Merge branch 'dev' into ahmed_file
This commit is contained in:
commit
81c51483e9
@ -11,6 +11,7 @@ cli: Named pipes support, by @bimbashrestha
|
||||
cli: short tar's extension support, by @stokito
|
||||
cli: command --output-dir-flat= , generates target files into requested directory, by @senhuang42
|
||||
cli: commands --stream-size=# and --size-hint=#, by @nmagerko
|
||||
cli: command --exclude-compressed, by @shashank0791
|
||||
cli: faster `-t` test mode
|
||||
cli: improved some error messages, by @vangyzen
|
||||
cli: rare deadlock condition within dictionary builder, by @terrelln
|
||||
|
@ -43,7 +43,7 @@ on the [Silesia compression corpus].
|
||||
|
||||
| Compressor name | Ratio | Compression| Decompress.|
|
||||
| --------------- | ------| -----------| ---------- |
|
||||
| **zstd 1.4.0 -1** | 2.884 | 530 MB/s | 1360 MB/s |
|
||||
| **zstd 1.4.4 -1** | 2.884 | 520 MB/s | 1600 MB/s |
|
||||
| zlib 1.2.11 -1 | 2.743 | 110 MB/s | 440 MB/s |
|
||||
| brotli 1.0.7 -0 | 2.701 | 430 MB/s | 470 MB/s |
|
||||
| quicklz 1.5.0 -1 | 2.238 | 600 MB/s | 800 MB/s |
|
||||
|
@ -173,8 +173,6 @@
|
||||
sh -e playTests.sh --test-large-data &&
|
||||
fullbench.exe -i1 &&
|
||||
fullbench.exe -i1 -P0 &&
|
||||
fuzzer_VS2008_%PLATFORM%_Release.exe %FUZZERTEST% &&
|
||||
fuzzer_VS2010_%PLATFORM%_Release.exe %FUZZERTEST% &&
|
||||
fuzzer_VS2012_%PLATFORM%_Release.exe %FUZZERTEST% &&
|
||||
fuzzer_VS2013_%PLATFORM%_Release.exe %FUZZERTEST% &&
|
||||
fuzzer_VS2015_%PLATFORM%_Release.exe %FUZZERTEST%
|
||||
|
@ -124,9 +124,13 @@ endif ()
|
||||
#-----------------------------------------------------------------------------
|
||||
add_subdirectory(lib)
|
||||
|
||||
option(ZSTD_PROGRAMS_LINK_SHARED "PROGRAMS LINK SHARED" OFF)
|
||||
|
||||
if (ZSTD_BUILD_PROGRAMS)
|
||||
if (NOT ZSTD_BUILD_STATIC)
|
||||
if (NOT ZSTD_BUILD_STATIC AND NOT ZSTD_PROGRAMS_LINK_SHARED)
|
||||
message(SEND_ERROR "You need to build static library to build zstd CLI")
|
||||
elseif(NOT ZSTD_BUILD_SHARED AND ZSTD_PROGRAMS_LINK_SHARED)
|
||||
message(SEND_ERROR "You need to build shared library to build zstd CLI")
|
||||
endif ()
|
||||
|
||||
add_subdirectory(programs)
|
||||
|
@ -21,13 +21,19 @@ if (ZSTD_LEGACY_SUPPORT)
|
||||
include_directories(${PROGRAMS_LEGACY_DIR} ${LIBRARY_DIR}/legacy)
|
||||
endif ()
|
||||
|
||||
if (ZSTD_PROGRAMS_LINK_SHARED)
|
||||
set(PROGRAMS_ZSTD_LINK_TARGET libzstd_shared)
|
||||
else ()
|
||||
set(PROGRAMS_ZSTD_LINK_TARGET libzstd_static)
|
||||
endif ()
|
||||
|
||||
if (MSVC)
|
||||
set(MSVC_RESOURCE_DIR ${ZSTD_SOURCE_DIR}/build/VS2010/zstd)
|
||||
set(PlatformDependResources ${MSVC_RESOURCE_DIR}/zstd.rc)
|
||||
endif ()
|
||||
|
||||
add_executable(zstd ${PROGRAMS_DIR}/zstdcli.c ${PROGRAMS_DIR}/util.c ${PROGRAMS_DIR}/timefn.c ${PROGRAMS_DIR}/fileio.c ${PROGRAMS_DIR}/benchfn.c ${PROGRAMS_DIR}/benchzstd.c ${PROGRAMS_DIR}/datagen.c ${PROGRAMS_DIR}/dibio.c ${PlatformDependResources})
|
||||
target_link_libraries(zstd libzstd_static)
|
||||
target_link_libraries(zstd ${PROGRAMS_ZSTD_LINK_TARGET})
|
||||
if (CMAKE_SYSTEM_NAME MATCHES "(Solaris|SunOS)")
|
||||
target_link_libraries(zstd rt)
|
||||
endif ()
|
||||
@ -68,7 +74,7 @@ if (UNIX)
|
||||
DESTINATION "${MAN_INSTALL_DIR}")
|
||||
|
||||
add_executable(zstd-frugal ${PROGRAMS_DIR}/zstdcli.c ${PROGRAMS_DIR}/util.c ${PROGRAMS_DIR}/timefn.c ${PROGRAMS_DIR}/fileio.c)
|
||||
target_link_libraries(zstd-frugal libzstd_static)
|
||||
target_link_libraries(zstd-frugal ${PROGRAMS_ZSTD_LINK_TARGET})
|
||||
set_property(TARGET zstd-frugal APPEND PROPERTY COMPILE_DEFINITIONS "ZSTD_NOBENCH;ZSTD_NODICT")
|
||||
endif ()
|
||||
|
||||
|
@ -2771,7 +2771,7 @@ static size_t ZSTD_checkDictNCount(short* normalizedCounter, unsigned dictMaxSym
|
||||
/*! ZSTD_loadZstdDictionary() :
|
||||
* @return : dictID, or an error code
|
||||
* assumptions : magic number supposed already checked
|
||||
* dictSize supposed > 8
|
||||
* dictSize supposed >= 8
|
||||
*/
|
||||
static size_t ZSTD_loadZstdDictionary(ZSTD_compressedBlockState_t* bs,
|
||||
ZSTD_matchState_t* ms,
|
||||
@ -2788,7 +2788,7 @@ static size_t ZSTD_loadZstdDictionary(ZSTD_compressedBlockState_t* bs,
|
||||
size_t dictID;
|
||||
|
||||
ZSTD_STATIC_ASSERT(HUF_WORKSPACE_SIZE >= (1<<MAX(MLFSELog,LLFSELog)));
|
||||
assert(dictSize > 8);
|
||||
assert(dictSize >= 8);
|
||||
assert(MEM_readLE32(dictPtr) == ZSTD_MAGIC_DICTIONARY);
|
||||
|
||||
dictPtr += 4; /* skip magic number */
|
||||
@ -2890,7 +2890,10 @@ ZSTD_compress_insertDictionary(ZSTD_compressedBlockState_t* bs,
|
||||
void* workspace)
|
||||
{
|
||||
DEBUGLOG(4, "ZSTD_compress_insertDictionary (dictSize=%u)", (U32)dictSize);
|
||||
if ((dict==NULL) || (dictSize<=8)) return 0;
|
||||
if ((dict==NULL) || (dictSize<8)) {
|
||||
RETURN_ERROR_IF(dictContentType == ZSTD_dct_fullDict, dictionary_wrong);
|
||||
return 0;
|
||||
}
|
||||
|
||||
ZSTD_reset_compressedBlockState(bs);
|
||||
|
||||
@ -2934,6 +2937,7 @@ static size_t ZSTD_compressBegin_internal(ZSTD_CCtx* cctx,
|
||||
&& (cdict->dictContentSize > 0)
|
||||
&& ( pledgedSrcSize < ZSTD_USE_CDICT_PARAMS_SRCSIZE_CUTOFF
|
||||
|| pledgedSrcSize < cdict->dictContentSize * ZSTD_USE_CDICT_PARAMS_DICTSIZE_MULTIPLIER
|
||||
|| pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN
|
||||
|| cdict->compressionLevel == 0)
|
||||
&& (params->attachDictPref != ZSTD_dictForceLoad) ) {
|
||||
return ZSTD_resetCCtx_usingCDict(cctx, cdict, params, pledgedSrcSize, zbuff);
|
||||
@ -2941,7 +2945,7 @@ static size_t ZSTD_compressBegin_internal(ZSTD_CCtx* cctx,
|
||||
|
||||
FORWARD_IF_ERROR( ZSTD_resetCCtx_internal(cctx, *params, pledgedSrcSize,
|
||||
ZSTDcrp_makeClean, zbuff) );
|
||||
{ size_t const dictID = cdict ?
|
||||
{ size_t const dictID = cdict ?
|
||||
ZSTD_compress_insertDictionary(
|
||||
cctx->blockState.prevCBlock, &cctx->blockState.matchState,
|
||||
&cctx->workspace, params, cdict->dictContent, cdict->dictContentSize,
|
||||
@ -3218,7 +3222,7 @@ static size_t ZSTD_initCDict_internal(
|
||||
ZSTDirp_reset,
|
||||
ZSTD_resetTarget_CDict));
|
||||
/* (Maybe) load the dictionary
|
||||
* Skips loading the dictionary if it is <= 8 bytes.
|
||||
* Skips loading the dictionary if it is < 8 bytes.
|
||||
*/
|
||||
{ ZSTD_CCtx_params params;
|
||||
memset(¶ms, 0, sizeof(params));
|
||||
@ -3381,6 +3385,7 @@ size_t ZSTD_compressBegin_usingCDict_advanced(
|
||||
{ ZSTD_CCtx_params params = cctx->requestedParams;
|
||||
params.cParams = ( pledgedSrcSize < ZSTD_USE_CDICT_PARAMS_SRCSIZE_CUTOFF
|
||||
|| pledgedSrcSize < cdict->dictContentSize * ZSTD_USE_CDICT_PARAMS_DICTSIZE_MULTIPLIER
|
||||
|| pledgedSrcSize == ZSTD_CONTENTSIZE_UNKNOWN
|
||||
|| cdict->compressionLevel == 0 )
|
||||
&& (params.attachDictPref != ZSTD_dictForceLoad) ?
|
||||
ZSTD_getCParamsFromCDict(cdict)
|
||||
|
@ -1096,7 +1096,7 @@ ZSTD_loadDEntropy(ZSTD_entropyDTables_t* entropy,
|
||||
size_t const dictContentSize = (size_t)(dictEnd - (dictPtr+12));
|
||||
for (i=0; i<3; i++) {
|
||||
U32 const rep = MEM_readLE32(dictPtr); dictPtr += 4;
|
||||
RETURN_ERROR_IF(rep==0 || rep >= dictContentSize,
|
||||
RETURN_ERROR_IF(rep==0 || rep > dictContentSize,
|
||||
dictionary_corrupted);
|
||||
entropy->rep[i] = rep;
|
||||
} }
|
||||
@ -1265,7 +1265,7 @@ size_t ZSTD_DCtx_loadDictionary_advanced(ZSTD_DCtx* dctx,
|
||||
{
|
||||
RETURN_ERROR_IF(dctx->streamStage != zdss_init, stage_wrong);
|
||||
ZSTD_clearDict(dctx);
|
||||
if (dict && dictSize >= 8) {
|
||||
if (dict && dictSize != 0) {
|
||||
dctx->ddictLocal = ZSTD_createDDict_advanced(dict, dictSize, dictLoadMethod, dictContentType, dctx->customMem);
|
||||
RETURN_ERROR_IF(dctx->ddictLocal == NULL, memory_allocation);
|
||||
dctx->ddict = dctx->ddictLocal;
|
||||
|
@ -617,7 +617,7 @@ static void ZSTD_safecopy(BYTE* op, BYTE* const oend_w, BYTE const* ip, ptrdiff_
|
||||
ptrdiff_t const diff = op - ip;
|
||||
BYTE* const oend = op + length;
|
||||
|
||||
assert((ovtype == ZSTD_no_overlap && (diff <= -8 || diff >= 8)) ||
|
||||
assert((ovtype == ZSTD_no_overlap && (diff <= -8 || diff >= 8 || op >= oend_w)) ||
|
||||
(ovtype == ZSTD_overlap_src_before_dst && diff >= 0));
|
||||
|
||||
if (length < 8) {
|
||||
|
@ -319,6 +319,8 @@ struct FIO_prefs_s {
|
||||
/* Computation resources preferences */
|
||||
unsigned memLimit;
|
||||
int nbWorkers;
|
||||
|
||||
int excludeCompressedFiles;
|
||||
};
|
||||
|
||||
|
||||
@ -359,6 +361,7 @@ FIO_prefs_t* FIO_createPreferences(void)
|
||||
ret->srcSizeHint = 0;
|
||||
ret->testMode = 0;
|
||||
ret->literalCompressionMode = ZSTD_lcm_auto;
|
||||
ret->excludeCompressedFiles = 0;
|
||||
return ret;
|
||||
}
|
||||
|
||||
@ -402,6 +405,8 @@ void FIO_setNbWorkers(FIO_prefs_t* const prefs, int nbWorkers) {
|
||||
prefs->nbWorkers = nbWorkers;
|
||||
}
|
||||
|
||||
void FIO_setExcludeCompressedFile(FIO_prefs_t* const prefs, int excludeCompressedFiles) { prefs->excludeCompressedFiles = excludeCompressedFiles; }
|
||||
|
||||
void FIO_setBlockSize(FIO_prefs_t* const prefs, int blockSize) {
|
||||
if (blockSize && prefs->nbWorkers==0)
|
||||
DISPLAYLEVEL(2, "Setting block size is useless in single-thread mode \n");
|
||||
@ -520,7 +525,11 @@ static FILE* FIO_openSrcFile(const char* srcFileName)
|
||||
return NULL;
|
||||
}
|
||||
|
||||
if (!UTIL_isRegularFile(srcFileName) && !UTIL_isFIFO(srcFileName)) {
|
||||
if (!UTIL_isRegularFile(srcFileName)
|
||||
#ifndef _MSC_VER
|
||||
&& !UTIL_isFIFO(srcFileName)
|
||||
#endif /* _MSC_VER */
|
||||
) {
|
||||
DISPLAYLEVEL(1, "zstd: %s is not a regular file -- ignored \n",
|
||||
srcFileName);
|
||||
return NULL;
|
||||
@ -1425,6 +1434,21 @@ static int FIO_compressFilename_dstFile(FIO_prefs_t* const prefs,
|
||||
return result;
|
||||
}
|
||||
|
||||
/* List used to compare file extensions (used with --exclude-compressed flag)
|
||||
* Different from the suffixList and should only apply to ZSTD compress operationResult
|
||||
*/
|
||||
static const char *compressedFileExtensions[] = {
|
||||
ZSTD_EXTENSION,
|
||||
TZSTD_EXTENSION,
|
||||
GZ_EXTENSION,
|
||||
TGZ_EXTENSION,
|
||||
LZMA_EXTENSION,
|
||||
XZ_EXTENSION,
|
||||
TXZ_EXTENSION,
|
||||
LZ4_EXTENSION,
|
||||
TLZ4_EXTENSION,
|
||||
NULL
|
||||
};
|
||||
|
||||
/*! FIO_compressFilename_srcFile() :
|
||||
* @return : 0 : compression completed correctly,
|
||||
@ -1451,6 +1475,15 @@ FIO_compressFilename_srcFile(FIO_prefs_t* const prefs,
|
||||
return 1;
|
||||
}
|
||||
|
||||
/* Check if "srcFile" is compressed. Only done if --exclude-compressed flag is used
|
||||
* YES => ZSTD will skip compression of the file and will return 0.
|
||||
* NO => ZSTD will resume with compress operation.
|
||||
*/
|
||||
if (prefs->excludeCompressedFiles == 1 && UTIL_isCompressedFile(srcFileName, compressedFileExtensions)) {
|
||||
DISPLAYLEVEL(4, "File is already compressed : %s \n", srcFileName);
|
||||
return 0;
|
||||
}
|
||||
|
||||
ress.srcFile = FIO_openSrcFile(srcFileName);
|
||||
if (ress.srcFile == NULL) return 1; /* srcFile could not be opened */
|
||||
|
||||
|
@ -93,6 +93,7 @@ void FIO_setLiteralCompressionMode(
|
||||
|
||||
void FIO_setNoProgress(unsigned noProgress);
|
||||
void FIO_setNotificationLevel(int level);
|
||||
void FIO_setExcludeCompressedFile(FIO_prefs_t* const prefs, int excludeCompressedFiles);
|
||||
|
||||
/*-*************************************
|
||||
* Single File functions
|
||||
|
@ -152,6 +152,8 @@ int UTIL_isSameFile(const char* fName1, const char* fName2)
|
||||
#endif
|
||||
}
|
||||
|
||||
#ifndef _MSC_VER
|
||||
/* Using this to distinguish named pipes */
|
||||
U32 UTIL_isFIFO(const char* infilename)
|
||||
{
|
||||
/* macro guards, as defined in : https://linux.die.net/man/2/lstat */
|
||||
@ -163,7 +165,7 @@ U32 UTIL_isFIFO(const char* infilename)
|
||||
(void)infilename;
|
||||
return 0;
|
||||
}
|
||||
|
||||
#endif
|
||||
|
||||
U32 UTIL_isLink(const char* infilename)
|
||||
{
|
||||
@ -525,6 +527,27 @@ int UTIL_prepareFileList(const char *dirName, char** bufStart, size_t* pos, char
|
||||
|
||||
#endif /* #ifdef _WIN32 */
|
||||
|
||||
int UTIL_isCompressedFile(const char *inputName, const char *extensionList[])
|
||||
{
|
||||
const char* ext = UTIL_getFileExtension(inputName);
|
||||
while(*extensionList!=NULL)
|
||||
{
|
||||
const int isCompressedExtension = strcmp(ext,*extensionList);
|
||||
if(isCompressedExtension==0)
|
||||
return 1;
|
||||
++extensionList;
|
||||
}
|
||||
return 0;
|
||||
}
|
||||
|
||||
/*Utility function to get file extension from file */
|
||||
const char* UTIL_getFileExtension(const char* infilename)
|
||||
{
|
||||
const char* extension = strrchr(infilename, '.');
|
||||
if(!extension || extension==infilename) return "";
|
||||
return extension;
|
||||
}
|
||||
|
||||
/*
|
||||
* UTIL_createFileList - takes a list of files and directories (params: inputNames, inputNamesNb), scans directories,
|
||||
* and returns a new list of files (params: return value, allocatedBuffer, allocatedNamesNb).
|
||||
|
@ -40,7 +40,6 @@ extern "C" {
|
||||
#include <time.h> /* clock_t, clock, CLOCKS_PER_SEC, nanosleep */
|
||||
#include "mem.h" /* U32, U64 */
|
||||
|
||||
|
||||
/*-************************************************************
|
||||
* Avoid fseek()'s 2GiB barrier with MSVC, macOS, *BSD, MinGW
|
||||
***************************************************************/
|
||||
@ -135,8 +134,12 @@ U32 UTIL_isDirectory(const char* infilename);
|
||||
int UTIL_getFileStat(const char* infilename, stat_t* statbuf);
|
||||
int UTIL_isSameFile(const char* file1, const char* file2);
|
||||
int UTIL_compareStr(const void *p1, const void *p2);
|
||||
int UTIL_isCompressedFile(const char* infilename, const char *extensionList[]);
|
||||
const char* UTIL_getFileExtension(const char* infilename);
|
||||
|
||||
#ifndef _MSC_VER
|
||||
U32 UTIL_isFIFO(const char* infilename);
|
||||
#endif
|
||||
U32 UTIL_isLink(const char* infilename);
|
||||
#define UTIL_FILESIZE_UNKNOWN ((U64)(-1))
|
||||
U64 UTIL_getFileSize(const char* infilename);
|
||||
|
@ -135,6 +135,7 @@ static int usage_advanced(const char* programName)
|
||||
DISPLAY( " -q : suppress warnings; specify twice to suppress errors too\n");
|
||||
DISPLAY( " -c : force write to standard output, even if it is the console\n");
|
||||
DISPLAY( " -l : print information about zstd compressed files \n");
|
||||
DISPLAY( "--exclude-compressed: only compress files that are not previously compressed \n");
|
||||
#ifndef ZSTD_NOCOMPRESS
|
||||
DISPLAY( "--ultra : enable levels beyond %i, up to %i (requires more memory)\n", ZSTDCLI_CLEVEL_MAX, ZSTD_maxCLevel());
|
||||
DISPLAY( "--long[=#]: enable long distance matching with given window log (default: %u)\n", g_defaultMaxWindowLog);
|
||||
@ -710,7 +711,7 @@ int main(int argCount, const char* argv[])
|
||||
if (!strcmp(argument, "--compress-literals")) { literalCompressionMode = ZSTD_lcm_huffman; continue; }
|
||||
if (!strcmp(argument, "--no-compress-literals")) { literalCompressionMode = ZSTD_lcm_uncompressed; continue; }
|
||||
if (!strcmp(argument, "--no-progress")) { FIO_setNoProgress(1); continue; }
|
||||
|
||||
if (!strcmp(argument, "--exclude-compressed")) { FIO_setExcludeCompressedFile(prefs, 1); continue; }
|
||||
/* long commands with arguments */
|
||||
#ifndef ZSTD_NODICT
|
||||
if (longCommandWArg(&argument, "--train-cover")) {
|
||||
@ -1057,7 +1058,11 @@ int main(int argCount, const char* argv[])
|
||||
if (!followLinks) {
|
||||
unsigned u;
|
||||
for (u=0, fileNamesNb=0; u<filenameIdx; u++) {
|
||||
if (UTIL_isLink(filenameTable[u]) && !UTIL_isFIFO(filenameTable[u])) {
|
||||
if (UTIL_isLink(filenameTable[u])
|
||||
#ifndef _MSC_VER
|
||||
&& !UTIL_isFIFO(filenameTable[u])
|
||||
#endif /* _MSC_VER */
|
||||
) {
|
||||
DISPLAYLEVEL(2, "Warning : %s is a symbolic link, ignoring\n", filenameTable[u]);
|
||||
} else {
|
||||
filenameTable[fileNamesNb++] = filenameTable[u];
|
||||
|
@ -73,7 +73,8 @@ FUZZ_TARGETS := \
|
||||
dictionary_round_trip \
|
||||
dictionary_decompress \
|
||||
zstd_frame_info \
|
||||
simple_compress
|
||||
simple_compress \
|
||||
dictionary_loader
|
||||
|
||||
all: $(FUZZ_TARGETS)
|
||||
|
||||
@ -110,6 +111,9 @@ simple_compress: $(FUZZ_HEADERS) $(FUZZ_OBJ) simple_compress.o
|
||||
zstd_frame_info: $(FUZZ_HEADERS) $(FUZZ_OBJ) zstd_frame_info.o
|
||||
$(CXX) $(FUZZ_TARGET_FLAGS) $(FUZZ_OBJ) zstd_frame_info.o $(LIB_FUZZING_ENGINE) -o $@
|
||||
|
||||
dictionary_loader: $(FUZZ_HEADERS) $(FUZZ_OBJ) dictionary_loader.o
|
||||
$(CXX) $(FUZZ_TARGET_FLAGS) $(FUZZ_OBJ) dictionary_loader.o $(LIB_FUZZING_ENGINE) -o $@
|
||||
|
||||
libregression.a: $(FUZZ_HEADERS) $(PRGDIR)/util.h $(PRGDIR)/util.c regression_driver.o
|
||||
$(AR) $(FUZZ_ARFLAGS) $@ regression_driver.o
|
||||
|
||||
|
93
tests/fuzz/dictionary_loader.c
Normal file
93
tests/fuzz/dictionary_loader.c
Normal file
@ -0,0 +1,93 @@
|
||||
/*
|
||||
* Copyright (c) 2016-present, Facebook, Inc.
|
||||
* All rights reserved.
|
||||
*
|
||||
* This source code is licensed under both the BSD-style license (found in the
|
||||
* LICENSE file in the root directory of this source tree) and the GPLv2 (found
|
||||
* in the COPYING file in the root directory of this source tree).
|
||||
*/
|
||||
|
||||
/**
|
||||
* This fuzz target makes sure that whenever a compression dictionary can be
|
||||
* loaded, the data can be round tripped.
|
||||
*/
|
||||
|
||||
#include <stddef.h>
|
||||
#include <stdlib.h>
|
||||
#include <stdio.h>
|
||||
#include <string.h>
|
||||
#include "fuzz_helpers.h"
|
||||
#include "zstd_helpers.h"
|
||||
#include "fuzz_data_producer.h"
|
||||
|
||||
/**
|
||||
* Compresses the data and returns the compressed size or an error.
|
||||
*/
|
||||
static size_t compress(void* compressed, size_t compressedCapacity,
|
||||
void const* source, size_t sourceSize,
|
||||
void const* dict, size_t dictSize,
|
||||
ZSTD_dictLoadMethod_e dictLoadMethod,
|
||||
ZSTD_dictContentType_e dictContentType)
|
||||
{
|
||||
ZSTD_CCtx* cctx = ZSTD_createCCtx();
|
||||
FUZZ_ZASSERT(ZSTD_CCtx_loadDictionary_advanced(
|
||||
cctx, dict, dictSize, dictLoadMethod, dictContentType));
|
||||
size_t const compressedSize = ZSTD_compress2(
|
||||
cctx, compressed, compressedCapacity, source, sourceSize);
|
||||
ZSTD_freeCCtx(cctx);
|
||||
return compressedSize;
|
||||
}
|
||||
|
||||
static size_t decompress(void* result, size_t resultCapacity,
|
||||
void const* compressed, size_t compressedSize,
|
||||
void const* dict, size_t dictSize,
|
||||
ZSTD_dictLoadMethod_e dictLoadMethod,
|
||||
ZSTD_dictContentType_e dictContentType)
|
||||
{
|
||||
ZSTD_DCtx* dctx = ZSTD_createDCtx();
|
||||
FUZZ_ZASSERT(ZSTD_DCtx_loadDictionary_advanced(
|
||||
dctx, dict, dictSize, dictLoadMethod, dictContentType));
|
||||
size_t const resultSize = ZSTD_decompressDCtx(
|
||||
dctx, result, resultCapacity, compressed, compressedSize);
|
||||
FUZZ_ZASSERT(resultSize);
|
||||
ZSTD_freeDCtx(dctx);
|
||||
return resultSize;
|
||||
}
|
||||
|
||||
int LLVMFuzzerTestOneInput(const uint8_t *src, size_t size)
|
||||
{
|
||||
FUZZ_dataProducer_t *producer = FUZZ_dataProducer_create(src, size);
|
||||
ZSTD_dictLoadMethod_e const dlm =
|
||||
size = FUZZ_dataProducer_uint32Range(producer, 0, 1);
|
||||
ZSTD_dictContentType_e const dct =
|
||||
FUZZ_dataProducer_uint32Range(producer, 0, 2);
|
||||
size = FUZZ_dataProducer_remainingBytes(producer);
|
||||
|
||||
DEBUGLOG(2, "Dict load method %d", dlm);
|
||||
DEBUGLOG(2, "Dict content type %d", dct);
|
||||
DEBUGLOG(2, "Dict size %u", (unsigned)size);
|
||||
|
||||
void* const rBuf = malloc(size);
|
||||
FUZZ_ASSERT(rBuf);
|
||||
size_t const cBufSize = ZSTD_compressBound(size);
|
||||
void* const cBuf = malloc(cBufSize);
|
||||
FUZZ_ASSERT(cBuf);
|
||||
|
||||
size_t const cSize =
|
||||
compress(cBuf, cBufSize, src, size, src, size, dlm, dct);
|
||||
/* compression failing is okay */
|
||||
if (ZSTD_isError(cSize)) {
|
||||
FUZZ_ASSERT_MSG(dct != ZSTD_dct_rawContent, "Raw must always succeed!");
|
||||
goto out;
|
||||
}
|
||||
size_t const rSize =
|
||||
decompress(rBuf, size, cBuf, cSize, src, size, dlm, dct);
|
||||
FUZZ_ASSERT_MSG(rSize == size, "Incorrect regenerated size");
|
||||
FUZZ_ASSERT_MSG(!memcmp(src, rBuf, size), "Corruption!");
|
||||
|
||||
out:
|
||||
free(cBuf);
|
||||
free(rBuf);
|
||||
FUZZ_dataProducer_free(producer);
|
||||
return 0;
|
||||
}
|
@ -27,6 +27,7 @@ def abs_join(a, *p):
|
||||
class InputType(object):
|
||||
RAW_DATA = 1
|
||||
COMPRESSED_DATA = 2
|
||||
DICTIONARY_DATA = 3
|
||||
|
||||
|
||||
class FrameType(object):
|
||||
@ -54,6 +55,7 @@ TARGET_INFO = {
|
||||
'dictionary_decompress': TargetInfo(InputType.COMPRESSED_DATA),
|
||||
'zstd_frame_info': TargetInfo(InputType.COMPRESSED_DATA),
|
||||
'simple_compress': TargetInfo(InputType.RAW_DATA),
|
||||
'dictionary_loader': TargetInfo(InputType.DICTIONARY_DATA),
|
||||
}
|
||||
TARGETS = list(TARGET_INFO.keys())
|
||||
ALL_TARGETS = TARGETS + ['all']
|
||||
@ -73,6 +75,7 @@ LIB_FUZZING_ENGINE = os.environ.get('LIB_FUZZING_ENGINE', 'libregression.a')
|
||||
AFL_FUZZ = os.environ.get('AFL_FUZZ', 'afl-fuzz')
|
||||
DECODECORPUS = os.environ.get('DECODECORPUS',
|
||||
abs_join(FUZZ_DIR, '..', 'decodecorpus'))
|
||||
ZSTD = os.environ.get('ZSTD', abs_join(FUZZ_DIR, '..', '..', 'zstd'))
|
||||
|
||||
# Sanitizer environment variables
|
||||
MSAN_EXTRA_CPPFLAGS = os.environ.get('MSAN_EXTRA_CPPFLAGS', '')
|
||||
@ -673,6 +676,11 @@ def gen_parser(args):
|
||||
default=DECODECORPUS,
|
||||
help="decodecorpus binary (default: $DECODECORPUS='{}')".format(
|
||||
DECODECORPUS))
|
||||
parser.add_argument(
|
||||
'--zstd',
|
||||
type=str,
|
||||
default=ZSTD,
|
||||
help="zstd binary (default: $ZSTD='{}')".format(ZSTD))
|
||||
parser.add_argument(
|
||||
'--fuzz-rng-seed-size',
|
||||
type=int,
|
||||
@ -707,46 +715,66 @@ def gen(args):
|
||||
return 1
|
||||
|
||||
seed = create(args.seed)
|
||||
with tmpdir() as compressed:
|
||||
with tmpdir() as decompressed:
|
||||
cmd = [
|
||||
args.decodecorpus,
|
||||
'-n{}'.format(args.number),
|
||||
'-p{}/'.format(compressed),
|
||||
'-o{}'.format(decompressed),
|
||||
with tmpdir() as compressed, tmpdir() as decompressed, tmpdir() as dict:
|
||||
info = TARGET_INFO[args.TARGET]
|
||||
|
||||
if info.input_type == InputType.DICTIONARY_DATA:
|
||||
number = max(args.number, 1000)
|
||||
else:
|
||||
number = args.number
|
||||
cmd = [
|
||||
args.decodecorpus,
|
||||
'-n{}'.format(args.number),
|
||||
'-p{}/'.format(compressed),
|
||||
'-o{}'.format(decompressed),
|
||||
]
|
||||
|
||||
if info.frame_type == FrameType.BLOCK:
|
||||
cmd += [
|
||||
'--gen-blocks',
|
||||
'--max-block-size-log={}'.format(min(args.max_size_log, 17))
|
||||
]
|
||||
else:
|
||||
cmd += ['--max-content-size-log={}'.format(args.max_size_log)]
|
||||
|
||||
info = TARGET_INFO[args.TARGET]
|
||||
if info.frame_type == FrameType.BLOCK:
|
||||
cmd += [
|
||||
'--gen-blocks',
|
||||
'--max-block-size-log={}'.format(min(args.max_size_log, 17))
|
||||
print(' '.join(cmd))
|
||||
subprocess.check_call(cmd)
|
||||
|
||||
if info.input_type == InputType.RAW_DATA:
|
||||
print('using decompressed data in {}'.format(decompressed))
|
||||
samples = decompressed
|
||||
elif info.input_type == InputType.COMPRESSED_DATA:
|
||||
print('using compressed data in {}'.format(compressed))
|
||||
samples = compressed
|
||||
else:
|
||||
assert info.input_type == InputType.DICTIONARY_DATA
|
||||
print('making dictionary data from {}'.format(decompressed))
|
||||
samples = dict
|
||||
min_dict_size_log = 9
|
||||
max_dict_size_log = max(min_dict_size_log + 1, args.max_size_log)
|
||||
for dict_size_log in range(min_dict_size_log, max_dict_size_log):
|
||||
dict_size = 1 << dict_size_log
|
||||
cmd = [
|
||||
args.zstd,
|
||||
'--train',
|
||||
'-r', decompressed,
|
||||
'--maxdict={}'.format(dict_size),
|
||||
'-o', abs_join(dict, '{}.zstd-dict'.format(dict_size))
|
||||
]
|
||||
else:
|
||||
cmd += ['--max-content-size-log={}'.format(args.max_size_log)]
|
||||
print(' '.join(cmd))
|
||||
subprocess.check_call(cmd)
|
||||
|
||||
print(' '.join(cmd))
|
||||
subprocess.check_call(cmd)
|
||||
|
||||
if info.input_type == InputType.RAW_DATA:
|
||||
print('using decompressed data in {}'.format(decompressed))
|
||||
samples = decompressed
|
||||
else:
|
||||
assert info.input_type == InputType.COMPRESSED_DATA
|
||||
print('using compressed data in {}'.format(compressed))
|
||||
samples = compressed
|
||||
|
||||
# Copy the samples over and prepend the RNG seeds
|
||||
for name in os.listdir(samples):
|
||||
samplename = abs_join(samples, name)
|
||||
outname = abs_join(seed, name)
|
||||
with open(samplename, 'rb') as sample:
|
||||
with open(outname, 'wb') as out:
|
||||
CHUNK_SIZE = 131072
|
||||
# Copy the samples over and prepend the RNG seeds
|
||||
for name in os.listdir(samples):
|
||||
samplename = abs_join(samples, name)
|
||||
outname = abs_join(seed, name)
|
||||
with open(samplename, 'rb') as sample:
|
||||
with open(outname, 'wb') as out:
|
||||
CHUNK_SIZE = 131072
|
||||
chunk = sample.read(CHUNK_SIZE)
|
||||
while len(chunk) > 0:
|
||||
out.write(chunk)
|
||||
chunk = sample.read(CHUNK_SIZE)
|
||||
while len(chunk) > 0:
|
||||
out.write(chunk)
|
||||
chunk = sample.read(CHUNK_SIZE)
|
||||
return 0
|
||||
|
||||
|
||||
|
@ -215,6 +215,37 @@ $ZSTD tmp -c --compress-literals -19 | $ZSTD -t
|
||||
$ZSTD -b --fast=1 -i0e1 tmp --compress-literals
|
||||
$ZSTD -b --fast=1 -i0e1 tmp --no-compress-literals
|
||||
|
||||
println "test: --exclude-compressed flag"
|
||||
rm -rf precompressedFilterTestDir
|
||||
mkdir -p precompressedFilterTestDir
|
||||
./datagen $size > precompressedFilterTestDir/input.5
|
||||
./datagen $size > precompressedFilterTestDir/input.6
|
||||
$ZSTD --exclude-compressed --long --rm -r precompressedFilterTestDir
|
||||
sleep 5
|
||||
./datagen $size > precompressedFilterTestDir/input.7
|
||||
./datagen $size > precompressedFilterTestDir/input.8
|
||||
$ZSTD --exclude-compressed --long --rm -r precompressedFilterTestDir
|
||||
test ! -f precompressedFilterTestDir/input.5.zst.zst
|
||||
test ! -f precompressedFilterTestDir/input.6.zst.zst
|
||||
file1timestamp=`date -r precompressedFilterTestDir/input.5.zst +%s`
|
||||
file2timestamp=`date -r precompressedFilterTestDir/input.7.zst +%s`
|
||||
if [[ $file2timestamp -ge $file1timestamp ]]; then
|
||||
println "Test is successful. input.5.zst is precompressed and therefore not compressed/modified again."
|
||||
else
|
||||
println "Test is not successful"
|
||||
fi
|
||||
#File Extension check.
|
||||
./datagen $size > precompressedFilterTestDir/input.zstbar
|
||||
$ZSTD --exclude-compressed --long --rm -r precompressedFilterTestDir
|
||||
#ZSTD should compress input.zstbar
|
||||
test -f precompressedFilterTestDir/input.zstbar.zst
|
||||
#Check without the --exclude-compressed flag
|
||||
$ZSTD --long --rm -r precompressedFilterTestDir
|
||||
#Files should get compressed again without the --exclude-compressed flag.
|
||||
test -f precompressedFilterTestDir/input.5.zst.zst
|
||||
test -f precompressedFilterTestDir/input.6.zst.zst
|
||||
println "Test completed"
|
||||
|
||||
println "test : file removal"
|
||||
$ZSTD -f --rm tmp
|
||||
test ! -f tmp # tmp should no longer be present
|
||||
@ -1207,6 +1238,8 @@ test -f dictionary
|
||||
rm -f tmp* dictionary
|
||||
|
||||
|
||||
if [ "$isWindows" = false ] ; then
|
||||
|
||||
println "\n===> zstd fifo named pipe test "
|
||||
head -c 10 /dev/zero > tmp_original
|
||||
mkfifo named_pipe
|
||||
@ -1217,4 +1250,6 @@ $DIFF -s tmp_original tmp_decompressed
|
||||
rm -rf tmp*
|
||||
rm -rf named_pipe
|
||||
|
||||
fi
|
||||
|
||||
rm -f tmp*
|
||||
|
@ -103,6 +103,7 @@ zstd_zlibwrapper.o: $(ZLIBWRAPPER_PATH)/zstd_zlibwrapper.c $(ZLIBWRAPPER_PATH)/z
|
||||
|
||||
zstdTurnedOn_zlibwrapper.o: CPPFLAGS += -DZWRAP_USE_ZSTD=1
|
||||
zstdTurnedOn_zlibwrapper.o: $(ZLIBWRAPPER_PATH)/zstd_zlibwrapper.c $(ZLIBWRAPPER_PATH)/zstd_zlibwrapper.h
|
||||
$(CC) $(CPPFLAGS) $(CFLAGS) $< -c -o $@
|
||||
|
||||
$(ZSTDLIBDIR)/libzstd.a:
|
||||
$(MAKE) -C $(ZSTDLIBDIR) libzstd.a
|
||||
|
Loading…
Reference in New Issue
Block a user