2020-02-08 13:56:59 +00:00
/*
2023-02-10 19:25:40 +00:00
* Copyright ( c ) Meta Platforms , Inc . and affiliates .
2020-02-08 13:56:59 +00:00
* 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 ) .
* You may select , at your option , one of the above - listed licenses .
*/
/*-*************************************
* Dependencies
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
2021-05-14 22:48:15 +00:00
# include "../common/zstd_deps.h" /* INT_MAX, ZSTD_memset, ZSTD_memcpy */
# include "../common/mem.h"
2020-02-08 13:56:59 +00:00
# include "hist.h" /* HIST_countFast_wksp */
# define FSE_STATIC_LINKING_ONLY /* FSE_encodeSymbol */
2021-05-14 22:48:15 +00:00
# include "../common/fse.h"
# include "../common/huf.h"
2020-02-08 13:56:59 +00:00
# include "zstd_compress_internal.h"
# include "zstd_compress_sequences.h"
# include "zstd_compress_literals.h"
# include "zstd_fast.h"
# include "zstd_double_fast.h"
# include "zstd_lazy.h"
# include "zstd_opt.h"
# include "zstd_ldm.h"
# include "zstd_compress_superblock.h"
2023-02-10 19:25:40 +00:00
# include "../common/bits.h" /* ZSTD_highbit32 */
2020-02-08 13:56:59 +00:00
2020-12-18 14:46:27 +00:00
/* ***************************************************************
* Tuning parameters
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
/*!
* COMPRESS_HEAPMODE :
* Select how default decompression function ZSTD_compress ( ) allocates its context ,
* on stack ( 0 , default ) , or into heap ( 1 ) .
* Note that functions with explicit context such as ZSTD_compressCCtx ( ) are unaffected .
*/
# ifndef ZSTD_COMPRESS_HEAPMODE
# define ZSTD_COMPRESS_HEAPMODE 0
# endif
2021-12-21 11:21:29 +00:00
/*!
* ZSTD_HASHLOG3_MAX :
* Maximum size of the hash table dedicated to find 3 - bytes matches ,
* in log format , aka 17 = > 1 < < 17 = = 128 Ki positions .
* This structure is only used in zstd_opt .
* Since allocation is centralized for all strategies , it has to be known here .
* The actual ( selected ) size of the hash table is then stored in ZSTD_matchState_t . hashLog3 ,
* so that zstd_opt . c doesn ' t need to know about this constant .
*/
# ifndef ZSTD_HASHLOG3_MAX
# define ZSTD_HASHLOG3_MAX 17
# endif
2020-02-08 13:56:59 +00:00
/*-*************************************
* Helper functions
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
/* ZSTD_compressBound()
2023-02-10 19:25:40 +00:00
* Note that the result from this function is only valid for
* the one - pass compression functions .
* When employing the streaming mode ,
* if flushes are frequently altering the size of blocks ,
* the overhead from block headers can make the compressed data larger
* than the return value of ZSTD_compressBound ( ) .
2020-02-08 13:56:59 +00:00
*/
size_t ZSTD_compressBound ( size_t srcSize ) {
2023-02-10 19:25:40 +00:00
size_t const r = ZSTD_COMPRESSBOUND ( srcSize ) ;
if ( r = = 0 ) return ERROR ( srcSize_wrong ) ;
return r ;
2020-02-08 13:56:59 +00:00
}
/*-*************************************
* Context memory management
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
struct ZSTD_CDict_s {
const void * dictContent ;
size_t dictContentSize ;
2020-12-18 14:46:27 +00:00
ZSTD_dictContentType_e dictContentType ; /* The dictContentType the CDict was created with */
2020-02-08 13:56:59 +00:00
U32 * entropyWorkspace ; /* entropy workspace of HUF_WORKSPACE_SIZE bytes */
ZSTD_cwksp workspace ;
ZSTD_matchState_t matchState ;
ZSTD_compressedBlockState_t cBlockState ;
ZSTD_customMem customMem ;
U32 dictID ;
int compressionLevel ; /* 0 indicates that advanced API was used to select CDict params */
2021-12-21 11:21:29 +00:00
ZSTD_paramSwitch_e useRowMatchFinder ; /* Indicates whether the CDict was created with params that would use
* row - based matchfinder . Unless the cdict is reloaded , we will use
* the same greedy / lazy matchfinder at compression time .
*/
2020-02-08 13:56:59 +00:00
} ; /* typedef'd to ZSTD_CDict within "zstd.h" */
ZSTD_CCtx * ZSTD_createCCtx ( void )
{
return ZSTD_createCCtx_advanced ( ZSTD_defaultCMem ) ;
}
static void ZSTD_initCCtx ( ZSTD_CCtx * cctx , ZSTD_customMem memManager )
{
assert ( cctx ! = NULL ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memset ( cctx , 0 , sizeof ( * cctx ) ) ;
2020-02-08 13:56:59 +00:00
cctx - > customMem = memManager ;
2021-12-21 11:21:29 +00:00
cctx - > bmi2 = ZSTD_cpuSupportsBmi2 ( ) ;
2020-02-08 13:56:59 +00:00
{ size_t const err = ZSTD_CCtx_reset ( cctx , ZSTD_reset_parameters ) ;
assert ( ! ZSTD_isError ( err ) ) ;
( void ) err ;
}
}
ZSTD_CCtx * ZSTD_createCCtx_advanced ( ZSTD_customMem customMem )
{
ZSTD_STATIC_ASSERT ( zcss_init = = 0 ) ;
ZSTD_STATIC_ASSERT ( ZSTD_CONTENTSIZE_UNKNOWN = = ( 0ULL - 1 ) ) ;
2020-12-18 14:46:27 +00:00
if ( ( ! customMem . customAlloc ) ^ ( ! customMem . customFree ) ) return NULL ;
{ ZSTD_CCtx * const cctx = ( ZSTD_CCtx * ) ZSTD_customMalloc ( sizeof ( ZSTD_CCtx ) , customMem ) ;
2020-02-08 13:56:59 +00:00
if ( ! cctx ) return NULL ;
ZSTD_initCCtx ( cctx , customMem ) ;
return cctx ;
}
}
2020-06-04 19:03:27 +00:00
ZSTD_CCtx * ZSTD_initStaticCCtx ( void * workspace , size_t workspaceSize )
2020-02-08 13:56:59 +00:00
{
ZSTD_cwksp ws ;
ZSTD_CCtx * cctx ;
if ( workspaceSize < = sizeof ( ZSTD_CCtx ) ) return NULL ; /* minimum size */
if ( ( size_t ) workspace & 7 ) return NULL ; /* must be 8-aligned */
2020-12-18 14:46:27 +00:00
ZSTD_cwksp_init ( & ws , workspace , workspaceSize , ZSTD_cwksp_static_alloc ) ;
2020-02-08 13:56:59 +00:00
cctx = ( ZSTD_CCtx * ) ZSTD_cwksp_reserve_object ( & ws , sizeof ( ZSTD_CCtx ) ) ;
2020-06-04 19:03:27 +00:00
if ( cctx = = NULL ) return NULL ;
2020-12-18 14:46:27 +00:00
ZSTD_memset ( cctx , 0 , sizeof ( ZSTD_CCtx ) ) ;
2020-02-08 13:56:59 +00:00
ZSTD_cwksp_move ( & cctx - > workspace , & ws ) ;
cctx - > staticSize = workspaceSize ;
/* statically sized space. entropyWorkspace never moves (but prev/next block swap places) */
2020-12-18 14:46:27 +00:00
if ( ! ZSTD_cwksp_check_available ( & cctx - > workspace , ENTROPY_WORKSPACE_SIZE + 2 * sizeof ( ZSTD_compressedBlockState_t ) ) ) return NULL ;
2020-02-08 13:56:59 +00:00
cctx - > blockState . prevCBlock = ( ZSTD_compressedBlockState_t * ) ZSTD_cwksp_reserve_object ( & cctx - > workspace , sizeof ( ZSTD_compressedBlockState_t ) ) ;
cctx - > blockState . nextCBlock = ( ZSTD_compressedBlockState_t * ) ZSTD_cwksp_reserve_object ( & cctx - > workspace , sizeof ( ZSTD_compressedBlockState_t ) ) ;
2020-12-18 14:46:27 +00:00
cctx - > entropyWorkspace = ( U32 * ) ZSTD_cwksp_reserve_object ( & cctx - > workspace , ENTROPY_WORKSPACE_SIZE ) ;
2020-02-08 13:56:59 +00:00
cctx - > bmi2 = ZSTD_cpuid_bmi2 ( ZSTD_cpuid ( ) ) ;
return cctx ;
}
/**
* Clears and frees all of the dictionaries in the CCtx .
*/
static void ZSTD_clearAllDicts ( ZSTD_CCtx * cctx )
{
2020-12-18 14:46:27 +00:00
ZSTD_customFree ( cctx - > localDict . dictBuffer , cctx - > customMem ) ;
2020-02-08 13:56:59 +00:00
ZSTD_freeCDict ( cctx - > localDict . cdict ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memset ( & cctx - > localDict , 0 , sizeof ( cctx - > localDict ) ) ;
ZSTD_memset ( & cctx - > prefixDict , 0 , sizeof ( cctx - > prefixDict ) ) ;
2020-02-08 13:56:59 +00:00
cctx - > cdict = NULL ;
}
static size_t ZSTD_sizeof_localDict ( ZSTD_localDict dict )
{
size_t const bufferSize = dict . dictBuffer ! = NULL ? dict . dictSize : 0 ;
size_t const cdictSize = ZSTD_sizeof_CDict ( dict . cdict ) ;
return bufferSize + cdictSize ;
}
static void ZSTD_freeCCtxContent ( ZSTD_CCtx * cctx )
{
assert ( cctx ! = NULL ) ;
assert ( cctx - > staticSize = = 0 ) ;
ZSTD_clearAllDicts ( cctx ) ;
# ifdef ZSTD_MULTITHREAD
ZSTDMT_freeCCtx ( cctx - > mtctx ) ; cctx - > mtctx = NULL ;
# endif
ZSTD_cwksp_free ( & cctx - > workspace , cctx - > customMem ) ;
}
size_t ZSTD_freeCCtx ( ZSTD_CCtx * cctx )
{
if ( cctx = = NULL ) return 0 ; /* support free on NULL */
RETURN_ERROR_IF ( cctx - > staticSize , memory_allocation ,
" not compatible with static CCtx " ) ;
2023-02-10 19:25:40 +00:00
{ int cctxInWorkspace = ZSTD_cwksp_owns_buffer ( & cctx - > workspace , cctx ) ;
2020-02-08 13:56:59 +00:00
ZSTD_freeCCtxContent ( cctx ) ;
2023-02-10 19:25:40 +00:00
if ( ! cctxInWorkspace ) ZSTD_customFree ( cctx , cctx - > customMem ) ;
2020-02-08 13:56:59 +00:00
}
return 0 ;
}
static size_t ZSTD_sizeof_mtctx ( const ZSTD_CCtx * cctx )
{
# ifdef ZSTD_MULTITHREAD
return ZSTDMT_sizeof_CCtx ( cctx - > mtctx ) ;
# else
( void ) cctx ;
return 0 ;
# endif
}
size_t ZSTD_sizeof_CCtx ( const ZSTD_CCtx * cctx )
{
if ( cctx = = NULL ) return 0 ; /* support sizeof on NULL */
/* cctx may be in the workspace */
return ( cctx - > workspace . workspace = = cctx ? 0 : sizeof ( * cctx ) )
+ ZSTD_cwksp_sizeof ( & cctx - > workspace )
+ ZSTD_sizeof_localDict ( cctx - > localDict )
+ ZSTD_sizeof_mtctx ( cctx ) ;
}
size_t ZSTD_sizeof_CStream ( const ZSTD_CStream * zcs )
{
return ZSTD_sizeof_CCtx ( zcs ) ; /* same object */
}
/* private API call, for dictBuilder only */
const seqStore_t * ZSTD_getSeqStore ( const ZSTD_CCtx * ctx ) { return & ( ctx - > seqStore ) ; }
2021-05-14 22:48:15 +00:00
/* Returns true if the strategy supports using a row based matchfinder */
static int ZSTD_rowMatchFinderSupported ( const ZSTD_strategy strategy ) {
return ( strategy > = ZSTD_greedy & & strategy < = ZSTD_lazy2 ) ;
}
/* Returns true if the strategy and useRowMatchFinder mode indicate that we will use the row based matchfinder
* for this compression .
*/
2021-12-21 11:21:29 +00:00
static int ZSTD_rowMatchFinderUsed ( const ZSTD_strategy strategy , const ZSTD_paramSwitch_e mode ) {
assert ( mode ! = ZSTD_ps_auto ) ;
return ZSTD_rowMatchFinderSupported ( strategy ) & & ( mode = = ZSTD_ps_enable ) ;
2021-05-14 22:48:15 +00:00
}
2021-12-21 11:21:29 +00:00
/* Returns row matchfinder usage given an initial mode and cParams */
static ZSTD_paramSwitch_e ZSTD_resolveRowMatchFinderMode ( ZSTD_paramSwitch_e mode ,
const ZSTD_compressionParameters * const cParams ) {
# if defined(ZSTD_ARCH_X86_SSE2) || defined(ZSTD_ARCH_ARM_NEON)
2021-05-14 22:48:15 +00:00
int const kHasSIMD128 = 1 ;
# else
int const kHasSIMD128 = 0 ;
# endif
2021-12-21 11:21:29 +00:00
if ( mode ! = ZSTD_ps_auto ) return mode ; /* if requested enabled, but no SIMD, we still will use row matchfinder */
mode = ZSTD_ps_disable ;
2021-05-14 22:48:15 +00:00
if ( ! ZSTD_rowMatchFinderSupported ( cParams - > strategy ) ) return mode ;
if ( kHasSIMD128 ) {
2021-12-21 11:21:29 +00:00
if ( cParams - > windowLog > 14 ) mode = ZSTD_ps_enable ;
2021-05-14 22:48:15 +00:00
} else {
2021-12-21 11:21:29 +00:00
if ( cParams - > windowLog > 17 ) mode = ZSTD_ps_enable ;
2021-05-14 22:48:15 +00:00
}
return mode ;
}
2021-12-21 11:21:29 +00:00
/* Returns block splitter usage (generally speaking, when using slower/stronger compression modes) */
static ZSTD_paramSwitch_e ZSTD_resolveBlockSplitterMode ( ZSTD_paramSwitch_e mode ,
const ZSTD_compressionParameters * const cParams ) {
if ( mode ! = ZSTD_ps_auto ) return mode ;
return ( cParams - > strategy > = ZSTD_btopt & & cParams - > windowLog > = 17 ) ? ZSTD_ps_enable : ZSTD_ps_disable ;
}
2021-05-14 22:48:15 +00:00
/* Returns 1 if the arguments indicate that we should allocate a chainTable, 0 otherwise */
static int ZSTD_allocateChainTable ( const ZSTD_strategy strategy ,
2021-12-21 11:21:29 +00:00
const ZSTD_paramSwitch_e useRowMatchFinder ,
2021-05-14 22:48:15 +00:00
const U32 forDDSDict ) {
2021-12-21 11:21:29 +00:00
assert ( useRowMatchFinder ! = ZSTD_ps_auto ) ;
2021-05-14 22:48:15 +00:00
/* We always should allocate a chaintable if we are allocating a matchstate for a DDS dictionary matchstate.
* We do not allocate a chaintable if we are using ZSTD_fast , or are using the row - based matchfinder .
*/
return forDDSDict | | ( ( strategy ! = ZSTD_fast ) & & ! ZSTD_rowMatchFinderUsed ( strategy , useRowMatchFinder ) ) ;
}
2023-02-10 19:25:40 +00:00
/* Returns ZSTD_ps_enable if compression parameters are such that we should
2020-12-18 14:46:27 +00:00
* enable long distance matching ( wlog > = 27 , strategy > = btopt ) .
2023-02-10 19:25:40 +00:00
* Returns ZSTD_ps_disable otherwise .
2020-12-18 14:46:27 +00:00
*/
2021-12-21 11:21:29 +00:00
static ZSTD_paramSwitch_e ZSTD_resolveEnableLdm ( ZSTD_paramSwitch_e mode ,
const ZSTD_compressionParameters * const cParams ) {
if ( mode ! = ZSTD_ps_auto ) return mode ;
return ( cParams - > strategy > = ZSTD_btopt & & cParams - > windowLog > = 27 ) ? ZSTD_ps_enable : ZSTD_ps_disable ;
2021-05-14 22:48:15 +00:00
}
2023-02-10 19:25:40 +00:00
static int ZSTD_resolveExternalSequenceValidation ( int mode ) {
return mode ;
}
/* Resolves maxBlockSize to the default if no value is present. */
static size_t ZSTD_resolveMaxBlockSize ( size_t maxBlockSize ) {
if ( maxBlockSize = = 0 ) {
return ZSTD_BLOCKSIZE_MAX ;
} else {
return maxBlockSize ;
}
}
static ZSTD_paramSwitch_e ZSTD_resolveExternalRepcodeSearch ( ZSTD_paramSwitch_e value , int cLevel ) {
if ( value ! = ZSTD_ps_auto ) return value ;
if ( cLevel < 10 ) {
return ZSTD_ps_disable ;
} else {
return ZSTD_ps_enable ;
}
}
/* Returns 1 if compression parameters are such that CDict hashtable and chaintable indices are tagged.
* If so , the tags need to be removed in ZSTD_resetCCtx_byCopyingCDict . */
static int ZSTD_CDictIndicesAreTagged ( const ZSTD_compressionParameters * const cParams ) {
return cParams - > strategy = = ZSTD_fast | | cParams - > strategy = = ZSTD_dfast ;
}
2020-02-08 13:56:59 +00:00
static ZSTD_CCtx_params ZSTD_makeCCtxParamsFromCParams (
ZSTD_compressionParameters cParams )
{
ZSTD_CCtx_params cctxParams ;
2020-12-18 14:46:27 +00:00
/* should not matter, as all cParams are presumed properly defined */
ZSTD_CCtxParams_init ( & cctxParams , ZSTD_CLEVEL_DEFAULT ) ;
2020-02-08 13:56:59 +00:00
cctxParams . cParams = cParams ;
2020-12-18 14:46:27 +00:00
2021-05-14 22:48:15 +00:00
/* Adjust advanced params according to cParams */
2021-12-21 11:21:29 +00:00
cctxParams . ldmParams . enableLdm = ZSTD_resolveEnableLdm ( cctxParams . ldmParams . enableLdm , & cParams ) ;
if ( cctxParams . ldmParams . enableLdm = = ZSTD_ps_enable ) {
2020-12-18 14:46:27 +00:00
ZSTD_ldm_adjustParameters ( & cctxParams . ldmParams , & cParams ) ;
assert ( cctxParams . ldmParams . hashLog > = cctxParams . ldmParams . bucketSizeLog ) ;
assert ( cctxParams . ldmParams . hashRateLog < 32 ) ;
}
2021-12-21 11:21:29 +00:00
cctxParams . useBlockSplitter = ZSTD_resolveBlockSplitterMode ( cctxParams . useBlockSplitter , & cParams ) ;
2021-05-14 22:48:15 +00:00
cctxParams . useRowMatchFinder = ZSTD_resolveRowMatchFinderMode ( cctxParams . useRowMatchFinder , & cParams ) ;
2023-02-10 19:25:40 +00:00
cctxParams . validateSequences = ZSTD_resolveExternalSequenceValidation ( cctxParams . validateSequences ) ;
cctxParams . maxBlockSize = ZSTD_resolveMaxBlockSize ( cctxParams . maxBlockSize ) ;
cctxParams . searchForExternalRepcodes = ZSTD_resolveExternalRepcodeSearch ( cctxParams . searchForExternalRepcodes ,
cctxParams . compressionLevel ) ;
2020-02-08 13:56:59 +00:00
assert ( ! ZSTD_checkCParams ( cParams ) ) ;
return cctxParams ;
}
static ZSTD_CCtx_params * ZSTD_createCCtxParams_advanced (
ZSTD_customMem customMem )
{
ZSTD_CCtx_params * params ;
2020-12-18 14:46:27 +00:00
if ( ( ! customMem . customAlloc ) ^ ( ! customMem . customFree ) ) return NULL ;
params = ( ZSTD_CCtx_params * ) ZSTD_customCalloc (
2020-02-08 13:56:59 +00:00
sizeof ( ZSTD_CCtx_params ) , customMem ) ;
if ( ! params ) { return NULL ; }
2020-12-18 14:46:27 +00:00
ZSTD_CCtxParams_init ( params , ZSTD_CLEVEL_DEFAULT ) ;
2020-02-08 13:56:59 +00:00
params - > customMem = customMem ;
return params ;
}
ZSTD_CCtx_params * ZSTD_createCCtxParams ( void )
{
return ZSTD_createCCtxParams_advanced ( ZSTD_defaultCMem ) ;
}
size_t ZSTD_freeCCtxParams ( ZSTD_CCtx_params * params )
{
if ( params = = NULL ) { return 0 ; }
2020-12-18 14:46:27 +00:00
ZSTD_customFree ( params , params - > customMem ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
size_t ZSTD_CCtxParams_reset ( ZSTD_CCtx_params * params )
{
return ZSTD_CCtxParams_init ( params , ZSTD_CLEVEL_DEFAULT ) ;
}
size_t ZSTD_CCtxParams_init ( ZSTD_CCtx_params * cctxParams , int compressionLevel ) {
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( ! cctxParams , GENERIC , " NULL pointer! " ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memset ( cctxParams , 0 , sizeof ( * cctxParams ) ) ;
2020-02-08 13:56:59 +00:00
cctxParams - > compressionLevel = compressionLevel ;
cctxParams - > fParams . contentSizeFlag = 1 ;
return 0 ;
}
2021-05-14 22:48:15 +00:00
# define ZSTD_NO_CLEVEL 0
/**
2023-02-10 19:25:40 +00:00
* Initializes ` cctxParams ` from ` params ` and ` compressionLevel ` .
2021-05-14 22:48:15 +00:00
* @ param compressionLevel If params are derived from a compression level then that compression level , otherwise ZSTD_NO_CLEVEL .
*/
2023-02-10 19:25:40 +00:00
static void
ZSTD_CCtxParams_init_internal ( ZSTD_CCtx_params * cctxParams ,
const ZSTD_parameters * params ,
int compressionLevel )
2021-05-14 22:48:15 +00:00
{
assert ( ! ZSTD_checkCParams ( params - > cParams ) ) ;
ZSTD_memset ( cctxParams , 0 , sizeof ( * cctxParams ) ) ;
cctxParams - > cParams = params - > cParams ;
cctxParams - > fParams = params - > fParams ;
/* Should not matter, as all cParams are presumed properly defined.
* But , set it for tracing anyway .
*/
cctxParams - > compressionLevel = compressionLevel ;
cctxParams - > useRowMatchFinder = ZSTD_resolveRowMatchFinderMode ( cctxParams - > useRowMatchFinder , & params - > cParams ) ;
2021-12-21 11:21:29 +00:00
cctxParams - > useBlockSplitter = ZSTD_resolveBlockSplitterMode ( cctxParams - > useBlockSplitter , & params - > cParams ) ;
cctxParams - > ldmParams . enableLdm = ZSTD_resolveEnableLdm ( cctxParams - > ldmParams . enableLdm , & params - > cParams ) ;
2023-02-10 19:25:40 +00:00
cctxParams - > validateSequences = ZSTD_resolveExternalSequenceValidation ( cctxParams - > validateSequences ) ;
cctxParams - > maxBlockSize = ZSTD_resolveMaxBlockSize ( cctxParams - > maxBlockSize ) ;
cctxParams - > searchForExternalRepcodes = ZSTD_resolveExternalRepcodeSearch ( cctxParams - > searchForExternalRepcodes , compressionLevel ) ;
2021-12-21 11:21:29 +00:00
DEBUGLOG ( 4 , " ZSTD_CCtxParams_init_internal: useRowMatchFinder=%d, useBlockSplitter=%d ldm=%d " ,
cctxParams - > useRowMatchFinder , cctxParams - > useBlockSplitter , cctxParams - > ldmParams . enableLdm ) ;
2021-05-14 22:48:15 +00:00
}
2020-02-08 13:56:59 +00:00
size_t ZSTD_CCtxParams_init_advanced ( ZSTD_CCtx_params * cctxParams , ZSTD_parameters params )
{
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( ! cctxParams , GENERIC , " NULL pointer! " ) ;
FORWARD_IF_ERROR ( ZSTD_checkCParams ( params . cParams ) , " " ) ;
2021-05-14 22:48:15 +00:00
ZSTD_CCtxParams_init_internal ( cctxParams , & params , ZSTD_NO_CLEVEL ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
2021-05-14 22:48:15 +00:00
/**
* Sets cctxParams ' cParams and fParams from params , but otherwise leaves them alone .
2023-02-10 19:25:40 +00:00
* @ param params Validated zstd parameters .
2021-05-14 22:48:15 +00:00
*/
static void ZSTD_CCtxParams_setZstdParams (
ZSTD_CCtx_params * cctxParams , const ZSTD_parameters * params )
2020-02-08 13:56:59 +00:00
{
assert ( ! ZSTD_checkCParams ( params - > cParams ) ) ;
2021-05-14 22:48:15 +00:00
cctxParams - > cParams = params - > cParams ;
cctxParams - > fParams = params - > fParams ;
/* Should not matter, as all cParams are presumed properly defined.
* But , set it for tracing anyway .
*/
cctxParams - > compressionLevel = ZSTD_NO_CLEVEL ;
2020-02-08 13:56:59 +00:00
}
ZSTD_bounds ZSTD_cParam_getBounds ( ZSTD_cParameter param )
{
ZSTD_bounds bounds = { 0 , 0 , 0 } ;
switch ( param )
{
case ZSTD_c_compressionLevel :
bounds . lowerBound = ZSTD_minCLevel ( ) ;
bounds . upperBound = ZSTD_maxCLevel ( ) ;
return bounds ;
case ZSTD_c_windowLog :
bounds . lowerBound = ZSTD_WINDOWLOG_MIN ;
bounds . upperBound = ZSTD_WINDOWLOG_MAX ;
return bounds ;
case ZSTD_c_hashLog :
bounds . lowerBound = ZSTD_HASHLOG_MIN ;
bounds . upperBound = ZSTD_HASHLOG_MAX ;
return bounds ;
case ZSTD_c_chainLog :
bounds . lowerBound = ZSTD_CHAINLOG_MIN ;
bounds . upperBound = ZSTD_CHAINLOG_MAX ;
return bounds ;
case ZSTD_c_searchLog :
bounds . lowerBound = ZSTD_SEARCHLOG_MIN ;
bounds . upperBound = ZSTD_SEARCHLOG_MAX ;
return bounds ;
case ZSTD_c_minMatch :
bounds . lowerBound = ZSTD_MINMATCH_MIN ;
bounds . upperBound = ZSTD_MINMATCH_MAX ;
return bounds ;
case ZSTD_c_targetLength :
bounds . lowerBound = ZSTD_TARGETLENGTH_MIN ;
bounds . upperBound = ZSTD_TARGETLENGTH_MAX ;
return bounds ;
case ZSTD_c_strategy :
bounds . lowerBound = ZSTD_STRATEGY_MIN ;
bounds . upperBound = ZSTD_STRATEGY_MAX ;
return bounds ;
case ZSTD_c_contentSizeFlag :
bounds . lowerBound = 0 ;
bounds . upperBound = 1 ;
return bounds ;
case ZSTD_c_checksumFlag :
bounds . lowerBound = 0 ;
bounds . upperBound = 1 ;
return bounds ;
case ZSTD_c_dictIDFlag :
bounds . lowerBound = 0 ;
bounds . upperBound = 1 ;
return bounds ;
case ZSTD_c_nbWorkers :
bounds . lowerBound = 0 ;
# ifdef ZSTD_MULTITHREAD
bounds . upperBound = ZSTDMT_NBWORKERS_MAX ;
# else
bounds . upperBound = 0 ;
# endif
return bounds ;
case ZSTD_c_jobSize :
bounds . lowerBound = 0 ;
# ifdef ZSTD_MULTITHREAD
bounds . upperBound = ZSTDMT_JOBSIZE_MAX ;
# else
bounds . upperBound = 0 ;
# endif
return bounds ;
case ZSTD_c_overlapLog :
# ifdef ZSTD_MULTITHREAD
bounds . lowerBound = ZSTD_OVERLAPLOG_MIN ;
bounds . upperBound = ZSTD_OVERLAPLOG_MAX ;
# else
bounds . lowerBound = 0 ;
bounds . upperBound = 0 ;
# endif
return bounds ;
2020-12-18 14:46:27 +00:00
case ZSTD_c_enableDedicatedDictSearch :
bounds . lowerBound = 0 ;
bounds . upperBound = 1 ;
return bounds ;
2020-02-08 13:56:59 +00:00
case ZSTD_c_enableLongDistanceMatching :
2023-02-10 19:25:40 +00:00
bounds . lowerBound = ( int ) ZSTD_ps_auto ;
bounds . upperBound = ( int ) ZSTD_ps_disable ;
2020-02-08 13:56:59 +00:00
return bounds ;
case ZSTD_c_ldmHashLog :
bounds . lowerBound = ZSTD_LDM_HASHLOG_MIN ;
bounds . upperBound = ZSTD_LDM_HASHLOG_MAX ;
return bounds ;
case ZSTD_c_ldmMinMatch :
bounds . lowerBound = ZSTD_LDM_MINMATCH_MIN ;
bounds . upperBound = ZSTD_LDM_MINMATCH_MAX ;
return bounds ;
case ZSTD_c_ldmBucketSizeLog :
bounds . lowerBound = ZSTD_LDM_BUCKETSIZELOG_MIN ;
bounds . upperBound = ZSTD_LDM_BUCKETSIZELOG_MAX ;
return bounds ;
case ZSTD_c_ldmHashRateLog :
bounds . lowerBound = ZSTD_LDM_HASHRATELOG_MIN ;
bounds . upperBound = ZSTD_LDM_HASHRATELOG_MAX ;
return bounds ;
/* experimental parameters */
case ZSTD_c_rsyncable :
bounds . lowerBound = 0 ;
bounds . upperBound = 1 ;
return bounds ;
case ZSTD_c_forceMaxWindow :
bounds . lowerBound = 0 ;
bounds . upperBound = 1 ;
return bounds ;
case ZSTD_c_format :
ZSTD_STATIC_ASSERT ( ZSTD_f_zstd1 < ZSTD_f_zstd1_magicless ) ;
bounds . lowerBound = ZSTD_f_zstd1 ;
bounds . upperBound = ZSTD_f_zstd1_magicless ; /* note : how to ensure at compile time that this is the highest value enum ? */
return bounds ;
case ZSTD_c_forceAttachDict :
2020-12-18 14:46:27 +00:00
ZSTD_STATIC_ASSERT ( ZSTD_dictDefaultAttach < ZSTD_dictForceLoad ) ;
2020-02-08 13:56:59 +00:00
bounds . lowerBound = ZSTD_dictDefaultAttach ;
bounds . upperBound = ZSTD_dictForceLoad ; /* note : how to ensure at compile time that this is the highest value enum ? */
return bounds ;
case ZSTD_c_literalCompressionMode :
2021-12-21 11:21:29 +00:00
ZSTD_STATIC_ASSERT ( ZSTD_ps_auto < ZSTD_ps_enable & & ZSTD_ps_enable < ZSTD_ps_disable ) ;
bounds . lowerBound = ( int ) ZSTD_ps_auto ;
bounds . upperBound = ( int ) ZSTD_ps_disable ;
2020-02-08 13:56:59 +00:00
return bounds ;
case ZSTD_c_targetCBlockSize :
bounds . lowerBound = ZSTD_TARGETCBLOCKSIZE_MIN ;
bounds . upperBound = ZSTD_TARGETCBLOCKSIZE_MAX ;
return bounds ;
case ZSTD_c_srcSizeHint :
bounds . lowerBound = ZSTD_SRCSIZEHINT_MIN ;
bounds . upperBound = ZSTD_SRCSIZEHINT_MAX ;
return bounds ;
2020-12-18 14:46:27 +00:00
case ZSTD_c_stableInBuffer :
case ZSTD_c_stableOutBuffer :
bounds . lowerBound = ( int ) ZSTD_bm_buffered ;
bounds . upperBound = ( int ) ZSTD_bm_stable ;
return bounds ;
case ZSTD_c_blockDelimiters :
bounds . lowerBound = ( int ) ZSTD_sf_noBlockDelimiters ;
bounds . upperBound = ( int ) ZSTD_sf_explicitBlockDelimiters ;
return bounds ;
case ZSTD_c_validateSequences :
bounds . lowerBound = 0 ;
bounds . upperBound = 1 ;
return bounds ;
2021-12-21 11:21:29 +00:00
case ZSTD_c_useBlockSplitter :
bounds . lowerBound = ( int ) ZSTD_ps_auto ;
bounds . upperBound = ( int ) ZSTD_ps_disable ;
2021-05-14 22:48:15 +00:00
return bounds ;
case ZSTD_c_useRowMatchFinder :
2021-12-21 11:21:29 +00:00
bounds . lowerBound = ( int ) ZSTD_ps_auto ;
bounds . upperBound = ( int ) ZSTD_ps_disable ;
2021-05-14 22:48:15 +00:00
return bounds ;
case ZSTD_c_deterministicRefPrefix :
bounds . lowerBound = 0 ;
bounds . upperBound = 1 ;
return bounds ;
2023-02-10 19:25:40 +00:00
case ZSTD_c_prefetchCDictTables :
bounds . lowerBound = ( int ) ZSTD_ps_auto ;
bounds . upperBound = ( int ) ZSTD_ps_disable ;
return bounds ;
case ZSTD_c_enableSeqProducerFallback :
bounds . lowerBound = 0 ;
bounds . upperBound = 1 ;
return bounds ;
case ZSTD_c_maxBlockSize :
bounds . lowerBound = ZSTD_BLOCKSIZE_MAX_MIN ;
bounds . upperBound = ZSTD_BLOCKSIZE_MAX ;
return bounds ;
case ZSTD_c_searchForExternalRepcodes :
bounds . lowerBound = ( int ) ZSTD_ps_auto ;
bounds . upperBound = ( int ) ZSTD_ps_disable ;
return bounds ;
2020-02-08 13:56:59 +00:00
default :
2020-06-04 19:03:27 +00:00
bounds . error = ERROR ( parameter_unsupported ) ;
return bounds ;
2020-02-08 13:56:59 +00:00
}
}
/* ZSTD_cParam_clampBounds:
* Clamps the value into the bounded range .
*/
static size_t ZSTD_cParam_clampBounds ( ZSTD_cParameter cParam , int * value )
{
ZSTD_bounds const bounds = ZSTD_cParam_getBounds ( cParam ) ;
if ( ZSTD_isError ( bounds . error ) ) return bounds . error ;
if ( * value < bounds . lowerBound ) * value = bounds . lowerBound ;
if ( * value > bounds . upperBound ) * value = bounds . upperBound ;
return 0 ;
}
# define BOUNDCHECK(cParam, val) { \
RETURN_ERROR_IF ( ! ZSTD_cParam_withinBounds ( cParam , val ) , \
2020-06-04 19:03:27 +00:00
parameter_outOfBound , " Param out of bounds " ) ; \
2020-02-08 13:56:59 +00:00
}
static int ZSTD_isUpdateAuthorized ( ZSTD_cParameter param )
{
switch ( param )
{
case ZSTD_c_compressionLevel :
case ZSTD_c_hashLog :
case ZSTD_c_chainLog :
case ZSTD_c_searchLog :
case ZSTD_c_minMatch :
case ZSTD_c_targetLength :
case ZSTD_c_strategy :
return 1 ;
case ZSTD_c_format :
case ZSTD_c_windowLog :
case ZSTD_c_contentSizeFlag :
case ZSTD_c_checksumFlag :
case ZSTD_c_dictIDFlag :
case ZSTD_c_forceMaxWindow :
case ZSTD_c_nbWorkers :
case ZSTD_c_jobSize :
case ZSTD_c_overlapLog :
case ZSTD_c_rsyncable :
2020-12-18 14:46:27 +00:00
case ZSTD_c_enableDedicatedDictSearch :
2020-02-08 13:56:59 +00:00
case ZSTD_c_enableLongDistanceMatching :
case ZSTD_c_ldmHashLog :
case ZSTD_c_ldmMinMatch :
case ZSTD_c_ldmBucketSizeLog :
case ZSTD_c_ldmHashRateLog :
case ZSTD_c_forceAttachDict :
case ZSTD_c_literalCompressionMode :
case ZSTD_c_targetCBlockSize :
case ZSTD_c_srcSizeHint :
2020-12-18 14:46:27 +00:00
case ZSTD_c_stableInBuffer :
case ZSTD_c_stableOutBuffer :
case ZSTD_c_blockDelimiters :
case ZSTD_c_validateSequences :
2021-12-21 11:21:29 +00:00
case ZSTD_c_useBlockSplitter :
2021-05-14 22:48:15 +00:00
case ZSTD_c_useRowMatchFinder :
case ZSTD_c_deterministicRefPrefix :
2023-02-10 19:25:40 +00:00
case ZSTD_c_prefetchCDictTables :
case ZSTD_c_enableSeqProducerFallback :
case ZSTD_c_maxBlockSize :
case ZSTD_c_searchForExternalRepcodes :
2020-02-08 13:56:59 +00:00
default :
return 0 ;
}
}
size_t ZSTD_CCtx_setParameter ( ZSTD_CCtx * cctx , ZSTD_cParameter param , int value )
{
DEBUGLOG ( 4 , " ZSTD_CCtx_setParameter (%i, %i) " , ( int ) param , value ) ;
if ( cctx - > streamStage ! = zcss_init ) {
if ( ZSTD_isUpdateAuthorized ( param ) ) {
cctx - > cParamsChanged = 1 ;
} else {
2023-02-10 19:25:40 +00:00
RETURN_ERROR ( stage_wrong , " can only set params in cctx init stage " ) ;
2020-02-08 13:56:59 +00:00
} }
switch ( param )
{
case ZSTD_c_nbWorkers :
RETURN_ERROR_IF ( ( value ! = 0 ) & & cctx - > staticSize , parameter_unsupported ,
" MT not compatible with static alloc " ) ;
break ;
case ZSTD_c_compressionLevel :
case ZSTD_c_windowLog :
case ZSTD_c_hashLog :
case ZSTD_c_chainLog :
case ZSTD_c_searchLog :
case ZSTD_c_minMatch :
case ZSTD_c_targetLength :
case ZSTD_c_strategy :
case ZSTD_c_ldmHashRateLog :
case ZSTD_c_format :
case ZSTD_c_contentSizeFlag :
case ZSTD_c_checksumFlag :
case ZSTD_c_dictIDFlag :
case ZSTD_c_forceMaxWindow :
case ZSTD_c_forceAttachDict :
case ZSTD_c_literalCompressionMode :
case ZSTD_c_jobSize :
case ZSTD_c_overlapLog :
case ZSTD_c_rsyncable :
2020-12-18 14:46:27 +00:00
case ZSTD_c_enableDedicatedDictSearch :
2020-02-08 13:56:59 +00:00
case ZSTD_c_enableLongDistanceMatching :
case ZSTD_c_ldmHashLog :
case ZSTD_c_ldmMinMatch :
case ZSTD_c_ldmBucketSizeLog :
case ZSTD_c_targetCBlockSize :
case ZSTD_c_srcSizeHint :
2020-12-18 14:46:27 +00:00
case ZSTD_c_stableInBuffer :
case ZSTD_c_stableOutBuffer :
case ZSTD_c_blockDelimiters :
case ZSTD_c_validateSequences :
2021-12-21 11:21:29 +00:00
case ZSTD_c_useBlockSplitter :
2021-05-14 22:48:15 +00:00
case ZSTD_c_useRowMatchFinder :
case ZSTD_c_deterministicRefPrefix :
2023-02-10 19:25:40 +00:00
case ZSTD_c_prefetchCDictTables :
case ZSTD_c_enableSeqProducerFallback :
case ZSTD_c_maxBlockSize :
case ZSTD_c_searchForExternalRepcodes :
2020-02-08 13:56:59 +00:00
break ;
2020-06-04 19:03:27 +00:00
default : RETURN_ERROR ( parameter_unsupported , " unknown parameter " ) ;
2020-02-08 13:56:59 +00:00
}
return ZSTD_CCtxParams_setParameter ( & cctx - > requestedParams , param , value ) ;
}
size_t ZSTD_CCtxParams_setParameter ( ZSTD_CCtx_params * CCtxParams ,
ZSTD_cParameter param , int value )
{
DEBUGLOG ( 4 , " ZSTD_CCtxParams_setParameter (%i, %i) " , ( int ) param , value ) ;
switch ( param )
{
case ZSTD_c_format :
BOUNDCHECK ( ZSTD_c_format , value ) ;
CCtxParams - > format = ( ZSTD_format_e ) value ;
return ( size_t ) CCtxParams - > format ;
case ZSTD_c_compressionLevel : {
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_cParam_clampBounds ( param , & value ) , " " ) ;
2020-12-18 14:46:27 +00:00
if ( value = = 0 )
CCtxParams - > compressionLevel = ZSTD_CLEVEL_DEFAULT ; /* 0 == default */
else
2020-02-08 13:56:59 +00:00
CCtxParams - > compressionLevel = value ;
if ( CCtxParams - > compressionLevel > = 0 ) return ( size_t ) CCtxParams - > compressionLevel ;
return 0 ; /* return type (size_t) cannot represent negative values */
}
case ZSTD_c_windowLog :
if ( value ! = 0 ) /* 0 => use default */
BOUNDCHECK ( ZSTD_c_windowLog , value ) ;
CCtxParams - > cParams . windowLog = ( U32 ) value ;
return CCtxParams - > cParams . windowLog ;
case ZSTD_c_hashLog :
if ( value ! = 0 ) /* 0 => use default */
BOUNDCHECK ( ZSTD_c_hashLog , value ) ;
CCtxParams - > cParams . hashLog = ( U32 ) value ;
return CCtxParams - > cParams . hashLog ;
case ZSTD_c_chainLog :
if ( value ! = 0 ) /* 0 => use default */
BOUNDCHECK ( ZSTD_c_chainLog , value ) ;
CCtxParams - > cParams . chainLog = ( U32 ) value ;
return CCtxParams - > cParams . chainLog ;
case ZSTD_c_searchLog :
if ( value ! = 0 ) /* 0 => use default */
BOUNDCHECK ( ZSTD_c_searchLog , value ) ;
CCtxParams - > cParams . searchLog = ( U32 ) value ;
return ( size_t ) value ;
case ZSTD_c_minMatch :
if ( value ! = 0 ) /* 0 => use default */
BOUNDCHECK ( ZSTD_c_minMatch , value ) ;
2023-02-10 19:25:40 +00:00
CCtxParams - > cParams . minMatch = ( U32 ) value ;
2020-02-08 13:56:59 +00:00
return CCtxParams - > cParams . minMatch ;
case ZSTD_c_targetLength :
BOUNDCHECK ( ZSTD_c_targetLength , value ) ;
2023-02-10 19:25:40 +00:00
CCtxParams - > cParams . targetLength = ( U32 ) value ;
2020-02-08 13:56:59 +00:00
return CCtxParams - > cParams . targetLength ;
case ZSTD_c_strategy :
if ( value ! = 0 ) /* 0 => use default */
BOUNDCHECK ( ZSTD_c_strategy , value ) ;
CCtxParams - > cParams . strategy = ( ZSTD_strategy ) value ;
return ( size_t ) CCtxParams - > cParams . strategy ;
case ZSTD_c_contentSizeFlag :
/* Content size written in frame header _when known_ (default:1) */
DEBUGLOG ( 4 , " set content size flag = %u " , ( value ! = 0 ) ) ;
CCtxParams - > fParams . contentSizeFlag = value ! = 0 ;
2023-02-10 19:25:40 +00:00
return ( size_t ) CCtxParams - > fParams . contentSizeFlag ;
2020-02-08 13:56:59 +00:00
case ZSTD_c_checksumFlag :
/* A 32-bits content checksum will be calculated and written at end of frame (default:0) */
CCtxParams - > fParams . checksumFlag = value ! = 0 ;
2023-02-10 19:25:40 +00:00
return ( size_t ) CCtxParams - > fParams . checksumFlag ;
2020-02-08 13:56:59 +00:00
case ZSTD_c_dictIDFlag : /* When applicable, dictionary's dictID is provided in frame header (default:1) */
DEBUGLOG ( 4 , " set dictIDFlag = %u " , ( value ! = 0 ) ) ;
CCtxParams - > fParams . noDictIDFlag = ! value ;
return ! CCtxParams - > fParams . noDictIDFlag ;
case ZSTD_c_forceMaxWindow :
CCtxParams - > forceWindow = ( value ! = 0 ) ;
2023-02-10 19:25:40 +00:00
return ( size_t ) CCtxParams - > forceWindow ;
2020-02-08 13:56:59 +00:00
case ZSTD_c_forceAttachDict : {
const ZSTD_dictAttachPref_e pref = ( ZSTD_dictAttachPref_e ) value ;
2023-02-10 19:25:40 +00:00
BOUNDCHECK ( ZSTD_c_forceAttachDict , ( int ) pref ) ;
2020-02-08 13:56:59 +00:00
CCtxParams - > attachDictPref = pref ;
return CCtxParams - > attachDictPref ;
}
case ZSTD_c_literalCompressionMode : {
2021-12-21 11:21:29 +00:00
const ZSTD_paramSwitch_e lcm = ( ZSTD_paramSwitch_e ) value ;
2023-02-10 19:25:40 +00:00
BOUNDCHECK ( ZSTD_c_literalCompressionMode , ( int ) lcm ) ;
2020-02-08 13:56:59 +00:00
CCtxParams - > literalCompressionMode = lcm ;
return CCtxParams - > literalCompressionMode ;
}
case ZSTD_c_nbWorkers :
# ifndef ZSTD_MULTITHREAD
RETURN_ERROR_IF ( value ! = 0 , parameter_unsupported , " not compiled with multithreading " ) ;
return 0 ;
# else
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_cParam_clampBounds ( param , & value ) , " " ) ;
2020-02-08 13:56:59 +00:00
CCtxParams - > nbWorkers = value ;
return CCtxParams - > nbWorkers ;
# endif
case ZSTD_c_jobSize :
# ifndef ZSTD_MULTITHREAD
RETURN_ERROR_IF ( value ! = 0 , parameter_unsupported , " not compiled with multithreading " ) ;
return 0 ;
# else
/* Adjust to the minimum non-default value. */
if ( value ! = 0 & & value < ZSTDMT_JOBSIZE_MIN )
value = ZSTDMT_JOBSIZE_MIN ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_cParam_clampBounds ( param , & value ) , " " ) ;
2020-02-08 13:56:59 +00:00
assert ( value > = 0 ) ;
CCtxParams - > jobSize = value ;
return CCtxParams - > jobSize ;
# endif
case ZSTD_c_overlapLog :
# ifndef ZSTD_MULTITHREAD
RETURN_ERROR_IF ( value ! = 0 , parameter_unsupported , " not compiled with multithreading " ) ;
return 0 ;
# else
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_cParam_clampBounds ( ZSTD_c_overlapLog , & value ) , " " ) ;
2020-02-08 13:56:59 +00:00
CCtxParams - > overlapLog = value ;
return CCtxParams - > overlapLog ;
# endif
case ZSTD_c_rsyncable :
# ifndef ZSTD_MULTITHREAD
RETURN_ERROR_IF ( value ! = 0 , parameter_unsupported , " not compiled with multithreading " ) ;
return 0 ;
# else
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_cParam_clampBounds ( ZSTD_c_overlapLog , & value ) , " " ) ;
2020-02-08 13:56:59 +00:00
CCtxParams - > rsyncable = value ;
return CCtxParams - > rsyncable ;
# endif
2020-12-18 14:46:27 +00:00
case ZSTD_c_enableDedicatedDictSearch :
CCtxParams - > enableDedicatedDictSearch = ( value ! = 0 ) ;
2023-02-10 19:25:40 +00:00
return ( size_t ) CCtxParams - > enableDedicatedDictSearch ;
2020-12-18 14:46:27 +00:00
2020-02-08 13:56:59 +00:00
case ZSTD_c_enableLongDistanceMatching :
2023-02-10 19:25:40 +00:00
BOUNDCHECK ( ZSTD_c_enableLongDistanceMatching , value ) ;
2021-12-21 11:21:29 +00:00
CCtxParams - > ldmParams . enableLdm = ( ZSTD_paramSwitch_e ) value ;
2020-02-08 13:56:59 +00:00
return CCtxParams - > ldmParams . enableLdm ;
case ZSTD_c_ldmHashLog :
if ( value ! = 0 ) /* 0 ==> auto */
BOUNDCHECK ( ZSTD_c_ldmHashLog , value ) ;
2023-02-10 19:25:40 +00:00
CCtxParams - > ldmParams . hashLog = ( U32 ) value ;
2020-02-08 13:56:59 +00:00
return CCtxParams - > ldmParams . hashLog ;
case ZSTD_c_ldmMinMatch :
if ( value ! = 0 ) /* 0 ==> default */
BOUNDCHECK ( ZSTD_c_ldmMinMatch , value ) ;
2023-02-10 19:25:40 +00:00
CCtxParams - > ldmParams . minMatchLength = ( U32 ) value ;
2020-02-08 13:56:59 +00:00
return CCtxParams - > ldmParams . minMatchLength ;
case ZSTD_c_ldmBucketSizeLog :
if ( value ! = 0 ) /* 0 ==> default */
BOUNDCHECK ( ZSTD_c_ldmBucketSizeLog , value ) ;
2023-02-10 19:25:40 +00:00
CCtxParams - > ldmParams . bucketSizeLog = ( U32 ) value ;
2020-02-08 13:56:59 +00:00
return CCtxParams - > ldmParams . bucketSizeLog ;
case ZSTD_c_ldmHashRateLog :
2021-12-21 11:21:29 +00:00
if ( value ! = 0 ) /* 0 ==> default */
BOUNDCHECK ( ZSTD_c_ldmHashRateLog , value ) ;
2023-02-10 19:25:40 +00:00
CCtxParams - > ldmParams . hashRateLog = ( U32 ) value ;
2020-02-08 13:56:59 +00:00
return CCtxParams - > ldmParams . hashRateLog ;
case ZSTD_c_targetCBlockSize :
if ( value ! = 0 ) /* 0 ==> default */
BOUNDCHECK ( ZSTD_c_targetCBlockSize , value ) ;
2023-02-10 19:25:40 +00:00
CCtxParams - > targetCBlockSize = ( U32 ) value ;
2020-02-08 13:56:59 +00:00
return CCtxParams - > targetCBlockSize ;
case ZSTD_c_srcSizeHint :
if ( value ! = 0 ) /* 0 ==> default */
BOUNDCHECK ( ZSTD_c_srcSizeHint , value ) ;
CCtxParams - > srcSizeHint = value ;
2023-02-10 19:25:40 +00:00
return ( size_t ) CCtxParams - > srcSizeHint ;
2020-02-08 13:56:59 +00:00
2020-12-18 14:46:27 +00:00
case ZSTD_c_stableInBuffer :
BOUNDCHECK ( ZSTD_c_stableInBuffer , value ) ;
CCtxParams - > inBufferMode = ( ZSTD_bufferMode_e ) value ;
return CCtxParams - > inBufferMode ;
case ZSTD_c_stableOutBuffer :
BOUNDCHECK ( ZSTD_c_stableOutBuffer , value ) ;
CCtxParams - > outBufferMode = ( ZSTD_bufferMode_e ) value ;
return CCtxParams - > outBufferMode ;
case ZSTD_c_blockDelimiters :
BOUNDCHECK ( ZSTD_c_blockDelimiters , value ) ;
CCtxParams - > blockDelimiters = ( ZSTD_sequenceFormat_e ) value ;
return CCtxParams - > blockDelimiters ;
case ZSTD_c_validateSequences :
BOUNDCHECK ( ZSTD_c_validateSequences , value ) ;
CCtxParams - > validateSequences = value ;
return CCtxParams - > validateSequences ;
2021-12-21 11:21:29 +00:00
case ZSTD_c_useBlockSplitter :
BOUNDCHECK ( ZSTD_c_useBlockSplitter , value ) ;
CCtxParams - > useBlockSplitter = ( ZSTD_paramSwitch_e ) value ;
return CCtxParams - > useBlockSplitter ;
2021-05-14 22:48:15 +00:00
case ZSTD_c_useRowMatchFinder :
BOUNDCHECK ( ZSTD_c_useRowMatchFinder , value ) ;
2021-12-21 11:21:29 +00:00
CCtxParams - > useRowMatchFinder = ( ZSTD_paramSwitch_e ) value ;
2021-05-14 22:48:15 +00:00
return CCtxParams - > useRowMatchFinder ;
case ZSTD_c_deterministicRefPrefix :
BOUNDCHECK ( ZSTD_c_deterministicRefPrefix , value ) ;
CCtxParams - > deterministicRefPrefix = ! ! value ;
return CCtxParams - > deterministicRefPrefix ;
2023-02-10 19:25:40 +00:00
case ZSTD_c_prefetchCDictTables :
BOUNDCHECK ( ZSTD_c_prefetchCDictTables , value ) ;
CCtxParams - > prefetchCDictTables = ( ZSTD_paramSwitch_e ) value ;
return CCtxParams - > prefetchCDictTables ;
case ZSTD_c_enableSeqProducerFallback :
BOUNDCHECK ( ZSTD_c_enableSeqProducerFallback , value ) ;
CCtxParams - > enableMatchFinderFallback = value ;
return CCtxParams - > enableMatchFinderFallback ;
case ZSTD_c_maxBlockSize :
if ( value ! = 0 ) /* 0 ==> default */
BOUNDCHECK ( ZSTD_c_maxBlockSize , value ) ;
CCtxParams - > maxBlockSize = value ;
return CCtxParams - > maxBlockSize ;
case ZSTD_c_searchForExternalRepcodes :
BOUNDCHECK ( ZSTD_c_searchForExternalRepcodes , value ) ;
CCtxParams - > searchForExternalRepcodes = ( ZSTD_paramSwitch_e ) value ;
return CCtxParams - > searchForExternalRepcodes ;
2020-02-08 13:56:59 +00:00
default : RETURN_ERROR ( parameter_unsupported , " unknown parameter " ) ;
}
}
2021-05-14 22:48:15 +00:00
size_t ZSTD_CCtx_getParameter ( ZSTD_CCtx const * cctx , ZSTD_cParameter param , int * value )
2020-02-08 13:56:59 +00:00
{
return ZSTD_CCtxParams_getParameter ( & cctx - > requestedParams , param , value ) ;
}
size_t ZSTD_CCtxParams_getParameter (
2021-05-14 22:48:15 +00:00
ZSTD_CCtx_params const * CCtxParams , ZSTD_cParameter param , int * value )
2020-02-08 13:56:59 +00:00
{
switch ( param )
{
case ZSTD_c_format :
* value = CCtxParams - > format ;
break ;
case ZSTD_c_compressionLevel :
* value = CCtxParams - > compressionLevel ;
break ;
case ZSTD_c_windowLog :
* value = ( int ) CCtxParams - > cParams . windowLog ;
break ;
case ZSTD_c_hashLog :
* value = ( int ) CCtxParams - > cParams . hashLog ;
break ;
case ZSTD_c_chainLog :
* value = ( int ) CCtxParams - > cParams . chainLog ;
break ;
case ZSTD_c_searchLog :
* value = CCtxParams - > cParams . searchLog ;
break ;
case ZSTD_c_minMatch :
* value = CCtxParams - > cParams . minMatch ;
break ;
case ZSTD_c_targetLength :
* value = CCtxParams - > cParams . targetLength ;
break ;
case ZSTD_c_strategy :
* value = ( unsigned ) CCtxParams - > cParams . strategy ;
break ;
case ZSTD_c_contentSizeFlag :
* value = CCtxParams - > fParams . contentSizeFlag ;
break ;
case ZSTD_c_checksumFlag :
* value = CCtxParams - > fParams . checksumFlag ;
break ;
case ZSTD_c_dictIDFlag :
* value = ! CCtxParams - > fParams . noDictIDFlag ;
break ;
case ZSTD_c_forceMaxWindow :
* value = CCtxParams - > forceWindow ;
break ;
case ZSTD_c_forceAttachDict :
* value = CCtxParams - > attachDictPref ;
break ;
case ZSTD_c_literalCompressionMode :
* value = CCtxParams - > literalCompressionMode ;
break ;
case ZSTD_c_nbWorkers :
# ifndef ZSTD_MULTITHREAD
assert ( CCtxParams - > nbWorkers = = 0 ) ;
# endif
* value = CCtxParams - > nbWorkers ;
break ;
case ZSTD_c_jobSize :
# ifndef ZSTD_MULTITHREAD
RETURN_ERROR ( parameter_unsupported , " not compiled with multithreading " ) ;
# else
assert ( CCtxParams - > jobSize < = INT_MAX ) ;
* value = ( int ) CCtxParams - > jobSize ;
break ;
# endif
case ZSTD_c_overlapLog :
# ifndef ZSTD_MULTITHREAD
RETURN_ERROR ( parameter_unsupported , " not compiled with multithreading " ) ;
# else
* value = CCtxParams - > overlapLog ;
break ;
# endif
case ZSTD_c_rsyncable :
# ifndef ZSTD_MULTITHREAD
RETURN_ERROR ( parameter_unsupported , " not compiled with multithreading " ) ;
# else
* value = CCtxParams - > rsyncable ;
break ;
# endif
2020-12-18 14:46:27 +00:00
case ZSTD_c_enableDedicatedDictSearch :
* value = CCtxParams - > enableDedicatedDictSearch ;
break ;
2020-02-08 13:56:59 +00:00
case ZSTD_c_enableLongDistanceMatching :
* value = CCtxParams - > ldmParams . enableLdm ;
break ;
case ZSTD_c_ldmHashLog :
* value = CCtxParams - > ldmParams . hashLog ;
break ;
case ZSTD_c_ldmMinMatch :
* value = CCtxParams - > ldmParams . minMatchLength ;
break ;
case ZSTD_c_ldmBucketSizeLog :
* value = CCtxParams - > ldmParams . bucketSizeLog ;
break ;
case ZSTD_c_ldmHashRateLog :
* value = CCtxParams - > ldmParams . hashRateLog ;
break ;
case ZSTD_c_targetCBlockSize :
* value = ( int ) CCtxParams - > targetCBlockSize ;
break ;
case ZSTD_c_srcSizeHint :
* value = ( int ) CCtxParams - > srcSizeHint ;
break ;
2020-12-18 14:46:27 +00:00
case ZSTD_c_stableInBuffer :
* value = ( int ) CCtxParams - > inBufferMode ;
break ;
case ZSTD_c_stableOutBuffer :
* value = ( int ) CCtxParams - > outBufferMode ;
break ;
case ZSTD_c_blockDelimiters :
* value = ( int ) CCtxParams - > blockDelimiters ;
break ;
case ZSTD_c_validateSequences :
* value = ( int ) CCtxParams - > validateSequences ;
break ;
2021-12-21 11:21:29 +00:00
case ZSTD_c_useBlockSplitter :
* value = ( int ) CCtxParams - > useBlockSplitter ;
2021-05-14 22:48:15 +00:00
break ;
case ZSTD_c_useRowMatchFinder :
* value = ( int ) CCtxParams - > useRowMatchFinder ;
break ;
case ZSTD_c_deterministicRefPrefix :
* value = ( int ) CCtxParams - > deterministicRefPrefix ;
break ;
2023-02-10 19:25:40 +00:00
case ZSTD_c_prefetchCDictTables :
* value = ( int ) CCtxParams - > prefetchCDictTables ;
break ;
case ZSTD_c_enableSeqProducerFallback :
* value = CCtxParams - > enableMatchFinderFallback ;
break ;
case ZSTD_c_maxBlockSize :
* value = ( int ) CCtxParams - > maxBlockSize ;
break ;
case ZSTD_c_searchForExternalRepcodes :
* value = ( int ) CCtxParams - > searchForExternalRepcodes ;
break ;
2020-02-08 13:56:59 +00:00
default : RETURN_ERROR ( parameter_unsupported , " unknown parameter " ) ;
}
return 0 ;
}
/** ZSTD_CCtx_setParametersUsingCCtxParams() :
* just applies ` params ` into ` cctx `
* no action is performed , parameters are merely stored .
* If ZSTDMT is enabled , parameters are pushed to cctx - > mtctx .
* This is possible even if a compression is ongoing .
* In which case , new parameters will be applied on the fly , starting with next compression job .
*/
size_t ZSTD_CCtx_setParametersUsingCCtxParams (
ZSTD_CCtx * cctx , const ZSTD_CCtx_params * params )
{
DEBUGLOG ( 4 , " ZSTD_CCtx_setParametersUsingCCtxParams " ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( cctx - > streamStage ! = zcss_init , stage_wrong ,
" The context is in the wrong stage! " ) ;
RETURN_ERROR_IF ( cctx - > cdict , stage_wrong ,
" Can't override parameters with cdict attached (some must "
" be inherited from the cdict). " ) ;
2020-02-08 13:56:59 +00:00
cctx - > requestedParams = * params ;
return 0 ;
}
2023-02-10 19:25:40 +00:00
size_t ZSTD_CCtx_setCParams ( ZSTD_CCtx * cctx , ZSTD_compressionParameters cparams )
{
DEBUGLOG ( 4 , " ZSTD_CCtx_setCParams " ) ;
assert ( cctx ! = NULL ) ;
if ( cctx - > streamStage ! = zcss_init ) {
/* All parameters in @cparams are allowed to be updated during MT compression.
* This must be signaled , so that MT compression picks up the changes */
cctx - > cParamsChanged = 1 ;
}
/* only update if parameters are valid */
FORWARD_IF_ERROR ( ZSTD_checkCParams ( cparams ) , " " ) ;
cctx - > requestedParams . cParams = cparams ;
return 0 ;
}
2021-12-21 11:21:29 +00:00
size_t ZSTD_CCtx_setPledgedSrcSize ( ZSTD_CCtx * cctx , unsigned long long pledgedSrcSize )
2020-02-08 13:56:59 +00:00
{
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 4 , " ZSTD_CCtx_setPledgedSrcSize to %llu bytes " , pledgedSrcSize ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( cctx - > streamStage ! = zcss_init , stage_wrong ,
" Can't set pledgedSrcSize when not in init stage. " ) ;
2020-02-08 13:56:59 +00:00
cctx - > pledgedSrcSizePlusOne = pledgedSrcSize + 1 ;
return 0 ;
}
2020-12-18 14:46:27 +00:00
static ZSTD_compressionParameters ZSTD_dedicatedDictSearch_getCParams (
int const compressionLevel ,
size_t const dictSize ) ;
static int ZSTD_dedicatedDictSearch_isSupported (
const ZSTD_compressionParameters * cParams ) ;
static void ZSTD_dedicatedDictSearch_revertCParams (
ZSTD_compressionParameters * cParams ) ;
2020-02-08 13:56:59 +00:00
/**
* Initializes the local dict using the requested parameters .
* NOTE : This does not use the pledged src size , because it may be used for more
* than one compression .
*/
static size_t ZSTD_initLocalDict ( ZSTD_CCtx * cctx )
{
ZSTD_localDict * const dl = & cctx - > localDict ;
if ( dl - > dict = = NULL ) {
/* No local dictionary. */
assert ( dl - > dictBuffer = = NULL ) ;
assert ( dl - > cdict = = NULL ) ;
assert ( dl - > dictSize = = 0 ) ;
return 0 ;
}
if ( dl - > cdict ! = NULL ) {
assert ( cctx - > cdict = = dl - > cdict ) ;
/* Local dictionary already initialized. */
return 0 ;
}
assert ( dl - > dictSize > 0 ) ;
assert ( cctx - > cdict = = NULL ) ;
assert ( cctx - > prefixDict . dict = = NULL ) ;
2020-12-18 14:46:27 +00:00
dl - > cdict = ZSTD_createCDict_advanced2 (
2020-02-08 13:56:59 +00:00
dl - > dict ,
dl - > dictSize ,
ZSTD_dlm_byRef ,
dl - > dictContentType ,
2020-12-18 14:46:27 +00:00
& cctx - > requestedParams ,
2020-02-08 13:56:59 +00:00
cctx - > customMem ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( ! dl - > cdict , memory_allocation , " ZSTD_createCDict_advanced failed " ) ;
2020-02-08 13:56:59 +00:00
cctx - > cdict = dl - > cdict ;
return 0 ;
}
size_t ZSTD_CCtx_loadDictionary_advanced (
ZSTD_CCtx * cctx , const void * dict , size_t dictSize ,
ZSTD_dictLoadMethod_e dictLoadMethod , ZSTD_dictContentType_e dictContentType )
{
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( cctx - > streamStage ! = zcss_init , stage_wrong ,
" Can't load a dictionary when ctx is not in init stage. " ) ;
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 4 , " ZSTD_CCtx_loadDictionary_advanced (size: %u) " , ( U32 ) dictSize ) ;
ZSTD_clearAllDicts ( cctx ) ; /* in case one already exists */
if ( dict = = NULL | | dictSize = = 0 ) /* no dictionary mode */
return 0 ;
if ( dictLoadMethod = = ZSTD_dlm_byRef ) {
cctx - > localDict . dict = dict ;
} else {
2020-12-18 14:46:27 +00:00
void * dictBuffer ;
RETURN_ERROR_IF ( cctx - > staticSize , memory_allocation ,
" no malloc for static CCtx " ) ;
dictBuffer = ZSTD_customMalloc ( dictSize , cctx - > customMem ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( ! dictBuffer , memory_allocation , " NULL pointer! " ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( dictBuffer , dict , dictSize ) ;
2020-02-08 13:56:59 +00:00
cctx - > localDict . dictBuffer = dictBuffer ;
cctx - > localDict . dict = dictBuffer ;
}
cctx - > localDict . dictSize = dictSize ;
cctx - > localDict . dictContentType = dictContentType ;
return 0 ;
}
2021-12-21 11:21:29 +00:00
size_t ZSTD_CCtx_loadDictionary_byReference (
2020-02-08 13:56:59 +00:00
ZSTD_CCtx * cctx , const void * dict , size_t dictSize )
{
return ZSTD_CCtx_loadDictionary_advanced (
cctx , dict , dictSize , ZSTD_dlm_byRef , ZSTD_dct_auto ) ;
}
2021-12-21 11:21:29 +00:00
size_t ZSTD_CCtx_loadDictionary ( ZSTD_CCtx * cctx , const void * dict , size_t dictSize )
2020-02-08 13:56:59 +00:00
{
return ZSTD_CCtx_loadDictionary_advanced (
cctx , dict , dictSize , ZSTD_dlm_byCopy , ZSTD_dct_auto ) ;
}
size_t ZSTD_CCtx_refCDict ( ZSTD_CCtx * cctx , const ZSTD_CDict * cdict )
{
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( cctx - > streamStage ! = zcss_init , stage_wrong ,
" Can't ref a dict when ctx not in init stage. " ) ;
2020-02-08 13:56:59 +00:00
/* Free the existing local cdict (if any) to save memory. */
ZSTD_clearAllDicts ( cctx ) ;
cctx - > cdict = cdict ;
return 0 ;
}
2020-12-18 14:46:27 +00:00
size_t ZSTD_CCtx_refThreadPool ( ZSTD_CCtx * cctx , ZSTD_threadPool * pool )
{
RETURN_ERROR_IF ( cctx - > streamStage ! = zcss_init , stage_wrong ,
" Can't ref a pool when ctx not in init stage. " ) ;
cctx - > pool = pool ;
return 0 ;
}
2020-02-08 13:56:59 +00:00
size_t ZSTD_CCtx_refPrefix ( ZSTD_CCtx * cctx , const void * prefix , size_t prefixSize )
{
return ZSTD_CCtx_refPrefix_advanced ( cctx , prefix , prefixSize , ZSTD_dct_rawContent ) ;
}
size_t ZSTD_CCtx_refPrefix_advanced (
ZSTD_CCtx * cctx , const void * prefix , size_t prefixSize , ZSTD_dictContentType_e dictContentType )
{
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( cctx - > streamStage ! = zcss_init , stage_wrong ,
" Can't ref a prefix when ctx not in init stage. " ) ;
2020-02-08 13:56:59 +00:00
ZSTD_clearAllDicts ( cctx ) ;
2020-06-04 19:03:27 +00:00
if ( prefix ! = NULL & & prefixSize > 0 ) {
cctx - > prefixDict . dict = prefix ;
cctx - > prefixDict . dictSize = prefixSize ;
cctx - > prefixDict . dictContentType = dictContentType ;
}
2020-02-08 13:56:59 +00:00
return 0 ;
}
/*! ZSTD_CCtx_reset() :
* Also dumps dictionary */
size_t ZSTD_CCtx_reset ( ZSTD_CCtx * cctx , ZSTD_ResetDirective reset )
{
if ( ( reset = = ZSTD_reset_session_only )
| | ( reset = = ZSTD_reset_session_and_parameters ) ) {
cctx - > streamStage = zcss_init ;
cctx - > pledgedSrcSizePlusOne = 0 ;
}
if ( ( reset = = ZSTD_reset_parameters )
| | ( reset = = ZSTD_reset_session_and_parameters ) ) {
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( cctx - > streamStage ! = zcss_init , stage_wrong ,
" Can't reset parameters only when not in init stage. " ) ;
2020-02-08 13:56:59 +00:00
ZSTD_clearAllDicts ( cctx ) ;
2023-02-10 19:25:40 +00:00
ZSTD_memset ( & cctx - > externalMatchCtx , 0 , sizeof ( cctx - > externalMatchCtx ) ) ;
2020-02-08 13:56:59 +00:00
return ZSTD_CCtxParams_reset ( & cctx - > requestedParams ) ;
}
return 0 ;
}
/** ZSTD_checkCParams() :
control CParam values remain within authorized range .
@ return : 0 , or an error code if one value is beyond authorized range */
size_t ZSTD_checkCParams ( ZSTD_compressionParameters cParams )
{
BOUNDCHECK ( ZSTD_c_windowLog , ( int ) cParams . windowLog ) ;
BOUNDCHECK ( ZSTD_c_chainLog , ( int ) cParams . chainLog ) ;
BOUNDCHECK ( ZSTD_c_hashLog , ( int ) cParams . hashLog ) ;
BOUNDCHECK ( ZSTD_c_searchLog , ( int ) cParams . searchLog ) ;
BOUNDCHECK ( ZSTD_c_minMatch , ( int ) cParams . minMatch ) ;
BOUNDCHECK ( ZSTD_c_targetLength , ( int ) cParams . targetLength ) ;
BOUNDCHECK ( ZSTD_c_strategy , cParams . strategy ) ;
return 0 ;
}
/** ZSTD_clampCParams() :
* make CParam values within valid range .
* @ return : valid CParams */
static ZSTD_compressionParameters
ZSTD_clampCParams ( ZSTD_compressionParameters cParams )
{
# define CLAMP_TYPE(cParam, val, type) { \
ZSTD_bounds const bounds = ZSTD_cParam_getBounds ( cParam ) ; \
if ( ( int ) val < bounds . lowerBound ) val = ( type ) bounds . lowerBound ; \
else if ( ( int ) val > bounds . upperBound ) val = ( type ) bounds . upperBound ; \
}
# define CLAMP(cParam, val) CLAMP_TYPE(cParam, val, unsigned)
CLAMP ( ZSTD_c_windowLog , cParams . windowLog ) ;
CLAMP ( ZSTD_c_chainLog , cParams . chainLog ) ;
CLAMP ( ZSTD_c_hashLog , cParams . hashLog ) ;
CLAMP ( ZSTD_c_searchLog , cParams . searchLog ) ;
CLAMP ( ZSTD_c_minMatch , cParams . minMatch ) ;
CLAMP ( ZSTD_c_targetLength , cParams . targetLength ) ;
CLAMP_TYPE ( ZSTD_c_strategy , cParams . strategy , ZSTD_strategy ) ;
return cParams ;
}
/** ZSTD_cycleLog() :
* condition for correct operation : hashLog > 1 */
2020-06-04 19:03:27 +00:00
U32 ZSTD_cycleLog ( U32 hashLog , ZSTD_strategy strat )
2020-02-08 13:56:59 +00:00
{
U32 const btScale = ( ( U32 ) strat > = ( U32 ) ZSTD_btlazy2 ) ;
return hashLog - btScale ;
}
2020-12-18 14:46:27 +00:00
/** ZSTD_dictAndWindowLog() :
* Returns an adjusted window log that is large enough to fit the source and the dictionary .
* The zstd format says that the entire dictionary is valid if one byte of the dictionary
* is within the window . So the hashLog and chainLog should be large enough to reference both
* the dictionary and the window . So we must use this adjusted dictAndWindowLog when downsizing
* the hashLog and windowLog .
* NOTE : srcSize must not be ZSTD_CONTENTSIZE_UNKNOWN .
*/
static U32 ZSTD_dictAndWindowLog ( U32 windowLog , U64 srcSize , U64 dictSize )
{
const U64 maxWindowSize = 1ULL < < ZSTD_WINDOWLOG_MAX ;
/* No dictionary ==> No change */
if ( dictSize = = 0 ) {
return windowLog ;
}
assert ( windowLog < = ZSTD_WINDOWLOG_MAX ) ;
assert ( srcSize ! = ZSTD_CONTENTSIZE_UNKNOWN ) ; /* Handled in ZSTD_adjustCParams_internal() */
{
U64 const windowSize = 1ULL < < windowLog ;
U64 const dictAndWindowSize = dictSize + windowSize ;
/* If the window size is already large enough to fit both the source and the dictionary
* then just use the window size . Otherwise adjust so that it fits the dictionary and
* the window .
*/
if ( windowSize > = dictSize + srcSize ) {
return windowLog ; /* Window size large enough already */
} else if ( dictAndWindowSize > = maxWindowSize ) {
return ZSTD_WINDOWLOG_MAX ; /* Larger than max window log */
} else {
return ZSTD_highbit32 ( ( U32 ) dictAndWindowSize - 1 ) + 1 ;
}
}
}
2020-02-08 13:56:59 +00:00
/** ZSTD_adjustCParams_internal() :
* optimize ` cPar ` for a specified input ( ` srcSize ` and ` dictSize ` ) .
* mostly downsize to reduce memory consumption and initialization latency .
* ` srcSize ` can be ZSTD_CONTENTSIZE_UNKNOWN when not known .
2020-12-18 14:46:27 +00:00
* ` mode ` is the mode for parameter adjustment . See docs for ` ZSTD_cParamMode_e ` .
2020-02-08 13:56:59 +00:00
* note : ` srcSize = = 0 ` means 0 !
* condition : cPar is presumed validated ( can be checked using ZSTD_checkCParams ( ) ) . */
static ZSTD_compressionParameters
ZSTD_adjustCParams_internal ( ZSTD_compressionParameters cPar ,
unsigned long long srcSize ,
2020-12-18 14:46:27 +00:00
size_t dictSize ,
2023-02-10 19:25:40 +00:00
ZSTD_cParamMode_e mode ,
ZSTD_paramSwitch_e useRowMatchFinder )
2020-02-08 13:56:59 +00:00
{
2020-12-18 14:46:27 +00:00
const U64 minSrcSize = 513 ; /* (1<<9) + 1 */
const U64 maxWindowResize = 1ULL < < ( ZSTD_WINDOWLOG_MAX - 1 ) ;
2020-02-08 13:56:59 +00:00
assert ( ZSTD_checkCParams ( cPar ) = = 0 ) ;
2020-12-18 14:46:27 +00:00
switch ( mode ) {
case ZSTD_cpm_unknown :
2021-05-14 22:48:15 +00:00
case ZSTD_cpm_noAttachDict :
/* If we don't know the source size, don't make any
* assumptions about it . We will already have selected
* smaller parameters if a dictionary is in use .
*/
break ;
2020-12-18 14:46:27 +00:00
case ZSTD_cpm_createCDict :
2021-05-14 22:48:15 +00:00
/* Assume a small source size when creating a dictionary
2021-12-21 11:21:29 +00:00
* with an unknown source size .
2021-05-14 22:48:15 +00:00
*/
if ( dictSize & & srcSize = = ZSTD_CONTENTSIZE_UNKNOWN )
srcSize = minSrcSize ;
2020-12-18 14:46:27 +00:00
break ;
case ZSTD_cpm_attachDict :
2021-05-14 22:48:15 +00:00
/* Dictionary has its own dedicated parameters which have
* already been selected . We are selecting parameters
* for only the source .
*/
2020-12-18 14:46:27 +00:00
dictSize = 0 ;
break ;
default :
assert ( 0 ) ;
break ;
}
2020-02-08 13:56:59 +00:00
/* resize windowLog if input is small enough, to use less memory */
2023-02-10 19:25:40 +00:00
if ( ( srcSize < = maxWindowResize )
& & ( dictSize < = maxWindowResize ) ) {
2020-02-08 13:56:59 +00:00
U32 const tSize = ( U32 ) ( srcSize + dictSize ) ;
static U32 const hashSizeMin = 1 < < ZSTD_HASHLOG_MIN ;
U32 const srcLog = ( tSize < hashSizeMin ) ? ZSTD_HASHLOG_MIN :
ZSTD_highbit32 ( tSize - 1 ) + 1 ;
if ( cPar . windowLog > srcLog ) cPar . windowLog = srcLog ;
}
2021-05-14 22:48:15 +00:00
if ( srcSize ! = ZSTD_CONTENTSIZE_UNKNOWN ) {
U32 const dictAndWindowLog = ZSTD_dictAndWindowLog ( cPar . windowLog , ( U64 ) srcSize , ( U64 ) dictSize ) ;
2020-12-18 14:46:27 +00:00
U32 const cycleLog = ZSTD_cycleLog ( cPar . chainLog , cPar . strategy ) ;
if ( cPar . hashLog > dictAndWindowLog + 1 ) cPar . hashLog = dictAndWindowLog + 1 ;
if ( cycleLog > dictAndWindowLog )
cPar . chainLog - = ( cycleLog - dictAndWindowLog ) ;
2020-02-08 13:56:59 +00:00
}
if ( cPar . windowLog < ZSTD_WINDOWLOG_ABSOLUTEMIN )
cPar . windowLog = ZSTD_WINDOWLOG_ABSOLUTEMIN ; /* minimum wlog required for valid frame header */
2023-02-10 19:25:40 +00:00
/* We can't use more than 32 bits of hash in total, so that means that we require:
* ( hashLog + 8 ) < = 32 & & ( chainLog + 8 ) < = 32
*/
if ( mode = = ZSTD_cpm_createCDict & & ZSTD_CDictIndicesAreTagged ( & cPar ) ) {
U32 const maxShortCacheHashLog = 32 - ZSTD_SHORT_CACHE_TAG_BITS ;
if ( cPar . hashLog > maxShortCacheHashLog ) {
cPar . hashLog = maxShortCacheHashLog ;
}
if ( cPar . chainLog > maxShortCacheHashLog ) {
cPar . chainLog = maxShortCacheHashLog ;
}
}
/* At this point, we aren't 100% sure if we are using the row match finder.
* Unless it is explicitly disabled , conservatively assume that it is enabled .
* In this case it will only be disabled for small sources , so shrinking the
* hash log a little bit shouldn ' t result in any ratio loss .
*/
if ( useRowMatchFinder = = ZSTD_ps_auto )
useRowMatchFinder = ZSTD_ps_enable ;
/* We can't hash more than 32-bits in total. So that means that we require:
* ( hashLog - rowLog + 8 ) < = 32
*/
if ( ZSTD_rowMatchFinderUsed ( cPar . strategy , useRowMatchFinder ) ) {
/* Switch to 32-entry rows if searchLog is 5 (or more) */
U32 const rowLog = BOUNDED ( 4 , cPar . searchLog , 6 ) ;
U32 const maxRowHashLog = 32 - ZSTD_ROW_HASH_TAG_BITS ;
U32 const maxHashLog = maxRowHashLog + rowLog ;
assert ( cPar . hashLog > = rowLog ) ;
if ( cPar . hashLog > maxHashLog ) {
cPar . hashLog = maxHashLog ;
}
}
2020-02-08 13:56:59 +00:00
return cPar ;
}
ZSTD_compressionParameters
ZSTD_adjustCParams ( ZSTD_compressionParameters cPar ,
unsigned long long srcSize ,
size_t dictSize )
{
cPar = ZSTD_clampCParams ( cPar ) ; /* resulting cPar is necessarily valid (all parameters within range) */
if ( srcSize = = 0 ) srcSize = ZSTD_CONTENTSIZE_UNKNOWN ;
2023-02-10 19:25:40 +00:00
return ZSTD_adjustCParams_internal ( cPar , srcSize , dictSize , ZSTD_cpm_unknown , ZSTD_ps_auto ) ;
2020-02-08 13:56:59 +00:00
}
2020-12-18 14:46:27 +00:00
static ZSTD_compressionParameters ZSTD_getCParams_internal ( int compressionLevel , unsigned long long srcSizeHint , size_t dictSize , ZSTD_cParamMode_e mode ) ;
static ZSTD_parameters ZSTD_getParams_internal ( int compressionLevel , unsigned long long srcSizeHint , size_t dictSize , ZSTD_cParamMode_e mode ) ;
static void ZSTD_overrideCParams (
ZSTD_compressionParameters * cParams ,
const ZSTD_compressionParameters * overrides )
{
if ( overrides - > windowLog ) cParams - > windowLog = overrides - > windowLog ;
if ( overrides - > hashLog ) cParams - > hashLog = overrides - > hashLog ;
if ( overrides - > chainLog ) cParams - > chainLog = overrides - > chainLog ;
if ( overrides - > searchLog ) cParams - > searchLog = overrides - > searchLog ;
if ( overrides - > minMatch ) cParams - > minMatch = overrides - > minMatch ;
if ( overrides - > targetLength ) cParams - > targetLength = overrides - > targetLength ;
if ( overrides - > strategy ) cParams - > strategy = overrides - > strategy ;
}
2020-02-08 13:56:59 +00:00
ZSTD_compressionParameters ZSTD_getCParamsFromCCtxParams (
2020-12-18 14:46:27 +00:00
const ZSTD_CCtx_params * CCtxParams , U64 srcSizeHint , size_t dictSize , ZSTD_cParamMode_e mode )
2020-02-08 13:56:59 +00:00
{
ZSTD_compressionParameters cParams ;
if ( srcSizeHint = = ZSTD_CONTENTSIZE_UNKNOWN & & CCtxParams - > srcSizeHint > 0 ) {
srcSizeHint = CCtxParams - > srcSizeHint ;
}
2020-12-18 14:46:27 +00:00
cParams = ZSTD_getCParams_internal ( CCtxParams - > compressionLevel , srcSizeHint , dictSize , mode ) ;
2021-12-21 11:21:29 +00:00
if ( CCtxParams - > ldmParams . enableLdm = = ZSTD_ps_enable ) cParams . windowLog = ZSTD_LDM_DEFAULT_WINDOW_LOG ;
2020-12-18 14:46:27 +00:00
ZSTD_overrideCParams ( & cParams , & CCtxParams - > cParams ) ;
2020-02-08 13:56:59 +00:00
assert ( ! ZSTD_checkCParams ( cParams ) ) ;
/* srcSizeHint == 0 means 0 */
2023-02-10 19:25:40 +00:00
return ZSTD_adjustCParams_internal ( cParams , srcSizeHint , dictSize , mode , CCtxParams - > useRowMatchFinder ) ;
2020-02-08 13:56:59 +00:00
}
static size_t
ZSTD_sizeof_matchState ( const ZSTD_compressionParameters * const cParams ,
2021-12-21 11:21:29 +00:00
const ZSTD_paramSwitch_e useRowMatchFinder ,
2021-05-14 22:48:15 +00:00
const U32 enableDedicatedDictSearch ,
2020-02-08 13:56:59 +00:00
const U32 forCCtx )
{
2021-05-14 22:48:15 +00:00
/* chain table size should be 0 for fast or row-hash strategies */
size_t const chainSize = ZSTD_allocateChainTable ( cParams - > strategy , useRowMatchFinder , enableDedicatedDictSearch & & ! forCCtx )
? ( ( size_t ) 1 < < cParams - > chainLog )
: 0 ;
2020-02-08 13:56:59 +00:00
size_t const hSize = ( ( size_t ) 1 ) < < cParams - > hashLog ;
U32 const hashLog3 = ( forCCtx & & cParams - > minMatch = = 3 ) ? MIN ( ZSTD_HASHLOG3_MAX , cParams - > windowLog ) : 0 ;
size_t const h3Size = hashLog3 ? ( ( size_t ) 1 ) < < hashLog3 : 0 ;
/* We don't use ZSTD_cwksp_alloc_size() here because the tables aren't
* surrounded by redzones in ASAN . */
size_t const tableSpace = chainSize * sizeof ( U32 )
+ hSize * sizeof ( U32 )
+ h3Size * sizeof ( U32 ) ;
size_t const optPotentialSpace =
2021-05-14 22:48:15 +00:00
ZSTD_cwksp_aligned_alloc_size ( ( MaxML + 1 ) * sizeof ( U32 ) )
+ ZSTD_cwksp_aligned_alloc_size ( ( MaxLL + 1 ) * sizeof ( U32 ) )
+ ZSTD_cwksp_aligned_alloc_size ( ( MaxOff + 1 ) * sizeof ( U32 ) )
+ ZSTD_cwksp_aligned_alloc_size ( ( 1 < < Litbits ) * sizeof ( U32 ) )
+ ZSTD_cwksp_aligned_alloc_size ( ( ZSTD_OPT_NUM + 1 ) * sizeof ( ZSTD_match_t ) )
+ ZSTD_cwksp_aligned_alloc_size ( ( ZSTD_OPT_NUM + 1 ) * sizeof ( ZSTD_optimal_t ) ) ;
size_t const lazyAdditionalSpace = ZSTD_rowMatchFinderUsed ( cParams - > strategy , useRowMatchFinder )
? ZSTD_cwksp_aligned_alloc_size ( hSize * sizeof ( U16 ) )
: 0 ;
2020-02-08 13:56:59 +00:00
size_t const optSpace = ( forCCtx & & ( cParams - > strategy > = ZSTD_btopt ) )
? optPotentialSpace
: 0 ;
2021-05-14 22:48:15 +00:00
size_t const slackSpace = ZSTD_cwksp_slack_space_required ( ) ;
/* tables are guaranteed to be sized in multiples of 64 bytes (or 16 uint32_t) */
ZSTD_STATIC_ASSERT ( ZSTD_HASHLOG_MIN > = 4 & & ZSTD_WINDOWLOG_MIN > = 4 & & ZSTD_CHAINLOG_MIN > = 4 ) ;
2021-12-21 11:21:29 +00:00
assert ( useRowMatchFinder ! = ZSTD_ps_auto ) ;
2021-05-14 22:48:15 +00:00
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 4 , " chainSize: %u - hSize: %u - h3Size: %u " ,
( U32 ) chainSize , ( U32 ) hSize , ( U32 ) h3Size ) ;
2021-05-14 22:48:15 +00:00
return tableSpace + optSpace + slackSpace + lazyAdditionalSpace ;
2020-02-08 13:56:59 +00:00
}
2023-02-10 19:25:40 +00:00
/* Helper function for calculating memory requirements.
* Gives a tighter bound than ZSTD_sequenceBound ( ) by taking minMatch into account . */
static size_t ZSTD_maxNbSeq ( size_t blockSize , unsigned minMatch , int useSequenceProducer ) {
U32 const divider = ( minMatch = = 3 | | useSequenceProducer ) ? 3 : 4 ;
return blockSize / divider ;
}
2020-12-18 14:46:27 +00:00
static size_t ZSTD_estimateCCtxSize_usingCCtxParams_internal (
const ZSTD_compressionParameters * cParams ,
const ldmParams_t * ldmParams ,
const int isStatic ,
2021-12-21 11:21:29 +00:00
const ZSTD_paramSwitch_e useRowMatchFinder ,
2020-12-18 14:46:27 +00:00
const size_t buffInSize ,
const size_t buffOutSize ,
2023-02-10 19:25:40 +00:00
const U64 pledgedSrcSize ,
int useSequenceProducer ,
size_t maxBlockSize )
2020-02-08 13:56:59 +00:00
{
2021-12-21 11:21:29 +00:00
size_t const windowSize = ( size_t ) BOUNDED ( 1ULL , 1ULL < < cParams - > windowLog , pledgedSrcSize ) ;
2023-02-10 19:25:40 +00:00
size_t const blockSize = MIN ( ZSTD_resolveMaxBlockSize ( maxBlockSize ) , windowSize ) ;
size_t const maxNbSeq = ZSTD_maxNbSeq ( blockSize , cParams - > minMatch , useSequenceProducer ) ;
2020-12-18 14:46:27 +00:00
size_t const tokenSpace = ZSTD_cwksp_alloc_size ( WILDCOPY_OVERLENGTH + blockSize )
2021-05-14 22:48:15 +00:00
+ ZSTD_cwksp_aligned_alloc_size ( maxNbSeq * sizeof ( seqDef ) )
2020-12-18 14:46:27 +00:00
+ 3 * ZSTD_cwksp_alloc_size ( maxNbSeq * sizeof ( BYTE ) ) ;
size_t const entropySpace = ZSTD_cwksp_alloc_size ( ENTROPY_WORKSPACE_SIZE ) ;
size_t const blockStateSpace = 2 * ZSTD_cwksp_alloc_size ( sizeof ( ZSTD_compressedBlockState_t ) ) ;
2021-05-14 22:48:15 +00:00
size_t const matchStateSize = ZSTD_sizeof_matchState ( cParams , useRowMatchFinder , /* enableDedicatedDictSearch */ 0 , /* forCCtx */ 1 ) ;
2020-02-08 13:56:59 +00:00
2020-12-18 14:46:27 +00:00
size_t const ldmSpace = ZSTD_ldm_getTableSize ( * ldmParams ) ;
size_t const maxNbLdmSeq = ZSTD_ldm_getMaxNbSeq ( * ldmParams , blockSize ) ;
2021-12-21 11:21:29 +00:00
size_t const ldmSeqSpace = ldmParams - > enableLdm = = ZSTD_ps_enable ?
2021-05-14 22:48:15 +00:00
ZSTD_cwksp_aligned_alloc_size ( maxNbLdmSeq * sizeof ( rawSeq ) ) : 0 ;
2020-02-08 13:56:59 +00:00
2020-06-04 19:03:27 +00:00
2020-12-18 14:46:27 +00:00
size_t const bufferSpace = ZSTD_cwksp_alloc_size ( buffInSize )
+ ZSTD_cwksp_alloc_size ( buffOutSize ) ;
2020-02-08 13:56:59 +00:00
2020-12-18 14:46:27 +00:00
size_t const cctxSpace = isStatic ? ZSTD_cwksp_alloc_size ( sizeof ( ZSTD_CCtx ) ) : 0 ;
2020-06-04 19:03:27 +00:00
2023-02-10 19:25:40 +00:00
size_t const maxNbExternalSeq = ZSTD_sequenceBound ( blockSize ) ;
size_t const externalSeqSpace = useSequenceProducer
? ZSTD_cwksp_aligned_alloc_size ( maxNbExternalSeq * sizeof ( ZSTD_Sequence ) )
: 0 ;
2020-12-18 14:46:27 +00:00
size_t const neededSpace =
cctxSpace +
entropySpace +
blockStateSpace +
ldmSpace +
ldmSeqSpace +
matchStateSize +
tokenSpace +
2023-02-10 19:25:40 +00:00
bufferSpace +
externalSeqSpace ;
2020-12-18 14:46:27 +00:00
DEBUGLOG ( 5 , " estimate workspace : %u " , ( U32 ) neededSpace ) ;
return neededSpace ;
}
size_t ZSTD_estimateCCtxSize_usingCCtxParams ( const ZSTD_CCtx_params * params )
{
ZSTD_compressionParameters const cParams =
ZSTD_getCParamsFromCCtxParams ( params , ZSTD_CONTENTSIZE_UNKNOWN , 0 , ZSTD_cpm_noAttachDict ) ;
2021-12-21 11:21:29 +00:00
ZSTD_paramSwitch_e const useRowMatchFinder = ZSTD_resolveRowMatchFinderMode ( params - > useRowMatchFinder ,
& cParams ) ;
2020-12-18 14:46:27 +00:00
RETURN_ERROR_IF ( params - > nbWorkers > 0 , GENERIC , " Estimate CCtx size is supported for single-threaded compression only. " ) ;
/* estimateCCtxSize is for one-shot compression. So no buffers should
* be needed . However , we still allocate two 0 - sized buffers , which can
* take space under ASAN . */
return ZSTD_estimateCCtxSize_usingCCtxParams_internal (
2023-02-10 19:25:40 +00:00
& cParams , & params - > ldmParams , 1 , useRowMatchFinder , 0 , 0 , ZSTD_CONTENTSIZE_UNKNOWN , params - > useSequenceProducer , params - > maxBlockSize ) ;
2020-02-08 13:56:59 +00:00
}
size_t ZSTD_estimateCCtxSize_usingCParams ( ZSTD_compressionParameters cParams )
{
2021-05-14 22:48:15 +00:00
ZSTD_CCtx_params initialParams = ZSTD_makeCCtxParamsFromCParams ( cParams ) ;
if ( ZSTD_rowMatchFinderSupported ( cParams . strategy ) ) {
/* Pick bigger of not using and using row-based matchfinder for greedy and lazy strategies */
size_t noRowCCtxSize ;
size_t rowCCtxSize ;
2021-12-21 11:21:29 +00:00
initialParams . useRowMatchFinder = ZSTD_ps_disable ;
2021-05-14 22:48:15 +00:00
noRowCCtxSize = ZSTD_estimateCCtxSize_usingCCtxParams ( & initialParams ) ;
2021-12-21 11:21:29 +00:00
initialParams . useRowMatchFinder = ZSTD_ps_enable ;
2021-05-14 22:48:15 +00:00
rowCCtxSize = ZSTD_estimateCCtxSize_usingCCtxParams ( & initialParams ) ;
return MAX ( noRowCCtxSize , rowCCtxSize ) ;
} else {
return ZSTD_estimateCCtxSize_usingCCtxParams ( & initialParams ) ;
}
2020-02-08 13:56:59 +00:00
}
static size_t ZSTD_estimateCCtxSize_internal ( int compressionLevel )
{
2021-05-14 22:48:15 +00:00
int tier = 0 ;
size_t largestSize = 0 ;
static const unsigned long long srcSizeTiers [ 4 ] = { 16 KB , 128 KB , 256 KB , ZSTD_CONTENTSIZE_UNKNOWN } ;
for ( ; tier < 4 ; + + tier ) {
/* Choose the set of cParams for a given level across all srcSizes that give the largest cctxSize */
ZSTD_compressionParameters const cParams = ZSTD_getCParams_internal ( compressionLevel , srcSizeTiers [ tier ] , 0 , ZSTD_cpm_noAttachDict ) ;
largestSize = MAX ( ZSTD_estimateCCtxSize_usingCParams ( cParams ) , largestSize ) ;
}
return largestSize ;
2020-02-08 13:56:59 +00:00
}
size_t ZSTD_estimateCCtxSize ( int compressionLevel )
{
int level ;
size_t memBudget = 0 ;
for ( level = MIN ( compressionLevel , 1 ) ; level < = compressionLevel ; level + + ) {
2021-05-14 22:48:15 +00:00
/* Ensure monotonically increasing memory usage as compression level increases */
2020-02-08 13:56:59 +00:00
size_t const newMB = ZSTD_estimateCCtxSize_internal ( level ) ;
if ( newMB > memBudget ) memBudget = newMB ;
}
return memBudget ;
}
size_t ZSTD_estimateCStreamSize_usingCCtxParams ( const ZSTD_CCtx_params * params )
{
RETURN_ERROR_IF ( params - > nbWorkers > 0 , GENERIC , " Estimate CCtx size is supported for single-threaded compression only. " ) ;
{ ZSTD_compressionParameters const cParams =
2020-12-18 14:46:27 +00:00
ZSTD_getCParamsFromCCtxParams ( params , ZSTD_CONTENTSIZE_UNKNOWN , 0 , ZSTD_cpm_noAttachDict ) ;
2023-02-10 19:25:40 +00:00
size_t const blockSize = MIN ( ZSTD_resolveMaxBlockSize ( params - > maxBlockSize ) , ( size_t ) 1 < < cParams . windowLog ) ;
2020-12-18 14:46:27 +00:00
size_t const inBuffSize = ( params - > inBufferMode = = ZSTD_bm_buffered )
? ( ( size_t ) 1 < < cParams . windowLog ) + blockSize
: 0 ;
size_t const outBuffSize = ( params - > outBufferMode = = ZSTD_bm_buffered )
? ZSTD_compressBound ( blockSize ) + 1
: 0 ;
2021-12-21 11:21:29 +00:00
ZSTD_paramSwitch_e const useRowMatchFinder = ZSTD_resolveRowMatchFinderMode ( params - > useRowMatchFinder , & params - > cParams ) ;
2020-12-18 14:46:27 +00:00
return ZSTD_estimateCCtxSize_usingCCtxParams_internal (
2021-05-14 22:48:15 +00:00
& cParams , & params - > ldmParams , 1 , useRowMatchFinder , inBuffSize , outBuffSize ,
2023-02-10 19:25:40 +00:00
ZSTD_CONTENTSIZE_UNKNOWN , params - > useSequenceProducer , params - > maxBlockSize ) ;
2020-02-08 13:56:59 +00:00
}
}
size_t ZSTD_estimateCStreamSize_usingCParams ( ZSTD_compressionParameters cParams )
{
2021-05-14 22:48:15 +00:00
ZSTD_CCtx_params initialParams = ZSTD_makeCCtxParamsFromCParams ( cParams ) ;
if ( ZSTD_rowMatchFinderSupported ( cParams . strategy ) ) {
/* Pick bigger of not using and using row-based matchfinder for greedy and lazy strategies */
size_t noRowCCtxSize ;
size_t rowCCtxSize ;
2021-12-21 11:21:29 +00:00
initialParams . useRowMatchFinder = ZSTD_ps_disable ;
2021-05-14 22:48:15 +00:00
noRowCCtxSize = ZSTD_estimateCStreamSize_usingCCtxParams ( & initialParams ) ;
2021-12-21 11:21:29 +00:00
initialParams . useRowMatchFinder = ZSTD_ps_enable ;
2021-05-14 22:48:15 +00:00
rowCCtxSize = ZSTD_estimateCStreamSize_usingCCtxParams ( & initialParams ) ;
return MAX ( noRowCCtxSize , rowCCtxSize ) ;
} else {
return ZSTD_estimateCStreamSize_usingCCtxParams ( & initialParams ) ;
}
2020-02-08 13:56:59 +00:00
}
static size_t ZSTD_estimateCStreamSize_internal ( int compressionLevel )
{
2020-12-18 14:46:27 +00:00
ZSTD_compressionParameters const cParams = ZSTD_getCParams_internal ( compressionLevel , ZSTD_CONTENTSIZE_UNKNOWN , 0 , ZSTD_cpm_noAttachDict ) ;
2020-02-08 13:56:59 +00:00
return ZSTD_estimateCStreamSize_usingCParams ( cParams ) ;
}
size_t ZSTD_estimateCStreamSize ( int compressionLevel )
{
int level ;
size_t memBudget = 0 ;
for ( level = MIN ( compressionLevel , 1 ) ; level < = compressionLevel ; level + + ) {
size_t const newMB = ZSTD_estimateCStreamSize_internal ( level ) ;
if ( newMB > memBudget ) memBudget = newMB ;
}
return memBudget ;
}
/* ZSTD_getFrameProgression():
* tells how much data has been consumed ( input ) and produced ( output ) for current frame .
* able to count progression inside worker threads ( non - blocking mode ) .
*/
ZSTD_frameProgression ZSTD_getFrameProgression ( const ZSTD_CCtx * cctx )
{
# ifdef ZSTD_MULTITHREAD
if ( cctx - > appliedParams . nbWorkers > 0 ) {
return ZSTDMT_getFrameProgression ( cctx - > mtctx ) ;
}
# endif
{ ZSTD_frameProgression fp ;
size_t const buffered = ( cctx - > inBuff = = NULL ) ? 0 :
cctx - > inBuffPos - cctx - > inToCompress ;
if ( buffered ) assert ( cctx - > inBuffPos > = cctx - > inToCompress ) ;
assert ( buffered < = ZSTD_BLOCKSIZE_MAX ) ;
fp . ingested = cctx - > consumedSrcSize + buffered ;
fp . consumed = cctx - > consumedSrcSize ;
fp . produced = cctx - > producedCSize ;
fp . flushed = cctx - > producedCSize ; /* simplified; some data might still be left within streaming output buffer */
fp . currentJobID = 0 ;
fp . nbActiveWorkers = 0 ;
return fp ;
} }
/*! ZSTD_toFlushNow()
* Only useful for multithreading scenarios currently ( nbWorkers > = 1 ) .
*/
size_t ZSTD_toFlushNow ( ZSTD_CCtx * cctx )
{
# ifdef ZSTD_MULTITHREAD
if ( cctx - > appliedParams . nbWorkers > 0 ) {
return ZSTDMT_toFlushNow ( cctx - > mtctx ) ;
}
# endif
( void ) cctx ;
return 0 ; /* over-simplification; could also check if context is currently running in streaming mode, and in which case, report how many bytes are left to be flushed within output buffer */
}
static void ZSTD_assertEqualCParams ( ZSTD_compressionParameters cParams1 ,
ZSTD_compressionParameters cParams2 )
{
( void ) cParams1 ;
( void ) cParams2 ;
assert ( cParams1 . windowLog = = cParams2 . windowLog ) ;
assert ( cParams1 . chainLog = = cParams2 . chainLog ) ;
assert ( cParams1 . hashLog = = cParams2 . hashLog ) ;
assert ( cParams1 . searchLog = = cParams2 . searchLog ) ;
assert ( cParams1 . minMatch = = cParams2 . minMatch ) ;
assert ( cParams1 . targetLength = = cParams2 . targetLength ) ;
assert ( cParams1 . strategy = = cParams2 . strategy ) ;
}
void ZSTD_reset_compressedBlockState ( ZSTD_compressedBlockState_t * bs )
{
int i ;
for ( i = 0 ; i < ZSTD_REP_NUM ; + + i )
bs - > rep [ i ] = repStartValue [ i ] ;
bs - > entropy . huf . repeatMode = HUF_repeat_none ;
bs - > entropy . fse . offcode_repeatMode = FSE_repeat_none ;
bs - > entropy . fse . matchlength_repeatMode = FSE_repeat_none ;
bs - > entropy . fse . litlength_repeatMode = FSE_repeat_none ;
}
/*! ZSTD_invalidateMatchState()
* Invalidate all the matches in the match finder tables .
* Requires nextSrc and base to be set ( can be NULL ) .
*/
static void ZSTD_invalidateMatchState ( ZSTD_matchState_t * ms )
{
ZSTD_window_clear ( & ms - > window ) ;
ms - > nextToUpdate = ms - > window . dictLimit ;
ms - > loadedDictEnd = 0 ;
ms - > opt . litLengthSum = 0 ; /* force reset of btopt stats */
ms - > dictMatchState = NULL ;
}
/**
* Controls , for this matchState reset , whether the tables need to be cleared /
* prepared for the coming compression ( ZSTDcrp_makeClean ) , or whether the
* tables can be left unclean ( ZSTDcrp_leaveDirty ) , because we know that a
* subsequent operation will overwrite the table space anyways ( e . g . , copying
* the matchState contents in from a CDict ) .
*/
typedef enum {
ZSTDcrp_makeClean ,
ZSTDcrp_leaveDirty
} ZSTD_compResetPolicy_e ;
/**
* Controls , for this matchState reset , whether indexing can continue where it
* left off ( ZSTDirp_continue ) , or whether it needs to be restarted from zero
* ( ZSTDirp_reset ) .
*/
typedef enum {
ZSTDirp_continue ,
ZSTDirp_reset
} ZSTD_indexResetPolicy_e ;
typedef enum {
ZSTD_resetTarget_CDict ,
ZSTD_resetTarget_CCtx
} ZSTD_resetTarget_e ;
2021-05-14 22:48:15 +00:00
2020-02-08 13:56:59 +00:00
static size_t
ZSTD_reset_matchState ( ZSTD_matchState_t * ms ,
ZSTD_cwksp * ws ,
const ZSTD_compressionParameters * cParams ,
2021-12-21 11:21:29 +00:00
const ZSTD_paramSwitch_e useRowMatchFinder ,
2020-02-08 13:56:59 +00:00
const ZSTD_compResetPolicy_e crp ,
const ZSTD_indexResetPolicy_e forceResetIndex ,
const ZSTD_resetTarget_e forWho )
{
2021-05-14 22:48:15 +00:00
/* disable chain table allocation for fast or row-based strategies */
size_t const chainSize = ZSTD_allocateChainTable ( cParams - > strategy , useRowMatchFinder ,
ms - > dedicatedDictSearch & & ( forWho = = ZSTD_resetTarget_CDict ) )
? ( ( size_t ) 1 < < cParams - > chainLog )
: 0 ;
2020-02-08 13:56:59 +00:00
size_t const hSize = ( ( size_t ) 1 ) < < cParams - > hashLog ;
U32 const hashLog3 = ( ( forWho = = ZSTD_resetTarget_CCtx ) & & cParams - > minMatch = = 3 ) ? MIN ( ZSTD_HASHLOG3_MAX , cParams - > windowLog ) : 0 ;
size_t const h3Size = hashLog3 ? ( ( size_t ) 1 ) < < hashLog3 : 0 ;
DEBUGLOG ( 4 , " reset indices : %u " , forceResetIndex = = ZSTDirp_reset ) ;
2021-12-21 11:21:29 +00:00
assert ( useRowMatchFinder ! = ZSTD_ps_auto ) ;
2020-02-08 13:56:59 +00:00
if ( forceResetIndex = = ZSTDirp_reset ) {
ZSTD_window_init ( & ms - > window ) ;
ZSTD_cwksp_mark_tables_dirty ( ws ) ;
}
ms - > hashLog3 = hashLog3 ;
ZSTD_invalidateMatchState ( ms ) ;
assert ( ! ZSTD_cwksp_reserve_failed ( ws ) ) ; /* check that allocation hasn't already failed */
ZSTD_cwksp_clear_tables ( ws ) ;
DEBUGLOG ( 5 , " reserving table space " ) ;
/* table Space */
ms - > hashTable = ( U32 * ) ZSTD_cwksp_reserve_table ( ws , hSize * sizeof ( U32 ) ) ;
ms - > chainTable = ( U32 * ) ZSTD_cwksp_reserve_table ( ws , chainSize * sizeof ( U32 ) ) ;
ms - > hashTable3 = ( U32 * ) ZSTD_cwksp_reserve_table ( ws , h3Size * sizeof ( U32 ) ) ;
RETURN_ERROR_IF ( ZSTD_cwksp_reserve_failed ( ws ) , memory_allocation ,
" failed a workspace allocation in ZSTD_reset_matchState " ) ;
DEBUGLOG ( 4 , " reset table : %u " , crp ! = ZSTDcrp_leaveDirty ) ;
if ( crp ! = ZSTDcrp_leaveDirty ) {
/* reset tables only */
ZSTD_cwksp_clean_tables ( ws ) ;
}
/* opt parser space */
if ( ( forWho = = ZSTD_resetTarget_CCtx ) & & ( cParams - > strategy > = ZSTD_btopt ) ) {
DEBUGLOG ( 4 , " reserving optimal parser space " ) ;
ms - > opt . litFreq = ( unsigned * ) ZSTD_cwksp_reserve_aligned ( ws , ( 1 < < Litbits ) * sizeof ( unsigned ) ) ;
ms - > opt . litLengthFreq = ( unsigned * ) ZSTD_cwksp_reserve_aligned ( ws , ( MaxLL + 1 ) * sizeof ( unsigned ) ) ;
ms - > opt . matchLengthFreq = ( unsigned * ) ZSTD_cwksp_reserve_aligned ( ws , ( MaxML + 1 ) * sizeof ( unsigned ) ) ;
ms - > opt . offCodeFreq = ( unsigned * ) ZSTD_cwksp_reserve_aligned ( ws , ( MaxOff + 1 ) * sizeof ( unsigned ) ) ;
ms - > opt . matchTable = ( ZSTD_match_t * ) ZSTD_cwksp_reserve_aligned ( ws , ( ZSTD_OPT_NUM + 1 ) * sizeof ( ZSTD_match_t ) ) ;
ms - > opt . priceTable = ( ZSTD_optimal_t * ) ZSTD_cwksp_reserve_aligned ( ws , ( ZSTD_OPT_NUM + 1 ) * sizeof ( ZSTD_optimal_t ) ) ;
}
2021-05-14 22:48:15 +00:00
if ( ZSTD_rowMatchFinderUsed ( cParams - > strategy , useRowMatchFinder ) ) {
{ /* Row match finder needs an additional table of hashes ("tags") */
size_t const tagTableSize = hSize * sizeof ( U16 ) ;
ms - > tagTable = ( U16 * ) ZSTD_cwksp_reserve_aligned ( ws , tagTableSize ) ;
if ( ms - > tagTable ) ZSTD_memset ( ms - > tagTable , 0 , tagTableSize ) ;
}
{ /* Switch to 32-entry rows if searchLog is 5 (or more) */
2021-12-21 11:21:29 +00:00
U32 const rowLog = BOUNDED ( 4 , cParams - > searchLog , 6 ) ;
assert ( cParams - > hashLog > = rowLog ) ;
2021-05-14 22:48:15 +00:00
ms - > rowHashLog = cParams - > hashLog - rowLog ;
}
}
2020-02-08 13:56:59 +00:00
ms - > cParams = * cParams ;
RETURN_ERROR_IF ( ZSTD_cwksp_reserve_failed ( ws ) , memory_allocation ,
" failed a workspace allocation in ZSTD_reset_matchState " ) ;
return 0 ;
}
/* ZSTD_indexTooCloseToMax() :
* minor optimization : prefer memset ( ) rather than reduceIndex ( )
* which is measurably slow in some circumstances ( reported for Visual Studio ) .
* Works when re - using a context for a lot of smallish inputs :
* if all inputs are smaller than ZSTD_INDEXOVERFLOW_MARGIN ,
* memset ( ) will be triggered before reduceIndex ( ) .
*/
# define ZSTD_INDEXOVERFLOW_MARGIN (16 MB)
static int ZSTD_indexTooCloseToMax ( ZSTD_window_t w )
{
return ( size_t ) ( w . nextSrc - w . base ) > ( ZSTD_CURRENT_MAX - ZSTD_INDEXOVERFLOW_MARGIN ) ;
}
2021-05-14 22:48:15 +00:00
/** ZSTD_dictTooBig():
* When dictionaries are larger than ZSTD_CHUNKSIZE_MAX they can ' t be loaded in
* one go generically . So we ensure that in that case we reset the tables to zero ,
* so that we can load as much of the dictionary as possible .
*/
static int ZSTD_dictTooBig ( size_t const loadedDictSize )
{
return loadedDictSize > ZSTD_CHUNKSIZE_MAX ;
}
2020-02-08 13:56:59 +00:00
/*! ZSTD_resetCCtx_internal() :
2021-05-14 22:48:15 +00:00
* @ param loadedDictSize The size of the dictionary to be loaded
* into the context , if any . If no dictionary is used , or the
* dictionary is being attached / copied , then pass 0.
* note : ` params ` are assumed fully validated at this stage .
*/
2020-02-08 13:56:59 +00:00
static size_t ZSTD_resetCCtx_internal ( ZSTD_CCtx * zc ,
2021-05-14 22:48:15 +00:00
ZSTD_CCtx_params const * params ,
2020-02-08 13:56:59 +00:00
U64 const pledgedSrcSize ,
2021-05-14 22:48:15 +00:00
size_t const loadedDictSize ,
2020-02-08 13:56:59 +00:00
ZSTD_compResetPolicy_e const crp ,
ZSTD_buffered_policy_e const zbuff )
{
ZSTD_cwksp * const ws = & zc - > workspace ;
2021-12-21 11:21:29 +00:00
DEBUGLOG ( 4 , " ZSTD_resetCCtx_internal: pledgedSrcSize=%u, wlog=%u, useRowMatchFinder=%d useBlockSplitter=%d " ,
( U32 ) pledgedSrcSize , params - > cParams . windowLog , ( int ) params - > useRowMatchFinder , ( int ) params - > useBlockSplitter ) ;
2021-05-14 22:48:15 +00:00
assert ( ! ZSTD_isError ( ZSTD_checkCParams ( params - > cParams ) ) ) ;
2020-02-08 13:56:59 +00:00
zc - > isFirstBlock = 1 ;
2021-05-14 22:48:15 +00:00
/* Set applied params early so we can modify them for LDM,
* and point params at the applied params .
*/
zc - > appliedParams = * params ;
params = & zc - > appliedParams ;
2021-12-21 11:21:29 +00:00
assert ( params - > useRowMatchFinder ! = ZSTD_ps_auto ) ;
assert ( params - > useBlockSplitter ! = ZSTD_ps_auto ) ;
assert ( params - > ldmParams . enableLdm ! = ZSTD_ps_auto ) ;
2023-02-10 19:25:40 +00:00
assert ( params - > maxBlockSize ! = 0 ) ;
2021-12-21 11:21:29 +00:00
if ( params - > ldmParams . enableLdm = = ZSTD_ps_enable ) {
2020-02-08 13:56:59 +00:00
/* Adjust long distance matching parameters */
2021-05-14 22:48:15 +00:00
ZSTD_ldm_adjustParameters ( & zc - > appliedParams . ldmParams , & params - > cParams ) ;
assert ( params - > ldmParams . hashLog > = params - > ldmParams . bucketSizeLog ) ;
assert ( params - > ldmParams . hashRateLog < 32 ) ;
2020-02-08 13:56:59 +00:00
}
2021-05-14 22:48:15 +00:00
{ size_t const windowSize = MAX ( 1 , ( size_t ) MIN ( ( ( U64 ) 1 < < params - > cParams . windowLog ) , pledgedSrcSize ) ) ;
2023-02-10 19:25:40 +00:00
size_t const blockSize = MIN ( params - > maxBlockSize , windowSize ) ;
size_t const maxNbSeq = ZSTD_maxNbSeq ( blockSize , params - > cParams . minMatch , params - > useSequenceProducer ) ;
2021-05-14 22:48:15 +00:00
size_t const buffOutSize = ( zbuff = = ZSTDb_buffered & & params - > outBufferMode = = ZSTD_bm_buffered )
2020-12-18 14:46:27 +00:00
? ZSTD_compressBound ( blockSize ) + 1
: 0 ;
2021-05-14 22:48:15 +00:00
size_t const buffInSize = ( zbuff = = ZSTDb_buffered & & params - > inBufferMode = = ZSTD_bm_buffered )
2020-12-18 14:46:27 +00:00
? windowSize + blockSize
: 0 ;
2021-05-14 22:48:15 +00:00
size_t const maxNbLdmSeq = ZSTD_ldm_getMaxNbSeq ( params - > ldmParams , blockSize ) ;
2020-02-08 13:56:59 +00:00
2020-12-18 14:46:27 +00:00
int const indexTooClose = ZSTD_indexTooCloseToMax ( zc - > blockState . matchState . window ) ;
2021-05-14 22:48:15 +00:00
int const dictTooBig = ZSTD_dictTooBig ( loadedDictSize ) ;
2020-12-18 14:46:27 +00:00
ZSTD_indexResetPolicy_e needsIndexReset =
2021-05-14 22:48:15 +00:00
( indexTooClose | | dictTooBig | | ! zc - > initialized ) ? ZSTDirp_reset : ZSTDirp_continue ;
2020-02-08 13:56:59 +00:00
2020-12-18 14:46:27 +00:00
size_t const neededSpace =
ZSTD_estimateCCtxSize_usingCCtxParams_internal (
2021-05-14 22:48:15 +00:00
& params - > cParams , & params - > ldmParams , zc - > staticSize ! = 0 , params - > useRowMatchFinder ,
2023-02-10 19:25:40 +00:00
buffInSize , buffOutSize , pledgedSrcSize , params - > useSequenceProducer , params - > maxBlockSize ) ;
2021-05-14 22:48:15 +00:00
int resizeWorkspace ;
2020-12-18 14:46:27 +00:00
FORWARD_IF_ERROR ( neededSpace , " cctx size estimate failed! " ) ;
2020-02-08 13:56:59 +00:00
2020-06-04 19:03:27 +00:00
if ( ! zc - > staticSize ) ZSTD_cwksp_bump_oversized_duration ( ws , 0 ) ;
2020-02-08 13:56:59 +00:00
2021-05-14 22:48:15 +00:00
{ /* Check if workspace is large enough, alloc a new one if needed */
2020-02-08 13:56:59 +00:00
int const workspaceTooSmall = ZSTD_cwksp_sizeof ( ws ) < neededSpace ;
int const workspaceWasteful = ZSTD_cwksp_check_wasteful ( ws , neededSpace ) ;
2021-05-14 22:48:15 +00:00
resizeWorkspace = workspaceTooSmall | | workspaceWasteful ;
2020-12-18 14:46:27 +00:00
DEBUGLOG ( 4 , " Need %zu B workspace " , neededSpace ) ;
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 4 , " windowSize: %zu - blockSize: %zu " , windowSize , blockSize ) ;
2021-05-14 22:48:15 +00:00
if ( resizeWorkspace ) {
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 4 , " Resize workspaceSize from %zuKB to %zuKB " ,
ZSTD_cwksp_sizeof ( ws ) > > 10 ,
neededSpace > > 10 ) ;
RETURN_ERROR_IF ( zc - > staticSize , memory_allocation , " static cctx : no resize " ) ;
needsIndexReset = ZSTDirp_reset ;
ZSTD_cwksp_free ( ws , zc - > customMem ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_cwksp_create ( ws , neededSpace , zc - > customMem ) , " " ) ;
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 5 , " reserving object space " ) ;
/* Statically sized space.
* entropyWorkspace never moves ,
* though prev / next block swap places */
assert ( ZSTD_cwksp_check_available ( ws , 2 * sizeof ( ZSTD_compressedBlockState_t ) ) ) ;
zc - > blockState . prevCBlock = ( ZSTD_compressedBlockState_t * ) ZSTD_cwksp_reserve_object ( ws , sizeof ( ZSTD_compressedBlockState_t ) ) ;
RETURN_ERROR_IF ( zc - > blockState . prevCBlock = = NULL , memory_allocation , " couldn't allocate prevCBlock " ) ;
zc - > blockState . nextCBlock = ( ZSTD_compressedBlockState_t * ) ZSTD_cwksp_reserve_object ( ws , sizeof ( ZSTD_compressedBlockState_t ) ) ;
RETURN_ERROR_IF ( zc - > blockState . nextCBlock = = NULL , memory_allocation , " couldn't allocate nextCBlock " ) ;
2020-12-18 14:46:27 +00:00
zc - > entropyWorkspace = ( U32 * ) ZSTD_cwksp_reserve_object ( ws , ENTROPY_WORKSPACE_SIZE ) ;
2021-12-21 11:21:29 +00:00
RETURN_ERROR_IF ( zc - > entropyWorkspace = = NULL , memory_allocation , " couldn't allocate entropyWorkspace " ) ;
2020-02-08 13:56:59 +00:00
} }
ZSTD_cwksp_clear ( ws ) ;
/* init params */
2021-05-14 22:48:15 +00:00
zc - > blockState . matchState . cParams = params - > cParams ;
2023-02-10 19:25:40 +00:00
zc - > blockState . matchState . prefetchCDictTables = params - > prefetchCDictTables = = ZSTD_ps_enable ;
2020-02-08 13:56:59 +00:00
zc - > pledgedSrcSizePlusOne = pledgedSrcSize + 1 ;
zc - > consumedSrcSize = 0 ;
zc - > producedCSize = 0 ;
if ( pledgedSrcSize = = ZSTD_CONTENTSIZE_UNKNOWN )
zc - > appliedParams . fParams . contentSizeFlag = 0 ;
DEBUGLOG ( 4 , " pledged content size : %u ; flag : %u " ,
( unsigned ) pledgedSrcSize , zc - > appliedParams . fParams . contentSizeFlag ) ;
zc - > blockSize = blockSize ;
XXH64_reset ( & zc - > xxhState , 0 ) ;
zc - > stage = ZSTDcs_init ;
zc - > dictID = 0 ;
2021-05-14 22:48:15 +00:00
zc - > dictContentSize = 0 ;
2020-02-08 13:56:59 +00:00
ZSTD_reset_compressedBlockState ( zc - > blockState . prevCBlock ) ;
/* ZSTD_wildcopy() is used to copy into the literals buffer,
* so we have to oversize the buffer by WILDCOPY_OVERLENGTH bytes .
*/
zc - > seqStore . litStart = ZSTD_cwksp_reserve_buffer ( ws , blockSize + WILDCOPY_OVERLENGTH ) ;
zc - > seqStore . maxNbLit = blockSize ;
/* buffers */
2020-12-18 14:46:27 +00:00
zc - > bufferedPolicy = zbuff ;
2020-02-08 13:56:59 +00:00
zc - > inBuffSize = buffInSize ;
zc - > inBuff = ( char * ) ZSTD_cwksp_reserve_buffer ( ws , buffInSize ) ;
zc - > outBuffSize = buffOutSize ;
zc - > outBuff = ( char * ) ZSTD_cwksp_reserve_buffer ( ws , buffOutSize ) ;
/* ldm bucketOffsets table */
2021-12-21 11:21:29 +00:00
if ( params - > ldmParams . enableLdm = = ZSTD_ps_enable ) {
2020-02-08 13:56:59 +00:00
/* TODO: avoid memset? */
2021-05-14 22:48:15 +00:00
size_t const numBuckets =
( ( size_t ) 1 ) < < ( params - > ldmParams . hashLog -
params - > ldmParams . bucketSizeLog ) ;
zc - > ldmState . bucketOffsets = ZSTD_cwksp_reserve_buffer ( ws , numBuckets ) ;
ZSTD_memset ( zc - > ldmState . bucketOffsets , 0 , numBuckets ) ;
2020-02-08 13:56:59 +00:00
}
/* sequences storage */
ZSTD_referenceExternalSequences ( zc , NULL , 0 ) ;
zc - > seqStore . maxNbSeq = maxNbSeq ;
zc - > seqStore . llCode = ZSTD_cwksp_reserve_buffer ( ws , maxNbSeq * sizeof ( BYTE ) ) ;
zc - > seqStore . mlCode = ZSTD_cwksp_reserve_buffer ( ws , maxNbSeq * sizeof ( BYTE ) ) ;
zc - > seqStore . ofCode = ZSTD_cwksp_reserve_buffer ( ws , maxNbSeq * sizeof ( BYTE ) ) ;
zc - > seqStore . sequencesStart = ( seqDef * ) ZSTD_cwksp_reserve_aligned ( ws , maxNbSeq * sizeof ( seqDef ) ) ;
FORWARD_IF_ERROR ( ZSTD_reset_matchState (
& zc - > blockState . matchState ,
ws ,
2021-05-14 22:48:15 +00:00
& params - > cParams ,
params - > useRowMatchFinder ,
2020-02-08 13:56:59 +00:00
crp ,
needsIndexReset ,
2020-06-04 19:03:27 +00:00
ZSTD_resetTarget_CCtx ) , " " ) ;
2020-02-08 13:56:59 +00:00
/* ldm hash table */
2021-12-21 11:21:29 +00:00
if ( params - > ldmParams . enableLdm = = ZSTD_ps_enable ) {
2020-02-08 13:56:59 +00:00
/* TODO: avoid memset? */
2021-05-14 22:48:15 +00:00
size_t const ldmHSize = ( ( size_t ) 1 ) < < params - > ldmParams . hashLog ;
2020-02-08 13:56:59 +00:00
zc - > ldmState . hashTable = ( ldmEntry_t * ) ZSTD_cwksp_reserve_aligned ( ws , ldmHSize * sizeof ( ldmEntry_t ) ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memset ( zc - > ldmState . hashTable , 0 , ldmHSize * sizeof ( ldmEntry_t ) ) ;
2020-02-08 13:56:59 +00:00
zc - > ldmSequences = ( rawSeq * ) ZSTD_cwksp_reserve_aligned ( ws , maxNbLdmSeq * sizeof ( rawSeq ) ) ;
zc - > maxNbLdmSequences = maxNbLdmSeq ;
ZSTD_window_init ( & zc - > ldmState . window ) ;
2020-06-04 19:03:27 +00:00
zc - > ldmState . loadedDictEnd = 0 ;
2020-02-08 13:56:59 +00:00
}
2023-02-10 19:25:40 +00:00
/* reserve space for block-level external sequences */
if ( params - > useSequenceProducer ) {
size_t const maxNbExternalSeq = ZSTD_sequenceBound ( blockSize ) ;
zc - > externalMatchCtx . seqBufferCapacity = maxNbExternalSeq ;
zc - > externalMatchCtx . seqBuffer =
( ZSTD_Sequence * ) ZSTD_cwksp_reserve_aligned ( ws , maxNbExternalSeq * sizeof ( ZSTD_Sequence ) ) ;
}
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 3 , " wksp: finished allocating, %zd bytes remain available " , ZSTD_cwksp_available_space ( ws ) ) ;
2021-12-21 11:21:29 +00:00
assert ( ZSTD_cwksp_estimated_space_within_bounds ( ws , neededSpace , resizeWorkspace ) ) ;
2021-05-14 22:48:15 +00:00
2020-02-08 13:56:59 +00:00
zc - > initialized = 1 ;
return 0 ;
}
}
/* ZSTD_invalidateRepCodes() :
* ensures next compression will not use repcodes from previous block .
* Note : only works with regular variant ;
* do not use with extDict variant ! */
void ZSTD_invalidateRepCodes ( ZSTD_CCtx * cctx ) {
int i ;
for ( i = 0 ; i < ZSTD_REP_NUM ; i + + ) cctx - > blockState . prevCBlock - > rep [ i ] = 0 ;
assert ( ! ZSTD_window_hasExtDict ( cctx - > blockState . matchState . window ) ) ;
}
/* These are the approximate sizes for each strategy past which copying the
* dictionary tables into the working context is faster than using them
* in - place .
*/
static const size_t attachDictSizeCutoffs [ ZSTD_STRATEGY_MAX + 1 ] = {
8 KB , /* unused */
8 KB , /* ZSTD_fast */
16 KB , /* ZSTD_dfast */
32 KB , /* ZSTD_greedy */
32 KB , /* ZSTD_lazy */
32 KB , /* ZSTD_lazy2 */
32 KB , /* ZSTD_btlazy2 */
32 KB , /* ZSTD_btopt */
8 KB , /* ZSTD_btultra */
8 KB /* ZSTD_btultra2 */
} ;
static int ZSTD_shouldAttachDict ( const ZSTD_CDict * cdict ,
const ZSTD_CCtx_params * params ,
U64 pledgedSrcSize )
{
size_t cutoff = attachDictSizeCutoffs [ cdict - > matchState . cParams . strategy ] ;
2020-12-18 14:46:27 +00:00
int const dedicatedDictSearch = cdict - > matchState . dedicatedDictSearch ;
return dedicatedDictSearch
| | ( ( pledgedSrcSize < = cutoff
| | pledgedSrcSize = = ZSTD_CONTENTSIZE_UNKNOWN
| | params - > attachDictPref = = ZSTD_dictForceAttach )
& & params - > attachDictPref ! = ZSTD_dictForceCopy
& & ! params - > forceWindow ) ; /* dictMatchState isn't correctly
* handled in _enforceMaxDist */
2020-02-08 13:56:59 +00:00
}
static size_t
ZSTD_resetCCtx_byAttachingCDict ( ZSTD_CCtx * cctx ,
const ZSTD_CDict * cdict ,
ZSTD_CCtx_params params ,
U64 pledgedSrcSize ,
ZSTD_buffered_policy_e zbuff )
{
2021-05-14 22:48:15 +00:00
DEBUGLOG ( 4 , " ZSTD_resetCCtx_byAttachingCDict() pledgedSrcSize=%llu " ,
( unsigned long long ) pledgedSrcSize ) ;
2020-12-18 14:46:27 +00:00
{
ZSTD_compressionParameters adjusted_cdict_cParams = cdict - > matchState . cParams ;
2020-02-08 13:56:59 +00:00
unsigned const windowLog = params . cParams . windowLog ;
assert ( windowLog ! = 0 ) ;
/* Resize working context table params for input only, since the dict
* has its own tables . */
2020-12-18 14:46:27 +00:00
/* pledgedSrcSize == 0 means 0! */
if ( cdict - > matchState . dedicatedDictSearch ) {
ZSTD_dedicatedDictSearch_revertCParams ( & adjusted_cdict_cParams ) ;
}
params . cParams = ZSTD_adjustCParams_internal ( adjusted_cdict_cParams , pledgedSrcSize ,
2023-02-10 19:25:40 +00:00
cdict - > dictContentSize , ZSTD_cpm_attachDict ,
params . useRowMatchFinder ) ;
2020-02-08 13:56:59 +00:00
params . cParams . windowLog = windowLog ;
2021-05-14 22:48:15 +00:00
params . useRowMatchFinder = cdict - > useRowMatchFinder ; /* cdict overrides */
FORWARD_IF_ERROR ( ZSTD_resetCCtx_internal ( cctx , & params , pledgedSrcSize ,
/* loadedDictSize */ 0 ,
2020-06-04 19:03:27 +00:00
ZSTDcrp_makeClean , zbuff ) , " " ) ;
2020-12-18 14:46:27 +00:00
assert ( cctx - > appliedParams . cParams . strategy = = adjusted_cdict_cParams . strategy ) ;
2020-02-08 13:56:59 +00:00
}
{ const U32 cdictEnd = ( U32 ) ( cdict - > matchState . window . nextSrc
- cdict - > matchState . window . base ) ;
const U32 cdictLen = cdictEnd - cdict - > matchState . window . dictLimit ;
if ( cdictLen = = 0 ) {
/* don't even attach dictionaries with no contents */
DEBUGLOG ( 4 , " skipping attaching empty dictionary " ) ;
} else {
DEBUGLOG ( 4 , " attaching dictionary into context " ) ;
cctx - > blockState . matchState . dictMatchState = & cdict - > matchState ;
/* prep working match state so dict matches never have negative indices
* when they are translated to the working context ' s index space . */
if ( cctx - > blockState . matchState . window . dictLimit < cdictEnd ) {
cctx - > blockState . matchState . window . nextSrc =
cctx - > blockState . matchState . window . base + cdictEnd ;
ZSTD_window_clear ( & cctx - > blockState . matchState . window ) ;
}
/* loadedDictEnd is expressed within the referential of the active context */
cctx - > blockState . matchState . loadedDictEnd = cctx - > blockState . matchState . window . dictLimit ;
} }
cctx - > dictID = cdict - > dictID ;
2021-05-14 22:48:15 +00:00
cctx - > dictContentSize = cdict - > dictContentSize ;
2020-02-08 13:56:59 +00:00
/* copy block state */
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( cctx - > blockState . prevCBlock , & cdict - > cBlockState , sizeof ( cdict - > cBlockState ) ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
2023-02-10 19:25:40 +00:00
static void ZSTD_copyCDictTableIntoCCtx ( U32 * dst , U32 const * src , size_t tableSize ,
ZSTD_compressionParameters const * cParams ) {
if ( ZSTD_CDictIndicesAreTagged ( cParams ) ) {
/* Remove tags from the CDict table if they are present.
* See docs on " short cache " in zstd_compress_internal . h for context . */
size_t i ;
for ( i = 0 ; i < tableSize ; i + + ) {
U32 const taggedIndex = src [ i ] ;
U32 const index = taggedIndex > > ZSTD_SHORT_CACHE_TAG_BITS ;
dst [ i ] = index ;
}
} else {
ZSTD_memcpy ( dst , src , tableSize * sizeof ( U32 ) ) ;
}
}
2020-02-08 13:56:59 +00:00
static size_t ZSTD_resetCCtx_byCopyingCDict ( ZSTD_CCtx * cctx ,
const ZSTD_CDict * cdict ,
ZSTD_CCtx_params params ,
U64 pledgedSrcSize ,
ZSTD_buffered_policy_e zbuff )
{
const ZSTD_compressionParameters * cdict_cParams = & cdict - > matchState . cParams ;
2020-12-18 14:46:27 +00:00
assert ( ! cdict - > matchState . dedicatedDictSearch ) ;
2021-05-14 22:48:15 +00:00
DEBUGLOG ( 4 , " ZSTD_resetCCtx_byCopyingCDict() pledgedSrcSize=%llu " ,
( unsigned long long ) pledgedSrcSize ) ;
2020-02-08 13:56:59 +00:00
{ unsigned const windowLog = params . cParams . windowLog ;
assert ( windowLog ! = 0 ) ;
/* Copy only compression parameters related to tables. */
params . cParams = * cdict_cParams ;
params . cParams . windowLog = windowLog ;
2021-05-14 22:48:15 +00:00
params . useRowMatchFinder = cdict - > useRowMatchFinder ;
FORWARD_IF_ERROR ( ZSTD_resetCCtx_internal ( cctx , & params , pledgedSrcSize ,
/* loadedDictSize */ 0 ,
2020-06-04 19:03:27 +00:00
ZSTDcrp_leaveDirty , zbuff ) , " " ) ;
2020-02-08 13:56:59 +00:00
assert ( cctx - > appliedParams . cParams . strategy = = cdict_cParams - > strategy ) ;
assert ( cctx - > appliedParams . cParams . hashLog = = cdict_cParams - > hashLog ) ;
assert ( cctx - > appliedParams . cParams . chainLog = = cdict_cParams - > chainLog ) ;
}
ZSTD_cwksp_mark_tables_dirty ( & cctx - > workspace ) ;
2021-12-21 11:21:29 +00:00
assert ( params . useRowMatchFinder ! = ZSTD_ps_auto ) ;
2020-02-08 13:56:59 +00:00
/* copy tables */
2021-05-14 22:48:15 +00:00
{ size_t const chainSize = ZSTD_allocateChainTable ( cdict_cParams - > strategy , cdict - > useRowMatchFinder , 0 /* DDS guaranteed disabled */ )
? ( ( size_t ) 1 < < cdict_cParams - > chainLog )
: 0 ;
2020-02-08 13:56:59 +00:00
size_t const hSize = ( size_t ) 1 < < cdict_cParams - > hashLog ;
2023-02-10 19:25:40 +00:00
ZSTD_copyCDictTableIntoCCtx ( cctx - > blockState . matchState . hashTable ,
cdict - > matchState . hashTable ,
hSize , cdict_cParams ) ;
2021-05-14 22:48:15 +00:00
/* Do not copy cdict's chainTable if cctx has parameters such that it would not use chainTable */
if ( ZSTD_allocateChainTable ( cctx - > appliedParams . cParams . strategy , cctx - > appliedParams . useRowMatchFinder , 0 /* forDDSDict */ ) ) {
2023-02-10 19:25:40 +00:00
ZSTD_copyCDictTableIntoCCtx ( cctx - > blockState . matchState . chainTable ,
cdict - > matchState . chainTable ,
chainSize , cdict_cParams ) ;
2021-05-14 22:48:15 +00:00
}
/* copy tag table */
if ( ZSTD_rowMatchFinderUsed ( cdict_cParams - > strategy , cdict - > useRowMatchFinder ) ) {
size_t const tagTableSize = hSize * sizeof ( U16 ) ;
ZSTD_memcpy ( cctx - > blockState . matchState . tagTable ,
cdict - > matchState . tagTable ,
tagTableSize ) ;
}
2020-02-08 13:56:59 +00:00
}
/* Zero the hashTable3, since the cdict never fills it */
{ int const h3log = cctx - > blockState . matchState . hashLog3 ;
size_t const h3Size = h3log ? ( ( size_t ) 1 < < h3log ) : 0 ;
assert ( cdict - > matchState . hashLog3 = = 0 ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memset ( cctx - > blockState . matchState . hashTable3 , 0 , h3Size * sizeof ( U32 ) ) ;
2020-02-08 13:56:59 +00:00
}
ZSTD_cwksp_mark_tables_clean ( & cctx - > workspace ) ;
/* copy dictionary offsets */
{ ZSTD_matchState_t const * srcMatchState = & cdict - > matchState ;
ZSTD_matchState_t * dstMatchState = & cctx - > blockState . matchState ;
dstMatchState - > window = srcMatchState - > window ;
dstMatchState - > nextToUpdate = srcMatchState - > nextToUpdate ;
dstMatchState - > loadedDictEnd = srcMatchState - > loadedDictEnd ;
}
cctx - > dictID = cdict - > dictID ;
2021-05-14 22:48:15 +00:00
cctx - > dictContentSize = cdict - > dictContentSize ;
2020-02-08 13:56:59 +00:00
/* copy block state */
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( cctx - > blockState . prevCBlock , & cdict - > cBlockState , sizeof ( cdict - > cBlockState ) ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
/* We have a choice between copying the dictionary context into the working
* context , or referencing the dictionary context from the working context
* in - place . We decide here which strategy to use . */
static size_t ZSTD_resetCCtx_usingCDict ( ZSTD_CCtx * cctx ,
const ZSTD_CDict * cdict ,
const ZSTD_CCtx_params * params ,
U64 pledgedSrcSize ,
ZSTD_buffered_policy_e zbuff )
{
DEBUGLOG ( 4 , " ZSTD_resetCCtx_usingCDict (pledgedSrcSize=%u) " ,
( unsigned ) pledgedSrcSize ) ;
if ( ZSTD_shouldAttachDict ( cdict , params , pledgedSrcSize ) ) {
return ZSTD_resetCCtx_byAttachingCDict (
cctx , cdict , * params , pledgedSrcSize , zbuff ) ;
} else {
return ZSTD_resetCCtx_byCopyingCDict (
cctx , cdict , * params , pledgedSrcSize , zbuff ) ;
}
}
/*! ZSTD_copyCCtx_internal() :
* Duplicate an existing context ` srcCCtx ` into another one ` dstCCtx ` .
* Only works during stage ZSTDcs_init ( i . e . after creation , but before first call to ZSTD_compressContinue ( ) ) .
* The " context " , in this case , refers to the hash and chain tables ,
* entropy tables , and dictionary references .
* ` windowLog ` value is enforced if ! = 0 , otherwise value is copied from srcCCtx .
* @ return : 0 , or an error code */
static size_t ZSTD_copyCCtx_internal ( ZSTD_CCtx * dstCCtx ,
const ZSTD_CCtx * srcCCtx ,
ZSTD_frameParameters fParams ,
U64 pledgedSrcSize ,
ZSTD_buffered_policy_e zbuff )
{
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( srcCCtx - > stage ! = ZSTDcs_init , stage_wrong ,
" Can't copy a ctx that's not in init stage. " ) ;
2021-05-14 22:48:15 +00:00
DEBUGLOG ( 5 , " ZSTD_copyCCtx_internal " ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( & dstCCtx - > customMem , & srcCCtx - > customMem , sizeof ( ZSTD_customMem ) ) ;
2020-02-08 13:56:59 +00:00
{ ZSTD_CCtx_params params = dstCCtx - > requestedParams ;
/* Copy only compression parameters related to tables. */
params . cParams = srcCCtx - > appliedParams . cParams ;
2021-12-21 11:21:29 +00:00
assert ( srcCCtx - > appliedParams . useRowMatchFinder ! = ZSTD_ps_auto ) ;
assert ( srcCCtx - > appliedParams . useBlockSplitter ! = ZSTD_ps_auto ) ;
assert ( srcCCtx - > appliedParams . ldmParams . enableLdm ! = ZSTD_ps_auto ) ;
2021-05-14 22:48:15 +00:00
params . useRowMatchFinder = srcCCtx - > appliedParams . useRowMatchFinder ;
2021-12-21 11:21:29 +00:00
params . useBlockSplitter = srcCCtx - > appliedParams . useBlockSplitter ;
params . ldmParams = srcCCtx - > appliedParams . ldmParams ;
2020-02-08 13:56:59 +00:00
params . fParams = fParams ;
2023-02-10 19:25:40 +00:00
params . maxBlockSize = srcCCtx - > appliedParams . maxBlockSize ;
2021-05-14 22:48:15 +00:00
ZSTD_resetCCtx_internal ( dstCCtx , & params , pledgedSrcSize ,
/* loadedDictSize */ 0 ,
2020-02-08 13:56:59 +00:00
ZSTDcrp_leaveDirty , zbuff ) ;
assert ( dstCCtx - > appliedParams . cParams . windowLog = = srcCCtx - > appliedParams . cParams . windowLog ) ;
assert ( dstCCtx - > appliedParams . cParams . strategy = = srcCCtx - > appliedParams . cParams . strategy ) ;
assert ( dstCCtx - > appliedParams . cParams . hashLog = = srcCCtx - > appliedParams . cParams . hashLog ) ;
assert ( dstCCtx - > appliedParams . cParams . chainLog = = srcCCtx - > appliedParams . cParams . chainLog ) ;
assert ( dstCCtx - > blockState . matchState . hashLog3 = = srcCCtx - > blockState . matchState . hashLog3 ) ;
}
ZSTD_cwksp_mark_tables_dirty ( & dstCCtx - > workspace ) ;
/* copy tables */
2021-05-14 22:48:15 +00:00
{ size_t const chainSize = ZSTD_allocateChainTable ( srcCCtx - > appliedParams . cParams . strategy ,
srcCCtx - > appliedParams . useRowMatchFinder ,
0 /* forDDSDict */ )
? ( ( size_t ) 1 < < srcCCtx - > appliedParams . cParams . chainLog )
: 0 ;
2020-02-08 13:56:59 +00:00
size_t const hSize = ( size_t ) 1 < < srcCCtx - > appliedParams . cParams . hashLog ;
int const h3log = srcCCtx - > blockState . matchState . hashLog3 ;
size_t const h3Size = h3log ? ( ( size_t ) 1 < < h3log ) : 0 ;
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( dstCCtx - > blockState . matchState . hashTable ,
2020-02-08 13:56:59 +00:00
srcCCtx - > blockState . matchState . hashTable ,
hSize * sizeof ( U32 ) ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( dstCCtx - > blockState . matchState . chainTable ,
2020-02-08 13:56:59 +00:00
srcCCtx - > blockState . matchState . chainTable ,
chainSize * sizeof ( U32 ) ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( dstCCtx - > blockState . matchState . hashTable3 ,
2020-02-08 13:56:59 +00:00
srcCCtx - > blockState . matchState . hashTable3 ,
h3Size * sizeof ( U32 ) ) ;
}
ZSTD_cwksp_mark_tables_clean ( & dstCCtx - > workspace ) ;
/* copy dictionary offsets */
{
const ZSTD_matchState_t * srcMatchState = & srcCCtx - > blockState . matchState ;
ZSTD_matchState_t * dstMatchState = & dstCCtx - > blockState . matchState ;
dstMatchState - > window = srcMatchState - > window ;
dstMatchState - > nextToUpdate = srcMatchState - > nextToUpdate ;
dstMatchState - > loadedDictEnd = srcMatchState - > loadedDictEnd ;
}
dstCCtx - > dictID = srcCCtx - > dictID ;
2021-05-14 22:48:15 +00:00
dstCCtx - > dictContentSize = srcCCtx - > dictContentSize ;
2020-02-08 13:56:59 +00:00
/* copy block state */
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( dstCCtx - > blockState . prevCBlock , srcCCtx - > blockState . prevCBlock , sizeof ( * srcCCtx - > blockState . prevCBlock ) ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
/*! ZSTD_copyCCtx() :
* Duplicate an existing context ` srcCCtx ` into another one ` dstCCtx ` .
* Only works during stage ZSTDcs_init ( i . e . after creation , but before first call to ZSTD_compressContinue ( ) ) .
* pledgedSrcSize = = 0 means " unknown " .
* @ return : 0 , or an error code */
size_t ZSTD_copyCCtx ( ZSTD_CCtx * dstCCtx , const ZSTD_CCtx * srcCCtx , unsigned long long pledgedSrcSize )
{
ZSTD_frameParameters fParams = { 1 /*content*/ , 0 /*checksum*/ , 0 /*noDictID*/ } ;
2020-12-18 14:46:27 +00:00
ZSTD_buffered_policy_e const zbuff = srcCCtx - > bufferedPolicy ;
2020-02-08 13:56:59 +00:00
ZSTD_STATIC_ASSERT ( ( U32 ) ZSTDb_buffered = = 1 ) ;
if ( pledgedSrcSize = = 0 ) pledgedSrcSize = ZSTD_CONTENTSIZE_UNKNOWN ;
fParams . contentSizeFlag = ( pledgedSrcSize ! = ZSTD_CONTENTSIZE_UNKNOWN ) ;
return ZSTD_copyCCtx_internal ( dstCCtx , srcCCtx ,
fParams , pledgedSrcSize ,
zbuff ) ;
}
# define ZSTD_ROWSIZE 16
/*! ZSTD_reduceTable() :
* reduce table indexes by ` reducerValue ` , or squash to zero .
* PreserveMark preserves " unsorted mark " for btlazy2 strategy .
* It must be set to a clear 0 / 1 value , to remove branch during inlining .
* Presume table size is a multiple of ZSTD_ROWSIZE
* to help auto - vectorization */
FORCE_INLINE_TEMPLATE void
ZSTD_reduceTable_internal ( U32 * const table , U32 const size , U32 const reducerValue , int const preserveMark )
{
int const nbRows = ( int ) size / ZSTD_ROWSIZE ;
int cellNb = 0 ;
int rowNb ;
2021-12-21 11:21:29 +00:00
/* Protect special index values < ZSTD_WINDOW_START_INDEX. */
U32 const reducerThreshold = reducerValue + ZSTD_WINDOW_START_INDEX ;
2020-02-08 13:56:59 +00:00
assert ( ( size & ( ZSTD_ROWSIZE - 1 ) ) = = 0 ) ; /* multiple of ZSTD_ROWSIZE */
assert ( size < ( 1U < < 31 ) ) ; /* can be casted to int */
2020-12-18 14:46:27 +00:00
# if ZSTD_MEMORY_SANITIZER && !defined (ZSTD_MSAN_DONT_POISON_WORKSPACE)
2020-02-08 13:56:59 +00:00
/* To validate that the table re-use logic is sound, and that we don't
* access table space that we haven ' t cleaned , we re - " poison " the table
* space every time we mark it dirty .
*
* This function however is intended to operate on those dirty tables and
* re - clean them . So when this function is used correctly , we can unpoison
* the memory it operated on . This introduces a blind spot though , since
* if we now try to operate on __actually__ poisoned memory , we will not
* detect that . */
__msan_unpoison ( table , size * sizeof ( U32 ) ) ;
# endif
for ( rowNb = 0 ; rowNb < nbRows ; rowNb + + ) {
int column ;
for ( column = 0 ; column < ZSTD_ROWSIZE ; column + + ) {
2021-12-21 11:21:29 +00:00
U32 newVal ;
if ( preserveMark & & table [ cellNb ] = = ZSTD_DUBT_UNSORTED_MARK ) {
/* This write is pointless, but is required(?) for the compiler
* to auto - vectorize the loop . */
newVal = ZSTD_DUBT_UNSORTED_MARK ;
} else if ( table [ cellNb ] < reducerThreshold ) {
newVal = 0 ;
} else {
newVal = table [ cellNb ] - reducerValue ;
2020-02-08 13:56:59 +00:00
}
2021-12-21 11:21:29 +00:00
table [ cellNb ] = newVal ;
2020-02-08 13:56:59 +00:00
cellNb + + ;
} }
}
static void ZSTD_reduceTable ( U32 * const table , U32 const size , U32 const reducerValue )
{
ZSTD_reduceTable_internal ( table , size , reducerValue , 0 ) ;
}
static void ZSTD_reduceTable_btlazy2 ( U32 * const table , U32 const size , U32 const reducerValue )
{
ZSTD_reduceTable_internal ( table , size , reducerValue , 1 ) ;
}
/*! ZSTD_reduceIndex() :
* rescale all indexes to avoid future overflow ( indexes are U32 ) */
static void ZSTD_reduceIndex ( ZSTD_matchState_t * ms , ZSTD_CCtx_params const * params , const U32 reducerValue )
{
{ U32 const hSize = ( U32 ) 1 < < params - > cParams . hashLog ;
ZSTD_reduceTable ( ms - > hashTable , hSize , reducerValue ) ;
}
2021-05-14 22:48:15 +00:00
if ( ZSTD_allocateChainTable ( params - > cParams . strategy , params - > useRowMatchFinder , ( U32 ) ms - > dedicatedDictSearch ) ) {
2020-02-08 13:56:59 +00:00
U32 const chainSize = ( U32 ) 1 < < params - > cParams . chainLog ;
if ( params - > cParams . strategy = = ZSTD_btlazy2 )
ZSTD_reduceTable_btlazy2 ( ms - > chainTable , chainSize , reducerValue ) ;
else
ZSTD_reduceTable ( ms - > chainTable , chainSize , reducerValue ) ;
}
if ( ms - > hashLog3 ) {
U32 const h3Size = ( U32 ) 1 < < ms - > hashLog3 ;
ZSTD_reduceTable ( ms - > hashTable3 , h3Size , reducerValue ) ;
}
}
/*-*******************************************************
* Block entropic compression
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
/* See doc/zstd_compression_format.md for detailed format description */
2023-02-10 19:25:40 +00:00
int ZSTD_seqToCodes ( const seqStore_t * seqStorePtr )
2020-02-08 13:56:59 +00:00
{
const seqDef * const sequences = seqStorePtr - > sequencesStart ;
BYTE * const llCodeTable = seqStorePtr - > llCode ;
BYTE * const ofCodeTable = seqStorePtr - > ofCode ;
BYTE * const mlCodeTable = seqStorePtr - > mlCode ;
U32 const nbSeq = ( U32 ) ( seqStorePtr - > sequences - seqStorePtr - > sequencesStart ) ;
U32 u ;
2023-02-10 19:25:40 +00:00
int longOffsets = 0 ;
2020-02-08 13:56:59 +00:00
assert ( nbSeq < = seqStorePtr - > maxNbSeq ) ;
for ( u = 0 ; u < nbSeq ; u + + ) {
U32 const llv = sequences [ u ] . litLength ;
2023-02-10 19:25:40 +00:00
U32 const ofCode = ZSTD_highbit32 ( sequences [ u ] . offBase ) ;
2022-03-18 00:52:11 +00:00
U32 const mlv = sequences [ u ] . mlBase ;
2020-02-08 13:56:59 +00:00
llCodeTable [ u ] = ( BYTE ) ZSTD_LLcode ( llv ) ;
2023-02-10 19:25:40 +00:00
ofCodeTable [ u ] = ( BYTE ) ofCode ;
2020-02-08 13:56:59 +00:00
mlCodeTable [ u ] = ( BYTE ) ZSTD_MLcode ( mlv ) ;
2023-02-10 19:25:40 +00:00
assert ( ! ( MEM_64bits ( ) & & ofCode > = STREAM_ACCUMULATOR_MIN ) ) ;
if ( MEM_32bits ( ) & & ofCode > = STREAM_ACCUMULATOR_MIN )
longOffsets = 1 ;
2020-02-08 13:56:59 +00:00
}
2021-05-14 22:48:15 +00:00
if ( seqStorePtr - > longLengthType = = ZSTD_llt_literalLength )
2020-02-08 13:56:59 +00:00
llCodeTable [ seqStorePtr - > longLengthPos ] = MaxLL ;
2021-05-14 22:48:15 +00:00
if ( seqStorePtr - > longLengthType = = ZSTD_llt_matchLength )
2020-02-08 13:56:59 +00:00
mlCodeTable [ seqStorePtr - > longLengthPos ] = MaxML ;
2023-02-10 19:25:40 +00:00
return longOffsets ;
2020-02-08 13:56:59 +00:00
}
/* ZSTD_useTargetCBlockSize():
* Returns if target compressed block size param is being used .
* If used , compression will do best effort to make a compressed block size to be around targetCBlockSize .
* Returns 1 if true , 0 otherwise . */
static int ZSTD_useTargetCBlockSize ( const ZSTD_CCtx_params * cctxParams )
{
DEBUGLOG ( 5 , " ZSTD_useTargetCBlockSize (targetCBlockSize=%zu) " , cctxParams - > targetCBlockSize ) ;
return ( cctxParams - > targetCBlockSize ! = 0 ) ;
}
2021-05-14 22:48:15 +00:00
/* ZSTD_blockSplitterEnabled():
* Returns if block splitting param is being used
* If used , compression will do best effort to split a block in order to improve compression ratio .
2021-12-21 11:21:29 +00:00
* At the time this function is called , the parameter must be finalized .
2021-05-14 22:48:15 +00:00
* Returns 1 if true , 0 otherwise . */
static int ZSTD_blockSplitterEnabled ( ZSTD_CCtx_params * cctxParams )
{
2021-12-21 11:21:29 +00:00
DEBUGLOG ( 5 , " ZSTD_blockSplitterEnabled (useBlockSplitter=%d) " , cctxParams - > useBlockSplitter ) ;
assert ( cctxParams - > useBlockSplitter ! = ZSTD_ps_auto ) ;
return ( cctxParams - > useBlockSplitter = = ZSTD_ps_enable ) ;
2021-05-14 22:48:15 +00:00
}
/* Type returned by ZSTD_buildSequencesStatistics containing finalized symbol encoding types
* and size of the sequences statistics
*/
typedef struct {
U32 LLtype ;
U32 Offtype ;
U32 MLtype ;
size_t size ;
size_t lastCountSize ; /* Accounts for bug in 1.3.4. More detail in ZSTD_entropyCompressSeqStore_internal() */
2023-02-10 19:25:40 +00:00
int longOffsets ;
2021-05-14 22:48:15 +00:00
} ZSTD_symbolEncodingTypeStats_t ;
/* ZSTD_buildSequencesStatistics():
* Returns a ZSTD_symbolEncodingTypeStats_t , or a zstd error code in the ` size ` field .
* Modifies ` nextEntropy ` to have the appropriate values as a side effect .
* nbSeq must be greater than 0.
*
* entropyWkspSize must be of size at least ENTROPY_WORKSPACE_SIZE - ( MaxSeq + 1 ) * sizeof ( U32 )
*/
static ZSTD_symbolEncodingTypeStats_t
2023-02-10 19:25:40 +00:00
ZSTD_buildSequencesStatistics (
const seqStore_t * seqStorePtr , size_t nbSeq ,
const ZSTD_fseCTables_t * prevEntropy , ZSTD_fseCTables_t * nextEntropy ,
BYTE * dst , const BYTE * const dstEnd ,
ZSTD_strategy strategy , unsigned * countWorkspace ,
void * entropyWorkspace , size_t entropyWkspSize )
{
2021-05-14 22:48:15 +00:00
BYTE * const ostart = dst ;
const BYTE * const oend = dstEnd ;
BYTE * op = ostart ;
FSE_CTable * CTable_LitLength = nextEntropy - > litlengthCTable ;
FSE_CTable * CTable_OffsetBits = nextEntropy - > offcodeCTable ;
FSE_CTable * CTable_MatchLength = nextEntropy - > matchlengthCTable ;
const BYTE * const ofCodeTable = seqStorePtr - > ofCode ;
const BYTE * const llCodeTable = seqStorePtr - > llCode ;
const BYTE * const mlCodeTable = seqStorePtr - > mlCode ;
ZSTD_symbolEncodingTypeStats_t stats ;
stats . lastCountSize = 0 ;
/* convert length/distances into codes */
2023-02-10 19:25:40 +00:00
stats . longOffsets = ZSTD_seqToCodes ( seqStorePtr ) ;
2021-05-14 22:48:15 +00:00
assert ( op < = oend ) ;
assert ( nbSeq ! = 0 ) ; /* ZSTD_selectEncodingType() divides by nbSeq */
/* build CTable for Literal Lengths */
{ unsigned max = MaxLL ;
size_t const mostFrequent = HIST_countFast_wksp ( countWorkspace , & max , llCodeTable , nbSeq , entropyWorkspace , entropyWkspSize ) ; /* can't fail */
DEBUGLOG ( 5 , " Building LL table " ) ;
nextEntropy - > litlength_repeatMode = prevEntropy - > litlength_repeatMode ;
stats . LLtype = ZSTD_selectEncodingType ( & nextEntropy - > litlength_repeatMode ,
countWorkspace , max , mostFrequent , nbSeq ,
LLFSELog , prevEntropy - > litlengthCTable ,
LL_defaultNorm , LL_defaultNormLog ,
ZSTD_defaultAllowed , strategy ) ;
assert ( set_basic < set_compressed & & set_rle < set_compressed ) ;
assert ( ! ( stats . LLtype < set_compressed & & nextEntropy - > litlength_repeatMode ! = FSE_repeat_none ) ) ; /* We don't copy tables */
{ size_t const countSize = ZSTD_buildCTable (
op , ( size_t ) ( oend - op ) ,
CTable_LitLength , LLFSELog , ( symbolEncodingType_e ) stats . LLtype ,
countWorkspace , max , llCodeTable , nbSeq ,
LL_defaultNorm , LL_defaultNormLog , MaxLL ,
prevEntropy - > litlengthCTable ,
sizeof ( prevEntropy - > litlengthCTable ) ,
entropyWorkspace , entropyWkspSize ) ;
if ( ZSTD_isError ( countSize ) ) {
DEBUGLOG ( 3 , " ZSTD_buildCTable for LitLens failed " ) ;
stats . size = countSize ;
return stats ;
}
if ( stats . LLtype = = set_compressed )
stats . lastCountSize = countSize ;
op + = countSize ;
assert ( op < = oend ) ;
} }
/* build CTable for Offsets */
{ unsigned max = MaxOff ;
size_t const mostFrequent = HIST_countFast_wksp (
countWorkspace , & max , ofCodeTable , nbSeq , entropyWorkspace , entropyWkspSize ) ; /* can't fail */
/* We can only use the basic table if max <= DefaultMaxOff, otherwise the offsets are too large */
ZSTD_defaultPolicy_e const defaultPolicy = ( max < = DefaultMaxOff ) ? ZSTD_defaultAllowed : ZSTD_defaultDisallowed ;
DEBUGLOG ( 5 , " Building OF table " ) ;
nextEntropy - > offcode_repeatMode = prevEntropy - > offcode_repeatMode ;
stats . Offtype = ZSTD_selectEncodingType ( & nextEntropy - > offcode_repeatMode ,
countWorkspace , max , mostFrequent , nbSeq ,
OffFSELog , prevEntropy - > offcodeCTable ,
OF_defaultNorm , OF_defaultNormLog ,
defaultPolicy , strategy ) ;
assert ( ! ( stats . Offtype < set_compressed & & nextEntropy - > offcode_repeatMode ! = FSE_repeat_none ) ) ; /* We don't copy tables */
{ size_t const countSize = ZSTD_buildCTable (
op , ( size_t ) ( oend - op ) ,
CTable_OffsetBits , OffFSELog , ( symbolEncodingType_e ) stats . Offtype ,
countWorkspace , max , ofCodeTable , nbSeq ,
OF_defaultNorm , OF_defaultNormLog , DefaultMaxOff ,
prevEntropy - > offcodeCTable ,
sizeof ( prevEntropy - > offcodeCTable ) ,
entropyWorkspace , entropyWkspSize ) ;
if ( ZSTD_isError ( countSize ) ) {
DEBUGLOG ( 3 , " ZSTD_buildCTable for Offsets failed " ) ;
stats . size = countSize ;
return stats ;
}
if ( stats . Offtype = = set_compressed )
stats . lastCountSize = countSize ;
op + = countSize ;
assert ( op < = oend ) ;
} }
/* build CTable for MatchLengths */
{ unsigned max = MaxML ;
size_t const mostFrequent = HIST_countFast_wksp (
countWorkspace , & max , mlCodeTable , nbSeq , entropyWorkspace , entropyWkspSize ) ; /* can't fail */
DEBUGLOG ( 5 , " Building ML table (remaining space : %i) " , ( int ) ( oend - op ) ) ;
nextEntropy - > matchlength_repeatMode = prevEntropy - > matchlength_repeatMode ;
stats . MLtype = ZSTD_selectEncodingType ( & nextEntropy - > matchlength_repeatMode ,
countWorkspace , max , mostFrequent , nbSeq ,
MLFSELog , prevEntropy - > matchlengthCTable ,
ML_defaultNorm , ML_defaultNormLog ,
ZSTD_defaultAllowed , strategy ) ;
assert ( ! ( stats . MLtype < set_compressed & & nextEntropy - > matchlength_repeatMode ! = FSE_repeat_none ) ) ; /* We don't copy tables */
{ size_t const countSize = ZSTD_buildCTable (
op , ( size_t ) ( oend - op ) ,
CTable_MatchLength , MLFSELog , ( symbolEncodingType_e ) stats . MLtype ,
countWorkspace , max , mlCodeTable , nbSeq ,
ML_defaultNorm , ML_defaultNormLog , MaxML ,
prevEntropy - > matchlengthCTable ,
sizeof ( prevEntropy - > matchlengthCTable ) ,
entropyWorkspace , entropyWkspSize ) ;
if ( ZSTD_isError ( countSize ) ) {
DEBUGLOG ( 3 , " ZSTD_buildCTable for MatchLengths failed " ) ;
stats . size = countSize ;
return stats ;
}
if ( stats . MLtype = = set_compressed )
stats . lastCountSize = countSize ;
op + = countSize ;
assert ( op < = oend ) ;
} }
stats . size = ( size_t ) ( op - ostart ) ;
return stats ;
}
/* ZSTD_entropyCompressSeqStore_internal():
* compresses both literals and sequences
* Returns compressed size of block , or a zstd error .
*/
2021-12-21 11:21:29 +00:00
# define SUSPECT_UNCOMPRESSIBLE_LITERAL_RATIO 20
2020-02-08 13:56:59 +00:00
MEM_STATIC size_t
2023-02-10 19:25:40 +00:00
ZSTD_entropyCompressSeqStore_internal (
const seqStore_t * seqStorePtr ,
const ZSTD_entropyCTables_t * prevEntropy ,
ZSTD_entropyCTables_t * nextEntropy ,
const ZSTD_CCtx_params * cctxParams ,
void * dst , size_t dstCapacity ,
void * entropyWorkspace , size_t entropyWkspSize ,
const int bmi2 )
2020-02-08 13:56:59 +00:00
{
ZSTD_strategy const strategy = cctxParams - > cParams . strategy ;
2020-12-18 14:46:27 +00:00
unsigned * count = ( unsigned * ) entropyWorkspace ;
2020-02-08 13:56:59 +00:00
FSE_CTable * CTable_LitLength = nextEntropy - > fse . litlengthCTable ;
FSE_CTable * CTable_OffsetBits = nextEntropy - > fse . offcodeCTable ;
FSE_CTable * CTable_MatchLength = nextEntropy - > fse . matchlengthCTable ;
const seqDef * const sequences = seqStorePtr - > sequencesStart ;
2023-02-10 19:25:40 +00:00
const size_t nbSeq = ( size_t ) ( seqStorePtr - > sequences - seqStorePtr - > sequencesStart ) ;
2020-02-08 13:56:59 +00:00
const BYTE * const ofCodeTable = seqStorePtr - > ofCode ;
const BYTE * const llCodeTable = seqStorePtr - > llCode ;
const BYTE * const mlCodeTable = seqStorePtr - > mlCode ;
BYTE * const ostart = ( BYTE * ) dst ;
BYTE * const oend = ostart + dstCapacity ;
BYTE * op = ostart ;
2021-05-14 22:48:15 +00:00
size_t lastCountSize ;
2023-02-10 19:25:40 +00:00
int longOffsets = 0 ;
2020-02-08 13:56:59 +00:00
2020-12-18 14:46:27 +00:00
entropyWorkspace = count + ( MaxSeq + 1 ) ;
entropyWkspSize - = ( MaxSeq + 1 ) * sizeof ( * count ) ;
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " ZSTD_entropyCompressSeqStore_internal (nbSeq=%zu, dstCapacity=%zu) " , nbSeq , dstCapacity ) ;
2020-02-08 13:56:59 +00:00
ZSTD_STATIC_ASSERT ( HUF_WORKSPACE_SIZE > = ( 1 < < MAX ( MLFSELog , LLFSELog ) ) ) ;
2020-12-18 14:46:27 +00:00
assert ( entropyWkspSize > = HUF_WORKSPACE_SIZE ) ;
2020-02-08 13:56:59 +00:00
/* Compress literals */
{ const BYTE * const literals = seqStorePtr - > litStart ;
2023-02-10 19:25:40 +00:00
size_t const numSequences = ( size_t ) ( seqStorePtr - > sequences - seqStorePtr - > sequencesStart ) ;
size_t const numLiterals = ( size_t ) ( seqStorePtr - > lit - seqStorePtr - > litStart ) ;
2021-12-21 11:21:29 +00:00
/* Base suspicion of uncompressibility on ratio of literals to sequences */
unsigned const suspectUncompressible = ( numSequences = = 0 ) | | ( numLiterals / numSequences > = SUSPECT_UNCOMPRESSIBLE_LITERAL_RATIO ) ;
2020-02-08 13:56:59 +00:00
size_t const litSize = ( size_t ) ( seqStorePtr - > lit - literals ) ;
2023-02-10 19:25:40 +00:00
2020-02-08 13:56:59 +00:00
size_t const cSize = ZSTD_compressLiterals (
op , dstCapacity ,
literals , litSize ,
entropyWorkspace , entropyWkspSize ,
2023-02-10 19:25:40 +00:00
& prevEntropy - > huf , & nextEntropy - > huf ,
cctxParams - > cParams . strategy ,
ZSTD_literalsCompressionIsDisabled ( cctxParams ) ,
suspectUncompressible , bmi2 ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_compressLiterals failed " ) ;
2020-02-08 13:56:59 +00:00
assert ( cSize < = dstCapacity ) ;
op + = cSize ;
}
/* Sequences Header */
RETURN_ERROR_IF ( ( oend - op ) < 3 /*max nbSeq Size*/ + 1 /*seqHead*/ ,
2020-06-04 19:03:27 +00:00
dstSize_tooSmall , " Can't fit seq hdr in output buf! " ) ;
2020-02-08 13:56:59 +00:00
if ( nbSeq < 128 ) {
* op + + = ( BYTE ) nbSeq ;
} else if ( nbSeq < LONGNBSEQ ) {
op [ 0 ] = ( BYTE ) ( ( nbSeq > > 8 ) + 0x80 ) ;
op [ 1 ] = ( BYTE ) nbSeq ;
op + = 2 ;
} else {
op [ 0 ] = 0xFF ;
MEM_writeLE16 ( op + 1 , ( U16 ) ( nbSeq - LONGNBSEQ ) ) ;
op + = 3 ;
}
assert ( op < = oend ) ;
if ( nbSeq = = 0 ) {
/* Copy the old tables over as if we repeated them */
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( & nextEntropy - > fse , & prevEntropy - > fse , sizeof ( prevEntropy - > fse ) ) ;
2020-02-08 13:56:59 +00:00
return ( size_t ) ( op - ostart ) ;
}
2023-02-10 19:25:40 +00:00
{ BYTE * const seqHead = op + + ;
2021-05-14 22:48:15 +00:00
/* build stats for sequences */
2023-02-10 19:25:40 +00:00
const ZSTD_symbolEncodingTypeStats_t stats =
ZSTD_buildSequencesStatistics ( seqStorePtr , nbSeq ,
2021-05-14 22:48:15 +00:00
& prevEntropy - > fse , & nextEntropy - > fse ,
op , oend ,
strategy , count ,
entropyWorkspace , entropyWkspSize ) ;
FORWARD_IF_ERROR ( stats . size , " ZSTD_buildSequencesStatistics failed! " ) ;
* seqHead = ( BYTE ) ( ( stats . LLtype < < 6 ) + ( stats . Offtype < < 4 ) + ( stats . MLtype < < 2 ) ) ;
lastCountSize = stats . lastCountSize ;
op + = stats . size ;
2023-02-10 19:25:40 +00:00
longOffsets = stats . longOffsets ;
2021-05-14 22:48:15 +00:00
}
2020-02-08 13:56:59 +00:00
{ size_t const bitstreamSize = ZSTD_encodeSequences (
op , ( size_t ) ( oend - op ) ,
CTable_MatchLength , mlCodeTable ,
CTable_OffsetBits , ofCodeTable ,
CTable_LitLength , llCodeTable ,
sequences , nbSeq ,
longOffsets , bmi2 ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( bitstreamSize , " ZSTD_encodeSequences failed " ) ;
2020-02-08 13:56:59 +00:00
op + = bitstreamSize ;
assert ( op < = oend ) ;
/* zstd versions <= 1.3.4 mistakenly report corruption when
* FSE_readNCount ( ) receives a buffer < 4 bytes .
* Fixed by https : //github.com/facebook/zstd/pull/1146.
* This can happen when the last set_compressed table present is 2
* bytes and the bitstream is only one byte .
* In this exceedingly rare case , we will simply emit an uncompressed
* block , since it isn ' t worth optimizing .
*/
2021-05-14 22:48:15 +00:00
if ( lastCountSize & & ( lastCountSize + bitstreamSize ) < 4 ) {
/* lastCountSize >= 2 && bitstreamSize > 0 ==> lastCountSize == 3 */
assert ( lastCountSize + bitstreamSize = = 3 ) ;
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 5 , " Avoiding bug in zstd decoder in versions <= 1.3.4 by "
" emitting an uncompressed block. " ) ;
return 0 ;
}
}
DEBUGLOG ( 5 , " compressed block size : %u " , ( unsigned ) ( op - ostart ) ) ;
return ( size_t ) ( op - ostart ) ;
}
MEM_STATIC size_t
2023-02-10 19:25:40 +00:00
ZSTD_entropyCompressSeqStore (
const seqStore_t * seqStorePtr ,
const ZSTD_entropyCTables_t * prevEntropy ,
ZSTD_entropyCTables_t * nextEntropy ,
const ZSTD_CCtx_params * cctxParams ,
void * dst , size_t dstCapacity ,
size_t srcSize ,
void * entropyWorkspace , size_t entropyWkspSize ,
int bmi2 )
2020-02-08 13:56:59 +00:00
{
2021-05-14 22:48:15 +00:00
size_t const cSize = ZSTD_entropyCompressSeqStore_internal (
2020-02-08 13:56:59 +00:00
seqStorePtr , prevEntropy , nextEntropy , cctxParams ,
dst , dstCapacity ,
entropyWorkspace , entropyWkspSize , bmi2 ) ;
if ( cSize = = 0 ) return 0 ;
/* When srcSize <= dstCapacity, there is enough space to write a raw uncompressed block.
* Since we ran out of space , block must be not compressible , so fall back to raw uncompressed block .
*/
2023-02-10 19:25:40 +00:00
if ( ( cSize = = ERROR ( dstSize_tooSmall ) ) & ( srcSize < = dstCapacity ) ) {
DEBUGLOG ( 4 , " not enough dstCapacity (%zu) for ZSTD_entropyCompressSeqStore_internal()=> do not compress block " , dstCapacity ) ;
2020-02-08 13:56:59 +00:00
return 0 ; /* block not compressed */
2023-02-10 19:25:40 +00:00
}
2021-05-14 22:48:15 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_entropyCompressSeqStore_internal failed " ) ;
2020-02-08 13:56:59 +00:00
/* Check compressibility */
{ size_t const maxCSize = srcSize - ZSTD_minGain ( srcSize , cctxParams - > cParams . strategy ) ;
if ( cSize > = maxCSize ) return 0 ; /* block not compressed */
}
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " ZSTD_entropyCompressSeqStore() cSize: %zu " , cSize ) ;
/* libzstd decoder before > v1.5.4 is not compatible with compressed blocks of size ZSTD_BLOCKSIZE_MAX exactly.
* This restriction is indirectly already fulfilled by respecting ZSTD_minGain ( ) condition above .
*/
assert ( cSize < ZSTD_BLOCKSIZE_MAX ) ;
2020-02-08 13:56:59 +00:00
return cSize ;
}
/* ZSTD_selectBlockCompressor() :
* Not static , but internal use only ( used by long distance matcher )
* assumption : strat is a valid strategy */
2021-12-21 11:21:29 +00:00
ZSTD_blockCompressor ZSTD_selectBlockCompressor ( ZSTD_strategy strat , ZSTD_paramSwitch_e useRowMatchFinder , ZSTD_dictMode_e dictMode )
2020-02-08 13:56:59 +00:00
{
2020-12-18 14:46:27 +00:00
static const ZSTD_blockCompressor blockCompressor [ 4 ] [ ZSTD_STRATEGY_MAX + 1 ] = {
2020-02-08 13:56:59 +00:00
{ ZSTD_compressBlock_fast /* default for 0 */ ,
ZSTD_compressBlock_fast ,
ZSTD_compressBlock_doubleFast ,
ZSTD_compressBlock_greedy ,
ZSTD_compressBlock_lazy ,
ZSTD_compressBlock_lazy2 ,
ZSTD_compressBlock_btlazy2 ,
ZSTD_compressBlock_btopt ,
ZSTD_compressBlock_btultra ,
ZSTD_compressBlock_btultra2 } ,
{ ZSTD_compressBlock_fast_extDict /* default for 0 */ ,
ZSTD_compressBlock_fast_extDict ,
ZSTD_compressBlock_doubleFast_extDict ,
ZSTD_compressBlock_greedy_extDict ,
ZSTD_compressBlock_lazy_extDict ,
ZSTD_compressBlock_lazy2_extDict ,
ZSTD_compressBlock_btlazy2_extDict ,
ZSTD_compressBlock_btopt_extDict ,
ZSTD_compressBlock_btultra_extDict ,
ZSTD_compressBlock_btultra_extDict } ,
{ ZSTD_compressBlock_fast_dictMatchState /* default for 0 */ ,
ZSTD_compressBlock_fast_dictMatchState ,
ZSTD_compressBlock_doubleFast_dictMatchState ,
ZSTD_compressBlock_greedy_dictMatchState ,
ZSTD_compressBlock_lazy_dictMatchState ,
ZSTD_compressBlock_lazy2_dictMatchState ,
ZSTD_compressBlock_btlazy2_dictMatchState ,
ZSTD_compressBlock_btopt_dictMatchState ,
ZSTD_compressBlock_btultra_dictMatchState ,
2020-12-18 14:46:27 +00:00
ZSTD_compressBlock_btultra_dictMatchState } ,
{ NULL /* default for 0 */ ,
NULL ,
NULL ,
ZSTD_compressBlock_greedy_dedicatedDictSearch ,
ZSTD_compressBlock_lazy_dedicatedDictSearch ,
ZSTD_compressBlock_lazy2_dedicatedDictSearch ,
NULL ,
NULL ,
NULL ,
NULL }
2020-02-08 13:56:59 +00:00
} ;
ZSTD_blockCompressor selectedCompressor ;
ZSTD_STATIC_ASSERT ( ( unsigned ) ZSTD_fast = = 1 ) ;
assert ( ZSTD_cParam_withinBounds ( ZSTD_c_strategy , strat ) ) ;
2021-05-14 22:48:15 +00:00
DEBUGLOG ( 4 , " Selected block compressor: dictMode=%d strat=%d rowMatchfinder=%d " , ( int ) dictMode , ( int ) strat , ( int ) useRowMatchFinder ) ;
if ( ZSTD_rowMatchFinderUsed ( strat , useRowMatchFinder ) ) {
static const ZSTD_blockCompressor rowBasedBlockCompressors [ 4 ] [ 3 ] = {
{ ZSTD_compressBlock_greedy_row ,
ZSTD_compressBlock_lazy_row ,
ZSTD_compressBlock_lazy2_row } ,
{ ZSTD_compressBlock_greedy_extDict_row ,
ZSTD_compressBlock_lazy_extDict_row ,
ZSTD_compressBlock_lazy2_extDict_row } ,
{ ZSTD_compressBlock_greedy_dictMatchState_row ,
ZSTD_compressBlock_lazy_dictMatchState_row ,
ZSTD_compressBlock_lazy2_dictMatchState_row } ,
{ ZSTD_compressBlock_greedy_dedicatedDictSearch_row ,
ZSTD_compressBlock_lazy_dedicatedDictSearch_row ,
ZSTD_compressBlock_lazy2_dedicatedDictSearch_row }
} ;
DEBUGLOG ( 4 , " Selecting a row-based matchfinder " ) ;
2021-12-21 11:21:29 +00:00
assert ( useRowMatchFinder ! = ZSTD_ps_auto ) ;
2021-05-14 22:48:15 +00:00
selectedCompressor = rowBasedBlockCompressors [ ( int ) dictMode ] [ ( int ) strat - ( int ) ZSTD_greedy ] ;
} else {
selectedCompressor = blockCompressor [ ( int ) dictMode ] [ ( int ) strat ] ;
}
2020-02-08 13:56:59 +00:00
assert ( selectedCompressor ! = NULL ) ;
return selectedCompressor ;
}
static void ZSTD_storeLastLiterals ( seqStore_t * seqStorePtr ,
const BYTE * anchor , size_t lastLLSize )
{
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( seqStorePtr - > lit , anchor , lastLLSize ) ;
2020-02-08 13:56:59 +00:00
seqStorePtr - > lit + = lastLLSize ;
}
void ZSTD_resetSeqStore ( seqStore_t * ssPtr )
{
ssPtr - > lit = ssPtr - > litStart ;
ssPtr - > sequences = ssPtr - > sequencesStart ;
2021-05-14 22:48:15 +00:00
ssPtr - > longLengthType = ZSTD_llt_none ;
2020-02-08 13:56:59 +00:00
}
2023-02-10 19:25:40 +00:00
/* ZSTD_postProcessSequenceProducerResult() :
* Validates and post - processes sequences obtained through the external matchfinder API :
* - Checks whether nbExternalSeqs represents an error condition .
* - Appends a block delimiter to outSeqs if one is not already present .
* See zstd . h for context regarding block delimiters .
* Returns the number of sequences after post - processing , or an error code . */
static size_t ZSTD_postProcessSequenceProducerResult (
ZSTD_Sequence * outSeqs , size_t nbExternalSeqs , size_t outSeqsCapacity , size_t srcSize
) {
RETURN_ERROR_IF (
nbExternalSeqs > outSeqsCapacity ,
sequenceProducer_failed ,
" External sequence producer returned error code %lu " ,
( unsigned long ) nbExternalSeqs
) ;
RETURN_ERROR_IF (
nbExternalSeqs = = 0 & & srcSize > 0 ,
sequenceProducer_failed ,
" Got zero sequences from external sequence producer for a non-empty src buffer! "
) ;
if ( srcSize = = 0 ) {
ZSTD_memset ( & outSeqs [ 0 ] , 0 , sizeof ( ZSTD_Sequence ) ) ;
return 1 ;
}
{
ZSTD_Sequence const lastSeq = outSeqs [ nbExternalSeqs - 1 ] ;
/* We can return early if lastSeq is already a block delimiter. */
if ( lastSeq . offset = = 0 & & lastSeq . matchLength = = 0 ) {
return nbExternalSeqs ;
}
/* This error condition is only possible if the external matchfinder
* produced an invalid parse , by definition of ZSTD_sequenceBound ( ) . */
RETURN_ERROR_IF (
nbExternalSeqs = = outSeqsCapacity ,
sequenceProducer_failed ,
" nbExternalSeqs == outSeqsCapacity but lastSeq is not a block delimiter! "
) ;
/* lastSeq is not a block delimiter, so we need to append one. */
ZSTD_memset ( & outSeqs [ nbExternalSeqs ] , 0 , sizeof ( ZSTD_Sequence ) ) ;
return nbExternalSeqs + 1 ;
}
}
/* ZSTD_fastSequenceLengthSum() :
* Returns sum ( litLen ) + sum ( matchLen ) + lastLits for * seqBuf * .
* Similar to another function in zstd_compress . c ( determine_blockSize ) ,
* except it doesn ' t check for a block delimiter to end summation .
* Removing the early exit allows the compiler to auto - vectorize ( https : //godbolt.org/z/cY1cajz9P).
* This function can be deleted and replaced by determine_blockSize after we resolve issue # 3456. */
static size_t ZSTD_fastSequenceLengthSum ( ZSTD_Sequence const * seqBuf , size_t seqBufSize ) {
size_t matchLenSum , litLenSum , i ;
matchLenSum = 0 ;
litLenSum = 0 ;
for ( i = 0 ; i < seqBufSize ; i + + ) {
litLenSum + = seqBuf [ i ] . litLength ;
matchLenSum + = seqBuf [ i ] . matchLength ;
}
return litLenSum + matchLenSum ;
}
2020-02-08 13:56:59 +00:00
typedef enum { ZSTDbss_compress , ZSTDbss_noCompress } ZSTD_buildSeqStore_e ;
static size_t ZSTD_buildSeqStore ( ZSTD_CCtx * zc , const void * src , size_t srcSize )
{
ZSTD_matchState_t * const ms = & zc - > blockState . matchState ;
DEBUGLOG ( 5 , " ZSTD_buildSeqStore (srcSize=%zu) " , srcSize ) ;
assert ( srcSize < = ZSTD_BLOCKSIZE_MAX ) ;
/* Assert that we have correctly flushed the ctx params into the ms's copy */
ZSTD_assertEqualCParams ( zc - > appliedParams . cParams , ms - > cParams ) ;
2023-02-10 19:25:40 +00:00
/* TODO: See 3090. We reduced MIN_CBLOCK_SIZE from 3 to 2 so to compensate we are adding
* additional 1. We need to revisit and change this logic to be more consistent */
if ( srcSize < MIN_CBLOCK_SIZE + ZSTD_blockHeaderSize + 1 + 1 ) {
2020-12-18 14:46:27 +00:00
if ( zc - > appliedParams . cParams . strategy > = ZSTD_btopt ) {
ZSTD_ldm_skipRawSeqStoreBytes ( & zc - > externSeqStore , srcSize ) ;
} else {
ZSTD_ldm_skipSequences ( & zc - > externSeqStore , srcSize , zc - > appliedParams . cParams . minMatch ) ;
}
2020-02-08 13:56:59 +00:00
return ZSTDbss_noCompress ; /* don't even attempt compression below a certain srcSize */
}
ZSTD_resetSeqStore ( & ( zc - > seqStore ) ) ;
/* required for optimal parser to read stats from dictionary */
ms - > opt . symbolCosts = & zc - > blockState . prevCBlock - > entropy ;
/* tell the optimal parser how we expect to compress literals */
ms - > opt . literalCompressionMode = zc - > appliedParams . literalCompressionMode ;
/* a gap between an attached dict and the current window is not safe,
* they must remain adjacent ,
* and when that stops being the case , the dict must be unset */
assert ( ms - > dictMatchState = = NULL | | ms - > loadedDictEnd = = ms - > window . dictLimit ) ;
/* limited update after a very long match */
{ const BYTE * const base = ms - > window . base ;
const BYTE * const istart = ( const BYTE * ) src ;
2020-12-18 14:46:27 +00:00
const U32 curr = ( U32 ) ( istart - base ) ;
2020-02-08 13:56:59 +00:00
if ( sizeof ( ptrdiff_t ) = = 8 ) assert ( istart - base < ( ptrdiff_t ) ( U32 ) ( - 1 ) ) ; /* ensure no overflow */
2020-12-18 14:46:27 +00:00
if ( curr > ms - > nextToUpdate + 384 )
ms - > nextToUpdate = curr - MIN ( 192 , ( U32 ) ( curr - ms - > nextToUpdate - 384 ) ) ;
2020-02-08 13:56:59 +00:00
}
/* select and store sequences */
{ ZSTD_dictMode_e const dictMode = ZSTD_matchState_dictMode ( ms ) ;
size_t lastLLSize ;
{ int i ;
for ( i = 0 ; i < ZSTD_REP_NUM ; + + i )
zc - > blockState . nextCBlock - > rep [ i ] = zc - > blockState . prevCBlock - > rep [ i ] ;
}
if ( zc - > externSeqStore . pos < zc - > externSeqStore . size ) {
2021-12-21 11:21:29 +00:00
assert ( zc - > appliedParams . ldmParams . enableLdm = = ZSTD_ps_disable ) ;
2023-02-10 19:25:40 +00:00
/* External matchfinder + LDM is technically possible, just not implemented yet.
* We need to revisit soon and implement it . */
RETURN_ERROR_IF (
zc - > appliedParams . useSequenceProducer ,
parameter_combination_unsupported ,
" Long-distance matching with external sequence producer enabled is not currently supported. "
) ;
2020-02-08 13:56:59 +00:00
/* Updates ldmSeqStore.pos */
lastLLSize =
ZSTD_ldm_blockCompress ( & zc - > externSeqStore ,
ms , & zc - > seqStore ,
zc - > blockState . nextCBlock - > rep ,
2021-05-14 22:48:15 +00:00
zc - > appliedParams . useRowMatchFinder ,
2020-02-08 13:56:59 +00:00
src , srcSize ) ;
assert ( zc - > externSeqStore . pos < = zc - > externSeqStore . size ) ;
2021-12-21 11:21:29 +00:00
} else if ( zc - > appliedParams . ldmParams . enableLdm = = ZSTD_ps_enable ) {
2020-12-18 14:46:27 +00:00
rawSeqStore_t ldmSeqStore = kNullRawSeqStore ;
2020-02-08 13:56:59 +00:00
2023-02-10 19:25:40 +00:00
/* External matchfinder + LDM is technically possible, just not implemented yet.
* We need to revisit soon and implement it . */
RETURN_ERROR_IF (
zc - > appliedParams . useSequenceProducer ,
parameter_combination_unsupported ,
" Long-distance matching with external sequence producer enabled is not currently supported. "
) ;
2020-02-08 13:56:59 +00:00
ldmSeqStore . seq = zc - > ldmSequences ;
ldmSeqStore . capacity = zc - > maxNbLdmSequences ;
/* Updates ldmSeqStore.size */
FORWARD_IF_ERROR ( ZSTD_ldm_generateSequences ( & zc - > ldmState , & ldmSeqStore ,
& zc - > appliedParams . ldmParams ,
2020-06-04 19:03:27 +00:00
src , srcSize ) , " " ) ;
2020-02-08 13:56:59 +00:00
/* Updates ldmSeqStore.pos */
lastLLSize =
ZSTD_ldm_blockCompress ( & ldmSeqStore ,
ms , & zc - > seqStore ,
zc - > blockState . nextCBlock - > rep ,
2021-05-14 22:48:15 +00:00
zc - > appliedParams . useRowMatchFinder ,
2020-02-08 13:56:59 +00:00
src , srcSize ) ;
assert ( ldmSeqStore . pos = = ldmSeqStore . size ) ;
2023-02-10 19:25:40 +00:00
} else if ( zc - > appliedParams . useSequenceProducer ) {
assert (
zc - > externalMatchCtx . seqBufferCapacity > = ZSTD_sequenceBound ( srcSize )
) ;
assert ( zc - > externalMatchCtx . mFinder ! = NULL ) ;
{ U32 const windowSize = ( U32 ) 1 < < zc - > appliedParams . cParams . windowLog ;
size_t const nbExternalSeqs = ( zc - > externalMatchCtx . mFinder ) (
zc - > externalMatchCtx . mState ,
zc - > externalMatchCtx . seqBuffer ,
zc - > externalMatchCtx . seqBufferCapacity ,
src , srcSize ,
NULL , 0 , /* dict and dictSize, currently not supported */
zc - > appliedParams . compressionLevel ,
windowSize
) ;
size_t const nbPostProcessedSeqs = ZSTD_postProcessSequenceProducerResult (
zc - > externalMatchCtx . seqBuffer ,
nbExternalSeqs ,
zc - > externalMatchCtx . seqBufferCapacity ,
srcSize
) ;
/* Return early if there is no error, since we don't need to worry about last literals */
if ( ! ZSTD_isError ( nbPostProcessedSeqs ) ) {
ZSTD_sequencePosition seqPos = { 0 , 0 , 0 } ;
size_t const seqLenSum = ZSTD_fastSequenceLengthSum ( zc - > externalMatchCtx . seqBuffer , nbPostProcessedSeqs ) ;
RETURN_ERROR_IF ( seqLenSum > srcSize , externalSequences_invalid , " External sequences imply too large a block! " ) ;
FORWARD_IF_ERROR (
ZSTD_copySequencesToSeqStoreExplicitBlockDelim (
zc , & seqPos ,
zc - > externalMatchCtx . seqBuffer , nbPostProcessedSeqs ,
src , srcSize ,
zc - > appliedParams . searchForExternalRepcodes
) ,
" Failed to copy external sequences to seqStore! "
) ;
ms - > ldmSeqStore = NULL ;
DEBUGLOG ( 5 , " Copied %lu sequences from external sequence producer to internal seqStore. " , ( unsigned long ) nbExternalSeqs ) ;
return ZSTDbss_compress ;
}
/* Propagate the error if fallback is disabled */
if ( ! zc - > appliedParams . enableMatchFinderFallback ) {
return nbPostProcessedSeqs ;
}
/* Fallback to software matchfinder */
{ ZSTD_blockCompressor const blockCompressor = ZSTD_selectBlockCompressor ( zc - > appliedParams . cParams . strategy ,
zc - > appliedParams . useRowMatchFinder ,
dictMode ) ;
ms - > ldmSeqStore = NULL ;
DEBUGLOG (
5 ,
" External sequence producer returned error code %lu. Falling back to internal parser. " ,
( unsigned long ) nbExternalSeqs
) ;
lastLLSize = blockCompressor ( ms , & zc - > seqStore , zc - > blockState . nextCBlock - > rep , src , srcSize ) ;
} }
} else { /* not long range mode and no external matchfinder */
2021-05-14 22:48:15 +00:00
ZSTD_blockCompressor const blockCompressor = ZSTD_selectBlockCompressor ( zc - > appliedParams . cParams . strategy ,
zc - > appliedParams . useRowMatchFinder ,
dictMode ) ;
2020-12-18 14:46:27 +00:00
ms - > ldmSeqStore = NULL ;
2020-02-08 13:56:59 +00:00
lastLLSize = blockCompressor ( ms , & zc - > seqStore , zc - > blockState . nextCBlock - > rep , src , srcSize ) ;
}
{ const BYTE * const lastLiterals = ( const BYTE * ) src + srcSize - lastLLSize ;
ZSTD_storeLastLiterals ( & zc - > seqStore , lastLiterals , lastLLSize ) ;
} }
return ZSTDbss_compress ;
}
static void ZSTD_copyBlockSequences ( ZSTD_CCtx * zc )
{
const seqStore_t * seqStore = ZSTD_getSeqStore ( zc ) ;
2020-12-18 14:46:27 +00:00
const seqDef * seqStoreSeqs = seqStore - > sequencesStart ;
size_t seqStoreSeqSize = seqStore - > sequences - seqStoreSeqs ;
size_t seqStoreLiteralsSize = ( size_t ) ( seqStore - > lit - seqStore - > litStart ) ;
size_t literalsRead = 0 ;
size_t lastLLSize ;
2020-02-08 13:56:59 +00:00
ZSTD_Sequence * outSeqs = & zc - > seqCollector . seqStart [ zc - > seqCollector . seqIndex ] ;
2020-12-18 14:46:27 +00:00
size_t i ;
repcodes_t updatedRepcodes ;
2020-02-08 13:56:59 +00:00
assert ( zc - > seqCollector . seqIndex + 1 < zc - > seqCollector . maxSequences ) ;
2020-12-18 14:46:27 +00:00
/* Ensure we have enough space for last literals "sequence" */
assert ( zc - > seqCollector . maxSequences > = seqStoreSeqSize + 1 ) ;
ZSTD_memcpy ( updatedRepcodes . rep , zc - > blockState . prevCBlock - > rep , sizeof ( repcodes_t ) ) ;
for ( i = 0 ; i < seqStoreSeqSize ; + + i ) {
2022-03-18 00:52:11 +00:00
U32 rawOffset = seqStoreSeqs [ i ] . offBase - ZSTD_REP_NUM ;
2020-12-18 14:46:27 +00:00
outSeqs [ i ] . litLength = seqStoreSeqs [ i ] . litLength ;
2022-03-18 00:52:11 +00:00
outSeqs [ i ] . matchLength = seqStoreSeqs [ i ] . mlBase + MINMATCH ;
2020-12-18 14:46:27 +00:00
outSeqs [ i ] . rep = 0 ;
2020-02-08 13:56:59 +00:00
if ( i = = seqStore - > longLengthPos ) {
2021-05-14 22:48:15 +00:00
if ( seqStore - > longLengthType = = ZSTD_llt_literalLength ) {
2020-02-08 13:56:59 +00:00
outSeqs [ i ] . litLength + = 0x10000 ;
2021-05-14 22:48:15 +00:00
} else if ( seqStore - > longLengthType = = ZSTD_llt_matchLength ) {
2020-02-08 13:56:59 +00:00
outSeqs [ i ] . matchLength + = 0x10000 ;
}
}
2022-03-18 00:52:11 +00:00
if ( seqStoreSeqs [ i ] . offBase < = ZSTD_REP_NUM ) {
2020-12-18 14:46:27 +00:00
/* Derive the correct offset corresponding to a repcode */
2022-03-18 00:52:11 +00:00
outSeqs [ i ] . rep = seqStoreSeqs [ i ] . offBase ;
2020-12-18 14:46:27 +00:00
if ( outSeqs [ i ] . litLength ! = 0 ) {
rawOffset = updatedRepcodes . rep [ outSeqs [ i ] . rep - 1 ] ;
} else {
if ( outSeqs [ i ] . rep = = 3 ) {
rawOffset = updatedRepcodes . rep [ 0 ] - 1 ;
2020-02-08 13:56:59 +00:00
} else {
2020-12-18 14:46:27 +00:00
rawOffset = updatedRepcodes . rep [ outSeqs [ i ] . rep ] ;
2020-02-08 13:56:59 +00:00
}
}
}
2020-12-18 14:46:27 +00:00
outSeqs [ i ] . offset = rawOffset ;
/* seqStoreSeqs[i].offset == offCode+1, and ZSTD_updateRep() expects offCode
so we provide seqStoreSeqs [ i ] . offset - 1 */
2022-03-18 00:52:11 +00:00
ZSTD_updateRep ( updatedRepcodes . rep ,
2023-02-10 19:25:40 +00:00
seqStoreSeqs [ i ] . offBase ,
2022-03-18 00:52:11 +00:00
seqStoreSeqs [ i ] . litLength = = 0 ) ;
2020-12-18 14:46:27 +00:00
literalsRead + = outSeqs [ i ] . litLength ;
2020-02-08 13:56:59 +00:00
}
2020-12-18 14:46:27 +00:00
/* Insert last literals (if any exist) in the block as a sequence with ml == off == 0.
* If there are no last literals , then we ' ll emit ( of : 0 , ml : 0 , ll : 0 ) , which is a marker
* for the block boundary , according to the API .
*/
assert ( seqStoreLiteralsSize > = literalsRead ) ;
lastLLSize = seqStoreLiteralsSize - literalsRead ;
outSeqs [ i ] . litLength = ( U32 ) lastLLSize ;
outSeqs [ i ] . matchLength = outSeqs [ i ] . offset = outSeqs [ i ] . rep = 0 ;
seqStoreSeqSize + + ;
zc - > seqCollector . seqIndex + = seqStoreSeqSize ;
2020-02-08 13:56:59 +00:00
}
2023-02-10 19:25:40 +00:00
size_t ZSTD_sequenceBound ( size_t srcSize ) {
return ( srcSize / ZSTD_MINMATCH_MIN ) + 1 ;
}
2020-12-18 14:46:27 +00:00
size_t ZSTD_generateSequences ( ZSTD_CCtx * zc , ZSTD_Sequence * outSeqs ,
size_t outSeqsSize , const void * src , size_t srcSize )
2020-02-08 13:56:59 +00:00
{
const size_t dstCapacity = ZSTD_compressBound ( srcSize ) ;
2020-12-18 14:46:27 +00:00
void * dst = ZSTD_customMalloc ( dstCapacity , ZSTD_defaultCMem ) ;
2020-02-08 13:56:59 +00:00
SeqCollector seqCollector ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( dst = = NULL , memory_allocation , " NULL pointer! " ) ;
2020-02-08 13:56:59 +00:00
seqCollector . collectSequences = 1 ;
seqCollector . seqStart = outSeqs ;
seqCollector . seqIndex = 0 ;
seqCollector . maxSequences = outSeqsSize ;
zc - > seqCollector = seqCollector ;
ZSTD_compress2 ( zc , dst , dstCapacity , src , srcSize ) ;
2020-12-18 14:46:27 +00:00
ZSTD_customFree ( dst , ZSTD_defaultCMem ) ;
2020-02-08 13:56:59 +00:00
return zc - > seqCollector . seqIndex ;
}
2020-12-18 14:46:27 +00:00
size_t ZSTD_mergeBlockDelimiters ( ZSTD_Sequence * sequences , size_t seqsSize ) {
size_t in = 0 ;
size_t out = 0 ;
for ( ; in < seqsSize ; + + in ) {
if ( sequences [ in ] . offset = = 0 & & sequences [ in ] . matchLength = = 0 ) {
if ( in ! = seqsSize - 1 ) {
sequences [ in + 1 ] . litLength + = sequences [ in ] . litLength ;
}
} else {
sequences [ out ] = sequences [ in ] ;
+ + out ;
}
}
return out ;
}
/* Unrolled loop to read four size_ts of input at a time. Returns 1 if is RLE, 0 if not. */
static int ZSTD_isRLE ( const BYTE * src , size_t length ) {
const BYTE * ip = src ;
const BYTE value = ip [ 0 ] ;
const size_t valueST = ( size_t ) ( ( U64 ) value * 0x0101010101010101ULL ) ;
const size_t unrollSize = sizeof ( size_t ) * 4 ;
const size_t unrollMask = unrollSize - 1 ;
const size_t prefixLength = length & unrollMask ;
2020-02-08 13:56:59 +00:00
size_t i ;
2020-12-18 14:46:27 +00:00
if ( length = = 1 ) return 1 ;
/* Check if prefix is RLE first before using unrolled loop */
if ( prefixLength & & ZSTD_count ( ip + 1 , ip , ip + prefixLength ) ! = prefixLength - 1 ) {
return 0 ;
}
for ( i = prefixLength ; i ! = length ; i + = unrollSize ) {
2023-02-10 19:25:40 +00:00
size_t u ;
2020-12-18 14:46:27 +00:00
for ( u = 0 ; u < unrollSize ; u + = sizeof ( size_t ) ) {
if ( MEM_readST ( ip + i + u ) ! = valueST ) {
return 0 ;
2023-02-10 19:25:40 +00:00
} } }
2020-02-08 13:56:59 +00:00
return 1 ;
}
2020-06-04 19:03:27 +00:00
/* Returns true if the given block may be RLE.
* This is just a heuristic based on the compressibility .
* It may return both false positives and false negatives .
*/
static int ZSTD_maybeRLE ( seqStore_t const * seqStore )
{
size_t const nbSeqs = ( size_t ) ( seqStore - > sequences - seqStore - > sequencesStart ) ;
size_t const nbLits = ( size_t ) ( seqStore - > lit - seqStore - > litStart ) ;
return nbSeqs < 4 & & nbLits < 10 ;
}
2023-02-10 19:25:40 +00:00
static void
ZSTD_blockState_confirmRepcodesAndEntropyTables ( ZSTD_blockState_t * const bs )
2020-02-08 13:56:59 +00:00
{
2021-05-14 22:48:15 +00:00
ZSTD_compressedBlockState_t * const tmp = bs - > prevCBlock ;
bs - > prevCBlock = bs - > nextCBlock ;
bs - > nextCBlock = tmp ;
}
/* Writes the block header */
2023-02-10 19:25:40 +00:00
static void
writeBlockHeader ( void * op , size_t cSize , size_t blockSize , U32 lastBlock )
{
2021-05-14 22:48:15 +00:00
U32 const cBlockHeader = cSize = = 1 ?
lastBlock + ( ( ( U32 ) bt_rle ) < < 1 ) + ( U32 ) ( blockSize < < 3 ) :
lastBlock + ( ( ( U32 ) bt_compressed ) < < 1 ) + ( U32 ) ( cSize < < 3 ) ;
MEM_writeLE24 ( op , cBlockHeader ) ;
DEBUGLOG ( 3 , " writeBlockHeader: cSize: %zu blockSize: %zu lastBlock: %u " , cSize , blockSize , lastBlock ) ;
}
/** ZSTD_buildBlockEntropyStats_literals() :
* Builds entropy for the literals .
* Stores literals block type ( raw , rle , compressed , repeat ) and
* huffman description table to hufMetadata .
* Requires ENTROPY_WORKSPACE_SIZE workspace
2023-02-10 19:25:40 +00:00
* @ return : size of huffman description table , or an error code
*/
static size_t
ZSTD_buildBlockEntropyStats_literals ( void * const src , size_t srcSize ,
const ZSTD_hufCTables_t * prevHuf ,
ZSTD_hufCTables_t * nextHuf ,
ZSTD_hufCTablesMetadata_t * hufMetadata ,
const int literalsCompressionIsDisabled ,
void * workspace , size_t wkspSize ,
int hufFlags )
2021-05-14 22:48:15 +00:00
{
BYTE * const wkspStart = ( BYTE * ) workspace ;
BYTE * const wkspEnd = wkspStart + wkspSize ;
BYTE * const countWkspStart = wkspStart ;
unsigned * const countWksp = ( unsigned * ) workspace ;
const size_t countWkspSize = ( HUF_SYMBOLVALUE_MAX + 1 ) * sizeof ( unsigned ) ;
BYTE * const nodeWksp = countWkspStart + countWkspSize ;
2023-02-10 19:25:40 +00:00
const size_t nodeWkspSize = ( size_t ) ( wkspEnd - nodeWksp ) ;
2021-05-14 22:48:15 +00:00
unsigned maxSymbolValue = HUF_SYMBOLVALUE_MAX ;
2023-02-10 19:25:40 +00:00
unsigned huffLog = LitHufLog ;
2021-05-14 22:48:15 +00:00
HUF_repeat repeat = prevHuf - > repeatMode ;
DEBUGLOG ( 5 , " ZSTD_buildBlockEntropyStats_literals (srcSize=%zu) " , srcSize ) ;
/* Prepare nextEntropy assuming reusing the existing table */
ZSTD_memcpy ( nextHuf , prevHuf , sizeof ( * prevHuf ) ) ;
2021-12-21 11:21:29 +00:00
if ( literalsCompressionIsDisabled ) {
2021-05-14 22:48:15 +00:00
DEBUGLOG ( 5 , " set_basic - disabled " ) ;
hufMetadata - > hType = set_basic ;
return 0 ;
}
/* small ? don't even attempt compression (speed opt) */
# ifndef COMPRESS_LITERALS_SIZE_MIN
2023-02-10 19:25:40 +00:00
# define COMPRESS_LITERALS_SIZE_MIN 63 /* heuristic */
2021-05-14 22:48:15 +00:00
# endif
{ size_t const minLitSize = ( prevHuf - > repeatMode = = HUF_repeat_valid ) ? 6 : COMPRESS_LITERALS_SIZE_MIN ;
if ( srcSize < = minLitSize ) {
DEBUGLOG ( 5 , " set_basic - too small " ) ;
hufMetadata - > hType = set_basic ;
return 0 ;
2023-02-10 19:25:40 +00:00
} }
2021-05-14 22:48:15 +00:00
/* Scan input and build symbol stats */
2023-02-10 19:25:40 +00:00
{ size_t const largest =
HIST_count_wksp ( countWksp , & maxSymbolValue ,
( const BYTE * ) src , srcSize ,
workspace , wkspSize ) ;
2021-05-14 22:48:15 +00:00
FORWARD_IF_ERROR ( largest , " HIST_count_wksp failed " ) ;
if ( largest = = srcSize ) {
2023-02-10 19:25:40 +00:00
/* only one literal symbol */
2021-05-14 22:48:15 +00:00
DEBUGLOG ( 5 , " set_rle " ) ;
hufMetadata - > hType = set_rle ;
return 0 ;
}
if ( largest < = ( srcSize > > 7 ) + 4 ) {
2023-02-10 19:25:40 +00:00
/* heuristic: likely not compressible */
2021-05-14 22:48:15 +00:00
DEBUGLOG ( 5 , " set_basic - no gain " ) ;
hufMetadata - > hType = set_basic ;
return 0 ;
2023-02-10 19:25:40 +00:00
} }
2021-05-14 22:48:15 +00:00
/* Validate the previous Huffman table */
2023-02-10 19:25:40 +00:00
if ( repeat = = HUF_repeat_check
& & ! HUF_validateCTable ( ( HUF_CElt const * ) prevHuf - > CTable , countWksp , maxSymbolValue ) ) {
2021-05-14 22:48:15 +00:00
repeat = HUF_repeat_none ;
}
/* Build Huffman Tree */
ZSTD_memset ( nextHuf - > CTable , 0 , sizeof ( nextHuf - > CTable ) ) ;
2023-02-10 19:25:40 +00:00
huffLog = HUF_optimalTableLog ( huffLog , srcSize , maxSymbolValue , nodeWksp , nodeWkspSize , nextHuf - > CTable , countWksp , hufFlags ) ;
assert ( huffLog < = LitHufLog ) ;
2021-05-14 22:48:15 +00:00
{ size_t const maxBits = HUF_buildCTable_wksp ( ( HUF_CElt * ) nextHuf - > CTable , countWksp ,
maxSymbolValue , huffLog ,
nodeWksp , nodeWkspSize ) ;
FORWARD_IF_ERROR ( maxBits , " HUF_buildCTable_wksp " ) ;
huffLog = ( U32 ) maxBits ;
2023-02-10 19:25:40 +00:00
}
{ /* Build and write the CTable */
size_t const newCSize = HUF_estimateCompressedSize (
( HUF_CElt * ) nextHuf - > CTable , countWksp , maxSymbolValue ) ;
size_t const hSize = HUF_writeCTable_wksp (
hufMetadata - > hufDesBuffer , sizeof ( hufMetadata - > hufDesBuffer ) ,
( HUF_CElt * ) nextHuf - > CTable , maxSymbolValue , huffLog ,
nodeWksp , nodeWkspSize ) ;
/* Check against repeating the previous CTable */
if ( repeat ! = HUF_repeat_none ) {
size_t const oldCSize = HUF_estimateCompressedSize (
( HUF_CElt const * ) prevHuf - > CTable , countWksp , maxSymbolValue ) ;
if ( oldCSize < srcSize & & ( oldCSize < = hSize + newCSize | | hSize + 12 > = srcSize ) ) {
DEBUGLOG ( 5 , " set_repeat - smaller " ) ;
2021-05-14 22:48:15 +00:00
ZSTD_memcpy ( nextHuf , prevHuf , sizeof ( * prevHuf ) ) ;
2023-02-10 19:25:40 +00:00
hufMetadata - > hType = set_repeat ;
2021-05-14 22:48:15 +00:00
return 0 ;
2023-02-10 19:25:40 +00:00
} }
if ( newCSize + hSize > = srcSize ) {
DEBUGLOG ( 5 , " set_basic - no gains " ) ;
ZSTD_memcpy ( nextHuf , prevHuf , sizeof ( * prevHuf ) ) ;
hufMetadata - > hType = set_basic ;
return 0 ;
2021-05-14 22:48:15 +00:00
}
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " set_compressed (hSize=%u) " , ( U32 ) hSize ) ;
hufMetadata - > hType = set_compressed ;
nextHuf - > repeatMode = HUF_repeat_check ;
return hSize ;
2021-05-14 22:48:15 +00:00
}
}
/* ZSTD_buildDummySequencesStatistics():
* Returns a ZSTD_symbolEncodingTypeStats_t with all encoding types as set_basic ,
* and updates nextEntropy to the appropriate repeatMode .
*/
static ZSTD_symbolEncodingTypeStats_t
2023-02-10 19:25:40 +00:00
ZSTD_buildDummySequencesStatistics ( ZSTD_fseCTables_t * nextEntropy )
{
ZSTD_symbolEncodingTypeStats_t stats = { set_basic , set_basic , set_basic , 0 , 0 , 0 } ;
2021-05-14 22:48:15 +00:00
nextEntropy - > litlength_repeatMode = FSE_repeat_none ;
nextEntropy - > offcode_repeatMode = FSE_repeat_none ;
nextEntropy - > matchlength_repeatMode = FSE_repeat_none ;
return stats ;
}
/** ZSTD_buildBlockEntropyStats_sequences() :
* Builds entropy for the sequences .
* Stores symbol compression modes and fse table to fseMetadata .
* Requires ENTROPY_WORKSPACE_SIZE wksp .
2023-02-10 19:25:40 +00:00
* @ return : size of fse tables or error code */
static size_t
ZSTD_buildBlockEntropyStats_sequences (
const seqStore_t * seqStorePtr ,
const ZSTD_fseCTables_t * prevEntropy ,
ZSTD_fseCTables_t * nextEntropy ,
const ZSTD_CCtx_params * cctxParams ,
ZSTD_fseCTablesMetadata_t * fseMetadata ,
void * workspace , size_t wkspSize )
2021-05-14 22:48:15 +00:00
{
ZSTD_strategy const strategy = cctxParams - > cParams . strategy ;
2023-02-10 19:25:40 +00:00
size_t const nbSeq = ( size_t ) ( seqStorePtr - > sequences - seqStorePtr - > sequencesStart ) ;
2021-05-14 22:48:15 +00:00
BYTE * const ostart = fseMetadata - > fseTablesBuffer ;
BYTE * const oend = ostart + sizeof ( fseMetadata - > fseTablesBuffer ) ;
BYTE * op = ostart ;
unsigned * countWorkspace = ( unsigned * ) workspace ;
unsigned * entropyWorkspace = countWorkspace + ( MaxSeq + 1 ) ;
size_t entropyWorkspaceSize = wkspSize - ( MaxSeq + 1 ) * sizeof ( * countWorkspace ) ;
ZSTD_symbolEncodingTypeStats_t stats ;
DEBUGLOG ( 5 , " ZSTD_buildBlockEntropyStats_sequences (nbSeq=%zu) " , nbSeq ) ;
stats = nbSeq ! = 0 ? ZSTD_buildSequencesStatistics ( seqStorePtr , nbSeq ,
prevEntropy , nextEntropy , op , oend ,
strategy , countWorkspace ,
entropyWorkspace , entropyWorkspaceSize )
: ZSTD_buildDummySequencesStatistics ( nextEntropy ) ;
FORWARD_IF_ERROR ( stats . size , " ZSTD_buildSequencesStatistics failed! " ) ;
fseMetadata - > llType = ( symbolEncodingType_e ) stats . LLtype ;
fseMetadata - > ofType = ( symbolEncodingType_e ) stats . Offtype ;
fseMetadata - > mlType = ( symbolEncodingType_e ) stats . MLtype ;
fseMetadata - > lastCountSize = stats . lastCountSize ;
return stats . size ;
}
/** ZSTD_buildBlockEntropyStats() :
* Builds entropy for the block .
* Requires workspace size ENTROPY_WORKSPACE_SIZE
2023-02-10 19:25:40 +00:00
* @ return : 0 on success , or an error code
* Note : also employed in superblock
2021-05-14 22:48:15 +00:00
*/
2023-02-10 19:25:40 +00:00
size_t ZSTD_buildBlockEntropyStats (
const seqStore_t * seqStorePtr ,
const ZSTD_entropyCTables_t * prevEntropy ,
ZSTD_entropyCTables_t * nextEntropy ,
const ZSTD_CCtx_params * cctxParams ,
ZSTD_entropyCTablesMetadata_t * entropyMetadata ,
void * workspace , size_t wkspSize )
{
size_t const litSize = ( size_t ) ( seqStorePtr - > lit - seqStorePtr - > litStart ) ;
int const huf_useOptDepth = ( cctxParams - > cParams . strategy > = HUF_OPTIMAL_DEPTH_THRESHOLD ) ;
int const hufFlags = huf_useOptDepth ? HUF_flags_optimalDepth : 0 ;
2021-05-14 22:48:15 +00:00
entropyMetadata - > hufMetadata . hufDesSize =
ZSTD_buildBlockEntropyStats_literals ( seqStorePtr - > litStart , litSize ,
& prevEntropy - > huf , & nextEntropy - > huf ,
& entropyMetadata - > hufMetadata ,
2021-12-21 11:21:29 +00:00
ZSTD_literalsCompressionIsDisabled ( cctxParams ) ,
2023-02-10 19:25:40 +00:00
workspace , wkspSize , hufFlags ) ;
2021-05-14 22:48:15 +00:00
FORWARD_IF_ERROR ( entropyMetadata - > hufMetadata . hufDesSize , " ZSTD_buildBlockEntropyStats_literals failed " ) ;
entropyMetadata - > fseMetadata . fseTablesSize =
ZSTD_buildBlockEntropyStats_sequences ( seqStorePtr ,
& prevEntropy - > fse , & nextEntropy - > fse ,
cctxParams ,
& entropyMetadata - > fseMetadata ,
workspace , wkspSize ) ;
FORWARD_IF_ERROR ( entropyMetadata - > fseMetadata . fseTablesSize , " ZSTD_buildBlockEntropyStats_sequences failed " ) ;
return 0 ;
}
/* Returns the size estimate for the literals section (header + content) of a block */
2023-02-10 19:25:40 +00:00
static size_t
ZSTD_estimateBlockSize_literal ( const BYTE * literals , size_t litSize ,
const ZSTD_hufCTables_t * huf ,
const ZSTD_hufCTablesMetadata_t * hufMetadata ,
void * workspace , size_t wkspSize ,
int writeEntropy )
2021-05-14 22:48:15 +00:00
{
unsigned * const countWksp = ( unsigned * ) workspace ;
unsigned maxSymbolValue = HUF_SYMBOLVALUE_MAX ;
size_t literalSectionHeaderSize = 3 + ( litSize > = 1 KB ) + ( litSize > = 16 KB ) ;
U32 singleStream = litSize < 256 ;
if ( hufMetadata - > hType = = set_basic ) return litSize ;
else if ( hufMetadata - > hType = = set_rle ) return 1 ;
else if ( hufMetadata - > hType = = set_compressed | | hufMetadata - > hType = = set_repeat ) {
size_t const largest = HIST_count_wksp ( countWksp , & maxSymbolValue , ( const BYTE * ) literals , litSize , workspace , wkspSize ) ;
if ( ZSTD_isError ( largest ) ) return litSize ;
{ size_t cLitSizeEstimate = HUF_estimateCompressedSize ( ( const HUF_CElt * ) huf - > CTable , countWksp , maxSymbolValue ) ;
if ( writeEntropy ) cLitSizeEstimate + = hufMetadata - > hufDesSize ;
if ( ! singleStream ) cLitSizeEstimate + = 6 ; /* multi-stream huffman uses 6-byte jump table */
return cLitSizeEstimate + literalSectionHeaderSize ;
} }
assert ( 0 ) ; /* impossible */
return 0 ;
}
/* Returns the size estimate for the FSE-compressed symbols (of, ml, ll) of a block */
2023-02-10 19:25:40 +00:00
static size_t
ZSTD_estimateBlockSize_symbolType ( symbolEncodingType_e type ,
const BYTE * codeTable , size_t nbSeq , unsigned maxCode ,
const FSE_CTable * fseCTable ,
const U8 * additionalBits ,
short const * defaultNorm , U32 defaultNormLog , U32 defaultMax ,
void * workspace , size_t wkspSize )
2021-05-14 22:48:15 +00:00
{
unsigned * const countWksp = ( unsigned * ) workspace ;
const BYTE * ctp = codeTable ;
const BYTE * const ctStart = ctp ;
const BYTE * const ctEnd = ctStart + nbSeq ;
size_t cSymbolTypeSizeEstimateInBits = 0 ;
unsigned max = maxCode ;
HIST_countFast_wksp ( countWksp , & max , codeTable , nbSeq , workspace , wkspSize ) ; /* can't fail */
if ( type = = set_basic ) {
/* We selected this encoding type, so it must be valid. */
assert ( max < = defaultMax ) ;
( void ) defaultMax ;
cSymbolTypeSizeEstimateInBits = ZSTD_crossEntropyCost ( defaultNorm , defaultNormLog , countWksp , max ) ;
} else if ( type = = set_rle ) {
cSymbolTypeSizeEstimateInBits = 0 ;
} else if ( type = = set_compressed | | type = = set_repeat ) {
cSymbolTypeSizeEstimateInBits = ZSTD_fseBitCost ( fseCTable , countWksp , max ) ;
}
if ( ZSTD_isError ( cSymbolTypeSizeEstimateInBits ) ) {
return nbSeq * 10 ;
}
while ( ctp < ctEnd ) {
if ( additionalBits ) cSymbolTypeSizeEstimateInBits + = additionalBits [ * ctp ] ;
else cSymbolTypeSizeEstimateInBits + = * ctp ; /* for offset, offset code is also the number of additional bits */
ctp + + ;
}
return cSymbolTypeSizeEstimateInBits > > 3 ;
}
/* Returns the size estimate for the sequences section (header + content) of a block */
2023-02-10 19:25:40 +00:00
static size_t
ZSTD_estimateBlockSize_sequences ( const BYTE * ofCodeTable ,
const BYTE * llCodeTable ,
const BYTE * mlCodeTable ,
size_t nbSeq ,
const ZSTD_fseCTables_t * fseTables ,
const ZSTD_fseCTablesMetadata_t * fseMetadata ,
void * workspace , size_t wkspSize ,
int writeEntropy )
2021-05-14 22:48:15 +00:00
{
size_t sequencesSectionHeaderSize = 1 /* seqHead */ + 1 /* min seqSize size */ + ( nbSeq > = 128 ) + ( nbSeq > = LONGNBSEQ ) ;
size_t cSeqSizeEstimate = 0 ;
cSeqSizeEstimate + = ZSTD_estimateBlockSize_symbolType ( fseMetadata - > ofType , ofCodeTable , nbSeq , MaxOff ,
2023-02-10 19:25:40 +00:00
fseTables - > offcodeCTable , NULL ,
OF_defaultNorm , OF_defaultNormLog , DefaultMaxOff ,
workspace , wkspSize ) ;
2021-05-14 22:48:15 +00:00
cSeqSizeEstimate + = ZSTD_estimateBlockSize_symbolType ( fseMetadata - > llType , llCodeTable , nbSeq , MaxLL ,
2023-02-10 19:25:40 +00:00
fseTables - > litlengthCTable , LL_bits ,
LL_defaultNorm , LL_defaultNormLog , MaxLL ,
workspace , wkspSize ) ;
2021-05-14 22:48:15 +00:00
cSeqSizeEstimate + = ZSTD_estimateBlockSize_symbolType ( fseMetadata - > mlType , mlCodeTable , nbSeq , MaxML ,
2023-02-10 19:25:40 +00:00
fseTables - > matchlengthCTable , ML_bits ,
ML_defaultNorm , ML_defaultNormLog , MaxML ,
workspace , wkspSize ) ;
2021-05-14 22:48:15 +00:00
if ( writeEntropy ) cSeqSizeEstimate + = fseMetadata - > fseTablesSize ;
return cSeqSizeEstimate + sequencesSectionHeaderSize ;
}
/* Returns the size estimate for a given stream of literals, of, ll, ml */
2023-02-10 19:25:40 +00:00
static size_t
ZSTD_estimateBlockSize ( const BYTE * literals , size_t litSize ,
const BYTE * ofCodeTable ,
const BYTE * llCodeTable ,
const BYTE * mlCodeTable ,
size_t nbSeq ,
const ZSTD_entropyCTables_t * entropy ,
const ZSTD_entropyCTablesMetadata_t * entropyMetadata ,
void * workspace , size_t wkspSize ,
int writeLitEntropy , int writeSeqEntropy )
{
2021-05-14 22:48:15 +00:00
size_t const literalsSize = ZSTD_estimateBlockSize_literal ( literals , litSize ,
2023-02-10 19:25:40 +00:00
& entropy - > huf , & entropyMetadata - > hufMetadata ,
workspace , wkspSize , writeLitEntropy ) ;
2021-05-14 22:48:15 +00:00
size_t const seqSize = ZSTD_estimateBlockSize_sequences ( ofCodeTable , llCodeTable , mlCodeTable ,
2023-02-10 19:25:40 +00:00
nbSeq , & entropy - > fse , & entropyMetadata - > fseMetadata ,
workspace , wkspSize , writeSeqEntropy ) ;
2021-05-14 22:48:15 +00:00
return seqSize + literalsSize + ZSTD_blockHeaderSize ;
}
/* Builds entropy statistics and uses them for blocksize estimation.
*
2023-02-10 19:25:40 +00:00
* @ return : estimated compressed size of the seqStore , or a zstd error .
2021-05-14 22:48:15 +00:00
*/
2023-02-10 19:25:40 +00:00
static size_t
ZSTD_buildEntropyStatisticsAndEstimateSubBlockSize ( seqStore_t * seqStore , ZSTD_CCtx * zc )
{
ZSTD_entropyCTablesMetadata_t * const entropyMetadata = & zc - > blockSplitCtx . entropyMetadata ;
2021-12-21 11:21:29 +00:00
DEBUGLOG ( 6 , " ZSTD_buildEntropyStatisticsAndEstimateSubBlockSize() " ) ;
2021-05-14 22:48:15 +00:00
FORWARD_IF_ERROR ( ZSTD_buildBlockEntropyStats ( seqStore ,
& zc - > blockState . prevCBlock - > entropy ,
& zc - > blockState . nextCBlock - > entropy ,
& zc - > appliedParams ,
2021-12-21 11:21:29 +00:00
entropyMetadata ,
2023-02-10 19:25:40 +00:00
zc - > entropyWorkspace , ENTROPY_WORKSPACE_SIZE ) , " " ) ;
return ZSTD_estimateBlockSize (
seqStore - > litStart , ( size_t ) ( seqStore - > lit - seqStore - > litStart ) ,
2021-05-14 22:48:15 +00:00
seqStore - > ofCode , seqStore - > llCode , seqStore - > mlCode ,
( size_t ) ( seqStore - > sequences - seqStore - > sequencesStart ) ,
2023-02-10 19:25:40 +00:00
& zc - > blockState . nextCBlock - > entropy ,
entropyMetadata ,
zc - > entropyWorkspace , ENTROPY_WORKSPACE_SIZE ,
2021-12-21 11:21:29 +00:00
( int ) ( entropyMetadata - > hufMetadata . hType = = set_compressed ) , 1 ) ;
2021-05-14 22:48:15 +00:00
}
/* Returns literals bytes represented in a seqStore */
2023-02-10 19:25:40 +00:00
static size_t ZSTD_countSeqStoreLiteralsBytes ( const seqStore_t * const seqStore )
{
2021-05-14 22:48:15 +00:00
size_t literalsBytes = 0 ;
2023-02-10 19:25:40 +00:00
size_t const nbSeqs = ( size_t ) ( seqStore - > sequences - seqStore - > sequencesStart ) ;
2021-05-14 22:48:15 +00:00
size_t i ;
for ( i = 0 ; i < nbSeqs ; + + i ) {
2023-02-10 19:25:40 +00:00
seqDef const seq = seqStore - > sequencesStart [ i ] ;
2021-05-14 22:48:15 +00:00
literalsBytes + = seq . litLength ;
if ( i = = seqStore - > longLengthPos & & seqStore - > longLengthType = = ZSTD_llt_literalLength ) {
literalsBytes + = 0x10000 ;
2023-02-10 19:25:40 +00:00
} }
2021-05-14 22:48:15 +00:00
return literalsBytes ;
}
/* Returns match bytes represented in a seqStore */
2023-02-10 19:25:40 +00:00
static size_t ZSTD_countSeqStoreMatchBytes ( const seqStore_t * const seqStore )
{
2021-05-14 22:48:15 +00:00
size_t matchBytes = 0 ;
2023-02-10 19:25:40 +00:00
size_t const nbSeqs = ( size_t ) ( seqStore - > sequences - seqStore - > sequencesStart ) ;
2021-05-14 22:48:15 +00:00
size_t i ;
for ( i = 0 ; i < nbSeqs ; + + i ) {
seqDef seq = seqStore - > sequencesStart [ i ] ;
2022-03-18 00:52:11 +00:00
matchBytes + = seq . mlBase + MINMATCH ;
2021-05-14 22:48:15 +00:00
if ( i = = seqStore - > longLengthPos & & seqStore - > longLengthType = = ZSTD_llt_matchLength ) {
matchBytes + = 0x10000 ;
2023-02-10 19:25:40 +00:00
} }
2021-05-14 22:48:15 +00:00
return matchBytes ;
}
/* Derives the seqStore that is a chunk of the originalSeqStore from [startIdx, endIdx).
* Stores the result in resultSeqStore .
*/
static void ZSTD_deriveSeqStoreChunk ( seqStore_t * resultSeqStore ,
const seqStore_t * originalSeqStore ,
2023-02-10 19:25:40 +00:00
size_t startIdx , size_t endIdx )
{
2021-05-14 22:48:15 +00:00
* resultSeqStore = * originalSeqStore ;
if ( startIdx > 0 ) {
resultSeqStore - > sequences = originalSeqStore - > sequencesStart + startIdx ;
2023-02-10 19:25:40 +00:00
resultSeqStore - > litStart + = ZSTD_countSeqStoreLiteralsBytes ( resultSeqStore ) ;
2021-05-14 22:48:15 +00:00
}
/* Move longLengthPos into the correct position if necessary */
if ( originalSeqStore - > longLengthType ! = ZSTD_llt_none ) {
if ( originalSeqStore - > longLengthPos < startIdx | | originalSeqStore - > longLengthPos > endIdx ) {
resultSeqStore - > longLengthType = ZSTD_llt_none ;
} else {
resultSeqStore - > longLengthPos - = ( U32 ) startIdx ;
}
}
resultSeqStore - > sequencesStart = originalSeqStore - > sequencesStart + startIdx ;
resultSeqStore - > sequences = originalSeqStore - > sequencesStart + endIdx ;
if ( endIdx = = ( size_t ) ( originalSeqStore - > sequences - originalSeqStore - > sequencesStart ) ) {
/* This accounts for possible last literals if the derived chunk reaches the end of the block */
2023-02-10 19:25:40 +00:00
assert ( resultSeqStore - > lit = = originalSeqStore - > lit ) ;
2021-05-14 22:48:15 +00:00
} else {
2023-02-10 19:25:40 +00:00
size_t const literalsBytes = ZSTD_countSeqStoreLiteralsBytes ( resultSeqStore ) ;
resultSeqStore - > lit = resultSeqStore - > litStart + literalsBytes ;
2021-05-14 22:48:15 +00:00
}
resultSeqStore - > llCode + = startIdx ;
resultSeqStore - > mlCode + = startIdx ;
resultSeqStore - > ofCode + = startIdx ;
}
/**
2023-02-10 19:25:40 +00:00
* Returns the raw offset represented by the combination of offBase , ll0 , and repcode history .
* offBase must represent a repcode in the numeric representation of ZSTD_storeSeq ( ) .
2021-05-14 22:48:15 +00:00
*/
2022-03-18 00:52:11 +00:00
static U32
2023-02-10 19:25:40 +00:00
ZSTD_resolveRepcodeToRawOffset ( const U32 rep [ ZSTD_REP_NUM ] , const U32 offBase , const U32 ll0 )
{
U32 const adjustedRepCode = OFFBASE_TO_REPCODE ( offBase ) - 1 + ll0 ; /* [ 0 - 3 ] */
assert ( OFFBASE_IS_REPCODE ( offBase ) ) ;
if ( adjustedRepCode = = ZSTD_REP_NUM ) {
assert ( ll0 ) ;
/* litlength == 0 and offCode == 2 implies selection of first repcode - 1
* This is only valid if it results in a valid offset value , aka > 0.
* Note : it may happen that ` rep [ 0 ] = = 1 ` in exceptional circumstances .
* In which case this function will return 0 , which is an invalid offset .
* It ' s not an issue though , since this value will be
* compared and discarded within ZSTD_seqStore_resolveOffCodes ( ) .
*/
2021-05-14 22:48:15 +00:00
return rep [ 0 ] - 1 ;
}
2023-02-10 19:25:40 +00:00
return rep [ adjustedRepCode ] ;
2021-05-14 22:48:15 +00:00
}
/**
* ZSTD_seqStore_resolveOffCodes ( ) reconciles any possible divergences in offset history that may arise
2022-03-18 00:52:11 +00:00
* due to emission of RLE / raw blocks that disturb the offset history ,
* and replaces any repcodes within the seqStore that may be invalid .
*
* dRepcodes are updated as would be on the decompression side .
* cRepcodes are updated exactly in accordance with the seqStore .
2021-05-14 22:48:15 +00:00
*
2022-03-18 00:52:11 +00:00
* Note : this function assumes seq - > offBase respects the following numbering scheme :
* 0 : invalid
* 1 - 3 : repcode 1 - 3
* 4 + : real_offset + 3
2021-05-14 22:48:15 +00:00
*/
2023-02-10 19:25:40 +00:00
static void
ZSTD_seqStore_resolveOffCodes ( repcodes_t * const dRepcodes , repcodes_t * const cRepcodes ,
const seqStore_t * const seqStore , U32 const nbSeq )
{
2021-05-14 22:48:15 +00:00
U32 idx = 0 ;
for ( ; idx < nbSeq ; + + idx ) {
seqDef * const seq = seqStore - > sequencesStart + idx ;
U32 const ll0 = ( seq - > litLength = = 0 ) ;
2023-02-10 19:25:40 +00:00
U32 const offBase = seq - > offBase ;
assert ( offBase > 0 ) ;
if ( OFFBASE_IS_REPCODE ( offBase ) ) {
U32 const dRawOffset = ZSTD_resolveRepcodeToRawOffset ( dRepcodes - > rep , offBase , ll0 ) ;
U32 const cRawOffset = ZSTD_resolveRepcodeToRawOffset ( cRepcodes - > rep , offBase , ll0 ) ;
2021-05-14 22:48:15 +00:00
/* Adjust simulated decompression repcode history if we come across a mismatch. Replace
* the repcode with the offset it actually references , determined by the compression
* repcode history .
*/
if ( dRawOffset ! = cRawOffset ) {
2023-02-10 19:25:40 +00:00
seq - > offBase = OFFSET_TO_OFFBASE ( cRawOffset ) ;
2021-05-14 22:48:15 +00:00
}
}
/* Compression repcode history is always updated with values directly from the unmodified seqStore.
* Decompression repcode history may use modified seq - > offset value taken from compression repcode history .
*/
2023-02-10 19:25:40 +00:00
ZSTD_updateRep ( dRepcodes - > rep , seq - > offBase , ll0 ) ;
ZSTD_updateRep ( cRepcodes - > rep , offBase , ll0 ) ;
2021-05-14 22:48:15 +00:00
}
}
/* ZSTD_compressSeqStore_singleBlock():
* Compresses a seqStore into a block with a block header , into the buffer dst .
*
* Returns the total size of that block ( including header ) or a ZSTD error code .
*/
2022-03-18 00:52:11 +00:00
static size_t
2023-02-10 19:25:40 +00:00
ZSTD_compressSeqStore_singleBlock ( ZSTD_CCtx * zc ,
const seqStore_t * const seqStore ,
2022-03-18 00:52:11 +00:00
repcodes_t * const dRep , repcodes_t * const cRep ,
void * dst , size_t dstCapacity ,
2023-02-10 19:25:40 +00:00
const void * src , size_t srcSize ,
2022-03-18 00:52:11 +00:00
U32 lastBlock , U32 isPartition )
{
2021-05-14 22:48:15 +00:00
const U32 rleMaxLength = 25 ;
BYTE * op = ( BYTE * ) dst ;
const BYTE * ip = ( const BYTE * ) src ;
size_t cSize ;
size_t cSeqsSize ;
/* In case of an RLE or raw block, the simulated decompression repcode history must be reset */
repcodes_t const dRepOriginal = * dRep ;
2022-03-18 00:52:11 +00:00
DEBUGLOG ( 5 , " ZSTD_compressSeqStore_singleBlock " ) ;
2021-05-14 22:48:15 +00:00
if ( isPartition )
ZSTD_seqStore_resolveOffCodes ( dRep , cRep , seqStore , ( U32 ) ( seqStore - > sequences - seqStore - > sequencesStart ) ) ;
2021-12-21 11:21:29 +00:00
RETURN_ERROR_IF ( dstCapacity < ZSTD_blockHeaderSize , dstSize_tooSmall , " Block header doesn't fit " ) ;
2021-05-14 22:48:15 +00:00
cSeqsSize = ZSTD_entropyCompressSeqStore ( seqStore ,
& zc - > blockState . prevCBlock - > entropy , & zc - > blockState . nextCBlock - > entropy ,
& zc - > appliedParams ,
op + ZSTD_blockHeaderSize , dstCapacity - ZSTD_blockHeaderSize ,
srcSize ,
zc - > entropyWorkspace , ENTROPY_WORKSPACE_SIZE /* statically allocated in resetCCtx */ ,
zc - > bmi2 ) ;
FORWARD_IF_ERROR ( cSeqsSize , " ZSTD_entropyCompressSeqStore failed! " ) ;
if ( ! zc - > isFirstBlock & &
cSeqsSize < rleMaxLength & &
ZSTD_isRLE ( ( BYTE const * ) src , srcSize ) ) {
/* We don't want to emit our first block as a RLE even if it qualifies because
* doing so will cause the decoder ( cli only ) to throw a " should consume all input error. "
* This is only an issue for zstd < = v1 .4 .3
*/
cSeqsSize = 1 ;
}
if ( zc - > seqCollector . collectSequences ) {
ZSTD_copyBlockSequences ( zc ) ;
ZSTD_blockState_confirmRepcodesAndEntropyTables ( & zc - > blockState ) ;
return 0 ;
}
if ( cSeqsSize = = 0 ) {
cSize = ZSTD_noCompressBlock ( op , dstCapacity , ip , srcSize , lastBlock ) ;
FORWARD_IF_ERROR ( cSize , " Nocompress block failed " ) ;
DEBUGLOG ( 4 , " Writing out nocompress block, size: %zu " , cSize ) ;
* dRep = dRepOriginal ; /* reset simulated decompression repcode history */
} else if ( cSeqsSize = = 1 ) {
cSize = ZSTD_rleCompressBlock ( op , dstCapacity , * ip , srcSize , lastBlock ) ;
FORWARD_IF_ERROR ( cSize , " RLE compress block failed " ) ;
DEBUGLOG ( 4 , " Writing out RLE block, size: %zu " , cSize ) ;
* dRep = dRepOriginal ; /* reset simulated decompression repcode history */
} else {
ZSTD_blockState_confirmRepcodesAndEntropyTables ( & zc - > blockState ) ;
writeBlockHeader ( op , cSeqsSize , srcSize , lastBlock ) ;
cSize = ZSTD_blockHeaderSize + cSeqsSize ;
DEBUGLOG ( 4 , " Writing out compressed block, size: %zu " , cSize ) ;
}
2021-12-21 11:21:29 +00:00
if ( zc - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = = FSE_repeat_valid )
zc - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = FSE_repeat_check ;
2021-05-14 22:48:15 +00:00
return cSize ;
}
/* Struct to keep track of where we are in our recursive calls. */
typedef struct {
U32 * splitLocations ; /* Array of split indices */
size_t idx ; /* The current index within splitLocations being worked on */
} seqStoreSplits ;
# define MIN_SEQUENCES_BLOCK_SPLITTING 300
/* Helper function to perform the recursive search for block splits.
* Estimates the cost of seqStore prior to split , and estimates the cost of splitting the sequences in half .
2023-02-10 19:25:40 +00:00
* If advantageous to split , then we recurse down the two sub - blocks .
* If not , or if an error occurred in estimation , then we do not recurse .
2021-05-14 22:48:15 +00:00
*
2023-02-10 19:25:40 +00:00
* Note : The recursion depth is capped by a heuristic minimum number of sequences ,
* defined by MIN_SEQUENCES_BLOCK_SPLITTING .
2021-05-14 22:48:15 +00:00
* In theory , this means the absolute largest recursion depth is 10 = = log2 ( maxNbSeqInBlock / MIN_SEQUENCES_BLOCK_SPLITTING ) .
* In practice , recursion depth usually doesn ' t go beyond 4.
*
2023-02-10 19:25:40 +00:00
* Furthermore , the number of splits is capped by ZSTD_MAX_NB_BLOCK_SPLITS .
* At ZSTD_MAX_NB_BLOCK_SPLITS = = 196 with the current existing blockSize
2021-05-14 22:48:15 +00:00
* maximum of 128 KB , this value is actually impossible to reach .
*/
2022-03-18 00:52:11 +00:00
static void
ZSTD_deriveBlockSplitsHelper ( seqStoreSplits * splits , size_t startIdx , size_t endIdx ,
ZSTD_CCtx * zc , const seqStore_t * origSeqStore )
{
2023-02-10 19:25:40 +00:00
seqStore_t * const fullSeqStoreChunk = & zc - > blockSplitCtx . fullSeqStoreChunk ;
seqStore_t * const firstHalfSeqStore = & zc - > blockSplitCtx . firstHalfSeqStore ;
seqStore_t * const secondHalfSeqStore = & zc - > blockSplitCtx . secondHalfSeqStore ;
2021-05-14 22:48:15 +00:00
size_t estimatedOriginalSize ;
size_t estimatedFirstHalfSize ;
size_t estimatedSecondHalfSize ;
size_t midIdx = ( startIdx + endIdx ) / 2 ;
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " ZSTD_deriveBlockSplitsHelper: startIdx=%zu endIdx=%zu " , startIdx , endIdx ) ;
assert ( endIdx > = startIdx ) ;
2021-12-21 11:21:29 +00:00
if ( endIdx - startIdx < MIN_SEQUENCES_BLOCK_SPLITTING | | splits - > idx > = ZSTD_MAX_NB_BLOCK_SPLITS ) {
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 6 , " ZSTD_deriveBlockSplitsHelper: Too few sequences (%zu) " , endIdx - startIdx ) ;
2021-05-14 22:48:15 +00:00
return ;
}
2021-12-21 11:21:29 +00:00
ZSTD_deriveSeqStoreChunk ( fullSeqStoreChunk , origSeqStore , startIdx , endIdx ) ;
ZSTD_deriveSeqStoreChunk ( firstHalfSeqStore , origSeqStore , startIdx , midIdx ) ;
ZSTD_deriveSeqStoreChunk ( secondHalfSeqStore , origSeqStore , midIdx , endIdx ) ;
estimatedOriginalSize = ZSTD_buildEntropyStatisticsAndEstimateSubBlockSize ( fullSeqStoreChunk , zc ) ;
estimatedFirstHalfSize = ZSTD_buildEntropyStatisticsAndEstimateSubBlockSize ( firstHalfSeqStore , zc ) ;
estimatedSecondHalfSize = ZSTD_buildEntropyStatisticsAndEstimateSubBlockSize ( secondHalfSeqStore , zc ) ;
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " Estimated original block size: %zu -- First half split: %zu -- Second half split: %zu " ,
2021-05-14 22:48:15 +00:00
estimatedOriginalSize , estimatedFirstHalfSize , estimatedSecondHalfSize ) ;
if ( ZSTD_isError ( estimatedOriginalSize ) | | ZSTD_isError ( estimatedFirstHalfSize ) | | ZSTD_isError ( estimatedSecondHalfSize ) ) {
return ;
}
if ( estimatedFirstHalfSize + estimatedSecondHalfSize < estimatedOriginalSize ) {
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " split decided at seqNb:%zu " , midIdx ) ;
2021-05-14 22:48:15 +00:00
ZSTD_deriveBlockSplitsHelper ( splits , startIdx , midIdx , zc , origSeqStore ) ;
splits - > splitLocations [ splits - > idx ] = ( U32 ) midIdx ;
splits - > idx + + ;
ZSTD_deriveBlockSplitsHelper ( splits , midIdx , endIdx , zc , origSeqStore ) ;
}
}
2023-02-10 19:25:40 +00:00
/* Base recursive function.
* Populates a table with intra - block partition indices that can improve compression ratio .
2021-05-14 22:48:15 +00:00
*
2023-02-10 19:25:40 +00:00
* @ return : number of splits made ( which equals the size of the partition table - 1 ) .
2021-05-14 22:48:15 +00:00
*/
2023-02-10 19:25:40 +00:00
static size_t ZSTD_deriveBlockSplits ( ZSTD_CCtx * zc , U32 partitions [ ] , U32 nbSeq )
{
seqStoreSplits splits ;
splits . splitLocations = partitions ;
splits . idx = 0 ;
2021-05-14 22:48:15 +00:00
if ( nbSeq < = 4 ) {
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " ZSTD_deriveBlockSplits: Too few sequences to split (%u <= 4) " , nbSeq ) ;
2021-05-14 22:48:15 +00:00
/* Refuse to try and split anything with less than 4 sequences */
return 0 ;
}
ZSTD_deriveBlockSplitsHelper ( & splits , 0 , nbSeq , zc , & zc - > seqStore ) ;
splits . splitLocations [ splits . idx ] = nbSeq ;
DEBUGLOG ( 5 , " ZSTD_deriveBlockSplits: final nb partitions: %zu " , splits . idx + 1 ) ;
return splits . idx ;
}
/* ZSTD_compressBlock_splitBlock():
* Attempts to split a given block into multiple blocks to improve compression ratio .
*
* Returns combined size of all blocks ( which includes headers ) , or a ZSTD error code .
*/
2022-03-18 00:52:11 +00:00
static size_t
2023-02-10 19:25:40 +00:00
ZSTD_compressBlock_splitBlock_internal ( ZSTD_CCtx * zc ,
void * dst , size_t dstCapacity ,
const void * src , size_t blockSize ,
U32 lastBlock , U32 nbSeq )
2022-03-18 00:52:11 +00:00
{
2021-05-14 22:48:15 +00:00
size_t cSize = 0 ;
const BYTE * ip = ( const BYTE * ) src ;
BYTE * op = ( BYTE * ) dst ;
size_t i = 0 ;
size_t srcBytesTotal = 0 ;
2023-02-10 19:25:40 +00:00
U32 * const partitions = zc - > blockSplitCtx . partitions ; /* size == ZSTD_MAX_NB_BLOCK_SPLITS */
seqStore_t * const nextSeqStore = & zc - > blockSplitCtx . nextSeqStore ;
seqStore_t * const currSeqStore = & zc - > blockSplitCtx . currSeqStore ;
size_t const numSplits = ZSTD_deriveBlockSplits ( zc , partitions , nbSeq ) ;
2021-05-14 22:48:15 +00:00
/* If a block is split and some partitions are emitted as RLE/uncompressed, then repcode history
* may become invalid . In order to reconcile potentially invalid repcodes , we keep track of two
* separate repcode histories that simulate repcode history on compression and decompression side ,
* and use the histories to determine whether we must replace a particular repcode with its raw offset .
*
* 1 ) cRep gets updated for each partition , regardless of whether the block was emitted as uncompressed
* or RLE . This allows us to retrieve the offset value that an invalid repcode references within
* a nocompress / RLE block .
* 2 ) dRep gets updated only for compressed partitions , and when a repcode gets replaced , will use
* the replacement offset value rather than the original repcode to update the repcode history .
* dRep also will be the final repcode history sent to the next block .
*
* See ZSTD_seqStore_resolveOffCodes ( ) for more details .
*/
repcodes_t dRep ;
repcodes_t cRep ;
ZSTD_memcpy ( dRep . rep , zc - > blockState . prevCBlock - > rep , sizeof ( repcodes_t ) ) ;
ZSTD_memcpy ( cRep . rep , zc - > blockState . prevCBlock - > rep , sizeof ( repcodes_t ) ) ;
2021-12-21 11:21:29 +00:00
ZSTD_memset ( nextSeqStore , 0 , sizeof ( seqStore_t ) ) ;
2021-05-14 22:48:15 +00:00
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " ZSTD_compressBlock_splitBlock_internal (dstCapacity=%u, dictLimit=%u, nextToUpdate=%u) " ,
2021-05-14 22:48:15 +00:00
( unsigned ) dstCapacity , ( unsigned ) zc - > blockState . matchState . window . dictLimit ,
( unsigned ) zc - > blockState . matchState . nextToUpdate ) ;
if ( numSplits = = 0 ) {
2023-02-10 19:25:40 +00:00
size_t cSizeSingleBlock =
ZSTD_compressSeqStore_singleBlock ( zc , & zc - > seqStore ,
& dRep , & cRep ,
op , dstCapacity ,
ip , blockSize ,
lastBlock , 0 /* isPartition */ ) ;
2021-05-14 22:48:15 +00:00
FORWARD_IF_ERROR ( cSizeSingleBlock , " Compressing single block from splitBlock_internal() failed! " ) ;
DEBUGLOG ( 5 , " ZSTD_compressBlock_splitBlock_internal: No splits " ) ;
2023-02-10 19:25:40 +00:00
assert ( zc - > blockSize < = ZSTD_BLOCKSIZE_MAX ) ;
assert ( cSizeSingleBlock < = zc - > blockSize + ZSTD_blockHeaderSize ) ;
2021-05-14 22:48:15 +00:00
return cSizeSingleBlock ;
}
2021-12-21 11:21:29 +00:00
ZSTD_deriveSeqStoreChunk ( currSeqStore , & zc - > seqStore , 0 , partitions [ 0 ] ) ;
2021-05-14 22:48:15 +00:00
for ( i = 0 ; i < = numSplits ; + + i ) {
size_t cSizeChunk ;
U32 const lastPartition = ( i = = numSplits ) ;
U32 lastBlockEntireSrc = 0 ;
2023-02-10 19:25:40 +00:00
size_t srcBytes = ZSTD_countSeqStoreLiteralsBytes ( currSeqStore ) + ZSTD_countSeqStoreMatchBytes ( currSeqStore ) ;
2021-05-14 22:48:15 +00:00
srcBytesTotal + = srcBytes ;
if ( lastPartition ) {
/* This is the final partition, need to account for possible last literals */
srcBytes + = blockSize - srcBytesTotal ;
lastBlockEntireSrc = lastBlock ;
} else {
2021-12-21 11:21:29 +00:00
ZSTD_deriveSeqStoreChunk ( nextSeqStore , & zc - > seqStore , partitions [ i ] , partitions [ i + 1 ] ) ;
2021-05-14 22:48:15 +00:00
}
2021-12-21 11:21:29 +00:00
cSizeChunk = ZSTD_compressSeqStore_singleBlock ( zc , currSeqStore ,
2021-05-14 22:48:15 +00:00
& dRep , & cRep ,
op , dstCapacity ,
ip , srcBytes ,
lastBlockEntireSrc , 1 /* isPartition */ ) ;
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " Estimated size: %zu vs %zu : actual size " ,
ZSTD_buildEntropyStatisticsAndEstimateSubBlockSize ( currSeqStore , zc ) , cSizeChunk ) ;
2021-05-14 22:48:15 +00:00
FORWARD_IF_ERROR ( cSizeChunk , " Compressing chunk failed! " ) ;
ip + = srcBytes ;
op + = cSizeChunk ;
dstCapacity - = cSizeChunk ;
cSize + = cSizeChunk ;
2021-12-21 11:21:29 +00:00
* currSeqStore = * nextSeqStore ;
2023-02-10 19:25:40 +00:00
assert ( cSizeChunk < = zc - > blockSize + ZSTD_blockHeaderSize ) ;
2021-05-14 22:48:15 +00:00
}
2023-02-10 19:25:40 +00:00
/* cRep and dRep may have diverged during the compression.
* If so , we use the dRep repcodes for the next block .
2021-05-14 22:48:15 +00:00
*/
ZSTD_memcpy ( zc - > blockState . prevCBlock - > rep , dRep . rep , sizeof ( repcodes_t ) ) ;
return cSize ;
}
2022-03-18 00:52:11 +00:00
static size_t
ZSTD_compressBlock_splitBlock ( ZSTD_CCtx * zc ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize , U32 lastBlock )
{
2021-05-14 22:48:15 +00:00
U32 nbSeq ;
size_t cSize ;
DEBUGLOG ( 4 , " ZSTD_compressBlock_splitBlock " ) ;
2021-12-21 11:21:29 +00:00
assert ( zc - > appliedParams . useBlockSplitter = = ZSTD_ps_enable ) ;
2021-05-14 22:48:15 +00:00
{ const size_t bss = ZSTD_buildSeqStore ( zc , src , srcSize ) ;
FORWARD_IF_ERROR ( bss , " ZSTD_buildSeqStore failed " ) ;
if ( bss = = ZSTDbss_noCompress ) {
if ( zc - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = = FSE_repeat_valid )
zc - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = FSE_repeat_check ;
2023-02-10 19:25:40 +00:00
cSize = ZSTD_noCompressBlock ( dst , dstCapacity , src , srcSize , lastBlock ) ;
2021-05-14 22:48:15 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_noCompressBlock failed " ) ;
DEBUGLOG ( 4 , " ZSTD_compressBlock_splitBlock: Nocompress block " ) ;
return cSize ;
}
nbSeq = ( U32 ) ( zc - > seqStore . sequences - zc - > seqStore . sequencesStart ) ;
}
cSize = ZSTD_compressBlock_splitBlock_internal ( zc , dst , dstCapacity , src , srcSize , lastBlock , nbSeq ) ;
FORWARD_IF_ERROR ( cSize , " Splitting blocks failed! " ) ;
return cSize ;
2020-02-08 13:56:59 +00:00
}
2022-03-18 00:52:11 +00:00
static size_t
ZSTD_compressBlock_internal ( ZSTD_CCtx * zc ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize , U32 frame )
2020-02-08 13:56:59 +00:00
{
2023-02-10 19:25:40 +00:00
/* This is an estimated upper bound for the length of an rle block.
* This isn ' t the actual upper bound .
* Finding the real threshold needs further investigation .
2020-02-08 13:56:59 +00:00
*/
const U32 rleMaxLength = 25 ;
size_t cSize ;
const BYTE * ip = ( const BYTE * ) src ;
BYTE * op = ( BYTE * ) dst ;
DEBUGLOG ( 5 , " ZSTD_compressBlock_internal (dstCapacity=%u, dictLimit=%u, nextToUpdate=%u) " ,
( unsigned ) dstCapacity , ( unsigned ) zc - > blockState . matchState . window . dictLimit ,
( unsigned ) zc - > blockState . matchState . nextToUpdate ) ;
{ const size_t bss = ZSTD_buildSeqStore ( zc , src , srcSize ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( bss , " ZSTD_buildSeqStore failed " ) ;
2020-02-08 13:56:59 +00:00
if ( bss = = ZSTDbss_noCompress ) { cSize = 0 ; goto out ; }
}
if ( zc - > seqCollector . collectSequences ) {
ZSTD_copyBlockSequences ( zc ) ;
2021-05-14 22:48:15 +00:00
ZSTD_blockState_confirmRepcodesAndEntropyTables ( & zc - > blockState ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
/* encode sequences and literals */
2021-05-14 22:48:15 +00:00
cSize = ZSTD_entropyCompressSeqStore ( & zc - > seqStore ,
2020-02-08 13:56:59 +00:00
& zc - > blockState . prevCBlock - > entropy , & zc - > blockState . nextCBlock - > entropy ,
& zc - > appliedParams ,
dst , dstCapacity ,
srcSize ,
2020-12-18 14:46:27 +00:00
zc - > entropyWorkspace , ENTROPY_WORKSPACE_SIZE /* statically allocated in resetCCtx */ ,
2020-02-08 13:56:59 +00:00
zc - > bmi2 ) ;
if ( frame & &
/* We don't want to emit our first block as a RLE even if it qualifies because
* doing so will cause the decoder ( cli only ) to throw a " should consume all input error. "
* This is only an issue for zstd < = v1 .4 .3
*/
! zc - > isFirstBlock & &
cSize < rleMaxLength & &
ZSTD_isRLE ( ip , srcSize ) )
{
cSize = 1 ;
op [ 0 ] = ip [ 0 ] ;
}
out :
if ( ! ZSTD_isError ( cSize ) & & cSize > 1 ) {
2021-05-14 22:48:15 +00:00
ZSTD_blockState_confirmRepcodesAndEntropyTables ( & zc - > blockState ) ;
2020-02-08 13:56:59 +00:00
}
/* We check that dictionaries have offset codes available for the first
* block . After the first block , the offcode table might not have large
* enough codes to represent the offsets in the data .
*/
if ( zc - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = = FSE_repeat_valid )
zc - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = FSE_repeat_check ;
return cSize ;
}
static size_t ZSTD_compressBlock_targetCBlockSize_body ( ZSTD_CCtx * zc ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
const size_t bss , U32 lastBlock )
{
DEBUGLOG ( 6 , " Attempting ZSTD_compressSuperBlock() " ) ;
if ( bss = = ZSTDbss_compress ) {
2020-06-04 19:03:27 +00:00
if ( /* We don't want to emit our first block as a RLE even if it qualifies because
* doing so will cause the decoder ( cli only ) to throw a " should consume all input error. "
* This is only an issue for zstd < = v1 .4 .3
*/
! zc - > isFirstBlock & &
ZSTD_maybeRLE ( & zc - > seqStore ) & &
ZSTD_isRLE ( ( BYTE const * ) src , srcSize ) )
{
return ZSTD_rleCompressBlock ( dst , dstCapacity , * ( BYTE const * ) src , srcSize , lastBlock ) ;
}
2020-02-08 13:56:59 +00:00
/* Attempt superblock compression.
*
* Note that compressed size of ZSTD_compressSuperBlock ( ) is not bound by the
* standard ZSTD_compressBound ( ) . This is a problem , because even if we have
* space now , taking an extra byte now could cause us to run out of space later
* and violate ZSTD_compressBound ( ) .
*
* Define blockBound ( blockSize ) = blockSize + ZSTD_blockHeaderSize .
*
* In order to respect ZSTD_compressBound ( ) we must attempt to emit a raw
* uncompressed block in these cases :
* * cSize = = 0 : Return code for an uncompressed block .
* * cSize = = dstSize_tooSmall : We may have expanded beyond blockBound ( srcSize ) .
* ZSTD_noCompressBlock ( ) will return dstSize_tooSmall if we are really out of
* output space .
* * cSize > = blockBound ( srcSize ) : We have expanded the block too much so
* emit an uncompressed block .
*/
2023-02-10 19:25:40 +00:00
{ size_t const cSize =
ZSTD_compressSuperBlock ( zc , dst , dstCapacity , src , srcSize , lastBlock ) ;
2020-06-04 19:03:27 +00:00
if ( cSize ! = ERROR ( dstSize_tooSmall ) ) {
2023-02-10 19:25:40 +00:00
size_t const maxCSize =
srcSize - ZSTD_minGain ( srcSize , zc - > appliedParams . cParams . strategy ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_compressSuperBlock failed " ) ;
if ( cSize ! = 0 & & cSize < maxCSize + ZSTD_blockHeaderSize ) {
2021-05-14 22:48:15 +00:00
ZSTD_blockState_confirmRepcodesAndEntropyTables ( & zc - > blockState ) ;
2020-06-04 19:03:27 +00:00
return cSize ;
}
2020-02-08 13:56:59 +00:00
}
}
2023-02-10 19:25:40 +00:00
} /* if (bss == ZSTDbss_compress)*/
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 6 , " Resorting to ZSTD_noCompressBlock() " ) ;
/* Superblock compression failed, attempt to emit a single no compress block.
* The decoder will be able to stream this block since it is uncompressed .
*/
return ZSTD_noCompressBlock ( dst , dstCapacity , src , srcSize , lastBlock ) ;
}
static size_t ZSTD_compressBlock_targetCBlockSize ( ZSTD_CCtx * zc ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
U32 lastBlock )
{
size_t cSize = 0 ;
const size_t bss = ZSTD_buildSeqStore ( zc , src , srcSize ) ;
DEBUGLOG ( 5 , " ZSTD_compressBlock_targetCBlockSize (dstCapacity=%u, dictLimit=%u, nextToUpdate=%u, srcSize=%zu) " ,
( unsigned ) dstCapacity , ( unsigned ) zc - > blockState . matchState . window . dictLimit , ( unsigned ) zc - > blockState . matchState . nextToUpdate , srcSize ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( bss , " ZSTD_buildSeqStore failed " ) ;
2020-02-08 13:56:59 +00:00
cSize = ZSTD_compressBlock_targetCBlockSize_body ( zc , dst , dstCapacity , src , srcSize , bss , lastBlock ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_compressBlock_targetCBlockSize_body failed " ) ;
2020-02-08 13:56:59 +00:00
if ( zc - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = = FSE_repeat_valid )
zc - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = FSE_repeat_check ;
return cSize ;
}
static void ZSTD_overflowCorrectIfNeeded ( ZSTD_matchState_t * ms ,
ZSTD_cwksp * ws ,
ZSTD_CCtx_params const * params ,
void const * ip ,
void const * iend )
{
2021-05-14 22:48:15 +00:00
U32 const cycleLog = ZSTD_cycleLog ( params - > cParams . chainLog , params - > cParams . strategy ) ;
U32 const maxDist = ( U32 ) 1 < < params - > cParams . windowLog ;
if ( ZSTD_window_needOverflowCorrection ( ms - > window , cycleLog , maxDist , ms - > loadedDictEnd , ip , iend ) ) {
2020-02-08 13:56:59 +00:00
U32 const correction = ZSTD_window_correctOverflow ( & ms - > window , cycleLog , maxDist , ip ) ;
ZSTD_STATIC_ASSERT ( ZSTD_CHAINLOG_MAX < = 30 ) ;
ZSTD_STATIC_ASSERT ( ZSTD_WINDOWLOG_MAX_32 < = 30 ) ;
ZSTD_STATIC_ASSERT ( ZSTD_WINDOWLOG_MAX < = 31 ) ;
ZSTD_cwksp_mark_tables_dirty ( ws ) ;
ZSTD_reduceIndex ( ms , params , correction ) ;
ZSTD_cwksp_mark_tables_clean ( ws ) ;
if ( ms - > nextToUpdate < correction ) ms - > nextToUpdate = 0 ;
else ms - > nextToUpdate - = correction ;
/* invalidate dictionaries on overflow correction */
ms - > loadedDictEnd = 0 ;
ms - > dictMatchState = NULL ;
}
}
/*! ZSTD_compress_frameChunk() :
* Compress a chunk of data into one or multiple blocks .
* All blocks will be terminated , all input will be consumed .
* Function will issue an error if there is not enough ` dstCapacity ` to hold the compressed content .
* Frame is supposed already started ( header already produced )
2023-02-10 19:25:40 +00:00
* @ return : compressed size , or an error code
2020-02-08 13:56:59 +00:00
*/
2021-05-14 22:48:15 +00:00
static size_t ZSTD_compress_frameChunk ( ZSTD_CCtx * cctx ,
2020-02-08 13:56:59 +00:00
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
U32 lastFrameChunk )
{
size_t blockSize = cctx - > blockSize ;
size_t remaining = srcSize ;
const BYTE * ip = ( const BYTE * ) src ;
BYTE * const ostart = ( BYTE * ) dst ;
BYTE * op = ostart ;
U32 const maxDist = ( U32 ) 1 < < cctx - > appliedParams . cParams . windowLog ;
assert ( cctx - > appliedParams . cParams . windowLog < = ZSTD_WINDOWLOG_MAX ) ;
2020-12-18 14:46:27 +00:00
DEBUGLOG ( 4 , " ZSTD_compress_frameChunk (blockSize=%u) " , ( unsigned ) blockSize ) ;
2020-02-08 13:56:59 +00:00
if ( cctx - > appliedParams . fParams . checksumFlag & & srcSize )
XXH64_update ( & cctx - > xxhState , src , srcSize ) ;
while ( remaining ) {
ZSTD_matchState_t * const ms = & cctx - > blockState . matchState ;
U32 const lastBlock = lastFrameChunk & ( blockSize > = remaining ) ;
2023-02-10 19:25:40 +00:00
/* TODO: See 3090. We reduced MIN_CBLOCK_SIZE from 3 to 2 so to compensate we are adding
* additional 1. We need to revisit and change this logic to be more consistent */
RETURN_ERROR_IF ( dstCapacity < ZSTD_blockHeaderSize + MIN_CBLOCK_SIZE + 1 ,
2020-02-08 13:56:59 +00:00
dstSize_tooSmall ,
" not enough space to store compressed block " ) ;
if ( remaining < blockSize ) blockSize = remaining ;
ZSTD_overflowCorrectIfNeeded (
ms , & cctx - > workspace , & cctx - > appliedParams , ip , ip + blockSize ) ;
ZSTD_checkDictValidity ( & ms - > window , ip + blockSize , maxDist , & ms - > loadedDictEnd , & ms - > dictMatchState ) ;
2021-12-21 11:21:29 +00:00
ZSTD_window_enforceMaxDist ( & ms - > window , ip , maxDist , & ms - > loadedDictEnd , & ms - > dictMatchState ) ;
2020-02-08 13:56:59 +00:00
/* Ensure hash/chain table insertion resumes no sooner than lowlimit */
if ( ms - > nextToUpdate < ms - > window . lowLimit ) ms - > nextToUpdate = ms - > window . lowLimit ;
{ size_t cSize ;
if ( ZSTD_useTargetCBlockSize ( & cctx - > appliedParams ) ) {
cSize = ZSTD_compressBlock_targetCBlockSize ( cctx , op , dstCapacity , ip , blockSize , lastBlock ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_compressBlock_targetCBlockSize failed " ) ;
2020-02-08 13:56:59 +00:00
assert ( cSize > 0 ) ;
assert ( cSize < = blockSize + ZSTD_blockHeaderSize ) ;
2021-05-14 22:48:15 +00:00
} else if ( ZSTD_blockSplitterEnabled ( & cctx - > appliedParams ) ) {
cSize = ZSTD_compressBlock_splitBlock ( cctx , op , dstCapacity , ip , blockSize , lastBlock ) ;
FORWARD_IF_ERROR ( cSize , " ZSTD_compressBlock_splitBlock failed " ) ;
assert ( cSize > 0 | | cctx - > seqCollector . collectSequences = = 1 ) ;
2020-02-08 13:56:59 +00:00
} else {
cSize = ZSTD_compressBlock_internal ( cctx ,
op + ZSTD_blockHeaderSize , dstCapacity - ZSTD_blockHeaderSize ,
ip , blockSize , 1 /* frame */ ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_compressBlock_internal failed " ) ;
2020-02-08 13:56:59 +00:00
if ( cSize = = 0 ) { /* block is not compressible */
cSize = ZSTD_noCompressBlock ( op , dstCapacity , ip , blockSize , lastBlock ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_noCompressBlock failed " ) ;
2020-02-08 13:56:59 +00:00
} else {
U32 const cBlockHeader = cSize = = 1 ?
lastBlock + ( ( ( U32 ) bt_rle ) < < 1 ) + ( U32 ) ( blockSize < < 3 ) :
lastBlock + ( ( ( U32 ) bt_compressed ) < < 1 ) + ( U32 ) ( cSize < < 3 ) ;
MEM_writeLE24 ( op , cBlockHeader ) ;
cSize + = ZSTD_blockHeaderSize ;
}
2023-02-10 19:25:40 +00:00
} /* if (ZSTD_useTargetCBlockSize(&cctx->appliedParams))*/
2020-02-08 13:56:59 +00:00
ip + = blockSize ;
assert ( remaining > = blockSize ) ;
remaining - = blockSize ;
op + = cSize ;
assert ( dstCapacity > = cSize ) ;
dstCapacity - = cSize ;
cctx - > isFirstBlock = 0 ;
DEBUGLOG ( 5 , " ZSTD_compress_frameChunk: adding a block of size %u " ,
( unsigned ) cSize ) ;
} }
if ( lastFrameChunk & & ( op > ostart ) ) cctx - > stage = ZSTDcs_ending ;
return ( size_t ) ( op - ostart ) ;
}
static size_t ZSTD_writeFrameHeader ( void * dst , size_t dstCapacity ,
const ZSTD_CCtx_params * params , U64 pledgedSrcSize , U32 dictID )
{ BYTE * const op = ( BYTE * ) dst ;
U32 const dictIDSizeCodeLength = ( dictID > 0 ) + ( dictID > = 256 ) + ( dictID > = 65536 ) ; /* 0-3 */
U32 const dictIDSizeCode = params - > fParams . noDictIDFlag ? 0 : dictIDSizeCodeLength ; /* 0-3 */
U32 const checksumFlag = params - > fParams . checksumFlag > 0 ;
U32 const windowSize = ( U32 ) 1 < < params - > cParams . windowLog ;
U32 const singleSegment = params - > fParams . contentSizeFlag & & ( windowSize > = pledgedSrcSize ) ;
BYTE const windowLogByte = ( BYTE ) ( ( params - > cParams . windowLog - ZSTD_WINDOWLOG_ABSOLUTEMIN ) < < 3 ) ;
U32 const fcsCode = params - > fParams . contentSizeFlag ?
( pledgedSrcSize > = 256 ) + ( pledgedSrcSize > = 65536 + 256 ) + ( pledgedSrcSize > = 0xFFFFFFFFU ) : 0 ; /* 0-3 */
BYTE const frameHeaderDescriptionByte = ( BYTE ) ( dictIDSizeCode + ( checksumFlag < < 2 ) + ( singleSegment < < 5 ) + ( fcsCode < < 6 ) ) ;
size_t pos = 0 ;
assert ( ! ( params - > fParams . contentSizeFlag & & pledgedSrcSize = = ZSTD_CONTENTSIZE_UNKNOWN ) ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( dstCapacity < ZSTD_FRAMEHEADERSIZE_MAX , dstSize_tooSmall ,
" dst buf is too small to fit worst-case frame header size. " ) ;
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 4 , " ZSTD_writeFrameHeader : dictIDFlag : %u ; dictID : %u ; dictIDSizeCode : %u " ,
! params - > fParams . noDictIDFlag , ( unsigned ) dictID , ( unsigned ) dictIDSizeCode ) ;
if ( params - > format = = ZSTD_f_zstd1 ) {
MEM_writeLE32 ( dst , ZSTD_MAGICNUMBER ) ;
pos = 4 ;
}
op [ pos + + ] = frameHeaderDescriptionByte ;
if ( ! singleSegment ) op [ pos + + ] = windowLogByte ;
switch ( dictIDSizeCode )
{
2021-12-21 11:21:29 +00:00
default :
assert ( 0 ) ; /* impossible */
ZSTD_FALLTHROUGH ;
2020-02-08 13:56:59 +00:00
case 0 : break ;
case 1 : op [ pos ] = ( BYTE ) ( dictID ) ; pos + + ; break ;
case 2 : MEM_writeLE16 ( op + pos , ( U16 ) dictID ) ; pos + = 2 ; break ;
case 3 : MEM_writeLE32 ( op + pos , dictID ) ; pos + = 4 ; break ;
}
switch ( fcsCode )
{
2021-12-21 11:21:29 +00:00
default :
assert ( 0 ) ; /* impossible */
ZSTD_FALLTHROUGH ;
2020-02-08 13:56:59 +00:00
case 0 : if ( singleSegment ) op [ pos + + ] = ( BYTE ) ( pledgedSrcSize ) ; break ;
case 1 : MEM_writeLE16 ( op + pos , ( U16 ) ( pledgedSrcSize - 256 ) ) ; pos + = 2 ; break ;
case 2 : MEM_writeLE32 ( op + pos , ( U32 ) ( pledgedSrcSize ) ) ; pos + = 4 ; break ;
case 3 : MEM_writeLE64 ( op + pos , ( U64 ) ( pledgedSrcSize ) ) ; pos + = 8 ; break ;
}
return pos ;
}
2021-05-14 22:48:15 +00:00
/* ZSTD_writeSkippableFrame_advanced() :
* Writes out a skippable frame with the specified magic number variant ( 16 are supported ) ,
* from ZSTD_MAGIC_SKIPPABLE_START to ZSTD_MAGIC_SKIPPABLE_START + 15 , and the desired source data .
*
* Returns the total number of bytes written , or a ZSTD error code .
*/
size_t ZSTD_writeSkippableFrame ( void * dst , size_t dstCapacity ,
const void * src , size_t srcSize , unsigned magicVariant ) {
BYTE * op = ( BYTE * ) dst ;
RETURN_ERROR_IF ( dstCapacity < srcSize + ZSTD_SKIPPABLEHEADERSIZE /* Skippable frame overhead */ ,
dstSize_tooSmall , " Not enough room for skippable frame " ) ;
RETURN_ERROR_IF ( srcSize > ( unsigned ) 0xFFFFFFFF , srcSize_wrong , " Src size too large for skippable frame " ) ;
RETURN_ERROR_IF ( magicVariant > 15 , parameter_outOfBound , " Skippable frame magic number variant not supported " ) ;
MEM_writeLE32 ( op , ( U32 ) ( ZSTD_MAGIC_SKIPPABLE_START + magicVariant ) ) ;
MEM_writeLE32 ( op + 4 , ( U32 ) srcSize ) ;
ZSTD_memcpy ( op + 8 , src , srcSize ) ;
return srcSize + ZSTD_SKIPPABLEHEADERSIZE ;
}
2020-02-08 13:56:59 +00:00
/* ZSTD_writeLastEmptyBlock() :
* output an empty Block with end - of - frame mark to complete a frame
* @ return : size of data written into ` dst ` ( = = ZSTD_blockHeaderSize ( defined in zstd_internal . h ) )
* or an error code if ` dstCapacity ` is too small ( < ZSTD_blockHeaderSize )
*/
size_t ZSTD_writeLastEmptyBlock ( void * dst , size_t dstCapacity )
{
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( dstCapacity < ZSTD_blockHeaderSize , dstSize_tooSmall ,
" dst buf is too small to write frame trailer empty block. " ) ;
2020-02-08 13:56:59 +00:00
{ U32 const cBlockHeader24 = 1 /*lastBlock*/ + ( ( ( U32 ) bt_raw ) < < 1 ) ; /* 0 size */
MEM_writeLE24 ( dst , cBlockHeader24 ) ;
return ZSTD_blockHeaderSize ;
}
}
size_t ZSTD_referenceExternalSequences ( ZSTD_CCtx * cctx , rawSeq * seq , size_t nbSeq )
{
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( cctx - > stage ! = ZSTDcs_init , stage_wrong ,
" wrong cctx stage " ) ;
2021-12-21 11:21:29 +00:00
RETURN_ERROR_IF ( cctx - > appliedParams . ldmParams . enableLdm = = ZSTD_ps_enable ,
2020-06-04 19:03:27 +00:00
parameter_unsupported ,
" incompatible with ldm " ) ;
2020-02-08 13:56:59 +00:00
cctx - > externSeqStore . seq = seq ;
cctx - > externSeqStore . size = nbSeq ;
cctx - > externSeqStore . capacity = nbSeq ;
cctx - > externSeqStore . pos = 0 ;
2020-12-18 14:46:27 +00:00
cctx - > externSeqStore . posInSequence = 0 ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
static size_t ZSTD_compressContinue_internal ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
U32 frame , U32 lastFrameChunk )
{
ZSTD_matchState_t * const ms = & cctx - > blockState . matchState ;
size_t fhSize = 0 ;
DEBUGLOG ( 5 , " ZSTD_compressContinue_internal, stage: %u, srcSize: %u " ,
cctx - > stage , ( unsigned ) srcSize ) ;
RETURN_ERROR_IF ( cctx - > stage = = ZSTDcs_created , stage_wrong ,
" missing init (ZSTD_compressBegin) " ) ;
if ( frame & & ( cctx - > stage = = ZSTDcs_init ) ) {
fhSize = ZSTD_writeFrameHeader ( dst , dstCapacity , & cctx - > appliedParams ,
cctx - > pledgedSrcSizePlusOne - 1 , cctx - > dictID ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( fhSize , " ZSTD_writeFrameHeader failed " ) ;
2020-02-08 13:56:59 +00:00
assert ( fhSize < = dstCapacity ) ;
dstCapacity - = fhSize ;
dst = ( char * ) dst + fhSize ;
cctx - > stage = ZSTDcs_ongoing ;
}
if ( ! srcSize ) return fhSize ; /* do not generate an empty block if no input */
2021-05-14 22:48:15 +00:00
if ( ! ZSTD_window_update ( & ms - > window , src , srcSize , ms - > forceNonContiguous ) ) {
ms - > forceNonContiguous = 0 ;
2020-02-08 13:56:59 +00:00
ms - > nextToUpdate = ms - > window . dictLimit ;
}
2021-12-21 11:21:29 +00:00
if ( cctx - > appliedParams . ldmParams . enableLdm = = ZSTD_ps_enable ) {
2021-05-14 22:48:15 +00:00
ZSTD_window_update ( & cctx - > ldmState . window , src , srcSize , /* forceNonContiguous */ 0 ) ;
2020-02-08 13:56:59 +00:00
}
if ( ! frame ) {
/* overflow check and correction for block mode */
ZSTD_overflowCorrectIfNeeded (
ms , & cctx - > workspace , & cctx - > appliedParams ,
src , ( BYTE const * ) src + srcSize ) ;
}
DEBUGLOG ( 5 , " ZSTD_compressContinue_internal (blockSize=%u) " , ( unsigned ) cctx - > blockSize ) ;
{ size_t const cSize = frame ?
ZSTD_compress_frameChunk ( cctx , dst , dstCapacity , src , srcSize , lastFrameChunk ) :
ZSTD_compressBlock_internal ( cctx , dst , dstCapacity , src , srcSize , 0 /* frame */ ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( cSize , " %s " , frame ? " ZSTD_compress_frameChunk failed " : " ZSTD_compressBlock_internal failed " ) ;
2020-02-08 13:56:59 +00:00
cctx - > consumedSrcSize + = srcSize ;
cctx - > producedCSize + = ( cSize + fhSize ) ;
assert ( ! ( cctx - > appliedParams . fParams . contentSizeFlag & & cctx - > pledgedSrcSizePlusOne = = 0 ) ) ;
if ( cctx - > pledgedSrcSizePlusOne ! = 0 ) { /* control src size */
ZSTD_STATIC_ASSERT ( ZSTD_CONTENTSIZE_UNKNOWN = = ( unsigned long long ) - 1 ) ;
RETURN_ERROR_IF (
cctx - > consumedSrcSize + 1 > cctx - > pledgedSrcSizePlusOne ,
srcSize_wrong ,
" error : pledgedSrcSize = %u, while realSrcSize >= %u " ,
( unsigned ) cctx - > pledgedSrcSizePlusOne - 1 ,
( unsigned ) cctx - > consumedSrcSize ) ;
}
return cSize + fhSize ;
}
}
size_t ZSTD_compressContinue ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize )
{
DEBUGLOG ( 5 , " ZSTD_compressContinue (srcSize=%u) " , ( unsigned ) srcSize ) ;
return ZSTD_compressContinue_internal ( cctx , dst , dstCapacity , src , srcSize , 1 /* frame mode */ , 0 /* last chunk */ ) ;
}
size_t ZSTD_getBlockSize ( const ZSTD_CCtx * cctx )
{
ZSTD_compressionParameters const cParams = cctx - > appliedParams . cParams ;
assert ( ! ZSTD_checkCParams ( cParams ) ) ;
2023-02-10 19:25:40 +00:00
return MIN ( cctx - > appliedParams . maxBlockSize , ( size_t ) 1 < < cParams . windowLog ) ;
2020-02-08 13:56:59 +00:00
}
size_t ZSTD_compressBlock ( ZSTD_CCtx * cctx , void * dst , size_t dstCapacity , const void * src , size_t srcSize )
{
DEBUGLOG ( 5 , " ZSTD_compressBlock: srcSize = %u " , ( unsigned ) srcSize ) ;
{ size_t const blockSizeMax = ZSTD_getBlockSize ( cctx ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( srcSize > blockSizeMax , srcSize_wrong , " input is larger than a block " ) ; }
2020-02-08 13:56:59 +00:00
return ZSTD_compressContinue_internal ( cctx , dst , dstCapacity , src , srcSize , 0 /* frame mode */ , 0 /* last chunk */ ) ;
}
/*! ZSTD_loadDictionaryContent() :
* @ return : 0 , or an error code
*/
static size_t ZSTD_loadDictionaryContent ( ZSTD_matchState_t * ms ,
2020-06-04 19:03:27 +00:00
ldmState_t * ls ,
2020-02-08 13:56:59 +00:00
ZSTD_cwksp * ws ,
ZSTD_CCtx_params const * params ,
const void * src , size_t srcSize ,
2023-02-10 19:25:40 +00:00
ZSTD_dictTableLoadMethod_e dtlm ,
ZSTD_tableFillPurpose_e tfp )
2020-02-08 13:56:59 +00:00
{
const BYTE * ip = ( const BYTE * ) src ;
const BYTE * const iend = ip + srcSize ;
2021-12-21 11:21:29 +00:00
int const loadLdmDict = params - > ldmParams . enableLdm = = ZSTD_ps_enable & & ls ! = NULL ;
2021-05-14 22:48:15 +00:00
2023-02-10 19:25:40 +00:00
/* Assert that the ms params match the params we're being given */
2021-05-14 22:48:15 +00:00
ZSTD_assertEqualCParams ( params - > cParams , ms - > cParams ) ;
2020-02-08 13:56:59 +00:00
2023-02-10 19:25:40 +00:00
{ /* Ensure large dictionaries can't cause index overflow */
2021-05-14 22:48:15 +00:00
/* Allow the dictionary to set indices up to exactly ZSTD_CURRENT_MAX.
* Dictionaries right at the edge will immediately trigger overflow
* correction , but I don ' t want to insert extra constraints here .
*/
2023-02-10 19:25:40 +00:00
U32 maxDictSize = ZSTD_CURRENT_MAX - ZSTD_WINDOW_START_INDEX ;
int const CDictTaggedIndices = ZSTD_CDictIndicesAreTagged ( & params - > cParams ) ;
if ( CDictTaggedIndices & & tfp = = ZSTD_tfp_forCDict ) {
/* Some dictionary matchfinders in zstd use "short cache",
* which treats the lower ZSTD_SHORT_CACHE_TAG_BITS of each
* CDict hashtable entry as a tag rather than as part of an index .
* When short cache is used , we need to truncate the dictionary
* so that its indices don ' t overlap with the tag . */
U32 const shortCacheMaxDictSize = ( 1u < < ( 32 - ZSTD_SHORT_CACHE_TAG_BITS ) ) - ZSTD_WINDOW_START_INDEX ;
maxDictSize = MIN ( maxDictSize , shortCacheMaxDictSize ) ;
assert ( ! loadLdmDict ) ;
}
2021-05-14 22:48:15 +00:00
/* If the dictionary is too large, only load the suffix of the dictionary. */
if ( srcSize > maxDictSize ) {
ip = iend - maxDictSize ;
src = ip ;
srcSize = maxDictSize ;
2023-02-10 19:25:40 +00:00
} }
if ( srcSize > ZSTD_CHUNKSIZE_MAX ) {
/* We must have cleared our windows when our source is this large. */
assert ( ZSTD_window_isEmpty ( ms - > window ) ) ;
if ( loadLdmDict ) assert ( ZSTD_window_isEmpty ( ls - > window ) ) ;
2021-05-14 22:48:15 +00:00
}
DEBUGLOG ( 4 , " ZSTD_loadDictionaryContent(): useRowMatchFinder=%d " , ( int ) params - > useRowMatchFinder ) ;
ZSTD_window_update ( & ms - > window , src , srcSize , /* forceNonContiguous */ 0 ) ;
2020-02-08 13:56:59 +00:00
ms - > loadedDictEnd = params - > forceWindow ? 0 : ( U32 ) ( iend - ms - > window . base ) ;
2021-05-14 22:48:15 +00:00
ms - > forceNonContiguous = params - > deterministicRefPrefix ;
2020-02-08 13:56:59 +00:00
2021-05-14 22:48:15 +00:00
if ( loadLdmDict ) {
ZSTD_window_update ( & ls - > window , src , srcSize , /* forceNonContiguous */ 0 ) ;
2020-06-04 19:03:27 +00:00
ls - > loadedDictEnd = params - > forceWindow ? 0 : ( U32 ) ( iend - ls - > window . base ) ;
}
2020-02-08 13:56:59 +00:00
if ( srcSize < = HASH_READ_SIZE ) return 0 ;
2021-05-14 22:48:15 +00:00
ZSTD_overflowCorrectIfNeeded ( ms , ws , params , ip , iend ) ;
2020-02-08 13:56:59 +00:00
2021-05-14 22:48:15 +00:00
if ( loadLdmDict )
ZSTD_ldm_fillHashTable ( ls , ip , iend , & params - > ldmParams ) ;
2020-02-08 13:56:59 +00:00
2021-05-14 22:48:15 +00:00
switch ( params - > cParams . strategy )
{
case ZSTD_fast :
2023-02-10 19:25:40 +00:00
ZSTD_fillHashTable ( ms , iend , dtlm , tfp ) ;
2021-05-14 22:48:15 +00:00
break ;
case ZSTD_dfast :
2023-02-10 19:25:40 +00:00
ZSTD_fillDoubleHashTable ( ms , iend , dtlm , tfp ) ;
2021-05-14 22:48:15 +00:00
break ;
2020-02-08 13:56:59 +00:00
2021-05-14 22:48:15 +00:00
case ZSTD_greedy :
case ZSTD_lazy :
case ZSTD_lazy2 :
assert ( srcSize > = HASH_READ_SIZE ) ;
if ( ms - > dedicatedDictSearch ) {
assert ( ms - > chainTable ! = NULL ) ;
ZSTD_dedicatedDictSearch_lazy_loadDictionary ( ms , iend - HASH_READ_SIZE ) ;
} else {
2021-12-21 11:21:29 +00:00
assert ( params - > useRowMatchFinder ! = ZSTD_ps_auto ) ;
if ( params - > useRowMatchFinder = = ZSTD_ps_enable ) {
2021-05-14 22:48:15 +00:00
size_t const tagTableSize = ( ( size_t ) 1 < < params - > cParams . hashLog ) * sizeof ( U16 ) ;
ZSTD_memset ( ms - > tagTable , 0 , tagTableSize ) ;
ZSTD_row_update ( ms , iend - HASH_READ_SIZE ) ;
DEBUGLOG ( 4 , " Using row-based hash table for lazy dict " ) ;
} else {
ZSTD_insertAndFindFirstIndex ( ms , iend - HASH_READ_SIZE ) ;
DEBUGLOG ( 4 , " Using chain-based hash table for lazy dict " ) ;
2020-12-18 14:46:27 +00:00
}
2020-02-08 13:56:59 +00:00
}
2021-05-14 22:48:15 +00:00
break ;
case ZSTD_btlazy2 : /* we want the dictionary table fully sorted */
case ZSTD_btopt :
case ZSTD_btultra :
case ZSTD_btultra2 :
assert ( srcSize > = HASH_READ_SIZE ) ;
ZSTD_updateTree ( ms , iend - HASH_READ_SIZE , iend ) ;
break ;
2020-02-08 13:56:59 +00:00
2021-05-14 22:48:15 +00:00
default :
assert ( 0 ) ; /* not possible : not a valid strategy id */
2020-02-08 13:56:59 +00:00
}
ms - > nextToUpdate = ( U32 ) ( iend - ms - > window . base ) ;
return 0 ;
}
/* Dictionaries that assign zero probability to symbols that show up causes problems
2020-12-18 14:46:27 +00:00
* when FSE encoding . Mark dictionaries with zero probability symbols as FSE_repeat_check
* and only dictionaries with 100 % valid symbols can be assumed valid .
*/
static FSE_repeat ZSTD_dictNCountRepeat ( short * normalizedCounter , unsigned dictMaxSymbolValue , unsigned maxSymbolValue )
{
2020-02-08 13:56:59 +00:00
U32 s ;
2020-12-18 14:46:27 +00:00
if ( dictMaxSymbolValue < maxSymbolValue ) {
return FSE_repeat_check ;
}
2020-02-08 13:56:59 +00:00
for ( s = 0 ; s < = maxSymbolValue ; + + s ) {
2020-12-18 14:46:27 +00:00
if ( normalizedCounter [ s ] = = 0 ) {
return FSE_repeat_check ;
}
2020-02-08 13:56:59 +00:00
}
2020-12-18 14:46:27 +00:00
return FSE_repeat_valid ;
2020-02-08 13:56:59 +00:00
}
size_t ZSTD_loadCEntropy ( ZSTD_compressedBlockState_t * bs , void * workspace ,
const void * const dict , size_t dictSize )
{
2020-12-18 14:46:27 +00:00
short offcodeNCount [ MaxOff + 1 ] ;
unsigned offcodeMaxValue = MaxOff ;
2020-02-08 13:56:59 +00:00
const BYTE * dictPtr = ( const BYTE * ) dict ; /* skip magic num and dict ID */
const BYTE * const dictEnd = dictPtr + dictSize ;
dictPtr + = 8 ;
bs - > entropy . huf . repeatMode = HUF_repeat_check ;
{ unsigned maxSymbolValue = 255 ;
unsigned hasZeroWeights = 1 ;
size_t const hufHeaderSize = HUF_readCTable ( ( HUF_CElt * ) bs - > entropy . huf . CTable , & maxSymbolValue , dictPtr ,
dictEnd - dictPtr , & hasZeroWeights ) ;
/* We only set the loaded table as valid if it contains all non-zero
* weights . Otherwise , we set it to check */
if ( ! hasZeroWeights )
bs - > entropy . huf . repeatMode = HUF_repeat_valid ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( HUF_isError ( hufHeaderSize ) , dictionary_corrupted , " " ) ;
RETURN_ERROR_IF ( maxSymbolValue < 255 , dictionary_corrupted , " " ) ;
2020-02-08 13:56:59 +00:00
dictPtr + = hufHeaderSize ;
}
{ unsigned offcodeLog ;
2020-12-18 14:46:27 +00:00
size_t const offcodeHeaderSize = FSE_readNCount ( offcodeNCount , & offcodeMaxValue , & offcodeLog , dictPtr , dictEnd - dictPtr ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( FSE_isError ( offcodeHeaderSize ) , dictionary_corrupted , " " ) ;
RETURN_ERROR_IF ( offcodeLog > OffFSELog , dictionary_corrupted , " " ) ;
2020-02-08 13:56:59 +00:00
/* fill all offset symbols to avoid garbage at end of table */
RETURN_ERROR_IF ( FSE_isError ( FSE_buildCTable_wksp (
bs - > entropy . fse . offcodeCTable ,
offcodeNCount , MaxOff , offcodeLog ,
workspace , HUF_WORKSPACE_SIZE ) ) ,
2020-06-04 19:03:27 +00:00
dictionary_corrupted , " " ) ;
2020-12-18 14:46:27 +00:00
/* Defer checking offcodeMaxValue because we need to know the size of the dictionary content */
2020-02-08 13:56:59 +00:00
dictPtr + = offcodeHeaderSize ;
}
{ short matchlengthNCount [ MaxML + 1 ] ;
unsigned matchlengthMaxValue = MaxML , matchlengthLog ;
size_t const matchlengthHeaderSize = FSE_readNCount ( matchlengthNCount , & matchlengthMaxValue , & matchlengthLog , dictPtr , dictEnd - dictPtr ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( FSE_isError ( matchlengthHeaderSize ) , dictionary_corrupted , " " ) ;
RETURN_ERROR_IF ( matchlengthLog > MLFSELog , dictionary_corrupted , " " ) ;
2020-02-08 13:56:59 +00:00
RETURN_ERROR_IF ( FSE_isError ( FSE_buildCTable_wksp (
bs - > entropy . fse . matchlengthCTable ,
matchlengthNCount , matchlengthMaxValue , matchlengthLog ,
workspace , HUF_WORKSPACE_SIZE ) ) ,
2020-06-04 19:03:27 +00:00
dictionary_corrupted , " " ) ;
2020-12-18 14:46:27 +00:00
bs - > entropy . fse . matchlength_repeatMode = ZSTD_dictNCountRepeat ( matchlengthNCount , matchlengthMaxValue , MaxML ) ;
2020-02-08 13:56:59 +00:00
dictPtr + = matchlengthHeaderSize ;
}
{ short litlengthNCount [ MaxLL + 1 ] ;
unsigned litlengthMaxValue = MaxLL , litlengthLog ;
size_t const litlengthHeaderSize = FSE_readNCount ( litlengthNCount , & litlengthMaxValue , & litlengthLog , dictPtr , dictEnd - dictPtr ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( FSE_isError ( litlengthHeaderSize ) , dictionary_corrupted , " " ) ;
RETURN_ERROR_IF ( litlengthLog > LLFSELog , dictionary_corrupted , " " ) ;
2020-02-08 13:56:59 +00:00
RETURN_ERROR_IF ( FSE_isError ( FSE_buildCTable_wksp (
bs - > entropy . fse . litlengthCTable ,
litlengthNCount , litlengthMaxValue , litlengthLog ,
workspace , HUF_WORKSPACE_SIZE ) ) ,
2020-06-04 19:03:27 +00:00
dictionary_corrupted , " " ) ;
2020-12-18 14:46:27 +00:00
bs - > entropy . fse . litlength_repeatMode = ZSTD_dictNCountRepeat ( litlengthNCount , litlengthMaxValue , MaxLL ) ;
2020-02-08 13:56:59 +00:00
dictPtr + = litlengthHeaderSize ;
}
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( dictPtr + 12 > dictEnd , dictionary_corrupted , " " ) ;
2020-02-08 13:56:59 +00:00
bs - > rep [ 0 ] = MEM_readLE32 ( dictPtr + 0 ) ;
bs - > rep [ 1 ] = MEM_readLE32 ( dictPtr + 4 ) ;
bs - > rep [ 2 ] = MEM_readLE32 ( dictPtr + 8 ) ;
dictPtr + = 12 ;
2020-12-18 14:46:27 +00:00
{ size_t const dictContentSize = ( size_t ) ( dictEnd - dictPtr ) ;
U32 offcodeMax = MaxOff ;
if ( dictContentSize < = ( ( U32 ) - 1 ) - 128 KB ) {
U32 const maxOffset = ( U32 ) dictContentSize + 128 KB ; /* The maximum offset that must be supported */
offcodeMax = ZSTD_highbit32 ( maxOffset ) ; /* Calculate minimum offset code required to represent maxOffset */
}
/* All offset values <= dictContentSize + 128 KB must be representable for a valid table */
bs - > entropy . fse . offcode_repeatMode = ZSTD_dictNCountRepeat ( offcodeNCount , offcodeMaxValue , MIN ( offcodeMax , MaxOff ) ) ;
/* All repCodes must be <= dictContentSize and != 0 */
{ U32 u ;
for ( u = 0 ; u < 3 ; u + + ) {
RETURN_ERROR_IF ( bs - > rep [ u ] = = 0 , dictionary_corrupted , " " ) ;
RETURN_ERROR_IF ( bs - > rep [ u ] > dictContentSize , dictionary_corrupted , " " ) ;
} } }
2020-02-08 13:56:59 +00:00
return dictPtr - ( const BYTE * ) dict ;
}
/* Dictionary format :
* See :
2021-05-14 22:48:15 +00:00
* https : //github.com/facebook/zstd/blob/release/doc/zstd_compression_format.md#dictionary-format
2020-02-08 13:56:59 +00:00
*/
/*! ZSTD_loadZstdDictionary() :
* @ return : dictID , or an error code
* assumptions : magic number supposed already checked
* dictSize supposed > = 8
*/
static size_t ZSTD_loadZstdDictionary ( ZSTD_compressedBlockState_t * bs ,
ZSTD_matchState_t * ms ,
ZSTD_cwksp * ws ,
ZSTD_CCtx_params const * params ,
const void * dict , size_t dictSize ,
ZSTD_dictTableLoadMethod_e dtlm ,
2023-02-10 19:25:40 +00:00
ZSTD_tableFillPurpose_e tfp ,
2020-02-08 13:56:59 +00:00
void * workspace )
{
const BYTE * dictPtr = ( const BYTE * ) dict ;
const BYTE * const dictEnd = dictPtr + dictSize ;
size_t dictID ;
size_t eSize ;
ZSTD_STATIC_ASSERT ( HUF_WORKSPACE_SIZE > = ( 1 < < MAX ( MLFSELog , LLFSELog ) ) ) ;
assert ( dictSize > = 8 ) ;
assert ( MEM_readLE32 ( dictPtr ) = = ZSTD_MAGIC_DICTIONARY ) ;
dictID = params - > fParams . noDictIDFlag ? 0 : MEM_readLE32 ( dictPtr + 4 /* skip magic number */ ) ;
2020-12-18 14:46:27 +00:00
eSize = ZSTD_loadCEntropy ( bs , workspace , dict , dictSize ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( eSize , " ZSTD_loadCEntropy failed " ) ;
2020-02-08 13:56:59 +00:00
dictPtr + = eSize ;
2020-12-18 14:46:27 +00:00
{
size_t const dictContentSize = ( size_t ) ( dictEnd - dictPtr ) ;
2020-02-08 13:56:59 +00:00
FORWARD_IF_ERROR ( ZSTD_loadDictionaryContent (
2023-02-10 19:25:40 +00:00
ms , NULL , ws , params , dictPtr , dictContentSize , dtlm , tfp ) , " " ) ;
2020-02-08 13:56:59 +00:00
}
2020-12-18 14:46:27 +00:00
return dictID ;
2020-02-08 13:56:59 +00:00
}
/** ZSTD_compress_insertDictionary() :
* @ return : dictID , or an error code */
static size_t
ZSTD_compress_insertDictionary ( ZSTD_compressedBlockState_t * bs ,
ZSTD_matchState_t * ms ,
2020-06-04 19:03:27 +00:00
ldmState_t * ls ,
2020-02-08 13:56:59 +00:00
ZSTD_cwksp * ws ,
const ZSTD_CCtx_params * params ,
const void * dict , size_t dictSize ,
ZSTD_dictContentType_e dictContentType ,
ZSTD_dictTableLoadMethod_e dtlm ,
2023-02-10 19:25:40 +00:00
ZSTD_tableFillPurpose_e tfp ,
2020-02-08 13:56:59 +00:00
void * workspace )
{
DEBUGLOG ( 4 , " ZSTD_compress_insertDictionary (dictSize=%u) " , ( U32 ) dictSize ) ;
if ( ( dict = = NULL ) | | ( dictSize < 8 ) ) {
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( dictContentType = = ZSTD_dct_fullDict , dictionary_wrong , " " ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
ZSTD_reset_compressedBlockState ( bs ) ;
/* dict restricted modes */
if ( dictContentType = = ZSTD_dct_rawContent )
2023-02-10 19:25:40 +00:00
return ZSTD_loadDictionaryContent ( ms , ls , ws , params , dict , dictSize , dtlm , tfp ) ;
2020-02-08 13:56:59 +00:00
if ( MEM_readLE32 ( dict ) ! = ZSTD_MAGIC_DICTIONARY ) {
if ( dictContentType = = ZSTD_dct_auto ) {
DEBUGLOG ( 4 , " raw content dictionary detected " ) ;
return ZSTD_loadDictionaryContent (
2023-02-10 19:25:40 +00:00
ms , ls , ws , params , dict , dictSize , dtlm , tfp ) ;
2020-02-08 13:56:59 +00:00
}
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( dictContentType = = ZSTD_dct_fullDict , dictionary_wrong , " " ) ;
2020-02-08 13:56:59 +00:00
assert ( 0 ) ; /* impossible */
}
/* dict as full zstd dictionary */
return ZSTD_loadZstdDictionary (
2023-02-10 19:25:40 +00:00
bs , ms , ws , params , dict , dictSize , dtlm , tfp , workspace ) ;
2020-02-08 13:56:59 +00:00
}
# define ZSTD_USE_CDICT_PARAMS_SRCSIZE_CUTOFF (128 KB)
2020-12-18 14:46:27 +00:00
# define ZSTD_USE_CDICT_PARAMS_DICTSIZE_MULTIPLIER (6ULL)
2020-02-08 13:56:59 +00:00
/*! ZSTD_compressBegin_internal() :
2023-02-10 19:25:40 +00:00
* Assumption : either @ dict OR @ cdict ( or none ) is non - NULL , never both
2020-02-08 13:56:59 +00:00
* @ return : 0 , or an error code */
static size_t ZSTD_compressBegin_internal ( ZSTD_CCtx * cctx ,
const void * dict , size_t dictSize ,
ZSTD_dictContentType_e dictContentType ,
ZSTD_dictTableLoadMethod_e dtlm ,
const ZSTD_CDict * cdict ,
const ZSTD_CCtx_params * params , U64 pledgedSrcSize ,
ZSTD_buffered_policy_e zbuff )
{
2021-05-14 22:48:15 +00:00
size_t const dictContentSize = cdict ? cdict - > dictContentSize : dictSize ;
# if ZSTD_TRACE
cctx - > traceCtx = ( ZSTD_trace_compress_begin ! = NULL ) ? ZSTD_trace_compress_begin ( cctx ) : 0 ;
# endif
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 4 , " ZSTD_compressBegin_internal: wlog=%u " , params - > cParams . windowLog ) ;
/* params are supposed to be fully validated at this point */
assert ( ! ZSTD_isError ( ZSTD_checkCParams ( params - > cParams ) ) ) ;
assert ( ! ( ( dict ) & & ( cdict ) ) ) ; /* either dict or cdict, not both */
if ( ( cdict )
& & ( 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 ) ;
}
2021-05-14 22:48:15 +00:00
FORWARD_IF_ERROR ( ZSTD_resetCCtx_internal ( cctx , params , pledgedSrcSize ,
dictContentSize ,
2020-06-04 19:03:27 +00:00
ZSTDcrp_makeClean , zbuff ) , " " ) ;
2020-02-08 13:56:59 +00:00
{ size_t const dictID = cdict ?
ZSTD_compress_insertDictionary (
cctx - > blockState . prevCBlock , & cctx - > blockState . matchState ,
2020-06-04 19:03:27 +00:00
& cctx - > ldmState , & cctx - > workspace , & cctx - > appliedParams , cdict - > dictContent ,
2020-12-18 14:46:27 +00:00
cdict - > dictContentSize , cdict - > dictContentType , dtlm ,
2023-02-10 19:25:40 +00:00
ZSTD_tfp_forCCtx , cctx - > entropyWorkspace )
2020-02-08 13:56:59 +00:00
: ZSTD_compress_insertDictionary (
cctx - > blockState . prevCBlock , & cctx - > blockState . matchState ,
2020-06-04 19:03:27 +00:00
& cctx - > ldmState , & cctx - > workspace , & cctx - > appliedParams , dict , dictSize ,
2023-02-10 19:25:40 +00:00
dictContentType , dtlm , ZSTD_tfp_forCCtx , cctx - > entropyWorkspace ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( dictID , " ZSTD_compress_insertDictionary failed " ) ;
2020-02-08 13:56:59 +00:00
assert ( dictID < = UINT_MAX ) ;
cctx - > dictID = ( U32 ) dictID ;
2021-05-14 22:48:15 +00:00
cctx - > dictContentSize = dictContentSize ;
2020-02-08 13:56:59 +00:00
}
return 0 ;
}
size_t ZSTD_compressBegin_advanced_internal ( ZSTD_CCtx * cctx ,
const void * dict , size_t dictSize ,
ZSTD_dictContentType_e dictContentType ,
ZSTD_dictTableLoadMethod_e dtlm ,
const ZSTD_CDict * cdict ,
const ZSTD_CCtx_params * params ,
unsigned long long pledgedSrcSize )
{
DEBUGLOG ( 4 , " ZSTD_compressBegin_advanced_internal: wlog=%u " , params - > cParams . windowLog ) ;
/* compression parameters verification and optimization */
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_checkCParams ( params - > cParams ) , " " ) ;
2020-02-08 13:56:59 +00:00
return ZSTD_compressBegin_internal ( cctx ,
dict , dictSize , dictContentType , dtlm ,
cdict ,
params , pledgedSrcSize ,
ZSTDb_not_buffered ) ;
}
/*! ZSTD_compressBegin_advanced() :
* @ return : 0 , or an error code */
size_t ZSTD_compressBegin_advanced ( ZSTD_CCtx * cctx ,
const void * dict , size_t dictSize ,
ZSTD_parameters params , unsigned long long pledgedSrcSize )
{
2021-05-14 22:48:15 +00:00
ZSTD_CCtx_params cctxParams ;
ZSTD_CCtxParams_init_internal ( & cctxParams , & params , ZSTD_NO_CLEVEL ) ;
2020-02-08 13:56:59 +00:00
return ZSTD_compressBegin_advanced_internal ( cctx ,
dict , dictSize , ZSTD_dct_auto , ZSTD_dtlm_fast ,
NULL /*cdict*/ ,
& cctxParams , pledgedSrcSize ) ;
}
2023-02-10 19:25:40 +00:00
size_t
ZSTD_compressBegin_usingDict ( ZSTD_CCtx * cctx , const void * dict , size_t dictSize , int compressionLevel )
2020-02-08 13:56:59 +00:00
{
2021-05-14 22:48:15 +00:00
ZSTD_CCtx_params cctxParams ;
2023-02-10 19:25:40 +00:00
{ ZSTD_parameters const params = ZSTD_getParams_internal ( compressionLevel , ZSTD_CONTENTSIZE_UNKNOWN , dictSize , ZSTD_cpm_noAttachDict ) ;
2021-05-14 22:48:15 +00:00
ZSTD_CCtxParams_init_internal ( & cctxParams , & params , ( compressionLevel = = 0 ) ? ZSTD_CLEVEL_DEFAULT : compressionLevel ) ;
}
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 4 , " ZSTD_compressBegin_usingDict (dictSize=%u) " , ( unsigned ) dictSize ) ;
return ZSTD_compressBegin_internal ( cctx , dict , dictSize , ZSTD_dct_auto , ZSTD_dtlm_fast , NULL ,
& cctxParams , ZSTD_CONTENTSIZE_UNKNOWN , ZSTDb_not_buffered ) ;
}
size_t ZSTD_compressBegin ( ZSTD_CCtx * cctx , int compressionLevel )
{
return ZSTD_compressBegin_usingDict ( cctx , NULL , 0 , compressionLevel ) ;
}
/*! ZSTD_writeEpilogue() :
* Ends a frame .
* @ return : nb of bytes written into dst ( or an error code ) */
static size_t ZSTD_writeEpilogue ( ZSTD_CCtx * cctx , void * dst , size_t dstCapacity )
{
BYTE * const ostart = ( BYTE * ) dst ;
BYTE * op = ostart ;
size_t fhSize = 0 ;
DEBUGLOG ( 4 , " ZSTD_writeEpilogue " ) ;
RETURN_ERROR_IF ( cctx - > stage = = ZSTDcs_created , stage_wrong , " init missing " ) ;
/* special case : empty frame */
if ( cctx - > stage = = ZSTDcs_init ) {
fhSize = ZSTD_writeFrameHeader ( dst , dstCapacity , & cctx - > appliedParams , 0 , 0 ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( fhSize , " ZSTD_writeFrameHeader failed " ) ;
2020-02-08 13:56:59 +00:00
dstCapacity - = fhSize ;
op + = fhSize ;
cctx - > stage = ZSTDcs_ongoing ;
}
if ( cctx - > stage ! = ZSTDcs_ending ) {
/* write one last empty block, make it the "last" block */
U32 const cBlockHeader24 = 1 /* last block */ + ( ( ( U32 ) bt_raw ) < < 1 ) + 0 ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( dstCapacity < 4 , dstSize_tooSmall , " no room for epilogue " ) ;
2020-02-08 13:56:59 +00:00
MEM_writeLE32 ( op , cBlockHeader24 ) ;
op + = ZSTD_blockHeaderSize ;
dstCapacity - = ZSTD_blockHeaderSize ;
}
if ( cctx - > appliedParams . fParams . checksumFlag ) {
U32 const checksum = ( U32 ) XXH64_digest ( & cctx - > xxhState ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( dstCapacity < 4 , dstSize_tooSmall , " no room for checksum " ) ;
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 4 , " ZSTD_writeEpilogue: write checksum : %08X " , ( unsigned ) checksum ) ;
MEM_writeLE32 ( op , checksum ) ;
op + = 4 ;
}
cctx - > stage = ZSTDcs_created ; /* return to "created but no init" status */
return op - ostart ;
}
2021-05-14 22:48:15 +00:00
void ZSTD_CCtx_trace ( ZSTD_CCtx * cctx , size_t extraCSize )
{
# if ZSTD_TRACE
if ( cctx - > traceCtx & & ZSTD_trace_compress_end ! = NULL ) {
int const streaming = cctx - > inBuffSize > 0 | | cctx - > outBuffSize > 0 | | cctx - > appliedParams . nbWorkers > 0 ;
ZSTD_Trace trace ;
ZSTD_memset ( & trace , 0 , sizeof ( trace ) ) ;
trace . version = ZSTD_VERSION_NUMBER ;
trace . streaming = streaming ;
trace . dictionaryID = cctx - > dictID ;
trace . dictionarySize = cctx - > dictContentSize ;
trace . uncompressedSize = cctx - > consumedSrcSize ;
trace . compressedSize = cctx - > producedCSize + extraCSize ;
trace . params = & cctx - > appliedParams ;
trace . cctx = cctx ;
ZSTD_trace_compress_end ( cctx - > traceCtx , & trace ) ;
}
cctx - > traceCtx = 0 ;
# else
( void ) cctx ;
( void ) extraCSize ;
# endif
}
2020-02-08 13:56:59 +00:00
size_t ZSTD_compressEnd ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize )
{
size_t endResult ;
size_t const cSize = ZSTD_compressContinue_internal ( cctx ,
dst , dstCapacity , src , srcSize ,
1 /* frame mode */ , 1 /* last chunk */ ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_compressContinue_internal failed " ) ;
2020-02-08 13:56:59 +00:00
endResult = ZSTD_writeEpilogue ( cctx , ( char * ) dst + cSize , dstCapacity - cSize ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( endResult , " ZSTD_writeEpilogue failed " ) ;
2020-02-08 13:56:59 +00:00
assert ( ! ( cctx - > appliedParams . fParams . contentSizeFlag & & cctx - > pledgedSrcSizePlusOne = = 0 ) ) ;
if ( cctx - > pledgedSrcSizePlusOne ! = 0 ) { /* control src size */
ZSTD_STATIC_ASSERT ( ZSTD_CONTENTSIZE_UNKNOWN = = ( unsigned long long ) - 1 ) ;
DEBUGLOG ( 4 , " end of frame : controlling src size " ) ;
RETURN_ERROR_IF (
cctx - > pledgedSrcSizePlusOne ! = cctx - > consumedSrcSize + 1 ,
srcSize_wrong ,
" error : pledgedSrcSize = %u, while realSrcSize = %u " ,
( unsigned ) cctx - > pledgedSrcSizePlusOne - 1 ,
( unsigned ) cctx - > consumedSrcSize ) ;
}
2021-05-14 22:48:15 +00:00
ZSTD_CCtx_trace ( cctx , endResult ) ;
2020-02-08 13:56:59 +00:00
return cSize + endResult ;
}
size_t ZSTD_compress_advanced ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
const void * dict , size_t dictSize ,
ZSTD_parameters params )
{
DEBUGLOG ( 4 , " ZSTD_compress_advanced " ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_checkCParams ( params . cParams ) , " " ) ;
2021-05-14 22:48:15 +00:00
ZSTD_CCtxParams_init_internal ( & cctx - > simpleApiParams , & params , ZSTD_NO_CLEVEL ) ;
return ZSTD_compress_advanced_internal ( cctx ,
dst , dstCapacity ,
src , srcSize ,
dict , dictSize ,
& cctx - > simpleApiParams ) ;
2020-02-08 13:56:59 +00:00
}
/* Internal */
size_t ZSTD_compress_advanced_internal (
ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
const void * dict , size_t dictSize ,
const ZSTD_CCtx_params * params )
{
DEBUGLOG ( 4 , " ZSTD_compress_advanced_internal (srcSize:%u) " , ( unsigned ) srcSize ) ;
FORWARD_IF_ERROR ( ZSTD_compressBegin_internal ( cctx ,
dict , dictSize , ZSTD_dct_auto , ZSTD_dtlm_fast , NULL ,
2020-06-04 19:03:27 +00:00
params , srcSize , ZSTDb_not_buffered ) , " " ) ;
2020-02-08 13:56:59 +00:00
return ZSTD_compressEnd ( cctx , dst , dstCapacity , src , srcSize ) ;
}
size_t ZSTD_compress_usingDict ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
const void * dict , size_t dictSize ,
int compressionLevel )
{
2021-05-14 22:48:15 +00:00
{
ZSTD_parameters const params = ZSTD_getParams_internal ( compressionLevel , srcSize , dict ? dictSize : 0 , ZSTD_cpm_noAttachDict ) ;
assert ( params . fParams . contentSizeFlag = = 1 ) ;
ZSTD_CCtxParams_init_internal ( & cctx - > simpleApiParams , & params , ( compressionLevel = = 0 ) ? ZSTD_CLEVEL_DEFAULT : compressionLevel ) ;
}
2020-06-04 19:03:27 +00:00
DEBUGLOG ( 4 , " ZSTD_compress_usingDict (srcSize=%u) " , ( unsigned ) srcSize ) ;
2021-05-14 22:48:15 +00:00
return ZSTD_compress_advanced_internal ( cctx , dst , dstCapacity , src , srcSize , dict , dictSize , & cctx - > simpleApiParams ) ;
2020-02-08 13:56:59 +00:00
}
size_t ZSTD_compressCCtx ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
int compressionLevel )
{
DEBUGLOG ( 4 , " ZSTD_compressCCtx (srcSize=%u) " , ( unsigned ) srcSize ) ;
assert ( cctx ! = NULL ) ;
return ZSTD_compress_usingDict ( cctx , dst , dstCapacity , src , srcSize , NULL , 0 , compressionLevel ) ;
}
size_t ZSTD_compress ( void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
int compressionLevel )
{
size_t result ;
2020-12-18 14:46:27 +00:00
# if ZSTD_COMPRESS_HEAPMODE
ZSTD_CCtx * cctx = ZSTD_createCCtx ( ) ;
RETURN_ERROR_IF ( ! cctx , memory_allocation , " ZSTD_createCCtx failed " ) ;
result = ZSTD_compressCCtx ( cctx , dst , dstCapacity , src , srcSize , compressionLevel ) ;
ZSTD_freeCCtx ( cctx ) ;
# else
2020-02-08 13:56:59 +00:00
ZSTD_CCtx ctxBody ;
ZSTD_initCCtx ( & ctxBody , ZSTD_defaultCMem ) ;
result = ZSTD_compressCCtx ( & ctxBody , dst , dstCapacity , src , srcSize , compressionLevel ) ;
ZSTD_freeCCtxContent ( & ctxBody ) ; /* can't free ctxBody itself, as it's on stack; free only heap content */
2020-12-18 14:46:27 +00:00
# endif
2020-02-08 13:56:59 +00:00
return result ;
}
/* ===== Dictionary API ===== */
/*! ZSTD_estimateCDictSize_advanced() :
* Estimate amount of memory that will be needed to create a dictionary with following arguments */
size_t ZSTD_estimateCDictSize_advanced (
size_t dictSize , ZSTD_compressionParameters cParams ,
ZSTD_dictLoadMethod_e dictLoadMethod )
{
DEBUGLOG ( 5 , " sizeof(ZSTD_CDict) : %u " , ( unsigned ) sizeof ( ZSTD_CDict ) ) ;
return ZSTD_cwksp_alloc_size ( sizeof ( ZSTD_CDict ) )
+ ZSTD_cwksp_alloc_size ( HUF_WORKSPACE_SIZE )
2021-05-14 22:48:15 +00:00
/* enableDedicatedDictSearch == 1 ensures that CDict estimation will not be too small
* in case we are using DDS with row - hash . */
2021-12-21 11:21:29 +00:00
+ ZSTD_sizeof_matchState ( & cParams , ZSTD_resolveRowMatchFinderMode ( ZSTD_ps_auto , & cParams ) ,
2021-05-14 22:48:15 +00:00
/* enableDedicatedDictSearch */ 1 , /* forCCtx */ 0 )
2020-02-08 13:56:59 +00:00
+ ( dictLoadMethod = = ZSTD_dlm_byRef ? 0
: ZSTD_cwksp_alloc_size ( ZSTD_cwksp_align ( dictSize , sizeof ( void * ) ) ) ) ;
}
size_t ZSTD_estimateCDictSize ( size_t dictSize , int compressionLevel )
{
2020-12-18 14:46:27 +00:00
ZSTD_compressionParameters const cParams = ZSTD_getCParams_internal ( compressionLevel , ZSTD_CONTENTSIZE_UNKNOWN , dictSize , ZSTD_cpm_createCDict ) ;
2020-02-08 13:56:59 +00:00
return ZSTD_estimateCDictSize_advanced ( dictSize , cParams , ZSTD_dlm_byCopy ) ;
}
size_t ZSTD_sizeof_CDict ( const ZSTD_CDict * cdict )
{
if ( cdict = = NULL ) return 0 ; /* support sizeof on NULL */
DEBUGLOG ( 5 , " sizeof(*cdict) : %u " , ( unsigned ) sizeof ( * cdict ) ) ;
/* cdict may be in the workspace */
return ( cdict - > workspace . workspace = = cdict ? 0 : sizeof ( * cdict ) )
+ ZSTD_cwksp_sizeof ( & cdict - > workspace ) ;
}
static size_t ZSTD_initCDict_internal (
ZSTD_CDict * cdict ,
const void * dictBuffer , size_t dictSize ,
ZSTD_dictLoadMethod_e dictLoadMethod ,
ZSTD_dictContentType_e dictContentType ,
2020-12-18 14:46:27 +00:00
ZSTD_CCtx_params params )
2020-02-08 13:56:59 +00:00
{
DEBUGLOG ( 3 , " ZSTD_initCDict_internal (dictContentType:%u) " , ( unsigned ) dictContentType ) ;
2020-12-18 14:46:27 +00:00
assert ( ! ZSTD_checkCParams ( params . cParams ) ) ;
cdict - > matchState . cParams = params . cParams ;
cdict - > matchState . dedicatedDictSearch = params . enableDedicatedDictSearch ;
2020-02-08 13:56:59 +00:00
if ( ( dictLoadMethod = = ZSTD_dlm_byRef ) | | ( ! dictBuffer ) | | ( ! dictSize ) ) {
cdict - > dictContent = dictBuffer ;
} else {
void * internalBuffer = ZSTD_cwksp_reserve_object ( & cdict - > workspace , ZSTD_cwksp_align ( dictSize , sizeof ( void * ) ) ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( ! internalBuffer , memory_allocation , " NULL pointer! " ) ;
2020-02-08 13:56:59 +00:00
cdict - > dictContent = internalBuffer ;
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( internalBuffer , dictBuffer , dictSize ) ;
2020-02-08 13:56:59 +00:00
}
cdict - > dictContentSize = dictSize ;
2020-12-18 14:46:27 +00:00
cdict - > dictContentType = dictContentType ;
2020-02-08 13:56:59 +00:00
cdict - > entropyWorkspace = ( U32 * ) ZSTD_cwksp_reserve_object ( & cdict - > workspace , HUF_WORKSPACE_SIZE ) ;
/* Reset the state to no dictionary */
ZSTD_reset_compressedBlockState ( & cdict - > cBlockState ) ;
FORWARD_IF_ERROR ( ZSTD_reset_matchState (
& cdict - > matchState ,
& cdict - > workspace ,
2020-12-18 14:46:27 +00:00
& params . cParams ,
2021-05-14 22:48:15 +00:00
params . useRowMatchFinder ,
2020-02-08 13:56:59 +00:00
ZSTDcrp_makeClean ,
ZSTDirp_reset ,
2020-06-04 19:03:27 +00:00
ZSTD_resetTarget_CDict ) , " " ) ;
2020-02-08 13:56:59 +00:00
/* (Maybe) load the dictionary
* Skips loading the dictionary if it is < 8 bytes .
*/
2020-12-18 14:46:27 +00:00
{ params . compressionLevel = ZSTD_CLEVEL_DEFAULT ;
2020-02-08 13:56:59 +00:00
params . fParams . contentSizeFlag = 1 ;
{ size_t const dictID = ZSTD_compress_insertDictionary (
2020-06-04 19:03:27 +00:00
& cdict - > cBlockState , & cdict - > matchState , NULL , & cdict - > workspace ,
2020-02-08 13:56:59 +00:00
& params , cdict - > dictContent , cdict - > dictContentSize ,
2023-02-10 19:25:40 +00:00
dictContentType , ZSTD_dtlm_full , ZSTD_tfp_forCDict , cdict - > entropyWorkspace ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( dictID , " ZSTD_compress_insertDictionary failed " ) ;
2020-02-08 13:56:59 +00:00
assert ( dictID < = ( size_t ) ( U32 ) - 1 ) ;
cdict - > dictID = ( U32 ) dictID ;
}
}
return 0 ;
}
2020-12-18 14:46:27 +00:00
static ZSTD_CDict * ZSTD_createCDict_advanced_internal ( size_t dictSize ,
2020-02-08 13:56:59 +00:00
ZSTD_dictLoadMethod_e dictLoadMethod ,
2021-05-14 22:48:15 +00:00
ZSTD_compressionParameters cParams ,
2021-12-21 11:21:29 +00:00
ZSTD_paramSwitch_e useRowMatchFinder ,
2021-05-14 22:48:15 +00:00
U32 enableDedicatedDictSearch ,
ZSTD_customMem customMem )
2020-02-08 13:56:59 +00:00
{
2020-12-18 14:46:27 +00:00
if ( ( ! customMem . customAlloc ) ^ ( ! customMem . customFree ) ) return NULL ;
2020-02-08 13:56:59 +00:00
{ size_t const workspaceSize =
ZSTD_cwksp_alloc_size ( sizeof ( ZSTD_CDict ) ) +
ZSTD_cwksp_alloc_size ( HUF_WORKSPACE_SIZE ) +
2021-05-14 22:48:15 +00:00
ZSTD_sizeof_matchState ( & cParams , useRowMatchFinder , enableDedicatedDictSearch , /* forCCtx */ 0 ) +
2020-02-08 13:56:59 +00:00
( dictLoadMethod = = ZSTD_dlm_byRef ? 0
: ZSTD_cwksp_alloc_size ( ZSTD_cwksp_align ( dictSize , sizeof ( void * ) ) ) ) ;
2020-12-18 14:46:27 +00:00
void * const workspace = ZSTD_customMalloc ( workspaceSize , customMem ) ;
2020-02-08 13:56:59 +00:00
ZSTD_cwksp ws ;
ZSTD_CDict * cdict ;
if ( ! workspace ) {
2020-12-18 14:46:27 +00:00
ZSTD_customFree ( workspace , customMem ) ;
2020-02-08 13:56:59 +00:00
return NULL ;
}
2020-12-18 14:46:27 +00:00
ZSTD_cwksp_init ( & ws , workspace , workspaceSize , ZSTD_cwksp_dynamic_alloc ) ;
2020-02-08 13:56:59 +00:00
cdict = ( ZSTD_CDict * ) ZSTD_cwksp_reserve_object ( & ws , sizeof ( ZSTD_CDict ) ) ;
assert ( cdict ! = NULL ) ;
ZSTD_cwksp_move ( & cdict - > workspace , & ws ) ;
cdict - > customMem = customMem ;
2021-05-14 22:48:15 +00:00
cdict - > compressionLevel = ZSTD_NO_CLEVEL ; /* signals advanced API usage */
cdict - > useRowMatchFinder = useRowMatchFinder ;
2020-02-08 13:56:59 +00:00
return cdict ;
}
}
2020-12-18 14:46:27 +00:00
ZSTD_CDict * ZSTD_createCDict_advanced ( const void * dictBuffer , size_t dictSize ,
ZSTD_dictLoadMethod_e dictLoadMethod ,
ZSTD_dictContentType_e dictContentType ,
ZSTD_compressionParameters cParams ,
ZSTD_customMem customMem )
{
ZSTD_CCtx_params cctxParams ;
ZSTD_memset ( & cctxParams , 0 , sizeof ( cctxParams ) ) ;
ZSTD_CCtxParams_init ( & cctxParams , 0 ) ;
cctxParams . cParams = cParams ;
cctxParams . customMem = customMem ;
return ZSTD_createCDict_advanced2 (
dictBuffer , dictSize ,
dictLoadMethod , dictContentType ,
& cctxParams , customMem ) ;
}
2021-12-21 11:21:29 +00:00
ZSTD_CDict * ZSTD_createCDict_advanced2 (
2020-12-18 14:46:27 +00:00
const void * dict , size_t dictSize ,
ZSTD_dictLoadMethod_e dictLoadMethod ,
ZSTD_dictContentType_e dictContentType ,
const ZSTD_CCtx_params * originalCctxParams ,
ZSTD_customMem customMem )
{
ZSTD_CCtx_params cctxParams = * originalCctxParams ;
ZSTD_compressionParameters cParams ;
ZSTD_CDict * cdict ;
DEBUGLOG ( 3 , " ZSTD_createCDict_advanced2, mode %u " , ( unsigned ) dictContentType ) ;
if ( ! customMem . customAlloc ^ ! customMem . customFree ) return NULL ;
if ( cctxParams . enableDedicatedDictSearch ) {
cParams = ZSTD_dedicatedDictSearch_getCParams (
cctxParams . compressionLevel , dictSize ) ;
ZSTD_overrideCParams ( & cParams , & cctxParams . cParams ) ;
} else {
cParams = ZSTD_getCParamsFromCCtxParams (
& cctxParams , ZSTD_CONTENTSIZE_UNKNOWN , dictSize , ZSTD_cpm_createCDict ) ;
}
if ( ! ZSTD_dedicatedDictSearch_isSupported ( & cParams ) ) {
/* Fall back to non-DDSS params */
cctxParams . enableDedicatedDictSearch = 0 ;
cParams = ZSTD_getCParamsFromCCtxParams (
& cctxParams , ZSTD_CONTENTSIZE_UNKNOWN , dictSize , ZSTD_cpm_createCDict ) ;
}
2021-05-14 22:48:15 +00:00
DEBUGLOG ( 3 , " ZSTD_createCDict_advanced2: DDS: %u " , cctxParams . enableDedicatedDictSearch ) ;
2020-12-18 14:46:27 +00:00
cctxParams . cParams = cParams ;
2021-05-14 22:48:15 +00:00
cctxParams . useRowMatchFinder = ZSTD_resolveRowMatchFinderMode ( cctxParams . useRowMatchFinder , & cParams ) ;
2020-12-18 14:46:27 +00:00
cdict = ZSTD_createCDict_advanced_internal ( dictSize ,
dictLoadMethod , cctxParams . cParams ,
2021-05-14 22:48:15 +00:00
cctxParams . useRowMatchFinder , cctxParams . enableDedicatedDictSearch ,
2020-12-18 14:46:27 +00:00
customMem ) ;
if ( ZSTD_isError ( ZSTD_initCDict_internal ( cdict ,
dict , dictSize ,
dictLoadMethod , dictContentType ,
cctxParams ) ) ) {
ZSTD_freeCDict ( cdict ) ;
return NULL ;
}
return cdict ;
}
2020-02-08 13:56:59 +00:00
ZSTD_CDict * ZSTD_createCDict ( const void * dict , size_t dictSize , int compressionLevel )
{
2020-12-18 14:46:27 +00:00
ZSTD_compressionParameters cParams = ZSTD_getCParams_internal ( compressionLevel , ZSTD_CONTENTSIZE_UNKNOWN , dictSize , ZSTD_cpm_createCDict ) ;
ZSTD_CDict * const cdict = ZSTD_createCDict_advanced ( dict , dictSize ,
2020-02-08 13:56:59 +00:00
ZSTD_dlm_byCopy , ZSTD_dct_auto ,
cParams , ZSTD_defaultCMem ) ;
if ( cdict )
2020-12-18 14:46:27 +00:00
cdict - > compressionLevel = ( compressionLevel = = 0 ) ? ZSTD_CLEVEL_DEFAULT : compressionLevel ;
2020-02-08 13:56:59 +00:00
return cdict ;
}
ZSTD_CDict * ZSTD_createCDict_byReference ( const void * dict , size_t dictSize , int compressionLevel )
{
2020-12-18 14:46:27 +00:00
ZSTD_compressionParameters cParams = ZSTD_getCParams_internal ( compressionLevel , ZSTD_CONTENTSIZE_UNKNOWN , dictSize , ZSTD_cpm_createCDict ) ;
ZSTD_CDict * const cdict = ZSTD_createCDict_advanced ( dict , dictSize ,
2020-02-08 13:56:59 +00:00
ZSTD_dlm_byRef , ZSTD_dct_auto ,
cParams , ZSTD_defaultCMem ) ;
2020-12-18 14:46:27 +00:00
if ( cdict )
cdict - > compressionLevel = ( compressionLevel = = 0 ) ? ZSTD_CLEVEL_DEFAULT : compressionLevel ;
return cdict ;
2020-02-08 13:56:59 +00:00
}
size_t ZSTD_freeCDict ( ZSTD_CDict * cdict )
{
if ( cdict = = NULL ) return 0 ; /* support free on NULL */
{ ZSTD_customMem const cMem = cdict - > customMem ;
int cdictInWorkspace = ZSTD_cwksp_owns_buffer ( & cdict - > workspace , cdict ) ;
ZSTD_cwksp_free ( & cdict - > workspace , cMem ) ;
if ( ! cdictInWorkspace ) {
2020-12-18 14:46:27 +00:00
ZSTD_customFree ( cdict , cMem ) ;
2020-02-08 13:56:59 +00:00
}
return 0 ;
}
}
/*! ZSTD_initStaticCDict_advanced() :
* Generate a digested dictionary in provided memory area .
* workspace : The memory area to emplace the dictionary into .
* Provided pointer must 8 - bytes aligned .
* It must outlive dictionary usage .
* workspaceSize : Use ZSTD_estimateCDictSize ( )
* to determine how large workspace must be .
* cParams : use ZSTD_getCParams ( ) to transform a compression level
* into its relevants cParams .
* @ return : pointer to ZSTD_CDict * , or NULL if error ( size too small )
* Note : there is no corresponding " free " function .
* Since workspace was allocated externally , it must be freed externally .
*/
const ZSTD_CDict * ZSTD_initStaticCDict (
void * workspace , size_t workspaceSize ,
const void * dict , size_t dictSize ,
ZSTD_dictLoadMethod_e dictLoadMethod ,
ZSTD_dictContentType_e dictContentType ,
ZSTD_compressionParameters cParams )
{
2021-12-21 11:21:29 +00:00
ZSTD_paramSwitch_e const useRowMatchFinder = ZSTD_resolveRowMatchFinderMode ( ZSTD_ps_auto , & cParams ) ;
2021-05-14 22:48:15 +00:00
/* enableDedicatedDictSearch == 1 ensures matchstate is not too small in case this CDict will be used for DDS + row hash */
size_t const matchStateSize = ZSTD_sizeof_matchState ( & cParams , useRowMatchFinder , /* enableDedicatedDictSearch */ 1 , /* forCCtx */ 0 ) ;
2020-02-08 13:56:59 +00:00
size_t const neededSize = ZSTD_cwksp_alloc_size ( sizeof ( ZSTD_CDict ) )
+ ( dictLoadMethod = = ZSTD_dlm_byRef ? 0
: ZSTD_cwksp_alloc_size ( ZSTD_cwksp_align ( dictSize , sizeof ( void * ) ) ) )
+ ZSTD_cwksp_alloc_size ( HUF_WORKSPACE_SIZE )
+ matchStateSize ;
ZSTD_CDict * cdict ;
2020-12-18 14:46:27 +00:00
ZSTD_CCtx_params params ;
2020-02-08 13:56:59 +00:00
if ( ( size_t ) workspace & 7 ) return NULL ; /* 8-aligned */
{
ZSTD_cwksp ws ;
2020-12-18 14:46:27 +00:00
ZSTD_cwksp_init ( & ws , workspace , workspaceSize , ZSTD_cwksp_static_alloc ) ;
2020-02-08 13:56:59 +00:00
cdict = ( ZSTD_CDict * ) ZSTD_cwksp_reserve_object ( & ws , sizeof ( ZSTD_CDict ) ) ;
if ( cdict = = NULL ) return NULL ;
ZSTD_cwksp_move ( & cdict - > workspace , & ws ) ;
}
DEBUGLOG ( 4 , " (workspaceSize < neededSize) : (%u < %u) => %u " ,
( unsigned ) workspaceSize , ( unsigned ) neededSize , ( unsigned ) ( workspaceSize < neededSize ) ) ;
if ( workspaceSize < neededSize ) return NULL ;
2020-12-18 14:46:27 +00:00
ZSTD_CCtxParams_init ( & params , 0 ) ;
params . cParams = cParams ;
2021-05-14 22:48:15 +00:00
params . useRowMatchFinder = useRowMatchFinder ;
cdict - > useRowMatchFinder = useRowMatchFinder ;
2020-12-18 14:46:27 +00:00
2020-02-08 13:56:59 +00:00
if ( ZSTD_isError ( ZSTD_initCDict_internal ( cdict ,
dict , dictSize ,
dictLoadMethod , dictContentType ,
2020-12-18 14:46:27 +00:00
params ) ) )
2020-02-08 13:56:59 +00:00
return NULL ;
return cdict ;
}
ZSTD_compressionParameters ZSTD_getCParamsFromCDict ( const ZSTD_CDict * cdict )
{
assert ( cdict ! = NULL ) ;
return cdict - > matchState . cParams ;
}
2020-12-18 14:46:27 +00:00
/*! ZSTD_getDictID_fromCDict() :
* Provides the dictID of the dictionary loaded into ` cdict ` .
* If @ return = = 0 , the dictionary is not conformant to Zstandard specification , or empty .
* Non - conformant dictionaries can still be loaded , but as content - only dictionaries . */
unsigned ZSTD_getDictID_fromCDict ( const ZSTD_CDict * cdict )
{
if ( cdict = = NULL ) return 0 ;
return cdict - > dictID ;
}
2021-05-14 22:48:15 +00:00
/* ZSTD_compressBegin_usingCDict_internal() :
* Implementation of various ZSTD_compressBegin_usingCDict * functions .
*/
static size_t ZSTD_compressBegin_usingCDict_internal (
2020-02-08 13:56:59 +00:00
ZSTD_CCtx * const cctx , const ZSTD_CDict * const cdict ,
ZSTD_frameParameters const fParams , unsigned long long const pledgedSrcSize )
{
2021-05-14 22:48:15 +00:00
ZSTD_CCtx_params cctxParams ;
DEBUGLOG ( 4 , " ZSTD_compressBegin_usingCDict_internal " ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR_IF ( cdict = = NULL , dictionary_wrong , " NULL pointer! " ) ;
2021-05-14 22:48:15 +00:00
/* Initialize the cctxParams from the cdict */
{
ZSTD_parameters params ;
params . fParams = fParams ;
2020-02-08 13:56:59 +00:00
params . cParams = ( pledgedSrcSize < ZSTD_USE_CDICT_PARAMS_SRCSIZE_CUTOFF
| | pledgedSrcSize < cdict - > dictContentSize * ZSTD_USE_CDICT_PARAMS_DICTSIZE_MULTIPLIER
| | pledgedSrcSize = = ZSTD_CONTENTSIZE_UNKNOWN
2021-05-14 22:48:15 +00:00
| | cdict - > compressionLevel = = 0 ) ?
2020-02-08 13:56:59 +00:00
ZSTD_getCParamsFromCDict ( cdict )
: ZSTD_getCParams ( cdict - > compressionLevel ,
pledgedSrcSize ,
cdict - > dictContentSize ) ;
2021-05-14 22:48:15 +00:00
ZSTD_CCtxParams_init_internal ( & cctxParams , & params , cdict - > compressionLevel ) ;
}
/* Increase window log to fit the entire dictionary and source if the
* source size is known . Limit the increase to 19 , which is the
* window log for compression level 1 with the largest source size .
*/
if ( pledgedSrcSize ! = ZSTD_CONTENTSIZE_UNKNOWN ) {
U32 const limitedSrcSize = ( U32 ) MIN ( pledgedSrcSize , 1U < < 19 ) ;
U32 const limitedSrcLog = limitedSrcSize > 1 ? ZSTD_highbit32 ( limitedSrcSize - 1 ) + 1 : 1 ;
cctxParams . cParams . windowLog = MAX ( cctxParams . cParams . windowLog , limitedSrcLog ) ;
2020-02-08 13:56:59 +00:00
}
2021-05-14 22:48:15 +00:00
return ZSTD_compressBegin_internal ( cctx ,
NULL , 0 , ZSTD_dct_auto , ZSTD_dtlm_fast ,
cdict ,
& cctxParams , pledgedSrcSize ,
ZSTDb_not_buffered ) ;
}
/* ZSTD_compressBegin_usingCDict_advanced() :
* This function is DEPRECATED .
* cdict must be ! = NULL */
size_t ZSTD_compressBegin_usingCDict_advanced (
ZSTD_CCtx * const cctx , const ZSTD_CDict * const cdict ,
ZSTD_frameParameters const fParams , unsigned long long const pledgedSrcSize )
{
return ZSTD_compressBegin_usingCDict_internal ( cctx , cdict , fParams , pledgedSrcSize ) ;
2020-02-08 13:56:59 +00:00
}
/* ZSTD_compressBegin_usingCDict() :
2021-05-14 22:48:15 +00:00
* cdict must be ! = NULL */
2020-02-08 13:56:59 +00:00
size_t ZSTD_compressBegin_usingCDict ( ZSTD_CCtx * cctx , const ZSTD_CDict * cdict )
{
ZSTD_frameParameters const fParams = { 0 /*content*/ , 0 /*checksum*/ , 0 /*noDictID*/ } ;
2021-05-14 22:48:15 +00:00
return ZSTD_compressBegin_usingCDict_internal ( cctx , cdict , fParams , ZSTD_CONTENTSIZE_UNKNOWN ) ;
2020-02-08 13:56:59 +00:00
}
2021-05-14 22:48:15 +00:00
/*! ZSTD_compress_usingCDict_internal():
* Implementation of various ZSTD_compress_usingCDict * functions .
*/
static size_t ZSTD_compress_usingCDict_internal ( ZSTD_CCtx * cctx ,
2020-02-08 13:56:59 +00:00
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
const ZSTD_CDict * cdict , ZSTD_frameParameters fParams )
{
2021-05-14 22:48:15 +00:00
FORWARD_IF_ERROR ( ZSTD_compressBegin_usingCDict_internal ( cctx , cdict , fParams , srcSize ) , " " ) ; /* will check if cdict != NULL */
2020-02-08 13:56:59 +00:00
return ZSTD_compressEnd ( cctx , dst , dstCapacity , src , srcSize ) ;
}
2021-05-14 22:48:15 +00:00
/*! ZSTD_compress_usingCDict_advanced():
* This function is DEPRECATED .
*/
size_t ZSTD_compress_usingCDict_advanced ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
const ZSTD_CDict * cdict , ZSTD_frameParameters fParams )
{
return ZSTD_compress_usingCDict_internal ( cctx , dst , dstCapacity , src , srcSize , cdict , fParams ) ;
}
2020-02-08 13:56:59 +00:00
/*! ZSTD_compress_usingCDict() :
* Compression using a digested Dictionary .
* Faster startup than ZSTD_compress_usingDict ( ) , recommended when same dictionary is used multiple times .
* Note that compression parameters are decided at CDict creation time
* while frame parameters are hardcoded */
size_t ZSTD_compress_usingCDict ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize ,
const ZSTD_CDict * cdict )
{
ZSTD_frameParameters const fParams = { 1 /*content*/ , 0 /*checksum*/ , 0 /*noDictID*/ } ;
2021-05-14 22:48:15 +00:00
return ZSTD_compress_usingCDict_internal ( cctx , dst , dstCapacity , src , srcSize , cdict , fParams ) ;
2020-02-08 13:56:59 +00:00
}
/* ******************************************************************
* Streaming
* * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * */
ZSTD_CStream * ZSTD_createCStream ( void )
{
DEBUGLOG ( 3 , " ZSTD_createCStream " ) ;
return ZSTD_createCStream_advanced ( ZSTD_defaultCMem ) ;
}
ZSTD_CStream * ZSTD_initStaticCStream ( void * workspace , size_t workspaceSize )
{
return ZSTD_initStaticCCtx ( workspace , workspaceSize ) ;
}
ZSTD_CStream * ZSTD_createCStream_advanced ( ZSTD_customMem customMem )
{ /* CStream and CCtx are now same object */
return ZSTD_createCCtx_advanced ( customMem ) ;
}
size_t ZSTD_freeCStream ( ZSTD_CStream * zcs )
{
return ZSTD_freeCCtx ( zcs ) ; /* same object */
}
/*====== Initialization ======*/
size_t ZSTD_CStreamInSize ( void ) { return ZSTD_BLOCKSIZE_MAX ; }
size_t ZSTD_CStreamOutSize ( void )
{
return ZSTD_compressBound ( ZSTD_BLOCKSIZE_MAX ) + ZSTD_blockHeaderSize + 4 /* 32-bits hash */ ;
}
2020-12-18 14:46:27 +00:00
static ZSTD_cParamMode_e ZSTD_getCParamMode ( ZSTD_CDict const * cdict , ZSTD_CCtx_params const * params , U64 pledgedSrcSize )
2020-02-08 13:56:59 +00:00
{
2020-12-18 14:46:27 +00:00
if ( cdict ! = NULL & & ZSTD_shouldAttachDict ( cdict , params , pledgedSrcSize ) )
return ZSTD_cpm_attachDict ;
else
return ZSTD_cpm_noAttachDict ;
2020-02-08 13:56:59 +00:00
}
/* ZSTD_resetCStream():
* pledgedSrcSize = = 0 means " unknown " */
size_t ZSTD_resetCStream ( ZSTD_CStream * zcs , unsigned long long pss )
{
/* temporary : 0 interpreted as "unknown" during transition period.
* Users willing to specify " unknown " * * must * * use ZSTD_CONTENTSIZE_UNKNOWN .
* 0 will be interpreted as " empty " in the future .
*/
U64 const pledgedSrcSize = ( pss = = 0 ) ? ZSTD_CONTENTSIZE_UNKNOWN : pss ;
DEBUGLOG ( 4 , " ZSTD_resetCStream: pledgedSrcSize = %u " , ( unsigned ) pledgedSrcSize ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_setPledgedSrcSize ( zcs , pledgedSrcSize ) , " " ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
/*! ZSTD_initCStream_internal() :
* Note : for lib / compress only . Used by zstdmt_compress . c .
* Assumption 1 : params are valid
* Assumption 2 : either dict , or cdict , is defined , not both */
size_t ZSTD_initCStream_internal ( ZSTD_CStream * zcs ,
const void * dict , size_t dictSize , const ZSTD_CDict * cdict ,
const ZSTD_CCtx_params * params ,
unsigned long long pledgedSrcSize )
{
DEBUGLOG ( 4 , " ZSTD_initCStream_internal " ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_setPledgedSrcSize ( zcs , pledgedSrcSize ) , " " ) ;
2020-02-08 13:56:59 +00:00
assert ( ! ZSTD_isError ( ZSTD_checkCParams ( params - > cParams ) ) ) ;
zcs - > requestedParams = * params ;
assert ( ! ( ( dict ) & & ( cdict ) ) ) ; /* either dict or cdict, not both */
if ( dict ) {
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_loadDictionary ( zcs , dict , dictSize ) , " " ) ;
2020-02-08 13:56:59 +00:00
} else {
/* Dictionary is cleared if !cdict */
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_refCDict ( zcs , cdict ) , " " ) ;
2020-02-08 13:56:59 +00:00
}
return 0 ;
}
/* ZSTD_initCStream_usingCDict_advanced() :
* same as ZSTD_initCStream_usingCDict ( ) , with control over frame parameters */
size_t ZSTD_initCStream_usingCDict_advanced ( ZSTD_CStream * zcs ,
const ZSTD_CDict * cdict ,
ZSTD_frameParameters fParams ,
unsigned long long pledgedSrcSize )
{
DEBUGLOG ( 4 , " ZSTD_initCStream_usingCDict_advanced " ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_setPledgedSrcSize ( zcs , pledgedSrcSize ) , " " ) ;
2020-02-08 13:56:59 +00:00
zcs - > requestedParams . fParams = fParams ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_refCDict ( zcs , cdict ) , " " ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
/* note : cdict must outlive compression session */
size_t ZSTD_initCStream_usingCDict ( ZSTD_CStream * zcs , const ZSTD_CDict * cdict )
{
DEBUGLOG ( 4 , " ZSTD_initCStream_usingCDict " ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_refCDict ( zcs , cdict ) , " " ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
/* ZSTD_initCStream_advanced() :
* pledgedSrcSize must be exact .
* if srcSize is not known at init time , use value ZSTD_CONTENTSIZE_UNKNOWN .
* dict is loaded with default parameters ZSTD_dct_auto and ZSTD_dlm_byCopy . */
size_t ZSTD_initCStream_advanced ( ZSTD_CStream * zcs ,
const void * dict , size_t dictSize ,
ZSTD_parameters params , unsigned long long pss )
{
/* for compatibility with older programs relying on this behavior.
* Users should now specify ZSTD_CONTENTSIZE_UNKNOWN .
* This line will be removed in the future .
*/
U64 const pledgedSrcSize = ( pss = = 0 & & params . fParams . contentSizeFlag = = 0 ) ? ZSTD_CONTENTSIZE_UNKNOWN : pss ;
DEBUGLOG ( 4 , " ZSTD_initCStream_advanced " ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_setPledgedSrcSize ( zcs , pledgedSrcSize ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_checkCParams ( params . cParams ) , " " ) ;
2021-05-14 22:48:15 +00:00
ZSTD_CCtxParams_setZstdParams ( & zcs - > requestedParams , & params ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_loadDictionary ( zcs , dict , dictSize ) , " " ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
size_t ZSTD_initCStream_usingDict ( ZSTD_CStream * zcs , const void * dict , size_t dictSize , int compressionLevel )
{
DEBUGLOG ( 4 , " ZSTD_initCStream_usingDict " ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_setParameter ( zcs , ZSTD_c_compressionLevel , compressionLevel ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_loadDictionary ( zcs , dict , dictSize ) , " " ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
size_t ZSTD_initCStream_srcSize ( ZSTD_CStream * zcs , int compressionLevel , unsigned long long pss )
{
/* temporary : 0 interpreted as "unknown" during transition period.
* Users willing to specify " unknown " * * must * * use ZSTD_CONTENTSIZE_UNKNOWN .
* 0 will be interpreted as " empty " in the future .
*/
U64 const pledgedSrcSize = ( pss = = 0 ) ? ZSTD_CONTENTSIZE_UNKNOWN : pss ;
DEBUGLOG ( 4 , " ZSTD_initCStream_srcSize " ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_refCDict ( zcs , NULL ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_setParameter ( zcs , ZSTD_c_compressionLevel , compressionLevel ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_setPledgedSrcSize ( zcs , pledgedSrcSize ) , " " ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
size_t ZSTD_initCStream ( ZSTD_CStream * zcs , int compressionLevel )
{
DEBUGLOG ( 4 , " ZSTD_initCStream " ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_refCDict ( zcs , NULL ) , " " ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_setParameter ( zcs , ZSTD_c_compressionLevel , compressionLevel ) , " " ) ;
2020-02-08 13:56:59 +00:00
return 0 ;
}
/*====== Compression ======*/
static size_t ZSTD_nextInputSizeHint ( const ZSTD_CCtx * cctx )
{
2023-02-10 19:25:40 +00:00
if ( cctx - > appliedParams . inBufferMode = = ZSTD_bm_stable ) {
return cctx - > blockSize - cctx - > stableIn_notConsumed ;
}
assert ( cctx - > appliedParams . inBufferMode = = ZSTD_bm_buffered ) ;
{ size_t hintInSize = cctx - > inBuffTarget - cctx - > inBuffPos ;
if ( hintInSize = = 0 ) hintInSize = cctx - > blockSize ;
return hintInSize ;
}
2020-02-08 13:56:59 +00:00
}
/** ZSTD_compressStream_generic():
* internal function for all * compressStream * ( ) variants
2023-02-10 19:25:40 +00:00
* @ return : hint size for next input to complete ongoing block */
2020-02-08 13:56:59 +00:00
static size_t ZSTD_compressStream_generic ( ZSTD_CStream * zcs ,
ZSTD_outBuffer * output ,
ZSTD_inBuffer * input ,
ZSTD_EndDirective const flushMode )
{
2023-02-10 19:25:40 +00:00
const char * const istart = ( assert ( input ! = NULL ) , ( const char * ) input - > src ) ;
const char * const iend = ( istart ! = NULL ) ? istart + input - > size : istart ;
const char * ip = ( istart ! = NULL ) ? istart + input - > pos : istart ;
char * const ostart = ( assert ( output ! = NULL ) , ( char * ) output - > dst ) ;
char * const oend = ( ostart ! = NULL ) ? ostart + output - > size : ostart ;
char * op = ( ostart ! = NULL ) ? ostart + output - > pos : ostart ;
2020-02-08 13:56:59 +00:00
U32 someMoreWork = 1 ;
/* check expectations */
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " ZSTD_compressStream_generic, flush=%i, srcSize = %zu " , ( int ) flushMode , input - > size - input - > pos ) ;
assert ( zcs ! = NULL ) ;
if ( zcs - > appliedParams . inBufferMode = = ZSTD_bm_stable ) {
assert ( input - > pos > = zcs - > stableIn_notConsumed ) ;
input - > pos - = zcs - > stableIn_notConsumed ;
ip - = zcs - > stableIn_notConsumed ;
zcs - > stableIn_notConsumed = 0 ;
}
2020-12-18 14:46:27 +00:00
if ( zcs - > appliedParams . inBufferMode = = ZSTD_bm_buffered ) {
assert ( zcs - > inBuff ! = NULL ) ;
assert ( zcs - > inBuffSize > 0 ) ;
}
if ( zcs - > appliedParams . outBufferMode = = ZSTD_bm_buffered ) {
assert ( zcs - > outBuff ! = NULL ) ;
assert ( zcs - > outBuffSize > 0 ) ;
}
2023-02-10 19:25:40 +00:00
if ( input - > src = = NULL ) assert ( input - > size = = 0 ) ;
2020-02-08 13:56:59 +00:00
assert ( input - > pos < = input - > size ) ;
2023-02-10 19:25:40 +00:00
if ( output - > dst = = NULL ) assert ( output - > size = = 0 ) ;
assert ( output - > pos < = output - > size ) ;
2020-12-18 14:46:27 +00:00
assert ( ( U32 ) flushMode < = ( U32 ) ZSTD_e_end ) ;
2020-02-08 13:56:59 +00:00
while ( someMoreWork ) {
switch ( zcs - > streamStage )
{
case zcss_init :
RETURN_ERROR ( init_missing , " call ZSTD_initCStream() first! " ) ;
case zcss_load :
if ( ( flushMode = = ZSTD_e_end )
2020-12-18 14:46:27 +00:00
& & ( ( size_t ) ( oend - op ) > = ZSTD_compressBound ( iend - ip ) /* Enough output space */
| | zcs - > appliedParams . outBufferMode = = ZSTD_bm_stable ) /* OR we are allowed to return dstSizeTooSmall */
2020-02-08 13:56:59 +00:00
& & ( zcs - > inBuffPos = = 0 ) ) {
/* shortcut to compression pass directly into output buffer */
size_t const cSize = ZSTD_compressEnd ( zcs ,
op , oend - op , ip , iend - ip ) ;
DEBUGLOG ( 4 , " ZSTD_compressEnd : cSize=%u " , ( unsigned ) cSize ) ;
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( cSize , " ZSTD_compressEnd failed " ) ;
2020-02-08 13:56:59 +00:00
ip = iend ;
op + = cSize ;
zcs - > frameEnded = 1 ;
ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) ;
someMoreWork = 0 ; break ;
}
2020-12-18 14:46:27 +00:00
/* complete loading into inBuffer in buffered mode */
if ( zcs - > appliedParams . inBufferMode = = ZSTD_bm_buffered ) {
size_t const toLoad = zcs - > inBuffTarget - zcs - > inBuffPos ;
2020-02-08 13:56:59 +00:00
size_t const loaded = ZSTD_limitCopy (
zcs - > inBuff + zcs - > inBuffPos , toLoad ,
ip , iend - ip ) ;
zcs - > inBuffPos + = loaded ;
2023-02-10 19:25:40 +00:00
if ( ip ) ip + = loaded ;
2020-02-08 13:56:59 +00:00
if ( ( flushMode = = ZSTD_e_continue )
& & ( zcs - > inBuffPos < zcs - > inBuffTarget ) ) {
/* not enough input to fill full block : stop here */
someMoreWork = 0 ; break ;
}
if ( ( flushMode = = ZSTD_e_flush )
& & ( zcs - > inBuffPos = = zcs - > inToCompress ) ) {
/* empty */
someMoreWork = 0 ; break ;
}
2023-02-10 19:25:40 +00:00
} else {
assert ( zcs - > appliedParams . inBufferMode = = ZSTD_bm_stable ) ;
if ( ( flushMode = = ZSTD_e_continue )
& & ( ( size_t ) ( iend - ip ) < zcs - > blockSize ) ) {
/* can't compress a full block : stop here */
zcs - > stableIn_notConsumed = ( size_t ) ( iend - ip ) ;
ip = iend ; /* pretend to have consumed input */
someMoreWork = 0 ; break ;
}
if ( ( flushMode = = ZSTD_e_flush )
& & ( ip = = iend ) ) {
/* empty */
someMoreWork = 0 ; break ;
}
2020-02-08 13:56:59 +00:00
}
/* compress current block (note : this stage cannot be stopped in the middle) */
DEBUGLOG ( 5 , " stream compression stage (flushMode==%u) " , flushMode ) ;
2020-12-18 14:46:27 +00:00
{ int const inputBuffered = ( zcs - > appliedParams . inBufferMode = = ZSTD_bm_buffered ) ;
void * cDst ;
2020-02-08 13:56:59 +00:00
size_t cSize ;
size_t oSize = oend - op ;
2023-02-10 19:25:40 +00:00
size_t const iSize = inputBuffered ? zcs - > inBuffPos - zcs - > inToCompress
: MIN ( ( size_t ) ( iend - ip ) , zcs - > blockSize ) ;
2020-12-18 14:46:27 +00:00
if ( oSize > = ZSTD_compressBound ( iSize ) | | zcs - > appliedParams . outBufferMode = = ZSTD_bm_stable )
2020-02-08 13:56:59 +00:00
cDst = op ; /* compress into output buffer, to skip flush stage */
else
cDst = zcs - > outBuff , oSize = zcs - > outBuffSize ;
2020-12-18 14:46:27 +00:00
if ( inputBuffered ) {
unsigned const lastBlock = ( flushMode = = ZSTD_e_end ) & & ( ip = = iend ) ;
cSize = lastBlock ?
ZSTD_compressEnd ( zcs , cDst , oSize ,
zcs - > inBuff + zcs - > inToCompress , iSize ) :
ZSTD_compressContinue ( zcs , cDst , oSize ,
zcs - > inBuff + zcs - > inToCompress , iSize ) ;
FORWARD_IF_ERROR ( cSize , " %s " , lastBlock ? " ZSTD_compressEnd failed " : " ZSTD_compressContinue failed " ) ;
zcs - > frameEnded = lastBlock ;
/* prepare next block */
zcs - > inBuffTarget = zcs - > inBuffPos + zcs - > blockSize ;
if ( zcs - > inBuffTarget > zcs - > inBuffSize )
zcs - > inBuffPos = 0 , zcs - > inBuffTarget = zcs - > blockSize ;
DEBUGLOG ( 5 , " inBuffTarget:%u / inBuffSize:%u " ,
( unsigned ) zcs - > inBuffTarget , ( unsigned ) zcs - > inBuffSize ) ;
if ( ! lastBlock )
assert ( zcs - > inBuffTarget < = zcs - > inBuffSize ) ;
zcs - > inToCompress = zcs - > inBuffPos ;
2023-02-10 19:25:40 +00:00
} else { /* !inputBuffered, hence ZSTD_bm_stable */
unsigned const lastBlock = ( flushMode = = ZSTD_e_end ) & & ( ip + iSize = = iend ) ;
2020-12-18 14:46:27 +00:00
cSize = lastBlock ?
ZSTD_compressEnd ( zcs , cDst , oSize , ip , iSize ) :
ZSTD_compressContinue ( zcs , cDst , oSize , ip , iSize ) ;
/* Consume the input prior to error checking to mirror buffered mode. */
2023-02-10 19:25:40 +00:00
if ( ip ) ip + = iSize ;
2020-12-18 14:46:27 +00:00
FORWARD_IF_ERROR ( cSize , " %s " , lastBlock ? " ZSTD_compressEnd failed " : " ZSTD_compressContinue failed " ) ;
zcs - > frameEnded = lastBlock ;
2023-02-10 19:25:40 +00:00
if ( lastBlock ) assert ( ip = = iend ) ;
2020-12-18 14:46:27 +00:00
}
2020-02-08 13:56:59 +00:00
if ( cDst = = op ) { /* no need to flush */
op + = cSize ;
if ( zcs - > frameEnded ) {
DEBUGLOG ( 5 , " Frame completed directly in outBuffer " ) ;
someMoreWork = 0 ;
ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) ;
}
break ;
}
zcs - > outBuffContentSize = cSize ;
zcs - > outBuffFlushedSize = 0 ;
zcs - > streamStage = zcss_flush ; /* pass-through to flush stage */
}
2021-12-21 11:21:29 +00:00
ZSTD_FALLTHROUGH ;
2020-02-08 13:56:59 +00:00
case zcss_flush :
DEBUGLOG ( 5 , " flush stage " ) ;
2020-12-18 14:46:27 +00:00
assert ( zcs - > appliedParams . outBufferMode = = ZSTD_bm_buffered ) ;
2020-02-08 13:56:59 +00:00
{ size_t const toFlush = zcs - > outBuffContentSize - zcs - > outBuffFlushedSize ;
size_t const flushed = ZSTD_limitCopy ( op , ( size_t ) ( oend - op ) ,
zcs - > outBuff + zcs - > outBuffFlushedSize , toFlush ) ;
DEBUGLOG ( 5 , " toFlush: %u into %u ==> flushed: %u " ,
( unsigned ) toFlush , ( unsigned ) ( oend - op ) , ( unsigned ) flushed ) ;
if ( flushed )
op + = flushed ;
zcs - > outBuffFlushedSize + = flushed ;
if ( toFlush ! = flushed ) {
/* flush not fully completed, presumably because dst is too small */
assert ( op = = oend ) ;
someMoreWork = 0 ;
break ;
}
zcs - > outBuffContentSize = zcs - > outBuffFlushedSize = 0 ;
if ( zcs - > frameEnded ) {
DEBUGLOG ( 5 , " Frame completed on flush " ) ;
someMoreWork = 0 ;
ZSTD_CCtx_reset ( zcs , ZSTD_reset_session_only ) ;
break ;
}
zcs - > streamStage = zcss_load ;
break ;
}
default : /* impossible */
assert ( 0 ) ;
}
}
input - > pos = ip - istart ;
output - > pos = op - ostart ;
if ( zcs - > frameEnded ) return 0 ;
return ZSTD_nextInputSizeHint ( zcs ) ;
}
static size_t ZSTD_nextInputSizeHint_MTorST ( const ZSTD_CCtx * cctx )
{
# ifdef ZSTD_MULTITHREAD
if ( cctx - > appliedParams . nbWorkers > = 1 ) {
assert ( cctx - > mtctx ! = NULL ) ;
return ZSTDMT_nextInputSizeHint ( cctx - > mtctx ) ;
}
# endif
return ZSTD_nextInputSizeHint ( cctx ) ;
}
size_t ZSTD_compressStream ( ZSTD_CStream * zcs , ZSTD_outBuffer * output , ZSTD_inBuffer * input )
{
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_compressStream2 ( zcs , output , input , ZSTD_e_continue ) , " " ) ;
2020-02-08 13:56:59 +00:00
return ZSTD_nextInputSizeHint_MTorST ( zcs ) ;
}
2020-12-18 14:46:27 +00:00
/* After a compression call set the expected input/output buffer.
* This is validated at the start of the next compression call .
*/
2023-02-10 19:25:40 +00:00
static void
ZSTD_setBufferExpectations ( ZSTD_CCtx * cctx , const ZSTD_outBuffer * output , const ZSTD_inBuffer * input )
2020-12-18 14:46:27 +00:00
{
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " ZSTD_setBufferExpectations (for advanced stable in/out modes) " ) ;
2020-12-18 14:46:27 +00:00
if ( cctx - > appliedParams . inBufferMode = = ZSTD_bm_stable ) {
cctx - > expectedInBuffer = * input ;
}
if ( cctx - > appliedParams . outBufferMode = = ZSTD_bm_stable ) {
cctx - > expectedOutBufferSize = output - > size - output - > pos ;
}
}
/* Validate that the input/output buffers match the expectations set by
* ZSTD_setBufferExpectations .
*/
static size_t ZSTD_checkBufferStability ( ZSTD_CCtx const * cctx ,
ZSTD_outBuffer const * output ,
ZSTD_inBuffer const * input ,
ZSTD_EndDirective endOp )
{
if ( cctx - > appliedParams . inBufferMode = = ZSTD_bm_stable ) {
ZSTD_inBuffer const expect = cctx - > expectedInBuffer ;
2023-02-10 19:25:40 +00:00
if ( expect . src ! = input - > src | | expect . pos ! = input - > pos )
RETURN_ERROR ( stabilityCondition_notRespected , " ZSTD_c_stableInBuffer enabled but input differs! " ) ;
2020-12-18 14:46:27 +00:00
}
2023-02-10 19:25:40 +00:00
( void ) endOp ;
2020-12-18 14:46:27 +00:00
if ( cctx - > appliedParams . outBufferMode = = ZSTD_bm_stable ) {
size_t const outBufferSize = output - > size - output - > pos ;
if ( cctx - > expectedOutBufferSize ! = outBufferSize )
2023-02-10 19:25:40 +00:00
RETURN_ERROR ( stabilityCondition_notRespected , " ZSTD_c_stableOutBuffer enabled but output size differs! " ) ;
2020-12-18 14:46:27 +00:00
}
return 0 ;
}
static size_t ZSTD_CCtx_init_compressStream2 ( ZSTD_CCtx * cctx ,
ZSTD_EndDirective endOp ,
2023-02-10 19:25:40 +00:00
size_t inSize )
{
2020-12-18 14:46:27 +00:00
ZSTD_CCtx_params params = cctx - > requestedParams ;
ZSTD_prefixDict const prefixDict = cctx - > prefixDict ;
FORWARD_IF_ERROR ( ZSTD_initLocalDict ( cctx ) , " " ) ; /* Init the local dict if present. */
ZSTD_memset ( & cctx - > prefixDict , 0 , sizeof ( cctx - > prefixDict ) ) ; /* single usage */
assert ( prefixDict . dict = = NULL | | cctx - > cdict = = NULL ) ; /* only one can be set */
2021-05-14 22:48:15 +00:00
if ( cctx - > cdict & & ! cctx - > localDict . cdict ) {
/* Let the cdict's compression level take priority over the requested params.
* But do not take the cdict ' s compression level if the " cdict " is actually a localDict
* generated from ZSTD_initLocalDict ( ) .
*/
params . compressionLevel = cctx - > cdict - > compressionLevel ;
}
2020-12-18 14:46:27 +00:00
DEBUGLOG ( 4 , " ZSTD_compressStream2 : transparent init stage " ) ;
2023-02-10 19:25:40 +00:00
if ( endOp = = ZSTD_e_end ) cctx - > pledgedSrcSizePlusOne = inSize + 1 ; /* auto-determine pledgedSrcSize */
{ size_t const dictSize = prefixDict . dict
2020-12-18 14:46:27 +00:00
? prefixDict . dictSize
: ( cctx - > cdict ? cctx - > cdict - > dictContentSize : 0 ) ;
ZSTD_cParamMode_e const mode = ZSTD_getCParamMode ( cctx - > cdict , & params , cctx - > pledgedSrcSizePlusOne - 1 ) ;
params . cParams = ZSTD_getCParamsFromCCtxParams (
& params , cctx - > pledgedSrcSizePlusOne - 1 ,
dictSize , mode ) ;
}
2021-12-21 11:21:29 +00:00
params . useBlockSplitter = ZSTD_resolveBlockSplitterMode ( params . useBlockSplitter , & params . cParams ) ;
params . ldmParams . enableLdm = ZSTD_resolveEnableLdm ( params . ldmParams . enableLdm , & params . cParams ) ;
2021-05-14 22:48:15 +00:00
params . useRowMatchFinder = ZSTD_resolveRowMatchFinderMode ( params . useRowMatchFinder , & params . cParams ) ;
2023-02-10 19:25:40 +00:00
params . validateSequences = ZSTD_resolveExternalSequenceValidation ( params . validateSequences ) ;
params . maxBlockSize = ZSTD_resolveMaxBlockSize ( params . maxBlockSize ) ;
params . searchForExternalRepcodes = ZSTD_resolveExternalRepcodeSearch ( params . searchForExternalRepcodes , params . compressionLevel ) ;
2021-05-14 22:48:15 +00:00
2020-12-18 14:46:27 +00:00
# ifdef ZSTD_MULTITHREAD
2023-02-10 19:25:40 +00:00
/* If external matchfinder is enabled, make sure to fail before checking job size (for consistency) */
RETURN_ERROR_IF (
params . useSequenceProducer = = 1 & & params . nbWorkers > = 1 ,
parameter_combination_unsupported ,
" External sequence producer isn't supported with nbWorkers >= 1 "
) ;
2020-12-18 14:46:27 +00:00
if ( ( cctx - > pledgedSrcSizePlusOne - 1 ) < = ZSTDMT_JOBSIZE_MIN ) {
params . nbWorkers = 0 ; /* do not invoke multi-threading when src size is too small */
}
if ( params . nbWorkers > 0 ) {
2021-05-14 22:48:15 +00:00
# if ZSTD_TRACE
cctx - > traceCtx = ( ZSTD_trace_compress_begin ! = NULL ) ? ZSTD_trace_compress_begin ( cctx ) : 0 ;
# endif
2020-12-18 14:46:27 +00:00
/* mt context creation */
if ( cctx - > mtctx = = NULL ) {
DEBUGLOG ( 4 , " ZSTD_compressStream2: creating new mtctx for nbWorkers=%u " ,
params . nbWorkers ) ;
cctx - > mtctx = ZSTDMT_createCCtx_advanced ( ( U32 ) params . nbWorkers , cctx - > customMem , cctx - > pool ) ;
RETURN_ERROR_IF ( cctx - > mtctx = = NULL , memory_allocation , " NULL pointer! " ) ;
}
/* mt compression */
DEBUGLOG ( 4 , " call ZSTDMT_initCStream_internal as nbWorkers=%u " , params . nbWorkers ) ;
FORWARD_IF_ERROR ( ZSTDMT_initCStream_internal (
cctx - > mtctx ,
prefixDict . dict , prefixDict . dictSize , prefixDict . dictContentType ,
cctx - > cdict , params , cctx - > pledgedSrcSizePlusOne - 1 ) , " " ) ;
2021-05-14 22:48:15 +00:00
cctx - > dictID = cctx - > cdict ? cctx - > cdict - > dictID : 0 ;
cctx - > dictContentSize = cctx - > cdict ? cctx - > cdict - > dictContentSize : prefixDict . dictSize ;
cctx - > consumedSrcSize = 0 ;
cctx - > producedCSize = 0 ;
2020-12-18 14:46:27 +00:00
cctx - > streamStage = zcss_load ;
cctx - > appliedParams = params ;
} else
2023-02-10 19:25:40 +00:00
# endif /* ZSTD_MULTITHREAD */
2020-12-18 14:46:27 +00:00
{ U64 const pledgedSrcSize = cctx - > pledgedSrcSizePlusOne - 1 ;
assert ( ! ZSTD_isError ( ZSTD_checkCParams ( params . cParams ) ) ) ;
FORWARD_IF_ERROR ( ZSTD_compressBegin_internal ( cctx ,
prefixDict . dict , prefixDict . dictSize , prefixDict . dictContentType , ZSTD_dtlm_fast ,
cctx - > cdict ,
& params , pledgedSrcSize ,
ZSTDb_buffered ) , " " ) ;
assert ( cctx - > appliedParams . nbWorkers = = 0 ) ;
cctx - > inToCompress = 0 ;
cctx - > inBuffPos = 0 ;
if ( cctx - > appliedParams . inBufferMode = = ZSTD_bm_buffered ) {
/* for small input: avoid automatic flush on reaching end of block, since
* it would require to add a 3 - bytes null block to end frame
*/
cctx - > inBuffTarget = cctx - > blockSize + ( cctx - > blockSize = = pledgedSrcSize ) ;
} else {
cctx - > inBuffTarget = 0 ;
}
cctx - > outBuffContentSize = cctx - > outBuffFlushedSize = 0 ;
cctx - > streamStage = zcss_load ;
cctx - > frameEnded = 0 ;
}
return 0 ;
}
2020-02-08 13:56:59 +00:00
2023-02-10 19:25:40 +00:00
/* @return provides a minimum amount of data remaining to be flushed from internal buffers
*/
2020-02-08 13:56:59 +00:00
size_t ZSTD_compressStream2 ( ZSTD_CCtx * cctx ,
ZSTD_outBuffer * output ,
ZSTD_inBuffer * input ,
ZSTD_EndDirective endOp )
{
DEBUGLOG ( 5 , " ZSTD_compressStream2, endOp=%u " , ( unsigned ) endOp ) ;
/* check conditions */
2020-12-18 14:46:27 +00:00
RETURN_ERROR_IF ( output - > pos > output - > size , dstSize_tooSmall , " invalid output buffer " ) ;
RETURN_ERROR_IF ( input - > pos > input - > size , srcSize_wrong , " invalid input buffer " ) ;
RETURN_ERROR_IF ( ( U32 ) endOp > ( U32 ) ZSTD_e_end , parameter_outOfBound , " invalid endDirective " ) ;
assert ( cctx ! = NULL ) ;
2020-02-08 13:56:59 +00:00
/* transparent initialization stage */
if ( cctx - > streamStage = = zcss_init ) {
2023-02-10 19:25:40 +00:00
size_t const inputSize = input - > size - input - > pos ; /* no obligation to start from pos==0 */
size_t const totalInputSize = inputSize + cctx - > stableIn_notConsumed ;
if ( ( cctx - > requestedParams . inBufferMode = = ZSTD_bm_stable ) /* input is presumed stable, across invocations */
& & ( endOp = = ZSTD_e_continue ) /* no flush requested, more input to come */
& & ( totalInputSize < ZSTD_BLOCKSIZE_MAX ) ) { /* not even reached one block yet */
if ( cctx - > stableIn_notConsumed ) { /* not the first time */
/* check stable source guarantees */
RETURN_ERROR_IF ( input - > src ! = cctx - > expectedInBuffer . src , stabilityCondition_notRespected , " stableInBuffer condition not respected: wrong src pointer " ) ;
RETURN_ERROR_IF ( input - > pos ! = cctx - > expectedInBuffer . size , stabilityCondition_notRespected , " stableInBuffer condition not respected: externally modified pos " ) ;
}
/* pretend input was consumed, to give a sense forward progress */
input - > pos = input - > size ;
/* save stable inBuffer, for later control, and flush/end */
cctx - > expectedInBuffer = * input ;
/* but actually input wasn't consumed, so keep track of position from where compression shall resume */
cctx - > stableIn_notConsumed + = inputSize ;
/* don't initialize yet, wait for the first block of flush() order, for better parameters adaptation */
return ZSTD_FRAMEHEADERSIZE_MIN ( cctx - > requestedParams . format ) ; /* at least some header to produce */
}
FORWARD_IF_ERROR ( ZSTD_CCtx_init_compressStream2 ( cctx , endOp , totalInputSize ) , " compressStream2 initialization failed " ) ;
ZSTD_setBufferExpectations ( cctx , output , input ) ; /* Set initial buffer expectations now that we've initialized */
2020-12-18 14:46:27 +00:00
}
2020-02-08 13:56:59 +00:00
/* end of transparent initialization stage */
2020-12-18 14:46:27 +00:00
FORWARD_IF_ERROR ( ZSTD_checkBufferStability ( cctx , output , input , endOp ) , " invalid buffers " ) ;
2020-02-08 13:56:59 +00:00
/* compression stage */
# ifdef ZSTD_MULTITHREAD
if ( cctx - > appliedParams . nbWorkers > 0 ) {
size_t flushMin ;
if ( cctx - > cParamsChanged ) {
ZSTDMT_updateCParams_whileCompressing ( cctx - > mtctx , & cctx - > requestedParams ) ;
cctx - > cParamsChanged = 0 ;
}
2023-02-10 19:25:40 +00:00
if ( cctx - > stableIn_notConsumed ) {
assert ( cctx - > appliedParams . inBufferMode = = ZSTD_bm_stable ) ;
/* some early data was skipped - make it available for consumption */
assert ( input - > pos > = cctx - > stableIn_notConsumed ) ;
input - > pos - = cctx - > stableIn_notConsumed ;
cctx - > stableIn_notConsumed = 0 ;
}
2020-12-18 14:46:27 +00:00
for ( ; ; ) {
size_t const ipos = input - > pos ;
size_t const opos = output - > pos ;
2020-02-08 13:56:59 +00:00
flushMin = ZSTDMT_compressStream_generic ( cctx - > mtctx , output , input , endOp ) ;
2021-05-14 22:48:15 +00:00
cctx - > consumedSrcSize + = ( U64 ) ( input - > pos - ipos ) ;
cctx - > producedCSize + = ( U64 ) ( output - > pos - opos ) ;
2020-02-08 13:56:59 +00:00
if ( ZSTD_isError ( flushMin )
| | ( endOp = = ZSTD_e_end & & flushMin = = 0 ) ) { /* compression completed */
2021-05-14 22:48:15 +00:00
if ( flushMin = = 0 )
ZSTD_CCtx_trace ( cctx , 0 ) ;
2020-02-08 13:56:59 +00:00
ZSTD_CCtx_reset ( cctx , ZSTD_reset_session_only ) ;
}
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( flushMin , " ZSTDMT_compressStream_generic failed " ) ;
2020-12-18 14:46:27 +00:00
if ( endOp = = ZSTD_e_continue ) {
/* We only require some progress with ZSTD_e_continue, not maximal progress.
* We ' re done if we ' ve consumed or produced any bytes , or either buffer is
* full .
*/
if ( input - > pos ! = ipos | | output - > pos ! = opos | | input - > pos = = input - > size | | output - > pos = = output - > size )
break ;
} else {
assert ( endOp = = ZSTD_e_flush | | endOp = = ZSTD_e_end ) ;
/* We require maximal progress. We're done when the flush is complete or the
* output buffer is full .
*/
if ( flushMin = = 0 | | output - > pos = = output - > size )
break ;
}
}
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 5 , " completed ZSTD_compressStream2 delegating to ZSTDMT_compressStream_generic " ) ;
/* Either we don't require maximum forward progress, we've finished the
* flush , or we are out of output space .
*/
2020-12-18 14:46:27 +00:00
assert ( endOp = = ZSTD_e_continue | | flushMin = = 0 | | output - > pos = = output - > size ) ;
ZSTD_setBufferExpectations ( cctx , output , input ) ;
2020-02-08 13:56:59 +00:00
return flushMin ;
}
2023-02-10 19:25:40 +00:00
# endif /* ZSTD_MULTITHREAD */
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( ZSTD_compressStream_generic ( cctx , output , input , endOp ) , " " ) ;
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 5 , " completed ZSTD_compressStream2 " ) ;
2020-12-18 14:46:27 +00:00
ZSTD_setBufferExpectations ( cctx , output , input ) ;
2020-02-08 13:56:59 +00:00
return cctx - > outBuffContentSize - cctx - > outBuffFlushedSize ; /* remaining to flush */
}
size_t ZSTD_compressStream2_simpleArgs (
ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity , size_t * dstPos ,
const void * src , size_t srcSize , size_t * srcPos ,
ZSTD_EndDirective endOp )
{
2023-02-10 19:25:40 +00:00
ZSTD_outBuffer output ;
ZSTD_inBuffer input ;
output . dst = dst ;
output . size = dstCapacity ;
output . pos = * dstPos ;
input . src = src ;
input . size = srcSize ;
input . pos = * srcPos ;
2020-02-08 13:56:59 +00:00
/* ZSTD_compressStream2() will check validity of dstPos and srcPos */
2023-02-10 19:25:40 +00:00
{ size_t const cErr = ZSTD_compressStream2 ( cctx , & output , & input , endOp ) ;
* dstPos = output . pos ;
* srcPos = input . pos ;
return cErr ;
}
2020-02-08 13:56:59 +00:00
}
size_t ZSTD_compress2 ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const void * src , size_t srcSize )
{
2020-12-18 14:46:27 +00:00
ZSTD_bufferMode_e const originalInBufferMode = cctx - > requestedParams . inBufferMode ;
ZSTD_bufferMode_e const originalOutBufferMode = cctx - > requestedParams . outBufferMode ;
2020-06-04 19:03:27 +00:00
DEBUGLOG ( 4 , " ZSTD_compress2 (srcSize=%u) " , ( unsigned ) srcSize ) ;
2020-02-08 13:56:59 +00:00
ZSTD_CCtx_reset ( cctx , ZSTD_reset_session_only ) ;
2020-12-18 14:46:27 +00:00
/* Enable stable input/output buffers. */
cctx - > requestedParams . inBufferMode = ZSTD_bm_stable ;
cctx - > requestedParams . outBufferMode = ZSTD_bm_stable ;
2020-02-08 13:56:59 +00:00
{ size_t oPos = 0 ;
size_t iPos = 0 ;
size_t const result = ZSTD_compressStream2_simpleArgs ( cctx ,
dst , dstCapacity , & oPos ,
src , srcSize , & iPos ,
ZSTD_e_end ) ;
2020-12-18 14:46:27 +00:00
/* Reset to the original values. */
cctx - > requestedParams . inBufferMode = originalInBufferMode ;
cctx - > requestedParams . outBufferMode = originalOutBufferMode ;
2023-02-10 19:25:40 +00:00
2020-06-04 19:03:27 +00:00
FORWARD_IF_ERROR ( result , " ZSTD_compressStream2_simpleArgs failed " ) ;
2020-02-08 13:56:59 +00:00
if ( result ! = 0 ) { /* compression not completed, due to lack of output space */
assert ( oPos = = dstCapacity ) ;
2020-06-04 19:03:27 +00:00
RETURN_ERROR ( dstSize_tooSmall , " " ) ;
2020-02-08 13:56:59 +00:00
}
assert ( iPos = = srcSize ) ; /* all input is expected consumed */
return oPos ;
}
}
2022-03-18 00:52:11 +00:00
/* ZSTD_validateSequence() :
* @ offCode : is presumed to follow format required by ZSTD_storeSeq ( )
* @ returns a ZSTD error code if sequence is not valid
*/
static size_t
2023-02-10 19:25:40 +00:00
ZSTD_validateSequence ( U32 offCode , U32 matchLength , U32 minMatch ,
size_t posInSrc , U32 windowLog , size_t dictSize , int useSequenceProducer )
2022-03-18 00:52:11 +00:00
{
2023-02-10 19:25:40 +00:00
U32 const windowSize = 1u < < windowLog ;
/* posInSrc represents the amount of data the decoder would decode up to this point.
2020-12-18 14:46:27 +00:00
* As long as the amount of data decoded is less than or equal to window size , offsets may be
* larger than the total length of output decoded in order to reference the dict , even larger than
* window size . After output surpasses windowSize , we ' re limited to windowSize offsets again .
*/
2022-03-18 00:52:11 +00:00
size_t const offsetBound = posInSrc > windowSize ? ( size_t ) windowSize : posInSrc + ( size_t ) dictSize ;
2023-02-10 19:25:40 +00:00
size_t const matchLenLowerBound = ( minMatch = = 3 | | useSequenceProducer ) ? 3 : 4 ;
RETURN_ERROR_IF ( offCode > OFFSET_TO_OFFBASE ( offsetBound ) , externalSequences_invalid , " Offset too large! " ) ;
/* Validate maxNbSeq is large enough for the given matchLength and minMatch */
RETURN_ERROR_IF ( matchLength < matchLenLowerBound , externalSequences_invalid , " Matchlength too small for the minMatch " ) ;
2020-12-18 14:46:27 +00:00
return 0 ;
}
/* Returns an offset code, given a sequence's raw offset, the ongoing repcode array, and whether litLength == 0 */
2023-02-10 19:25:40 +00:00
static U32 ZSTD_finalizeOffBase ( U32 rawOffset , const U32 rep [ ZSTD_REP_NUM ] , U32 ll0 )
2022-03-18 00:52:11 +00:00
{
2023-02-10 19:25:40 +00:00
U32 offBase = OFFSET_TO_OFFBASE ( rawOffset ) ;
2020-12-18 14:46:27 +00:00
if ( ! ll0 & & rawOffset = = rep [ 0 ] ) {
2023-02-10 19:25:40 +00:00
offBase = REPCODE1_TO_OFFBASE ;
2020-12-18 14:46:27 +00:00
} else if ( rawOffset = = rep [ 1 ] ) {
2023-02-10 19:25:40 +00:00
offBase = REPCODE_TO_OFFBASE ( 2 - ll0 ) ;
2020-12-18 14:46:27 +00:00
} else if ( rawOffset = = rep [ 2 ] ) {
2023-02-10 19:25:40 +00:00
offBase = REPCODE_TO_OFFBASE ( 3 - ll0 ) ;
2020-12-18 14:46:27 +00:00
} else if ( ll0 & & rawOffset = = rep [ 0 ] - 1 ) {
2023-02-10 19:25:40 +00:00
offBase = REPCODE3_TO_OFFBASE ;
2020-12-18 14:46:27 +00:00
}
2023-02-10 19:25:40 +00:00
return offBase ;
2020-12-18 14:46:27 +00:00
}
2023-02-10 19:25:40 +00:00
size_t
2022-03-18 00:52:11 +00:00
ZSTD_copySequencesToSeqStoreExplicitBlockDelim ( ZSTD_CCtx * cctx ,
ZSTD_sequencePosition * seqPos ,
const ZSTD_Sequence * const inSeqs , size_t inSeqsSize ,
2023-02-10 19:25:40 +00:00
const void * src , size_t blockSize ,
ZSTD_paramSwitch_e externalRepSearch )
2022-03-18 00:52:11 +00:00
{
2020-12-18 14:46:27 +00:00
U32 idx = seqPos - > idx ;
2023-02-10 19:25:40 +00:00
U32 const startIdx = idx ;
2020-12-18 14:46:27 +00:00
BYTE const * ip = ( BYTE const * ) ( src ) ;
const BYTE * const iend = ip + blockSize ;
repcodes_t updatedRepcodes ;
U32 dictSize ;
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " ZSTD_copySequencesToSeqStoreExplicitBlockDelim (blockSize = %zu) " , blockSize ) ;
2020-12-18 14:46:27 +00:00
if ( cctx - > cdict ) {
dictSize = ( U32 ) cctx - > cdict - > dictContentSize ;
} else if ( cctx - > prefixDict . dict ) {
dictSize = ( U32 ) cctx - > prefixDict . dictSize ;
} else {
dictSize = 0 ;
}
ZSTD_memcpy ( updatedRepcodes . rep , cctx - > blockState . prevCBlock - > rep , sizeof ( repcodes_t ) ) ;
2023-02-10 19:25:40 +00:00
for ( ; idx < inSeqsSize & & ( inSeqs [ idx ] . matchLength ! = 0 | | inSeqs [ idx ] . offset ! = 0 ) ; + + idx ) {
2022-03-18 00:52:11 +00:00
U32 const litLength = inSeqs [ idx ] . litLength ;
U32 const matchLength = inSeqs [ idx ] . matchLength ;
2023-02-10 19:25:40 +00:00
U32 offBase ;
if ( externalRepSearch = = ZSTD_ps_disable ) {
offBase = OFFSET_TO_OFFBASE ( inSeqs [ idx ] . offset ) ;
} else {
U32 const ll0 = ( litLength = = 0 ) ;
offBase = ZSTD_finalizeOffBase ( inSeqs [ idx ] . offset , updatedRepcodes . rep , ll0 ) ;
ZSTD_updateRep ( updatedRepcodes . rep , offBase , ll0 ) ;
}
2020-12-18 14:46:27 +00:00
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 6 , " Storing sequence: (of: %u, ml: %u, ll: %u) " , offBase , matchLength , litLength ) ;
2020-12-18 14:46:27 +00:00
if ( cctx - > appliedParams . validateSequences ) {
seqPos - > posInSrc + = litLength + matchLength ;
2023-02-10 19:25:40 +00:00
FORWARD_IF_ERROR ( ZSTD_validateSequence ( offBase , matchLength , cctx - > appliedParams . cParams . minMatch , seqPos - > posInSrc ,
cctx - > appliedParams . cParams . windowLog , dictSize , cctx - > appliedParams . useSequenceProducer ) ,
2020-12-18 14:46:27 +00:00
" Sequence validation failed " ) ;
}
2023-02-10 19:25:40 +00:00
RETURN_ERROR_IF ( idx - seqPos - > idx > = cctx - > seqStore . maxNbSeq , externalSequences_invalid ,
2020-12-18 14:46:27 +00:00
" Not enough memory allocated. Try adjusting ZSTD_c_minMatch. " ) ;
2023-02-10 19:25:40 +00:00
ZSTD_storeSeq ( & cctx - > seqStore , litLength , ip , iend , offBase , matchLength ) ;
2020-12-18 14:46:27 +00:00
ip + = matchLength + litLength ;
}
2023-02-10 19:25:40 +00:00
/* If we skipped repcode search while parsing, we need to update repcodes now */
assert ( externalRepSearch ! = ZSTD_ps_auto ) ;
assert ( idx > = startIdx ) ;
if ( externalRepSearch = = ZSTD_ps_disable & & idx ! = startIdx ) {
U32 * const rep = updatedRepcodes . rep ;
U32 lastSeqIdx = idx - 1 ; /* index of last non-block-delimiter sequence */
if ( lastSeqIdx > = startIdx + 2 ) {
rep [ 2 ] = inSeqs [ lastSeqIdx - 2 ] . offset ;
rep [ 1 ] = inSeqs [ lastSeqIdx - 1 ] . offset ;
rep [ 0 ] = inSeqs [ lastSeqIdx ] . offset ;
} else if ( lastSeqIdx = = startIdx + 1 ) {
rep [ 2 ] = rep [ 0 ] ;
rep [ 1 ] = inSeqs [ lastSeqIdx - 1 ] . offset ;
rep [ 0 ] = inSeqs [ lastSeqIdx ] . offset ;
} else {
assert ( lastSeqIdx = = startIdx ) ;
rep [ 2 ] = rep [ 1 ] ;
rep [ 1 ] = rep [ 0 ] ;
rep [ 0 ] = inSeqs [ lastSeqIdx ] . offset ;
}
}
2020-12-18 14:46:27 +00:00
ZSTD_memcpy ( cctx - > blockState . nextCBlock - > rep , updatedRepcodes . rep , sizeof ( repcodes_t ) ) ;
if ( inSeqs [ idx ] . litLength ) {
DEBUGLOG ( 6 , " Storing last literals of size: %u " , inSeqs [ idx ] . litLength ) ;
ZSTD_storeLastLiterals ( & cctx - > seqStore , ip , inSeqs [ idx ] . litLength ) ;
ip + = inSeqs [ idx ] . litLength ;
seqPos - > posInSrc + = inSeqs [ idx ] . litLength ;
}
2023-02-10 19:25:40 +00:00
RETURN_ERROR_IF ( ip ! = iend , externalSequences_invalid , " Blocksize doesn't agree with block delimiter! " ) ;
2020-12-18 14:46:27 +00:00
seqPos - > idx = idx + 1 ;
return 0 ;
}
2023-02-10 19:25:40 +00:00
size_t
2022-03-18 00:52:11 +00:00
ZSTD_copySequencesToSeqStoreNoBlockDelim ( ZSTD_CCtx * cctx , ZSTD_sequencePosition * seqPos ,
const ZSTD_Sequence * const inSeqs , size_t inSeqsSize ,
2023-02-10 19:25:40 +00:00
const void * src , size_t blockSize , ZSTD_paramSwitch_e externalRepSearch )
2022-03-18 00:52:11 +00:00
{
2020-12-18 14:46:27 +00:00
U32 idx = seqPos - > idx ;
U32 startPosInSequence = seqPos - > posInSequence ;
U32 endPosInSequence = seqPos - > posInSequence + ( U32 ) blockSize ;
size_t dictSize ;
BYTE const * ip = ( BYTE const * ) ( src ) ;
BYTE const * iend = ip + blockSize ; /* May be adjusted if we decide to process fewer than blockSize bytes */
repcodes_t updatedRepcodes ;
U32 bytesAdjustment = 0 ;
U32 finalMatchSplit = 0 ;
2023-02-10 19:25:40 +00:00
/* TODO(embg) support fast parsing mode in noBlockDelim mode */
( void ) externalRepSearch ;
2020-12-18 14:46:27 +00:00
if ( cctx - > cdict ) {
dictSize = cctx - > cdict - > dictContentSize ;
} else if ( cctx - > prefixDict . dict ) {
dictSize = cctx - > prefixDict . dictSize ;
} else {
dictSize = 0 ;
}
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " ZSTD_copySequencesToSeqStoreNoBlockDelim: idx: %u PIS: %u blockSize: %zu " , idx , startPosInSequence , blockSize ) ;
2020-12-18 14:46:27 +00:00
DEBUGLOG ( 5 , " Start seq: idx: %u (of: %u ml: %u ll: %u) " , idx , inSeqs [ idx ] . offset , inSeqs [ idx ] . matchLength , inSeqs [ idx ] . litLength ) ;
ZSTD_memcpy ( updatedRepcodes . rep , cctx - > blockState . prevCBlock - > rep , sizeof ( repcodes_t ) ) ;
while ( endPosInSequence & & idx < inSeqsSize & & ! finalMatchSplit ) {
const ZSTD_Sequence currSeq = inSeqs [ idx ] ;
2022-03-18 00:52:11 +00:00
U32 litLength = currSeq . litLength ;
U32 matchLength = currSeq . matchLength ;
U32 const rawOffset = currSeq . offset ;
2023-02-10 19:25:40 +00:00
U32 offBase ;
2020-12-18 14:46:27 +00:00
/* Modify the sequence depending on where endPosInSequence lies */
if ( endPosInSequence > = currSeq . litLength + currSeq . matchLength ) {
if ( startPosInSequence > = litLength ) {
startPosInSequence - = litLength ;
litLength = 0 ;
matchLength - = startPosInSequence ;
} else {
litLength - = startPosInSequence ;
}
/* Move to the next sequence */
endPosInSequence - = currSeq . litLength + currSeq . matchLength ;
startPosInSequence = 0 ;
} else {
/* This is the final (partial) sequence we're adding from inSeqs, and endPosInSequence
does not reach the end of the match . So , we have to split the sequence */
DEBUGLOG ( 6 , " Require a split: diff: %u, idx: %u PIS: %u " ,
currSeq . litLength + currSeq . matchLength - endPosInSequence , idx , endPosInSequence ) ;
if ( endPosInSequence > litLength ) {
U32 firstHalfMatchLength ;
litLength = startPosInSequence > = litLength ? 0 : litLength - startPosInSequence ;
firstHalfMatchLength = endPosInSequence - startPosInSequence - litLength ;
if ( matchLength > blockSize & & firstHalfMatchLength > = cctx - > appliedParams . cParams . minMatch ) {
/* Only ever split the match if it is larger than the block size */
U32 secondHalfMatchLength = currSeq . matchLength + currSeq . litLength - endPosInSequence ;
if ( secondHalfMatchLength < cctx - > appliedParams . cParams . minMatch ) {
/* Move the endPosInSequence backward so that it creates match of minMatch length */
endPosInSequence - = cctx - > appliedParams . cParams . minMatch - secondHalfMatchLength ;
bytesAdjustment = cctx - > appliedParams . cParams . minMatch - secondHalfMatchLength ;
firstHalfMatchLength - = bytesAdjustment ;
}
matchLength = firstHalfMatchLength ;
/* Flag that we split the last match - after storing the sequence, exit the loop,
but keep the value of endPosInSequence */
finalMatchSplit = 1 ;
} else {
/* Move the position in sequence backwards so that we don't split match, and break to store
* the last literals . We use the original currSeq . litLength as a marker for where endPosInSequence
* should go . We prefer to do this whenever it is not necessary to split the match , or if doing so
* would cause the first half of the match to be too small
*/
bytesAdjustment = endPosInSequence - currSeq . litLength ;
endPosInSequence = currSeq . litLength ;
break ;
}
} else {
/* This sequence ends inside the literals, break to store the last literals */
break ;
}
}
/* Check if this offset can be represented with a repcode */
2022-03-18 00:52:11 +00:00
{ U32 const ll0 = ( litLength = = 0 ) ;
2023-02-10 19:25:40 +00:00
offBase = ZSTD_finalizeOffBase ( rawOffset , updatedRepcodes . rep , ll0 ) ;
ZSTD_updateRep ( updatedRepcodes . rep , offBase , ll0 ) ;
2020-12-18 14:46:27 +00:00
}
if ( cctx - > appliedParams . validateSequences ) {
seqPos - > posInSrc + = litLength + matchLength ;
2023-02-10 19:25:40 +00:00
FORWARD_IF_ERROR ( ZSTD_validateSequence ( offBase , matchLength , cctx - > appliedParams . cParams . minMatch , seqPos - > posInSrc ,
cctx - > appliedParams . cParams . windowLog , dictSize , cctx - > appliedParams . useSequenceProducer ) ,
2020-12-18 14:46:27 +00:00
" Sequence validation failed " ) ;
}
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 6 , " Storing sequence: (of: %u, ml: %u, ll: %u) " , offBase , matchLength , litLength ) ;
RETURN_ERROR_IF ( idx - seqPos - > idx > = cctx - > seqStore . maxNbSeq , externalSequences_invalid ,
2020-12-18 14:46:27 +00:00
" Not enough memory allocated. Try adjusting ZSTD_c_minMatch. " ) ;
2023-02-10 19:25:40 +00:00
ZSTD_storeSeq ( & cctx - > seqStore , litLength , ip , iend , offBase , matchLength ) ;
2020-12-18 14:46:27 +00:00
ip + = matchLength + litLength ;
2023-02-10 19:25:40 +00:00
if ( ! finalMatchSplit )
idx + + ; /* Next Sequence */
2020-12-18 14:46:27 +00:00
}
DEBUGLOG ( 5 , " Ending seq: idx: %u (of: %u ml: %u ll: %u) " , idx , inSeqs [ idx ] . offset , inSeqs [ idx ] . matchLength , inSeqs [ idx ] . litLength ) ;
assert ( idx = = inSeqsSize | | endPosInSequence < = inSeqs [ idx ] . litLength + inSeqs [ idx ] . matchLength ) ;
seqPos - > idx = idx ;
seqPos - > posInSequence = endPosInSequence ;
ZSTD_memcpy ( cctx - > blockState . nextCBlock - > rep , updatedRepcodes . rep , sizeof ( repcodes_t ) ) ;
iend - = bytesAdjustment ;
if ( ip ! = iend ) {
/* Store any last literals */
U32 lastLLSize = ( U32 ) ( iend - ip ) ;
assert ( ip < = iend ) ;
DEBUGLOG ( 6 , " Storing last literals of size: %u " , lastLLSize ) ;
ZSTD_storeLastLiterals ( & cctx - > seqStore , ip , lastLLSize ) ;
seqPos - > posInSrc + = lastLLSize ;
}
return bytesAdjustment ;
}
typedef size_t ( * ZSTD_sequenceCopier ) ( ZSTD_CCtx * cctx , ZSTD_sequencePosition * seqPos ,
const ZSTD_Sequence * const inSeqs , size_t inSeqsSize ,
2023-02-10 19:25:40 +00:00
const void * src , size_t blockSize , ZSTD_paramSwitch_e externalRepSearch ) ;
2022-03-18 00:52:11 +00:00
static ZSTD_sequenceCopier ZSTD_selectSequenceCopier ( ZSTD_sequenceFormat_e mode )
{
2020-12-18 14:46:27 +00:00
ZSTD_sequenceCopier sequenceCopier = NULL ;
assert ( ZSTD_cParam_withinBounds ( ZSTD_c_blockDelimiters , mode ) ) ;
if ( mode = = ZSTD_sf_explicitBlockDelimiters ) {
return ZSTD_copySequencesToSeqStoreExplicitBlockDelim ;
} else if ( mode = = ZSTD_sf_noBlockDelimiters ) {
return ZSTD_copySequencesToSeqStoreNoBlockDelim ;
}
assert ( sequenceCopier ! = NULL ) ;
return sequenceCopier ;
}
2023-02-10 19:25:40 +00:00
/* Discover the size of next block by searching for the delimiter.
* Note that a block delimiter * * must * * exist in this mode ,
* otherwise it ' s an input error .
* The block size retrieved will be later compared to ensure it remains within bounds */
static size_t
blockSize_explicitDelimiter ( const ZSTD_Sequence * inSeqs , size_t inSeqsSize , ZSTD_sequencePosition seqPos )
{
int end = 0 ;
size_t blockSize = 0 ;
size_t spos = seqPos . idx ;
DEBUGLOG ( 6 , " blockSize_explicitDelimiter : seq %zu / %zu " , spos , inSeqsSize ) ;
assert ( spos < = inSeqsSize ) ;
while ( spos < inSeqsSize ) {
end = ( inSeqs [ spos ] . offset = = 0 ) ;
blockSize + = inSeqs [ spos ] . litLength + inSeqs [ spos ] . matchLength ;
if ( end ) {
if ( inSeqs [ spos ] . matchLength ! = 0 )
RETURN_ERROR ( externalSequences_invalid , " delimiter format error : both matchlength and offset must be == 0 " ) ;
break ;
}
spos + + ;
}
if ( ! end )
RETURN_ERROR ( externalSequences_invalid , " Reached end of sequences without finding a block delimiter " ) ;
return blockSize ;
}
/* More a "target" block size */
static size_t blockSize_noDelimiter ( size_t blockSize , size_t remaining )
{
int const lastBlock = ( remaining < = blockSize ) ;
return lastBlock ? remaining : blockSize ;
}
static size_t determine_blockSize ( ZSTD_sequenceFormat_e mode ,
size_t blockSize , size_t remaining ,
const ZSTD_Sequence * inSeqs , size_t inSeqsSize , ZSTD_sequencePosition seqPos )
{
DEBUGLOG ( 6 , " determine_blockSize : remainingSize = %zu " , remaining ) ;
if ( mode = = ZSTD_sf_noBlockDelimiters )
return blockSize_noDelimiter ( blockSize , remaining ) ;
{ size_t const explicitBlockSize = blockSize_explicitDelimiter ( inSeqs , inSeqsSize , seqPos ) ;
FORWARD_IF_ERROR ( explicitBlockSize , " Error while determining block size with explicit delimiters " ) ;
if ( explicitBlockSize > blockSize )
RETURN_ERROR ( externalSequences_invalid , " sequences incorrectly define a too large block " ) ;
if ( explicitBlockSize > remaining )
RETURN_ERROR ( externalSequences_invalid , " sequences define a frame longer than source " ) ;
return explicitBlockSize ;
}
}
2020-12-18 14:46:27 +00:00
/* Compress, block-by-block, all of the sequences given.
*
2022-03-18 00:52:11 +00:00
* Returns the cumulative size of all compressed blocks ( including their headers ) ,
* otherwise a ZSTD error .
2020-12-18 14:46:27 +00:00
*/
2022-03-18 00:52:11 +00:00
static size_t
ZSTD_compressSequences_internal ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
const ZSTD_Sequence * inSeqs , size_t inSeqsSize ,
const void * src , size_t srcSize )
{
2020-12-18 14:46:27 +00:00
size_t cSize = 0 ;
size_t remaining = srcSize ;
ZSTD_sequencePosition seqPos = { 0 , 0 , 0 } ;
BYTE const * ip = ( BYTE const * ) src ;
BYTE * op = ( BYTE * ) dst ;
2022-03-18 00:52:11 +00:00
ZSTD_sequenceCopier const sequenceCopier = ZSTD_selectSequenceCopier ( cctx - > appliedParams . blockDelimiters ) ;
2020-12-18 14:46:27 +00:00
DEBUGLOG ( 4 , " ZSTD_compressSequences_internal srcSize: %zu, inSeqsSize: %zu " , srcSize , inSeqsSize ) ;
/* Special case: empty frame */
if ( remaining = = 0 ) {
U32 const cBlockHeader24 = 1 /* last block */ + ( ( ( U32 ) bt_raw ) < < 1 ) ;
RETURN_ERROR_IF ( dstCapacity < 4 , dstSize_tooSmall , " No room for empty frame block header " ) ;
MEM_writeLE32 ( op , cBlockHeader24 ) ;
op + = ZSTD_blockHeaderSize ;
dstCapacity - = ZSTD_blockHeaderSize ;
cSize + = ZSTD_blockHeaderSize ;
}
while ( remaining ) {
2023-02-10 19:25:40 +00:00
size_t compressedSeqsSize ;
2020-12-18 14:46:27 +00:00
size_t cBlockSize ;
size_t additionalByteAdjustment ;
2023-02-10 19:25:40 +00:00
size_t blockSize = determine_blockSize ( cctx - > appliedParams . blockDelimiters ,
cctx - > blockSize , remaining ,
inSeqs , inSeqsSize , seqPos ) ;
U32 const lastBlock = ( blockSize = = remaining ) ;
FORWARD_IF_ERROR ( blockSize , " Error while trying to determine block size " ) ;
assert ( blockSize < = remaining ) ;
2020-12-18 14:46:27 +00:00
ZSTD_resetSeqStore ( & cctx - > seqStore ) ;
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " Working on new block. Blocksize: %zu (total:%zu) " , blockSize , ( ip - ( const BYTE * ) src ) + blockSize ) ;
2020-12-18 14:46:27 +00:00
2023-02-10 19:25:40 +00:00
additionalByteAdjustment = sequenceCopier ( cctx , & seqPos , inSeqs , inSeqsSize , ip , blockSize , cctx - > appliedParams . searchForExternalRepcodes ) ;
2020-12-18 14:46:27 +00:00
FORWARD_IF_ERROR ( additionalByteAdjustment , " Bad sequence copy " ) ;
blockSize - = additionalByteAdjustment ;
/* If blocks are too small, emit as a nocompress block */
2023-02-10 19:25:40 +00:00
/* TODO: See 3090. We reduced MIN_CBLOCK_SIZE from 3 to 2 so to compensate we are adding
* additional 1. We need to revisit and change this logic to be more consistent */
if ( blockSize < MIN_CBLOCK_SIZE + ZSTD_blockHeaderSize + 1 + 1 ) {
2020-12-18 14:46:27 +00:00
cBlockSize = ZSTD_noCompressBlock ( op , dstCapacity , ip , blockSize , lastBlock ) ;
FORWARD_IF_ERROR ( cBlockSize , " Nocompress block failed " ) ;
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " Block too small, writing out nocompress block: cSize: %zu " , cBlockSize ) ;
2020-12-18 14:46:27 +00:00
cSize + = cBlockSize ;
ip + = blockSize ;
op + = cBlockSize ;
remaining - = blockSize ;
dstCapacity - = cBlockSize ;
continue ;
}
2023-02-10 19:25:40 +00:00
RETURN_ERROR_IF ( dstCapacity < ZSTD_blockHeaderSize , dstSize_tooSmall , " not enough dstCapacity to write a new compressed block " ) ;
2021-05-14 22:48:15 +00:00
compressedSeqsSize = ZSTD_entropyCompressSeqStore ( & cctx - > seqStore ,
2020-12-18 14:46:27 +00:00
& cctx - > blockState . prevCBlock - > entropy , & cctx - > blockState . nextCBlock - > entropy ,
& cctx - > appliedParams ,
op + ZSTD_blockHeaderSize /* Leave space for block header */ , dstCapacity - ZSTD_blockHeaderSize ,
blockSize ,
cctx - > entropyWorkspace , ENTROPY_WORKSPACE_SIZE /* statically allocated in resetCCtx */ ,
cctx - > bmi2 ) ;
FORWARD_IF_ERROR ( compressedSeqsSize , " Compressing sequences of block failed " ) ;
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " Compressed sequences size: %zu " , compressedSeqsSize ) ;
2020-12-18 14:46:27 +00:00
if ( ! cctx - > isFirstBlock & &
ZSTD_maybeRLE ( & cctx - > seqStore ) & &
2023-02-10 19:25:40 +00:00
ZSTD_isRLE ( ip , blockSize ) ) {
2020-12-18 14:46:27 +00:00
/* We don't want to emit our first block as a RLE even if it qualifies because
* doing so will cause the decoder ( cli only ) to throw a " should consume all input error. "
* This is only an issue for zstd < = v1 .4 .3
*/
compressedSeqsSize = 1 ;
}
if ( compressedSeqsSize = = 0 ) {
/* ZSTD_noCompressBlock writes the block header as well */
cBlockSize = ZSTD_noCompressBlock ( op , dstCapacity , ip , blockSize , lastBlock ) ;
2023-02-10 19:25:40 +00:00
FORWARD_IF_ERROR ( cBlockSize , " ZSTD_noCompressBlock failed " ) ;
DEBUGLOG ( 5 , " Writing out nocompress block, size: %zu " , cBlockSize ) ;
2020-12-18 14:46:27 +00:00
} else if ( compressedSeqsSize = = 1 ) {
cBlockSize = ZSTD_rleCompressBlock ( op , dstCapacity , * ip , blockSize , lastBlock ) ;
2023-02-10 19:25:40 +00:00
FORWARD_IF_ERROR ( cBlockSize , " ZSTD_rleCompressBlock failed " ) ;
DEBUGLOG ( 5 , " Writing out RLE block, size: %zu " , cBlockSize ) ;
2020-12-18 14:46:27 +00:00
} else {
U32 cBlockHeader ;
/* Error checking and repcodes update */
2021-05-14 22:48:15 +00:00
ZSTD_blockState_confirmRepcodesAndEntropyTables ( & cctx - > blockState ) ;
2020-12-18 14:46:27 +00:00
if ( cctx - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = = FSE_repeat_valid )
cctx - > blockState . prevCBlock - > entropy . fse . offcode_repeatMode = FSE_repeat_check ;
/* Write block header into beginning of block*/
cBlockHeader = lastBlock + ( ( ( U32 ) bt_compressed ) < < 1 ) + ( U32 ) ( compressedSeqsSize < < 3 ) ;
MEM_writeLE24 ( op , cBlockHeader ) ;
cBlockSize = ZSTD_blockHeaderSize + compressedSeqsSize ;
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " Writing out compressed block, size: %zu " , cBlockSize ) ;
2020-12-18 14:46:27 +00:00
}
cSize + = cBlockSize ;
if ( lastBlock ) {
break ;
} else {
ip + = blockSize ;
op + = cBlockSize ;
remaining - = blockSize ;
dstCapacity - = cBlockSize ;
cctx - > isFirstBlock = 0 ;
}
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 5 , " cSize running total: %zu (remaining dstCapacity=%zu) " , cSize , dstCapacity ) ;
2020-12-18 14:46:27 +00:00
}
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 4 , " cSize final total: %zu " , cSize ) ;
2020-12-18 14:46:27 +00:00
return cSize ;
}
2023-02-10 19:25:40 +00:00
size_t ZSTD_compressSequences ( ZSTD_CCtx * cctx ,
void * dst , size_t dstCapacity ,
2020-12-18 14:46:27 +00:00
const ZSTD_Sequence * inSeqs , size_t inSeqsSize ,
2022-03-18 00:52:11 +00:00
const void * src , size_t srcSize )
{
2020-12-18 14:46:27 +00:00
BYTE * op = ( BYTE * ) dst ;
size_t cSize = 0 ;
size_t compressedBlocksSize = 0 ;
size_t frameHeaderSize = 0 ;
/* Transparent initialization stage, same as compressStream2() */
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 4 , " ZSTD_compressSequences (dstCapacity=%zu) " , dstCapacity ) ;
2020-12-18 14:46:27 +00:00
assert ( cctx ! = NULL ) ;
FORWARD_IF_ERROR ( ZSTD_CCtx_init_compressStream2 ( cctx , ZSTD_e_end , srcSize ) , " CCtx initialization failed " ) ;
/* Begin writing output, starting with frame header */
frameHeaderSize = ZSTD_writeFrameHeader ( op , dstCapacity , & cctx - > appliedParams , srcSize , cctx - > dictID ) ;
op + = frameHeaderSize ;
dstCapacity - = frameHeaderSize ;
cSize + = frameHeaderSize ;
if ( cctx - > appliedParams . fParams . checksumFlag & & srcSize ) {
XXH64_update ( & cctx - > xxhState , src , srcSize ) ;
}
/* cSize includes block header size and compressed sequences size */
compressedBlocksSize = ZSTD_compressSequences_internal ( cctx ,
op , dstCapacity ,
inSeqs , inSeqsSize ,
src , srcSize ) ;
FORWARD_IF_ERROR ( compressedBlocksSize , " Compressing blocks failed! " ) ;
cSize + = compressedBlocksSize ;
dstCapacity - = compressedBlocksSize ;
if ( cctx - > appliedParams . fParams . checksumFlag ) {
U32 const checksum = ( U32 ) XXH64_digest ( & cctx - > xxhState ) ;
RETURN_ERROR_IF ( dstCapacity < 4 , dstSize_tooSmall , " no room for checksum " ) ;
DEBUGLOG ( 4 , " Write checksum : %08X " , ( unsigned ) checksum ) ;
MEM_writeLE32 ( ( char * ) dst + cSize , checksum ) ;
cSize + = 4 ;
}
2023-02-10 19:25:40 +00:00
DEBUGLOG ( 4 , " Final compressed size: %zu " , cSize ) ;
2020-12-18 14:46:27 +00:00
return cSize ;
}
2020-02-08 13:56:59 +00:00
/*====== Finalize ======*/
2023-02-10 19:25:40 +00:00
static ZSTD_inBuffer inBuffer_forEndFlush ( const ZSTD_CStream * zcs )
{
const ZSTD_inBuffer nullInput = { NULL , 0 , 0 } ;
const int stableInput = ( zcs - > appliedParams . inBufferMode = = ZSTD_bm_stable ) ;
return stableInput ? zcs - > expectedInBuffer : nullInput ;
}
2020-02-08 13:56:59 +00:00
/*! ZSTD_flushStream() :
* @ return : amount of data remaining to flush */
size_t ZSTD_flushStream ( ZSTD_CStream * zcs , ZSTD_outBuffer * output )
{
2023-02-10 19:25:40 +00:00
ZSTD_inBuffer input = inBuffer_forEndFlush ( zcs ) ;
input . size = input . pos ; /* do not ingest more input during flush */
2020-02-08 13:56:59 +00:00
return ZSTD_compressStream2 ( zcs , output , & input , ZSTD_e_flush ) ;
}
size_t ZSTD_endStream ( ZSTD_CStream * zcs , ZSTD_outBuffer * output )
{
2023-02-10 19:25:40 +00:00
ZSTD_inBuffer input = inBuffer_forEndFlush ( zcs ) ;
2020-02-08 13:56:59 +00:00
size_t const remainingToFlush = ZSTD_compressStream2 ( zcs , output , & input , ZSTD_e_end ) ;
2023-02-10 19:25:40 +00:00
FORWARD_IF_ERROR ( remainingToFlush , " ZSTD_compressStream2(,,ZSTD_e_end) failed " ) ;
2020-02-08 13:56:59 +00:00
if ( zcs - > appliedParams . nbWorkers > 0 ) return remainingToFlush ; /* minimal estimation */
/* single thread mode : attempt to calculate remaining to flush more precisely */
{ size_t const lastBlockSize = zcs - > frameEnded ? 0 : ZSTD_BLOCKHEADERSIZE ;
size_t const checksumSize = ( size_t ) ( zcs - > frameEnded ? 0 : zcs - > appliedParams . fParams . checksumFlag * 4 ) ;
size_t const toFlush = remainingToFlush + lastBlockSize + checksumSize ;
DEBUGLOG ( 4 , " ZSTD_endStream : remaining to flush : %u " , ( unsigned ) toFlush ) ;
return toFlush ;
}
}
/*-===== Pre-defined compression levels =====-*/
2021-12-21 11:21:29 +00:00
# include "clevels.h"
2020-02-08 13:56:59 +00:00
int ZSTD_maxCLevel ( void ) { return ZSTD_MAX_CLEVEL ; }
int ZSTD_minCLevel ( void ) { return ( int ) - ZSTD_TARGETLENGTH_MAX ; }
2021-05-14 22:48:15 +00:00
int ZSTD_defaultCLevel ( void ) { return ZSTD_CLEVEL_DEFAULT ; }
2020-02-08 13:56:59 +00:00
2020-12-18 14:46:27 +00:00
static ZSTD_compressionParameters ZSTD_dedicatedDictSearch_getCParams ( int const compressionLevel , size_t const dictSize )
{
ZSTD_compressionParameters cParams = ZSTD_getCParams_internal ( compressionLevel , 0 , dictSize , ZSTD_cpm_createCDict ) ;
switch ( cParams . strategy ) {
case ZSTD_fast :
case ZSTD_dfast :
break ;
case ZSTD_greedy :
case ZSTD_lazy :
case ZSTD_lazy2 :
cParams . hashLog + = ZSTD_LAZY_DDSS_BUCKET_LOG ;
break ;
case ZSTD_btlazy2 :
case ZSTD_btopt :
case ZSTD_btultra :
case ZSTD_btultra2 :
break ;
}
return cParams ;
}
static int ZSTD_dedicatedDictSearch_isSupported (
ZSTD_compressionParameters const * cParams )
{
2021-05-14 22:48:15 +00:00
return ( cParams - > strategy > = ZSTD_greedy )
& & ( cParams - > strategy < = ZSTD_lazy2 )
& & ( cParams - > hashLog > cParams - > chainLog )
& & ( cParams - > chainLog < = 24 ) ;
2020-12-18 14:46:27 +00:00
}
/**
* Reverses the adjustment applied to cparams when enabling dedicated dict
* search . This is used to recover the params set to be used in the working
* context . ( Otherwise , those tables would also grow . )
*/
static void ZSTD_dedicatedDictSearch_revertCParams (
ZSTD_compressionParameters * cParams ) {
switch ( cParams - > strategy ) {
case ZSTD_fast :
case ZSTD_dfast :
break ;
case ZSTD_greedy :
case ZSTD_lazy :
case ZSTD_lazy2 :
cParams - > hashLog - = ZSTD_LAZY_DDSS_BUCKET_LOG ;
2021-05-14 22:48:15 +00:00
if ( cParams - > hashLog < ZSTD_HASHLOG_MIN ) {
cParams - > hashLog = ZSTD_HASHLOG_MIN ;
}
2020-12-18 14:46:27 +00:00
break ;
case ZSTD_btlazy2 :
case ZSTD_btopt :
case ZSTD_btultra :
case ZSTD_btultra2 :
break ;
}
}
static U64 ZSTD_getCParamRowSize ( U64 srcSizeHint , size_t dictSize , ZSTD_cParamMode_e mode )
{
switch ( mode ) {
case ZSTD_cpm_unknown :
case ZSTD_cpm_noAttachDict :
case ZSTD_cpm_createCDict :
break ;
case ZSTD_cpm_attachDict :
dictSize = 0 ;
break ;
default :
assert ( 0 ) ;
break ;
}
{ int const unknown = srcSizeHint = = ZSTD_CONTENTSIZE_UNKNOWN ;
size_t const addedSize = unknown & & dictSize > 0 ? 500 : 0 ;
return unknown & & dictSize = = 0 ? ZSTD_CONTENTSIZE_UNKNOWN : srcSizeHint + dictSize + addedSize ;
}
}
2020-02-08 13:56:59 +00:00
/*! ZSTD_getCParams_internal() :
* @ return ZSTD_compressionParameters structure for a selected compression level , srcSize and dictSize .
* Note : srcSizeHint 0 means 0 , use ZSTD_CONTENTSIZE_UNKNOWN for unknown .
2020-12-18 14:46:27 +00:00
* Use dictSize = = 0 for unknown or unused .
* Note : ` mode ` controls how we treat the ` dictSize ` . See docs for ` ZSTD_cParamMode_e ` . */
static ZSTD_compressionParameters ZSTD_getCParams_internal ( int compressionLevel , unsigned long long srcSizeHint , size_t dictSize , ZSTD_cParamMode_e mode )
2020-02-08 13:56:59 +00:00
{
2020-12-18 14:46:27 +00:00
U64 const rSize = ZSTD_getCParamRowSize ( srcSizeHint , dictSize , mode ) ;
2020-02-08 13:56:59 +00:00
U32 const tableID = ( rSize < = 256 KB ) + ( rSize < = 128 KB ) + ( rSize < = 16 KB ) ;
2020-12-18 14:46:27 +00:00
int row ;
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 5 , " ZSTD_getCParams_internal (cLevel=%i) " , compressionLevel ) ;
2020-12-18 14:46:27 +00:00
/* row */
2020-02-08 13:56:59 +00:00
if ( compressionLevel = = 0 ) row = ZSTD_CLEVEL_DEFAULT ; /* 0 == default */
2020-12-18 14:46:27 +00:00
else if ( compressionLevel < 0 ) row = 0 ; /* entry 0 is baseline for fast mode */
else if ( compressionLevel > ZSTD_MAX_CLEVEL ) row = ZSTD_MAX_CLEVEL ;
else row = compressionLevel ;
2020-02-08 13:56:59 +00:00
{ ZSTD_compressionParameters cp = ZSTD_defaultCParameters [ tableID ] [ row ] ;
2021-05-14 22:48:15 +00:00
DEBUGLOG ( 5 , " ZSTD_getCParams_internal selected tableID: %u row: %u strat: %u " , tableID , row , ( U32 ) cp . strategy ) ;
2020-12-18 14:46:27 +00:00
/* acceleration factor */
if ( compressionLevel < 0 ) {
int const clampedCompressionLevel = MAX ( ZSTD_minCLevel ( ) , compressionLevel ) ;
cp . targetLength = ( unsigned ) ( - clampedCompressionLevel ) ;
}
2020-02-08 13:56:59 +00:00
/* refine parameters based on srcSize & dictSize */
2023-02-10 19:25:40 +00:00
return ZSTD_adjustCParams_internal ( cp , srcSizeHint , dictSize , mode , ZSTD_ps_auto ) ;
2020-02-08 13:56:59 +00:00
}
}
/*! ZSTD_getCParams() :
* @ return ZSTD_compressionParameters structure for a selected compression level , srcSize and dictSize .
* Size values are optional , provide 0 if not known or unused */
ZSTD_compressionParameters ZSTD_getCParams ( int compressionLevel , unsigned long long srcSizeHint , size_t dictSize )
{
if ( srcSizeHint = = 0 ) srcSizeHint = ZSTD_CONTENTSIZE_UNKNOWN ;
2020-12-18 14:46:27 +00:00
return ZSTD_getCParams_internal ( compressionLevel , srcSizeHint , dictSize , ZSTD_cpm_unknown ) ;
2020-02-08 13:56:59 +00:00
}
/*! ZSTD_getParams() :
* same idea as ZSTD_getCParams ( )
* @ return a ` ZSTD_parameters ` structure ( instead of ` ZSTD_compressionParameters ` ) .
* Fields of ` ZSTD_frameParameters ` are set to default values */
2020-12-18 14:46:27 +00:00
static ZSTD_parameters ZSTD_getParams_internal ( int compressionLevel , unsigned long long srcSizeHint , size_t dictSize , ZSTD_cParamMode_e mode ) {
2020-02-08 13:56:59 +00:00
ZSTD_parameters params ;
2020-12-18 14:46:27 +00:00
ZSTD_compressionParameters const cParams = ZSTD_getCParams_internal ( compressionLevel , srcSizeHint , dictSize , mode ) ;
2020-02-08 13:56:59 +00:00
DEBUGLOG ( 5 , " ZSTD_getParams (cLevel=%i) " , compressionLevel ) ;
2020-12-18 14:46:27 +00:00
ZSTD_memset ( & params , 0 , sizeof ( params ) ) ;
2020-02-08 13:56:59 +00:00
params . cParams = cParams ;
params . fParams . contentSizeFlag = 1 ;
return params ;
}
/*! ZSTD_getParams() :
* same idea as ZSTD_getCParams ( )
* @ return a ` ZSTD_parameters ` structure ( instead of ` ZSTD_compressionParameters ` ) .
* Fields of ` ZSTD_frameParameters ` are set to default values */
ZSTD_parameters ZSTD_getParams ( int compressionLevel , unsigned long long srcSizeHint , size_t dictSize ) {
if ( srcSizeHint = = 0 ) srcSizeHint = ZSTD_CONTENTSIZE_UNKNOWN ;
2020-12-18 14:46:27 +00:00
return ZSTD_getParams_internal ( compressionLevel , srcSizeHint , dictSize , ZSTD_cpm_unknown ) ;
2020-02-08 13:56:59 +00:00
}
2023-02-10 19:25:40 +00:00
void ZSTD_registerSequenceProducer (
ZSTD_CCtx * zc , void * mState ,
ZSTD_sequenceProducer_F * mFinder
) {
if ( mFinder ! = NULL ) {
ZSTD_externalMatchCtx emctx ;
emctx . mState = mState ;
emctx . mFinder = mFinder ;
emctx . seqBuffer = NULL ;
emctx . seqBufferCapacity = 0 ;
zc - > externalMatchCtx = emctx ;
zc - > requestedParams . useSequenceProducer = 1 ;
} else {
ZSTD_memset ( & zc - > externalMatchCtx , 0 , sizeof ( zc - > externalMatchCtx ) ) ;
zc - > requestedParams . useSequenceProducer = 0 ;
}
}