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

Merge pull request #26 from Cyan4973/frame

Browse Source 
Frame
This commit is contained in:
Yann Collet 2014-09-14 00:30:01 +02:00
commit 3f80a22819
9 changed files with 2304 additions and 64 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
Makefile
View File

@ -31,7 +31,7 @@
# ################################################################
# Version numbers
VERSION=122
VERSION=123
export RELEASE=r$(VERSION)
LIBVER_MAJOR=`sed -n '/define LZ4_VERSION_MAJOR/s/.*[[:blank:]]\([0-9][0-9]*\).*/\1/p' < lz4.h`
LIBVER_MINOR=`sed -n '/define LZ4_VERSION_MINOR/s/.*[[:blank:]]\([0-9][0-9]*\).*/\1/p' < lz4.h`

27
lz4.c
View File

@ -855,12 +855,12 @@ int LZ4_saveDict (LZ4_stream_t* LZ4_dict, char* safeBuffer, int dictSize)
if ((U32)dictSize > 64 KB) dictSize = 64 KB; /* useless to define a dictionary > 64 KB */
if ((U32)dictSize > dict->dictSize) dictSize = dict->dictSize;
memcpy(safeBuffer, previousDictEnd - dictSize, dictSize);
memmove(safeBuffer, previousDictEnd - dictSize, dictSize);
dict->dictionary = (const BYTE*)safeBuffer;
dict->dictSize = (U32)dictSize;
return 1;
return dictSize;
}
@ -870,9 +870,9 @@ int LZ4_saveDict (LZ4_stream_t* LZ4_dict, char* safeBuffer, int dictSize)
****************************/
/*
* This generic decompression function cover all use cases.
* It shall be instanciated several times, using different sets of directives
* It shall be instantiated several times, using different sets of directives
* Note that it is essential this generic function is really inlined,
* in order to remove useless branches during compilation optimisation.
* in order to remove useless branches during compilation optimization.
*/
FORCE_INLINE int LZ4_decompress_generic(
const char* source,
@ -1153,14 +1153,29 @@ Advanced decoding functions :
the dictionary must be explicitly provided within parameters
*/
FORCE_INLINE int LZ4_decompress_usingDict_generic(const char* source, char* dest, int compressedSize, int maxOutputSize, int safe, const char* dictStart, int dictSize)
{
if ((dictStart+dictSize == source) && (dictSize >= (int)(64 KB - 1)))
return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, safe, full, 0, withPrefix64k, NULL, 64 KB);
return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, safe, full, 0, usingExtDict, dictStart, dictSize);
}
int LZ4_decompress_safe_usingDict(const char* source, char* dest, int compressedSize, int maxOutputSize, const char* dictStart, int dictSize)
{
return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, endOnInputSize, full, 0, usingExtDict, dictStart, dictSize);
//return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, endOnInputSize, full, 0, usingExtDict, dictStart, dictSize);
return LZ4_decompress_usingDict_generic(source, dest, compressedSize, maxOutputSize, 1, dictStart, dictSize);
}
int LZ4_decompress_fast_usingDict(const char* source, char* dest, int originalSize, const char* dictStart, int dictSize)
{
return LZ4_decompress_generic(source, dest, 0, originalSize, endOnOutputSize, full, 0, usingExtDict, dictStart, dictSize);
//return LZ4_decompress_generic(source, dest, 0, originalSize, endOnOutputSize, full, 0, usingExtDict, dictStart, dictSize);
return LZ4_decompress_usingDict_generic(source, dest, 0, originalSize, 0, dictStart, dictSize);
}
/* debug function */
int LZ4_decompress_safe_forceExtDict(const char* source, char* dest, int compressedSize, int maxOutputSize, const char* dictStart, int dictSize)
{
return LZ4_decompress_generic(source, dest, compressedSize, maxOutputSize, endOnInputSize, full, 0, usingExtDict, dictStart, dictSize);
}

27
lz4.h
View File

@ -37,13 +37,18 @@
extern "C" {
#endif
/*
* lz4.h provides raw compression format functions, for optimal performance and integration into programs.
* If you need to generate data using an inter-operable format (respecting the framing specification),
* please use lz4frame.h instead.
*/
/**************************************
Version
**************************************/
#define LZ4_VERSION_MAJOR 1 /* for major interface/format changes */
#define LZ4_VERSION_MINOR 3 /* for minor interface/format changes */
#define LZ4_VERSION_RELEASE 0 /* for tweaks, bug-fixes, or development */
#define LZ4_VERSION_RELEASE 1 /* for tweaks, bug-fixes, or development */
#define LZ4_VERSION_NUMBER (LZ4_VERSION_MAJOR *100*100 + LZ4_VERSION_MINOR *100 + LZ4_VERSION_RELEASE)
int LZ4_versionNumber (void);
@ -142,7 +147,7 @@ LZ4_decompress_fast() :
Destination buffer must be already allocated. Its size must be a minimum of 'originalSize' bytes.
note : This function fully respect memory boundaries for properly formed compressed data.
It is a bit faster than LZ4_decompress_safe().
However, it does not provide any protection against intentionnally modified data stream (malicious input).
However, it does not provide any protection against intentionally modified data stream (malicious input).
Use this function in trusted environment only (data to decode comes from a trusted source).
*/
int LZ4_decompress_fast (const char* source, char* dest, int originalSize);
@ -178,10 +183,9 @@ typedef struct { unsigned int table[LZ4_STREAMSIZE_U32]; } LZ4_stream_t;
/*
* LZ4_resetStream
* Use this function to init a newly allocated LZ4_stream_t structure
* You can also reset an existing LZ4_stream_t structure
* Use this function to init an allocated LZ4_stream_t structure
*/
void LZ4_resetStream (LZ4_stream_t* LZ4_stream);
void LZ4_resetStream (LZ4_stream_t* LZ4_streamPtr);
/*
* If you prefer dynamic allocation methods,
@ -217,10 +221,10 @@ int LZ4_compress_limitedOutput_continue (LZ4_stream_t* LZ4_stream, const char* s
/*
* LZ4_saveDict
* If previously compressed data block is not guaranteed to remain available at its memory location
* save it into a safe place (char* safeBuffer)
* save it into a safer place (char* safeBuffer)
* Note : you don't need to call LZ4_loadDict() afterwards,
* dictionary is immediately usable, you can therefore call again LZ4_compress_continue()
* Return : 1 if OK, 0 if error
* Return : dictionary size in bytes, or 0 if error
* Note : any dictSize > 64 KB will be interpreted as 64KB.
*/
int LZ4_saveDict (LZ4_stream_t* LZ4_stream, char* safeBuffer, int dictSize);
@ -262,9 +266,9 @@ int LZ4_freeStreamDecode (LZ4_streamDecode_t* LZ4_stream);
These decoding functions allow decompression of multiple blocks in "streaming" mode.
Previously decoded blocks must still be available at the memory position where they were decoded.
If it's not possible, save the relevant part of decoded data into a safe buffer,
and indicate where its new address using LZ4_setDictDecode()
and indicate where its new address using LZ4_setStreamDecode()
*/
int LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int compressedSize, int maxOutputSize);
int LZ4_decompress_safe_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int compressedSize, int maxDecompressedSize);
int LZ4_decompress_fast_continue (LZ4_streamDecode_t* LZ4_streamDecode, const char* source, char* dest, int originalSize);
@ -273,10 +277,9 @@ Advanced decoding functions :
*_usingDict() :
These decoding functions work the same as
a combination of LZ4_setDictDecode() followed by LZ4_decompress_x_continue()
all together into a single function call.
It doesn't use nor update an LZ4_streamDecode_t structure.
They don't use nor update an LZ4_streamDecode_t structure.
*/
int LZ4_decompress_safe_usingDict (const char* source, char* dest, int compressedSize, int maxOutputSize, const char* dictStart, int dictSize);
int LZ4_decompress_safe_usingDict (const char* source, char* dest, int compressedSize, int maxDecompressedSize, const char* dictStart, int dictSize);
int LZ4_decompress_fast_usingDict (const char* source, char* dest, int originalSize, const char* dictStart, int dictSize);

1127
lz4frame.c Normal file
View File

File diff suppressed because it is too large Load Diff

258
lz4frame.h Normal file
View File

@ -0,0 +1,258 @@
/*
LZ4 auto-framing library
Header File
Copyright (C) 2011-2014, 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 :
- LZ4 source repository : http://code.google.com/p/lz4/
- LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c
*/
/* LZ4F is a stand-alone API to create LZ4-compressed frames
* fully conformant to specification v1.4.1.
* All related operations, including memory management, are handled by the library.
* You don't need lz4.h when using lz4frame.h.
* */
#pragma once
#if defined (__cplusplus)
extern "C" {
#endif
/****************************************
Note : experimental API.
Not yet integrated within lz4 library.
****************************************/
/**************************************
Includes
**************************************/
#include <stddef.h> /* size_t */
/**************************************
Error management
**************************************/
typedef size_t LZ4F_errorCode_t;
typedef enum { OK_FrameEnd = 1 } LZ4F_successCodes;
typedef enum { OK_NoError = 0, ERROR_GENERIC = 1,
ERROR_maxBlockSize_invalid, ERROR_blockMode_invalid, ERROR_contentChecksumFlag_invalid,
ERROR_compressionLevel_invalid,
ERROR_allocation_failed,
ERROR_srcSize_tooLarge, ERROR_dstMaxSize_tooSmall,
ERROR_checksum_invalid,
ERROR_maxCode
} LZ4F_errorCodes; /* error codes are negative unsigned values.
Compare function result to (-specificCode) */
int LZ4F_isError(LZ4F_errorCode_t code); /* Basically : code > -ERROR_maxCode */
/**************************************
Framing compression functions
**************************************/
typedef enum { LZ4F_default=0, max64KB=4, max256KB=5, max1MB=6, max4MB=7} blockSizeID_t;
typedef enum { blockLinked=0, blockIndependent} blockMode_t;
typedef enum { noContentChecksum=0, contentChecksumEnabled } contentChecksum_t;
typedef struct {
blockSizeID_t blockSizeID; /* max64KB, max256KB, max1MB, max4MB ; 0 == default */
blockMode_t blockMode; /* blockLinked, blockIndependent ; 0 == default */
contentChecksum_t contentChecksumFlag; /* noContentChecksum, contentChecksumEnabled ; 0 == default */
unsigned reserved[5];
} LZ4F_frameInfo_t;
typedef struct {
LZ4F_frameInfo_t frameInfo;
unsigned compressionLevel; /* from 0 to 16 */
unsigned autoFlush; /* 1 == always flush; reduce need for tmp buffer */
unsigned reserved[4];
} LZ4F_preferences_t;
/***********************************
* Simple compression function
* *********************************/
size_t LZ4F_compressFrameBound(size_t srcSize, const LZ4F_preferences_t* preferencesPtr);
size_t LZ4F_compressFrame(void* dstBuffer, size_t dstMaxSize, const void* srcBuffer, size_t srcSize, const LZ4F_preferences_t* preferencesPtr);
/* LZ4F_compressFrame()
* Compress an entire srcBuffer into a valid LZ4 frame, as defined by specification v1.4.1, in a single step.
* The most important rule is that dstBuffer MUST be large enough (dstMaxSize) to ensure compression completion even in worst case.
* You can get the minimum value of dstMaxSize by using LZ4F_compressFrameBound()
* If this condition is not respected, LZ4F_compressFrame() will fail (result is an errorCode)
* The LZ4F_preferences_t structure is optional : you can provide NULL as argument. All preferences will be set to default.
* The result of the function is the number of bytes written into dstBuffer.
* The function outputs an error code if it fails (can be tested using LZ4F_isError())
*/
/**********************************
* Advanced compression functions
* *********************************/
typedef void* LZ4F_compressionContext_t;
typedef struct {
unsigned stableSrc; /* 1 == src content will remain available on future calls to LZ4F_compress(); avoid saving src content within tmp buffer as future dictionary */
unsigned reserved[3];
} LZ4F_compressOptions_t;
/* Resource Management */
#define LZ4F_VERSION 100
LZ4F_errorCode_t LZ4F_createCompressionContext(LZ4F_compressionContext_t* LZ4F_compressionContextPtr, unsigned version);
LZ4F_errorCode_t LZ4F_freeCompressionContext(LZ4F_compressionContext_t LZ4F_compressionContext);
/* LZ4F_createCompressionContext() :
* The first thing to do is to create a compressionContext object, which will be used in all compression operations.
* This is achieved using LZ4F_createCompressionContext(), which takes as argument a version and an LZ4F_preferences_t structure.
* The version provided MUST be LZ4F_VERSION. It is intended to track potential version differences between different binaries.
* The function will provide a pointer to a fully allocated LZ4F_compressionContext_t object.
* If the result LZ4F_errorCode_t is not zero, there was an error during context creation.
* Object can release its memory using LZ4F_freeCompressionContext();
*/
/* Compression */
size_t LZ4F_compressBegin(LZ4F_compressionContext_t compressionContext, void* dstBuffer, size_t dstMaxSize, const LZ4F_preferences_t* preferencesPtr);
/* LZ4F_compressBegin() :
* will write the frame header into dstBuffer.
* dstBuffer must be large enough to accommodate a header (dstMaxSize). Maximum header size is 19 bytes.
* The LZ4F_preferences_t structure is optional : you can provide NULL as argument, all preferences will then be set to default.
* The result of the function is the number of bytes written into dstBuffer for the header
* or an error code (can be tested using LZ4F_isError())
*/
size_t LZ4F_compressBound(size_t srcSize, const LZ4F_preferences_t* preferencesPtr);
/* LZ4F_compressBound() :
* Provides the minimum size of Dst buffer given srcSize to handle worst case situations.
* preferencesPtr is optional : you can provide NULL as argument, all preferences will then be set to default.
*/
size_t LZ4F_compressUpdate(LZ4F_compressionContext_t compressionContext, void* dstBuffer, size_t dstMaxSize, const void* srcBuffer, size_t srcSize, const LZ4F_compressOptions_t* compressOptionsPtr);
/* LZ4F_compressUpdate()
* LZ4F_compressUpdate() can be called repetitively to compress as much data as necessary.
* The most important rule is that dstBuffer MUST be large enough (dstMaxSize) to ensure compression completion even in worst case.
* If this condition is not respected, LZ4F_compress() will fail (result is an errorCode)
* You can get the minimum value of dstMaxSize by using LZ4F_compressBound()
* The LZ4F_compressOptions_t structure is optional : you can provide NULL as argument.
* The result of the function is the number of bytes written into dstBuffer : it can be zero, meaning input data was just buffered.
* The function outputs an error code if it fails (can be tested using LZ4F_isError())
*/
size_t LZ4F_flush(LZ4F_compressionContext_t compressionContext, void* dstBuffer, size_t dstMaxSize, const LZ4F_compressOptions_t* compressOptionsPtr);
/* LZ4F_flush()
* Should you need to create compressed data immediately, without waiting for a block to be filled,
* you can call LZ4_flush(), which will immediately compress any remaining data buffered within compressionContext.
* The LZ4F_compressOptions_t structure is optional : you can provide NULL as argument.
* The result of the function is the number of bytes written into dstBuffer
* (it can be zero, this means there was no data left within compressionContext)
* The function outputs an error code if it fails (can be tested using LZ4F_isError())
*/
size_t LZ4F_compressEnd(LZ4F_compressionContext_t compressionContext, void* dstBuffer, size_t dstMaxSize, const LZ4F_compressOptions_t* compressOptionsPtr);
/* LZ4F_compressEnd()
* When you want to properly finish the compressed frame, just call LZ4F_compressEnd().
* It will flush whatever data remained within compressionContext (like LZ4_flush())
* but also properly finalize the frame, with an endMark and a checksum.
* The result of the function is the number of bytes written into dstBuffer (necessarily >= 4 (endMark size))
* The function outputs an error code if it fails (can be tested using LZ4F_isError())
* The LZ4F_compressOptions_t structure is optional : you can provide NULL as argument.
* compressionContext can then be used again, starting with LZ4F_compressBegin(). The preferences will remain the same.
*/
/***********************************
* Decompression functions
* *********************************/
typedef void* LZ4F_decompressionContext_t;
typedef struct {
unsigned stableDst; /* unused for the time being, must be 0 */
unsigned reserved[3];
} LZ4F_decompressOptions_t;
/* Resource management */
LZ4F_errorCode_t LZ4F_createDecompressionContext(LZ4F_compressionContext_t* LZ4F_decompressionContextPtr, unsigned version);
LZ4F_errorCode_t LZ4F_freeDecompressionContext(LZ4F_compressionContext_t LZ4F_decompressionContext);
/* LZ4F_createDecompressionContext() :
* The first thing to do is to create a decompressionContext object, which will be used in all decompression operations.
* This is achieved using LZ4F_createDecompressionContext().
* The function will provide a pointer to a fully allocated and initialized LZ4F_decompressionContext object.
* If the result LZ4F_errorCode_t is not OK_NoError, there was an error during context creation.
* Object can release its memory using LZ4F_freeDecompressionContext();
*/
/* Decompression */
size_t LZ4F_getFrameInfo(LZ4F_decompressionContext_t decompressionContext, LZ4F_frameInfo_t* frameInfoPtr, const void* srcBuffer, size_t* srcSizePtr);
/* LZ4F_getFrameInfo()
* This function decodes frame header information, such as blockSize.
* It is optional : you could start by calling directly LZ4F_decompress() instead.
* The objective is to extract header information without starting decompression, typically for allocation purposes.
* LZ4F_getFrameInfo() can also be used *after* starting decompression, on a valid LZ4F_decompressionContext_t.
* The number of bytes read from srcBuffer will be provided within *srcSizePtr (necessarily <= original value).
* You are expected to resume decompression from where it stopped (srcBuffer + *srcSizePtr)
* The function result is an hint of the better srcSize to use for next call to LZ4F_decompress,
* or an error code which can be tested using LZ4F_isError().
*/
size_t LZ4F_decompress(LZ4F_decompressionContext_t decompressionContext, void* dstBuffer, size_t* dstSizePtr, const void* srcBuffer, size_t* srcSizePtr, const LZ4F_decompressOptions_t* decompressOptionsPtr);
/* LZ4F_decompress()
* Call this function repetitively to regenerate data compressed within srcBuffer.
* The function will attempt to decode *srcSizePtr bytes from srcBuffer, into dstBuffer of maximum size *dstSizePtr.
*
* The number of bytes regenerated into dstBuffer will be provided within *dstSizePtr (necessarily <= original value).
*
* The number of bytes effectively used from srcBuffer will be provided within *srcSizePtr (necessarily <= original value).
* If the number of bytes read is < number of bytes provided, then the decompression operation is not complete.
* This typically happens when dstBuffer is not large enough to contain all decoded data.
* LZ4F_decompress() will have to be called again, starting from where it stopped (srcBuffer + *srcSizePtr)
* The function will check this condition, and refuse to continue if it is not respected.
* dstBuffer is supposed to be flushed between calls to the function, since its content will be rewritten.
* Different dst arguments can be used between each calls.
*
* The function result is an hint of the better srcSize to use for next call to LZ4F_decompress.
* Basically, it's the size of the current (or remaining) compressed block + header of next block.
* Respecting the hint provides some boost to performance, since it allows less buffer shuffling.
* Note that this is just a hint, you can always provide any srcSize you want.
* When a frame is fully decoded, the function result will be 0.
* If decompression failed, function result is an error code which can be tested using LZ4F_isError().
*/
#if defined (__cplusplus)
}
#endif

22
programs/Makefile
View File

@ -30,7 +30,7 @@
# fullbench32: Same as fullbench, but forced to compile in 32-bits mode
# ##########################################################################
RELEASE=r122
RELEASE=r123
DESTDIR?=
PREFIX ?= /usr
@ -68,7 +68,7 @@ endif
default: lz4 lz4c
all: lz4 lz4c lz4c32 fullbench fullbench32 fuzzer fuzzer32 datagen
all: lz4 lz4c lz4c32 fullbench fullbench32 fuzzer fuzzer32 frametest datagen
lz4: $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c bench.c xxhash.c lz4io.c lz4cli.c
$(CC) $(FLAGS) -DDISABLE_LZ4C_LEGACY_OPTIONS $^ -o $@$(EXT)
@ -79,10 +79,10 @@ lz4c : $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c bench.c xxhash.c lz4io.c lz4cli.c
lz4c32: $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c bench.c xxhash.c lz4io.c lz4cli.c
$(CC) -m32 $(FLAGS) $^ -o $@$(EXT)
fullbench : $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c xxhash.c fullbench.c
fullbench : $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c $(LZ4DIR)/lz4frame.c xxhash.c fullbench.c
$(CC) $(FLAGS) $^ -o $@$(EXT)
fullbench32: $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c xxhash.c fullbench.c
fullbench32: $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c $(LZ4DIR)/lz4frame.c xxhash.c fullbench.c
$(CC) -m32 $(FLAGS) $^ -o $@$(EXT)
fuzzer : $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c xxhash.c fuzzer.c
@ -91,6 +91,9 @@ fuzzer : $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c xxhash.c fuzzer.c
fuzzer32: $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c xxhash.c fuzzer.c
$(CC) -m32 $(FLAGS) $^ -o $@$(EXT)
frametest: $(LZ4DIR)/lz4frame.c $(LZ4DIR)/lz4.c $(LZ4DIR)/lz4hc.c xxhash.c frametest.c
$(CC) $(FLAGS) $^ -o $@$(EXT)
datagen : datagen.c
$(CC) $(FLAGS) $^ -o $@$(EXT)
@ -99,7 +102,8 @@ clean:
@rm -f core *.o \
lz4$(EXT) lz4c$(EXT) lz4c32$(EXT) \
fullbench$(EXT) fullbench32$(EXT) \
fuzzer$(EXT) fuzzer32$(EXT) datagen$(EXT)
fuzzer$(EXT) fuzzer32$(EXT) \
frametest$(EXT) datagen$(EXT)
@echo Cleaning completed
@ -128,7 +132,7 @@ uninstall:
[ -f $(DESTDIR)$(MANDIR)/lz4cat.1 ] && rm -f $(DESTDIR)$(MANDIR)/lz4cat.1
@echo lz4 successfully uninstalled
test-native: test-lz4 test-lz4c test-fullbench test-fuzzer test-mem
test-native: test-lz4 test-lz4c test-frame test-fullbench test-fuzzer test-mem
test-force32: test-lz4c32 test-fullbench32 test-fuzzer32 test-mem32
@ -165,7 +169,10 @@ test-fuzzer: fuzzer
test-fuzzer32: fuzzer32
./fuzzer32 --no-prompt
test-mem: lz4 datagen
test-frame: frametest
./frametest
test-mem: lz4 datagen frametest
./datagen -g16KB > tmp
valgrind ./lz4 -9 -BD -f tmp /dev/null
./datagen -g16MB > tmp
@ -173,6 +180,7 @@ test-mem: lz4 datagen
./datagen -g256MB > tmp
valgrind ./lz4 -B4D -f tmp /dev/null
rm tmp
valgrind ./frametest -i100
test-mem32: lz4c32 datagen
# unfortunately, valgrind doesn't seem to work with non-native binary. If someone knows how to do a valgrind-test on a 32-bits exe with a 64-bits system...

619
programs/frametest.c Normal file
View File

@ -0,0 +1,619 @@
/*
frameTest - test tool for lz4frame
Copyright (C) Yann Collet 2014
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 source repository : http://code.google.com/p/lz4/
- LZ4 public forum : https://groups.google.com/forum/#!forum/lz4c
*/
/**************************************
Compiler specific
**************************************/
#define _CRT_SECURE_NO_WARNINGS // fgets
#ifdef _MSC_VER /* Visual Studio */
# pragma warning(disable : 4127) /* disable: C4127: conditional expression is constant */
# pragma warning(disable : 4146) /* disable: C4146: minus unsigned expression */
#endif
#define GCC_VERSION (__GNUC__ * 100 + __GNUC_MINOR__)
#ifdef __GNUC__
# pragma GCC diagnostic ignored "-Wmissing-braces" /* GCC bug 53119 : doesn't accept { 0 } as initializer (https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53119) */
# pragma GCC diagnostic ignored "-Wmissing-field-initializers" /* GCC bug 53119 : doesn't accept { 0 } as initializer (https://gcc.gnu.org/bugzilla/show_bug.cgi?id=53119) */
#endif
/**************************************
Includes
**************************************/
#include <stdlib.h>
#include <stdio.h> // fgets, sscanf
#include <sys/timeb.h> // timeb
#include <string.h> // strcmp
#include "lz4frame.h"
#include "xxhash.h" // XXH64
/**************************************
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
**************************************/
#ifndef LZ4_VERSION
# define LZ4_VERSION ""
#endif
#define KB *(1U<<10)
#define MB *(1U<<20)
#define GB *(1U<<30)
static const U32 nbTestsDefault = 128 KB;
#define COMPRESSIBLE_NOISE_LENGTH (2 MB)
#define FUZ_COMPRESSIBILITY_DEFAULT 50
static const U32 prime1 = 2654435761U;
static const U32 prime2 = 2246822519U;
/**************************************
Macros
**************************************/
#define DISPLAY(...) fprintf(stderr, __VA_ARGS__)
#define DISPLAYLEVEL(l, ...) if (displayLevel>=l) { DISPLAY(__VA_ARGS__); }
#define DISPLAYUPDATE(l, ...) if (displayLevel>=l) { \
if ((FUZ_GetMilliSpan(g_time) > refreshRate) || (displayLevel>=4)) \
{ g_time = FUZ_GetMilliStart(); DISPLAY(__VA_ARGS__); \
if (displayLevel>=4) fflush(stdout); } }
static const U32 refreshRate = 150;
static U32 g_time = 0;
/*****************************************
Local Parameters
*****************************************/
static U32 no_prompt = 0;
static char* programName;
static U32 displayLevel = 2;
/*********************************************************
Fuzzer functions
*********************************************************/
static U32 FUZ_GetMilliStart(void)
{
struct timeb tb;
U32 nCount;
ftime( &tb );
nCount = (U32) (((tb.time & 0xFFFFF) * 1000) + tb.millitm);
return nCount;
}
static U32 FUZ_GetMilliSpan(U32 nTimeStart)
{
U32 nCurrent = FUZ_GetMilliStart();
U32 nSpan = nCurrent - nTimeStart;
if (nTimeStart > nCurrent)
nSpan += 0x100000 * 1000;
return nSpan;
}
# define FUZ_rotl32(x,r) ((x << r) | (x >> (32 - r)))
unsigned int FUZ_rand(unsigned int* src)
{
U32 rand32 = *src;
rand32 *= prime1;
rand32 += prime2;
rand32 = FUZ_rotl32(rand32, 13);
*src = rand32;
return rand32 >> 5;
}
#define FUZ_RAND15BITS (FUZ_rand(seed) & 0x7FFF)
#define FUZ_RANDLENGTH ( (FUZ_rand(seed) & 3) ? (FUZ_rand(seed) % 15) : (FUZ_rand(seed) % 510) + 15)
static void FUZ_fillCompressibleNoiseBuffer(void* buffer, unsigned bufferSize, double proba, U32* seed)
{
BYTE* BBuffer = (BYTE*)buffer;
unsigned pos = 0;
U32 P32 = (U32)(32768 * proba);
// First Byte
BBuffer[pos++] = (BYTE)(FUZ_rand(seed));
while (pos < bufferSize)
{
// Select : Literal (noise) or copy (within 64K)
if (FUZ_RAND15BITS < P32)
{
// Copy (within 64K)
unsigned match, end;
unsigned length = FUZ_RANDLENGTH + 4;
unsigned offset = FUZ_RAND15BITS + 1;
if (offset > pos) offset = pos;
if (pos + length > bufferSize) length = bufferSize - pos;
match = pos - offset;
end = pos + length;
while (pos < end) BBuffer[pos++] = BBuffer[match++];
}
else
{
// Literal (noise)
unsigned end;
unsigned length = FUZ_RANDLENGTH;
if (pos + length > bufferSize) length = bufferSize - pos;
end = pos + length;
while (pos < end) BBuffer[pos++] = (BYTE)(FUZ_rand(seed) >> 5);
}
}
}
static unsigned FUZ_highbit(U32 v32)
{
unsigned nbBits = 0;
if (v32==0) return 0;
while (v32) { v32 >>= 1; nbBits ++; }
return nbBits;
}
int basicTests(U32 seed, double compressibility)
{
int testResult = 0;
void* CNBuffer;
void* compressedBuffer;
void* decodedBuffer;
U32 randState = seed;
size_t cSize, testSize;
LZ4F_preferences_t prefs = { 0 };
LZ4F_decompressionContext_t dCtx;
U64 crcOrig;
// Create compressible test buffer
CNBuffer = malloc(COMPRESSIBLE_NOISE_LENGTH);
compressedBuffer = malloc(LZ4F_compressFrameBound(COMPRESSIBLE_NOISE_LENGTH, NULL));
decodedBuffer = malloc(COMPRESSIBLE_NOISE_LENGTH);
FUZ_fillCompressibleNoiseBuffer(CNBuffer, COMPRESSIBLE_NOISE_LENGTH, compressibility, &randState);
crcOrig = XXH64(CNBuffer, COMPRESSIBLE_NOISE_LENGTH, 1);
// Trivial tests : one-step frame
testSize = COMPRESSIBLE_NOISE_LENGTH;
DISPLAYLEVEL(3, "Using NULL preferences : \n");
cSize = LZ4F_compressFrame(compressedBuffer, LZ4F_compressFrameBound(testSize, NULL), CNBuffer, testSize, NULL);
if (LZ4F_isError(cSize)) goto _output_error;
DISPLAYLEVEL(3, "Compressed %i bytes into a %i bytes frame \n", (int)testSize, (int)cSize);
DISPLAYLEVEL(3, "Decompression test : \n");
{
size_t decodedBufferSize = COMPRESSIBLE_NOISE_LENGTH;
size_t compressedBufferSize = cSize;
BYTE* op = (BYTE*)decodedBuffer;
BYTE* const oend = (BYTE*)decodedBuffer + COMPRESSIBLE_NOISE_LENGTH;
BYTE* ip = (BYTE*)compressedBuffer;
BYTE* const iend = (BYTE*)compressedBuffer + cSize;
U64 crcDest;
LZ4F_errorCode_t errorCode = LZ4F_createDecompressionContext(&dCtx, LZ4F_VERSION);
if (LZ4F_isError(errorCode)) goto _output_error;
DISPLAYLEVEL(3, "Single Block : \n");
errorCode = LZ4F_decompress(dCtx, decodedBuffer, &decodedBufferSize, compressedBuffer, &compressedBufferSize, NULL);
crcDest = XXH64(decodedBuffer, COMPRESSIBLE_NOISE_LENGTH, 1);
if (crcDest != crcOrig) goto _output_error;
DISPLAYLEVEL(3, "Regenerated %i bytes \n", (int)decodedBufferSize);
DISPLAYLEVEL(3, "Byte after byte : \n");
while (ip < iend)
{
size_t oSize = oend-op;
size_t iSize = 1;
//DISPLAY("%7i \n", (int)(ip-(BYTE*)compressedBuffer));
errorCode = LZ4F_decompress(dCtx, op, &oSize, ip, &iSize, NULL);
if (LZ4F_isError(errorCode)) goto _output_error;
op += oSize;
ip += iSize;
}
crcDest = XXH64(decodedBuffer, COMPRESSIBLE_NOISE_LENGTH, 1);
if (crcDest != crcOrig) goto _output_error;
DISPLAYLEVEL(3, "Regenerated %i bytes \n", (int)decodedBufferSize);
errorCode = LZ4F_freeDecompressionContext(dCtx);
if (LZ4F_isError(errorCode)) goto _output_error;
}
DISPLAYLEVEL(3, "Using 64 KB block : \n");
prefs.frameInfo.blockSizeID = max64KB;
prefs.frameInfo.contentChecksumFlag = contentChecksumEnabled;
cSize = LZ4F_compressFrame(compressedBuffer, LZ4F_compressFrameBound(testSize, &prefs), CNBuffer, testSize, &prefs);
if (LZ4F_isError(cSize)) goto _output_error;
DISPLAYLEVEL(3, "Compressed %i bytes into a %i bytes frame \n", (int)testSize, (int)cSize);
DISPLAYLEVEL(3, "without checksum : \n");
prefs.frameInfo.contentChecksumFlag = noContentChecksum;
cSize = LZ4F_compressFrame(compressedBuffer, LZ4F_compressFrameBound(testSize, &prefs), CNBuffer, testSize, &prefs);
if (LZ4F_isError(cSize)) goto _output_error;
DISPLAYLEVEL(3, "Compressed %i bytes into a %i bytes frame \n", (int)testSize, (int)cSize);
DISPLAYLEVEL(3, "Using 256 KB block : \n");
prefs.frameInfo.blockSizeID = max256KB;
prefs.frameInfo.contentChecksumFlag = contentChecksumEnabled;
cSize = LZ4F_compressFrame(compressedBuffer, LZ4F_compressFrameBound(testSize, &prefs), CNBuffer, testSize, &prefs);
if (LZ4F_isError(cSize)) goto _output_error;
DISPLAYLEVEL(3, "Compressed %i bytes into a %i bytes frame \n", (int)testSize, (int)cSize);
DISPLAYLEVEL(3, "Decompression test : \n");
{
size_t decodedBufferSize = COMPRESSIBLE_NOISE_LENGTH;
unsigned maxBits = FUZ_highbit(decodedBufferSize);
BYTE* op = (BYTE*)decodedBuffer;
BYTE* const oend = (BYTE*)decodedBuffer + COMPRESSIBLE_NOISE_LENGTH;
BYTE* ip = (BYTE*)compressedBuffer;
BYTE* const iend = (BYTE*)compressedBuffer + cSize;
U64 crcDest;
LZ4F_errorCode_t errorCode = LZ4F_createDecompressionContext(&dCtx, LZ4F_VERSION);
if (LZ4F_isError(errorCode)) goto _output_error;
DISPLAYLEVEL(3, "random segment sizes : \n");
while (ip < iend)
{
unsigned nbBits = FUZ_rand(&randState) % maxBits;
size_t iSize = (FUZ_rand(&randState) & ((1<<nbBits)-1)) + 1;
size_t oSize = oend-op;
if (iSize > (size_t)(iend-ip)) iSize = iend-ip;
//DISPLAY("%7i : + %6i\n", (int)(ip-(BYTE*)compressedBuffer), (int)iSize);
errorCode = LZ4F_decompress(dCtx, op, &oSize, ip, &iSize, NULL);
if (LZ4F_isError(errorCode)) goto _output_error;
op += oSize;
ip += iSize;
}
crcDest = XXH64(decodedBuffer, COMPRESSIBLE_NOISE_LENGTH, 1);
if (crcDest != crcOrig) goto _output_error;
DISPLAYLEVEL(3, "Regenerated %i bytes \n", (int)decodedBufferSize);
errorCode = LZ4F_freeDecompressionContext(dCtx);
if (LZ4F_isError(errorCode)) goto _output_error;
}
DISPLAYLEVEL(3, "without checksum : \n");
prefs.frameInfo.contentChecksumFlag = noContentChecksum;
cSize = LZ4F_compressFrame(compressedBuffer, LZ4F_compressFrameBound(testSize, &prefs), CNBuffer, testSize, &prefs);
if (LZ4F_isError(cSize)) goto _output_error;
DISPLAYLEVEL(3, "Compressed %i bytes into a %i bytes frame \n", (int)testSize, (int)cSize);
DISPLAYLEVEL(3, "Using 1 MB block : \n");
prefs.frameInfo.blockSizeID = max1MB;
prefs.frameInfo.contentChecksumFlag = contentChecksumEnabled;
cSize = LZ4F_compressFrame(compressedBuffer, LZ4F_compressFrameBound(testSize, &prefs), CNBuffer, testSize, &prefs);
if (LZ4F_isError(cSize)) goto _output_error;
DISPLAYLEVEL(3, "Compressed %i bytes into a %i bytes frame \n", (int)testSize, (int)cSize);
DISPLAYLEVEL(3, "without checksum : \n");
prefs.frameInfo.contentChecksumFlag = noContentChecksum;
cSize = LZ4F_compressFrame(compressedBuffer, LZ4F_compressFrameBound(testSize, &prefs), CNBuffer, testSize, &prefs);
if (LZ4F_isError(cSize)) goto _output_error;
DISPLAYLEVEL(3, "Compressed %i bytes into a %i bytes frame \n", (int)testSize, (int)cSize);
DISPLAYLEVEL(3, "Using 4 MB block : \n");
prefs.frameInfo.blockSizeID = max4MB;
prefs.frameInfo.contentChecksumFlag = contentChecksumEnabled;
cSize = LZ4F_compressFrame(compressedBuffer, LZ4F_compressFrameBound(testSize, &prefs), CNBuffer, testSize, &prefs);
if (LZ4F_isError(cSize)) goto _output_error;
DISPLAYLEVEL(3, "Compressed %i bytes into a %i bytes frame \n", (int)testSize, (int)cSize);
DISPLAYLEVEL(3, "without checksum : \n");
prefs.frameInfo.contentChecksumFlag = noContentChecksum;
cSize = LZ4F_compressFrame(compressedBuffer, LZ4F_compressFrameBound(testSize, &prefs), CNBuffer, testSize, &prefs);
if (LZ4F_isError(cSize)) goto _output_error;
DISPLAYLEVEL(3, "Compressed %i bytes into a %i bytes frame \n", (int)testSize, (int)cSize);
DISPLAY("Basic tests completed \n");
_end:
free(CNBuffer);
free(compressedBuffer);
free(decodedBuffer);
return testResult;
_output_error:
testResult = 1;
DISPLAY("Error detected ! \n");
goto _end;
}
static void locateBuffDiff(const void* buff1, const void* buff2, size_t size)
{
int p=0;
BYTE* b1=(BYTE*)buff1;
BYTE* b2=(BYTE*)buff2;
while (b1[p]==b2[p]) p++;
printf("Error at pos %i/%i : %02X != %02X \n", p, (int)size, b1[p], b2[p]);
}
static const U32 srcDataLength = 9 MB; /* needs to be > 2x4MB to test large blocks */
int fuzzerTests(U32 seed, unsigned nbTests, unsigned startTest, double compressibility)
{
unsigned testResult = 0;
unsigned testNb = 0;
void* srcBuffer = NULL;
void* compressedBuffer = NULL;
void* decodedBuffer = NULL;
U32 coreRand = seed;
LZ4F_decompressionContext_t dCtx;
LZ4F_compressionContext_t cCtx;
size_t result;
# define CHECK(cond, ...) if (cond) { DISPLAY("Error => "); DISPLAY(__VA_ARGS__); \
DISPLAY(" (seed %u, test nb %i) \n", seed, testNb); goto _output_error; }
// Create buffers
result = LZ4F_createDecompressionContext(&dCtx, LZ4F_VERSION);
CHECK(LZ4F_isError(result), "Allocation failed (error %i)", (int)result);
result = LZ4F_createCompressionContext(&cCtx, LZ4F_VERSION);
CHECK(LZ4F_isError(result), "Allocation failed (error %i)", (int)result);
srcBuffer = malloc(srcDataLength);
CHECK(srcBuffer==NULL, "srcBuffer Allocation failed");
compressedBuffer = malloc(LZ4F_compressFrameBound(srcDataLength, NULL));
CHECK(compressedBuffer==NULL, "compressedBuffer Allocation failed");
decodedBuffer = malloc(srcDataLength);
CHECK(decodedBuffer==NULL, "decodedBuffer Allocation failed");
FUZ_fillCompressibleNoiseBuffer(srcBuffer, srcDataLength, compressibility, &coreRand);
// jump to requested testNb
for (testNb =0; testNb < startTest; testNb++) (void)FUZ_rand(&coreRand); // sync randomizer
// main fuzzer loop
for ( ; testNb < nbTests; testNb++)
{
U32 randState = coreRand ^ prime1;
unsigned BSId = 4 + (FUZ_rand(&randState) & 3);
unsigned BMId = FUZ_rand(&randState) & 1;
unsigned CCflag = FUZ_rand(&randState) & 1;
unsigned autoflush = (FUZ_rand(&randState) & 3) == 2;
LZ4F_preferences_t prefs = { 0 };
LZ4F_compressOptions_t options = { 0 };
unsigned nbBits = (FUZ_rand(&randState) % (FUZ_highbit(srcDataLength-1) - 1)) + 1;
size_t srcSize = (FUZ_rand(&randState) & ((1<<nbBits)-1)) + 1;
size_t srcStart = FUZ_rand(&randState) % (srcDataLength - srcSize);
size_t cSize;
U64 crcOrig, crcDecoded;
prefs.frameInfo.blockMode = BMId;
prefs.frameInfo.blockSizeID = BSId;
prefs.frameInfo.contentChecksumFlag = CCflag;
prefs.autoFlush = autoflush;
DISPLAYUPDATE(2, "\r%5i ", testNb);
crcOrig = XXH64((BYTE*)srcBuffer+srcStart, srcSize, 1);
{
const BYTE* ip = (const BYTE*)srcBuffer + srcStart;
const BYTE* const iend = ip + srcSize;
BYTE* op = compressedBuffer;
BYTE* const oend = op + LZ4F_compressFrameBound(srcDataLength, NULL);
unsigned maxBits = FUZ_highbit(srcSize);
result = LZ4F_compressBegin(cCtx, op, oend-op, &prefs);
CHECK(LZ4F_isError(result), "Compression header failed (error %i)", (int)result);
op += result;
while (ip < iend)
{
unsigned nbBitsSeg = FUZ_rand(&randState) % maxBits;
size_t iSize = (FUZ_rand(&randState) & ((1<<nbBitsSeg)-1)) + 1;
size_t oSize = oend-op;
unsigned forceFlush = ((FUZ_rand(&randState) & 3) == 1);
if (iSize > (size_t)(iend-ip)) iSize = iend-ip;
options.stableSrc = ((FUZ_rand(&randState) && 3) == 2);
result = LZ4F_compressUpdate(cCtx, op, oSize, ip, iSize, &options);
CHECK(LZ4F_isError(result), "Compression failed (error %i)", (int)result);
op += result;
ip += iSize;
if (forceFlush)
{
result = LZ4F_flush(cCtx, op, oend-op, &options);
CHECK(LZ4F_isError(result), "Compression failed (error %i)", (int)result);
op += result;
}
}
result = LZ4F_compressEnd(cCtx, op, oend-op, &options);
CHECK(LZ4F_isError(result), "Compression completion failed (error %i)", (int)result);
op += result;
cSize = op-(BYTE*)compressedBuffer;
}
{
const BYTE* ip = compressedBuffer;
const BYTE* const iend = ip + cSize;
BYTE* op = decodedBuffer;
BYTE* const oend = op + srcDataLength;
unsigned maxBits = FUZ_highbit(cSize);
while (ip < iend)
{
unsigned nbBitsI = (FUZ_rand(&randState) % (maxBits-1)) + 1;
unsigned nbBitsO = (FUZ_rand(&randState) % (maxBits)) + 1;
size_t iSize = (FUZ_rand(&randState) & ((1<<nbBitsI)-1)) + 1;
size_t oSize = (FUZ_rand(&randState) & ((1<<nbBitsO)-1)) + 2;
if (iSize > (size_t)(iend-ip)) iSize = iend-ip;
if (oSize > (size_t)(oend-op)) oSize = oend-op;
result = LZ4F_decompress(dCtx, op, &oSize, ip, &iSize, NULL);
if (result == (size_t)-ERROR_checksum_invalid) locateBuffDiff((BYTE*)srcBuffer+srcStart, decodedBuffer, srcSize);
CHECK(LZ4F_isError(result), "Decompression failed (error %i)", (int)result);
op += oSize;
ip += iSize;
}
CHECK(result != 0, "Frame decompression failed (error %i)", (int)result);
crcDecoded = XXH64(decodedBuffer, op-(BYTE*)decodedBuffer, 1);
if (crcDecoded != crcOrig) locateBuffDiff((BYTE*)srcBuffer+srcStart, decodedBuffer, srcSize);
CHECK(crcDecoded != crcOrig, "Decompression corruption");
}
(void)FUZ_rand(&coreRand); // update rand seed
}
DISPLAYLEVEL(2, "\rAll tests completed \n");
_end:
LZ4F_freeDecompressionContext(dCtx);
LZ4F_freeCompressionContext(cCtx);
free(srcBuffer);
free(compressedBuffer);
free(decodedBuffer);
return testResult;
_output_error:
testResult = 1;
goto _end;
}
int FUZ_usage(void)
{
DISPLAY( "Usage :\n");
DISPLAY( " %s [args]\n", programName);
DISPLAY( "\n");
DISPLAY( "Arguments :\n");
DISPLAY( " -i# : Nb of tests (default:%i) \n", nbTestsDefault);
DISPLAY( " -s# : Select seed (default:prompt user)\n");
DISPLAY( " -t# : Select starting test number (default:0)\n");
DISPLAY( " -p# : Select compressibility in %% (default:%i%%)\n", FUZ_COMPRESSIBILITY_DEFAULT);
DISPLAY( " -v : verbose\n");
DISPLAY( " -h : display help and exit\n");
return 0;
}
int main(int argc, char** argv)
{
U32 seed=0;
int seedset=0;
int argNb;
int nbTests = nbTestsDefault;
int testNb = 0;
int proba = FUZ_COMPRESSIBILITY_DEFAULT;
int result=0;
// Check command line
programName = argv[0];
for(argNb=1; argNb<argc; argNb++)
{
char* argument = argv[argNb];
if(!argument) continue; // Protection if argument empty
// Decode command (note : aggregated commands are allowed)
if (argument[0]=='-')
{
if (!strcmp(argument, "--no-prompt")) { no_prompt=1; seedset=1; displayLevel=1; continue; }
while (argument[1]!=0)
{
argument++;
switch(*argument)
{
case 'h':
return FUZ_usage();
case 'v':
argument++;
displayLevel=4;
break;
case 'q':
argument++;
displayLevel--;
break;
case 'i':
argument++;
nbTests=0;
while ((*argument>='0') && (*argument<='9'))
{
nbTests *= 10;
nbTests += *argument - '0';
argument++;
}
break;
case 's':
argument++;
seed=0; seedset=1;
while ((*argument>='0') && (*argument<='9'))
{
seed *= 10;
seed += *argument - '0';
argument++;
}
break;
case 't':
argument++;
testNb=0;
while ((*argument>='0') && (*argument<='9'))
{
testNb *= 10;
testNb += *argument - '0';
argument++;
}
break;
case 'p':
argument++;
proba=0;
while ((*argument>='0') && (*argument<='9'))
{
proba *= 10;
proba += *argument - '0';
argument++;
}
if (proba<0) proba=0;
if (proba>100) proba=100;
break;
default: ;
}
}
}
}
// Get Seed
printf("Starting lz4frame tester (%i-bits, %s)\n", (int)(sizeof(size_t)*8), LZ4_VERSION);
if (!seedset) seed = FUZ_GetMilliStart() % 10000;
printf("Seed = %u\n", seed);
if (proba!=FUZ_COMPRESSIBILITY_DEFAULT) printf("Compressibility : %i%%\n", proba);
if (nbTests<=0) nbTests=1;
//if (testNb==0) result = basicTests(seed, ((double)proba) / 100);
if (result) return 1;
return fuzzerTests(seed, nbTests, testNb, ((double)proba) / 100);
}

118
programs/fullbench.c
View File

@ -60,10 +60,8 @@
#endif
#include "lz4.h"
#define COMPRESSOR0 LZ4_compress
#include "lz4hc.h"
#define COMPRESSOR1 LZ4_compressHC
#define DEFAULTCOMPRESSOR COMPRESSOR0
#include "lz4frame.h"
#include "xxhash.h"
@ -323,6 +321,11 @@ static int local_LZ4_compressHC_limitedOutput_continue(const char* in, char* out
return LZ4_compressHC_limitedOutput_continue(ctx, in, out, inSize, LZ4_compressBound(inSize));
}
static int local_LZ4F_compressFrame(const char* in, char* out, int inSize)
{
return LZ4F_compressFrame(out, 2*inSize + 16, in, inSize, NULL);
}
static int local_LZ4_decompress_fast(const char* in, char* out, int inSize, int outSize)
{
(void)inSize;
@ -351,26 +354,51 @@ static int local_LZ4_decompress_safe_usingDict(const char* in, char* out, int in
return outSize;
}
extern int LZ4_decompress_safe_forceExtDict(const char* in, char* out, int inSize, int outSize, const char* dict, int dictSize);
static int local_LZ4_decompress_safe_forceExtDict(const char* in, char* out, int inSize, int outSize)
{
(void)inSize;
LZ4_decompress_safe_forceExtDict(in, out, inSize, outSize, in - 65536, 65536);
return outSize;
}
static int local_LZ4_decompress_safe_partial(const char* in, char* out, int inSize, int outSize)
{
return LZ4_decompress_safe_partial(in, out, inSize, outSize - 5, outSize);
}
static LZ4F_decompressionContext_t g_dCtx;
static int local_LZ4F_decompress(const char* in, char* out, int inSize, int outSize)
{
size_t srcSize = inSize;
size_t dstSize = outSize;
size_t result;
result = LZ4F_decompress(g_dCtx, out, &dstSize, in, &srcSize, NULL);
if (result!=0) { DISPLAY("Error decompressing frame : unfinished frame\n"); exit(8); }
if (srcSize != (size_t)inSize) { DISPLAY("Error decompressing frame : read size incorrect\n"); exit(9); }
return dstSize;
}
int fullSpeedBench(char** fileNamesTable, int nbFiles)
{
int fileIdx=0;
char* orig_buff;
# define NB_COMPRESSION_ALGORITHMS 13
# define MINCOMPRESSIONCHAR '0'
# define NB_COMPRESSION_ALGORITHMS 14
double totalCTime[NB_COMPRESSION_ALGORITHMS+1] = {0};
double totalCSize[NB_COMPRESSION_ALGORITHMS+1] = {0};
# define NB_DECOMPRESSION_ALGORITHMS 7
# define MINDECOMPRESSIONCHAR '0'
# define MAXDECOMPRESSIONCHAR (MINDECOMPRESSIONCHAR + NB_DECOMPRESSION_ALGORITHMS)
static char* decompressionNames[] = { "LZ4_decompress_fast", "LZ4_decompress_fast_withPrefix64k", "LZ4_decompress_fast_usingDict",
"LZ4_decompress_safe", "LZ4_decompress_safe_withPrefix64k", "LZ4_decompress_safe_usingDict", "LZ4_decompress_safe_partial" };
# define NB_DECOMPRESSION_ALGORITHMS 9
double totalDTime[NB_DECOMPRESSION_ALGORITHMS+1] = {0};
size_t errorCode;
errorCode = LZ4F_createDecompressionContext(&g_dCtx, LZ4F_VERSION);
if (LZ4F_isError(errorCode))
{
DISPLAY("dctx allocation issue \n");
return 10;
}
// Loop for each file
while (fileIdx<nbFiles)
@ -428,22 +456,6 @@ int fullSpeedBench(char** fileNamesTable, int nbFiles)
return 12;
}
// Init chunks data
{
int i;
size_t remaining = benchedSize;
char* in = orig_buff;
char* out = compressed_buff;
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);
@ -471,10 +483,27 @@ int fullSpeedBench(char** fileNamesTable, int nbFiles)
DISPLAY("\r%79s\r", "");
DISPLAY(" %s : \n", inFileName);
// Init chunks data
{
int i;
size_t remaining = benchedSize;
char* in = orig_buff;
char* out = compressed_buff;
nbChunks = (int) ((int)benchedSize / chunkSize) + 1;
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;
}
}
// Compression Algorithms
for (cAlgNb=1; (cAlgNb <= NB_COMPRESSION_ALGORITHMS) && (compressionTest); cAlgNb++)
{
char* compressorName;
const char* compressorName;
int (*compressionFunction)(const char*, char*, int);
void* (*initFunction)(const char*) = NULL;
double bestTime = 100000000.;
@ -496,6 +525,9 @@ int fullSpeedBench(char** fileNamesTable, int nbFiles)
case 11: compressionFunction = local_LZ4_compressHC_continue; initFunction = LZ4_createHC; compressorName = "LZ4_compressHC_continue"; break;
case 12: compressionFunction = local_LZ4_compressHC_limitedOutput_continue; initFunction = LZ4_createHC; compressorName = "LZ4_compressHC_limitedOutput_continue"; break;
case 13: compressionFunction = local_LZ4_compress_forceDict; initFunction = local_LZ4_resetDictT; compressorName = "LZ4_compress_forceDict"; break;
case 14: compressionFunction = local_LZ4F_compressFrame; compressorName = "LZ4F_compressFrame";
chunkP[0].origSize = benchedSize; nbChunks=1;
break;
default : DISPLAY("ERROR ! Bad algorithm Id !! \n"); free(chunkP); return 1;
}
@ -546,29 +578,39 @@ int fullSpeedBench(char** fileNamesTable, int nbFiles)
chunkP[chunkNb].compressedSize = LZ4_compress(chunkP[chunkNb].origBuffer, chunkP[chunkNb].compressedBuffer, chunkP[chunkNb].origSize);
if (chunkP[chunkNb].compressedSize==0) DISPLAY("ERROR ! %s() = 0 !! \n", "LZ4_compress"), exit(1);
}
{ size_t i; for (i=0; i<benchedSize; i++) orig_buff[i]=0; } // zeroing source area, for CRC checking
// Decompression Algorithms
for (dAlgNb=0; (dAlgNb < NB_DECOMPRESSION_ALGORITHMS) && (decompressionTest); dAlgNb++)
for (dAlgNb=1; (dAlgNb <= NB_DECOMPRESSION_ALGORITHMS) && (decompressionTest); dAlgNb++)
{
char* dName = decompressionNames[dAlgNb];
//const char* dName = decompressionNames[dAlgNb];
const char* dName;
int (*decompressionFunction)(const char*, char*, int, int);
double bestTime = 100000000.;
if ((decompressionAlgo != ALL_DECOMPRESSORS) && (decompressionAlgo != dAlgNb+1)) continue;
if ((decompressionAlgo != ALL_DECOMPRESSORS) && (decompressionAlgo != dAlgNb)) continue;
switch(dAlgNb)
{
case 0: decompressionFunction = local_LZ4_decompress_fast; break;
case 1: decompressionFunction = local_LZ4_decompress_fast_withPrefix64k; break;
case 2: decompressionFunction = local_LZ4_decompress_fast_usingDict; break;
case 3: decompressionFunction = LZ4_decompress_safe; break;
case 4: decompressionFunction = LZ4_decompress_safe_withPrefix64k; break;
case 5: decompressionFunction = local_LZ4_decompress_safe_usingDict; break;
case 6: decompressionFunction = local_LZ4_decompress_safe_partial; break;
case 1: decompressionFunction = local_LZ4_decompress_fast; dName = "LZ4_decompress_fast"; break;
case 2: decompressionFunction = local_LZ4_decompress_fast_withPrefix64k; dName = "LZ4_decompress_fast_withPrefix64k"; break;
case 3: decompressionFunction = local_LZ4_decompress_fast_usingDict; dName = "LZ4_decompress_fast_usingDict"; break;
case 4: decompressionFunction = LZ4_decompress_safe; dName = "LZ4_decompress_safe"; break;
case 5: decompressionFunction = LZ4_decompress_safe_withPrefix64k; dName = "LZ4_decompress_safe_withPrefix64k"; break;
case 6: decompressionFunction = local_LZ4_decompress_safe_usingDict; dName = "LZ4_decompress_safe_usingDict"; break;
case 7: decompressionFunction = local_LZ4_decompress_safe_partial; dName = "LZ4_decompress_safe_partial"; break;
case 8: decompressionFunction = local_LZ4_decompress_safe_forceExtDict; dName = "LZ4_decompress_safe_forceExtDict"; break;
case 9: decompressionFunction = local_LZ4F_decompress; dName = "LZ4F_decompress";
errorCode = LZ4F_compressFrame(compressed_buff, compressedBuffSize, orig_buff, benchedSize, NULL);
if (LZ4F_isError(errorCode)) { DISPLAY("Preparation error compressing frame\n"); return 1; }
chunkP[0].origSize = benchedSize;
chunkP[0].compressedSize = errorCode;
nbChunks = 1;
break;
default : DISPLAY("ERROR ! Bad decompression algorithm Id !! \n"); free(chunkP); return 1;
}
{ size_t i; for (i=0; i<benchedSize; i++) orig_buff[i]=0; } // zeroing source area, for CRC checking
for (loopNb = 1; loopNb <= nbIterations; loopNb++)
{
double averageTime;

168
xxhash.h Normal file
View File

@ -0,0 +1,168 @@
/*
xxHash - Extremely 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, unsigned int len, unsigned int seed);
unsigned long long XXH64 (const void* input, unsigned int len, unsigned long long 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.
XXH64() :
Calculate the 64-bits hash of sequence of length "len" stored at memory address "input".
*/
/*****************************
Advanced Hash Functions
*****************************/
void* XXH32_init (unsigned int seed);
XXH_errorcode XXH32_update (void* state, const void* input, unsigned int len);
unsigned int XXH32_digest (void* state);
void* XXH64_init (unsigned long long seed);
XXH_errorcode XXH64_update (void* state, const void* input, unsigned int len);
unsigned long long XXH64_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* XXHnn_init()
The function returns a pointer which holds the state of calculation.
If the pointer is NULL, allocation has failed, so no state can be tracked.
The state pointer must be provided as "void* state" parameter for XXHnn_update().
XXHnn_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 XXHnn_digest().
This function returns the final nn-bits hash.
You must provide the same "void* state" parameter created by XXHnn_init().
Memory will be freed by XXHnn_digest().
*/
int XXH32_sizeofState(void);
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;
int XXH64_sizeofState(void);
XXH_errorcode XXH64_resetState(void* state, unsigned long long seed);
#define XXH64_SIZEOFSTATE 88
typedef struct { long long ll[(XXH64_SIZEOFSTATE+(sizeof(long long)-1))/sizeof(long long)]; } XXH64_stateSpace_t;
/*
These functions allow user application to make its own allocation for state.
XXHnn_sizeofState() is used to know how much space must be allocated for the xxHash nn-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 XXHnn_resetState(), which initializes the state.
For static allocation purposes (such as allocation on stack, or freestanding systems without malloc()),
use the structure XXHnn_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);
unsigned long long XXH64_intermediateDigest (void* state);
/*
These functions do the same as XXHnn_digest(), generating a nn-bit hash,
but preserve memory context.
This way, it becomes possible to generate intermediate hashes, and then continue feeding data with XXHnn_update().
To free memory context, use XXHnn_digest(), or free().
*/
#if defined (__cplusplus)
}
#endif