From 21fc1c831858d6e50771aadd8579aa2d782c3efa Mon Sep 17 00:00:00 2001 From: Tino Reichardt Date: Sun, 6 Dec 2020 11:07:42 +0100 Subject: [PATCH] Update lz4 to version 1.9.3 --- C/lz4/README.md | 37 +- C/lz4/lz4.c | 557 +++++++++++++++---------- C/lz4/lz4.h | 150 +++---- C/lz4/lz4frame.c | 147 ++++--- C/lz4/lz4frame.h | 36 +- C/lz4/lz4hc.c | 307 +++++++++----- C/lz4/lz4hc.h | 55 +-- CPP/7zip/Bundles/Codec_lz4/resource.rc | 2 +- README.md | 4 +- 9 files changed, 759 insertions(+), 536 deletions(-) diff --git a/C/lz4/README.md b/C/lz4/README.md index cba2c34ab..e2af868ff 100644 --- a/C/lz4/README.md +++ b/C/lz4/README.md @@ -35,21 +35,22 @@ So it's necessary to include all `*.c` and `*.h` files present in `/lib`. Definitions which are not guaranteed to remain stable in future versions, are protected behind macros, such as `LZ4_STATIC_LINKING_ONLY`. -As the name implies, these definitions can only be invoked +As the name strongly implies, these definitions should only be invoked in the context of static linking ***only***. Otherwise, dependent application may fail on API or ABI break in the future. -The associated symbols are also not present in dynamic library by default. +The associated symbols are also not exposed by the dynamic library by default. Should they be nonetheless needed, it's possible to force their publication -by using build macro `LZ4_PUBLISH_STATIC_FUNCTIONS`. +by using build macros `LZ4_PUBLISH_STATIC_FUNCTIONS` +and `LZ4F_PUBLISH_STATIC_FUNCTIONS`. #### Build macros -The following build macro can be selected at compilation time : +The following build macro can be selected to adjust source code behavior at compilation time : -- `LZ4_FAST_DEC_LOOP` : this triggers the optimized decompression loop. - This loops works great on x86/x64 cpus, and is automatically enabled on this platform. - It's possible to enable or disable it manually, by passing `LZ4_FAST_DEC_LOOP=1` or `0` to the preprocessor. +- `LZ4_FAST_DEC_LOOP` : this triggers a speed optimized decompression loop, more powerful on modern cpus. + This loop works great on `x86`, `x64` and `aarch64` cpus, and is automatically enabled for them. + It's also possible to enable or disable it manually, by passing `LZ4_FAST_DEC_LOOP=1` or `0` to the preprocessor. For example, with `gcc` : `-DLZ4_FAST_DEC_LOOP=1`, and with `make` : `CPPFLAGS+=-DLZ4_FAST_DEC_LOOP=1 make lz4`. @@ -65,8 +66,24 @@ The following build macro can be selected at compilation time : Should this be a problem, it's generally possible to make the compiler ignore these warnings, for example with `-Wno-deprecated-declarations` on `gcc`, or `_CRT_SECURE_NO_WARNINGS` for Visual Studio. - Another method is to define `LZ4_DISABLE_DEPRECATE_WARNINGS` - before including the LZ4 header files. + This build macro offers another project-specific method + by defining `LZ4_DISABLE_DEPRECATE_WARNINGS` before including the LZ4 header files. + +- `LZ4_USER_MEMORY_FUNCTIONS` : replace calls to 's `malloc`, `calloc` and `free` + by user-defined functions, which must be called `LZ4_malloc()`, `LZ4_calloc()` and `LZ4_free()`. + User functions must be available at link time. + +- `LZ4_FORCE_SW_BITCOUNT` : by default, the compression algorithm tries to determine lengths + by using bitcount instructions, generally implemented as fast single instructions in many cpus. + In case the target cpus doesn't support it, or compiler intrinsic doesn't work, or feature bad performance, + it's possible to use an optimized software path instead. + This is achieved by setting this build macros . + In most cases, it's not expected to be necessary, + but it can be legitimately considered for less common platforms. + +- `LZ4_ALIGN_TEST` : alignment test ensures that the memory area + passed as argument to become a compression state is suitably aligned. + This test can be disabled if it proves flaky, by setting this value to 0. #### Amalgamation @@ -102,7 +119,7 @@ The compiled executable will require LZ4 DLL which is available at `dll\liblz4.d #### Miscellaneous -Other files present in the directory are not source code. There are : +Other files present in the directory are not source code. They are : - `LICENSE` : contains the BSD license text - `Makefile` : `make` script to compile and install lz4 library (static and dynamic) diff --git a/C/lz4/lz4.c b/C/lz4/lz4.c index 9808d70ae..9f5e9bfa0 100644 --- a/C/lz4/lz4.c +++ b/C/lz4/lz4.c @@ -45,10 +45,16 @@ #endif /* - * ACCELERATION_DEFAULT : + * LZ4_ACCELERATION_DEFAULT : * Select "acceleration" for LZ4_compress_fast() when parameter value <= 0 */ -#define ACCELERATION_DEFAULT 1 +#define LZ4_ACCELERATION_DEFAULT 1 +/* + * LZ4_ACCELERATION_MAX : + * Any "acceleration" value higher than this threshold + * get treated as LZ4_ACCELERATION_MAX instead (fix #876) + */ +#define LZ4_ACCELERATION_MAX 65537 /*-************************************ @@ -82,6 +88,7 @@ * Define this parameter if your target system or compiler does not support hardware bit count */ #if defined(_MSC_VER) && defined(_WIN32_WCE) /* Visual Studio for WinCE doesn't support Hardware bit count */ +# undef LZ4_FORCE_SW_BITCOUNT /* avoid double def */ # define LZ4_FORCE_SW_BITCOUNT #endif @@ -114,10 +121,9 @@ /*-************************************ * Compiler Options **************************************/ -#ifdef _MSC_VER /* Visual Studio */ -# include -# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */ -# pragma warning(disable : 4293) /* disable: C4293: too large shift (32-bits) */ +#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 */ #endif /* _MSC_VER */ #ifndef LZ4_FORCE_INLINE @@ -136,7 +142,7 @@ # endif /* _MSC_VER */ #endif /* LZ4_FORCE_INLINE */ -/* LZ4_FORCE_O2_GCC_PPC64LE and LZ4_FORCE_O2_INLINE_GCC_PPC64LE +/* LZ4_FORCE_O2 and LZ4_FORCE_INLINE * gcc on ppc64le generates an unrolled SIMDized loop for LZ4_wildCopy8, * together with a simple 8-byte copy loop as a fall-back path. * However, this optimization hurts the decompression speed by >30%, @@ -151,11 +157,11 @@ * of LZ4_wildCopy8 does not affect the compression speed. */ #if defined(__PPC64__) && defined(__LITTLE_ENDIAN__) && defined(__GNUC__) && !defined(__clang__) -# define LZ4_FORCE_O2_GCC_PPC64LE __attribute__((optimize("O2"))) -# define LZ4_FORCE_O2_INLINE_GCC_PPC64LE __attribute__((optimize("O2"))) LZ4_FORCE_INLINE +# define LZ4_FORCE_O2 __attribute__((optimize("O2"))) +# undef LZ4_FORCE_INLINE +# define LZ4_FORCE_INLINE static __inline __attribute__((optimize("O2"),always_inline)) #else -# define LZ4_FORCE_O2_GCC_PPC64LE -# define LZ4_FORCE_O2_INLINE_GCC_PPC64LE static +# define LZ4_FORCE_O2 #endif #if (defined(__GNUC__) && (__GNUC__ >= 3)) || (defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 800)) || defined(__clang__) @@ -171,14 +177,33 @@ #define unlikely(expr) expect((expr) != 0, 0) #endif +/* Should the alignment test prove unreliable, for some reason, + * it can be disabled by setting LZ4_ALIGN_TEST to 0 */ +#ifndef LZ4_ALIGN_TEST /* can be externally provided */ +# define LZ4_ALIGN_TEST 1 +#endif + /*-************************************ * Memory routines **************************************/ -#include /* malloc, calloc, free */ -#define ALLOC(s) malloc(s) -#define ALLOC_AND_ZERO(s) calloc(1,s) -#define FREEMEM(p) free(p) +#ifdef 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 */ +void* LZ4_malloc(size_t s); +void* LZ4_calloc(size_t n, size_t s); +void LZ4_free(void* p); +# define ALLOC(s) LZ4_malloc(s) +# define ALLOC_AND_ZERO(s) LZ4_calloc(1,s) +# define FREEMEM(p) LZ4_free(p) +#else +# include /* malloc, calloc, free */ +# define ALLOC(s) malloc(s) +# define ALLOC_AND_ZERO(s) calloc(1,s) +# define FREEMEM(p) free(p) +#endif + #include /* memset, memcpy */ #define MEM_INIT(p,v,s) memset((p),(v),(s)) @@ -225,21 +250,27 @@ static const int LZ4_minLength = (MFLIMIT+1); #if defined(LZ4_DEBUG) && (LZ4_DEBUG>=2) # include -static int g_debuglog_enable = 1; -# define DEBUGLOG(l, ...) { \ - if ((g_debuglog_enable) && (l<=LZ4_DEBUG)) { \ - fprintf(stderr, __FILE__ ": "); \ - fprintf(stderr, __VA_ARGS__); \ - fprintf(stderr, " \n"); \ - } } + static int g_debuglog_enable = 1; +# define DEBUGLOG(l, ...) { \ + if ((g_debuglog_enable) && (l<=LZ4_DEBUG)) { \ + fprintf(stderr, __FILE__ ": "); \ + fprintf(stderr, __VA_ARGS__); \ + fprintf(stderr, " \n"); \ + } } #else -# define DEBUGLOG(l, ...) {} /* disabled */ +# define DEBUGLOG(l, ...) {} /* disabled */ #endif +static int LZ4_isAligned(const void* ptr, size_t alignment) +{ + return ((size_t)ptr & (alignment -1)) == 0; +} + /*-************************************ * Types **************************************/ +#include #if defined(__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) # include typedef uint8_t BYTE; @@ -249,6 +280,9 @@ static int g_debuglog_enable = 1; typedef uint64_t U64; typedef uintptr_t uptrval; #else +# if UINT_MAX != 4294967295UL +# error "LZ4 code (when not C++ or C99) assumes that sizeof(int) == 4" +# endif typedef unsigned char BYTE; typedef unsigned short U16; typedef unsigned int U32; @@ -273,6 +307,21 @@ typedef enum { /*-************************************ * Reading and writing into memory **************************************/ + +/** + * LZ4 relies on memcpy with a constant size being inlined. In freestanding + * environments, the compiler can't assume the implementation of memcpy() is + * standard compliant, so it can't apply its specialized memcpy() inlining + * logic. When possible, use __builtin_memcpy() to tell the compiler to analyze + * 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) +#endif + static unsigned LZ4_isLittleEndian(void) { const union { U32 u; BYTE c[4]; } one = { 1 }; /* don't use static : performance detrimental */ @@ -307,27 +356,27 @@ static void LZ4_write32(void* memPtr, U32 value) { ((unalign*)memPtr)->u32 = val static U16 LZ4_read16(const void* memPtr) { - U16 val; memcpy(&val, memPtr, sizeof(val)); return val; + U16 val; LZ4_memcpy(&val, memPtr, sizeof(val)); return val; } static U32 LZ4_read32(const void* memPtr) { - U32 val; memcpy(&val, memPtr, sizeof(val)); return val; + U32 val; LZ4_memcpy(&val, memPtr, sizeof(val)); return val; } static reg_t LZ4_read_ARCH(const void* memPtr) { - reg_t val; memcpy(&val, memPtr, sizeof(val)); return val; + reg_t val; LZ4_memcpy(&val, memPtr, sizeof(val)); return val; } static void LZ4_write16(void* memPtr, U16 value) { - memcpy(memPtr, &value, sizeof(value)); + LZ4_memcpy(memPtr, &value, sizeof(value)); } static void LZ4_write32(void* memPtr, U32 value) { - memcpy(memPtr, &value, sizeof(value)); + LZ4_memcpy(memPtr, &value, sizeof(value)); } #endif /* LZ4_FORCE_MEMORY_ACCESS */ @@ -355,14 +404,14 @@ static void LZ4_writeLE16(void* memPtr, U16 value) } /* customized variant of memcpy, which can overwrite up to 8 bytes beyond dstEnd */ -LZ4_FORCE_O2_INLINE_GCC_PPC64LE +LZ4_FORCE_INLINE void LZ4_wildCopy8(void* dstPtr, const void* srcPtr, void* dstEnd) { BYTE* d = (BYTE*)dstPtr; const BYTE* s = (const BYTE*)srcPtr; BYTE* const e = (BYTE*)dstEnd; - do { memcpy(d,s,8); d+=8; s+=8; } while (d= 16. */ -LZ4_FORCE_O2_INLINE_GCC_PPC64LE void +LZ4_FORCE_INLINE void LZ4_wildCopy32(void* dstPtr, const void* srcPtr, void* dstEnd) { BYTE* d = (BYTE*)dstPtr; const BYTE* s = (const BYTE*)srcPtr; BYTE* const e = (BYTE*)dstEnd; - do { memcpy(d,s,16); memcpy(d+16,s+16,16); d+=32; s+=32; } while (d= dstPtr + MINMATCH * - there is at least 8 bytes available to write after dstEnd */ -LZ4_FORCE_O2_INLINE_GCC_PPC64LE void +LZ4_FORCE_INLINE void LZ4_memcpy_using_offset(BYTE* dstPtr, const BYTE* srcPtr, BYTE* dstEnd, const size_t offset) { BYTE v[8]; assert(dstEnd >= dstPtr + MINMATCH); - LZ4_write32(dstPtr, 0); /* silence an msan warning when offset==0 */ switch(offset) { case 1: - memset(v, *srcPtr, 8); + MEM_INIT(v, *srcPtr, 8); break; case 2: - memcpy(v, srcPtr, 2); - memcpy(&v[2], srcPtr, 2); - memcpy(&v[4], &v[0], 4); + LZ4_memcpy(v, srcPtr, 2); + LZ4_memcpy(&v[2], srcPtr, 2); + LZ4_memcpy(&v[4], v, 4); break; case 4: - memcpy(v, srcPtr, 4); - memcpy(&v[4], srcPtr, 4); + LZ4_memcpy(v, srcPtr, 4); + LZ4_memcpy(&v[4], srcPtr, 4); break; default: LZ4_memcpy_using_offset_base(dstPtr, srcPtr, dstEnd, offset); return; } - memcpy(dstPtr, v, 8); + LZ4_memcpy(dstPtr, v, 8); dstPtr += 8; while (dstPtr < dstEnd) { - memcpy(dstPtr, v, 8); + LZ4_memcpy(dstPtr, v, 8); dstPtr += 8; } } @@ -462,75 +512,92 @@ LZ4_memcpy_using_offset(BYTE* dstPtr, const BYTE* srcPtr, BYTE* dstEnd, const si **************************************/ static unsigned LZ4_NbCommonBytes (reg_t val) { + assert(val != 0); if (LZ4_isLittleEndian()) { - if (sizeof(val)==8) { -# if defined(_MSC_VER) && defined(_WIN64) && !defined(LZ4_FORCE_SW_BITCOUNT) + if (sizeof(val) == 8) { +# if defined(_MSC_VER) && (_MSC_VER >= 1800) && defined(_M_AMD64) && !defined(LZ4_FORCE_SW_BITCOUNT) + /* x64 CPUS without BMI support interpret `TZCNT` as `REP BSF` */ + return (unsigned)_tzcnt_u64(val) >> 3; +# elif defined(_MSC_VER) && defined(_WIN64) && !defined(LZ4_FORCE_SW_BITCOUNT) unsigned long r = 0; - _BitScanForward64( &r, (U64)val ); - return (int)(r>>3); -# elif (defined(__clang__) || (defined(__GNUC__) && (__GNUC__>=3))) && !defined(LZ4_FORCE_SW_BITCOUNT) + _BitScanForward64(&r, (U64)val); + return (unsigned)r >> 3; +# elif (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(LZ4_FORCE_SW_BITCOUNT) return (unsigned)__builtin_ctzll((U64)val) >> 3; # else - static const int DeBruijnBytePos[64] = { 0, 0, 0, 0, 0, 1, 1, 2, - 0, 3, 1, 3, 1, 4, 2, 7, - 0, 2, 3, 6, 1, 5, 3, 5, - 1, 3, 4, 4, 2, 5, 6, 7, - 7, 0, 1, 2, 3, 3, 4, 6, - 2, 6, 5, 5, 3, 4, 5, 6, - 7, 1, 2, 4, 6, 4, 4, 5, - 7, 2, 6, 5, 7, 6, 7, 7 }; - return DeBruijnBytePos[((U64)((val & -(long long)val) * 0x0218A392CDABBD3FULL)) >> 58]; + const U64 m = 0x0101010101010101ULL; + val ^= val - 1; + return (unsigned)(((U64)((val & (m - 1)) * m)) >> 56); # endif } else /* 32 bits */ { -# if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT) +# if defined(_MSC_VER) && (_MSC_VER >= 1400) && !defined(LZ4_FORCE_SW_BITCOUNT) unsigned long r; - _BitScanForward( &r, (U32)val ); - return (int)(r>>3); -# elif (defined(__clang__) || (defined(__GNUC__) && (__GNUC__>=3))) && !defined(LZ4_FORCE_SW_BITCOUNT) + _BitScanForward(&r, (U32)val); + return (unsigned)r >> 3; +# elif (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(__TINYC__) && !defined(LZ4_FORCE_SW_BITCOUNT) return (unsigned)__builtin_ctz((U32)val) >> 3; # else - static const int DeBruijnBytePos[32] = { 0, 0, 3, 0, 3, 1, 3, 0, - 3, 2, 2, 1, 3, 2, 0, 1, - 3, 3, 1, 2, 2, 2, 2, 0, - 3, 1, 2, 0, 1, 0, 1, 1 }; - return DeBruijnBytePos[((U32)((val & -(S32)val) * 0x077CB531U)) >> 27]; + const U32 m = 0x01010101; + return (unsigned)((((val - 1) ^ val) & (m - 1)) * m) >> 24; # endif } } else /* Big Endian CPU */ { - if (sizeof(val)==8) { /* 64-bits */ -# if defined(_MSC_VER) && defined(_WIN64) && !defined(LZ4_FORCE_SW_BITCOUNT) - unsigned long r = 0; - _BitScanReverse64( &r, val ); - return (unsigned)(r>>3); -# elif (defined(__clang__) || (defined(__GNUC__) && (__GNUC__>=3))) && !defined(LZ4_FORCE_SW_BITCOUNT) + if (sizeof(val)==8) { +# if (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(__TINYC__) && !defined(LZ4_FORCE_SW_BITCOUNT) return (unsigned)__builtin_clzll((U64)val) >> 3; # else +#if 1 + /* this method is probably faster, + * but adds a 128 bytes lookup table */ + static const unsigned char ctz7_tab[128] = { + 7, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 5, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 6, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 5, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + 4, 0, 1, 0, 2, 0, 1, 0, 3, 0, 1, 0, 2, 0, 1, 0, + }; + U64 const mask = 0x0101010101010101ULL; + U64 const t = (((val >> 8) - mask) | val) & mask; + return ctz7_tab[(t * 0x0080402010080402ULL) >> 57]; +#else + /* this method doesn't consume memory space like the previous one, + * but it contains several branches, + * that may end up slowing execution */ static const U32 by32 = sizeof(val)*4; /* 32 on 64 bits (goal), 16 on 32 bits. - Just to avoid some static analyzer complaining about shift by 32 on 32-bits target. - Note that this code path is never triggered in 32-bits mode. */ + Just to avoid some static analyzer complaining about shift by 32 on 32-bits target. + Note that this code path is never triggered in 32-bits mode. */ unsigned r; if (!(val>>by32)) { r=4; } else { r=0; val>>=by32; } if (!(val>>16)) { r+=2; val>>=8; } else { val>>=24; } r += (!val); return r; +#endif # endif } else /* 32 bits */ { -# if defined(_MSC_VER) && !defined(LZ4_FORCE_SW_BITCOUNT) - unsigned long r = 0; - _BitScanReverse( &r, (unsigned long)val ); - return (unsigned)(r>>3); -# elif (defined(__clang__) || (defined(__GNUC__) && (__GNUC__>=3))) && !defined(LZ4_FORCE_SW_BITCOUNT) +# if (defined(__clang__) || (defined(__GNUC__) && ((__GNUC__ > 3) || \ + ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 4))))) && \ + !defined(LZ4_FORCE_SW_BITCOUNT) return (unsigned)__builtin_clz((U32)val) >> 3; # else - unsigned r; - if (!(val>>16)) { r=2; val>>=8; } else { r=0; val>>=24; } - r += (!val); - return r; + val >>= 8; + val = ((((val + 0x00FFFF00) | 0x00FFFFFF) + val) | + (val + 0x00FF0000)) >> 24; + return (unsigned)val ^ 3; # endif } } } + #define STEPSIZE sizeof(reg_t) LZ4_FORCE_INLINE unsigned LZ4_count(const BYTE* pIn, const BYTE* pMatch, const BYTE* pInLimit) @@ -605,7 +672,7 @@ typedef enum { noDictIssue = 0, dictSmall } dictIssue_directive; 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() { return LZ4_STREAMSIZE; } +int LZ4_sizeofState(void) { return LZ4_STREAMSIZE; } /*-************************************ @@ -628,7 +695,7 @@ int LZ4_decompress_safe_forceExtDict(const char* source, char* dest, /*-****************************** * Compression functions ********************************/ -static U32 LZ4_hash4(U32 sequence, tableType_t const tableType) +LZ4_FORCE_INLINE U32 LZ4_hash4(U32 sequence, tableType_t const tableType) { if (tableType == byU16) return ((sequence * 2654435761U) >> ((MINMATCH*8)-(LZ4_HASHLOG+1))); @@ -636,7 +703,7 @@ static U32 LZ4_hash4(U32 sequence, tableType_t const tableType) return ((sequence * 2654435761U) >> ((MINMATCH*8)-LZ4_HASHLOG)); } -static U32 LZ4_hash5(U64 sequence, tableType_t const tableType) +LZ4_FORCE_INLINE U32 LZ4_hash5(U64 sequence, tableType_t const tableType) { const U32 hashLog = (tableType == byU16) ? LZ4_HASHLOG+1 : LZ4_HASHLOG; if (LZ4_isLittleEndian()) { @@ -654,7 +721,7 @@ LZ4_FORCE_INLINE U32 LZ4_hashPosition(const void* const p, tableType_t const tab return LZ4_hash4(LZ4_read32(p), tableType); } -static void LZ4_clearHash(U32 h, void* tableBase, tableType_t const tableType) +LZ4_FORCE_INLINE void LZ4_clearHash(U32 h, void* tableBase, tableType_t const tableType) { switch (tableType) { @@ -666,7 +733,7 @@ static void LZ4_clearHash(U32 h, void* tableBase, tableType_t const tableType) } } -static void LZ4_putIndexOnHash(U32 idx, U32 h, void* tableBase, tableType_t const tableType) +LZ4_FORCE_INLINE void LZ4_putIndexOnHash(U32 idx, U32 h, void* tableBase, tableType_t const tableType) { switch (tableType) { @@ -678,7 +745,7 @@ static void LZ4_putIndexOnHash(U32 idx, U32 h, void* tableBase, tableType_t cons } } -static void LZ4_putPositionOnHash(const BYTE* p, U32 h, +LZ4_FORCE_INLINE void LZ4_putPositionOnHash(const BYTE* p, U32 h, void* tableBase, tableType_t const tableType, const BYTE* srcBase) { @@ -703,7 +770,7 @@ LZ4_FORCE_INLINE void LZ4_putPosition(const BYTE* p, void* tableBase, tableType_ * Assumption 1 : only valid if tableType == byU32 or byU16. * Assumption 2 : h is presumed valid (within limits of hash table) */ -static U32 LZ4_getIndexOnHash(U32 h, const void* tableBase, tableType_t tableType) +LZ4_FORCE_INLINE U32 LZ4_getIndexOnHash(U32 h, const void* tableBase, tableType_t tableType) { LZ4_STATIC_ASSERT(LZ4_MEMORY_USAGE > 2); if (tableType == byU32) { @@ -739,22 +806,13 @@ LZ4_FORCE_INLINE void LZ4_prepareTable(LZ4_stream_t_internal* const cctx, const int inputSize, const tableType_t tableType) { - /* If compression failed during the previous step, then the context - * is marked as dirty, therefore, it has to be fully reset. - */ - if (cctx->dirty) { - DEBUGLOG(5, "LZ4_prepareTable: Full reset for %p", cctx); - MEM_INIT(cctx, 0, sizeof(LZ4_stream_t_internal)); - return; - } - /* If the table hasn't been used, it's guaranteed to be zeroed out, and is * therefore safe to use no matter what mode we're in. Otherwise, we figure * out if it's safe to leave as is or whether it needs to be reset. */ - if (cctx->tableType != clearedTable) { + if ((tableType_t)cctx->tableType != clearedTable) { assert(inputSize >= 0); - if (cctx->tableType != tableType + if ((tableType_t)cctx->tableType != tableType || ((tableType == byU16) && cctx->currentOffset + (unsigned)inputSize >= 0xFFFFU) || ((tableType == byU32) && cctx->currentOffset > 1 GB) || tableType == byPtr @@ -763,7 +821,7 @@ LZ4_prepareTable(LZ4_stream_t_internal* const cctx, DEBUGLOG(4, "LZ4_prepareTable: Resetting table in %p", cctx); MEM_INIT(cctx->hashTable, 0, LZ4_HASHTABLESIZE); cctx->currentOffset = 0; - cctx->tableType = clearedTable; + cctx->tableType = (U32)clearedTable; } else { DEBUGLOG(4, "LZ4_prepareTable: Re-use hash table (no reset)"); } @@ -785,8 +843,12 @@ LZ4_prepareTable(LZ4_stream_t_internal* const cctx, } /** LZ4_compress_generic() : - inlined, to ensure branches are decided at compilation time */ -LZ4_FORCE_INLINE int LZ4_compress_generic( + * inlined, to ensure branches are decided at compilation time. + * Presumed already validated at this stage: + * - source != NULL + * - inputSize > 0 + */ +LZ4_FORCE_INLINE int LZ4_compress_generic_validated( LZ4_stream_t_internal* const cctx, const char* const source, char* const dest, @@ -815,7 +877,7 @@ LZ4_FORCE_INLINE int LZ4_compress_generic( int const maybe_extMem = (dictDirective == usingExtDict) || (dictDirective == usingDictCtx); U32 const prefixIdxLimit = startIndex - dictSize; /* used when dictDirective == dictSmall */ - const BYTE* const dictEnd = dictionary + dictSize; + const BYTE* const dictEnd = dictionary ? dictionary + dictSize : dictionary; const BYTE* anchor = (const BYTE*) source; const BYTE* const iend = ip + inputSize; const BYTE* const mflimitPlusOne = iend - MFLIMIT + 1; @@ -823,7 +885,7 @@ LZ4_FORCE_INLINE int LZ4_compress_generic( /* the dictCtx currentOffset is indexed on the start of the dictionary, * while a dictionary in the current context precedes the currentOffset */ - const BYTE* dictBase = (dictDirective == usingDictCtx) ? + const BYTE* dictBase = !dictionary ? NULL : (dictDirective == usingDictCtx) ? dictionary + dictSize - dictCtx->currentOffset : dictionary + dictSize - startIndex; @@ -833,11 +895,11 @@ LZ4_FORCE_INLINE int LZ4_compress_generic( U32 offset = 0; U32 forwardH; - DEBUGLOG(5, "LZ4_compress_generic: srcSize=%i, tableType=%u", inputSize, tableType); + DEBUGLOG(5, "LZ4_compress_generic_validated: srcSize=%i, tableType=%u", inputSize, tableType); + assert(ip != NULL); /* If init conditions are not met, we don't have to mark stream * as having dirty context, since no action was taken yet */ if (outputDirective == fillOutput && maxOutputSize < 1) { return 0; } /* Impossible to store anything */ - if ((U32)inputSize > (U32)LZ4_MAX_INPUT_SIZE) { return 0; } /* Unsupported inputSize, too large (or negative) */ if ((tableType == byU16) && (inputSize>=LZ4_64Klimit)) { return 0; } /* Size too large (not within 64K limit) */ if (tableType==byPtr) assert(dictDirective==noDict); /* only supported use case with byPtr */ assert(acceleration >= 1); @@ -854,7 +916,7 @@ LZ4_FORCE_INLINE int LZ4_compress_generic( cctx->dictSize += (U32)inputSize; } cctx->currentOffset += (U32)inputSize; - cctx->tableType = (U16)tableType; + cctx->tableType = (U32)tableType; if (inputSize= op); - lastRun = (size_t)(olimit-op) - 1; - lastRun -= (lastRun+240)/255; + lastRun = (size_t)(olimit-op) - 1/*token*/; + lastRun -= (lastRun + 256 - RUN_MASK) / 256; /*additional length tokens*/ } else { assert(outputDirective == limitedOutput); return 0; /* cannot compress within `dst` budget. Stored indexes in hash table are nonetheless fine */ } } + DEBUGLOG(6, "Final literal run : %i literals", (int)lastRun); if (lastRun >= RUN_MASK) { size_t accumulator = lastRun - RUN_MASK; *op++ = RUN_MASK << ML_BITS; @@ -1162,7 +1225,7 @@ LZ4_FORCE_INLINE int LZ4_compress_generic( } else { *op++ = (BYTE)(lastRun< 0); + DEBUGLOG(5, "LZ4_compress_generic: compressed %i bytes into %i bytes", inputSize, result); return result; } +/** LZ4_compress_generic() : + * inlined, to ensure branches are decided at compilation time; + * takes care of src == (NULL, 0) + * and forward the rest to LZ4_compress_generic_validated */ +LZ4_FORCE_INLINE int LZ4_compress_generic( + LZ4_stream_t_internal* const cctx, + const char* const src, + char* const dst, + const int srcSize, + int *inputConsumed, /* only written when outputDirective == fillOutput */ + const int dstCapacity, + const limitedOutput_directive outputDirective, + const tableType_t tableType, + const dict_directive dictDirective, + const dictIssue_directive dictIssue, + const int acceleration) +{ + DEBUGLOG(5, "LZ4_compress_generic: srcSize=%i, dstCapacity=%i", + srcSize, dstCapacity); + + if ((U32)srcSize > (U32)LZ4_MAX_INPUT_SIZE) { return 0; } /* Unsupported srcSize, too large (or negative) */ + if (srcSize == 0) { /* src == NULL supported if srcSize == 0 */ + if (outputDirective != notLimited && dstCapacity <= 0) return 0; /* no output, can't write anything */ + DEBUGLOG(5, "Generating an empty block"); + assert(outputDirective == notLimited || dstCapacity >= 1); + assert(dst != NULL); + dst[0] = 0; + if (outputDirective == fillOutput) { + assert (inputConsumed != NULL); + *inputConsumed = 0; + } + return 1; + } + assert(src != NULL); + + return LZ4_compress_generic_validated(cctx, src, dst, srcSize, + inputConsumed, /* only written into if outputDirective == fillOutput */ + dstCapacity, outputDirective, + tableType, dictDirective, dictIssue, acceleration); +} + int LZ4_compress_fast_extState(void* state, const char* source, char* dest, int inputSize, int maxOutputSize, int acceleration) { LZ4_stream_t_internal* const ctx = & LZ4_initStream(state, sizeof(LZ4_stream_t)) -> internal_donotuse; assert(ctx != NULL); - if (acceleration < 1) acceleration = ACCELERATION_DEFAULT; + if (acceleration < 1) acceleration = LZ4_ACCELERATION_DEFAULT; + if (acceleration > LZ4_ACCELERATION_MAX) acceleration = LZ4_ACCELERATION_MAX; if (maxOutputSize >= LZ4_compressBound(inputSize)) { if (inputSize < LZ4_64Klimit) { return LZ4_compress_generic(ctx, source, dest, inputSize, NULL, 0, notLimited, byU16, noDict, noDictIssue, acceleration); @@ -1211,7 +1316,8 @@ int LZ4_compress_fast_extState(void* state, const char* source, char* dest, int int LZ4_compress_fast_extState_fastReset(void* state, const char* src, char* dst, int srcSize, int dstCapacity, int acceleration) { LZ4_stream_t_internal* ctx = &((LZ4_stream_t*)state)->internal_donotuse; - if (acceleration < 1) acceleration = ACCELERATION_DEFAULT; + if (acceleration < 1) acceleration = LZ4_ACCELERATION_DEFAULT; + if (acceleration > LZ4_ACCELERATION_MAX) acceleration = LZ4_ACCELERATION_MAX; if (dstCapacity >= LZ4_compressBound(srcSize)) { if (srcSize < LZ4_64Klimit) { @@ -1270,22 +1376,6 @@ int LZ4_compress_default(const char* src, char* dst, int srcSize, int maxOutputS } -/* hidden debug function */ -/* strangely enough, gcc generates faster code when this function is uncommented, even if unused */ -int LZ4_compress_fast_force(const char* src, char* dst, int srcSize, int dstCapacity, int acceleration) -{ - LZ4_stream_t ctx; - LZ4_initStream(&ctx, sizeof(ctx)); - - if (srcSize < LZ4_64Klimit) { - return LZ4_compress_generic(&ctx.internal_donotuse, src, dst, srcSize, NULL, dstCapacity, limitedOutput, byU16, noDict, noDictIssue, acceleration); - } else { - tableType_t const addrMode = (sizeof(void*) > 4) ? byU32 : byPtr; - return LZ4_compress_generic(&ctx.internal_donotuse, src, dst, srcSize, NULL, dstCapacity, limitedOutput, addrMode, noDict, noDictIssue, acceleration); - } -} - - /* Note!: This function leaves the stream in an unclean/broken state! * It is not safe to subsequently use the same state with a _fastReset() or * _continue() call without resetting it. */ @@ -1340,27 +1430,23 @@ LZ4_stream_t* LZ4_createStream(void) return lz4s; } -#ifndef _MSC_VER /* for some reason, Visual fails the aligment test on 32-bit x86 : - it reports an aligment of 8-bytes, - while actually aligning LZ4_stream_t on 4 bytes. */ static size_t LZ4_stream_t_alignment(void) { - struct { char c; LZ4_stream_t t; } t_a; - return sizeof(t_a) - sizeof(t_a.t); -} +#if LZ4_ALIGN_TEST + typedef struct { char c; LZ4_stream_t t; } t_a; + return sizeof(t_a) - sizeof(LZ4_stream_t); +#else + return 1; /* effectively disabled */ #endif +} LZ4_stream_t* LZ4_initStream (void* buffer, size_t size) { DEBUGLOG(5, "LZ4_initStream"); if (buffer == NULL) { return NULL; } if (size < sizeof(LZ4_stream_t)) { return NULL; } -#ifndef _MSC_VER /* for some reason, Visual fails the aligment test on 32-bit x86 : - it reports an aligment of 8-bytes, - while actually aligning LZ4_stream_t on 4 bytes. */ - if (((size_t)buffer) & (LZ4_stream_t_alignment() - 1)) { return NULL; } /* alignment check */ -#endif - MEM_INIT(buffer, 0, sizeof(LZ4_stream_t)); + if (!LZ4_isAligned(buffer, LZ4_stream_t_alignment())) return NULL; + MEM_INIT(buffer, 0, sizeof(LZ4_stream_t_internal)); return (LZ4_stream_t*)buffer; } @@ -1369,7 +1455,7 @@ LZ4_stream_t* LZ4_initStream (void* buffer, size_t size) void LZ4_resetStream (LZ4_stream_t* LZ4_stream) { DEBUGLOG(5, "LZ4_resetStream (ctx:%p)", LZ4_stream); - MEM_INIT(LZ4_stream, 0, sizeof(LZ4_stream_t)); + MEM_INIT(LZ4_stream, 0, sizeof(LZ4_stream_t_internal)); } void LZ4_resetStream_fast(LZ4_stream_t* ctx) { @@ -1418,7 +1504,7 @@ int LZ4_loadDict (LZ4_stream_t* LZ4_dict, const char* dictionary, int dictSize) base = dictEnd - dict->currentOffset; dict->dictionary = p; dict->dictSize = (U32)(dictEnd - p); - dict->tableType = tableType; + dict->tableType = (U32)tableType; while (p <= dictEnd-HASH_UNIT) { LZ4_putPosition(p, dict->hashTable, tableType, base); @@ -1436,12 +1522,6 @@ void LZ4_attach_dictionary(LZ4_stream_t* workingStream, const LZ4_stream_t* dict workingStream, dictionaryStream, dictCtx != NULL ? dictCtx->dictSize : 0); - /* Calling LZ4_resetStream_fast() here makes sure that changes will not be - * erased by subsequent calls to LZ4_resetStream_fast() in case stream was - * marked as having dirty context, e.g. requiring full reset. - */ - LZ4_resetStream_fast(workingStream); - if (dictCtx != NULL) { /* If the current offset is zero, we will never look in the * external dictionary context, since there is no value a table @@ -1493,9 +1573,9 @@ int LZ4_compress_fast_continue (LZ4_stream_t* LZ4_stream, DEBUGLOG(5, "LZ4_compress_fast_continue (inputSize=%i)", inputSize); - if (streamPtr->dirty) { return 0; } /* Uninitialized structure detected */ LZ4_renormDictT(streamPtr, inputSize); /* avoid index overflow */ - if (acceleration < 1) acceleration = ACCELERATION_DEFAULT; + 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 */ @@ -1538,7 +1618,7 @@ int LZ4_compress_fast_continue (LZ4_stream_t* LZ4_stream, * cost to copy the dictionary's tables into the active context, * so that the compression loop is only looking into one table. */ - memcpy(streamPtr, streamPtr->dictCtx, sizeof(LZ4_stream_t)); + LZ4_memcpy(streamPtr, streamPtr->dictCtx, sizeof(*streamPtr)); result = LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, usingExtDict, noDictIssue, acceleration); } else { result = LZ4_compress_generic(streamPtr, source, dest, inputSize, NULL, maxOutputSize, limitedOutput, tableType, usingDictCtx, noDictIssue, acceleration); @@ -1593,7 +1673,9 @@ int LZ4_saveDict (LZ4_stream_t* LZ4_dict, char* safeBuffer, int dictSize) if ((U32)dictSize > 64 KB) { dictSize = 64 KB; } /* useless to define a dictionary > 64 KB */ if ((U32)dictSize > dict->dictSize) { dictSize = (int)dict->dictSize; } - memmove(safeBuffer, previousDictEnd - dictSize, dictSize); + if (safeBuffer == NULL) assert(dictSize == 0); + if (dictSize > 0) + memmove(safeBuffer, previousDictEnd - dictSize, dictSize); dict->dictionary = (const BYTE*)safeBuffer; dict->dictSize = (U32)dictSize; @@ -1623,25 +1705,27 @@ typedef enum { decode_full_block = 0, partial_decode = 1 } earlyEnd_directive; */ 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) -{ - unsigned length = 0; - unsigned s; - if (initial_check && unlikely((*ip) >= lencheck)) { /* overflow detection */ - *error = initial_error; - return length; - } - do { - s = **ip; - (*ip)++; - length += s; - if (loop_check && unlikely((*ip) >= lencheck)) { /* overflow detection */ - *error = loop_error; - return length; +read_variable_length(const BYTE**ip, const BYTE* lencheck, + int loop_check, int initial_check, + variable_length_error* error) +{ + U32 length = 0; + U32 s; + if (initial_check && unlikely((*ip) >= lencheck)) { /* overflow detection */ + *error = initial_error; + return length; } - } while (s==255); + do { + s = **ip; + (*ip)++; + length += s; + if (loop_check && unlikely((*ip) >= lencheck)) { /* overflow detection */ + *error = loop_error; + return length; + } + } while (s==255); - return length; + return length; } /*! LZ4_decompress_generic() : @@ -1722,7 +1806,7 @@ LZ4_decompress_generic( /* decode literal length */ if (length == RUN_MASK) { variable_length_error error = ok; - length += read_variable_length(&ip, iend-RUN_MASK, endOnInput, endOnInput, &error); + 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 */ @@ -1746,12 +1830,12 @@ LZ4_decompress_generic( /* 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 */ - memcpy(op, ip, 16); + 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 */ - memcpy(op, ip, 8); - if (length > 8) { memcpy(op+8, ip+8, 8); } + LZ4_memcpy(op, ip, 8); + if (length > 8) { LZ4_memcpy(op+8, ip+8, 8); } } ip += length; op = cpy; } @@ -1765,10 +1849,10 @@ LZ4_decompress_generic( 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, endOnInput, 0, &error); - if (error != ok) { goto _output_error; } + 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 */ length += MINMATCH; if (op + length >= oend - FASTLOOP_SAFE_DISTANCE) { @@ -1787,19 +1871,20 @@ LZ4_decompress_generic( assert(match <= op); assert(op + 18 <= oend); - memcpy(op, match, 8); - memcpy(op+8, match+8, 8); - memcpy(op+16, match+16, 2); + LZ4_memcpy(op, match, 8); + LZ4_memcpy(op+8, match+8, 8); + LZ4_memcpy(op+16, match+16, 2); op += length; continue; } } } - if ((checkOffset) && (unlikely(match + dictSize < lowPrefix))) { goto _output_error; } /* Error : offset outside buffers */ + if (checkOffset && (unlikely(match + dictSize < lowPrefix))) { goto _output_error; } /* Error : offset outside buffers */ /* match starting within external dictionary */ if ((dict==usingExtDict) && (match < lowPrefix)) { if (unlikely(op+length > oend-LASTLITERALS)) { if (partialDecoding) { - length = MIN(length, (size_t)(oend-op)); /* reach end of buffer */ + DEBUGLOG(7, "partialDecoding: dictionary match, close to dstEnd"); + length = MIN(length, (size_t)(oend-op)); } else { goto _output_error; /* end-of-block condition violated */ } } @@ -1812,14 +1897,14 @@ LZ4_decompress_generic( /* match stretches into both external dictionary and current block */ size_t const copySize = (size_t)(lowPrefix - match); size_t const restSize = length - copySize; - memcpy(op, dictEnd - copySize, copySize); + LZ4_memcpy(op, dictEnd - copySize, copySize); op += copySize; if (restSize > (size_t)(op - lowPrefix)) { /* overlap copy */ BYTE* const endOfMatch = op + restSize; const BYTE* copyFrom = lowPrefix; while (op < endOfMatch) { *op++ = *copyFrom++; } } else { - memcpy(op, lowPrefix, restSize); + LZ4_memcpy(op, lowPrefix, restSize); op += restSize; } } continue; @@ -1860,7 +1945,7 @@ LZ4_decompress_generic( /* strictly "less than" on input, to re-enter the loop with at least one byte */ && likely((endOnInput ? ip < shortiend : 1) & (op <= shortoend)) ) { /* Copy the literals */ - memcpy(op, ip, endOnInput ? 16 : 8); + LZ4_memcpy(op, ip, endOnInput ? 16 : 8); op += length; ip += length; /* The second stage: prepare for match copying, decode full info. @@ -1875,9 +1960,9 @@ LZ4_decompress_generic( && (offset >= 8) && (dict==withPrefix64k || match >= lowPrefix) ) { /* Copy the match. */ - memcpy(op + 0, match + 0, 8); - memcpy(op + 8, match + 8, 8); - memcpy(op +16, match +16, 2); + LZ4_memcpy(op + 0, match + 0, 8); + LZ4_memcpy(op + 8, match + 8, 8); + LZ4_memcpy(op +16, match +16, 2); op += length + MINMATCH; /* Both stages worked, load the next token. */ continue; @@ -1891,7 +1976,7 @@ LZ4_decompress_generic( /* decode literal length */ if (length == RUN_MASK) { variable_length_error error = ok; - length += read_variable_length(&ip, iend-RUN_MASK, endOnInput, endOnInput, &error); + 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 */ @@ -1907,29 +1992,34 @@ LZ4_decompress_generic( || ((!endOnInput) && (cpy>oend-WILDCOPYLENGTH)) ) { /* We've either hit the input parsing restriction or the output parsing restriction. - * If we've hit the input parsing condition then this must be the last sequence. - * If we've hit the output parsing condition then we are either using partialDecoding - * or we've hit the output parsing condition. + * In the normal scenario, decoding a full block, it must be the last sequence, + * otherwise it's an error (invalid input or dimensions). + * In partialDecoding scenario, it's necessary to ensure there is no buffer overflow. */ if (partialDecoding) { /* 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); - /* If we're in this block because of the input parsing condition, then we must be on the - * last sequence (or invalid), so we must check that we exactly consume the input. + 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)); + DEBUGLOG(7, "partialDecoding: remaining space in srcBuffer : %i", (int)(iend - ip)); + /* Finishing in the middle of a literals segment, + * due to lack of input. */ - if ((ip+length>iend-(2+1+LASTLITERALS)) && (ip+length != iend)) { goto _output_error; } - assert(ip+length <= iend); - /* We are finishing in the middle of a literals segment. - * Break after the copy. + if (ip+length > iend) { + length = (size_t)(iend-ip); + cpy = op + length; + } + /* Finishing in the middle of a literals segment, + * due to lack of output space. */ if (cpy > oend) { cpy = oend; assert(op<=oend); length = (size_t)(oend-op); } - assert(ip+length <= iend); } 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). @@ -1938,16 +2028,22 @@ LZ4_decompress_generic( /* 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))) { goto _output_error; } + if ((endOnInput) && ((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, which only matters for in-place decompression scenarios */ + memmove(op, ip, length); /* supports overlapping memory regions; only matters for in-place decompression scenarios */ ip += length; op += length; - /* Necessarily EOF when !partialDecoding. When partialDecoding - * it is EOF if we've either filled the output buffer or hit - * the input parsing restriction. + /* Necessarily EOF when !partialDecoding. + * When partialDecoding, it is EOF if we've either + * filled the output buffer or + * can't proceed with reading an offset for following match. */ - if (!partialDecoding || (cpy == oend) || (ip == iend)) { + if (!partialDecoding || (cpy == oend) || (ip >= (iend-2))) { break; } } else { @@ -1965,7 +2061,7 @@ LZ4_decompress_generic( _copy_match: if (length == ML_MASK) { variable_length_error error = ok; - length += read_variable_length(&ip, iend - LASTLITERALS + 1, endOnInput, 0, &error); + 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 */ } @@ -1990,14 +2086,14 @@ LZ4_decompress_generic( /* match stretches into both external dictionary and current block */ size_t const copySize = (size_t)(lowPrefix - match); size_t const restSize = length - copySize; - memcpy(op, dictEnd - copySize, copySize); + LZ4_memcpy(op, dictEnd - copySize, copySize); op += copySize; if (restSize > (size_t)(op - lowPrefix)) { /* overlap copy */ BYTE* const endOfMatch = op + restSize; const BYTE* copyFrom = lowPrefix; while (op < endOfMatch) *op++ = *copyFrom++; } else { - memcpy(op, lowPrefix, restSize); + LZ4_memcpy(op, lowPrefix, restSize); op += restSize; } } continue; @@ -2016,7 +2112,7 @@ LZ4_decompress_generic( if (matchEnd > op) { /* overlap copy */ while (op < copyEnd) { *op++ = *match++; } } else { - memcpy(op, match, mlen); + LZ4_memcpy(op, match, mlen); } op = copyEnd; if (op == oend) { break; } @@ -2030,10 +2126,10 @@ LZ4_decompress_generic( op[2] = match[2]; op[3] = match[3]; match += inc32table[offset]; - memcpy(op+4, match, 4); + LZ4_memcpy(op+4, match, 4); match -= dec64table[offset]; } else { - memcpy(op, match, 8); + LZ4_memcpy(op, match, 8); match += 8; } op += 8; @@ -2048,7 +2144,7 @@ LZ4_decompress_generic( } while (op < cpy) { *op++ = *match++; } } else { - memcpy(op, match, 8); + LZ4_memcpy(op, match, 8); if (length > 16) { LZ4_wildCopy8(op+8, match+8, cpy); } } op = cpy; /* wildcopy correction */ @@ -2056,6 +2152,7 @@ LZ4_decompress_generic( /* 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 */ @@ -2070,7 +2167,7 @@ LZ4_decompress_generic( /*===== Instantiate the API decoding functions. =====*/ -LZ4_FORCE_O2_GCC_PPC64LE +LZ4_FORCE_O2 int LZ4_decompress_safe(const char* source, char* dest, int compressedSize, int maxDecompressedSize) { return LZ4_decompress_generic(source, dest, compressedSize, maxDecompressedSize, @@ -2078,7 +2175,7 @@ int LZ4_decompress_safe(const char* source, char* dest, int compressedSize, int (BYTE*)dest, NULL, 0); } -LZ4_FORCE_O2_GCC_PPC64LE +LZ4_FORCE_O2 int LZ4_decompress_safe_partial(const char* src, char* dst, int compressedSize, int targetOutputSize, int dstCapacity) { dstCapacity = MIN(targetOutputSize, dstCapacity); @@ -2087,7 +2184,7 @@ int LZ4_decompress_safe_partial(const char* src, char* dst, int compressedSize, noDict, (BYTE*)dst, NULL, 0); } -LZ4_FORCE_O2_GCC_PPC64LE +LZ4_FORCE_O2 int LZ4_decompress_fast(const char* source, char* dest, int originalSize) { return LZ4_decompress_generic(source, dest, 0, originalSize, @@ -2097,7 +2194,7 @@ int LZ4_decompress_fast(const char* source, char* dest, int originalSize) /*===== Instantiate a few more decoding cases, used more than once. =====*/ -LZ4_FORCE_O2_GCC_PPC64LE /* Exported, an obsolete API function. */ +LZ4_FORCE_O2 /* Exported, an obsolete API function. */ int LZ4_decompress_safe_withPrefix64k(const char* source, char* dest, int compressedSize, int maxOutputSize) { return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, @@ -2113,7 +2210,7 @@ int LZ4_decompress_fast_withPrefix64k(const char* source, char* dest, int origin return LZ4_decompress_fast(source, dest, originalSize); } -LZ4_FORCE_O2_GCC_PPC64LE +LZ4_FORCE_O2 static int LZ4_decompress_safe_withSmallPrefix(const char* source, char* dest, int compressedSize, int maxOutputSize, size_t prefixSize) { @@ -2122,7 +2219,7 @@ static int LZ4_decompress_safe_withSmallPrefix(const char* source, char* dest, i (BYTE*)dest-prefixSize, NULL, 0); } -LZ4_FORCE_O2_GCC_PPC64LE +LZ4_FORCE_O2 int LZ4_decompress_safe_forceExtDict(const char* source, char* dest, int compressedSize, int maxOutputSize, const void* dictStart, size_t dictSize) @@ -2132,7 +2229,7 @@ int LZ4_decompress_safe_forceExtDict(const char* source, char* dest, (BYTE*)dest, (const BYTE*)dictStart, dictSize); } -LZ4_FORCE_O2_GCC_PPC64LE +LZ4_FORCE_O2 static int LZ4_decompress_fast_extDict(const char* source, char* dest, int originalSize, const void* dictStart, size_t dictSize) { @@ -2221,7 +2318,7 @@ int LZ4_decoderRingBufferSize(int maxBlockSize) If it's not possible, save the relevant part of decoded data into a safe buffer, and indicate where it stands using LZ4_setStreamDecode() */ -LZ4_FORCE_O2_GCC_PPC64LE +LZ4_FORCE_O2 int LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int compressedSize, int maxOutputSize) { LZ4_streamDecode_t_internal* lz4sd = &LZ4_streamDecode->internal_donotuse; @@ -2261,7 +2358,7 @@ int LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, const ch return result; } -LZ4_FORCE_O2_GCC_PPC64LE +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; @@ -2374,7 +2471,7 @@ int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, /* Obsolete Streaming functions */ -int LZ4_sizeofStreamState() { return LZ4_STREAMSIZE; } +int LZ4_sizeofStreamState(void) { return LZ4_STREAMSIZE; } int LZ4_resetStreamState(void* state, char* inputBuffer) { diff --git a/C/lz4/lz4.h b/C/lz4/lz4.h index 32108e232..7ab1e483a 100644 --- a/C/lz4/lz4.h +++ b/C/lz4/lz4.h @@ -100,7 +100,7 @@ extern "C" { /*------ 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 2 /* for tweaks, bug-fixes, or development */ +#define LZ4_VERSION_RELEASE 3 /* for tweaks, bug-fixes, or development */ #define LZ4_VERSION_NUMBER (LZ4_VERSION_MAJOR *100*100 + LZ4_VERSION_MINOR *100 + LZ4_VERSION_RELEASE) @@ -186,7 +186,8 @@ LZ4LIB_API int LZ4_compressBound(int inputSize); The larger the acceleration value, the faster the algorithm, but also the lesser the compression. It's a trade-off. It can be fine tuned, with each successive value providing roughly +~3% to speed. An acceleration value of "1" is the same as regular LZ4_compress_default() - Values <= 0 will be replaced by ACCELERATION_DEFAULT (currently == 1, see lz4.c). + Values <= 0 will be replaced by LZ4_ACCELERATION_DEFAULT (currently == 1, see lz4.c). + Values > LZ4_ACCELERATION_MAX will be replaced by LZ4_ACCELERATION_MAX (currently == 65537, see lz4.c). */ LZ4LIB_API int LZ4_compress_fast (const char* src, char* dst, int srcSize, int dstCapacity, int acceleration); @@ -212,7 +213,18 @@ LZ4LIB_API int LZ4_compress_fast_extState (void* state, const char* src, char* d * New value is necessarily <= input value. * @return : Nb bytes written into 'dst' (necessarily <= targetDestSize) * or 0 if compression fails. -*/ + * + * Note : from v1.8.2 to v1.9.1, this function had a bug (fixed un v1.9.2+): + * the produced compressed content could, in specific circumstances, + * require to be decompressed into a destination buffer larger + * by at least 1 byte than the content to decompress. + * If an application uses `LZ4_compress_destSize()`, + * it's highly recommended to update liblz4 to v1.9.2 or better. + * If this can't be done or ensured, + * the receiving decompression function should provide + * a dstCapacity which is > decompressedSize, by at least 1 byte. + * See https://github.com/lz4/lz4/issues/859 for details + */ LZ4LIB_API int LZ4_compress_destSize (const char* src, char* dst, int* srcSizePtr, int targetDstSize); @@ -220,25 +232,35 @@ LZ4LIB_API int LZ4_compress_destSize (const char* src, char* dst, int* srcSizePt * Decompress an LZ4 compressed block, of size 'srcSize' at position 'src', * into destination buffer 'dst' of size 'dstCapacity'. * Up to 'targetOutputSize' bytes will be decoded. - * The function stops decoding on reaching this objective, - * which can boost performance when only the beginning of a block is required. + * The function stops decoding on reaching this objective. + * This can be useful to boost performance + * whenever only the beginning of a block is required. * - * @return : the number of bytes decoded in `dst` (necessarily <= dstCapacity) + * @return : the number of bytes decoded in `dst` (necessarily <= targetOutputSize) * If source stream is detected malformed, function returns a negative result. * - * Note : @return can be < targetOutputSize, if compressed block contains less data. + * Note 1 : @return can be < targetOutputSize, if compressed block contains less data. + * + * Note 2 : targetOutputSize must be <= dstCapacity * - * Note 2 : this function features 2 parameters, targetOutputSize and dstCapacity, - * and expects targetOutputSize <= dstCapacity. - * It effectively stops decoding on reaching targetOutputSize, + * Note 3 : this function effectively stops decoding on reaching targetOutputSize, * so dstCapacity is kind of redundant. - * This is because in a previous version of this function, - * decoding operation would not "break" a sequence in the middle. - * As a consequence, there was no guarantee that decoding would stop at exactly targetOutputSize, + * This is because in older versions of this function, + * decoding operation would still write complete sequences. + * Therefore, there was no guarantee that it would stop writing at exactly targetOutputSize, * it could write more bytes, though only up to dstCapacity. * Some "margin" used to be required for this operation to work properly. - * This is no longer necessary. - * The function nonetheless keeps its signature, in an effort to not break API. + * Thankfully, this is no longer necessary. + * The function nonetheless keeps the same signature, in an effort to preserve API compatibility. + * + * Note 4 : If srcSize is the exact size of the block, + * then targetOutputSize can be any value, + * including larger than the block's decompressed size. + * The function will, at most, generate block's decompressed size. + * + * Note 5 : If srcSize is _larger_ than block's compressed size, + * then targetOutputSize **MUST** be <= block's decompressed size. + * Otherwise, *silent corruption will occur*. */ LZ4LIB_API int LZ4_decompress_safe_partial (const char* src, char* dst, int srcSize, int targetOutputSize, int dstCapacity); @@ -547,74 +569,64 @@ LZ4LIB_STATIC_API void LZ4_attach_dictionary(LZ4_stream_t* workingStream, const #define LZ4_H_98237428734687 /*-************************************************************ - * PRIVATE DEFINITIONS + * Private Definitions ************************************************************** * Do not use these definitions directly. * They are only exposed to allow static allocation of `LZ4_stream_t` and `LZ4_streamDecode_t`. - * Accessing members will expose code to API and/or ABI break in future versions of the library. + * Accessing members will expose user code to API and/or ABI break in future versions of the library. **************************************************************/ #define LZ4_HASHLOG (LZ4_MEMORY_USAGE-2) #define LZ4_HASHTABLESIZE (1 << LZ4_MEMORY_USAGE) #define LZ4_HASH_SIZE_U32 (1 << LZ4_HASHLOG) /* required as macro for static allocation */ #if defined(__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) -#include - -typedef struct LZ4_stream_t_internal LZ4_stream_t_internal; -struct LZ4_stream_t_internal { - uint32_t hashTable[LZ4_HASH_SIZE_U32]; - uint32_t currentOffset; - uint16_t dirty; - uint16_t tableType; - const uint8_t* dictionary; - const LZ4_stream_t_internal* dictCtx; - uint32_t dictSize; -}; - -typedef struct { - const uint8_t* externalDict; - size_t extDictSize; - const uint8_t* prefixEnd; - size_t prefixSize; -} LZ4_streamDecode_t_internal; - +# include + typedef int8_t LZ4_i8; + typedef uint8_t LZ4_byte; + typedef uint16_t LZ4_u16; + typedef uint32_t LZ4_u32; #else + typedef signed char LZ4_i8; + typedef unsigned char LZ4_byte; + typedef unsigned short LZ4_u16; + typedef unsigned int LZ4_u32; +#endif typedef struct LZ4_stream_t_internal LZ4_stream_t_internal; struct LZ4_stream_t_internal { - unsigned int hashTable[LZ4_HASH_SIZE_U32]; - unsigned int currentOffset; - unsigned short dirty; - unsigned short tableType; - const unsigned char* dictionary; + LZ4_u32 hashTable[LZ4_HASH_SIZE_U32]; + LZ4_u32 currentOffset; + LZ4_u32 tableType; + const LZ4_byte* dictionary; const LZ4_stream_t_internal* dictCtx; - unsigned int dictSize; + LZ4_u32 dictSize; }; typedef struct { - const unsigned char* externalDict; - const unsigned char* prefixEnd; + const LZ4_byte* externalDict; size_t extDictSize; + const LZ4_byte* prefixEnd; size_t prefixSize; } LZ4_streamDecode_t_internal; -#endif /*! LZ4_stream_t : - * information structure to track an LZ4 stream. + * 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 a future version. + * this definition is not API/ABI safe, and may change in future versions. */ -#define LZ4_STREAMSIZE_U64 ((1 << (LZ4_MEMORY_USAGE-3)) + 4 + ((sizeof(void*)==16) ? 4 : 0) /*AS-400*/ ) -#define LZ4_STREAMSIZE (LZ4_STREAMSIZE_U64 * sizeof(unsigned long long)) +#define LZ4_STREAMSIZE 16416 /* static size, for inter-version compatibility */ +#define LZ4_STREAMSIZE_VOIDP (LZ4_STREAMSIZE / sizeof(void*)) union LZ4_stream_u { - unsigned long long table[LZ4_STREAMSIZE_U64]; + void* table[LZ4_STREAMSIZE_VOIDP]; LZ4_stream_t_internal internal_donotuse; -} ; /* previously typedef'd to LZ4_stream_t */ +}; /* previously typedef'd to LZ4_stream_t */ + /*! LZ4_initStream() : v1.9.0+ * An LZ4_stream_t structure must be initialized at least once. @@ -667,22 +679,21 @@ union LZ4_streamDecode_u { #ifdef LZ4_DISABLE_DEPRECATE_WARNINGS # define LZ4_DEPRECATED(message) /* disable deprecation warnings */ #else -# define LZ4_GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__) # if defined (__cplusplus) && (__cplusplus >= 201402) /* C++14 or greater */ # define LZ4_DEPRECATED(message) [[deprecated(message)]] -# elif (LZ4_GCC_VERSION >= 405) || defined(__clang__) -# define LZ4_DEPRECATED(message) __attribute__((deprecated(message))) -# elif (LZ4_GCC_VERSION >= 301) -# define LZ4_DEPRECATED(message) __attribute__((deprecated)) # elif defined(_MSC_VER) # define LZ4_DEPRECATED(message) __declspec(deprecated(message)) +# elif defined(__clang__) || (defined(__GNUC__) && (__GNUC__ * 10 + __GNUC_MINOR__ >= 45)) +# define LZ4_DEPRECATED(message) __attribute__((deprecated(message))) +# elif defined(__GNUC__) && (__GNUC__ * 10 + __GNUC_MINOR__ >= 31) +# define LZ4_DEPRECATED(message) __attribute__((deprecated)) # else -# pragma message("WARNING: You need to implement LZ4_DEPRECATED for this compiler") -# define LZ4_DEPRECATED(message) +# pragma message("WARNING: LZ4_DEPRECATED needs custom implementation for this compiler") +# define LZ4_DEPRECATED(message) /* disabled */ # endif #endif /* LZ4_DISABLE_DEPRECATE_WARNINGS */ -/* Obsolete compression functions */ +/*! Obsolete compression functions (since v1.7.3) */ LZ4_DEPRECATED("use LZ4_compress_default() instead") LZ4LIB_API int LZ4_compress (const char* src, char* dest, int srcSize); LZ4_DEPRECATED("use LZ4_compress_default() instead") LZ4LIB_API int LZ4_compress_limitedOutput (const char* src, char* dest, int srcSize, int maxOutputSize); LZ4_DEPRECATED("use LZ4_compress_fast_extState() instead") LZ4LIB_API int LZ4_compress_withState (void* state, const char* source, char* dest, int inputSize); @@ -690,11 +701,12 @@ LZ4_DEPRECATED("use LZ4_compress_fast_extState() instead") LZ4LIB_API int LZ4_co LZ4_DEPRECATED("use LZ4_compress_fast_continue() instead") LZ4LIB_API int LZ4_compress_continue (LZ4_stream_t* LZ4_streamPtr, const char* source, char* dest, int inputSize); LZ4_DEPRECATED("use LZ4_compress_fast_continue() instead") LZ4LIB_API int LZ4_compress_limitedOutput_continue (LZ4_stream_t* LZ4_streamPtr, const char* source, char* dest, int inputSize, int maxOutputSize); -/* Obsolete decompression functions */ +/*! Obsolete decompression functions (since v1.8.0) */ LZ4_DEPRECATED("use LZ4_decompress_fast() instead") LZ4LIB_API int LZ4_uncompress (const char* source, char* dest, int outputSize); LZ4_DEPRECATED("use LZ4_decompress_safe() instead") LZ4LIB_API int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize); -/* Obsolete streaming functions; degraded functionality; do not use! +/* Obsolete streaming functions (since v1.7.0) + * degraded functionality; do not use! * * In order to perform streaming compression, these functions depended on data * that is no longer tracked in the state. They have been preserved as well as @@ -708,23 +720,22 @@ LZ4_DEPRECATED("Use LZ4_createStream() instead") LZ4LIB_API int LZ4_sizeofStre LZ4_DEPRECATED("Use LZ4_resetStream() instead") LZ4LIB_API int LZ4_resetStreamState(void* state, char* inputBuffer); LZ4_DEPRECATED("Use LZ4_saveDict() instead") LZ4LIB_API char* LZ4_slideInputBuffer (void* state); -/* Obsolete streaming decoding functions */ +/*! Obsolete streaming decoding functions (since v1.7.0) */ LZ4_DEPRECATED("use LZ4_decompress_safe_usingDict() instead") LZ4LIB_API int LZ4_decompress_safe_withPrefix64k (const char* src, char* dst, int compressedSize, int maxDstSize); LZ4_DEPRECATED("use LZ4_decompress_fast_usingDict() instead") LZ4LIB_API int LZ4_decompress_fast_withPrefix64k (const char* src, char* dst, int originalSize); -/*! LZ4_decompress_fast() : **unsafe!** +/*! Obsolete LZ4_decompress_fast variants (since v1.9.0) : * These functions used to be faster than LZ4_decompress_safe(), - * but it has changed, and they are now slower than LZ4_decompress_safe(). + * but this is no longer the case. They are now slower. * This is because LZ4_decompress_fast() doesn't know the input size, - * and therefore must progress more cautiously in the input buffer to not read beyond the end of block. + * and therefore must progress more cautiously into the input buffer to not read beyond the end of block. * On top of that `LZ4_decompress_fast()` is not protected vs malformed or malicious inputs, making it a security liability. * As a consequence, LZ4_decompress_fast() is strongly discouraged, and deprecated. * * The last remaining LZ4_decompress_fast() specificity is that * it can decompress a block without knowing its compressed size. - * Such functionality could be achieved in a more secure manner, - * by also providing the maximum size of input buffer, - * but it would require new prototypes, and adaptation of the implementation to this new use case. + * Such functionality can be achieved in a more secure manner + * by employing LZ4_decompress_safe_partial(). * * Parameters: * originalSize : is the uncompressed size to regenerate. @@ -739,7 +750,6 @@ LZ4_DEPRECATED("use LZ4_decompress_fast_usingDict() instead") LZ4LIB_API int LZ4 * But they may happen if input data is invalid (error or intentional tampering). * As a consequence, use these functions in trusted environments with trusted data **only**. */ - LZ4_DEPRECATED("This function is deprecated and unsafe. Consider using LZ4_decompress_safe() instead") LZ4LIB_API int LZ4_decompress_fast (const char* src, char* dst, int originalSize); LZ4_DEPRECATED("This function is deprecated and unsafe. Consider using LZ4_decompress_safe_continue() instead") diff --git a/C/lz4/lz4frame.c b/C/lz4/lz4frame.c index 9d29af0ed..b052f17ee 100644 --- a/C/lz4/lz4frame.c +++ b/C/lz4/lz4frame.c @@ -71,8 +71,8 @@ * towards another library or solution of their choice * by modifying below section. */ -#include /* malloc, calloc, free */ #ifndef LZ4_SRC_INCLUDED /* avoid redefinition when sources are coalesced */ +# include /* malloc, calloc, free */ # define ALLOC(s) malloc(s) # define ALLOC_AND_ZERO(s) calloc(1,(s)) # define FREEMEM(p) free(p) @@ -533,7 +533,7 @@ void LZ4F_freeCDict(LZ4F_CDict* cdict) * 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_compressionContext_t* 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); @@ -541,20 +541,18 @@ LZ4F_errorCode_t LZ4F_createCompressionContext(LZ4F_compressionContext_t* LZ4F_c cctxPtr->version = version; cctxPtr->cStage = 0; /* Next stage : init stream */ - *LZ4F_compressionContextPtr = (LZ4F_compressionContext_t)cctxPtr; + *LZ4F_compressionContextPtr = cctxPtr; return LZ4F_OK_NoError; } -LZ4F_errorCode_t LZ4F_freeCompressionContext(LZ4F_compressionContext_t LZ4F_compressionContext) +LZ4F_errorCode_t LZ4F_freeCompressionContext(LZ4F_cctx* cctxPtr) { - LZ4F_cctx_t* const cctxPtr = (LZ4F_cctx_t*)LZ4F_compressionContext; - if (cctxPtr != NULL) { /* support free on NULL */ - FREEMEM(cctxPtr->lz4CtxPtr); /* works because LZ4_streamHC_t and LZ4_stream_t are simple POD types */ + FREEMEM(cctxPtr->lz4CtxPtr); /* note: LZ4_streamHC_t and LZ4_stream_t are simple POD types */ FREEMEM(cctxPtr->tmpBuff); - FREEMEM(LZ4F_compressionContext); + FREEMEM(cctxPtr); } return LZ4F_OK_NoError; @@ -725,6 +723,9 @@ size_t LZ4F_compressBegin(LZ4F_cctx* cctxPtr, */ size_t LZ4F_compressBound(size_t srcSize, const LZ4F_preferences_t* preferencesPtr) { + if (preferencesPtr && preferencesPtr->autoFlush) { + return LZ4F_compressBound_internal(srcSize, preferencesPtr, 0); + } return LZ4F_compressBound_internal(srcSize, preferencesPtr, (size_t)-1); } @@ -747,6 +748,7 @@ static size_t LZ4F_makeBlock(void* dst, (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); cSize = (U32)srcSize; LZ4F_writeLE32(cSizePtr, cSize | LZ4F_BLOCKUNCOMPRESSED_FLAG); memcpy(cSizePtr+BHSize, src, srcSize); @@ -989,6 +991,7 @@ size_t LZ4F_compressEnd(LZ4F_cctx* cctxPtr, BYTE* dstPtr = dstStart; size_t const flushSize = LZ4F_flush(cctxPtr, dstBuffer, dstCapacity, compressOptionsPtr); + DEBUGLOG(5,"LZ4F_compressEnd: dstCapacity=%u", (unsigned)dstCapacity); if (LZ4F_isError(flushSize)) return flushSize; dstPtr += flushSize; @@ -1002,6 +1005,7 @@ size_t LZ4F_compressEnd(LZ4F_cctx* cctxPtr, if (cctxPtr->prefs.frameInfo.contentChecksumFlag == LZ4F_contentChecksumEnabled) { U32 const xxh = XXH32_digest(&(cctxPtr->xxh)); if (dstCapacity < 8) return err0r(LZ4F_ERROR_dstMaxSize_tooSmall); + DEBUGLOG(5,"Writing 32-bit content checksum"); LZ4F_writeLE32(dstPtr, xxh); dstPtr+=4; /* content Checksum */ } @@ -1112,6 +1116,7 @@ static size_t LZ4F_decodeHeader(LZ4F_dctx* dctx, const void* src, size_t srcSize size_t frameHeaderSize; const BYTE* srcPtr = (const BYTE*)src; + DEBUGLOG(5, "LZ4F_decodeHeader"); /* need to decode header to get frameInfo */ if (srcSize < minFHSize) return err0r(LZ4F_ERROR_frameHeader_incomplete); /* minimal frame header size */ MEM_INIT(&(dctx->frameInfo), 0, sizeof(dctx->frameInfo)); @@ -1132,8 +1137,10 @@ static size_t LZ4F_decodeHeader(LZ4F_dctx* dctx, const void* src, size_t srcSize /* control magic number */ #ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION - if (LZ4F_readLE32(srcPtr) != LZ4F_MAGICNUMBER) + if (LZ4F_readLE32(srcPtr) != LZ4F_MAGICNUMBER) { + DEBUGLOG(4, "frame header error : unknown magic number"); return err0r(LZ4F_ERROR_frameType_unknown); + } #endif dctx->frameInfo.frameType = LZ4F_frame; @@ -1282,15 +1289,20 @@ LZ4F_errorCode_t LZ4F_getFrameInfo(LZ4F_dctx* dctx, /* LZ4F_updateDict() : - * only used for LZ4F_blockLinked mode */ + * only used for LZ4F_blockLinked mode + * Condition : dstPtr != NULL + */ static void LZ4F_updateDict(LZ4F_dctx* dctx, const BYTE* dstPtr, size_t dstSize, const BYTE* dstBufferStart, unsigned withinTmp) { - if (dctx->dictSize==0) - dctx->dict = (const BYTE*)dstPtr; /* priority to dictionary continuity */ + assert(dstPtr != NULL); + if (dctx->dictSize==0) { + dctx->dict = (const BYTE*)dstPtr; /* priority to prefix mode */ + } + assert(dctx->dict != NULL); - if (dctx->dict + dctx->dictSize == dstPtr) { /* dictionary continuity, directly within dstBuffer */ + if (dctx->dict + dctx->dictSize == dstPtr) { /* prefix mode, everything within dstBuffer */ dctx->dictSize += dstSize; return; } @@ -1304,9 +1316,10 @@ static void LZ4F_updateDict(LZ4F_dctx* dctx, assert(dstSize < 64 KB); /* if dstSize >= 64 KB, dictionary would be set into dstBuffer directly */ - /* dstBuffer does not contain whole useful history (64 KB), so it must be saved within tmpOut */ + /* dstBuffer does not contain whole useful history (64 KB), so it must be saved within tmpOutBuffer */ + assert(dctx->tmpOutBuffer != NULL); - if ((withinTmp) && (dctx->dict == dctx->tmpOutBuffer)) { /* continue history within tmpOutBuffer */ + if (withinTmp && (dctx->dict == dctx->tmpOutBuffer)) { /* continue history within tmpOutBuffer */ /* withinTmp expectation : content of [dstPtr,dstSize] is same as [dict+dictSize,dstSize], so we just extend it */ assert(dctx->dict + dctx->dictSize == dctx->tmpOut + dctx->tmpOutStart); dctx->dictSize += dstSize; @@ -1378,17 +1391,21 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, const BYTE* const srcEnd = srcStart + *srcSizePtr; const BYTE* srcPtr = srcStart; BYTE* const dstStart = (BYTE*)dstBuffer; - BYTE* const dstEnd = dstStart + *dstSizePtr; + BYTE* const dstEnd = dstStart ? dstStart + *dstSizePtr : NULL; BYTE* dstPtr = dstStart; const BYTE* selectedIn = NULL; unsigned doAnotherStage = 1; size_t nextSrcSizeHint = 1; + DEBUGLOG(5, "LZ4F_decompress : %p,%u => %p,%u", + srcBuffer, (unsigned)*srcSizePtr, dstBuffer, (unsigned)*dstSizePtr); + if (dstBuffer == NULL) assert(*dstSizePtr == 0); MEM_INIT(&optionsNull, 0, sizeof(optionsNull)); if (decompressOptionsPtr==NULL) decompressOptionsPtr = &optionsNull; *srcSizePtr = 0; *dstSizePtr = 0; + assert(dctx != NULL); /* behaves as a state machine */ @@ -1398,6 +1415,7 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, { case dstage_getFrameHeader: + 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; @@ -1411,6 +1429,7 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, /* fall-through */ case dstage_storeFrameHeader: + DEBUGLOG(6, "dstage_storeFrameHeader"); { size_t const sizeToCopy = MIN(dctx->tmpInTarget - dctx->tmpInSize, (size_t)(srcEnd - srcPtr)); memcpy(dctx->header + dctx->tmpInSize, srcPtr, sizeToCopy); dctx->tmpInSize += sizeToCopy; @@ -1427,6 +1446,7 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, break; case dstage_init: + DEBUGLOG(6, "dstage_init"); if (dctx->frameInfo.contentChecksumFlag) (void)XXH32_reset(&(dctx->xxh), 0); /* internal buffers allocation */ { size_t const bufferNeeded = dctx->maxBlockSize @@ -1480,17 +1500,21 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, } /* if (dctx->dStage == dstage_storeBlockHeader) */ /* decode block header */ - { size_t const nextCBlockSize = LZ4F_readLE32(selectedIn) & 0x7FFFFFFFU; + { U32 const blockHeader = LZ4F_readLE32(selectedIn); + size_t const nextCBlockSize = blockHeader & 0x7FFFFFFFU; size_t const crcSize = dctx->frameInfo.blockChecksumFlag * BFSize; - if (nextCBlockSize==0) { /* frameEnd signal, no more block */ + if (blockHeader==0) { /* frameEnd signal, no more block */ + DEBUGLOG(5, "end of frame"); dctx->dStage = dstage_getSuffix; break; } - if (nextCBlockSize > dctx->maxBlockSize) + if (nextCBlockSize > dctx->maxBlockSize) { return err0r(LZ4F_ERROR_maxBlockSize_invalid); - if (LZ4F_readLE32(selectedIn) & LZ4F_BLOCKUNCOMPRESSED_FLAG) { + } + if (blockHeader & LZ4F_BLOCKUNCOMPRESSED_FLAG) { /* next block is uncompressed */ dctx->tmpInTarget = nextCBlockSize; + DEBUGLOG(5, "next block is uncompressed (size %u)", (U32)nextCBlockSize); if (dctx->frameInfo.blockChecksumFlag) { (void)XXH32_reset(&dctx->blockChecksum, 0); } @@ -1508,20 +1532,26 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, } case dstage_copyDirect: /* uncompressed block */ - { size_t const minBuffSize = MIN((size_t)(srcEnd-srcPtr), (size_t)(dstEnd-dstPtr)); - size_t const sizeToCopy = MIN(dctx->tmpInTarget, minBuffSize); - memcpy(dstPtr, srcPtr, sizeToCopy); - 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.contentSize) - dctx->frameRemainingSize -= sizeToCopy; + DEBUGLOG(6, "dstage_copyDirect"); + { size_t sizeToCopy; + if (dstPtr == NULL) { + sizeToCopy = 0; + } else { + 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->frameInfo.contentChecksumFlag) + (void)XXH32_update(&dctx->xxh, srcPtr, sizeToCopy); + if (dctx->frameInfo.contentSize) + dctx->frameRemainingSize -= sizeToCopy; - /* history management (linked blocks only)*/ - if (dctx->frameInfo.blockMode == LZ4F_blockLinked) - LZ4F_updateDict(dctx, dstPtr, sizeToCopy, dstStart, 0); + /* history management (linked blocks only)*/ + if (dctx->frameInfo.blockMode == LZ4F_blockLinked) { + LZ4F_updateDict(dctx, dstPtr, sizeToCopy, dstStart, 0); + } } srcPtr += sizeToCopy; dstPtr += sizeToCopy; @@ -1534,15 +1564,16 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, break; } dctx->tmpInTarget -= sizeToCopy; /* need to copy more */ - nextSrcSizeHint = dctx->tmpInTarget + - +(dctx->frameInfo.blockChecksumFlag ? BFSize : 0) - + BHSize /* next header size */; - doAnotherStage = 0; - break; } + nextSrcSizeHint = dctx->tmpInTarget + + +(dctx->frameInfo.blockChecksumFlag ? BFSize : 0) + + BHSize /* next header size */; + doAnotherStage = 0; + break; /* check block checksum for recently transferred uncompressed block */ case dstage_getBlockChecksum: + DEBUGLOG(6, "dstage_getBlockChecksum"); { const void* crcSrc; if ((srcEnd-srcPtr >= 4) && (dctx->tmpInSize==0)) { crcSrc = srcPtr; @@ -1562,8 +1593,12 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, { U32 const readCRC = LZ4F_readLE32(crcSrc); U32 const calcCRC = XXH32_digest(&dctx->blockChecksum); #ifndef FUZZING_BUILD_MODE_UNSAFE_FOR_PRODUCTION - if (readCRC != calcCRC) + DEBUGLOG(6, "compare block checksum"); + if (readCRC != calcCRC) { + DEBUGLOG(4, "incorrect block checksum: %08X != %08X", + readCRC, calcCRC); return err0r(LZ4F_ERROR_blockChecksum_invalid); + } #else (void)readCRC; (void)calcCRC; @@ -1573,6 +1608,7 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, break; case dstage_getCBlock: + DEBUGLOG(6, "dstage_getCBlock"); if ((size_t)(srcEnd-srcPtr) < dctx->tmpInTarget) { dctx->tmpInSize = 0; dctx->dStage = dstage_storeCBlock; @@ -1582,7 +1618,7 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, selectedIn = srcPtr; srcPtr += dctx->tmpInTarget; - if (0) /* jump over next block */ + if (0) /* always jump over next block */ case dstage_storeCBlock: { size_t const wantedData = dctx->tmpInTarget - dctx->tmpInSize; size_t const inputLeft = (size_t)(srcEnd-srcPtr); @@ -1619,6 +1655,7 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, 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 */ dict += dictSize - 64 KB; @@ -1636,8 +1673,9 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, dctx->frameRemainingSize -= (size_t)decodedSize; /* dictionary management */ - if (dctx->frameInfo.blockMode==LZ4F_blockLinked) + if (dctx->frameInfo.blockMode==LZ4F_blockLinked) { LZ4F_updateDict(dctx, dstPtr, (size_t)decodedSize, dstStart, 0); + } dstPtr += decodedSize; dctx->dStage = dstage_getBlockHeader; @@ -1684,7 +1722,9 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, /* fall-through */ case dstage_flushOut: /* flush decoded data from tmpOut to dstBuffer */ - { size_t const sizeToCopy = MIN(dctx->tmpOutSize - dctx->tmpOutStart, (size_t)(dstEnd-dstPtr)); + DEBUGLOG(6, "dstage_flushOut"); + if (dstPtr != NULL) { + size_t const sizeToCopy = MIN(dctx->tmpOutSize - dctx->tmpOutStart, (size_t)(dstEnd-dstPtr)); memcpy(dstPtr, dctx->tmpOut + dctx->tmpOutStart, sizeToCopy); /* dictionary management */ @@ -1693,16 +1733,15 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, dctx->tmpOutStart += sizeToCopy; dstPtr += sizeToCopy; - - if (dctx->tmpOutStart == dctx->tmpOutSize) { /* all flushed */ - dctx->dStage = dstage_getBlockHeader; /* get next block */ - break; - } - /* could not flush everything : stop there, just request a block header */ - doAnotherStage = 0; - nextSrcSizeHint = BHSize; + } + if (dctx->tmpOutStart == dctx->tmpOutSize) { /* all flushed */ + dctx->dStage = dstage_getBlockHeader; /* get next block */ break; } + /* could not flush everything : stop there, just request a block header */ + doAnotherStage = 0; + nextSrcSizeHint = BHSize; + break; case dstage_getSuffix: if (dctx->frameRemainingSize) @@ -1806,6 +1845,7 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, 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 */ + && (dctx->dict != NULL) /* dictionary exists */ && (!decompressOptionsPtr->stableDst) /* cannot rely on dst data to remain there for next call */ && ((unsigned)(dctx->dStage)-2 < (unsigned)(dstage_getSuffix)-2) ) /* valid stages : [init ... getSuffix[ */ { @@ -1815,9 +1855,9 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, const BYTE* oldDictEnd = dctx->dict + dctx->dictSize - dctx->tmpOutStart; if (dctx->tmpOutSize > 64 KB) copySize = 0; if (copySize > preserveSize) copySize = preserveSize; + assert(dctx->tmpOutBuffer != NULL); - if (copySize > 0) - memcpy(dctx->tmpOutBuffer + preserveSize - copySize, oldDictEnd - copySize, copySize); + memcpy(dctx->tmpOutBuffer + preserveSize - copySize, oldDictEnd - copySize, copySize); dctx->dict = dctx->tmpOutBuffer; dctx->dictSize = preserveSize + dctx->tmpOutStart; @@ -1825,8 +1865,7 @@ size_t LZ4F_decompress(LZ4F_dctx* dctx, const BYTE* const oldDictEnd = dctx->dict + dctx->dictSize; size_t const newDictSize = MIN(dctx->dictSize, 64 KB); - if (newDictSize > 0) - memcpy(dctx->tmpOutBuffer, oldDictEnd - newDictSize, newDictSize); + memcpy(dctx->tmpOutBuffer, oldDictEnd - newDictSize, newDictSize); dctx->dict = dctx->tmpOutBuffer; dctx->dictSize = newDictSize; diff --git a/C/lz4/lz4frame.h b/C/lz4/lz4frame.h index 391e48401..4573317ef 100644 --- a/C/lz4/lz4frame.h +++ b/C/lz4/lz4frame.h @@ -66,17 +66,22 @@ extern "C" { *****************************************************************/ /* LZ4_DLL_EXPORT : * Enable exporting of functions when building a Windows DLL - * LZ4FLIB_API : + * LZ4FLIB_VISIBILITY : * Control library symbols visibility. */ +#ifndef LZ4FLIB_VISIBILITY +# if defined(__GNUC__) && (__GNUC__ >= 4) +# define LZ4FLIB_VISIBILITY __attribute__ ((visibility ("default"))) +# else +# define LZ4FLIB_VISIBILITY +# endif +#endif #if defined(LZ4_DLL_EXPORT) && (LZ4_DLL_EXPORT==1) -# define LZ4FLIB_API __declspec(dllexport) +# define LZ4FLIB_API __declspec(dllexport) LZ4FLIB_VISIBILITY #elif defined(LZ4_DLL_IMPORT) && (LZ4_DLL_IMPORT==1) -# define LZ4FLIB_API __declspec(dllimport) -#elif defined(__GNUC__) && (__GNUC__ >= 4) -# define LZ4FLIB_API __attribute__ ((__visibility__ ("default"))) +# define LZ4FLIB_API __declspec(dllimport) LZ4FLIB_VISIBILITY #else -# define LZ4FLIB_API +# define LZ4FLIB_API LZ4FLIB_VISIBILITY #endif #ifdef LZ4F_DISABLE_DEPRECATE_WARNINGS @@ -103,7 +108,7 @@ LZ4FLIB_API const char* LZ4F_getErrorName(LZ4F_errorCode_t code); /**< return /*-************************************ * Frame compression types - **************************************/ + ************************************* */ /* #define LZ4F_ENABLE_OBSOLETE_ENUMS // uncomment to enable obsolete enums */ #ifdef LZ4F_ENABLE_OBSOLETE_ENUMS # define LZ4F_OBSOLETE_ENUM(x) , LZ4F_DEPRECATE(x) = LZ4F_##x @@ -113,7 +118,8 @@ LZ4FLIB_API const char* LZ4F_getErrorName(LZ4F_errorCode_t code); /**< return /* The larger the block size, the (slightly) better the compression ratio, * though there are diminishing returns. - * Larger blocks also increase memory usage on both compression and decompression sides. */ + * Larger blocks also increase memory usage on both compression and decompression sides. + */ typedef enum { LZ4F_default=0, LZ4F_max64KB=4, @@ -284,7 +290,7 @@ LZ4FLIB_API size_t LZ4F_compressBegin(LZ4F_cctx* cctx, * @return is always the same for a srcSize and prefsPtr. * prefsPtr is optional : when NULL is provided, preferences will be set to cover worst case scenario. * tech details : - * @return includes the possibility that internal buffer might already be filled by up to (blockSize-1) bytes. + * @return if automatic flushing is not enabled, includes the possibility that internal buffer might already be filled by up to (blockSize-1) bytes. * It also includes frame footer (ending + checksum), since it might be generated by LZ4F_compressEnd(). * @return doesn't include frame header, as it was already generated by LZ4F_compressBegin(). */ @@ -376,7 +382,7 @@ LZ4FLIB_API LZ4F_errorCode_t LZ4F_freeDecompressionContext(LZ4F_dctx* dctx); * note : Frame header size is variable, but is guaranteed to be * >= LZ4F_HEADER_SIZE_MIN bytes, and <= LZ4F_HEADER_SIZE_MAX bytes. */ -size_t LZ4F_headerSize(const void* src, size_t srcSize); +LZ4FLIB_API size_t LZ4F_headerSize(const void* src, size_t srcSize); /*! LZ4F_getFrameInfo() : * This function extracts frame parameters (max blockSize, dictID, etc.). @@ -426,8 +432,10 @@ LZ4FLIB_API size_t LZ4F_getFrameInfo(LZ4F_dctx* dctx, const void* srcBuffer, size_t* srcSizePtr); /*! LZ4F_decompress() : - * Call this function repetitively to regenerate compressed data from `srcBuffer`. - * The function will read up to *srcSizePtr bytes from srcBuffer, + * Call this function repetitively to regenerate data compressed in `srcBuffer`. + * + * The function requires a valid dctx state. + * It will read up to *srcSizePtr bytes from srcBuffer, * and decompress data into dstBuffer, of capacity *dstSizePtr. * * The nb of bytes consumed from srcBuffer will be written into *srcSizePtr (necessarily <= original value). @@ -493,9 +501,9 @@ extern "C" { * Use at your own risk. */ #ifdef LZ4F_PUBLISH_STATIC_FUNCTIONS -#define LZ4FLIB_STATIC_API LZ4FLIB_API +# define LZ4FLIB_STATIC_API LZ4FLIB_API #else -#define LZ4FLIB_STATIC_API +# define LZ4FLIB_STATIC_API #endif diff --git a/C/lz4/lz4hc.c b/C/lz4/lz4hc.c index 5922ed7b1..77c9f4305 100644 --- a/C/lz4/lz4hc.c +++ b/C/lz4/lz4hc.c @@ -53,7 +53,7 @@ #include "lz4hc.h" -/*=== Common LZ4 definitions ===*/ +/*=== Common definitions ===*/ #if defined(__GNUC__) # pragma GCC diagnostic ignored "-Wunused-function" #endif @@ -61,15 +61,16 @@ # pragma clang diagnostic ignored "-Wunused-function" #endif -/*=== Enums ===*/ -typedef enum { noDictCtx, usingDictCtxHc } dictCtx_directive; - - #define LZ4_COMMONDEFS_ONLY #ifndef LZ4_SRC_INCLUDED #include "lz4.c" /* LZ4_count, constants, mem */ #endif + +/*=== Enums ===*/ +typedef enum { noDictCtx, usingDictCtxHc } dictCtx_directive; + + /*=== Constants ===*/ #define OPTIMAL_ML (int)((ML_MASK-1)+MINMATCH) #define LZ4_OPT_NUM (1<<12) @@ -92,7 +93,7 @@ static U32 LZ4HC_hashPtr(const void* ptr) { return HASH_FUNCTION(LZ4_read32(ptr) **************************************/ static void LZ4HC_clearTables (LZ4HC_CCtx_internal* hc4) { - MEM_INIT((void*)hc4->hashTable, 0, sizeof(hc4->hashTable)); + MEM_INIT(hc4->hashTable, 0, sizeof(hc4->hashTable)); MEM_INIT(hc4->chainTable, 0xFF, sizeof(hc4->chainTable)); } @@ -161,8 +162,7 @@ int LZ4HC_countBack(const BYTE* const ip, const BYTE* const match, static U32 LZ4HC_rotatePattern(size_t const rotate, U32 const pattern) { size_t const bitsToRotate = (rotate & (sizeof(pattern) - 1)) << 3; - if (bitsToRotate == 0) - return pattern; + if (bitsToRotate == 0) return pattern; return LZ4HC_rotl32(pattern, (int)bitsToRotate); } @@ -172,7 +172,8 @@ static unsigned LZ4HC_countPattern(const BYTE* ip, const BYTE* const iEnd, U32 const pattern32) { const BYTE* const iStart = ip; - reg_t const pattern = (sizeof(pattern)==8) ? (reg_t)pattern32 + (((reg_t)pattern32) << 32) : pattern32; + reg_t const pattern = (sizeof(pattern)==8) ? + (reg_t)pattern32 + (((reg_t)pattern32) << (sizeof(pattern)*4)) : pattern32; while (likely(ip < iEnd-(sizeof(pattern)-1))) { reg_t const diff = LZ4_read_ARCH(ip) ^ pattern; @@ -270,7 +271,7 @@ LZ4HC_InsertAndGetWiderMatch ( DEBUGLOG(7, "First match at index %u / %u (lowestMatchIndex)", matchIndex, lowestMatchIndex); - while ((matchIndex>=lowestMatchIndex) && (nbAttempts)) { + while ((matchIndex>=lowestMatchIndex) && (nbAttempts>0)) { int matchLength=0; nbAttempts--; assert(matchIndex < ipIndex); @@ -389,8 +390,8 @@ LZ4HC_InsertAndGetWiderMatch ( if (lookBackLength==0) { /* no back possible */ size_t const maxML = MIN(currentSegmentLength, srcPatternLength); if ((size_t)longest < maxML) { - assert(base + matchIndex < ip); - if (ip - (base+matchIndex) > LZ4_DISTANCE_MAX) break; + assert(base + matchIndex != ip); + if ((size_t)(ip - base) - matchIndex > LZ4_DISTANCE_MAX) break; assert(maxML < 2 GB); longest = (int)maxML; *matchpos = base + matchIndex; /* virtual pos, relative to ip, to retrieve offset */ @@ -410,7 +411,7 @@ LZ4HC_InsertAndGetWiderMatch ( } /* while ((matchIndex>=lowestMatchIndex) && (nbAttempts)) */ if ( dict == usingDictCtxHc - && nbAttempts + && nbAttempts > 0 && ipIndex - lowestMatchIndex < LZ4_DISTANCE_MAX) { size_t const dictEndOffset = (size_t)(dictCtx->end - dictCtx->base); U32 dictMatchIndex = dictCtx->hashTable[LZ4HC_hashPtr(ip)]; @@ -460,74 +461,90 @@ int LZ4HC_InsertAndFindBestMatch(LZ4HC_CCtx_internal* const hc4, /* Index tabl * @return : 0 if ok, * 1 if buffer issue detected */ LZ4_FORCE_INLINE int LZ4HC_encodeSequence ( - const BYTE** ip, - BYTE** op, - const BYTE** anchor, + const BYTE** _ip, + BYTE** _op, + const BYTE** _anchor, int matchLength, const BYTE* const match, limitedOutput_directive limit, BYTE* oend) { +#define ip (*_ip) +#define op (*_op) +#define anchor (*_anchor) + size_t length; - BYTE* const token = (*op)++; + BYTE* const token = op++; #if defined(LZ4_DEBUG) && (LZ4_DEBUG >= 6) static const BYTE* start = NULL; static U32 totalCost = 0; - U32 const pos = (start==NULL) ? 0 : (U32)(*anchor - start); - U32 const ll = (U32)(*ip - *anchor); + U32 const pos = (start==NULL) ? 0 : (U32)(anchor - start); + U32 const ll = (U32)(ip - anchor); U32 const llAdd = (ll>=15) ? ((ll-15) / 255) + 1 : 0; U32 const mlAdd = (matchLength>=19) ? ((matchLength-19) / 255) + 1 : 0; U32 const cost = 1 + llAdd + ll + 2 + mlAdd; - if (start==NULL) start = *anchor; /* only works for single segment */ + if (start==NULL) start = anchor; /* only works for single segment */ /* g_debuglog_enable = (pos >= 2228) & (pos <= 2262); */ - DEBUGLOG(6, "pos:%7u -- literals:%3u, match:%4i, offset:%5u, cost:%3u + %u", + DEBUGLOG(6, "pos:%7u -- literals:%4u, match:%4i, offset:%5u, cost:%4u + %5u", pos, - (U32)(*ip - *anchor), matchLength, (U32)(*ip-match), + (U32)(ip - anchor), matchLength, (U32)(ip-match), cost, totalCost); totalCost += cost; #endif /* Encode Literal length */ - length = (size_t)(*ip - *anchor); - if ((limit) && ((*op + (length / 255) + length + (2 + 1 + LASTLITERALS)) > oend)) return 1; /* Check output limit */ + length = (size_t)(ip - anchor); + LZ4_STATIC_ASSERT(notLimited == 0); + /* Check output limit */ + if (limit && ((op + (length / 255) + length + (2 + 1 + LASTLITERALS)) > oend)) { + DEBUGLOG(6, "Not enough room to write %i literals (%i bytes remaining)", + (int)length, (int)(oend - op)); + return 1; + } if (length >= RUN_MASK) { size_t len = length - RUN_MASK; *token = (RUN_MASK << ML_BITS); - for(; len >= 255 ; len -= 255) *(*op)++ = 255; - *(*op)++ = (BYTE)len; + for(; len >= 255 ; len -= 255) *op++ = 255; + *op++ = (BYTE)len; } else { *token = (BYTE)(length << ML_BITS); } /* Copy Literals */ - LZ4_wildCopy8(*op, *anchor, (*op) + length); - *op += length; + LZ4_wildCopy8(op, anchor, op + length); + op += length; /* Encode Offset */ - assert( (*ip - match) <= LZ4_DISTANCE_MAX ); /* note : consider providing offset as a value, rather than as a pointer difference */ - LZ4_writeLE16(*op, (U16)(*ip-match)); *op += 2; + assert( (ip - match) <= LZ4_DISTANCE_MAX ); /* note : consider providing offset as a value, rather than as a pointer difference */ + LZ4_writeLE16(op, (U16)(ip - match)); op += 2; /* Encode MatchLength */ assert(matchLength >= MINMATCH); length = (size_t)matchLength - MINMATCH; - if ((limit) && (*op + (length / 255) + (1 + LASTLITERALS) > oend)) return 1; /* Check output limit */ + if (limit && (op + (length / 255) + (1 + LASTLITERALS) > oend)) { + DEBUGLOG(6, "Not enough room to write match length"); + return 1; /* Check output limit */ + } if (length >= ML_MASK) { *token += ML_MASK; length -= ML_MASK; - for(; length >= 510 ; length -= 510) { *(*op)++ = 255; *(*op)++ = 255; } - if (length >= 255) { length -= 255; *(*op)++ = 255; } - *(*op)++ = (BYTE)length; + for(; length >= 510 ; length -= 510) { *op++ = 255; *op++ = 255; } + if (length >= 255) { length -= 255; *op++ = 255; } + *op++ = (BYTE)length; } else { *token += (BYTE)(length); } /* Prepare next loop */ - *ip += matchLength; - *anchor = *ip; + ip += matchLength; + anchor = ip; return 0; } +#undef ip +#undef op +#undef anchor LZ4_FORCE_INLINE int LZ4HC_compress_hashChain ( LZ4HC_CCtx_internal* const ctx, @@ -535,7 +552,7 @@ LZ4_FORCE_INLINE int LZ4HC_compress_hashChain ( char* const dest, int* srcSizePtr, int const maxOutputSize, - unsigned maxNbAttempts, + int maxNbAttempts, const limitedOutput_directive limit, const dictCtx_directive dict ) @@ -565,7 +582,7 @@ LZ4_FORCE_INLINE int LZ4HC_compress_hashChain ( /* init */ *srcSizePtr = 0; if (limit == fillOutput) oend -= LASTLITERALS; /* Hack for support LZ4 format restriction */ - if (inputSize < LZ4_minLength) goto _last_literals; /* Input too small, no compression (all literals) */ + if (inputSize < LZ4_minLength) goto _last_literals; /* Input too small, no compression (all literals) */ /* Main Loop */ while (ip <= mflimit) { @@ -637,7 +654,11 @@ LZ4_FORCE_INLINE int LZ4HC_compress_hashChain ( if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ml, ref, limit, oend)) goto _dest_overflow; ip = start2; optr = op; - if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ml2, ref2, limit, oend)) goto _dest_overflow; + if (LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ml2, ref2, limit, oend)) { + ml = ml2; + ref = ref2; + goto _dest_overflow; + } continue; } @@ -709,17 +730,18 @@ LZ4_FORCE_INLINE int LZ4HC_compress_hashChain ( _last_literals: /* Encode Last Literals */ { size_t lastRunSize = (size_t)(iend - anchor); /* literals */ - size_t litLength = (lastRunSize + 255 - RUN_MASK) / 255; - size_t const totalSize = 1 + litLength + lastRunSize; + size_t llAdd = (lastRunSize + 255 - RUN_MASK) / 255; + size_t const totalSize = 1 + llAdd + lastRunSize; if (limit == fillOutput) oend += LASTLITERALS; /* restore correct value */ if (limit && (op + totalSize > oend)) { - if (limit == limitedOutput) return 0; /* Check output limit */ + if (limit == limitedOutput) return 0; /* adapt lastRunSize to fill 'dest' */ - lastRunSize = (size_t)(oend - op) - 1; - litLength = (lastRunSize + 255 - RUN_MASK) / 255; - lastRunSize -= litLength; + lastRunSize = (size_t)(oend - op) - 1 /*token*/; + llAdd = (lastRunSize + 256 - RUN_MASK) / 256; + lastRunSize -= llAdd; } - ip = anchor + lastRunSize; + DEBUGLOG(6, "Final literal run : %i literals", (int)lastRunSize); + ip = anchor + lastRunSize; /* can be != iend if limit==fillOutput */ if (lastRunSize >= RUN_MASK) { size_t accumulator = lastRunSize - RUN_MASK; @@ -739,9 +761,25 @@ LZ4_FORCE_INLINE int LZ4HC_compress_hashChain ( _dest_overflow: if (limit == fillOutput) { + /* Assumption : ip, anchor, ml and ref must be set correctly */ + size_t const ll = (size_t)(ip - anchor); + size_t const ll_addbytes = (ll + 240) / 255; + size_t const ll_totalCost = 1 + ll_addbytes + ll; + BYTE* const maxLitPos = oend - 3; /* 2 for offset, 1 for token */ + DEBUGLOG(6, "Last sequence overflowing"); op = optr; /* restore correct out pointer */ + if (op + ll_totalCost <= maxLitPos) { + /* ll validated; now adjust match length */ + size_t const bytesLeftForMl = (size_t)(maxLitPos - (op+ll_totalCost)); + size_t const maxMlSize = MINMATCH + (ML_MASK-1) + (bytesLeftForMl * 255); + assert(maxMlSize < INT_MAX); assert(ml >= 0); + if ((size_t)ml > maxMlSize) ml = (int)maxMlSize; + if ((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1 + ml >= MFLIMIT) { + LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ml, ref, notLimited, oend); + } } goto _last_literals; } + /* compression failed */ return 0; } @@ -752,7 +790,7 @@ static int LZ4HC_compress_optimal( LZ4HC_CCtx_internal* ctx, int const nbSearches, size_t sufficient_len, const limitedOutput_directive limit, int const fullUpdate, const dictCtx_directive dict, - HCfavor_e favorDecSpeed); + const HCfavor_e favorDecSpeed); LZ4_FORCE_INLINE int LZ4HC_compress_generic_internal ( @@ -769,7 +807,7 @@ LZ4_FORCE_INLINE int LZ4HC_compress_generic_internal ( typedef enum { lz4hc, lz4opt } lz4hc_strat_e; typedef struct { lz4hc_strat_e strat; - U32 nbSearches; + int nbSearches; U32 targetLength; } cParams_t; static const cParams_t clTable[LZ4HC_CLEVEL_MAX+1] = { @@ -788,7 +826,8 @@ LZ4_FORCE_INLINE int LZ4HC_compress_generic_internal ( { lz4opt,16384,LZ4_OPT_NUM }, /* 12==LZ4HC_CLEVEL_MAX */ }; - DEBUGLOG(4, "LZ4HC_compress_generic(ctx=%p, src=%p, srcSize=%d)", ctx, src, *srcSizePtr); + DEBUGLOG(4, "LZ4HC_compress_generic(ctx=%p, src=%p, srcSize=%d, limit=%d)", + ctx, src, *srcSizePtr, limit); if (limit == fillOutput && dstCapacity < 1) return 0; /* Impossible to store anything */ if ((U32)*srcSizePtr > (U32)LZ4_MAX_INPUT_SIZE) return 0; /* Unsupported input size (too large or negative) */ @@ -808,7 +847,7 @@ LZ4_FORCE_INLINE int LZ4HC_compress_generic_internal ( assert(cParam.strat == lz4opt); result = LZ4HC_compress_optimal(ctx, src, dst, srcSizePtr, dstCapacity, - (int)cParam.nbSearches, cParam.targetLength, limit, + cParam.nbSearches, cParam.targetLength, limit, cLevel == LZ4HC_CLEVEL_MAX, /* ultra mode */ dict, favor); } @@ -881,27 +920,22 @@ LZ4HC_compress_generic ( int LZ4_sizeofStateHC(void) { return (int)sizeof(LZ4_streamHC_t); } -#ifndef _MSC_VER /* for some reason, Visual fails the aligment test on 32-bit x86 : - * it reports an aligment of 8-bytes, - * while actually aligning LZ4_streamHC_t on 4 bytes. */ static size_t LZ4_streamHC_t_alignment(void) { - struct { char c; LZ4_streamHC_t t; } t_a; - return sizeof(t_a) - sizeof(t_a.t); -} +#if LZ4_ALIGN_TEST + typedef struct { char c; LZ4_streamHC_t t; } t_a; + return sizeof(t_a) - sizeof(LZ4_streamHC_t); +#else + return 1; /* effectively disabled */ #endif +} /* state is presumed correctly initialized, * in which case its size and alignment have already been validate */ int LZ4_compress_HC_extStateHC_fastReset (void* state, const char* src, char* dst, int srcSize, int dstCapacity, int compressionLevel) { LZ4HC_CCtx_internal* const ctx = &((LZ4_streamHC_t*)state)->internal_donotuse; -#ifndef _MSC_VER /* for some reason, Visual fails the aligment test on 32-bit x86 : - * it reports an aligment of 8-bytes, - * while actually aligning LZ4_streamHC_t on 4 bytes. */ - assert(((size_t)state & (LZ4_streamHC_t_alignment() - 1)) == 0); /* check alignment */ -#endif - if (((size_t)(state)&(sizeof(void*)-1)) != 0) return 0; /* Error : state is not aligned for pointers (32 or 64 bits) */ + if (!LZ4_isAligned(state, LZ4_streamHC_t_alignment())) return 0; LZ4_resetStreamHC_fast((LZ4_streamHC_t*)state, compressionLevel); LZ4HC_init_internal (ctx, (const BYTE*)src); if (dstCapacity < LZ4_compressBound(srcSize)) @@ -950,10 +984,11 @@ int LZ4_compress_HC_destSize(void* state, const char* source, char* dest, int* s /* allocation */ LZ4_streamHC_t* LZ4_createStreamHC(void) { - LZ4_streamHC_t* const LZ4_streamHCPtr = (LZ4_streamHC_t*)ALLOC(sizeof(LZ4_streamHC_t)); - if (LZ4_streamHCPtr==NULL) return NULL; - LZ4_initStreamHC(LZ4_streamHCPtr, sizeof(*LZ4_streamHCPtr)); /* full initialization, malloc'ed buffer can be full of garbage */ - return LZ4_streamHCPtr; + LZ4_streamHC_t* const state = + (LZ4_streamHC_t*)ALLOC_AND_ZERO(sizeof(LZ4_streamHC_t)); + if (state == NULL) return NULL; + LZ4_setCompressionLevel(state, LZ4HC_CLEVEL_DEFAULT); + return state; } int LZ4_freeStreamHC (LZ4_streamHC_t* LZ4_streamHCPtr) @@ -968,22 +1003,16 @@ int LZ4_freeStreamHC (LZ4_streamHC_t* LZ4_streamHCPtr) LZ4_streamHC_t* LZ4_initStreamHC (void* buffer, size_t size) { LZ4_streamHC_t* const LZ4_streamHCPtr = (LZ4_streamHC_t*)buffer; - if (buffer == NULL) return NULL; - if (size < sizeof(LZ4_streamHC_t)) return NULL; -#ifndef _MSC_VER /* for some reason, Visual fails the aligment test on 32-bit x86 : - * it reports an aligment of 8-bytes, - * while actually aligning LZ4_streamHC_t on 4 bytes. */ - if (((size_t)buffer) & (LZ4_streamHC_t_alignment() - 1)) return NULL; /* alignment check */ -#endif /* if compilation fails here, LZ4_STREAMHCSIZE must be increased */ LZ4_STATIC_ASSERT(sizeof(LZ4HC_CCtx_internal) <= LZ4_STREAMHCSIZE); - DEBUGLOG(4, "LZ4_initStreamHC(%p, %u)", LZ4_streamHCPtr, (unsigned)size); - /* end-base will trigger a clearTable on starting compression */ - LZ4_streamHCPtr->internal_donotuse.end = (const BYTE *)(ptrdiff_t)-1; - LZ4_streamHCPtr->internal_donotuse.base = NULL; - LZ4_streamHCPtr->internal_donotuse.dictCtx = NULL; - LZ4_streamHCPtr->internal_donotuse.favorDecSpeed = 0; - LZ4_streamHCPtr->internal_donotuse.dirty = 0; + DEBUGLOG(4, "LZ4_initStreamHC(%p, %u)", buffer, (unsigned)size); + /* check conditions */ + if (buffer == NULL) return NULL; + if (size < sizeof(LZ4_streamHC_t)) return NULL; + if (!LZ4_isAligned(buffer, LZ4_streamHC_t_alignment())) return NULL; + /* init */ + { LZ4HC_CCtx_internal* const hcstate = &(LZ4_streamHCPtr->internal_donotuse); + MEM_INIT(hcstate, 0, sizeof(*hcstate)); } LZ4_setCompressionLevel(LZ4_streamHCPtr, LZ4HC_CLEVEL_DEFAULT); return LZ4_streamHCPtr; } @@ -1028,7 +1057,7 @@ int LZ4_loadDictHC (LZ4_streamHC_t* LZ4_streamHCPtr, const char* dictionary, int dictSize) { LZ4HC_CCtx_internal* const ctxPtr = &LZ4_streamHCPtr->internal_donotuse; - DEBUGLOG(4, "LZ4_loadDictHC(%p, %p, %d)", LZ4_streamHCPtr, dictionary, dictSize); + DEBUGLOG(4, "LZ4_loadDictHC(ctx:%p, dict:%p, dictSize:%d)", LZ4_streamHCPtr, dictionary, dictSize); assert(LZ4_streamHCPtr != NULL); if (dictSize > 64 KB) { dictionary += (size_t)dictSize - 64 KB; @@ -1069,14 +1098,15 @@ static void LZ4HC_setExternalDict(LZ4HC_CCtx_internal* ctxPtr, const BYTE* newBl ctxPtr->dictCtx = NULL; } -static int LZ4_compressHC_continue_generic (LZ4_streamHC_t* LZ4_streamHCPtr, - const char* src, char* dst, - int* srcSizePtr, int dstCapacity, - limitedOutput_directive limit) +static int +LZ4_compressHC_continue_generic (LZ4_streamHC_t* LZ4_streamHCPtr, + const char* src, char* dst, + int* srcSizePtr, int dstCapacity, + limitedOutput_directive limit) { LZ4HC_CCtx_internal* const ctxPtr = &LZ4_streamHCPtr->internal_donotuse; - DEBUGLOG(4, "LZ4_compressHC_continue_generic(ctx=%p, src=%p, srcSize=%d)", - LZ4_streamHCPtr, src, *srcSizePtr); + DEBUGLOG(5, "LZ4_compressHC_continue_generic(ctx=%p, src=%p, srcSize=%d, limit=%d)", + LZ4_streamHCPtr, src, *srcSizePtr, limit); assert(ctxPtr != NULL); /* auto-init if forgotten */ if (ctxPtr->base == NULL) LZ4HC_init_internal (ctxPtr, (const BYTE*) src); @@ -1100,8 +1130,7 @@ static int LZ4_compressHC_continue_generic (LZ4_streamHC_t* LZ4_streamHCPtr, if (sourceEnd > dictEnd) sourceEnd = dictEnd; ctxPtr->lowLimit = (U32)(sourceEnd - ctxPtr->dictBase); if (ctxPtr->dictLimit - ctxPtr->lowLimit < 4) ctxPtr->lowLimit = ctxPtr->dictLimit; - } - } + } } return LZ4HC_compress_generic (ctxPtr, src, dst, srcSizePtr, dstCapacity, ctxPtr->compressionLevel, limit); } @@ -1121,23 +1150,30 @@ int LZ4_compress_HC_continue_destSize (LZ4_streamHC_t* LZ4_streamHCPtr, const ch -/* dictionary saving */ - +/* LZ4_saveDictHC : + * save history content + * into a user-provided buffer + * which is then used to continue compression + */ 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)); - DEBUGLOG(4, "LZ4_saveDictHC(%p, %p, %d)", LZ4_streamHCPtr, safeBuffer, dictSize); + DEBUGLOG(5, "LZ4_saveDictHC(%p, %p, %d)", LZ4_streamHCPtr, safeBuffer, dictSize); + assert(prefixSize >= 0); if (dictSize > 64 KB) dictSize = 64 KB; if (dictSize < 4) dictSize = 0; if (dictSize > prefixSize) dictSize = prefixSize; - memmove(safeBuffer, streamPtr->end - dictSize, dictSize); + if (safeBuffer == NULL) assert(dictSize == 0); + if (dictSize > 0) + memmove(safeBuffer, streamPtr->end - dictSize, dictSize); { U32 const endIndex = (U32)(streamPtr->end - streamPtr->base); streamPtr->end = (const BYTE*)safeBuffer + dictSize; streamPtr->base = streamPtr->end - endIndex; streamPtr->dictLimit = endIndex - (U32)dictSize; streamPtr->lowLimit = endIndex - (U32)dictSize; - if (streamPtr->nextToUpdate < streamPtr->dictLimit) streamPtr->nextToUpdate = streamPtr->dictLimit; + if (streamPtr->nextToUpdate < streamPtr->dictLimit) + streamPtr->nextToUpdate = streamPtr->dictLimit; } return dictSize; } @@ -1287,8 +1323,13 @@ static int LZ4HC_compress_optimal ( LZ4HC_CCtx_internal* ctx, const dictCtx_directive dict, const HCfavor_e favorDecSpeed) { + int retval = 0; #define TRAILING_LITERALS 3 +#ifdef LZ4HC_HEAPMODE + 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... */ +#endif const BYTE* ip = (const BYTE*) source; const BYTE* anchor = ip; @@ -1298,15 +1339,19 @@ static int LZ4HC_compress_optimal ( LZ4HC_CCtx_internal* ctx, BYTE* op = (BYTE*) dst; BYTE* opSaved = (BYTE*) dst; BYTE* oend = op + dstCapacity; + int ovml = MINMATCH; /* overflow - last sequence */ + const BYTE* ovref = NULL; /* init */ +#ifdef LZ4HC_HEAPMODE + if (opt == NULL) goto _return_label; +#endif DEBUGLOG(5, "LZ4HC_compress_optimal(dst=%p, dstCapa=%u)", dst, (unsigned)dstCapacity); *srcSizePtr = 0; if (limit == fillOutput) oend -= LASTLITERALS; /* Hack for support LZ4 format restriction */ if (sufficient_len >= LZ4_OPT_NUM) sufficient_len = LZ4_OPT_NUM-1; /* Main Loop */ - assert(ip - anchor < LZ4_MAX_INPUT_SIZE); while (ip <= mflimit) { int const llen = (int)(ip - anchor); int best_mlen, best_off; @@ -1320,8 +1365,11 @@ static int LZ4HC_compress_optimal ( LZ4HC_CCtx_internal* ctx, int const firstML = firstMatch.len; const BYTE* const matchPos = ip - firstMatch.off; opSaved = op; - if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), firstML, matchPos, limit, oend) ) /* updates ip, op and anchor */ + if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), firstML, matchPos, limit, oend) ) { /* updates ip, op and anchor */ + ovml = firstML; + ovref = matchPos; goto _dest_overflow; + } continue; } @@ -1463,7 +1511,7 @@ static int LZ4HC_compress_optimal ( LZ4HC_CCtx_internal* ctx, best_off = opt[last_match_pos].off; cur = last_match_pos - best_mlen; - encode: /* cur, last_match_pos, best_mlen, best_off must be set */ +encode: /* cur, last_match_pos, best_mlen, best_off must be set */ assert(cur < LZ4_OPT_NUM); assert(last_match_pos >= 1); /* == 1 when only one candidate */ DEBUGLOG(6, "reverse traversal, looking for shortest path (last_match_pos=%i)", last_match_pos); @@ -1493,25 +1541,31 @@ static int LZ4HC_compress_optimal ( LZ4HC_CCtx_internal* ctx, assert(ml >= MINMATCH); assert((offset >= 1) && (offset <= LZ4_DISTANCE_MAX)); opSaved = op; - if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ml, ip - offset, limit, oend) ) /* updates ip, op and anchor */ + if ( LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ml, ip - offset, limit, oend) ) { /* updates ip, op and anchor */ + ovml = ml; + ovref = ip - offset; goto _dest_overflow; - } } + } } } } /* while (ip <= mflimit) */ - _last_literals: +_last_literals: /* Encode Last Literals */ { size_t lastRunSize = (size_t)(iend - anchor); /* literals */ - size_t litLength = (lastRunSize + 255 - RUN_MASK) / 255; - size_t const totalSize = 1 + litLength + lastRunSize; + size_t llAdd = (lastRunSize + 255 - RUN_MASK) / 255; + size_t const totalSize = 1 + llAdd + lastRunSize; if (limit == fillOutput) oend += LASTLITERALS; /* restore correct value */ if (limit && (op + totalSize > oend)) { - if (limit == limitedOutput) return 0; /* Check output limit */ + if (limit == limitedOutput) { /* Check output limit */ + retval = 0; + goto _return_label; + } /* adapt lastRunSize to fill 'dst' */ - lastRunSize = (size_t)(oend - op) - 1; - litLength = (lastRunSize + 255 - RUN_MASK) / 255; - lastRunSize -= litLength; + lastRunSize = (size_t)(oend - op) - 1 /*token*/; + llAdd = (lastRunSize + 256 - RUN_MASK) / 256; + lastRunSize -= llAdd; } - ip = anchor + lastRunSize; + DEBUGLOG(6, "Final literal run : %i literals", (int)lastRunSize); + ip = anchor + lastRunSize; /* can be != iend if limit==fillOutput */ if (lastRunSize >= RUN_MASK) { size_t accumulator = lastRunSize - RUN_MASK; @@ -1527,12 +1581,35 @@ static int LZ4HC_compress_optimal ( LZ4HC_CCtx_internal* ctx, /* End */ *srcSizePtr = (int) (((const char*)ip) - source); - return (int) ((char*)op-dst); + retval = (int) ((char*)op-dst); + goto _return_label; - _dest_overflow: - if (limit == fillOutput) { - op = opSaved; /* restore correct out pointer */ - goto _last_literals; - } - return 0; - } +_dest_overflow: +if (limit == fillOutput) { + /* Assumption : ip, anchor, ovml and ovref must be set correctly */ + size_t const ll = (size_t)(ip - anchor); + size_t const ll_addbytes = (ll + 240) / 255; + size_t const ll_totalCost = 1 + ll_addbytes + ll; + BYTE* const maxLitPos = oend - 3; /* 2 for offset, 1 for token */ + DEBUGLOG(6, "Last sequence overflowing (only %i bytes remaining)", (int)(oend-1-opSaved)); + op = opSaved; /* restore correct out pointer */ + if (op + ll_totalCost <= maxLitPos) { + /* ll validated; now adjust match length */ + size_t const bytesLeftForMl = (size_t)(maxLitPos - (op+ll_totalCost)); + size_t const maxMlSize = MINMATCH + (ML_MASK-1) + (bytesLeftForMl * 255); + assert(maxMlSize < INT_MAX); assert(ovml >= 0); + if ((size_t)ovml > maxMlSize) ovml = (int)maxMlSize; + if ((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1 + ovml >= MFLIMIT) { + DEBUGLOG(6, "Space to end : %i + ml (%i)", (int)((oend + LASTLITERALS) - (op + ll_totalCost + 2) - 1), ovml); + DEBUGLOG(6, "Before : ip = %p, anchor = %p", ip, anchor); + LZ4HC_encodeSequence(UPDATABLE(ip, op, anchor), ovml, ovref, notLimited, oend); + DEBUGLOG(6, "After : ip = %p, anchor = %p", ip, anchor); + } } + goto _last_literals; +} +_return_label: +#ifdef LZ4HC_HEAPMODE + FREEMEM(opt); +#endif + return retval; +} diff --git a/C/lz4/lz4hc.h b/C/lz4/lz4hc.h index 44e35bbf6..3d441fb6f 100644 --- a/C/lz4/lz4hc.h +++ b/C/lz4/lz4hc.h @@ -198,57 +198,32 @@ LZ4LIB_API int LZ4_saveDictHC (LZ4_streamHC_t* streamHCPtr, char* safeBuffer, in #define LZ4HC_HASH_MASK (LZ4HC_HASHTABLESIZE - 1) -#if defined(__cplusplus) || (defined (__STDC_VERSION__) && (__STDC_VERSION__ >= 199901L) /* C99 */) -#include - -typedef struct LZ4HC_CCtx_internal LZ4HC_CCtx_internal; -struct LZ4HC_CCtx_internal -{ - uint32_t hashTable[LZ4HC_HASHTABLESIZE]; - uint16_t chainTable[LZ4HC_MAXD]; - const uint8_t* end; /* next block here to continue on current prefix */ - const uint8_t* base; /* All index relative to this position */ - const uint8_t* dictBase; /* alternate base for extDict */ - uint32_t dictLimit; /* below that point, need extDict */ - uint32_t lowLimit; /* below that point, no more dict */ - uint32_t nextToUpdate; /* index from which to continue dictionary update */ - short compressionLevel; - int8_t favorDecSpeed; /* favor decompression speed if this flag set, - otherwise, favor compression ratio */ - int8_t dirty; /* stream has to be fully reset if this flag is set */ - const LZ4HC_CCtx_internal* dictCtx; -}; - -#else - typedef struct LZ4HC_CCtx_internal LZ4HC_CCtx_internal; struct LZ4HC_CCtx_internal { - unsigned int hashTable[LZ4HC_HASHTABLESIZE]; - unsigned short chainTable[LZ4HC_MAXD]; - const unsigned char* end; /* next block here to continue on current prefix */ - const unsigned char* base; /* All index relative to this position */ - const unsigned char* dictBase; /* alternate base for extDict */ - unsigned int dictLimit; /* below that point, need extDict */ - unsigned int lowLimit; /* below that point, no more dict */ - unsigned int nextToUpdate; /* index from which to continue dictionary update */ - short compressionLevel; - char favorDecSpeed; /* favor decompression speed if this flag set, - otherwise, favor compression ratio */ - char dirty; /* stream has to be fully reset if this flag is set */ + 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 */ + 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 */ + short compressionLevel; + LZ4_i8 favorDecSpeed; /* favor decompression speed if this flag set, + otherwise, favor compression ratio */ + LZ4_i8 dirty; /* stream has to be fully reset if this flag is set */ const LZ4HC_CCtx_internal* dictCtx; }; -#endif - /* Do not use these definitions directly ! * Declare or allocate an LZ4_streamHC_t instead. */ -#define LZ4_STREAMHCSIZE (4*LZ4HC_HASHTABLESIZE + 2*LZ4HC_MAXD + 56 + ((sizeof(void*)==16) ? 56 : 0) /* AS400*/ ) /* 262200 or 262256*/ -#define LZ4_STREAMHCSIZE_SIZET (LZ4_STREAMHCSIZE / sizeof(size_t)) +#define LZ4_STREAMHCSIZE 262200 /* static size, for inter-version compatibility */ +#define LZ4_STREAMHCSIZE_VOIDP (LZ4_STREAMHCSIZE / sizeof(void*)) union LZ4_streamHC_u { - size_t table[LZ4_STREAMHCSIZE_SIZET]; + void* table[LZ4_STREAMHCSIZE_VOIDP]; LZ4HC_CCtx_internal internal_donotuse; }; /* previously typedef'd to LZ4_streamHC_t */ diff --git a/CPP/7zip/Bundles/Codec_lz4/resource.rc b/CPP/7zip/Bundles/Codec_lz4/resource.rc index 02dd1dc60..a9d43396b 100644 --- a/CPP/7zip/Bundles/Codec_lz4/resource.rc +++ b/CPP/7zip/Bundles/Codec_lz4/resource.rc @@ -1,6 +1,6 @@ #include "../../../../C/7zVersionTr.h" #include "../../../../C/7zVersion.rc" -MY_VERSION_INFO_DLL("7-Zip LZ4 Plugin v1.9.2", "Lz4") +MY_VERSION_INFO_DLL("7-Zip LZ4 Plugin v1.9.3", "Lz4") 101 ICON "../../Archive/Icons/lz4.ico" diff --git a/README.md b/README.md index a3aed16cd..b7abf1e4a 100644 --- a/README.md +++ b/README.md @@ -21,7 +21,7 @@ You can install it in two ways: 2. [Brotli] v.1.0.9 is a generic-purpose lossless compression algorithm that compresses data using a combination of a modern variant of the LZ77 algorithm, Huffman coding and 2nd order context modeling, with a compression ratio comparable to the best currently available general-purpose compression methods. It is similar in speed with deflate but offers more dense compression. - Levels: 0..11 -3. [LZ4] v1.9.2 is lossless compression algorithm, providing compression speed at 400 MB/s per core (0.16 Bytes/cycle). It features an extremely fast decoder, with speed in multiple GB/s per core (0.71 Bytes/cycle). A high compression derivative, called LZ4_HC, is available, trading customizable CPU time for compression ratio. +3. [LZ4] v1.9.3 is lossless compression algorithm, providing compression speed at 400 MB/s per core (0.16 Bytes/cycle). It features an extremely fast decoder, with speed in multiple GB/s per core (0.71 Bytes/cycle). A high compression derivative, called LZ4_HC, is available, trading customizable CPU time for compression ratio. - Levels: 1..12 4. [LZ5] v1.5 is a modification of LZ4 which gives a better ratio at cost of slower compression and decompression. @@ -280,7 +280,7 @@ You find this project useful, maybe you consider a donation ;-) - [Brotli] Version 1.0.9 - [Fast LZMA2] Version 1.0.1 - [Lizard] Version 1.0 - - [LZ4] Version 1.9.2 + - [LZ4] Version 1.9.3 - [LZ5] Version 1.5 - [Zstandard] Version 1.4.5