AuroraMiddleware/lz4
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

lz4 & lz4hc : added capability to allocate state & stream state with custom allocator (issue 99)

Browse Source 
fuzzer & fullbench : updated to test new functions
man : documented -l command (Legacy format, for Linux kernel compression) (issue 102)
cmake : improved version by Mika Attila, building programs and libraries (issue 100)
xxHash : updated to r33
Makefile : clean also delete local package .tar.gz


git-svn-id: https://lz4.googlecode.com/svn/trunk@110 650e7d94-2a16-8b24-b05c-7c0b3f6821cd
This commit is contained in:
yann.collet.73@gmail.com 2013-12-30 17:16:52 +00:00
parent 8ac549f10a
commit fb38ddaacb
15 changed files with 5711 additions and 5488 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
Makefile
View File

@ -30,14 +30,14 @@
# fullbench32: Same as fullbench, but forced to compile in 32-bits mode
# ################################################################
RELEASE=r109
RELEASE=r110
DESTDIR=
PREFIX=${DESTDIR}/usr
BINDIR=$(PREFIX)/bin
MANDIR=$(PREFIX)/share/man/man1
DISTRIBNAME=lz4-$(RELEASE).tar.gz
CC=gcc
CFLAGS=-I. -std=c99 -Wall -W -Wundef -DLZ4_VERSION=\"v1.0.9\"
CFLAGS=-I. -std=c99 -Wall -W -Wundef -DLZ4_VERSION=\"$(RELEASE)\"
# Define *.exe as extension for Windows systems
# ifeq ($(OS),Windows_NT)
@ -85,7 +85,8 @@ fullbench32: lz4.c lz4hc.c xxhash.c fullbench.c
clean:
@rm -f core *.o lz4$(EXT) lz4c$(EXT) lz4c32$(EXT) \
fuzzer$(EXT) fuzzer32$(EXT) fullbench$(EXT) fullbench32$(EXT)
fuzzer$(EXT) fuzzer32$(EXT) fullbench$(EXT) fullbench32$(EXT) \
$(DISTRIBNAME)
@echo Cleaning completed

68
NEWS
View File

@ -1,30 +1,38 @@
r109 :
lz4.c : corrected issue 98 (LZ4_compress_limitedOutput())
Makefile : can specify version number from makefile
r108 :
lz4.c : corrected compression efficiency issue 97 in 64-bits chained mode (-BD) for streams > 4 GB (thanks Roman Strashkin for reporting)
r107 :
Makefile : support DESTDIR for staged installs. Thanks Jorge Aparicio.
Makefile : make install installs both lz4 and lz4c (Jorge Aparicio)
Makefile : removed -Wno-implicit-declaration compilation switch
lz4cli.c : include <stduni.h> for isatty() (Luca Barbato)
lz4.h : introduced LZ4_MAX_INPUT_SIZE constant (Shay Green)
lz4.h : LZ4_compressBound() : unified macro and inline definitions (Shay Green)
lz4.h : LZ4_decompressSafe_partial() : clarify comments (Shay Green)
lz4.c : LZ4_compress() verify input size condition (Shay Green)
bench.c : corrected a bug in free memory size evaluation
cmake : install into bin/ directory (Richard Yao)
cmake : check for just C compiler (Elan Ruusamae)
r106 :
Makefile : make dist modify text files in the package to respect Unix EoL convention
lz4cli.c : corrected small display bug in HC mode
r105 :
Makefile : New install script and man page, contributed by Prasad Pandit
lz4cli.c : Minor modifications, for easier extensibility
COPYING : added license file
LZ4_Streaming_Format.odt : modified file name to remove white space characters
Makefile : .exe suffix now properly added only for Windows target
r110 :
lz4 & lz4hc : added capability to allocate state & stream state with custom allocator (issue 99)
fuzzer & fullbench : updated to test new functions
man : documented -l command (Legacy format, for Linux kernel compression) (issue 102)
cmake : improved version by Mika Attila, building programs and libraries (issue 100)
xxHash : updated to r33
Makefile : clean also delete local package .tar.gz
r109 :
lz4.c : corrected issue 98 (LZ4_compress_limitedOutput())
Makefile : can specify version number from makefile
r108 :
lz4.c : corrected compression efficiency issue 97 in 64-bits chained mode (-BD) for streams > 4 GB (thanks Roman Strashkin for reporting)
r107 :
Makefile : support DESTDIR for staged installs. Thanks Jorge Aparicio.
Makefile : make install installs both lz4 and lz4c (Jorge Aparicio)
Makefile : removed -Wno-implicit-declaration compilation switch
lz4cli.c : include <stduni.h> for isatty() (Luca Barbato)
lz4.h : introduced LZ4_MAX_INPUT_SIZE constant (Shay Green)
lz4.h : LZ4_compressBound() : unified macro and inline definitions (Shay Green)
lz4.h : LZ4_decompressSafe_partial() : clarify comments (Shay Green)
lz4.c : LZ4_compress() verify input size condition (Shay Green)
bench.c : corrected a bug in free memory size evaluation
cmake : install into bin/ directory (Richard Yao)
cmake : check for just C compiler (Elan Ruusamae)
r106 :
Makefile : make dist modify text files in the package to respect Unix EoL convention
lz4cli.c : corrected small display bug in HC mode
r105 :
Makefile : New install script and man page, contributed by Prasad Pandit
lz4cli.c : Minor modifications, for easier extensibility
COPYING : added license file
LZ4_Streaming_Format.odt : modified file name to remove white space characters
Makefile : .exe suffix now properly added only for Windows target

882
bench.c
View File

@ -1,441 +1,441 @@
/*
bench.c - Demo program to benchmark open-source compression algorithm
Copyright (C) Yann Collet 2012-2013
GPL v2 License
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
//**************************************
// Compiler Options
//**************************************
// Disable some Visual warning messages
#define _CRT_SECURE_NO_WARNINGS
#define _CRT_SECURE_NO_DEPRECATE // VS2005
// Unix Large Files support (>4GB)
#define _FILE_OFFSET_BITS 64
#if (defined(__sun__) && (!defined(__LP64__))) // Sun Solaris 32-bits requires specific definitions
# define _LARGEFILE_SOURCE
#elif ! defined(__LP64__) // No point defining Large file for 64 bit
# define _LARGEFILE64_SOURCE
#endif
// S_ISREG & gettimeofday() are not supported by MSVC
#if defined(_MSC_VER) || defined(_WIN32)
# define BMK_LEGACY_TIMER 1
#endif
//**************************************
// Includes
//**************************************
#include <stdlib.h> // malloc
#include <stdio.h> // fprintf, fopen, ftello64
#include <sys/types.h> // stat64
#include <sys/stat.h> // stat64
// Use ftime() if gettimeofday() is not available on your target
#if defined(BMK_LEGACY_TIMER)
# include <sys/timeb.h> // timeb, ftime
#else
# include <sys/time.h> // gettimeofday
#endif
#include "lz4.h"
#define COMPRESSOR0 LZ4_compress
#include "lz4hc.h"
#define COMPRESSOR1 LZ4_compressHC
#define DEFAULTCOMPRESSOR COMPRESSOR0
#include "xxhash.h"
//**************************************
// Compiler specifics
//**************************************
#if !defined(S_ISREG)
# define S_ISREG(x) (((x) & S_IFMT) == S_IFREG)
#endif
// GCC does not support _rotl outside of Windows
#if !defined(_WIN32)
# define _rotl(x,r) ((x << r) | (x >> (32 - r)))
#endif
//**************************************
// Basic Types
//**************************************
#if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L // C99
# include <stdint.h>
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef uint32_t U32;
typedef int32_t S32;
typedef uint64_t U64;
#else
typedef unsigned char BYTE;
typedef unsigned short U16;
typedef unsigned int U32;
typedef signed int S32;
typedef unsigned long long U64;
#endif
//**************************************
// Constants
//**************************************
#define NBLOOPS 3
#define TIMELOOP 2000
#define KB *(1U<<10)
#define MB *(1U<<20)
#define GB *(1U<<30)
#define KNUTH 2654435761U
#define MAX_MEM (2 GB - 64 MB)
#define DEFAULT_CHUNKSIZE (4 MB)
//**************************************
// Local structures
//**************************************
struct chunkParameters
{
U32 id;
char* origBuffer;
char* compressedBuffer;
int origSize;
int compressedSize;
};
struct compressionParameters
{
int (*compressionFunction)(const char*, char*, int);
int (*decompressionFunction)(const char*, char*, int);
};
//**************************************
// MACRO
//**************************************
#define DISPLAY(...) fprintf(stderr, __VA_ARGS__)
//**************************************
// Benchmark Parameters
//**************************************
static int chunkSize = DEFAULT_CHUNKSIZE;
static int nbIterations = NBLOOPS;
static int BMK_pause = 0;
void BMK_SetBlocksize(int bsize) { chunkSize = bsize; }
void BMK_SetNbIterations(int nbLoops)
{
nbIterations = nbLoops;
DISPLAY("- %i iterations -\n", nbIterations);
}
void BMK_SetPause() { BMK_pause = 1; }
//*********************************************************
// Private functions
//*********************************************************
#if defined(BMK_LEGACY_TIMER)
static int BMK_GetMilliStart()
{
// Based on Legacy ftime()
// Rolls over every ~ 12.1 days (0x100000/24/60/60)
// Use GetMilliSpan to correct for rollover
struct timeb tb;
int nCount;
ftime( &tb );
nCount = (int) (tb.millitm + (tb.time & 0xfffff) * 1000);
return nCount;
}
#else
static int BMK_GetMilliStart()
{
// Based on newer gettimeofday()
// Use GetMilliSpan to correct for rollover
struct timeval tv;
int nCount;
gettimeofday(&tv, NULL);
nCount = (int) (tv.tv_usec/1000 + (tv.tv_sec & 0xfffff) * 1000);
return nCount;
}
#endif
static int BMK_GetMilliSpan( int nTimeStart )
{
int nSpan = BMK_GetMilliStart() - nTimeStart;
if ( nSpan < 0 )
nSpan += 0x100000 * 1000;
return nSpan;
}
static size_t BMK_findMaxMem(U64 requiredMem)
{
size_t step = (64 MB);
BYTE* testmem=NULL;
requiredMem = (((requiredMem >> 26) + 1) << 26);
requiredMem += 2*step;
if (requiredMem > MAX_MEM) requiredMem = MAX_MEM;
while (!testmem)
{
requiredMem -= step;
testmem = (BYTE*) malloc ((size_t)requiredMem);
}
free (testmem);
return (size_t) (requiredMem - step);
}
static U64 BMK_GetFileSize(char* infilename)
{
int r;
#if defined(_MSC_VER)
struct _stat64 statbuf;
r = _stat64(infilename, &statbuf);
#else
struct stat statbuf;
r = stat(infilename, &statbuf);
#endif
if (r || !S_ISREG(statbuf.st_mode)) return 0; // No good...
return (U64)statbuf.st_size;
}
//*********************************************************
// Public function
//*********************************************************
int BMK_benchFile(char** fileNamesTable, int nbFiles, int cLevel)
{
int fileIdx=0;
char* orig_buff;
struct compressionParameters compP;
int cfunctionId;
U64 totals = 0;
U64 totalz = 0;
double totalc = 0.;
double totald = 0.;
// Init
if (cLevel <3) cfunctionId = 0; else cfunctionId = 1;
switch (cfunctionId)
{
#ifdef COMPRESSOR0
case 0 : compP.compressionFunction = COMPRESSOR0; break;
#endif
#ifdef COMPRESSOR1
case 1 : compP.compressionFunction = COMPRESSOR1; break;
#endif
default : compP.compressionFunction = DEFAULTCOMPRESSOR;
}
compP.decompressionFunction = LZ4_decompress_fast;
// Loop for each file
while (fileIdx<nbFiles)
{
FILE* inFile;
char* inFileName;
U64 inFileSize;
size_t benchedSize;
int nbChunks;
int maxCompressedChunkSize;
size_t readSize;
char* compressedBuffer; int compressedBuffSize;
struct chunkParameters* chunkP;
U32 crcOrig;
// Check file existence
inFileName = fileNamesTable[fileIdx++];
inFile = fopen( inFileName, "rb" );
if (inFile==NULL)
{
DISPLAY( "Pb opening %s\n", inFileName);
return 11;
}
// Memory allocation & restrictions
inFileSize = BMK_GetFileSize(inFileName);
benchedSize = (size_t) BMK_findMaxMem(inFileSize * 2) / 2;
if ((U64)benchedSize > inFileSize) benchedSize = (size_t)inFileSize;
if (benchedSize < inFileSize)
{
DISPLAY("Not enough memory for '%s' full size; testing %i MB only...\n", inFileName, (int)(benchedSize>>20));
}
// Alloc
chunkP = (struct chunkParameters*) malloc(((benchedSize / chunkSize)+1) * sizeof(struct chunkParameters));
orig_buff = (char*)malloc((size_t )benchedSize);
nbChunks = (int) (benchedSize / chunkSize) + 1;
maxCompressedChunkSize = LZ4_compressBound(chunkSize);
compressedBuffSize = nbChunks * maxCompressedChunkSize;
compressedBuffer = (char*)malloc((size_t )compressedBuffSize);
if (!orig_buff || !compressedBuffer)
{
DISPLAY("\nError: not enough memory!\n");
free(orig_buff);
free(compressedBuffer);
free(chunkP);
fclose(inFile);
return 12;
}
// Init chunks data
{
int i;
size_t remaining = benchedSize;
char* in = orig_buff;
char* out = compressedBuffer;
for (i=0; i<nbChunks; i++)
{
chunkP[i].id = i;
chunkP[i].origBuffer = in; in += chunkSize;
if ((int)remaining > chunkSize) { chunkP[i].origSize = chunkSize; remaining -= chunkSize; } else { chunkP[i].origSize = (int)remaining; remaining = 0; }
chunkP[i].compressedBuffer = out; out += maxCompressedChunkSize;
chunkP[i].compressedSize = 0;
}
}
// Fill input buffer
DISPLAY("Loading %s... \r", inFileName);
readSize = fread(orig_buff, 1, benchedSize, inFile);
fclose(inFile);
if (readSize != benchedSize)
{
DISPLAY("\nError: problem reading file '%s' !! \n", inFileName);
free(orig_buff);
free(compressedBuffer);
free(chunkP);
return 13;
}
// Calculating input Checksum
crcOrig = XXH32(orig_buff, (unsigned int)benchedSize,0);
// Bench
{
int loopNb, chunkNb;
size_t cSize=0;
double fastestC = 100000000., fastestD = 100000000.;
double ratio=0.;
U32 crcCheck=0;
DISPLAY("\r%79s\r", "");
for (loopNb = 1; loopNb <= nbIterations; loopNb++)
{
int nbLoops;
int milliTime;
// Compression
DISPLAY("%1i-%-14.14s : %9i ->\r", loopNb, inFileName, (int)benchedSize);
{ size_t i; for (i=0; i<benchedSize; i++) compressedBuffer[i]=(char)i; } // warmimg up memory
nbLoops = 0;
milliTime = BMK_GetMilliStart();
while(BMK_GetMilliStart() == milliTime);
milliTime = BMK_GetMilliStart();
while(BMK_GetMilliSpan(milliTime) < TIMELOOP)
{
for (chunkNb=0; chunkNb<nbChunks; chunkNb++)
chunkP[chunkNb].compressedSize = compP.compressionFunction(chunkP[chunkNb].origBuffer, chunkP[chunkNb].compressedBuffer, chunkP[chunkNb].origSize);
nbLoops++;
}
milliTime = BMK_GetMilliSpan(milliTime);
if ((double)milliTime < fastestC*nbLoops) fastestC = (double)milliTime/nbLoops;
cSize=0; for (chunkNb=0; chunkNb<nbChunks; chunkNb++) cSize += chunkP[chunkNb].compressedSize;
ratio = (double)cSize/(double)benchedSize*100.;
DISPLAY("%1i-%-14.14s : %9i -> %9i (%5.2f%%),%7.1f MB/s\r", loopNb, inFileName, (int)benchedSize, (int)cSize, ratio, (double)benchedSize / fastestC / 1000.);
// Decompression
{ size_t i; for (i=0; i<benchedSize; i++) orig_buff[i]=0; } // zeroing area, for CRC checking
nbLoops = 0;
milliTime = BMK_GetMilliStart();
while(BMK_GetMilliStart() == milliTime);
milliTime = BMK_GetMilliStart();
while(BMK_GetMilliSpan(milliTime) < TIMELOOP)
{
for (chunkNb=0; chunkNb<nbChunks; chunkNb++)
chunkP[chunkNb].compressedSize = LZ4_decompress_fast(chunkP[chunkNb].compressedBuffer, chunkP[chunkNb].origBuffer, chunkP[chunkNb].origSize);
nbLoops++;
}
milliTime = BMK_GetMilliSpan(milliTime);
if ((double)milliTime < fastestD*nbLoops) fastestD = (double)milliTime/nbLoops;
DISPLAY("%1i-%-14.14s : %9i -> %9i (%5.2f%%),%7.1f MB/s ,%7.1f MB/s\r", loopNb, inFileName, (int)benchedSize, (int)cSize, ratio, (double)benchedSize / fastestC / 1000., (double)benchedSize / fastestD / 1000.);
// CRC Checking
crcCheck = XXH32(orig_buff, (unsigned int)benchedSize,0);
if (crcOrig!=crcCheck) { DISPLAY("\n!!! WARNING !!! %14s : Invalid Checksum : %x != %x\n", inFileName, (unsigned)crcOrig, (unsigned)crcCheck); break; }
}
if (crcOrig==crcCheck)
{
if (ratio<100.)
DISPLAY("%-16.16s : %9i -> %9i (%5.2f%%),%7.1f MB/s ,%7.1f MB/s\n", inFileName, (int)benchedSize, (int)cSize, ratio, (double)benchedSize / fastestC / 1000., (double)benchedSize / fastestD / 1000.);
else
DISPLAY("%-16.16s : %9i -> %9i (%5.1f%%),%7.1f MB/s ,%7.1f MB/s \n", inFileName, (int)benchedSize, (int)cSize, ratio, (double)benchedSize / fastestC / 1000., (double)benchedSize / fastestD / 1000.);
}
totals += benchedSize;
totalz += cSize;
totalc += fastestC;
totald += fastestD;
}
free(orig_buff);
free(compressedBuffer);
free(chunkP);
}
if (nbFiles > 1)
DISPLAY("%-16.16s :%10llu ->%10llu (%5.2f%%), %6.1f MB/s , %6.1f MB/s\n", " TOTAL", (long long unsigned int)totals, (long long unsigned int)totalz, (double)totalz/(double)totals*100., (double)totals/totalc/1000., (double)totals/totald/1000.);
if (BMK_pause) { DISPLAY("press enter...\n"); getchar(); }
return 0;
}
/*
bench.c - Demo program to benchmark open-source compression algorithm
Copyright (C) Yann Collet 2012-2013
GPL v2 License
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
//**************************************
// Compiler Options
//**************************************
// Disable some Visual warning messages
#define _CRT_SECURE_NO_WARNINGS
#define _CRT_SECURE_NO_DEPRECATE // VS2005
// Unix Large Files support (>4GB)
#define _FILE_OFFSET_BITS 64
#if (defined(__sun__) && (!defined(__LP64__))) // Sun Solaris 32-bits requires specific definitions
# define _LARGEFILE_SOURCE
#elif ! defined(__LP64__) // No point defining Large file for 64 bit
# define _LARGEFILE64_SOURCE
#endif
// S_ISREG & gettimeofday() are not supported by MSVC
#if defined(_MSC_VER) || defined(_WIN32)
# define BMK_LEGACY_TIMER 1
#endif
//**************************************
// Includes
//**************************************
#include <stdlib.h> // malloc
#include <stdio.h> // fprintf, fopen, ftello64
#include <sys/types.h> // stat64
#include <sys/stat.h> // stat64
// Use ftime() if gettimeofday() is not available on your target
#if defined(BMK_LEGACY_TIMER)
# include <sys/timeb.h> // timeb, ftime
#else
# include <sys/time.h> // gettimeofday
#endif
#include "lz4.h"
#define COMPRESSOR0 LZ4_compress
#include "lz4hc.h"
#define COMPRESSOR1 LZ4_compressHC
#define DEFAULTCOMPRESSOR COMPRESSOR0
#include "xxhash.h"
//**************************************
// Compiler specifics
//**************************************
#if !defined(S_ISREG)
# define S_ISREG(x) (((x) & S_IFMT) == S_IFREG)
#endif
// GCC does not support _rotl outside of Windows
#if !defined(_WIN32)
# define _rotl(x,r) ((x << r) | (x >> (32 - r)))
#endif
//**************************************
// Basic Types
//**************************************
#if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L // C99
# include <stdint.h>
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef uint32_t U32;
typedef int32_t S32;
typedef uint64_t U64;
#else
typedef unsigned char BYTE;
typedef unsigned short U16;
typedef unsigned int U32;
typedef signed int S32;
typedef unsigned long long U64;
#endif
//**************************************
// Constants
//**************************************
#define NBLOOPS 3
#define TIMELOOP 2000
#define KB *(1U<<10)
#define MB *(1U<<20)
#define GB *(1U<<30)
#define KNUTH 2654435761U
#define MAX_MEM (2 GB - 64 MB)
#define DEFAULT_CHUNKSIZE (4 MB)
//**************************************
// Local structures
//**************************************
struct chunkParameters
{
U32 id;
char* origBuffer;
char* compressedBuffer;
int origSize;
int compressedSize;
};
struct compressionParameters
{
int (*compressionFunction)(const char*, char*, int);
int (*decompressionFunction)(const char*, char*, int);
};
//**************************************
// MACRO
//**************************************
#define DISPLAY(...) fprintf(stderr, __VA_ARGS__)
//**************************************
// Benchmark Parameters
//**************************************
static int chunkSize = DEFAULT_CHUNKSIZE;
static int nbIterations = NBLOOPS;
static int BMK_pause = 0;
void BMK_SetBlocksize(int bsize) { chunkSize = bsize; }
void BMK_SetNbIterations(int nbLoops)
{
nbIterations = nbLoops;
DISPLAY("- %i iterations -\n", nbIterations);
}
void BMK_SetPause() { BMK_pause = 1; }
//*********************************************************
// Private functions
//*********************************************************
#if defined(BMK_LEGACY_TIMER)
static int BMK_GetMilliStart()
{
// Based on Legacy ftime()
// Rolls over every ~ 12.1 days (0x100000/24/60/60)
// Use GetMilliSpan to correct for rollover
struct timeb tb;
int nCount;
ftime( &tb );
nCount = (int) (tb.millitm + (tb.time & 0xfffff) * 1000);
return nCount;
}
#else
static int BMK_GetMilliStart()
{
// Based on newer gettimeofday()
// Use GetMilliSpan to correct for rollover
struct timeval tv;
int nCount;
gettimeofday(&tv, NULL);
nCount = (int) (tv.tv_usec/1000 + (tv.tv_sec & 0xfffff) * 1000);
return nCount;
}
#endif
static int BMK_GetMilliSpan( int nTimeStart )
{
int nSpan = BMK_GetMilliStart() - nTimeStart;
if ( nSpan < 0 )
nSpan += 0x100000 * 1000;
return nSpan;
}
static size_t BMK_findMaxMem(U64 requiredMem)
{
size_t step = (64 MB);
BYTE* testmem=NULL;
requiredMem = (((requiredMem >> 26) + 1) << 26);
requiredMem += 2*step;
if (requiredMem > MAX_MEM) requiredMem = MAX_MEM;
while (!testmem)
{
requiredMem -= step;
testmem = (BYTE*) malloc ((size_t)requiredMem);
}
free (testmem);
return (size_t) (requiredMem - step);
}
static U64 BMK_GetFileSize(char* infilename)
{
int r;
#if defined(_MSC_VER)
struct _stat64 statbuf;
r = _stat64(infilename, &statbuf);
#else
struct stat statbuf;
r = stat(infilename, &statbuf);
#endif
if (r || !S_ISREG(statbuf.st_mode)) return 0; // No good...
return (U64)statbuf.st_size;
}
//*********************************************************
// Public function
//*********************************************************
int BMK_benchFile(char** fileNamesTable, int nbFiles, int cLevel)
{
int fileIdx=0;
char* orig_buff;
struct compressionParameters compP;
int cfunctionId;
U64 totals = 0;
U64 totalz = 0;
double totalc = 0.;
double totald = 0.;
// Init
if (cLevel <3) cfunctionId = 0; else cfunctionId = 1;
switch (cfunctionId)
{
#ifdef COMPRESSOR0
case 0 : compP.compressionFunction = COMPRESSOR0; break;
#endif
#ifdef COMPRESSOR1
case 1 : compP.compressionFunction = COMPRESSOR1; break;
#endif
default : compP.compressionFunction = DEFAULTCOMPRESSOR;
}
compP.decompressionFunction = LZ4_decompress_fast;
// Loop for each file
while (fileIdx<nbFiles)
{
FILE* inFile;
char* inFileName;
U64 inFileSize;
size_t benchedSize;
int nbChunks;
int maxCompressedChunkSize;
size_t readSize;
char* compressedBuffer; int compressedBuffSize;
struct chunkParameters* chunkP;
U32 crcOrig;
// Check file existence
inFileName = fileNamesTable[fileIdx++];
inFile = fopen( inFileName, "rb" );
if (inFile==NULL)
{
DISPLAY( "Pb opening %s\n", inFileName);
return 11;
}
// Memory allocation & restrictions
inFileSize = BMK_GetFileSize(inFileName);
benchedSize = (size_t) BMK_findMaxMem(inFileSize * 2) / 2;
if ((U64)benchedSize > inFileSize) benchedSize = (size_t)inFileSize;
if (benchedSize < inFileSize)
{
DISPLAY("Not enough memory for '%s' full size; testing %i MB only...\n", inFileName, (int)(benchedSize>>20));
}
// Alloc
chunkP = (struct chunkParameters*) malloc(((benchedSize / chunkSize)+1) * sizeof(struct chunkParameters));
orig_buff = (char*)malloc((size_t )benchedSize);
nbChunks = (int) (benchedSize / chunkSize) + 1;
maxCompressedChunkSize = LZ4_compressBound(chunkSize);
compressedBuffSize = nbChunks * maxCompressedChunkSize;
compressedBuffer = (char*)malloc((size_t )compressedBuffSize);
if (!orig_buff || !compressedBuffer)
{
DISPLAY("\nError: not enough memory!\n");
free(orig_buff);
free(compressedBuffer);
free(chunkP);
fclose(inFile);
return 12;
}
// Init chunks data
{
int i;
size_t remaining = benchedSize;
char* in = orig_buff;
char* out = compressedBuffer;
for (i=0; i<nbChunks; i++)
{
chunkP[i].id = i;
chunkP[i].origBuffer = in; in += chunkSize;
if ((int)remaining > chunkSize) { chunkP[i].origSize = chunkSize; remaining -= chunkSize; } else { chunkP[i].origSize = (int)remaining; remaining = 0; }
chunkP[i].compressedBuffer = out; out += maxCompressedChunkSize;
chunkP[i].compressedSize = 0;
}
}
// Fill input buffer
DISPLAY("Loading %s... \r", inFileName);
readSize = fread(orig_buff, 1, benchedSize, inFile);
fclose(inFile);
if (readSize != benchedSize)
{
DISPLAY("\nError: problem reading file '%s' !! \n", inFileName);
free(orig_buff);
free(compressedBuffer);
free(chunkP);
return 13;
}
// Calculating input Checksum
crcOrig = XXH32(orig_buff, (unsigned int)benchedSize,0);
// Bench
{
int loopNb, chunkNb;
size_t cSize=0;
double fastestC = 100000000., fastestD = 100000000.;
double ratio=0.;
U32 crcCheck=0;
DISPLAY("\r%79s\r", "");
for (loopNb = 1; loopNb <= nbIterations; loopNb++)
{
int nbLoops;
int milliTime;
// Compression
DISPLAY("%1i-%-14.14s : %9i ->\r", loopNb, inFileName, (int)benchedSize);
{ size_t i; for (i=0; i<benchedSize; i++) compressedBuffer[i]=(char)i; } // warmimg up memory
nbLoops = 0;
milliTime = BMK_GetMilliStart();
while(BMK_GetMilliStart() == milliTime);
milliTime = BMK_GetMilliStart();
while(BMK_GetMilliSpan(milliTime) < TIMELOOP)
{
for (chunkNb=0; chunkNb<nbChunks; chunkNb++)
chunkP[chunkNb].compressedSize = compP.compressionFunction(chunkP[chunkNb].origBuffer, chunkP[chunkNb].compressedBuffer, chunkP[chunkNb].origSize);
nbLoops++;
}
milliTime = BMK_GetMilliSpan(milliTime);
if ((double)milliTime < fastestC*nbLoops) fastestC = (double)milliTime/nbLoops;
cSize=0; for (chunkNb=0; chunkNb<nbChunks; chunkNb++) cSize += chunkP[chunkNb].compressedSize;
ratio = (double)cSize/(double)benchedSize*100.;
DISPLAY("%1i-%-14.14s : %9i -> %9i (%5.2f%%),%7.1f MB/s\r", loopNb, inFileName, (int)benchedSize, (int)cSize, ratio, (double)benchedSize / fastestC / 1000.);
// Decompression
{ size_t i; for (i=0; i<benchedSize; i++) orig_buff[i]=0; } // zeroing area, for CRC checking
nbLoops = 0;
milliTime = BMK_GetMilliStart();
while(BMK_GetMilliStart() == milliTime);
milliTime = BMK_GetMilliStart();
while(BMK_GetMilliSpan(milliTime) < TIMELOOP)
{
for (chunkNb=0; chunkNb<nbChunks; chunkNb++)
chunkP[chunkNb].compressedSize = LZ4_decompress_fast(chunkP[chunkNb].compressedBuffer, chunkP[chunkNb].origBuffer, chunkP[chunkNb].origSize);
nbLoops++;
}
milliTime = BMK_GetMilliSpan(milliTime);
if ((double)milliTime < fastestD*nbLoops) fastestD = (double)milliTime/nbLoops;
DISPLAY("%1i-%-14.14s : %9i -> %9i (%5.2f%%),%7.1f MB/s ,%7.1f MB/s\r", loopNb, inFileName, (int)benchedSize, (int)cSize, ratio, (double)benchedSize / fastestC / 1000., (double)benchedSize / fastestD / 1000.);
// CRC Checking
crcCheck = XXH32(orig_buff, (unsigned int)benchedSize,0);
if (crcOrig!=crcCheck) { DISPLAY("\n!!! WARNING !!! %14s : Invalid Checksum : %x != %x\n", inFileName, (unsigned)crcOrig, (unsigned)crcCheck); break; }
}
if (crcOrig==crcCheck)
{
if (ratio<100.)
DISPLAY("%-16.16s : %9i -> %9i (%5.2f%%),%7.1f MB/s ,%7.1f MB/s\n", inFileName, (int)benchedSize, (int)cSize, ratio, (double)benchedSize / fastestC / 1000., (double)benchedSize / fastestD / 1000.);
else
DISPLAY("%-16.16s : %9i -> %9i (%5.1f%%),%7.1f MB/s ,%7.1f MB/s \n", inFileName, (int)benchedSize, (int)cSize, ratio, (double)benchedSize / fastestC / 1000., (double)benchedSize / fastestD / 1000.);
}
totals += benchedSize;
totalz += cSize;
totalc += fastestC;
totald += fastestD;
}
free(orig_buff);
free(compressedBuffer);
free(chunkP);
}
if (nbFiles > 1)
DISPLAY("%-16.16s :%10llu ->%10llu (%5.2f%%), %6.1f MB/s , %6.1f MB/s\n", " TOTAL", (long long unsigned int)totals, (long long unsigned int)totalz, (double)totalz/(double)totals*100., (double)totals/totalc/1000., (double)totals/totald/1000.);
if (BMK_pause) { DISPLAY("\npress enter...\n"); getchar(); }
return 0;
}

82
bench.h
View File

@ -1,41 +1,41 @@
/*
bench.h - Demo program to benchmark open-source compression algorithm
Copyright (C) Yann Collet 2012-2013
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
#pragma once
#if defined (__cplusplus)
extern "C" {
#endif
int BMK_benchFile(char** fileNamesTable, int nbFiles, int cLevel);
// Parameters
void BMK_SetBlocksize(int bsize);
void BMK_SetNbIterations(int nbLoops);
void BMK_SetPause();
#if defined (__cplusplus)
}
#endif
/*
bench.h - Demo program to benchmark open-source compression algorithm
Copyright (C) Yann Collet 2012-2013
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
#pragma once
#if defined (__cplusplus)
extern "C" {
#endif
int BMK_benchFile(char** fileNamesTable, int nbFiles, int cLevel);
// Parameters
void BMK_SetBlocksize(int bsize);
void BMK_SetNbIterations(int nbLoops);
void BMK_SetPause();
#if defined (__cplusplus)
}
#endif

92
cmake/CMakeLists.txt
View File

@ -2,12 +2,11 @@ PROJECT(LZ4 C)
set(CPACK_PACKAGE_DESCRIPTION_SUMMARY "LZ4 compression library")
set(CPACK_PACKAGE_VERSION_MAJOR 0)
set(CPACK_PACKAGE_VERSION_MINOR 0)
set(CPACK_PACKAGE_VERSION_PATCH r107)
set(CPACK_PACKAGE_VERSION_PATCH r110)
#set(CPACK_RESOURCE_FILE_LICENSE ${CMAKE_CURRENT_BINARY_DIR}/COPYING_LGPL)
set(VERSION_STRING " \"${CPACK_PACKAGE_VERSION_MAJOR}.${CPACK_PACKAGE_VERSION_MINOR}.${CPACK_PACKAGE_VERSION_PATCH}\" ")
include(CPack)
cmake_minimum_required (VERSION 2.6)
INCLUDE (CheckTypeSize)
check_type_size("void *" SIZEOF_VOID_P)
@ -16,15 +15,10 @@ IF( ${SIZEOF_VOID_P} STREQUAL "8" )
MESSAGE( STATUS "64 bit architecture detected size of void * is " ${SIZEOF_VOID_P})
ENDIF()
set(BUILD_SHARED_LIBS OFF CACHE BOOL "Build shared libraries")
option(BUILD_TOOLS "Build the command line tools" ON)
option(BUILD_LIBS "Build the libraries in addition to the tools" OFF)
if (BUILD_SHARED_LIBS)
message(STATUS "Enable shared library building")
else(BUILD_SHARED_LIBS)
message(STATUS "Disable shared library building")
endif(BUILD_SHARED_LIBS)
if(UNIX AND BUILD_SHARED_LIBS)
if(UNIX AND BUILD_LIBS)
if(CMAKE_SYSTEM_PROCESSOR MATCHES "x86_64")
add_definitions(-fPIC)
endif(CMAKE_SYSTEM_PROCESSOR MATCHES "x86_64")
@ -34,62 +28,42 @@ set(SRC_DIR ../)
set(LZ4_SRCS_LIB ${SRC_DIR}lz4.c ${SRC_DIR}lz4hc.c ${SRC_DIR}lz4.h )
set(LZ4_SRCS ${SRC_DIR}xxhash.c ${SRC_DIR}bench.c ${SRC_DIR}lz4cli.c)
if(NOT BUILD_SHARED_LIBS)
if(BUILD_TOOLS AND NOT BUILD_LIBS)
set(LZ4_SRCS ${LZ4_SRCS} ${LZ4_SRCS_LIB})
endif()
if (CMAKE_SYSTEM_PROCESSOR STREQUAL "64bit")
message(STATUS "Build 64bit executable binary")
add_executable(lz4c64 ${LZ4_SRCS})
install(TARGETS lz4c64 RUNTIME DESTINATION "bin/")
if(NOT BUILD_SHARED_LIBS)
message(STATUS "Build 32bit executable binary")
add_executable(lz4c32 ${LZ4_SRCS})
install(TARGETS lz4c32 RUNTIME DESTINATION "bin/")
SET_TARGET_PROPERTIES(lz4c32 PROPERTIES
COMPILE_FLAGS PROPERTIES COMPILE_FLAGS "-m32" LINK_FLAGS "-m32")
endif()
else()
message(STATUS "Build 32bit executable binary")
add_executable(lz4c32 ${LZ4_SRCS})
install(TARGETS lz4c32 RUNTIME DESTINATION "bin/")
if(BUILD_TOOLS)
add_executable(lz4 ${LZ4_SRCS})
set_target_properties(lz4 PROPERTIES COMPILE_DEFINITIONS DISABLE_LZ4C_LEGACY_OPTIONS)
install(TARGETS lz4 RUNTIME DESTINATION "bin/")
add_executable(lz4c ${LZ4_SRCS})
install(TARGETS lz4c RUNTIME DESTINATION "bin/")
endif()
if(BUILD_SHARED_LIBS)
set(LZ4_SOVERSION "${CPACK_PACKAGE_VERSION_MAJOR}.${CPACK_PACKAGE_VERSION_MINOR}")
if (CMAKE_SYSTEM_PROCESSOR STREQUAL "64bit")
target_link_libraries(lz4c64 liblz4)
else()
target_link_libraries(lz4c32 liblz4)
endif()
# for static library
add_library(liblz4_static STATIC ${LZ4_SRCS_LIB})
set_target_properties(liblz4_static PROPERTIES OUTPUT_NAME lz4)
install(TARGETS liblz4_static
LIBRARY DESTINATION lib
ARCHIVE DESTINATION lib
if(BUILD_LIBS)
add_library(liblz4 ${LZ4_SRCS_LIB})
set_target_properties(liblz4 PROPERTIES
OUTPUT_NAME lz4
SOVERSION "${CPACK_PACKAGE_VERSION_MAJOR}.${CPACK_PACKAGE_VERSION_MINOR}"
)
install(TARGETS liblz4
LIBRARY DESTINATION lib
ARCHIVE DESTINATION lib
)
# for shared library o
add_library(liblz4 ${LZ4_SRCS_LIB})
set_target_properties(liblz4 PROPERTIES
OUTPUT_NAME lz4
SOVERSION ${LZ4_SOVERSION})
install(TARGETS liblz4
LIBRARY DESTINATION lib
ARCHIVE DESTINATION lib
install(FILES
${SRC_DIR}/lz4.h
${SRC_DIR}/lz4hc.h
DESTINATION include
)
install(FILES
${SRC_DIR}/lz4.h
${SRC_DIR}/lz4hc.h
DESTINATION include)
endif(BUILD_SHARED_LIBS)
if(BUILD_TOOLS)
target_link_libraries(lz4 liblz4)
target_link_libraries(lz4c liblz4)
endif()
endif()
#warnings
@ -98,8 +72,8 @@ ADD_DEFINITIONS("-Wall")
ADD_DEFINITIONS("-W")
ADD_DEFINITIONS("-Wundef")
ADD_DEFINITIONS("-Wcast-align")
ADD_DEFINITIONS("-Wno-implicit-function-declaration")
ADD_DEFINITIONS("-Os -march=native -std=c99")
ADD_DEFINITIONS("-std=c99")
ADD_DEFINITIONS("-DLZ4_VERSION=\"${CPACK_PACKAGE_VERSION_PATCH}\"")
INCLUDE_DIRECTORIES (${SRC_DIR})

1448
fullbench.c
View File

File diff suppressed because it is too large Load Diff

583
fuzzer.c
View File

@ -1,280 +1,303 @@
/*
fuzzer.c - Fuzzer test tool for LZ4
Copyright (C) Yann Collet - Andrew Mahone 2012-2013
Code started by Andrew Mahone, modified by Yann Collet
GPL v2 License
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
//**************************************
// Remove Visual warning messages
//**************************************
#define _CRT_SECURE_NO_WARNINGS // fgets
//**************************************
// Includes
//**************************************
#include <stdlib.h>
#include <stdio.h> // fgets, sscanf
#include <sys/timeb.h> // timeb
#include "lz4.h"
#include "lz4hc.h"
//**************************************
// Constants
//**************************************
#ifndef LZ4_VERSION
# define LZ4_VERSION ""
#endif
#define NB_ATTEMPTS (1<<17)
#define LEN ((1<<15))
#define SEQ_POW 2
#define NUM_SEQ (1 << SEQ_POW)
#define SEQ_MSK ((NUM_SEQ) - 1)
#define MOD_SEQ(x) ((((x) >> 8) & 255) == 0)
#define NEW_SEQ(x) ((((x) >> 10) %10) == 0)
#define PAGE_SIZE 4096
#define ROUND_PAGE(x) (((x) + PAGE_SIZE - 1) & ~(PAGE_SIZE - 1))
#define PRIME1 2654435761U
#define PRIME2 2246822519U
#define PRIME3 3266489917U
//*********************************************************
// Functions
//*********************************************************
static int FUZ_GetMilliStart()
{
struct timeb tb;
int nCount;
ftime( &tb );
nCount = (int) (tb.millitm + (tb.time & 0xfffff) * 1000);
return nCount;
}
static int FUZ_GetMilliSpan( int nTimeStart )
{
int nSpan = FUZ_GetMilliStart() - nTimeStart;
if ( nSpan < 0 )
nSpan += 0x100000 * 1000;
return nSpan;
}
unsigned int FUZ_rand(unsigned int* src)
{
*src = ((*src) * PRIME1) + PRIME2;
return *src;
}
int test_canary(unsigned char *buf)
{
int i;
for (i = 0; i < 2048; i++)
if (buf[i] != buf[i + 2048])
return 0;
return 1;
}
int FUZ_SecurityTest()
{
char* output;
char* input;
int i, r;
printf("Overflow test (issue 52)...\n");
input = (char*) malloc (20<<20);
output = (char*) malloc (20<<20);
input[0] = 0x0F;
input[1] = 0x00;
input[2] = 0x00;
for(i = 3; i < 16840000; i++)
input[i] = 0xff;
r = LZ4_decompress_fast(input, output, 20<<20);
free(input);
free(output);
printf(" Passed (return = %i < 0)\n",r);
return 0;
}
//int main(int argc, char *argv[]) {
int main() {
unsigned long long bytes = 0;
unsigned long long cbytes = 0;
unsigned long long hcbytes = 0;
unsigned char buf[LEN];
unsigned char testOut[LEN+1];
# define FUZ_max LZ4_COMPRESSBOUND(LEN)
# define FUZ_avail ROUND_PAGE(FUZ_max)
const int off_full = FUZ_avail - FUZ_max;
unsigned char cbuf[FUZ_avail + PAGE_SIZE];
unsigned int seed, randState, cur_seq=PRIME3, seeds[NUM_SEQ], timestamp=FUZ_GetMilliStart();
int i, j, k, ret, len, lenHC, attemptNb;
char userInput[30] = {0};
# define FUZ_CHECKTEST(cond, message) if (cond) { printf("Test %i : %s : seed %u, cycle %i \n", testNb, message, seed, attemptNb); goto _output_error; }
# define FUZ_DISPLAYTEST testNb++; printf("%2i\b\b", testNb);
printf("starting LZ4 fuzzer (%s)\n", LZ4_VERSION);
printf("Select an Initialisation number (default : random) : ");
fflush(stdout);
if ( fgets(userInput, sizeof userInput, stdin) )
{
if ( sscanf(userInput, "%d", &seed) == 1 ) {}
else seed = FUZ_GetMilliSpan(timestamp);
}
printf("Seed = %u\n", seed);
randState = seed;
//FUZ_SecurityTest();
for (i = 0; i < 2048; i++)
cbuf[FUZ_avail + i] = cbuf[FUZ_avail + 2048 + i] = FUZ_rand(&randState) >> 16;
for (attemptNb = 0; attemptNb < NB_ATTEMPTS; attemptNb++)
{
int testNb = 0;
printf("\r%7i /%7i - ", attemptNb, NB_ATTEMPTS);
for (j = 0; j < NUM_SEQ; j++) {
seeds[j] = FUZ_rand(&randState) << 8;
seeds[j] ^= (FUZ_rand(&randState) >> 8) & 65535;
}
for (j = 0; j < LEN; j++) {
k = FUZ_rand(&randState);
if (j == 0 || NEW_SEQ(k))
cur_seq = seeds[(FUZ_rand(&randState) >> 16) & SEQ_MSK];
if (MOD_SEQ(k)) {
k = (FUZ_rand(&randState) >> 16) & SEQ_MSK;
seeds[k] = FUZ_rand(&randState) << 8;
seeds[k] ^= (FUZ_rand(&randState) >> 8) & 65535;
}
buf[j] = FUZ_rand(&cur_seq) >> 16;
}
// Test compression HC
FUZ_DISPLAYTEST; // 1
ret = LZ4_compressHC_limitedOutput((const char*)buf, (char*)&cbuf[off_full], LEN, FUZ_max);
FUZ_CHECKTEST(ret==0, "HC compression failed despite sufficient space");
lenHC = ret;
// Test compression
FUZ_DISPLAYTEST; // 2
ret = LZ4_compress_limitedOutput((const char*)buf, (char*)&cbuf[off_full], LEN, FUZ_max);
FUZ_CHECKTEST(ret==0, "compression failed despite sufficient space");
len = ret;
// Test decoding with output size being exactly what's necessary => must work
FUZ_DISPLAYTEST; // 3
ret = LZ4_decompress_fast((char*)&cbuf[off_full], (char*)testOut, LEN);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_fast failed despite correct space");
// Test decoding with one byte missing => must fail
FUZ_DISPLAYTEST; // 4
ret = LZ4_decompress_fast((char*)&cbuf[off_full], (char*)testOut, LEN-1);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_fast should have failed, due to Output Size being too small");
// Test decoding with one byte too much => must fail
FUZ_DISPLAYTEST;
ret = LZ4_decompress_fast((char*)&cbuf[off_full], (char*)testOut, LEN+1);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_fast should have failed, due to Output Size being too large");
// Test decoding with enough output size => must work
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len, LEN+1);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_safe failed despite sufficient space");
// Test decoding with output size being exactly what's necessary => must work
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len, LEN);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_safe failed despite sufficient space");
// Test decoding with output size being one byte too short => must fail
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len, LEN-1);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_safe should have failed, due to Output Size being one byte too short");
// Test decoding with input size being one byte too short => must fail
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len-1, LEN);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_safe should have failed, due to input size being one byte too short");
// Test decoding with input size being one byte too large => must fail
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len+1, LEN);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_safe should have failed, due to input size being too large");
//if (ret>=0) { printf("Test 10 : decompression should have failed, due to input size being too large : seed %u, len %d\n", seed, LEN); goto _output_error; }
// Test partial decoding with target output size being max/2 => must work
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe_partial((char*)&cbuf[off_full], (char*)testOut, len, LEN/2, LEN);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_safe_partial failed despite sufficient space");
// Test partial decoding with target output size being just below max => must work
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe_partial((char*)&cbuf[off_full], (char*)testOut, len, LEN-3, LEN);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_safe_partial failed despite sufficient space");
// Test compression with output size being exactly what's necessary (should work)
FUZ_DISPLAYTEST;
ret = LZ4_compress_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-len], LEN, len);
FUZ_CHECKTEST(!test_canary(&cbuf[FUZ_avail]), "compression overran output buffer");
FUZ_CHECKTEST(ret==0, "compression failed despite sufficient space");
// Test HC compression with output size being exactly what's necessary (should work)
FUZ_DISPLAYTEST;
ret = LZ4_compressHC_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-len], LEN, lenHC);
FUZ_CHECKTEST(ret==0, "HC compression failed despite sufficient space");
// Test compression with just one missing byte into output buffer => must fail
FUZ_DISPLAYTEST;
ret = LZ4_compress_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-(len-1)], LEN, len-1);
FUZ_CHECKTEST(ret, "compression overran output buffer");
FUZ_CHECKTEST(!test_canary(&cbuf[FUZ_avail]), "compression overran output buffer");
// Test HC compression with just one missing byte into output buffer => must fail
FUZ_DISPLAYTEST;
ret = LZ4_compressHC_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-(len-1)], LEN, lenHC-1);
FUZ_CHECKTEST(ret, "HC compression overran output buffer");
bytes += LEN;
cbytes += len;
hcbytes += lenHC;
FUZ_rand(&randState);
}
printf("all tests completed successfully \n");
printf("compression ratio: %0.3f%%\n", (double)cbytes/bytes*100);
printf("HC compression ratio: %0.3f%%\n", (double)hcbytes/bytes*100);
getchar();
return 0;
_output_error:
getchar();
return 1;
}
/*
fuzzer.c - Fuzzer test tool for LZ4
Copyright (C) Yann Collet - Andrew Mahone 2012-2013
Code started by Andrew Mahone, modified by Yann Collet
GPL v2 License
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
//**************************************
// Remove Visual warning messages
//**************************************
#define _CRT_SECURE_NO_WARNINGS // fgets
//**************************************
// Includes
//**************************************
#include <stdlib.h>
#include <stdio.h> // fgets, sscanf
#include <sys/timeb.h> // timeb
#include "lz4.h"
#include "lz4hc.h"
//**************************************
// Constants
//**************************************
#ifndef LZ4_VERSION
# define LZ4_VERSION ""
#endif
#define NB_ATTEMPTS (1<<17)
#define LEN ((1<<15))
#define SEQ_POW 2
#define NUM_SEQ (1 << SEQ_POW)
#define SEQ_MSK ((NUM_SEQ) - 1)
#define MOD_SEQ(x) ((((x) >> 8) & 255) == 0)
#define NEW_SEQ(x) ((((x) >> 10) %10) == 0)
#define PAGE_SIZE 4096
#define ROUND_PAGE(x) (((x) + PAGE_SIZE - 1) & ~(PAGE_SIZE - 1))
#define PRIME1 2654435761U
#define PRIME2 2246822519U
#define PRIME3 3266489917U
//*********************************************************
// Functions
//*********************************************************
static int FUZ_GetMilliStart()
{
struct timeb tb;
int nCount;
ftime( &tb );
nCount = (int) (tb.millitm + (tb.time & 0xfffff) * 1000);
return nCount;
}
static int FUZ_GetMilliSpan( int nTimeStart )
{
int nSpan = FUZ_GetMilliStart() - nTimeStart;
if ( nSpan < 0 )
nSpan += 0x100000 * 1000;
return nSpan;
}
unsigned int FUZ_rand(unsigned int* src)
{
*src = ((*src) * PRIME1) + PRIME2;
return *src;
}
int test_canary(unsigned char *buf)
{
int i;
for (i = 0; i < 2048; i++)
if (buf[i] != buf[i + 2048])
return 0;
return 1;
}
int FUZ_SecurityTest()
{
char* output;
char* input;
int i, r;
printf("Overflow test (issue 52)...\n");
input = (char*) malloc (20<<20);
output = (char*) malloc (20<<20);
input[0] = 0x0F;
input[1] = 0x00;
input[2] = 0x00;
for(i = 3; i < 16840000; i++)
input[i] = 0xff;
r = LZ4_decompress_fast(input, output, 20<<20);
free(input);
free(output);
printf(" Passed (return = %i < 0)\n",r);
return 0;
}
//int main(int argc, char *argv[]) {
int main() {
unsigned long long bytes = 0;
unsigned long long cbytes = 0;
unsigned long long hcbytes = 0;
unsigned char buf[LEN];
unsigned char testOut[LEN+1];
# define FUZ_max LZ4_COMPRESSBOUND(LEN)
# define FUZ_avail ROUND_PAGE(FUZ_max)
const int off_full = FUZ_avail - FUZ_max;
unsigned char cbuf[FUZ_avail + PAGE_SIZE];
unsigned int seed, randState, cur_seq=PRIME3, seeds[NUM_SEQ], timestamp=FUZ_GetMilliStart();
int i, j, k, ret, len, lenHC, attemptNb;
char userInput[30] = {0};
# define FUZ_CHECKTEST(cond, message) if (cond) { printf("Test %i : %s : seed %u, cycle %i \n", testNb, message, seed, attemptNb); goto _output_error; }
# define FUZ_DISPLAYTEST testNb++; printf("%2i\b\b", testNb);
void* stateLZ4 = malloc(LZ4_sizeofState());
void* stateLZ4HC = malloc(LZ4_sizeofStateHC());
printf("starting LZ4 fuzzer (%s)\n", LZ4_VERSION);
printf("Select an Initialisation number (default : random) : ");
fflush(stdout);
if ( fgets(userInput, sizeof userInput, stdin) )
{
if ( sscanf(userInput, "%d", &seed) == 1 ) {}
else seed = FUZ_GetMilliSpan(timestamp);
}
printf("Seed = %u\n", seed);
randState = seed;
//FUZ_SecurityTest();
for (i = 0; i < 2048; i++)
cbuf[FUZ_avail + i] = cbuf[FUZ_avail + 2048 + i] = FUZ_rand(&randState) >> 16;
for (attemptNb = 0; attemptNb < NB_ATTEMPTS; attemptNb++)
{
int testNb = 0;
printf("\r%7i /%7i - ", attemptNb, NB_ATTEMPTS);
for (j = 0; j < NUM_SEQ; j++) {
seeds[j] = FUZ_rand(&randState) << 8;
seeds[j] ^= (FUZ_rand(&randState) >> 8) & 65535;
}
for (j = 0; j < LEN; j++) {
k = FUZ_rand(&randState);
if (j == 0 || NEW_SEQ(k))
cur_seq = seeds[(FUZ_rand(&randState) >> 16) & SEQ_MSK];
if (MOD_SEQ(k)) {
k = (FUZ_rand(&randState) >> 16) & SEQ_MSK;
seeds[k] = FUZ_rand(&randState) << 8;
seeds[k] ^= (FUZ_rand(&randState) >> 8) & 65535;
}
buf[j] = FUZ_rand(&cur_seq) >> 16;
}
// Test compression HC
FUZ_DISPLAYTEST; // 1
ret = LZ4_compressHC_limitedOutput((const char*)buf, (char*)&cbuf[off_full], LEN, FUZ_max);
FUZ_CHECKTEST(ret==0, "LZ4_compressHC_limitedOutput() failed despite sufficient space");
lenHC = ret;
// Test compression HC using external state
FUZ_DISPLAYTEST; // 1
ret = LZ4_compressHC_withStateHC(stateLZ4HC, (const char*)buf, (char*)&cbuf[off_full], LEN);
FUZ_CHECKTEST(ret==0, "LZ4_compressHC_withStateHC() failed");
// Test compression using external state
FUZ_DISPLAYTEST; // 2
ret = LZ4_compress_withState(stateLZ4, (const char*)buf, (char*)&cbuf[off_full], LEN);
FUZ_CHECKTEST(ret==0, "LZ4_compress_withState() failed");
// Test compression
FUZ_DISPLAYTEST; // 2
ret = LZ4_compress_limitedOutput((const char*)buf, (char*)&cbuf[off_full], LEN, FUZ_max);
FUZ_CHECKTEST(ret==0, "LZ4_compress_limitedOutput() failed despite sufficient space");
len = ret;
// Test decoding with output size being exactly what's necessary => must work
FUZ_DISPLAYTEST; // 3
ret = LZ4_decompress_fast((char*)&cbuf[off_full], (char*)testOut, LEN);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_fast failed despite correct space");
// Test decoding with one byte missing => must fail
FUZ_DISPLAYTEST; // 4
ret = LZ4_decompress_fast((char*)&cbuf[off_full], (char*)testOut, LEN-1);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_fast should have failed, due to Output Size being too small");
// Test decoding with one byte too much => must fail
FUZ_DISPLAYTEST;
ret = LZ4_decompress_fast((char*)&cbuf[off_full], (char*)testOut, LEN+1);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_fast should have failed, due to Output Size being too large");
// Test decoding with enough output size => must work
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len, LEN+1);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_safe failed despite sufficient space");
// Test decoding with output size being exactly what's necessary => must work
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len, LEN);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_safe failed despite sufficient space");
// Test decoding with output size being one byte too short => must fail
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len, LEN-1);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_safe should have failed, due to Output Size being one byte too short");
// Test decoding with input size being one byte too short => must fail
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len-1, LEN);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_safe should have failed, due to input size being one byte too short");
// Test decoding with input size being one byte too large => must fail
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe((char*)&cbuf[off_full], (char*)testOut, len+1, LEN);
FUZ_CHECKTEST(ret>=0, "LZ4_decompress_safe should have failed, due to input size being too large");
//if (ret>=0) { printf("Test 10 : decompression should have failed, due to input size being too large : seed %u, len %d\n", seed, LEN); goto _output_error; }
// Test partial decoding with target output size being max/2 => must work
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe_partial((char*)&cbuf[off_full], (char*)testOut, len, LEN/2, LEN);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_safe_partial failed despite sufficient space");
// Test partial decoding with target output size being just below max => must work
FUZ_DISPLAYTEST;
ret = LZ4_decompress_safe_partial((char*)&cbuf[off_full], (char*)testOut, len, LEN-3, LEN);
FUZ_CHECKTEST(ret<0, "LZ4_decompress_safe_partial failed despite sufficient space");
// Test compression with output size being exactly what's necessary (should work)
FUZ_DISPLAYTEST;
ret = LZ4_compress_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-len], LEN, len);
FUZ_CHECKTEST(ret==0, "LZ4_compress_limitedOutput() failed despite sufficient space");
FUZ_CHECKTEST(!test_canary(&cbuf[FUZ_avail]), "compression overran output buffer");
// Test compression with output size being exactly what's necessary and external state (should work)
FUZ_DISPLAYTEST; // 2
ret = LZ4_compress_limitedOutput_withState(stateLZ4, (const char*)buf, (char*)&cbuf[off_full], LEN, len);
FUZ_CHECKTEST(ret==0, "LZ4_compress_limitedOutput_withState() failed despite sufficient space");
FUZ_CHECKTEST(!test_canary(&cbuf[FUZ_avail]), "compression overran output buffer");
// Test HC compression with output size being exactly what's necessary (should work)
FUZ_DISPLAYTEST;
ret = LZ4_compressHC_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-len], LEN, lenHC);
FUZ_CHECKTEST(ret==0, "LZ4_compressHC_limitedOutput() failed despite sufficient space");
// Test HC compression with output size being exactly what's necessary (should work)
FUZ_DISPLAYTEST;
ret = LZ4_compressHC_limitedOutput_withStateHC(stateLZ4HC, (const char*)buf, (char*)&cbuf[FUZ_avail-len], LEN, lenHC);
FUZ_CHECKTEST(ret==0, "LZ4_compressHC_limitedOutput_withStateHC() failed despite sufficient space");
// Test compression with just one missing byte into output buffer => must fail
FUZ_DISPLAYTEST;
ret = LZ4_compress_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-(len-1)], LEN, len-1);
FUZ_CHECKTEST(ret, "compression overran output buffer");
FUZ_CHECKTEST(!test_canary(&cbuf[FUZ_avail]), "compression overran output buffer");
// Test HC compression with just one missing byte into output buffer => must fail
FUZ_DISPLAYTEST;
ret = LZ4_compressHC_limitedOutput((const char*)buf, (char*)&cbuf[FUZ_avail-(len-1)], LEN, lenHC-1);
FUZ_CHECKTEST(ret, "HC compression overran output buffer");
bytes += LEN;
cbytes += len;
hcbytes += lenHC;
FUZ_rand(&randState);
}
printf("all tests completed successfully \n");
printf("compression ratio: %0.3f%%\n", (double)cbytes/bytes*100);
printf("HC compression ratio: %0.3f%%\n", (double)hcbytes/bytes*100);
getchar();
return 0;
_output_error:
getchar();
return 1;
}

169
lz4.1
View File

@ -1,84 +1,87 @@
\"
\" lz4.1: This is a manual page for 'lz4' program. This file is part of the
\" lz4 <https://code.google.com/p/lz4/> project.
\"
\" No hyphenation
.hy 0
.nr HY 0
.TH lz4 man
.SH NAME
\fBlz4\fR - Extremely fast compression algorithm
.SH SYNOPSIS
.TP 5
\fBlz4\fR [\fBOPTIONS\fR] [-|INPUT-FILE] <OUTPUT-FILE>
.SH DESCRIPTION
.PP
\fBlz4\fR is an extremely fast lossless compression algorithm. It is based on
the \fBLZ77\fR family of compression scheme. At the compression speed of 400
MB/s per core, \fBlz4\fR is also scalable with multi-core CPUs. It features
an extremely fast decoder, with speed in multiple GB/s per core, typically
reaching the RAM speed limits on multi-core systems. \fBlz4\fR supports
following options
.SH OPTIONS
.TP
.B \-1
fast compression (default)
.TP
.B \-9
high compression
.TP
.B \-d
decompression
.TP
.B \-f
overwrite output without prompting
.TP
.B \-h/\-H
display help/long help and exit
.TP
.B \-V
display Version number and exit
.TP
.B \-v
verbose mode
.TP
.B \-q
suppress warnings; specify twice to suppress errors too
.TP
.B \-c
force write to standard output, even if it is the console
.TP
.B \-t
test compressed file integrity
.TP
.B \-z
force compression
.TP
.B \-B#
block size [4-7](default : 7)
.TP
.B \-BD
block dependency (improve compression ratio)
.TP
.B \-BX
enable block checksum (default:disabled)
.TP
.B \-Sx
disable stream checksum (default:enabled)
.TP
.B \-b
benchmark file(s)
.TP
.B \-i#
iteration loops [1-9](default : 3), benchmark mode only
.SH BUGS
Report bugs at:- https://code.google.com/p/lz4/
.SH AUTHOR
\"
\" lz4.1: This is a manual page for 'lz4' program. This file is part of the
\" lz4 <https://code.google.com/p/lz4/> project.
\"
\" No hyphenation
.hy 0
.nr HY 0
.TH lz4 man
.SH NAME
\fBlz4\fR - Extremely fast compression algorithm
.SH SYNOPSIS
.TP 5
\fBlz4\fR [\fBOPTIONS\fR] [-|INPUT-FILE] <OUTPUT-FILE>
.SH DESCRIPTION
.PP
\fBlz4\fR is an extremely fast lossless compression algorithm. It is based on
the \fBLZ77\fR family of compression scheme. At the compression speed of 400
MB/s per core, \fBlz4\fR is also scalable with multi-core CPUs. It features
an extremely fast decoder, with speed in multiple GB/s per core, typically
reaching the RAM speed limits on multi-core systems. \fBlz4\fR supports
following options
.SH OPTIONS
.TP
.B \-1
fast compression (default)
.TP
.B \-9
high compression
.TP
.B \-d
decompression
.TP
.B \-f
overwrite output without prompting
.TP
.B \-h/\-H
display help/long help and exit
.TP
.B \-V
display Version number and exit
.TP
.B \-v
verbose mode
.TP
.B \-q
suppress warnings; specify twice to suppress errors too
.TP
.B \-c
force write to standard output, even if it is the console
.TP
.B \-t
test compressed file integrity
.TP
.B \-z
force compression
.TP
.B \-l
use Legacy format (useful for Linux Kernel compression)
.TP
.B \-B#
block size [4-7](default : 7)
.TP
.B \-BD
block dependency (improve compression ratio)
.TP
.B \-BX
enable block checksum (default:disabled)
.TP
.B \-Sx
disable stream checksum (default:enabled)
.TP
.B \-b
benchmark file(s)
.TP
.B \-i#
iteration loops [1-9](default : 3), benchmark mode only
.SH BUGS
Report bugs at:- https://code.google.com/p/lz4/
.SH AUTHOR
Yann Collet

1687
lz4.c
View File

File diff suppressed because it is too large Load Diff

454
lz4.h
View File

@ -1,205 +1,249 @@
/*
LZ4 - Fast LZ compression algorithm
Header File
Copyright (C) 2011-2013, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
#pragma once
#if defined (__cplusplus)
extern "C" {
#endif
//**************************************
// Compiler Options
//**************************************
#if defined(_MSC_VER) && !defined(__cplusplus) // Visual Studio
# define inline __inline // Visual C is not C99, but supports some kind of inline
#endif
//****************************
// Simple Functions
//****************************
int LZ4_compress (const char* source, char* dest, int inputSize);
int LZ4_decompress_safe (const char* source, char* dest, int inputSize, int maxOutputSize);
/*
LZ4_compress() :
Compresses 'inputSize' bytes from 'source' into 'dest'.
Destination buffer must be already allocated,
and must be sized to handle worst cases situations (input data not compressible)
Worst case size evaluation is provided by function LZ4_compressBound()
inputSize : Max supported value is LZ4_MAX_INPUT_VALUE
return : the number of bytes written in buffer dest
or 0 if the compression fails
LZ4_decompress_safe() :
maxOutputSize : is the size of the destination buffer (which must be already allocated)
return : the number of bytes decoded in the destination buffer (necessarily <= maxOutputSize)
If the source stream is detected malformed, the function will stop decoding and return a negative result.
This function is protected against buffer overflow exploits (never writes outside of output buffer, and never reads outside of input buffer). Therefore, it is protected against malicious data packets
*/
//****************************
// Advanced Functions
//****************************
#define LZ4_MAX_INPUT_SIZE 0x7E000000 // 2 113 929 216 bytes
#define LZ4_COMPRESSBOUND(isize) ((unsigned int)(isize) > (unsigned int)LZ4_MAX_INPUT_SIZE ? 0 : (isize) + ((isize)/255) + 16)
static inline int LZ4_compressBound(int isize) { return LZ4_COMPRESSBOUND(isize); }
/*
LZ4_compressBound() :
Provides the maximum size that LZ4 may output in a "worst case" scenario (input data not compressible)
primarily useful for memory allocation of output buffer.
inline function is recommended for the general case,
macro is also provided when result needs to be evaluated at compilation (such as stack memory allocation).
isize : is the input size. Max supported value is LZ4_MAX_INPUT_SIZE
return : maximum output size in a "worst case" scenario
or 0, if input size is too large ( > LZ4_MAX_INPUT_SIZE)
*/
int LZ4_compress_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
/*
LZ4_compress_limitedOutput() :
Compress 'inputSize' bytes from 'source' into an output buffer 'dest' of maximum size 'maxOutputSize'.
If it cannot achieve it, compression will stop, and result of the function will be zero.
This function never writes outside of provided output buffer.
inputSize : Max supported value is LZ4_MAX_INPUT_VALUE
maxOutputSize : is the size of the destination buffer (which must be already allocated)
return : the number of bytes written in buffer 'dest'
or 0 if the compression fails
*/
int LZ4_decompress_fast (const char* source, char* dest, int outputSize);
/*
LZ4_decompress_fast() :
outputSize : is the original (uncompressed) size
return : the number of bytes read from the source buffer (in other words, the compressed size)
If the source stream is malformed, the function will stop decoding and return a negative result.
note : This function is a bit faster than LZ4_decompress_safe()
This function never writes outside of output buffers, but may read beyond input buffer in case of malicious data packet.
Use this function preferably into a trusted environment (data to decode comes from a trusted source).
Destination buffer must be already allocated. Its size must be a minimum of 'outputSize' bytes.
*/
int LZ4_decompress_safe_partial (const char* source, char* dest, int inputSize, int targetOutputSize, int maxOutputSize);
/*
LZ4_decompress_safe_partial() :
This function decompress a compressed block of size 'inputSize' at position 'source'
into output buffer 'dest' of size 'maxOutputSize'.
The function tries to stop decompressing operation as soon as 'targetOutputSize' has been reached,
reducing decompression time.
return : the number of bytes decoded in the destination buffer (necessarily <= maxOutputSize)
Note : this number can be < 'targetOutputSize' should the compressed block to decode be smaller.
Always control how many bytes were decoded.
If the source stream is detected malformed, the function will stop decoding and return a negative result.
This function never writes outside of output buffer, and never reads outside of input buffer. It is therefore protected against malicious data packets
*/
//****************************
// Stream Functions
//****************************
void* LZ4_create (const char* inputBuffer);
int LZ4_compress_continue (void* LZ4_Data, const char* source, char* dest, int inputSize);
int LZ4_compress_limitedOutput_continue (void* LZ4_Data, const char* source, char* dest, int inputSize, int maxOutputSize);
char* LZ4_slideInputBuffer (void* LZ4_Data);
int LZ4_free (void* LZ4_Data);
/*
These functions allow the compression of dependent blocks, where each block benefits from prior 64 KB within preceding blocks.
In order to achieve this, it is necessary to start creating the LZ4 Data Structure, thanks to the function :
void* LZ4_create (const char* inputBuffer);
The result of the function is the (void*) pointer on the LZ4 Data Structure.
This pointer will be needed in all other functions.
If the pointer returned is NULL, then the allocation has failed, and compression must be aborted.
The only parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
The input buffer must be already allocated, and size at least 192KB.
'inputBuffer' will also be the 'const char* source' of the first block.
All blocks are expected to lay next to each other within the input buffer, starting from 'inputBuffer'.
To compress each block, use either LZ4_compress_continue() or LZ4_compress_limitedOutput_continue().
Their behavior are identical to LZ4_compress() or LZ4_compress_limitedOutput(),
but require the LZ4 Data Structure as their first argument, and check that each block starts right after the previous one.
If next block does not begin immediately after the previous one, the compression will fail (return 0).
When it's no longer possible to lay the next block after the previous one (not enough space left into input buffer), a call to :
char* LZ4_slideInputBuffer(void* LZ4_Data);
must be performed. It will typically copy the latest 64KB of input at the beginning of input buffer.
Note that, for this function to work properly, minimum size of an input buffer must be 192KB.
==> The memory position where the next input data block must start is provided as the result of the function.
Compression can then resume, using LZ4_compress_continue() or LZ4_compress_limitedOutput_continue(), as usual.
When compression is completed, a call to LZ4_free() will release the memory used by the LZ4 Data Structure.
*/
int LZ4_decompress_safe_withPrefix64k (const char* source, char* dest, int inputSize, int maxOutputSize);
int LZ4_decompress_fast_withPrefix64k (const char* source, char* dest, int outputSize);
/*
*_withPrefix64k() :
These decoding functions work the same as their "normal name" versions,
but can use up to 64KB of data in front of 'char* dest'.
These functions are necessary to decode inter-dependant blocks.
*/
//****************************
// Obsolete Functions
//****************************
static inline int LZ4_uncompress (const char* source, char* dest, int outputSize) { return LZ4_decompress_fast(source, dest, outputSize); }
static inline int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize) { return LZ4_decompress_safe(source, dest, isize, maxOutputSize); }
/*
These functions are deprecated and should no longer be used.
They are provided here for compatibility with existing user programs.
*/
#if defined (__cplusplus)
}
#endif
/*
LZ4 - Fast LZ compression algorithm
Header File
Copyright (C) 2011-2013, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
#pragma once
#if defined (__cplusplus)
extern "C" {
#endif
//**************************************
// Compiler Options
//**************************************
#if defined(_MSC_VER) && !defined(__cplusplus) // Visual Studio
# define inline __inline // Visual C is not C99, but supports some kind of inline
#endif
//****************************
// Simple Functions
//****************************
int LZ4_compress (const char* source, char* dest, int inputSize);
int LZ4_decompress_safe (const char* source, char* dest, int inputSize, int maxOutputSize);
/*
LZ4_compress() :
Compresses 'inputSize' bytes from 'source' into 'dest'.
Destination buffer must be already allocated,
and must be sized to handle worst cases situations (input data not compressible)
Worst case size evaluation is provided by function LZ4_compressBound()
inputSize : Max supported value is LZ4_MAX_INPUT_VALUE
return : the number of bytes written in buffer dest
or 0 if the compression fails
LZ4_decompress_safe() :
maxOutputSize : is the size of the destination buffer (which must be already allocated)
return : the number of bytes decoded in the destination buffer (necessarily <= maxOutputSize)
If the source stream is detected malformed, the function will stop decoding and return a negative result.
This function is protected against buffer overflow exploits (never writes outside of output buffer, and never reads outside of input buffer). Therefore, it is protected against malicious data packets
*/
//****************************
// Advanced Functions
//****************************
#define LZ4_MAX_INPUT_SIZE 0x7E000000 // 2 113 929 216 bytes
#define LZ4_COMPRESSBOUND(isize) ((unsigned int)(isize) > (unsigned int)LZ4_MAX_INPUT_SIZE ? 0 : (isize) + ((isize)/255) + 16)
static inline int LZ4_compressBound(int isize) { return LZ4_COMPRESSBOUND(isize); }
/*
LZ4_compressBound() :
Provides the maximum size that LZ4 may output in a "worst case" scenario (input data not compressible)
primarily useful for memory allocation of output buffer.
inline function is recommended for the general case,
macro is also provided when result needs to be evaluated at compilation (such as stack memory allocation).
isize : is the input size. Max supported value is LZ4_MAX_INPUT_SIZE
return : maximum output size in a "worst case" scenario
or 0, if input size is too large ( > LZ4_MAX_INPUT_SIZE)
*/
int LZ4_compress_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
/*
LZ4_compress_limitedOutput() :
Compress 'inputSize' bytes from 'source' into an output buffer 'dest' of maximum size 'maxOutputSize'.
If it cannot achieve it, compression will stop, and result of the function will be zero.
This function never writes outside of provided output buffer.
inputSize : Max supported value is LZ4_MAX_INPUT_VALUE
maxOutputSize : is the size of the destination buffer (which must be already allocated)
return : the number of bytes written in buffer 'dest'
or 0 if the compression fails
*/
int LZ4_decompress_fast (const char* source, char* dest, int outputSize);
/*
LZ4_decompress_fast() :
outputSize : is the original (uncompressed) size
return : the number of bytes read from the source buffer (in other words, the compressed size)
If the source stream is malformed, the function will stop decoding and return a negative result.
note : This function is a bit faster than LZ4_decompress_safe()
This function never writes outside of output buffers, but may read beyond input buffer in case of malicious data packet.
Use this function preferably into a trusted environment (data to decode comes from a trusted source).
Destination buffer must be already allocated. Its size must be a minimum of 'outputSize' bytes.
*/
int LZ4_decompress_safe_partial (const char* source, char* dest, int inputSize, int targetOutputSize, int maxOutputSize);
/*
LZ4_decompress_safe_partial() :
This function decompress a compressed block of size 'inputSize' at position 'source'
into output buffer 'dest' of size 'maxOutputSize'.
The function tries to stop decompressing operation as soon as 'targetOutputSize' has been reached,
reducing decompression time.
return : the number of bytes decoded in the destination buffer (necessarily <= maxOutputSize)
Note : this number can be < 'targetOutputSize' should the compressed block to decode be smaller.
Always control how many bytes were decoded.
If the source stream is detected malformed, the function will stop decoding and return a negative result.
This function never writes outside of output buffer, and never reads outside of input buffer. It is therefore protected against malicious data packets
*/
//*****************************
// Using an external allocation
//*****************************
int LZ4_sizeofState();
int LZ4_compress_withState (void* state, const char* source, char* dest, int inputSize);
int LZ4_compress_limitedOutput_withState (void* state, const char* source, char* dest, int inputSize, int maxOutputSize);
/*
These functions are provided should you prefer to allocate memory for compression tables with your own allocation methods.
To know how much memory must be allocated for the compression tables, use :
int LZ4_sizeofState();
Note that tables must be aligned on 4-bytes boundaries, otherwise compression will fail (return code 0).
The allocated memory can be provided to the compressions functions using 'void* state' parameter.
LZ4_compress_withState() and LZ4_compress_limitedOutput_withState() are equivalent to previously described functions.
They just use the externally allocated memory area instead of allocating their own (on stack, or on heap).
*/
//****************************
// Streaming Functions
//****************************
void* LZ4_create (const char* inputBuffer);
int LZ4_compress_continue (void* LZ4_Data, const char* source, char* dest, int inputSize);
int LZ4_compress_limitedOutput_continue (void* LZ4_Data, const char* source, char* dest, int inputSize, int maxOutputSize);
char* LZ4_slideInputBuffer (void* LZ4_Data);
int LZ4_free (void* LZ4_Data);
/*
These functions allow the compression of dependent blocks, where each block benefits from prior 64 KB within preceding blocks.
In order to achieve this, it is necessary to start creating the LZ4 Data Structure, thanks to the function :
void* LZ4_create (const char* inputBuffer);
The result of the function is the (void*) pointer on the LZ4 Data Structure.
This pointer will be needed in all other functions.
If the pointer returned is NULL, then the allocation has failed, and compression must be aborted.
The only parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
The input buffer must be already allocated, and size at least 192KB.
'inputBuffer' will also be the 'const char* source' of the first block.
All blocks are expected to lay next to each other within the input buffer, starting from 'inputBuffer'.
To compress each block, use either LZ4_compress_continue() or LZ4_compress_limitedOutput_continue().
Their behavior are identical to LZ4_compress() or LZ4_compress_limitedOutput(),
but require the LZ4 Data Structure as their first argument, and check that each block starts right after the previous one.
If next block does not begin immediately after the previous one, the compression will fail (return 0).
When it's no longer possible to lay the next block after the previous one (not enough space left into input buffer), a call to :
char* LZ4_slideInputBuffer(void* LZ4_Data);
must be performed. It will typically copy the latest 64KB of input at the beginning of input buffer.
Note that, for this function to work properly, minimum size of an input buffer must be 192KB.
==> The memory position where the next input data block must start is provided as the result of the function.
Compression can then resume, using LZ4_compress_continue() or LZ4_compress_limitedOutput_continue(), as usual.
When compression is completed, a call to LZ4_free() will release the memory used by the LZ4 Data Structure.
*/
int LZ4_sizeofStreamState();
int LZ4_resetStreamState(void* state, const char* inputBuffer);
/*
These functions achieve the same result as :
void* LZ4_create (const char* inputBuffer);
They are provided here to allow the user program to allocate memory using its own routines.
To know how much space must be allocated, use LZ4_sizeofStreamState();
Note also that space must be 4-bytes aligned.
Once space is allocated, you must initialize it using : LZ4_resetStreamState(void* state, const char* inputBuffer);
void* state is a pointer to the space allocated.
It must be aligned on 4-bytes boundaries, and be large enough.
The parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
The input buffer must be already allocated, and size at least 192KB.
'inputBuffer' will also be the 'const char* source' of the first block.
The same space can be re-used multiple times, just by initializing it each time with LZ4_resetStreamState().
return value of LZ4_resetStreamState() must be 0 is OK.
Any other value means there was an error (typically, pointer is not aligned on 4-bytes boundaries).
*/
int LZ4_decompress_safe_withPrefix64k (const char* source, char* dest, int inputSize, int maxOutputSize);
int LZ4_decompress_fast_withPrefix64k (const char* source, char* dest, int outputSize);
/*
*_withPrefix64k() :
These decoding functions work the same as their "normal name" versions,
but can use up to 64KB of data in front of 'char* dest'.
These functions are necessary to decode inter-dependant blocks.
*/
//****************************
// Obsolete Functions
//****************************
static inline int LZ4_uncompress (const char* source, char* dest, int outputSize) { return LZ4_decompress_fast(source, dest, outputSize); }
static inline int LZ4_uncompress_unknownOutputSize (const char* source, char* dest, int isize, int maxOutputSize) { return LZ4_decompress_safe(source, dest, isize, maxOutputSize); }
/*
These functions are deprecated and should no longer be used.
They are provided here for compatibility with existing user programs.
*/
#if defined (__cplusplus)
}
#endif

2508
lz4cli.c
View File

File diff suppressed because it is too large Load Diff

1673
lz4hc.c
View File

File diff suppressed because it is too large Load Diff

268
lz4hc.h
View File

@ -1,111 +1,157 @@
/*
LZ4 HC - High Compression Mode of LZ4
Header File
Copyright (C) 2011-2013, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
#pragma once
#if defined (__cplusplus)
extern "C" {
#endif
int LZ4_compressHC (const char* source, char* dest, int inputSize);
/*
LZ4_compressHC :
return : the number of bytes in compressed buffer dest
or 0 if compression fails.
note : destination buffer must be already allocated.
To avoid any problem, size it to handle worst cases situations (input data not compressible)
Worst case size evaluation is provided by function LZ4_compressBound() (see "lz4.h")
*/
int LZ4_compressHC_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
/*
LZ4_compress_limitedOutput() :
Compress 'inputSize' bytes from 'source' into an output buffer 'dest' of maximum size 'maxOutputSize'.
If it cannot achieve it, compression will stop, and result of the function will be zero.
This function never writes outside of provided output buffer.
inputSize : Max supported value is 1 GB
maxOutputSize : is maximum allowed size into the destination buffer (which must be already allocated)
return : the number of output bytes written in buffer 'dest'
or 0 if compression fails.
*/
/* Note :
Decompression functions are provided within LZ4 source code (see "lz4.h") (BSD license)
*/
/* Advanced Functions */
void* LZ4_createHC (const char* inputBuffer);
int LZ4_compressHC_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize);
int LZ4_compressHC_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize);
char* LZ4_slideInputBufferHC (void* LZ4HC_Data);
int LZ4_freeHC (void* LZ4HC_Data);
/*
These functions allow the compression of dependent blocks, where each block benefits from prior 64 KB within preceding blocks.
In order to achieve this, it is necessary to start creating the LZ4HC Data Structure, thanks to the function :
void* LZ4_createHC (const char* inputBuffer);
The result of the function is the (void*) pointer on the LZ4HC Data Structure.
This pointer will be needed in all other functions.
If the pointer returned is NULL, then the allocation has failed, and compression must be aborted.
The only parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
The input buffer must be already allocated, and size at least 192KB.
'inputBuffer' will also be the 'const char* source' of the first block.
All blocks are expected to lay next to each other within the input buffer, starting from 'inputBuffer'.
To compress each block, use either LZ4_compressHC_continue() or LZ4_compressHC_limitedOutput_continue().
Their behavior are identical to LZ4_compressHC() or LZ4_compressHC_limitedOutput(),
but require the LZ4HC Data Structure as their first argument, and check that each block starts right after the previous one.
If next block does not begin immediately after the previous one, the compression will fail (return 0).
When it's no longer possible to lay the next block after the previous one (not enough space left into input buffer), a call to :
char* LZ4_slideInputBufferHC(void* LZ4HC_Data);
must be performed. It will typically copy the latest 64KB of input at the beginning of input buffer.
Note that, for this function to work properly, minimum size of an input buffer must be 192KB.
==> The memory position where the next input data block must start is provided as the result of the function.
Compression can then resume, using LZ4_compressHC_continue() or LZ4_compressHC_limitedOutput_continue(), as usual.
When compression is completed, a call to LZ4_freeHC() will release the memory used by the LZ4HC Data Structure.
*/
#if defined (__cplusplus)
}
#endif
/*
LZ4 HC - High Compression Mode of LZ4
Header File
Copyright (C) 2011-2013, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- LZ4 homepage : http://fastcompression.blogspot.com/p/lz4.html
- LZ4 source repository : http://code.google.com/p/lz4/
*/
#pragma once
#if defined (__cplusplus)
extern "C" {
#endif
int LZ4_compressHC (const char* source, char* dest, int inputSize);
/*
LZ4_compressHC :
return : the number of bytes in compressed buffer dest
or 0 if compression fails.
note : destination buffer must be already allocated.
To avoid any problem, size it to handle worst cases situations (input data not compressible)
Worst case size evaluation is provided by function LZ4_compressBound() (see "lz4.h")
*/
int LZ4_compressHC_limitedOutput (const char* source, char* dest, int inputSize, int maxOutputSize);
/*
LZ4_compress_limitedOutput() :
Compress 'inputSize' bytes from 'source' into an output buffer 'dest' of maximum size 'maxOutputSize'.
If it cannot achieve it, compression will stop, and result of the function will be zero.
This function never writes outside of provided output buffer.
inputSize : Max supported value is 1 GB
maxOutputSize : is maximum allowed size into the destination buffer (which must be already allocated)
return : the number of output bytes written in buffer 'dest'
or 0 if compression fails.
*/
/* Note :
Decompression functions are provided within LZ4 source code (see "lz4.h") (BSD license)
*/
//*****************************
// Using an external allocation
//*****************************
int LZ4_sizeofStateHC();
int LZ4_compressHC_withStateHC (void* state, const char* source, char* dest, int inputSize);
int LZ4_compressHC_limitedOutput_withStateHC (void* state, const char* source, char* dest, int inputSize, int maxOutputSize);
/*
These functions are provided should you prefer to allocate memory for compression tables with your own allocation methods.
To know how much memory must be allocated for the compression tables, use :
int LZ4_sizeofStateHC();
Note that tables must be aligned for pointer (32 or 64 bits), otherwise compression will fail (return code 0).
The allocated memory can be provided to the compressions functions using 'void* state' parameter.
LZ4_compress_withStateHC() and LZ4_compress_limitedOutput_withStateHC() are equivalent to previously described functions.
They just use the externally allocated memory area instead of allocating their own (on stack, or on heap).
*/
//****************************
// Streaming Functions
//****************************
void* LZ4_createHC (const char* inputBuffer);
int LZ4_compressHC_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize);
int LZ4_compressHC_limitedOutput_continue (void* LZ4HC_Data, const char* source, char* dest, int inputSize, int maxOutputSize);
char* LZ4_slideInputBufferHC (void* LZ4HC_Data);
int LZ4_freeHC (void* LZ4HC_Data);
/*
These functions allow the compression of dependent blocks, where each block benefits from prior 64 KB within preceding blocks.
In order to achieve this, it is necessary to start creating the LZ4HC Data Structure, thanks to the function :
void* LZ4_createHC (const char* inputBuffer);
The result of the function is the (void*) pointer on the LZ4HC Data Structure.
This pointer will be needed in all other functions.
If the pointer returned is NULL, then the allocation has failed, and compression must be aborted.
The only parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
The input buffer must be already allocated, and size at least 192KB.
'inputBuffer' will also be the 'const char* source' of the first block.
All blocks are expected to lay next to each other within the input buffer, starting from 'inputBuffer'.
To compress each block, use either LZ4_compressHC_continue() or LZ4_compressHC_limitedOutput_continue().
Their behavior are identical to LZ4_compressHC() or LZ4_compressHC_limitedOutput(),
but require the LZ4HC Data Structure as their first argument, and check that each block starts right after the previous one.
If next block does not begin immediately after the previous one, the compression will fail (return 0).
When it's no longer possible to lay the next block after the previous one (not enough space left into input buffer), a call to :
char* LZ4_slideInputBufferHC(void* LZ4HC_Data);
must be performed. It will typically copy the latest 64KB of input at the beginning of input buffer.
Note that, for this function to work properly, minimum size of an input buffer must be 192KB.
==> The memory position where the next input data block must start is provided as the result of the function.
Compression can then resume, using LZ4_compressHC_continue() or LZ4_compressHC_limitedOutput_continue(), as usual.
When compression is completed, a call to LZ4_freeHC() will release the memory used by the LZ4HC Data Structure.
*/
int LZ4_sizeofStreamStateHC();
int LZ4_resetStreamStateHC(void* state, const char* inputBuffer);
/*
These functions achieve the same result as :
void* LZ4_createHC (const char* inputBuffer);
They are provided here to allow the user program to allocate memory using its own routines.
To know how much space must be allocated, use LZ4_sizeofStreamStateHC();
Note also that space must be aligned for pointers (32 or 64 bits).
Once space is allocated, you must initialize it using : LZ4_resetStreamStateHC(void* state, const char* inputBuffer);
void* state is a pointer to the space allocated.
It must be aligned for pointers (32 or 64 bits), and be large enough.
The parameter 'const char* inputBuffer' must, obviously, point at the beginning of input buffer.
The input buffer must be already allocated, and size at least 192KB.
'inputBuffer' will also be the 'const char* source' of the first block.
The same space can be re-used multiple times, just by initializing it each time with LZ4_resetStreamState().
return value of LZ4_resetStreamStateHC() must be 0 is OK.
Any other value means there was an error (typically, state is not aligned for pointers (32 or 64 bits)).
*/
#if defined (__cplusplus)
}
#endif

950
xxhash.c
View File

@ -1,475 +1,475 @@
/*
xxHash - Fast Hash algorithm
Copyright (C) 2012-2013, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- xxHash source repository : http://code.google.com/p/xxhash/
*/
//**************************************
// Tuning parameters
//**************************************
// Unaligned memory access is automatically enabled for "common" CPU, such as x86.
// For others CPU, the compiler will be more cautious, and insert extra code to ensure aligned access is respected.
// If you know your target CPU supports unaligned memory access, you want to force this option manually to improve performance.
// You can also enable this parameter if you know your input data will always be aligned (boundaries of 4, for U32).
#if defined(__ARM_FEATURE_UNALIGNED) || defined(__i386) || defined(_M_IX86) || defined(__x86_64__) || defined(_M_X64)
# define XXH_USE_UNALIGNED_ACCESS 1
#endif
// XXH_ACCEPT_NULL_INPUT_POINTER :
// If the input pointer is a null pointer, xxHash default behavior is to trigger a memory access error, since it is a bad pointer.
// When this option is enabled, xxHash output for null input pointers will be the same as a null-length input.
// This option has a very small performance cost (only measurable on small inputs).
// By default, this option is disabled. To enable it, uncomment below define :
//#define XXH_ACCEPT_NULL_INPUT_POINTER 1
// XXH_FORCE_NATIVE_FORMAT :
// By default, xxHash library provides endian-independant Hash values, based on little-endian convention.
// Results are therefore identical for little-endian and big-endian CPU.
// This comes at a performance cost for big-endian CPU, since some swapping is required to emulate little-endian format.
// Should endian-independance be of no importance for your application, you may set the #define below to 1.
// It will improve speed for Big-endian CPU.
// This option has no impact on Little_Endian CPU.
#define XXH_FORCE_NATIVE_FORMAT 0
//**************************************
// Compiler Specific Options
//**************************************
// Disable some Visual warning messages
#ifdef _MSC_VER // Visual Studio
# pragma warning(disable : 4127) // disable: C4127: conditional expression is constant
#endif
#ifdef _MSC_VER // Visual Studio
# define forceinline static __forceinline
#else
# ifdef __GNUC__
# define forceinline static inline __attribute__((always_inline))
# else
# define forceinline static inline
# endif
#endif
//**************************************
// Includes & Memory related functions
//**************************************
#include "xxhash.h"
// Modify the local functions below should you wish to use some other memory related routines
// for malloc(), free()
#include <stdlib.h>
forceinline void* XXH_malloc(size_t s) { return malloc(s); }
forceinline void XXH_free (void* p) { free(p); }
// for memcpy()
#include <string.h>
forceinline void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcpy(dest,src,size); }
//**************************************
// Basic Types
//**************************************
#if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L // C99
# include <stdint.h>
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef uint32_t U32;
typedef int32_t S32;
typedef uint64_t U64;
#else
typedef unsigned char BYTE;
typedef unsigned short U16;
typedef unsigned int U32;
typedef signed int S32;
typedef unsigned long long U64;
#endif
#if defined(__GNUC__) && !defined(XXH_USE_UNALIGNED_ACCESS)
# define _PACKED __attribute__ ((packed))
#else
# define _PACKED
#endif
#if !defined(XXH_USE_UNALIGNED_ACCESS) && !defined(__GNUC__)
# ifdef __IBMC__
# pragma pack(1)
# else
# pragma pack(push, 1)
# endif
#endif
typedef struct _U32_S { U32 v; } _PACKED U32_S;
#if !defined(XXH_USE_UNALIGNED_ACCESS) && !defined(__GNUC__)
# pragma pack(pop)
#endif
#define A32(x) (((U32_S *)(x))->v)
//***************************************
// Compiler-specific Functions and Macros
//***************************************
#define GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
// Note : although _rotl exists for minGW (GCC under windows), performance seems poor
#if defined(_MSC_VER)
# define XXH_rotl32(x,r) _rotl(x,r)
#else
# define XXH_rotl32(x,r) ((x << r) | (x >> (32 - r)))
#endif
#if defined(_MSC_VER) // Visual Studio
# define XXH_swap32 _byteswap_ulong
#elif GCC_VERSION >= 403
# define XXH_swap32 __builtin_bswap32
#else
static inline U32 XXH_swap32 (U32 x) {
return ((x << 24) & 0xff000000 ) |
((x << 8) & 0x00ff0000 ) |
((x >> 8) & 0x0000ff00 ) |
((x >> 24) & 0x000000ff );}
#endif
//**************************************
// Constants
//**************************************
#define PRIME32_1 2654435761U
#define PRIME32_2 2246822519U
#define PRIME32_3 3266489917U
#define PRIME32_4 668265263U
#define PRIME32_5 374761393U
//**************************************
// Architecture Macros
//**************************************
typedef enum { XXH_bigEndian=0, XXH_littleEndian=1 } XXH_endianess;
#ifndef XXH_CPU_LITTLE_ENDIAN // It is possible to define XXH_CPU_LITTLE_ENDIAN externally, for example using a compiler switch
static const int one = 1;
# define XXH_CPU_LITTLE_ENDIAN (*(char*)(&one))
#endif
//**************************************
// Macros
//**************************************
#define XXH_STATIC_ASSERT(c) { enum { XXH_static_assert = 1/(!!(c)) }; } // use only *after* variable declarations
//****************************
// Memory reads
//****************************
typedef enum { XXH_aligned, XXH_unaligned } XXH_alignment;
forceinline U32 XXH_readLE32_align(const U32* ptr, XXH_endianess endian, XXH_alignment align)
{
if (align==XXH_unaligned)
return endian==XXH_littleEndian ? A32(ptr) : XXH_swap32(A32(ptr));
else
return endian==XXH_littleEndian ? *ptr : XXH_swap32(*ptr);
}
forceinline U32 XXH_readLE32(const U32* ptr, XXH_endianess endian) { return XXH_readLE32_align(ptr, endian, XXH_unaligned); }
//****************************
// Simple Hash Functions
//****************************
forceinline U32 XXH32_endian_align(const void* input, int len, U32 seed, XXH_endianess endian, XXH_alignment align)
{
const BYTE* p = (const BYTE*)input;
const BYTE* const bEnd = p + len;
U32 h32;
#ifdef XXH_ACCEPT_NULL_INPUT_POINTER
if (p==NULL) { len=0; p=(const BYTE*)(size_t)16; }
#endif
if (len>=16)
{
const BYTE* const limit = bEnd - 16;
U32 v1 = seed + PRIME32_1 + PRIME32_2;
U32 v2 = seed + PRIME32_2;
U32 v3 = seed + 0;
U32 v4 = seed - PRIME32_1;
do
{
v1 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_2; v1 = XXH_rotl32(v1, 13); v1 *= PRIME32_1; p+=4;
v2 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_2; v2 = XXH_rotl32(v2, 13); v2 *= PRIME32_1; p+=4;
v3 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_2; v3 = XXH_rotl32(v3, 13); v3 *= PRIME32_1; p+=4;
v4 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_2; v4 = XXH_rotl32(v4, 13); v4 *= PRIME32_1; p+=4;
} while (p<=limit);
h32 = XXH_rotl32(v1, 1) + XXH_rotl32(v2, 7) + XXH_rotl32(v3, 12) + XXH_rotl32(v4, 18);
}
else
{
h32 = seed + PRIME32_5;
}
h32 += (U32) len;
while (p<=bEnd-4)
{
h32 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_3;
h32 = XXH_rotl32(h32, 17) * PRIME32_4 ;
p+=4;
}
while (p<bEnd)
{
h32 += (*p) * PRIME32_5;
h32 = XXH_rotl32(h32, 11) * PRIME32_1 ;
p++;
}
h32 ^= h32 >> 15;
h32 *= PRIME32_2;
h32 ^= h32 >> 13;
h32 *= PRIME32_3;
h32 ^= h32 >> 16;
return h32;
}
U32 XXH32(const void* input, int len, U32 seed)
{
#if 0
// Simple version, good for code maintenance, but unfortunately slow for small inputs
void* state = XXH32_init(seed);
XXH32_update(state, input, len);
return XXH32_digest(state);
#else
XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
# if !defined(XXH_USE_UNALIGNED_ACCESS)
if ((((size_t)input) & 3)) // Input is aligned, let's leverage the speed advantage
{
if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned);
else
return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned);
}
# endif
if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned);
else
return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned);
#endif
}
//****************************
// Advanced Hash Functions
//****************************
struct XXH_state32_t
{
U64 total_len;
U32 seed;
U32 v1;
U32 v2;
U32 v3;
U32 v4;
int memsize;
char memory[16];
};
int XXH32_sizeofState()
{
XXH_STATIC_ASSERT(XXH32_SIZEOFSTATE >= sizeof(struct XXH_state32_t)); // A compilation error here means XXH32_SIZEOFSTATE is not large enough
return sizeof(struct XXH_state32_t);
}
XXH_errorcode XXH32_resetState(void* state_in, U32 seed)
{
struct XXH_state32_t * state = (struct XXH_state32_t *) state_in;
state->seed = seed;
state->v1 = seed + PRIME32_1 + PRIME32_2;
state->v2 = seed + PRIME32_2;
state->v3 = seed + 0;
state->v4 = seed - PRIME32_1;
state->total_len = 0;
state->memsize = 0;
return XXH_OK;
}
void* XXH32_init (U32 seed)
{
void* state = XXH_malloc (sizeof(struct XXH_state32_t));
XXH32_resetState(state, seed);
return state;
}
forceinline XXH_errorcode XXH32_update_endian (void* state_in, const void* input, int len, XXH_endianess endian)
{
struct XXH_state32_t * state = (struct XXH_state32_t *) state_in;
const BYTE* p = (const BYTE*)input;
const BYTE* const bEnd = p + len;
#ifdef XXH_ACCEPT_NULL_INPUT_POINTER
if (input==NULL) return XXH_ERROR;
#endif
state->total_len += len;
if (state->memsize + len < 16) // fill in tmp buffer
{
XXH_memcpy(state->memory + state->memsize, input, len);
state->memsize += len;
return XXH_OK;
}
if (state->memsize) // some data left from previous update
{
XXH_memcpy(state->memory + state->memsize, input, 16-state->memsize);
{
const U32* p32 = (const U32*)state->memory;
state->v1 += XXH_readLE32(p32, endian) * PRIME32_2; state->v1 = XXH_rotl32(state->v1, 13); state->v1 *= PRIME32_1; p32++;
state->v2 += XXH_readLE32(p32, endian) * PRIME32_2; state->v2 = XXH_rotl32(state->v2, 13); state->v2 *= PRIME32_1; p32++;
state->v3 += XXH_readLE32(p32, endian) * PRIME32_2; state->v3 = XXH_rotl32(state->v3, 13); state->v3 *= PRIME32_1; p32++;
state->v4 += XXH_readLE32(p32, endian) * PRIME32_2; state->v4 = XXH_rotl32(state->v4, 13); state->v4 *= PRIME32_1; p32++;
}
p += 16-state->memsize;
state->memsize = 0;
}
if (p <= bEnd-16)
{
const BYTE* const limit = bEnd - 16;
U32 v1 = state->v1;
U32 v2 = state->v2;
U32 v3 = state->v3;
U32 v4 = state->v4;
do
{
v1 += XXH_readLE32((const U32*)p, endian) * PRIME32_2; v1 = XXH_rotl32(v1, 13); v1 *= PRIME32_1; p+=4;
v2 += XXH_readLE32((const U32*)p, endian) * PRIME32_2; v2 = XXH_rotl32(v2, 13); v2 *= PRIME32_1; p+=4;
v3 += XXH_readLE32((const U32*)p, endian) * PRIME32_2; v3 = XXH_rotl32(v3, 13); v3 *= PRIME32_1; p+=4;
v4 += XXH_readLE32((const U32*)p, endian) * PRIME32_2; v4 = XXH_rotl32(v4, 13); v4 *= PRIME32_1; p+=4;
} while (p<=limit);
state->v1 = v1;
state->v2 = v2;
state->v3 = v3;
state->v4 = v4;
}
if (p < bEnd)
{
XXH_memcpy(state->memory, p, bEnd-p);
state->memsize = (int)(bEnd-p);
}
return XXH_OK;
}
XXH_errorcode XXH32_update (void* state_in, const void* input, int len)
{
XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
return XXH32_update_endian(state_in, input, len, XXH_littleEndian);
else
return XXH32_update_endian(state_in, input, len, XXH_bigEndian);
}
forceinline U32 XXH32_intermediateDigest_endian (void* state_in, XXH_endianess endian)
{
struct XXH_state32_t * state = (struct XXH_state32_t *) state_in;
const BYTE * p = (const BYTE*)state->memory;
BYTE* bEnd = (BYTE*)state->memory + state->memsize;
U32 h32;
if (state->total_len >= 16)
{
h32 = XXH_rotl32(state->v1, 1) + XXH_rotl32(state->v2, 7) + XXH_rotl32(state->v3, 12) + XXH_rotl32(state->v4, 18);
}
else
{
h32 = state->seed + PRIME32_5;
}
h32 += (U32) state->total_len;
while (p<=bEnd-4)
{
h32 += XXH_readLE32((const U32*)p, endian) * PRIME32_3;
h32 = XXH_rotl32(h32, 17) * PRIME32_4;
p+=4;
}
while (p<bEnd)
{
h32 += (*p) * PRIME32_5;
h32 = XXH_rotl32(h32, 11) * PRIME32_1;
p++;
}
h32 ^= h32 >> 15;
h32 *= PRIME32_2;
h32 ^= h32 >> 13;
h32 *= PRIME32_3;
h32 ^= h32 >> 16;
return h32;
}
U32 XXH32_intermediateDigest (void* state_in)
{
XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
return XXH32_intermediateDigest_endian(state_in, XXH_littleEndian);
else
return XXH32_intermediateDigest_endian(state_in, XXH_bigEndian);
}
U32 XXH32_digest (void* state_in)
{
U32 h32 = XXH32_intermediateDigest(state_in);
XXH_free(state_in);
return h32;
}
/*
xxHash - Fast Hash algorithm
Copyright (C) 2012-2014, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- xxHash source repository : http://code.google.com/p/xxhash/
*/
//**************************************
// Tuning parameters
//**************************************
// Unaligned memory access is automatically enabled for "common" CPU, such as x86.
// For others CPU, the compiler will be more cautious, and insert extra code to ensure aligned access is respected.
// If you know your target CPU supports unaligned memory access, you want to force this option manually to improve performance.
// You can also enable this parameter if you know your input data will always be aligned (boundaries of 4, for U32).
#if defined(__ARM_FEATURE_UNALIGNED) || defined(__i386) || defined(_M_IX86) || defined(__x86_64__) || defined(_M_X64)
# define XXH_USE_UNALIGNED_ACCESS 1
#endif
// XXH_ACCEPT_NULL_INPUT_POINTER :
// If the input pointer is a null pointer, xxHash default behavior is to trigger a memory access error, since it is a bad pointer.
// When this option is enabled, xxHash output for null input pointers will be the same as a null-length input.
// This option has a very small performance cost (only measurable on small inputs).
// By default, this option is disabled. To enable it, uncomment below define :
//#define XXH_ACCEPT_NULL_INPUT_POINTER 1
// XXH_FORCE_NATIVE_FORMAT :
// By default, xxHash library provides endian-independant Hash values, based on little-endian convention.
// Results are therefore identical for little-endian and big-endian CPU.
// This comes at a performance cost for big-endian CPU, since some swapping is required to emulate little-endian format.
// Should endian-independance be of no importance for your application, you may set the #define below to 1.
// It will improve speed for Big-endian CPU.
// This option has no impact on Little_Endian CPU.
#define XXH_FORCE_NATIVE_FORMAT 0
//**************************************
// Compiler Specific Options
//**************************************
// Disable some Visual warning messages
#ifdef _MSC_VER // Visual Studio
# pragma warning(disable : 4127) // disable: C4127: conditional expression is constant
#endif
#ifdef _MSC_VER // Visual Studio
# define FORCE_INLINE static __forceinline
#else
# ifdef __GNUC__
# define FORCE_INLINE static inline __attribute__((always_inline))
# else
# define FORCE_INLINE static inline
# endif
#endif
//**************************************
// Includes & Memory related functions
//**************************************
#include "xxhash.h"
// Modify the local functions below should you wish to use some other memory related routines
// for malloc(), free()
#include <stdlib.h>
FORCE_INLINE void* XXH_malloc(size_t s) { return malloc(s); }
FORCE_INLINE void XXH_free (void* p) { free(p); }
// for memcpy()
#include <string.h>
FORCE_INLINE void* XXH_memcpy(void* dest, const void* src, size_t size) { return memcpy(dest,src,size); }
//**************************************
// Basic Types
//**************************************
#if defined (__STDC_VERSION__) && __STDC_VERSION__ >= 199901L // C99
# include <stdint.h>
typedef uint8_t BYTE;
typedef uint16_t U16;
typedef uint32_t U32;
typedef int32_t S32;
typedef uint64_t U64;
#else
typedef unsigned char BYTE;
typedef unsigned short U16;
typedef unsigned int U32;
typedef signed int S32;
typedef unsigned long long U64;
#endif
#if defined(__GNUC__) && !defined(XXH_USE_UNALIGNED_ACCESS)
# define _PACKED __attribute__ ((packed))
#else
# define _PACKED
#endif
#if !defined(XXH_USE_UNALIGNED_ACCESS) && !defined(__GNUC__)
# ifdef __IBMC__
# pragma pack(1)
# else
# pragma pack(push, 1)
# endif
#endif
typedef struct _U32_S { U32 v; } _PACKED U32_S;
#if !defined(XXH_USE_UNALIGNED_ACCESS) && !defined(__GNUC__)
# pragma pack(pop)
#endif
#define A32(x) (((U32_S *)(x))->v)
//***************************************
// Compiler-specific Functions and Macros
//***************************************
#define GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
// Note : although _rotl exists for minGW (GCC under windows), performance seems poor
#if defined(_MSC_VER)
# define XXH_rotl32(x,r) _rotl(x,r)
#else
# define XXH_rotl32(x,r) ((x << r) | (x >> (32 - r)))
#endif
#if defined(_MSC_VER) // Visual Studio
# define XXH_swap32 _byteswap_ulong
#elif GCC_VERSION >= 403
# define XXH_swap32 __builtin_bswap32
#else
static inline U32 XXH_swap32 (U32 x) {
return ((x << 24) & 0xff000000 ) |
((x << 8) & 0x00ff0000 ) |
((x >> 8) & 0x0000ff00 ) |
((x >> 24) & 0x000000ff );}
#endif
//**************************************
// Constants
//**************************************
#define PRIME32_1 2654435761U
#define PRIME32_2 2246822519U
#define PRIME32_3 3266489917U
#define PRIME32_4 668265263U
#define PRIME32_5 374761393U
//**************************************
// Architecture Macros
//**************************************
typedef enum { XXH_bigEndian=0, XXH_littleEndian=1 } XXH_endianess;
#ifndef XXH_CPU_LITTLE_ENDIAN // It is possible to define XXH_CPU_LITTLE_ENDIAN externally, for example using a compiler switch
static const int one = 1;
# define XXH_CPU_LITTLE_ENDIAN (*(char*)(&one))
#endif
//**************************************
// Macros
//**************************************
#define XXH_STATIC_ASSERT(c) { enum { XXH_static_assert = 1/(!!(c)) }; } // use only *after* variable declarations
//****************************
// Memory reads
//****************************
typedef enum { XXH_aligned, XXH_unaligned } XXH_alignment;
FORCE_INLINE U32 XXH_readLE32_align(const U32* ptr, XXH_endianess endian, XXH_alignment align)
{
if (align==XXH_unaligned)
return endian==XXH_littleEndian ? A32(ptr) : XXH_swap32(A32(ptr));
else
return endian==XXH_littleEndian ? *ptr : XXH_swap32(*ptr);
}
FORCE_INLINE U32 XXH_readLE32(const U32* ptr, XXH_endianess endian) { return XXH_readLE32_align(ptr, endian, XXH_unaligned); }
//****************************
// Simple Hash Functions
//****************************
FORCE_INLINE U32 XXH32_endian_align(const void* input, int len, U32 seed, XXH_endianess endian, XXH_alignment align)
{
const BYTE* p = (const BYTE*)input;
const BYTE* const bEnd = p + len;
U32 h32;
#ifdef XXH_ACCEPT_NULL_INPUT_POINTER
if (p==NULL) { len=0; p=(const BYTE*)(size_t)16; }
#endif
if (len>=16)
{
const BYTE* const limit = bEnd - 16;
U32 v1 = seed + PRIME32_1 + PRIME32_2;
U32 v2 = seed + PRIME32_2;
U32 v3 = seed + 0;
U32 v4 = seed - PRIME32_1;
do
{
v1 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_2; v1 = XXH_rotl32(v1, 13); v1 *= PRIME32_1; p+=4;
v2 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_2; v2 = XXH_rotl32(v2, 13); v2 *= PRIME32_1; p+=4;
v3 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_2; v3 = XXH_rotl32(v3, 13); v3 *= PRIME32_1; p+=4;
v4 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_2; v4 = XXH_rotl32(v4, 13); v4 *= PRIME32_1; p+=4;
} while (p<=limit);
h32 = XXH_rotl32(v1, 1) + XXH_rotl32(v2, 7) + XXH_rotl32(v3, 12) + XXH_rotl32(v4, 18);
}
else
{
h32 = seed + PRIME32_5;
}
h32 += (U32) len;
while (p<=bEnd-4)
{
h32 += XXH_readLE32_align((const U32*)p, endian, align) * PRIME32_3;
h32 = XXH_rotl32(h32, 17) * PRIME32_4 ;
p+=4;
}
while (p<bEnd)
{
h32 += (*p) * PRIME32_5;
h32 = XXH_rotl32(h32, 11) * PRIME32_1 ;
p++;
}
h32 ^= h32 >> 15;
h32 *= PRIME32_2;
h32 ^= h32 >> 13;
h32 *= PRIME32_3;
h32 ^= h32 >> 16;
return h32;
}
U32 XXH32(const void* input, int len, U32 seed)
{
#if 0
// Simple version, good for code maintenance, but unfortunately slow for small inputs
void* state = XXH32_init(seed);
XXH32_update(state, input, len);
return XXH32_digest(state);
#else
XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
# if !defined(XXH_USE_UNALIGNED_ACCESS)
if ((((size_t)input) & 3)) // Input is aligned, let's leverage the speed advantage
{
if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_aligned);
else
return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_aligned);
}
# endif
if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
return XXH32_endian_align(input, len, seed, XXH_littleEndian, XXH_unaligned);
else
return XXH32_endian_align(input, len, seed, XXH_bigEndian, XXH_unaligned);
#endif
}
//****************************
// Advanced Hash Functions
//****************************
struct XXH_state32_t
{
U64 total_len;
U32 seed;
U32 v1;
U32 v2;
U32 v3;
U32 v4;
int memsize;
char memory[16];
};
int XXH32_sizeofState()
{
XXH_STATIC_ASSERT(XXH32_SIZEOFSTATE >= sizeof(struct XXH_state32_t)); // A compilation error here means XXH32_SIZEOFSTATE is not large enough
return sizeof(struct XXH_state32_t);
}
XXH_errorcode XXH32_resetState(void* state_in, U32 seed)
{
struct XXH_state32_t * state = (struct XXH_state32_t *) state_in;
state->seed = seed;
state->v1 = seed + PRIME32_1 + PRIME32_2;
state->v2 = seed + PRIME32_2;
state->v3 = seed + 0;
state->v4 = seed - PRIME32_1;
state->total_len = 0;
state->memsize = 0;
return XXH_OK;
}
void* XXH32_init (U32 seed)
{
void* state = XXH_malloc (sizeof(struct XXH_state32_t));
XXH32_resetState(state, seed);
return state;
}
FORCE_INLINE XXH_errorcode XXH32_update_endian (void* state_in, const void* input, int len, XXH_endianess endian)
{
struct XXH_state32_t * state = (struct XXH_state32_t *) state_in;
const BYTE* p = (const BYTE*)input;
const BYTE* const bEnd = p + len;
#ifdef XXH_ACCEPT_NULL_INPUT_POINTER
if (input==NULL) return XXH_ERROR;
#endif
state->total_len += len;
if (state->memsize + len < 16) // fill in tmp buffer
{
XXH_memcpy(state->memory + state->memsize, input, len);
state->memsize += len;
return XXH_OK;
}
if (state->memsize) // some data left from previous update
{
XXH_memcpy(state->memory + state->memsize, input, 16-state->memsize);
{
const U32* p32 = (const U32*)state->memory;
state->v1 += XXH_readLE32(p32, endian) * PRIME32_2; state->v1 = XXH_rotl32(state->v1, 13); state->v1 *= PRIME32_1; p32++;
state->v2 += XXH_readLE32(p32, endian) * PRIME32_2; state->v2 = XXH_rotl32(state->v2, 13); state->v2 *= PRIME32_1; p32++;
state->v3 += XXH_readLE32(p32, endian) * PRIME32_2; state->v3 = XXH_rotl32(state->v3, 13); state->v3 *= PRIME32_1; p32++;
state->v4 += XXH_readLE32(p32, endian) * PRIME32_2; state->v4 = XXH_rotl32(state->v4, 13); state->v4 *= PRIME32_1; p32++;
}
p += 16-state->memsize;
state->memsize = 0;
}
if (p <= bEnd-16)
{
const BYTE* const limit = bEnd - 16;
U32 v1 = state->v1;
U32 v2 = state->v2;
U32 v3 = state->v3;
U32 v4 = state->v4;
do
{
v1 += XXH_readLE32((const U32*)p, endian) * PRIME32_2; v1 = XXH_rotl32(v1, 13); v1 *= PRIME32_1; p+=4;
v2 += XXH_readLE32((const U32*)p, endian) * PRIME32_2; v2 = XXH_rotl32(v2, 13); v2 *= PRIME32_1; p+=4;
v3 += XXH_readLE32((const U32*)p, endian) * PRIME32_2; v3 = XXH_rotl32(v3, 13); v3 *= PRIME32_1; p+=4;
v4 += XXH_readLE32((const U32*)p, endian) * PRIME32_2; v4 = XXH_rotl32(v4, 13); v4 *= PRIME32_1; p+=4;
} while (p<=limit);
state->v1 = v1;
state->v2 = v2;
state->v3 = v3;
state->v4 = v4;
}
if (p < bEnd)
{
XXH_memcpy(state->memory, p, bEnd-p);
state->memsize = (int)(bEnd-p);
}
return XXH_OK;
}
XXH_errorcode XXH32_update (void* state_in, const void* input, int len)
{
XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
return XXH32_update_endian(state_in, input, len, XXH_littleEndian);
else
return XXH32_update_endian(state_in, input, len, XXH_bigEndian);
}
FORCE_INLINE U32 XXH32_intermediateDigest_endian (void* state_in, XXH_endianess endian)
{
struct XXH_state32_t * state = (struct XXH_state32_t *) state_in;
const BYTE * p = (const BYTE*)state->memory;
BYTE* bEnd = (BYTE*)state->memory + state->memsize;
U32 h32;
if (state->total_len >= 16)
{
h32 = XXH_rotl32(state->v1, 1) + XXH_rotl32(state->v2, 7) + XXH_rotl32(state->v3, 12) + XXH_rotl32(state->v4, 18);
}
else
{
h32 = state->seed + PRIME32_5;
}
h32 += (U32) state->total_len;
while (p<=bEnd-4)
{
h32 += XXH_readLE32((const U32*)p, endian) * PRIME32_3;
h32 = XXH_rotl32(h32, 17) * PRIME32_4;
p+=4;
}
while (p<bEnd)
{
h32 += (*p) * PRIME32_5;
h32 = XXH_rotl32(h32, 11) * PRIME32_1;
p++;
}
h32 ^= h32 >> 15;
h32 *= PRIME32_2;
h32 ^= h32 >> 13;
h32 *= PRIME32_3;
h32 ^= h32 >> 16;
return h32;
}
U32 XXH32_intermediateDigest (void* state_in)
{
XXH_endianess endian_detected = (XXH_endianess)XXH_CPU_LITTLE_ENDIAN;
if ((endian_detected==XXH_littleEndian) || XXH_FORCE_NATIVE_FORMAT)
return XXH32_intermediateDigest_endian(state_in, XXH_littleEndian);
else
return XXH32_intermediateDigest_endian(state_in, XXH_bigEndian);
}
U32 XXH32_digest (void* state_in)
{
U32 h32 = XXH32_intermediateDigest(state_in);
XXH_free(state_in);
return h32;
}

328
xxhash.h
View File

@ -1,164 +1,164 @@
/*
xxHash - Fast Hash algorithm
Header File
Copyright (C) 2012-2013, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- xxHash source repository : http://code.google.com/p/xxhash/
*/
/* Notice extracted from xxHash homepage :
xxHash is an extremely fast Hash algorithm, running at RAM speed limits.
It also successfully passes all tests from the SMHasher suite.
Comparison (single thread, Windows Seven 32 bits, using SMHasher on a Core 2 Duo @3GHz)
Name Speed Q.Score Author
xxHash 5.4 GB/s 10
CrapWow 3.2 GB/s 2 Andrew
MumurHash 3a 2.7 GB/s 10 Austin Appleby
SpookyHash 2.0 GB/s 10 Bob Jenkins
SBox 1.4 GB/s 9 Bret Mulvey
Lookup3 1.2 GB/s 9 Bob Jenkins
SuperFastHash 1.2 GB/s 1 Paul Hsieh
CityHash64 1.05 GB/s 10 Pike & Alakuijala
FNV 0.55 GB/s 5 Fowler, Noll, Vo
CRC32 0.43 GB/s 9
MD5-32 0.33 GB/s 10 Ronald L. Rivest
SHA1-32 0.28 GB/s 10
Q.Score is a measure of quality of the hash function.
It depends on successfully passing SMHasher test set.
10 is a perfect score.
*/
#pragma once
#if defined (__cplusplus)
extern "C" {
#endif
//****************************
// Type
//****************************
typedef enum { XXH_OK=0, XXH_ERROR } XXH_errorcode;
//****************************
// Simple Hash Functions
//****************************
unsigned int XXH32 (const void* input, int len, unsigned int seed);
/*
XXH32() :
Calculate the 32-bits hash of sequence of length "len" stored at memory address "input".
The memory between input & input+len must be valid (allocated and read-accessible).
"seed" can be used to alter the result predictably.
This function successfully passes all SMHasher tests.
Speed on Core 2 Duo @ 3 GHz (single thread, SMHasher benchmark) : 5.4 GB/s
Note that "len" is type "int", which means it is limited to 2^31-1.
If your data is larger, use the advanced functions below.
*/
//****************************
// Advanced Hash Functions
//****************************
void* XXH32_init (unsigned int seed);
XXH_errorcode XXH32_update (void* state, const void* input, int len);
unsigned int XXH32_digest (void* state);
/*
These functions calculate the xxhash of an input provided in several small packets,
as opposed to an input provided as a single block.
It must be started with :
void* XXH32_init()
The function returns a pointer which holds the state of calculation.
This pointer must be provided as "void* state" parameter for XXH32_update().
XXH32_update() can be called as many times as necessary.
The user must provide a valid (allocated) input.
The function returns an error code, with 0 meaning OK, and any other value meaning there is an error.
Note that "len" is type "int", which means it is limited to 2^31-1.
If your data is larger, it is recommended to chunk your data into blocks
of size for example 2^30 (1GB) to avoid any "int" overflow issue.
Finally, you can end the calculation anytime, by using XXH32_digest().
This function returns the final 32-bits hash.
You must provide the same "void* state" parameter created by XXH32_init().
Memory will be freed by XXH32_digest().
*/
int XXH32_sizeofState();
XXH_errorcode XXH32_resetState(void* state, unsigned int seed);
#define XXH32_SIZEOFSTATE 48
typedef struct { long long ll[(XXH32_SIZEOFSTATE+(sizeof(long long)-1))/sizeof(long long)]; } XXH32_stateSpace_t;
/*
These functions allow user application to make its own allocation for state.
XXH32_sizeofState() is used to know how much space must be allocated for the xxHash 32-bits state.
Note that the state must be aligned to access 'long long' fields. Memory must be allocated and referenced by a pointer.
This pointer must then be provided as 'state' into XXH32_resetState(), which initializes the state.
For static allocation purposes (such as allocation on stack, or freestanding systems without malloc()),
use the structure XXH32_stateSpace_t, which will ensure that memory space is large enough and correctly aligned to access 'long long' fields.
*/
unsigned int XXH32_intermediateDigest (void* state);
/*
This function does the same as XXH32_digest(), generating a 32-bit hash,
but preserve memory context.
This way, it becomes possible to generate intermediate hashes, and then continue feeding data with XXH32_update().
To free memory context, use XXH32_digest(), or free().
*/
//****************************
// Deprecated function names
//****************************
// The following translations are provided to ease code transition
// You are encouraged to no longer this function names
#define XXH32_feed XXH32_update
#define XXH32_result XXH32_digest
#define XXH32_getIntermediateResult XXH32_intermediateDigest
#if defined (__cplusplus)
}
#endif
/*
xxHash - Fast Hash algorithm
Header File
Copyright (C) 2012-2014, Yann Collet.
BSD 2-Clause License (http://www.opensource.org/licenses/bsd-license.php)
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
You can contact the author at :
- xxHash source repository : http://code.google.com/p/xxhash/
*/
/* Notice extracted from xxHash homepage :
xxHash is an extremely fast Hash algorithm, running at RAM speed limits.
It also successfully passes all tests from the SMHasher suite.
Comparison (single thread, Windows Seven 32 bits, using SMHasher on a Core 2 Duo @3GHz)
Name Speed Q.Score Author
xxHash 5.4 GB/s 10
CrapWow 3.2 GB/s 2 Andrew
MumurHash 3a 2.7 GB/s 10 Austin Appleby
SpookyHash 2.0 GB/s 10 Bob Jenkins
SBox 1.4 GB/s 9 Bret Mulvey
Lookup3 1.2 GB/s 9 Bob Jenkins
SuperFastHash 1.2 GB/s 1 Paul Hsieh
CityHash64 1.05 GB/s 10 Pike & Alakuijala
FNV 0.55 GB/s 5 Fowler, Noll, Vo
CRC32 0.43 GB/s 9
MD5-32 0.33 GB/s 10 Ronald L. Rivest
SHA1-32 0.28 GB/s 10
Q.Score is a measure of quality of the hash function.
It depends on successfully passing SMHasher test set.
10 is a perfect score.
*/
#pragma once
#if defined (__cplusplus)
extern "C" {
#endif
//****************************
// Type
//****************************
typedef enum { XXH_OK=0, XXH_ERROR } XXH_errorcode;
//****************************
// Simple Hash Functions
//****************************
unsigned int XXH32 (const void* input, int len, unsigned int seed);
/*
XXH32() :
Calculate the 32-bits hash of sequence of length "len" stored at memory address "input".
The memory between input & input+len must be valid (allocated and read-accessible).
"seed" can be used to alter the result predictably.
This function successfully passes all SMHasher tests.
Speed on Core 2 Duo @ 3 GHz (single thread, SMHasher benchmark) : 5.4 GB/s
Note that "len" is type "int", which means it is limited to 2^31-1.
If your data is larger, use the advanced functions below.
*/
//****************************
// Advanced Hash Functions
//****************************
void* XXH32_init (unsigned int seed);
XXH_errorcode XXH32_update (void* state, const void* input, int len);
unsigned int XXH32_digest (void* state);
/*
These functions calculate the xxhash of an input provided in several small packets,
as opposed to an input provided as a single block.
It must be started with :
void* XXH32_init()
The function returns a pointer which holds the state of calculation.
This pointer must be provided as "void* state" parameter for XXH32_update().
XXH32_update() can be called as many times as necessary.
The user must provide a valid (allocated) input.
The function returns an error code, with 0 meaning OK, and any other value meaning there is an error.
Note that "len" is type "int", which means it is limited to 2^31-1.
If your data is larger, it is recommended to chunk your data into blocks
of size for example 2^30 (1GB) to avoid any "int" overflow issue.
Finally, you can end the calculation anytime, by using XXH32_digest().
This function returns the final 32-bits hash.
You must provide the same "void* state" parameter created by XXH32_init().
Memory will be freed by XXH32_digest().
*/
int XXH32_sizeofState();
XXH_errorcode XXH32_resetState(void* state, unsigned int seed);
#define XXH32_SIZEOFSTATE 48
typedef struct { long long ll[(XXH32_SIZEOFSTATE+(sizeof(long long)-1))/sizeof(long long)]; } XXH32_stateSpace_t;
/*
These functions allow user application to make its own allocation for state.
XXH32_sizeofState() is used to know how much space must be allocated for the xxHash 32-bits state.
Note that the state must be aligned to access 'long long' fields. Memory must be allocated and referenced by a pointer.
This pointer must then be provided as 'state' into XXH32_resetState(), which initializes the state.
For static allocation purposes (such as allocation on stack, or freestanding systems without malloc()),
use the structure XXH32_stateSpace_t, which will ensure that memory space is large enough and correctly aligned to access 'long long' fields.
*/
unsigned int XXH32_intermediateDigest (void* state);
/*
This function does the same as XXH32_digest(), generating a 32-bit hash,
but preserve memory context.
This way, it becomes possible to generate intermediate hashes, and then continue feeding data with XXH32_update().
To free memory context, use XXH32_digest(), or free().
*/
//****************************
// Deprecated function names
//****************************
// The following translations are provided to ease code transition
// You are encouraged to no longer this function names
#define XXH32_feed XXH32_update
#define XXH32_result XXH32_digest
#define XXH32_getIntermediateResult XXH32_intermediateDigest
#if defined (__cplusplus)
}
#endif