KB from previous ones */ && (dctx->dict != dctx->tmpOutBuffer) /* dictionary is not already within tmp */ diff --git a/cbits/lz4frame.h b/cbits/lz4frame.h --- a/cbits/lz4frame.h +++ b/cbits/lz4frame.h @@ -1,7 +1,7 @@ /* - LZ4 auto-framing library + LZ4F - LZ4-Frame library Header File - Copyright (C) 2011-2017, Yann Collet. + Copyright (C) 2011-2020, Yann Collet. BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) Redistribution and use in source and binary forms, with or without @@ -39,7 +39,7 @@ * LZ4F also offers streaming capabilities. * * lz4.h is not required when using lz4frame.h, - * except to extract common constant such as LZ4_VERSION_NUMBER. + * except to extract common constants such as LZ4_VERSION_NUMBER. * */ #ifndef LZ4F_H_09782039843 @@ -54,12 +54,12 @@ /** - Introduction - - lz4frame.h implements LZ4 frame specification (doc/lz4_Frame_format.md). - lz4frame.h provides frame compression functions that take care - of encoding standard metadata alongside LZ4-compressed blocks. -*/ + * Introduction + * + * lz4frame.h implements LZ4 frame specification: see doc/lz4_Frame_format.md . + * LZ4 Frames are compatible with `lz4` CLI, + * and designed to be interoperable with any system. +**/ /*-*************************************************************** * Compiler specifics @@ -210,7 +210,7 @@ * Returns the maximum possible compressed size with LZ4F_compressFrame() given srcSize and preferences. * `preferencesPtr` is optional. It can be replaced by NULL, in which case, the function will assume default preferences. * Note : this result is only usable with LZ4F_compressFrame(). - * It may also be used with LZ4F_compressUpdate() _if no flush() operation_ is performed. + * It may also be relevant to LZ4F_compressUpdate() _only if_ no flush() operation is ever performed. */ LZ4FLIB_API size_t LZ4F_compressFrameBound(size_t srcSize, const LZ4F_preferences_t* preferencesPtr); @@ -230,7 +230,7 @@ * Advanced compression functions *************************************/ typedef struct LZ4F_cctx_s LZ4F_cctx; /* incomplete type */ -typedef LZ4F_cctx* LZ4F_compressionContext_t; /* for compatibility with previous API version */ +typedef LZ4F_cctx* LZ4F_compressionContext_t; /* for compatibility with older APIs, prefer using LZ4F_cctx */ typedef struct { unsigned stableSrc; /* 1 == src content will remain present on future calls to LZ4F_compress(); skip copying src content within tmp buffer */ @@ -243,20 +243,27 @@ LZ4FLIB_API unsigned LZ4F_getVersion(void); /*! LZ4F_createCompressionContext() : - * The first thing to do is to create a compressionContext object, which will be used in all compression operations. - * This is achieved using LZ4F_createCompressionContext(), which takes as argument a version. - * The version provided MUST be LZ4F_VERSION. It is intended to track potential version mismatch, notably when using DLL. - * The function will provide a pointer to a fully allocated LZ4F_cctx object. - * If @return != zero, there was an error during context creation. - * Object can release its memory using LZ4F_freeCompressionContext(); - */ + * The first thing to do is to create a compressionContext object, + * which will keep track of operation state during streaming compression. + * This is achieved using LZ4F_createCompressionContext(), which takes as argument a version, + * and a pointer to LZ4F_cctx*, to write the resulting pointer into. + * @version provided MUST be LZ4F_VERSION. It is intended to track potential version mismatch, notably when using DLL. + * The function provides a pointer to a fully allocated LZ4F_cctx object. + * @cctxPtr MUST be != NULL. + * If @return != zero, context creation failed. + * A created compression context can be employed multiple times for consecutive streaming operations. + * Once all streaming compression jobs are completed, + * the state object can be released using LZ4F_freeCompressionContext(). + * Note1 : LZ4F_freeCompressionContext() is always successful. Its return value can be ignored. + * Note2 : LZ4F_freeCompressionContext() works fine with NULL input pointers (do nothing). +**/ LZ4FLIB_API LZ4F_errorCode_t LZ4F_createCompressionContext(LZ4F_cctx** cctxPtr, unsigned version); LZ4FLIB_API LZ4F_errorCode_t LZ4F_freeCompressionContext(LZ4F_cctx* cctx); /*---- Compression ----*/ -#define LZ4F_HEADER_SIZE_MIN 7 /* LZ4 Frame header size can vary, depending on selected paramaters */ +#define LZ4F_HEADER_SIZE_MIN 7 /* LZ4 Frame header size can vary, depending on selected parameters */ #define LZ4F_HEADER_SIZE_MAX 19 /* Size in bytes of a block header in little-endian format. Highest bit indicates if block data is uncompressed */ @@ -301,8 +308,9 @@ * Important rule: dstCapacity MUST be large enough to ensure operation success even in worst case situations. * This value is provided by LZ4F_compressBound(). * If this condition is not respected, LZ4F_compress() will fail (result is an errorCode). - * LZ4F_compressUpdate() doesn't guarantee error recovery. - * When an error occurs, compression context must be freed or resized. + * After an error, the state is left in a UB state, and must be re-initialized or freed. + * If previously an uncompressed block was written, buffered data is flushed + * before appending compressed data is continued. * `cOptPtr` is optional : NULL can be provided, in which case all options are set to default. * @return : number of bytes written into `dstBuffer` (it can be zero, meaning input data was just buffered). * or an error code if it fails (which can be tested using LZ4F_isError()) @@ -347,8 +355,12 @@ typedef LZ4F_dctx* LZ4F_decompressionContext_t; /* compatibility with previous API versions */ typedef struct { - unsigned stableDst; /* pledges that last 64KB decompressed data will remain available unmodified. This optimization skips storage operations in tmp buffers. */ - unsigned reserved[3]; /* must be set to zero for forward compatibility */ + unsigned stableDst; /* pledges that last 64KB decompressed data will remain available unmodified between invocations. + * This optimization skips storage operations in tmp buffers. */ + unsigned skipChecksums; /* disable checksum calculation and verification, even when one is present in frame, to save CPU time. + * Setting this option to 1 once disables all checksums for the rest of the frame. */ + unsigned reserved1; /* must be set to zero for forward compatibility */ + unsigned reserved0; /* idem */ } LZ4F_decompressOptions_t; @@ -356,9 +368,10 @@ /*! LZ4F_createDecompressionContext() : * Create an LZ4F_dctx object, to track all decompression operations. - * The version provided MUST be LZ4F_VERSION. - * The function provides a pointer to an allocated and initialized LZ4F_dctx object. - * The result is an errorCode, which can be tested using LZ4F_isError(). + * @version provided MUST be LZ4F_VERSION. + * @dctxPtr MUST be valid. + * The function fills @dctxPtr with the value of a pointer to an allocated and initialized LZ4F_dctx object. + * The @return is an errorCode, which can be tested using LZ4F_isError(). * dctx memory can be released using LZ4F_freeDecompressionContext(); * Result of LZ4F_freeDecompressionContext() indicates current state of decompressionContext when being released. * That is, it should be == 0 if decompression has been completed fully and correctly. @@ -371,6 +384,8 @@ * Streaming decompression functions *************************************/ +#define LZ4F_MAGICNUMBER 0x184D2204U +#define LZ4F_MAGIC_SKIPPABLE_START 0x184D2A50U #define LZ4F_MIN_SIZE_TO_KNOW_HEADER_LENGTH 5 /*! LZ4F_headerSize() : v1.9.0+ @@ -386,7 +401,7 @@ /*! LZ4F_getFrameInfo() : * This function extracts frame parameters (max blockSize, dictID, etc.). - * Its usage is optional: user can call LZ4F_decompress() directly. + * Its usage is optional: user can also invoke LZ4F_decompress() directly. * * Extracted information will fill an existing LZ4F_frameInfo_t structure. * This can be useful for allocation and dictionary identification purposes. @@ -427,9 +442,10 @@ * note 1 : in case of error, dctx is not modified. Decoding operation can resume from beginning safely. * note 2 : frame parameters are *copied into* an already allocated LZ4F_frameInfo_t structure. */ -LZ4FLIB_API size_t LZ4F_getFrameInfo(LZ4F_dctx* dctx, - LZ4F_frameInfo_t* frameInfoPtr, - const void* srcBuffer, size_t* srcSizePtr); +LZ4FLIB_API size_t +LZ4F_getFrameInfo(LZ4F_dctx* dctx, + LZ4F_frameInfo_t* frameInfoPtr, + const void* srcBuffer, size_t* srcSizePtr); /*! LZ4F_decompress() : * Call this function repetitively to regenerate data compressed in `srcBuffer`. @@ -462,10 +478,11 @@ * * After a frame is fully decoded, dctx can be used again to decompress another frame. */ -LZ4FLIB_API size_t LZ4F_decompress(LZ4F_dctx* dctx, - void* dstBuffer, size_t* dstSizePtr, - const void* srcBuffer, size_t* srcSizePtr, - const LZ4F_decompressOptions_t* dOptPtr); +LZ4FLIB_API size_t +LZ4F_decompress(LZ4F_dctx* dctx, + void* dstBuffer, size_t* dstSizePtr, + const void* srcBuffer, size_t* srcSizePtr, + const LZ4F_decompressOptions_t* dOptPtr); /*! LZ4F_resetDecompressionContext() : added in v1.8.0 @@ -529,6 +546,8 @@ ITEM(ERROR_headerChecksum_invalid) \ ITEM(ERROR_contentChecksum_invalid) \ ITEM(ERROR_frameDecoding_alreadyStarted) \ + ITEM(ERROR_compressionState_uninitialized) \ + ITEM(ERROR_parameter_null) \ ITEM(ERROR_maxCode) #define LZ4F_GENERATE_ENUM(ENUM) LZ4F_##ENUM, @@ -539,8 +558,32 @@ LZ4FLIB_STATIC_API LZ4F_errorCodes LZ4F_getErrorCode(size_t functionResult); -LZ4FLIB_STATIC_API size_t LZ4F_getBlockSize(unsigned); +/*! LZ4F_getBlockSize() : + * Return, in scalar format (size_t), + * the maximum block size associated with blockSizeID. +**/ +LZ4FLIB_STATIC_API size_t LZ4F_getBlockSize(LZ4F_blockSizeID_t blockSizeID); + +/*! LZ4F_uncompressedUpdate() : + * LZ4F_uncompressedUpdate() can be called repetitively to add as much data uncompressed data as necessary. + * Important rule: dstCapacity MUST be large enough to store the entire source buffer as + * no compression is done for this operation + * If this condition is not respected, LZ4F_uncompressedUpdate() will fail (result is an errorCode). + * After an error, the state is left in a UB state, and must be re-initialized or freed. + * If previously a compressed block was written, buffered data is flushed + * before appending uncompressed data is continued. + * This is only supported when LZ4F_blockIndependent is used + * `cOptPtr` is optional : NULL can be provided, in which case all options are set to default. + * @return : number of bytes written into `dstBuffer` (it can be zero, meaning input data was just buffered). + * or an error code if it fails (which can be tested using LZ4F_isError()) + */ +LZ4FLIB_STATIC_API size_t +LZ4F_uncompressedUpdate(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const void* srcBuffer, size_t srcSize, + const LZ4F_compressOptions_t* cOptPtr); + /********************************** * Bulk processing dictionary API *********************************/ @@ -583,12 +626,12 @@ * but it's not recommended, as it's the only way to provide dictID in the frame header. * @return : number of bytes written into dstBuffer. * or an error code if it fails (can be tested using LZ4F_isError()) */ -LZ4FLIB_STATIC_API size_t LZ4F_compressFrame_usingCDict( - LZ4F_cctx* cctx, - void* dst, size_t dstCapacity, - const void* src, size_t srcSize, - const LZ4F_CDict* cdict, - const LZ4F_preferences_t* preferencesPtr); +LZ4FLIB_STATIC_API size_t +LZ4F_compressFrame_usingCDict(LZ4F_cctx* cctx, + void* dst, size_t dstCapacity, + const void* src, size_t srcSize, + const LZ4F_CDict* cdict, + const LZ4F_preferences_t* preferencesPtr); /*! LZ4F_compressBegin_usingCDict() : @@ -598,23 +641,49 @@ * however, it's the only way to provide dictID in the frame header. * @return : number of bytes written into dstBuffer for the header, * or an error code (which can be tested using LZ4F_isError()) */ -LZ4FLIB_STATIC_API size_t LZ4F_compressBegin_usingCDict( - LZ4F_cctx* cctx, - void* dstBuffer, size_t dstCapacity, - const LZ4F_CDict* cdict, - const LZ4F_preferences_t* prefsPtr); +LZ4FLIB_STATIC_API size_t +LZ4F_compressBegin_usingCDict(LZ4F_cctx* cctx, + void* dstBuffer, size_t dstCapacity, + const LZ4F_CDict* cdict, + const LZ4F_preferences_t* prefsPtr); /*! LZ4F_decompress_usingDict() : * Same as LZ4F_decompress(), using a predefined dictionary. * Dictionary is used "in place", without any preprocessing. - * It must remain accessible throughout the entire frame decoding. */ -LZ4FLIB_STATIC_API size_t LZ4F_decompress_usingDict( - LZ4F_dctx* dctxPtr, - void* dstBuffer, size_t* dstSizePtr, - const void* srcBuffer, size_t* srcSizePtr, - const void* dict, size_t dictSize, - const LZ4F_decompressOptions_t* decompressOptionsPtr); +** It must remain accessible throughout the entire frame decoding. */ +LZ4FLIB_STATIC_API size_t +LZ4F_decompress_usingDict(LZ4F_dctx* dctxPtr, + void* dstBuffer, size_t* dstSizePtr, + const void* srcBuffer, size_t* srcSizePtr, + const void* dict, size_t dictSize, + const LZ4F_decompressOptions_t* decompressOptionsPtr); + + +/*! Custom memory allocation : + * These prototypes make it possible to pass custom allocation/free functions. + * LZ4F_customMem is provided at state creation time, using LZ4F_create*_advanced() listed below. + * All allocation/free operations will be completed using these custom variants instead of regular ones. + */ +typedef void* (*LZ4F_AllocFunction) (void* opaqueState, size_t size); +typedef void* (*LZ4F_CallocFunction) (void* opaqueState, size_t size); +typedef void (*LZ4F_FreeFunction) (void* opaqueState, void* address); +typedef struct { + LZ4F_AllocFunction customAlloc; + LZ4F_CallocFunction customCalloc; /* optional; when not defined, uses customAlloc + memset */ + LZ4F_FreeFunction customFree; + void* opaqueState; +} LZ4F_CustomMem; +static +#ifdef __GNUC__ +__attribute__((__unused__)) +#endif +LZ4F_CustomMem const LZ4F_defaultCMem = { NULL, NULL, NULL, NULL }; /**< this constant defers to stdlib's functions */ + +LZ4FLIB_STATIC_API LZ4F_cctx* LZ4F_createCompressionContext_advanced(LZ4F_CustomMem customMem, unsigned version); +LZ4FLIB_STATIC_API LZ4F_dctx* LZ4F_createDecompressionContext_advanced(LZ4F_CustomMem customMem, unsigned version); +LZ4FLIB_STATIC_API LZ4F_CDict* LZ4F_createCDict_advanced(LZ4F_CustomMem customMem, const void* dictBuffer, size_t dictSize); + #if defined (__cplusplus) } diff --git a/cbits/lz4hc.c b/cbits/lz4hc.c --- a/cbits/lz4hc.c +++ b/cbits/lz4hc.c @@ -1,6 +1,6 @@ /* LZ4 HC - High Compression Mode of LZ4 - Copyright (C) 2011-2017, Yann Collet. + Copyright (C) 2011-2020, Yann Collet. BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) @@ -42,7 +42,7 @@ * Select how default compression function will allocate workplace memory, * in stack (0:fastest), or in heap (1:requires malloc()). * Since workplace is rather large, heap mode is recommended. - */ +**/ #ifndef LZ4HC_HEAPMODE # define LZ4HC_HEAPMODE 1 #endif @@ -99,18 +99,20 @@ static void LZ4HC_init_internal (LZ4HC_CCtx_internal* hc4, const BYTE* start) { - uptrval startingOffset = (uptrval)(hc4->end - hc4->base); - if (startingOffset > 1 GB) { + size_t const bufferSize = (size_t)(hc4->end - hc4->prefixStart); + size_t newStartingOffset = bufferSize + hc4->dictLimit; + assert(newStartingOffset >= bufferSize); /* check overflow */ + if (newStartingOffset > 1 GB) { LZ4HC_clearTables(hc4); - startingOffset = 0; + newStartingOffset = 0; } - startingOffset += 64 KB; - hc4->nextToUpdate = (U32) startingOffset; - hc4->base = start - startingOffset; + newStartingOffset += 64 KB; + hc4->nextToUpdate = (U32)newStartingOffset; + hc4->prefixStart = start; hc4->end = start; - hc4->dictBase = start - startingOffset; - hc4->dictLimit = (U32) startingOffset; - hc4->lowLimit = (U32) startingOffset; + hc4->dictStart = start; + hc4->dictLimit = (U32)newStartingOffset; + hc4->lowLimit = (U32)newStartingOffset; } @@ -119,12 +121,15 @@ { U16* const chainTable = hc4->chainTable; U32* const hashTable = hc4->hashTable; - const BYTE* const base = hc4->base; - U32 const target = (U32)(ip - base); + const BYTE* const prefixPtr = hc4->prefixStart; + U32 const prefixIdx = hc4->dictLimit; + U32 const target = (U32)(ip - prefixPtr) + prefixIdx; U32 idx = hc4->nextToUpdate; + assert(ip >= prefixPtr); + assert(target >= prefixIdx); while (idx < target) { - U32 const h = LZ4HC_hashPtr(base+idx); + U32 const h = LZ4HC_hashPtr(prefixPtr+idx-prefixIdx); size_t delta = idx - hashTable[h]; if (delta>LZ4_DISTANCE_MAX) delta = LZ4_DISTANCE_MAX; DELTANEXTU16(chainTable, idx) = (U16)delta; @@ -193,15 +198,14 @@ BYTE const byte = (BYTE)(pattern >> bitOffset); if (*ip != byte) break; ip ++; bitOffset -= 8; - } - } + } } return (unsigned)(ip - iStart); } /* LZ4HC_reverseCountPattern() : * pattern must be a sample of repetitive pattern of length 1, 2 or 4 (but not 3!) - * read using natural platform endianess */ + * read using natural platform endianness */ static unsigned LZ4HC_reverseCountPattern(const BYTE* ip, const BYTE* const iLow, U32 pattern) { @@ -211,7 +215,7 @@ if (LZ4_read32(ip-4) != pattern) break; ip -= 4; } - { const BYTE* bytePtr = (const BYTE*)(&pattern) + 3; /* works for any endianess */ + { const BYTE* bytePtr = (const BYTE*)(&pattern) + 3; /* works for any endianness */ while (likely(ip>iLow)) { if (ip[-1] != *bytePtr) break; ip--; bytePtr--; @@ -234,28 +238,28 @@ LZ4_FORCE_INLINE int LZ4HC_InsertAndGetWiderMatch ( - LZ4HC_CCtx_internal* hc4, - const BYTE* const ip, - const BYTE* const iLowLimit, - const BYTE* const iHighLimit, - int longest, - const BYTE** matchpos, - const BYTE** startpos, - const int maxNbAttempts, - const int patternAnalysis, - const int chainSwap, - const dictCtx_directive dict, - const HCfavor_e favorDecSpeed) + LZ4HC_CCtx_internal* const hc4, + const BYTE* const ip, + const BYTE* const iLowLimit, const BYTE* const iHighLimit, + int longest, + const BYTE** matchpos, + const BYTE** startpos, + const int maxNbAttempts, + const int patternAnalysis, const int chainSwap, + const dictCtx_directive dict, + const HCfavor_e favorDecSpeed) { U16* const chainTable = hc4->chainTable; U32* const HashTable = hc4->hashTable; const LZ4HC_CCtx_internal * const dictCtx = hc4->dictCtx; - const BYTE* const base = hc4->base; - const U32 dictLimit = hc4->dictLimit; - const BYTE* const lowPrefixPtr = base + dictLimit; - const U32 ipIndex = (U32)(ip - base); - const U32 lowestMatchIndex = (hc4->lowLimit + (LZ4_DISTANCE_MAX + 1) > ipIndex) ? hc4->lowLimit : ipIndex - LZ4_DISTANCE_MAX; - const BYTE* const dictBase = hc4->dictBase; + const BYTE* const prefixPtr = hc4->prefixStart; + const U32 prefixIdx = hc4->dictLimit; + const U32 ipIndex = (U32)(ip - prefixPtr) + prefixIdx; + const int withinStartDistance = (hc4->lowLimit + (LZ4_DISTANCE_MAX + 1) > ipIndex); + const U32 lowestMatchIndex = (withinStartDistance) ? hc4->lowLimit : ipIndex - LZ4_DISTANCE_MAX; + const BYTE* const dictStart = hc4->dictStart; + const U32 dictIdx = hc4->lowLimit; + const BYTE* const dictEnd = dictStart + prefixIdx - dictIdx; int const lookBackLength = (int)(ip-iLowLimit); int nbAttempts = maxNbAttempts; U32 matchChainPos = 0; @@ -277,14 +281,13 @@ assert(matchIndex < ipIndex); if (favorDecSpeed && (ipIndex - matchIndex < 8)) { /* do nothing */ - } else if (matchIndex >= dictLimit) { /* within current Prefix */ - const BYTE* const matchPtr = base + matchIndex; - assert(matchPtr >= lowPrefixPtr); + } else if (matchIndex >= prefixIdx) { /* within current Prefix */ + const BYTE* const matchPtr = prefixPtr + matchIndex - prefixIdx; assert(matchPtr < ip); assert(longest >= 1); if (LZ4_read16(iLowLimit + longest - 1) == LZ4_read16(matchPtr - lookBackLength + longest - 1)) { if (LZ4_read32(matchPtr) == pattern) { - int const back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, lowPrefixPtr) : 0; + int const back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, prefixPtr) : 0; matchLength = MINMATCH + (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, iHighLimit); matchLength -= back; if (matchLength > longest) { @@ -293,24 +296,25 @@ *startpos = ip + back; } } } } else { /* lowestMatchIndex <= matchIndex < dictLimit */ - const BYTE* const matchPtr = dictBase + matchIndex; - if (LZ4_read32(matchPtr) == pattern) { - const BYTE* const dictStart = dictBase + hc4->lowLimit; + const BYTE* const matchPtr = dictStart + (matchIndex - dictIdx); + assert(matchIndex >= dictIdx); + if ( likely(matchIndex <= prefixIdx - 4) + && (LZ4_read32(matchPtr) == pattern) ) { int back = 0; - const BYTE* vLimit = ip + (dictLimit - matchIndex); + const BYTE* vLimit = ip + (prefixIdx - matchIndex); if (vLimit > iHighLimit) vLimit = iHighLimit; matchLength = (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH; if ((ip+matchLength == vLimit) && (vLimit < iHighLimit)) - matchLength += LZ4_count(ip+matchLength, lowPrefixPtr, iHighLimit); + matchLength += LZ4_count(ip+matchLength, prefixPtr, iHighLimit); back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictStart) : 0; matchLength -= back; if (matchLength > longest) { longest = matchLength; - *matchpos = base + matchIndex + back; /* virtual pos, relative to ip, to retrieve offset */ + *matchpos = prefixPtr - prefixIdx + matchIndex + back; /* virtual pos, relative to ip, to retrieve offset */ *startpos = ip + back; } } } - if (chainSwap && matchLength==longest) { /* better match => select a better chain */ + if (chainSwap && matchLength==longest) { /* better match => select a better chain */ assert(lookBackLength==0); /* search forward only */ if (matchIndex + (U32)longest <= ipIndex) { int const kTrigger = 4; @@ -326,8 +330,7 @@ distanceToNextMatch = candidateDist; matchChainPos = (U32)pos; accel = 1 << kTrigger; - } - } + } } if (distanceToNextMatch > 1) { if (distanceToNextMatch > matchIndex) break; /* avoid overflow */ matchIndex -= distanceToNextMatch; @@ -347,23 +350,24 @@ repeat = rep_not; } } if ( (repeat == rep_confirmed) && (matchCandidateIdx >= lowestMatchIndex) - && LZ4HC_protectDictEnd(dictLimit, matchCandidateIdx) ) { - const int extDict = matchCandidateIdx < dictLimit; - const BYTE* const matchPtr = (extDict ? dictBase : base) + matchCandidateIdx; + && LZ4HC_protectDictEnd(prefixIdx, matchCandidateIdx) ) { + const int extDict = matchCandidateIdx < prefixIdx; + const BYTE* const matchPtr = (extDict ? dictStart - dictIdx : prefixPtr - prefixIdx) + matchCandidateIdx; if (LZ4_read32(matchPtr) == pattern) { /* good candidate */ - const BYTE* const dictStart = dictBase + hc4->lowLimit; - const BYTE* const iLimit = extDict ? dictBase + dictLimit : iHighLimit; + const BYTE* const iLimit = extDict ? dictEnd : iHighLimit; size_t forwardPatternLength = LZ4HC_countPattern(matchPtr+sizeof(pattern), iLimit, pattern) + sizeof(pattern); if (extDict && matchPtr + forwardPatternLength == iLimit) { U32 const rotatedPattern = LZ4HC_rotatePattern(forwardPatternLength, pattern); - forwardPatternLength += LZ4HC_countPattern(lowPrefixPtr, iHighLimit, rotatedPattern); + forwardPatternLength += LZ4HC_countPattern(prefixPtr, iHighLimit, rotatedPattern); } - { const BYTE* const lowestMatchPtr = extDict ? dictStart : lowPrefixPtr; + { const BYTE* const lowestMatchPtr = extDict ? dictStart : prefixPtr; size_t backLength = LZ4HC_reverseCountPattern(matchPtr, lowestMatchPtr, pattern); size_t currentSegmentLength; - if (!extDict && matchPtr - backLength == lowPrefixPtr && hc4->lowLimit < dictLimit) { + if (!extDict + && matchPtr - backLength == prefixPtr + && dictIdx < prefixIdx) { U32 const rotatedPattern = LZ4HC_rotatePattern((U32)(-(int)backLength), pattern); - backLength += LZ4HC_reverseCountPattern(dictBase + dictLimit, dictStart, rotatedPattern); + backLength += LZ4HC_reverseCountPattern(dictEnd, dictStart, rotatedPattern); } /* Limit backLength not go further than lowestMatchIndex */ backLength = matchCandidateIdx - MAX(matchCandidateIdx - (U32)backLength, lowestMatchIndex); @@ -373,28 +377,28 @@ if ( (currentSegmentLength >= srcPatternLength) /* current pattern segment large enough to contain full srcPatternLength */ && (forwardPatternLength <= srcPatternLength) ) { /* haven't reached this position yet */ U32 const newMatchIndex = matchCandidateIdx + (U32)forwardPatternLength - (U32)srcPatternLength; /* best position, full pattern, might be followed by more match */ - if (LZ4HC_protectDictEnd(dictLimit, newMatchIndex)) + if (LZ4HC_protectDictEnd(prefixIdx, newMatchIndex)) matchIndex = newMatchIndex; else { /* Can only happen if started in the prefix */ - assert(newMatchIndex >= dictLimit - 3 && newMatchIndex < dictLimit && !extDict); - matchIndex = dictLimit; + assert(newMatchIndex >= prefixIdx - 3 && newMatchIndex < prefixIdx && !extDict); + matchIndex = prefixIdx; } } else { U32 const newMatchIndex = matchCandidateIdx - (U32)backLength; /* farthest position in current segment, will find a match of length currentSegmentLength + maybe some back */ - if (!LZ4HC_protectDictEnd(dictLimit, newMatchIndex)) { - assert(newMatchIndex >= dictLimit - 3 && newMatchIndex < dictLimit && !extDict); - matchIndex = dictLimit; + if (!LZ4HC_protectDictEnd(prefixIdx, newMatchIndex)) { + assert(newMatchIndex >= prefixIdx - 3 && newMatchIndex < prefixIdx && !extDict); + matchIndex = prefixIdx; } else { matchIndex = newMatchIndex; if (lookBackLength==0) { /* no back possible */ size_t const maxML = MIN(currentSegmentLength, srcPatternLength); if ((size_t)longest < maxML) { - assert(base + matchIndex != ip); - if ((size_t)(ip - base) - matchIndex > LZ4_DISTANCE_MAX) break; + assert(prefixPtr - prefixIdx + matchIndex != ip); + if ((size_t)(ip - prefixPtr) + prefixIdx - matchIndex > LZ4_DISTANCE_MAX) break; assert(maxML < 2 GB); longest = (int)maxML; - *matchpos = base + matchIndex; /* virtual pos, relative to ip, to retrieve offset */ + *matchpos = prefixPtr - prefixIdx + matchIndex; /* virtual pos, relative to ip, to retrieve offset */ *startpos = ip; } { U32 const distToNextPattern = DELTANEXTU16(chainTable, matchIndex); @@ -413,12 +417,12 @@ if ( dict == usingDictCtxHc && nbAttempts > 0 && ipIndex - lowestMatchIndex < LZ4_DISTANCE_MAX) { - size_t const dictEndOffset = (size_t)(dictCtx->end - dictCtx->base); + size_t const dictEndOffset = (size_t)(dictCtx->end - dictCtx->prefixStart) + dictCtx->dictLimit; U32 dictMatchIndex = dictCtx->hashTable[LZ4HC_hashPtr(ip)]; assert(dictEndOffset <= 1 GB); matchIndex = dictMatchIndex + lowestMatchIndex - (U32)dictEndOffset; while (ipIndex - matchIndex <= LZ4_DISTANCE_MAX && nbAttempts--) { - const BYTE* const matchPtr = dictCtx->base + dictMatchIndex; + const BYTE* const matchPtr = dictCtx->prefixStart - dictCtx->dictLimit + dictMatchIndex; if (LZ4_read32(matchPtr) == pattern) { int mlt; @@ -426,11 +430,11 @@ const BYTE* vLimit = ip + (dictEndOffset - dictMatchIndex); if (vLimit > iHighLimit) vLimit = iHighLimit; mlt = (int)LZ4_count(ip+MINMATCH, matchPtr+MINMATCH, vLimit) + MINMATCH; - back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictCtx->base + dictCtx->dictLimit) : 0; + back = lookBackLength ? LZ4HC_countBack(ip, matchPtr, iLowLimit, dictCtx->prefixStart) : 0; mlt -= back; if (mlt > longest) { longest = mlt; - *matchpos = base + matchIndex + back; + *matchpos = prefixPtr - prefixIdx + matchIndex + back; *startpos = ip + back; } } @@ -442,13 +446,13 @@ return longest; } -LZ4_FORCE_INLINE -int LZ4HC_InsertAndFindBestMatch(LZ4HC_CCtx_internal* const hc4, /* Index table will be updated */ - const BYTE* const ip, const BYTE* const iLimit, - const BYTE** matchpos, - const int maxNbAttempts, - const int patternAnalysis, - const dictCtx_directive dict) +LZ4_FORCE_INLINE int +LZ4HC_InsertAndFindBestMatch(LZ4HC_CCtx_internal* const hc4, /* Index table will be updated */ + const BYTE* const ip, const BYTE* const iLimit, + const BYTE** matchpos, + const int maxNbAttempts, + const int patternAnalysis, + const dictCtx_directive dict) { const BYTE* uselessPtr = ip; /* note : LZ4HC_InsertAndGetWiderMatch() is able to modify the starting position of a match (*startpos), @@ -751,7 +755,7 @@ } else { *op++ = (BYTE)(lastRunSize << ML_BITS); } - memcpy(op, anchor, lastRunSize); + LZ4_memcpy(op, anchor, lastRunSize); op += lastRunSize; } @@ -884,13 +888,13 @@ limitedOutput_directive limit ) { - const size_t position = (size_t)(ctx->end - ctx->base) - ctx->lowLimit; + const size_t position = (size_t)(ctx->end - ctx->prefixStart) + (ctx->dictLimit - ctx->lowLimit); assert(ctx->dictCtx != NULL); if (position >= 64 KB) { ctx->dictCtx = NULL; return LZ4HC_compress_generic_noDictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit); } else if (position == 0 && *srcSizePtr > 4 KB) { - memcpy(ctx, ctx->dictCtx, sizeof(LZ4HC_CCtx_internal)); + LZ4_memcpy(ctx, ctx->dictCtx, sizeof(LZ4HC_CCtx_internal)); LZ4HC_setExternalDict(ctx, (const BYTE *)src); ctx->compressionLevel = (short)cLevel; return LZ4HC_compress_generic_noDictCtx(ctx, src, dst, srcSizePtr, dstCapacity, cLevel, limit); @@ -953,13 +957,15 @@ int LZ4_compress_HC(const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel) { + int cSize; #if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 LZ4_streamHC_t* const statePtr = (LZ4_streamHC_t*)ALLOC(sizeof(LZ4_streamHC_t)); + if (statePtr==NULL) return 0; #else LZ4_streamHC_t state; LZ4_streamHC_t* const statePtr = &state; #endif - int const cSize = LZ4_compress_HC_extStateHC(statePtr, src, dst, srcSize, dstCapacity, compressionLevel); + cSize = LZ4_compress_HC_extStateHC(statePtr, src, dst, srcSize, dstCapacity, compressionLevel); #if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 FREEMEM(statePtr); #endif @@ -982,6 +988,7 @@ * Streaming Functions **************************************/ /* allocation */ +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) LZ4_streamHC_t* LZ4_createStreamHC(void) { LZ4_streamHC_t* const state = @@ -998,13 +1005,12 @@ FREEMEM(LZ4_streamHCPtr); return 0; } +#endif LZ4_streamHC_t* LZ4_initStreamHC (void* buffer, size_t size) { LZ4_streamHC_t* const LZ4_streamHCPtr = (LZ4_streamHC_t*)buffer; - /* if compilation fails here, LZ4_STREAMHCSIZE must be increased */ - LZ4_STATIC_ASSERT(sizeof(LZ4HC_CCtx_internal) <= LZ4_STREAMHCSIZE); DEBUGLOG(4, "LZ4_initStreamHC(%p, %u)", buffer, (unsigned)size); /* check conditions */ if (buffer == NULL) return NULL; @@ -1030,9 +1036,13 @@ if (LZ4_streamHCPtr->internal_donotuse.dirty) { LZ4_initStreamHC(LZ4_streamHCPtr, sizeof(*LZ4_streamHCPtr)); } else { - /* preserve end - base : can trigger clearTable's threshold */ - LZ4_streamHCPtr->internal_donotuse.end -= (uptrval)LZ4_streamHCPtr->internal_donotuse.base; - LZ4_streamHCPtr->internal_donotuse.base = NULL; + /* preserve end - prefixStart : can trigger clearTable's threshold */ + if (LZ4_streamHCPtr->internal_donotuse.end != NULL) { + LZ4_streamHCPtr->internal_donotuse.end -= (uptrval)LZ4_streamHCPtr->internal_donotuse.prefixStart; + } else { + assert(LZ4_streamHCPtr->internal_donotuse.prefixStart == NULL); + } + LZ4_streamHCPtr->internal_donotuse.prefixStart = NULL; LZ4_streamHCPtr->internal_donotuse.dictCtx = NULL; } LZ4_setCompressionLevel(LZ4_streamHCPtr, compressionLevel); @@ -1083,14 +1093,14 @@ static void LZ4HC_setExternalDict(LZ4HC_CCtx_internal* ctxPtr, const BYTE* newBlock) { DEBUGLOG(4, "LZ4HC_setExternalDict(%p, %p)", ctxPtr, newBlock); - if (ctxPtr->end >= ctxPtr->base + ctxPtr->dictLimit + 4) + if (ctxPtr->end >= ctxPtr->prefixStart + 4) LZ4HC_Insert (ctxPtr, ctxPtr->end-3); /* Referencing remaining dictionary content */ /* Only one memory segment for extDict, so any previous extDict is lost at this stage */ ctxPtr->lowLimit = ctxPtr->dictLimit; - ctxPtr->dictLimit = (U32)(ctxPtr->end - ctxPtr->base); - ctxPtr->dictBase = ctxPtr->base; - ctxPtr->base = newBlock - ctxPtr->dictLimit; + ctxPtr->dictStart = ctxPtr->prefixStart; + ctxPtr->dictLimit += (U32)(ctxPtr->end - ctxPtr->prefixStart); + ctxPtr->prefixStart = newBlock; ctxPtr->end = newBlock; ctxPtr->nextToUpdate = ctxPtr->dictLimit; /* match referencing will resume from there */ @@ -1109,11 +1119,11 @@ LZ4_streamHCPtr, src, *srcSizePtr, limit); assert(ctxPtr != NULL); /* auto-init if forgotten */ - if (ctxPtr->base == NULL) LZ4HC_init_internal (ctxPtr, (const BYTE*) src); + if (ctxPtr->prefixStart == NULL) LZ4HC_init_internal (ctxPtr, (const BYTE*) src); /* Check overflow */ - if ((size_t)(ctxPtr->end - ctxPtr->base) > 2 GB) { - size_t dictSize = (size_t)(ctxPtr->end - ctxPtr->base) - ctxPtr->dictLimit; + if ((size_t)(ctxPtr->end - ctxPtr->prefixStart) + ctxPtr->dictLimit > 2 GB) { + size_t dictSize = (size_t)(ctxPtr->end - ctxPtr->prefixStart); if (dictSize > 64 KB) dictSize = 64 KB; LZ4_loadDictHC(LZ4_streamHCPtr, (const char*)(ctxPtr->end) - dictSize, (int)dictSize); } @@ -1124,13 +1134,16 @@ /* Check overlapping input/dictionary space */ { const BYTE* sourceEnd = (const BYTE*) src + *srcSizePtr; - const BYTE* const dictBegin = ctxPtr->dictBase + ctxPtr->lowLimit; - const BYTE* const dictEnd = ctxPtr->dictBase + ctxPtr->dictLimit; + const BYTE* const dictBegin = ctxPtr->dictStart; + const BYTE* const dictEnd = ctxPtr->dictStart + (ctxPtr->dictLimit - ctxPtr->lowLimit); if ((sourceEnd > dictBegin) && ((const BYTE*)src < dictEnd)) { if (sourceEnd > dictEnd) sourceEnd = dictEnd; - ctxPtr->lowLimit = (U32)(sourceEnd - ctxPtr->dictBase); - if (ctxPtr->dictLimit - ctxPtr->lowLimit < 4) ctxPtr->lowLimit = ctxPtr->dictLimit; - } } + ctxPtr->lowLimit += (U32)(sourceEnd - ctxPtr->dictStart); + ctxPtr->dictStart += (U32)(sourceEnd - ctxPtr->dictStart); + if (ctxPtr->dictLimit - ctxPtr->lowLimit < 4) { + ctxPtr->lowLimit = ctxPtr->dictLimit; + ctxPtr->dictStart = ctxPtr->prefixStart; + } } } return LZ4HC_compress_generic (ctxPtr, src, dst, srcSizePtr, dstCapacity, ctxPtr->compressionLevel, limit); } @@ -1158,7 +1171,7 @@ int LZ4_saveDictHC (LZ4_streamHC_t* LZ4_streamHCPtr, char* safeBuffer, int dictSize) { LZ4HC_CCtx_internal* const streamPtr = &LZ4_streamHCPtr->internal_donotuse; - int const prefixSize = (int)(streamPtr->end - (streamPtr->base + streamPtr->dictLimit)); + int const prefixSize = (int)(streamPtr->end - streamPtr->prefixStart); DEBUGLOG(5, "LZ4_saveDictHC(%p, %p, %d)", LZ4_streamHCPtr, safeBuffer, dictSize); assert(prefixSize >= 0); if (dictSize > 64 KB) dictSize = 64 KB; @@ -1166,12 +1179,13 @@ if (dictSize > prefixSize) dictSize = prefixSize; if (safeBuffer == NULL) assert(dictSize == 0); if (dictSize > 0) - memmove(safeBuffer, streamPtr->end - dictSize, dictSize); - { U32 const endIndex = (U32)(streamPtr->end - streamPtr->base); + LZ4_memmove(safeBuffer, streamPtr->end - dictSize, dictSize); + { U32 const endIndex = (U32)(streamPtr->end - streamPtr->prefixStart) + streamPtr->dictLimit; streamPtr->end = (const BYTE*)safeBuffer + dictSize; - streamPtr->base = streamPtr->end - endIndex; + streamPtr->prefixStart = streamPtr->end - dictSize; streamPtr->dictLimit = endIndex - (U32)dictSize; streamPtr->lowLimit = endIndex - (U32)dictSize; + streamPtr->dictStart = streamPtr->prefixStart; if (streamPtr->nextToUpdate < streamPtr->dictLimit) streamPtr->nextToUpdate = streamPtr->dictLimit; } @@ -1199,7 +1213,7 @@ /* Deprecated streaming functions */ -int LZ4_sizeofStreamStateHC(void) { return LZ4_STREAMHCSIZE; } +int LZ4_sizeofStreamStateHC(void) { return sizeof(LZ4_streamHC_t); } /* state is presumed correctly sized, aka >= sizeof(LZ4_streamHC_t) * @return : 0 on success, !=0 if error */ @@ -1211,6 +1225,7 @@ return 0; } +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) void* LZ4_createHC (const char* inputBuffer) { LZ4_streamHC_t* const hc4 = LZ4_createStreamHC(); @@ -1225,6 +1240,7 @@ FREEMEM(LZ4HC_Data); return 0; } +#endif int LZ4_compressHC2_continue (void* LZ4HC_Data, const char* src, char* dst, int srcSize, int cLevel) { @@ -1238,11 +1254,11 @@ char* LZ4_slideInputBufferHC(void* LZ4HC_Data) { - LZ4_streamHC_t *ctx = (LZ4_streamHC_t*)LZ4HC_Data; - const BYTE *bufferStart = ctx->internal_donotuse.base + ctx->internal_donotuse.lowLimit; + LZ4_streamHC_t* const ctx = (LZ4_streamHC_t*)LZ4HC_Data; + const BYTE* bufferStart = ctx->internal_donotuse.prefixStart - ctx->internal_donotuse.dictLimit + ctx->internal_donotuse.lowLimit; LZ4_resetStreamHC_fast(ctx, ctx->internal_donotuse.compressionLevel); /* avoid const char * -> char * conversion warning :( */ - return (char *)(uptrval)bufferStart; + return (char*)(uptrval)bufferStart; } @@ -1325,7 +1341,7 @@ { int retval = 0; #define TRAILING_LITERALS 3 -#ifdef LZ4HC_HEAPMODE +#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 LZ4HC_optimal_t* const opt = (LZ4HC_optimal_t*)ALLOC(sizeof(LZ4HC_optimal_t) * (LZ4_OPT_NUM + TRAILING_LITERALS)); #else LZ4HC_optimal_t opt[LZ4_OPT_NUM + TRAILING_LITERALS]; /* ~64 KB, which is a bit large for stack... */ @@ -1343,7 +1359,7 @@ const BYTE* ovref = NULL; /* init */ -#ifdef LZ4HC_HEAPMODE +#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 if (opt == NULL) goto _return_label; #endif DEBUGLOG(5, "LZ4HC_compress_optimal(dst=%p, dstCapa=%u)", dst, (unsigned)dstCapacity); @@ -1575,7 +1591,7 @@ } else { *op++ = (BYTE)(lastRunSize << ML_BITS); } - memcpy(op, anchor, lastRunSize); + LZ4_memcpy(op, anchor, lastRunSize); op += lastRunSize; } @@ -1608,7 +1624,7 @@ goto _last_literals; } _return_label: -#ifdef LZ4HC_HEAPMODE +#if defined(LZ4HC_HEAPMODE) && LZ4HC_HEAPMODE==1 FREEMEM(opt); #endif return retval; diff --git a/cbits/lz4hc.h b/cbits/lz4hc.h --- a/cbits/lz4hc.h +++ b/cbits/lz4hc.h @@ -1,7 +1,7 @@ /* LZ4 HC - High Compression Mode of LZ4 Header File - Copyright (C) 2011-2017, Yann Collet. + Copyright (C) 2011-2020, Yann Collet. BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php) Redistribution and use in source and binary forms, with or without @@ -198,14 +198,17 @@ #define LZ4HC_HASH_MASK (LZ4HC_HASHTABLESIZE - 1) +/* Never ever use these definitions directly ! + * Declare or allocate an LZ4_streamHC_t instead. +**/ typedef struct LZ4HC_CCtx_internal LZ4HC_CCtx_internal; struct LZ4HC_CCtx_internal { LZ4_u32 hashTable[LZ4HC_HASHTABLESIZE]; LZ4_u16 chainTable[LZ4HC_MAXD]; const LZ4_byte* end; /* next block here to continue on current prefix */ - const LZ4_byte* base; /* All index relative to this position */ - const LZ4_byte* dictBase; /* alternate base for extDict */ + const LZ4_byte* prefixStart; /* Indexes relative to this position */ + const LZ4_byte* dictStart; /* alternate reference for extDict */ LZ4_u32 dictLimit; /* below that point, need extDict */ LZ4_u32 lowLimit; /* below that point, no more dict */ LZ4_u32 nextToUpdate; /* index from which to continue dictionary update */ @@ -216,20 +219,15 @@ const LZ4HC_CCtx_internal* dictCtx; }; - -/* Do not use these definitions directly ! - * Declare or allocate an LZ4_streamHC_t instead. - */ -#define LZ4_STREAMHCSIZE 262200 /* static size, for inter-version compatibility */ -#define LZ4_STREAMHCSIZE_VOIDP (LZ4_STREAMHCSIZE / sizeof(void*)) +#define LZ4_STREAMHC_MINSIZE 262200 /* static size, for inter-version compatibility */ union LZ4_streamHC_u { - void* table[LZ4_STREAMHCSIZE_VOIDP]; + char minStateSize[LZ4_STREAMHC_MINSIZE]; LZ4HC_CCtx_internal internal_donotuse; }; /* previously typedef'd to LZ4_streamHC_t */ /* LZ4_streamHC_t : * This structure allows static allocation of LZ4 HC streaming state. - * This can be used to allocate statically, on state, or as part of a larger structure. + * This can be used to allocate statically on stack, or as part of a larger structure. * * Such state **must** be initialized using LZ4_initStreamHC() before first use. * @@ -244,7 +242,7 @@ * Required before first use of a statically allocated LZ4_streamHC_t. * Before v1.9.0 : use LZ4_resetStreamHC() instead */ -LZ4LIB_API LZ4_streamHC_t* LZ4_initStreamHC (void* buffer, size_t size); +LZ4LIB_API LZ4_streamHC_t* LZ4_initStreamHC(void* buffer, size_t size); /*-************************************ @@ -272,9 +270,11 @@ * LZ4_slideInputBufferHC() will truncate the history of the stream, rather * than preserve a window-sized chunk of history. */ +#if !defined(LZ4_STATIC_LINKING_ONLY_DISABLE_MEMORY_ALLOCATION) LZ4_DEPRECATED("use LZ4_createStreamHC() instead") LZ4LIB_API void* LZ4_createHC (const char* inputBuffer); -LZ4_DEPRECATED("use LZ4_saveDictHC() instead") LZ4LIB_API char* LZ4_slideInputBufferHC (void* LZ4HC_Data); LZ4_DEPRECATED("use LZ4_freeStreamHC() instead") LZ4LIB_API int LZ4_freeHC (void* LZ4HC_Data); +#endif +LZ4_DEPRECATED("use LZ4_saveDictHC() instead") LZ4LIB_API char* LZ4_slideInputBufferHC (void* LZ4HC_Data); LZ4_DEPRECATED("use LZ4_compress_HC_continue() instead") LZ4LIB_API int LZ4_compressHC2_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int compressionLevel); LZ4_DEPRECATED("use LZ4_compress_HC_continue() instead") LZ4LIB_API int LZ4_compressHC2_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize, int compressionLevel); LZ4_DEPRECATED("use LZ4_createStreamHC() instead") LZ4LIB_API int LZ4_sizeofStreamStateHC(void); @@ -305,7 +305,7 @@ * They should not be linked from DLL, * as there is no guarantee of API stability yet. * Prototypes will be promoted to "stable" status - * after successfull usage in real-life scenarios. + * after successful usage in real-life scenarios. ***************************************************/ #ifdef LZ4_HC_STATIC_LINKING_ONLY /* protection macro */ #ifndef LZ4_HC_SLO_098092834 diff --git a/lz4-hs.cabal b/lz4-hs.cabal --- a/lz4-hs.cabal +++ b/lz4-hs.cabal @@ -1,6 +1,6 @@ cabal-version: 2.2 name: lz4-hs -version: 0.1.5.1 +version: 0.1.5.2 license: BSD-3-Clause license-file: LICENSE copyright: Copyright: (c) 2020 Vanessa McHale diff --git a/mem/Mem.hs b/mem/Mem.hs --- a/mem/Mem.hs +++ b/mem/Mem.hs @@ -14,10 +14,8 @@ decompressDump :: IO () decompressDump = withSystemTempDirectory "lz4" $ - \fp -> BSL.writeFile (fp "valgrind-3.15.0.tar") =<< - (decompress <$> BSL.readFile "valgrind-3.15.0.tar.lz4") + \fp -> BSL.writeFile (fp "valgrind-3.15.0.tar") . decompress =<< BSL.readFile "valgrind-3.15.0.tar.lz4" compressDump :: IO () compressDump = withSystemTempDirectory "lz4" $ - \fp -> BSL.writeFile (fp "valgrind-3.15.0.tar.lz4") =<< - (compress <$> BSL.readFile "valgrind-3.15.0.tar") + \fp -> BSL.writeFile (fp "valgrind-3.15.0.tar.lz4") . compress =<< BSL.readFile "valgrind-3.15.0.tar"