From 31af8290d155091f99c42da7b3787964ca4b37d7 Mon Sep 17 00:00:00 2001
From: Yann Collet
size_t ZSTD_frameHeaderSize(const void* src, size_t srcSize); +`src` should point to the start of a ZSTD frame + `srcSize` must be >= ZSTD_frameHeaderSize_prefix. + @return : size of the Frame Header +
size_t ZSTD_sizeof_CCtx(const ZSTD_CCtx* cctx); @@ -401,12 +407,15 @@ size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict); Object memory usage can evolve if it's re-used multiple times.
size_t ZSTD_estimateCCtxSize(ZSTD_compressionParameters cParams); +size_t ZSTD_estimateCCtxSize(int compressionLevel); +size_t ZSTD_estimateCCtxSize_advanced(ZSTD_compressionParameters cParams); size_t ZSTD_estimateDCtxSize(void);These functions make it possible to estimate memory usage - of a future target object, before its allocation, - given a set of parameters, which vary depending on target object. + of a future {D,C}Ctx, before its creation. The objective is to guide decision before allocation. + ZSTD_estimateCCtxSize() will consider src size to be arbitrarily "large". + If srcSize is known to be small, ZSTD_estimateCCtxSize_advanced() will provide a better (smaller) estimation. + ZSTD_estimateCCtxSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel. Note : CCtx estimation is only correct for single-threaded compression
diff --git a/lib/compress/zstd_compress.c b/lib/compress/zstd_compress.c index e53496091..1bf676926 100644 --- a/lib/compress/zstd_compress.c +++ b/lib/compress/zstd_compress.c @@ -532,7 +532,7 @@ ZSTD_compressionParameters ZSTD_adjustCParams(ZSTD_compressionParameters cPar, u } -size_t ZSTD_estimateCCtxSize(ZSTD_compressionParameters cParams) +size_t ZSTD_estimateCCtxSize_advanced(ZSTD_compressionParameters cParams) { size_t const blockSize = MIN(ZSTD_BLOCKSIZE_MAX, (size_t)1 << cParams.windowLog); U32 const divider = (cParams.searchLength==3) ? 3 : 4; @@ -558,9 +558,15 @@ size_t ZSTD_estimateCCtxSize(ZSTD_compressionParameters cParams) return sizeof(ZSTD_CCtx) + neededSpace; } +size_t ZSTD_estimateCCtxSize(int compressionLevel) +{ + ZSTD_compressionParameters const cParams = ZSTD_getCParams(compressionLevel, 0, 0); + return ZSTD_estimateCCtxSize_advanced(cParams); +} + size_t ZSTD_estimateCStreamSize(ZSTD_compressionParameters cParams) { - size_t const CCtxSize = ZSTD_estimateCCtxSize(cParams); + size_t const CCtxSize = ZSTD_estimateCCtxSize_advanced(cParams); size_t const blockSize = MIN(ZSTD_BLOCKSIZE_MAX, (size_t)1 << cParams.windowLog); size_t const inBuffSize = ((size_t)1 << cParams.windowLog) + blockSize; size_t const outBuffSize = ZSTD_compressBound(blockSize) + 1; @@ -3355,8 +3361,8 @@ size_t ZSTD_compress(void* dst, size_t dstCapacity, const void* src, size_t srcS size_t ZSTD_estimateCDictSize(ZSTD_compressionParameters cParams, size_t dictSize, unsigned byReference) { DEBUGLOG(5, "sizeof(ZSTD_CDict) : %u", (U32)sizeof(ZSTD_CDict)); - DEBUGLOG(5, "CCtx estimate : %u", (U32)ZSTD_estimateCCtxSize(cParams)); - return sizeof(ZSTD_CDict) + ZSTD_estimateCCtxSize(cParams) + DEBUGLOG(5, "CCtx estimate : %u", (U32)ZSTD_estimateCCtxSize_advanced(cParams)); + return sizeof(ZSTD_CDict) + ZSTD_estimateCCtxSize_advanced(cParams) + (byReference ? 0 : dictSize); } @@ -3482,7 +3488,7 @@ ZSTD_CDict* ZSTD_initStaticCDict(void* workspace, size_t workspaceSize, unsigned byReference, ZSTD_dictMode_e dictMode, ZSTD_compressionParameters cParams) { - size_t const cctxSize = ZSTD_estimateCCtxSize(cParams); + size_t const cctxSize = ZSTD_estimateCCtxSize_advanced(cParams); size_t const neededSize = sizeof(ZSTD_CDict) + (byReference ? 0 : dictSize) + cctxSize; ZSTD_CDict* const cdict = (ZSTD_CDict*) workspace; diff --git a/lib/zstd.h b/lib/zstd.h index 8cf9ba75f..65932088c 100644 --- a/lib/zstd.h +++ b/lib/zstd.h @@ -495,11 +495,14 @@ ZSTDLIB_API size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict); /*! ZSTD_estimate*() : * These functions make it possible to estimate memory usage - * of a future target object, before its allocation, - * given a set of parameters, which vary depending on target object. + * of a future {D,C}Ctx, before its creation. * The objective is to guide decision before allocation. + * ZSTD_estimateCCtxSize() will consider src size to be arbitrarily "large". + * If srcSize is known to be small, ZSTD_estimateCCtxSize_advanced() will provide a better (smaller) estimation. + * ZSTD_estimateCCtxSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel. * Note : CCtx estimation is only correct for single-threaded compression */ -ZSTDLIB_API size_t ZSTD_estimateCCtxSize(ZSTD_compressionParameters cParams); +ZSTDLIB_API size_t ZSTD_estimateCCtxSize(int compressionLevel); +ZSTDLIB_API size_t ZSTD_estimateCCtxSize_advanced(ZSTD_compressionParameters cParams); ZSTDLIB_API size_t ZSTD_estimateDCtxSize(void); /*! ZSTD_estimate?StreamSize() : diff --git a/tests/paramgrill.c b/tests/paramgrill.c index 1185c6649..da06ccb52 100644 --- a/tests/paramgrill.c +++ b/tests/paramgrill.c @@ -390,8 +390,8 @@ static int BMK_seed(winnerInfo_t* winners, const ZSTD_compressionParameters para double W_DMemUsed_note = W_ratioNote * ( 40 + 9*cLevel) - log((double)W_DMemUsed); double O_DMemUsed_note = O_ratioNote * ( 40 + 9*cLevel) - log((double)O_DMemUsed); - size_t W_CMemUsed = (1 << params.windowLog) + ZSTD_estimateCCtxSize(params); - size_t O_CMemUsed = (1 << winners[cLevel].params.windowLog) + ZSTD_estimateCCtxSize(winners[cLevel].params); + size_t W_CMemUsed = (1 << params.windowLog) + ZSTD_estimateCCtxSize_advanced(params); + size_t O_CMemUsed = (1 << winners[cLevel].params.windowLog) + ZSTD_estimateCCtxSize_advanced(winners[cLevel].params); double W_CMemUsed_note = W_ratioNote * ( 50 + 13*cLevel) - log((double)W_CMemUsed); double O_CMemUsed_note = O_ratioNote * ( 50 + 13*cLevel) - log((double)O_CMemUsed); From 0c9a915a280d24aa887a4dd19a72195fb8ade5cb Mon Sep 17 00:00:00 2001 From: Yann ColletDate: Mon, 26 Jun 2017 16:02:25 -0700 Subject: [PATCH 2/8] ZSTD_estimateCStreamSize_advanced() --- doc/zstd_manual.html | 10 +++++++--- lib/compress/zstd_compress.c | 7 ++++++- lib/zstd.h | 8 ++++++-- tests/fuzzer.c | 3 +-- tests/zstreamtest.c | 2 +- 5 files changed, 21 insertions(+), 9 deletions(-) diff --git a/doc/zstd_manual.html b/doc/zstd_manual.html index 4397faf7a..2cd4f78c5 100644 --- a/doc/zstd_manual.html +++ b/doc/zstd_manual.html @@ -414,14 +414,18 @@ size_t ZSTD_estimateDCtxSize(void); of a future {D,C}Ctx, before its creation. The objective is to guide decision before allocation. ZSTD_estimateCCtxSize() will consider src size to be arbitrarily "large". - If srcSize is known to be small, ZSTD_estimateCCtxSize_advanced() will provide a better (smaller) estimation. + If srcSize is known to be small, ZSTD_estimateCCtxSize_advanced() can provide a tighter estimation. ZSTD_estimateCCtxSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel. Note : CCtx estimation is only correct for single-threaded compression
size_t ZSTD_estimateCStreamSize(ZSTD_compressionParameters cParams); +size_t ZSTD_estimateCStreamSize(int compressionLevel); +size_t ZSTD_estimateCStreamSize_advanced(ZSTD_compressionParameters cParams); size_t ZSTD_estimateDStreamSize(ZSTD_frameHeader fHeader); -Note : if streaming is init with function ZSTD_init?Stream_usingDict(), +
ZSTD_estimateCStreamSize() will consider src size to be arbitrarily "large". + If srcSize is known to be small, ZSTD_estimateCStreamSize_advanced() can provide a tighter estimation. + ZSTD_estimateCStreamSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel. + Note : if streaming is init with function ZSTD_init?Stream_usingDict(), an internal ?Dict will be created, which size is not estimated here. In this case, get total size by adding ZSTD_estimate?DictSize
size_t ZSTD_estimateCDictSize(ZSTD_compressionParameters cParams, size_t dictSize, unsigned byReference); +size_t ZSTD_estimateCDictSize(size_t dictSize, int compressionLevel); +size_t ZSTD_estimateCDictSize_advanced(size_t dictSize, ZSTD_compressionParameters cParams, unsigned byReference); size_t ZSTD_estimateDDictSize(size_t dictSize, unsigned byReference); -Note : dictionary created "byReference" are smaller +
ZSTD_estimateCDictSize() will bet that src size is relatively "small", and content is copied, like ZSTD_createCDict(). + ZSTD_estimateCStreamSize_advanced() makes it possible to control precisely compression parameters, like ZSTD_createCDict_advanced(). + Note : dictionary created "byReference" are smaller
typedef struct {
unsigned long long frameContentSize;
- unsigned windowSize;
+ size_t windowSize;
unsigned dictID;
unsigned checksumFlag;
} ZSTD_frameHeader;
@@ -421,12 +421,17 @@ size_t ZSTD_estimateDCtxSize(void);
size_t ZSTD_estimateCStreamSize(int compressionLevel);
size_t ZSTD_estimateCStreamSize_advanced(ZSTD_compressionParameters cParams);
-size_t ZSTD_estimateDStreamSize(ZSTD_frameHeader fHeader);
+size_t ZSTD_estimateDStreamSize(size_t windowSize);
+size_t ZSTD_estimateDStreamSize_fromFrame(const void* src, size_t srcSize);
ZSTD_estimateCStreamSize() will consider src size to be arbitrarily "large".
If srcSize is known to be small, ZSTD_estimateCStreamSize_advanced() can provide a tighter estimation.
ZSTD_estimateCStreamSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel.
+ Note : CStream estimation is only correct for single-threaded compression.
+ ZSTD_DStream memory budget depends on window Size.
+ This information can be passed manually, using ZSTD_estimateDStreamSize,
+ or deducted from a valid frame Header, using ZSTD_estimateDStreamSize_fromFrame();
Note : if streaming is init with function ZSTD_init?Stream_usingDict(),
- an internal ?Dict will be created, which size is not estimated here.
+ an internal ?Dict will be created, which additional size is not estimated here.
In this case, get total size by adding ZSTD_estimate?DictSize
@@ -622,6 +627,7 @@ size_t ZSTD_estimateDDictSize(size_t dictSize, unsigned byReference);
Advanced streaming functions
Advanced Streaming compression functions
ZSTD_CStream* ZSTD_createCStream_advanced(ZSTD_customMem customMem);
+ZSTD_CStream* ZSTD_initStaticCStream(void* workspace, size_t workspaceSize);
/**< same as ZSTD_initStaticCCtx() */
size_t ZSTD_initCStream_srcSize(ZSTD_CStream* zcs, int compressionLevel, unsigned long long pledgedSrcSize); /**< pledgedSrcSize must be correct, a size of 0 means unknown. for a frame size of 0 use initCStream_advanced */
size_t ZSTD_initCStream_usingDict(ZSTD_CStream* zcs, const void* dict, size_t dictSize, int compressionLevel); /**< creates of an internal CDict (incompatible with static CCtx), except if dict == NULL or dictSize < 8, in which case no dict is used. */
size_t ZSTD_initCStream_advanced(ZSTD_CStream* zcs, const void* dict, size_t dictSize,
@@ -640,6 +646,7 @@ size_t ZSTD_initCStream_usingCDict_advanced(ZSTD_CStream* zcs, const ZSTD_CDict*
Advanced Streaming decompression functions
typedef enum { DStream_p_maxWindowSize } ZSTD_DStreamParameter_e;
ZSTD_DStream* ZSTD_createDStream_advanced(ZSTD_customMem customMem);
+ZSTD_DStream* ZSTD_initStaticDStream(void* workspace, size_t workspaceSize); /**< same as ZSTD_initStaticDCtx() */
size_t ZSTD_setDStreamParameter(ZSTD_DStream* zds, ZSTD_DStreamParameter_e paramType, unsigned paramValue);
size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize); /**< note: a dict will not be used if dict == NULL or dictSize < 8 */
size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* zds, const ZSTD_DDict* ddict); /**< note : ddict will just be referenced, and must outlive decompression session */
@@ -694,16 +701,18 @@ size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* preparedCCtx, unsigned lo
A ZSTD_DCtx object can be re-used multiple times.
First typical operation is to retrieve frame parameters, using ZSTD_getFrameHeader().
- It fills a ZSTD_frameParams structure which provide important information to correctly decode the frame,
- such as the minimum rolling buffer size to allocate to decompress data (`windowSize`),
- and the dictionary ID used.
+ It fills a ZSTD_frameHeader structure with important information to correctly decode the frame,
+ such as minimum rolling buffer size to allocate to decompress data (`windowSize`),
+ and the dictionary ID in use.
(Note : content size is optional, it may not be present. 0 means : content size unknown).
Note that these values could be wrong, either because of data malformation, or because an attacker is spoofing deliberate false information.
As a consequence, check that values remain within valid application range, especially `windowSize`, before allocation.
- Each application can set its own limit, depending on local restrictions. For extended interoperability, it is recommended to support at least 8 MB.
- Frame parameters are extracted from the beginning of the compressed frame.
- Data fragment must be large enough to ensure successful decoding, typically `ZSTD_frameHeaderSize_max` bytes.
- @result : 0 : successful decoding, the `ZSTD_frameParams` structure is correctly filled.
+ Each application can set its own limit, depending on local restrictions.
+ For extended interoperability, it is recommended to support windowSize of at least 8 MB.
+ Frame header is extracted from the beginning of compressed frame, so providing only the frame's beginning is enough.
+ Data fragment must be large enough to ensure successful decoding.
+ `ZSTD_frameHeaderSize_max` bytes is guaranteed to always be large enough.
+ @result : 0 : successful decoding, the `ZSTD_frameHeader` structure is correctly filled.
>0 : `srcSize` is too small, please provide at least @result bytes on next attempt.
errorCode, which can be tested using ZSTD_isError().
@@ -747,7 +756,7 @@ size_t ZSTD_copyCCtx(ZSTD_CCtx* cctx, const ZSTD_CCtx* preparedCCtx, unsigned lo
It also returns Frame Size as fparamsPtr->frameContentSize.
-size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize);/**< doesn't consume input, see details below */ +
size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize);/**< doesn't consume input */ size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx); size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx* dctx, const void* dict, size_t dictSize); size_t ZSTD_decompressBegin_usingDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict); diff --git a/lib/compress/zstd_compress.c b/lib/compress/zstd_compress.c index 781cdcd42..ea37197ff 100644 --- a/lib/compress/zstd_compress.c +++ b/lib/compress/zstd_compress.c @@ -3594,6 +3594,11 @@ ZSTD_CStream* ZSTD_createCStream(void) 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); diff --git a/lib/decompress/zstd_decompress.c b/lib/decompress/zstd_decompress.c index 95d18d4b5..940c3e033 100644 --- a/lib/decompress/zstd_decompress.c +++ b/lib/decompress/zstd_decompress.c @@ -2143,6 +2143,11 @@ ZSTD_DStream* ZSTD_createDStream(void) return ZSTD_createDStream_advanced(ZSTD_defaultCMem); } +ZSTD_DStream* ZSTD_initStaticDStream(void *workspace, size_t workspaceSize) +{ + return ZSTD_initStaticDCtx(workspace, workspaceSize); +} + ZSTD_DStream* ZSTD_createDStream_advanced(ZSTD_customMem customMem) { return ZSTD_createDCtx_advanced(customMem); @@ -2214,15 +2219,23 @@ size_t ZSTD_sizeof_DStream(const ZSTD_DStream* zds) return ZSTD_sizeof_DCtx(zds); } -size_t ZSTD_estimateDStreamSize(ZSTD_frameHeader fHeader) +size_t ZSTD_estimateDStreamSize(size_t windowSize) { - size_t const windowSize = fHeader.windowSize; size_t const blockSize = MIN(windowSize, ZSTD_BLOCKSIZE_MAX); size_t const inBuffSize = blockSize; /* no block can be larger */ size_t const outBuffSize = windowSize + blockSize + (WILDCOPY_OVERLENGTH * 2); return sizeof(ZSTD_DStream) + ZSTD_estimateDCtxSize() + inBuffSize + outBuffSize; } +ZSTDLIB_API size_t ZSTD_estimateDStreamSize_fromFrame(const void* src, size_t srcSize) +{ + ZSTD_frameHeader fh; + size_t const err = ZSTD_getFrameHeader(&fh, src, srcSize); + if (ZSTD_isError(err)) return err; + if (err>0) return ERROR(srcSize_wrong); + return ZSTD_estimateDStreamSize(fh.windowSize); +} + /* ***** Decompression ***** */ diff --git a/lib/zstd.h b/lib/zstd.h index 492c3d43c..08358f0df 100644 --- a/lib/zstd.h +++ b/lib/zstd.h @@ -412,7 +412,7 @@ typedef struct { typedef struct { unsigned long long frameContentSize; - unsigned windowSize; + size_t windowSize; unsigned dictID; unsigned checksumFlag; } ZSTD_frameHeader; @@ -509,12 +509,17 @@ ZSTDLIB_API size_t ZSTD_estimateDCtxSize(void); * ZSTD_estimateCStreamSize() will consider src size to be arbitrarily "large". * If srcSize is known to be small, ZSTD_estimateCStreamSize_advanced() can provide a tighter estimation. * ZSTD_estimateCStreamSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel. + * Note : CStream estimation is only correct for single-threaded compression. + * ZSTD_DStream memory budget depends on window Size. + * This information can be passed manually, using ZSTD_estimateDStreamSize, + * or deducted from a valid frame Header, using ZSTD_estimateDStreamSize_fromFrame(); * Note : if streaming is init with function ZSTD_init?Stream_usingDict(), - * an internal ?Dict will be created, which size is not estimated here. + * an internal ?Dict will be created, which additional size is not estimated here. * In this case, get total size by adding ZSTD_estimate?DictSize */ ZSTDLIB_API size_t ZSTD_estimateCStreamSize(int compressionLevel); ZSTDLIB_API size_t ZSTD_estimateCStreamSize_advanced(ZSTD_compressionParameters cParams); -ZSTDLIB_API size_t ZSTD_estimateDStreamSize(ZSTD_frameHeader fHeader); +ZSTDLIB_API size_t ZSTD_estimateDStreamSize(size_t windowSize); +ZSTDLIB_API size_t ZSTD_estimateDStreamSize_fromFrame(const void* src, size_t srcSize); /*! ZSTD_estimate?DictSize() : * ZSTD_estimateCDictSize() will bet that src size is relatively "small", and content is copied, like ZSTD_createCDict(). @@ -717,6 +722,7 @@ ZSTDLIB_API unsigned ZSTD_getDictID_fromFrame(const void* src, size_t srcSize); /*===== Advanced Streaming compression functions =====*/ ZSTDLIB_API ZSTD_CStream* ZSTD_createCStream_advanced(ZSTD_customMem customMem); +ZSTDLIB_API ZSTD_CStream* ZSTD_initStaticCStream(void* workspace, size_t workspaceSize); /**< same as ZSTD_initStaticCCtx() */ ZSTDLIB_API size_t ZSTD_initCStream_srcSize(ZSTD_CStream* zcs, int compressionLevel, unsigned long long pledgedSrcSize); /**< pledgedSrcSize must be correct, a size of 0 means unknown. for a frame size of 0 use initCStream_advanced */ ZSTDLIB_API size_t ZSTD_initCStream_usingDict(ZSTD_CStream* zcs, const void* dict, size_t dictSize, int compressionLevel); /**< creates of an internal CDict (incompatible with static CCtx), except if dict == NULL or dictSize < 8, in which case no dict is used. */ ZSTDLIB_API size_t ZSTD_initCStream_advanced(ZSTD_CStream* zcs, const void* dict, size_t dictSize, @@ -737,6 +743,7 @@ ZSTDLIB_API size_t ZSTD_resetCStream(ZSTD_CStream* zcs, unsigned long long pledg /*===== Advanced Streaming decompression functions =====*/ typedef enum { DStream_p_maxWindowSize } ZSTD_DStreamParameter_e; ZSTDLIB_API ZSTD_DStream* ZSTD_createDStream_advanced(ZSTD_customMem customMem); +ZSTDLIB_API ZSTD_DStream* ZSTD_initStaticDStream(void* workspace, size_t workspaceSize); /**< same as ZSTD_initStaticDCtx() */ ZSTDLIB_API size_t ZSTD_setDStreamParameter(ZSTD_DStream* zds, ZSTD_DStreamParameter_e paramType, unsigned paramValue); ZSTDLIB_API size_t ZSTD_initDStream_usingDict(ZSTD_DStream* zds, const void* dict, size_t dictSize); /**< note: a dict will not be used if dict == NULL or dictSize < 8 */ ZSTDLIB_API size_t ZSTD_initDStream_usingDDict(ZSTD_DStream* zds, const ZSTD_DDict* ddict); /**< note : ddict will just be referenced, and must outlive decompression session */ @@ -803,16 +810,18 @@ ZSTDLIB_API size_t ZSTD_compressEnd(ZSTD_CCtx* cctx, void* dst, size_t dstCapaci A ZSTD_DCtx object can be re-used multiple times. First typical operation is to retrieve frame parameters, using ZSTD_getFrameHeader(). - It fills a ZSTD_frameParams structure which provide important information to correctly decode the frame, - such as the minimum rolling buffer size to allocate to decompress data (`windowSize`), - and the dictionary ID used. + It fills a ZSTD_frameHeader structure with important information to correctly decode the frame, + such as minimum rolling buffer size to allocate to decompress data (`windowSize`), + and the dictionary ID in use. (Note : content size is optional, it may not be present. 0 means : content size unknown). Note that these values could be wrong, either because of data malformation, or because an attacker is spoofing deliberate false information. As a consequence, check that values remain within valid application range, especially `windowSize`, before allocation. - Each application can set its own limit, depending on local restrictions. For extended interoperability, it is recommended to support at least 8 MB. - Frame parameters are extracted from the beginning of the compressed frame. - Data fragment must be large enough to ensure successful decoding, typically `ZSTD_frameHeaderSize_max` bytes. - @result : 0 : successful decoding, the `ZSTD_frameParams` structure is correctly filled. + Each application can set its own limit, depending on local restrictions. + For extended interoperability, it is recommended to support windowSize of at least 8 MB. + Frame header is extracted from the beginning of compressed frame, so providing only the frame's beginning is enough. + Data fragment must be large enough to ensure successful decoding. + `ZSTD_frameHeaderSize_max` bytes is guaranteed to always be large enough. + @result : 0 : successful decoding, the `ZSTD_frameHeader` structure is correctly filled. >0 : `srcSize` is too small, please provide at least @result bytes on next attempt. errorCode, which can be tested using ZSTD_isError(). @@ -857,7 +866,7 @@ ZSTDLIB_API size_t ZSTD_compressEnd(ZSTD_CCtx* cctx, void* dst, size_t dstCapaci */ /*===== Buffer-less streaming decompression functions =====*/ -ZSTDLIB_API size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize); /**< doesn't consume input, see details below */ +ZSTDLIB_API size_t ZSTD_getFrameHeader(ZSTD_frameHeader* zfhPtr, const void* src, size_t srcSize); /**< doesn't consume input */ ZSTDLIB_API size_t ZSTD_decompressBegin(ZSTD_DCtx* dctx); ZSTDLIB_API size_t ZSTD_decompressBegin_usingDict(ZSTD_DCtx* dctx, const void* dict, size_t dictSize); ZSTDLIB_API size_t ZSTD_decompressBegin_usingDDict(ZSTD_DCtx* dctx, const ZSTD_DDict* ddict); diff --git a/tests/zstreamtest.c b/tests/zstreamtest.c index 2c9cb5204..aae83972a 100644 --- a/tests/zstreamtest.c +++ b/tests/zstreamtest.c @@ -285,8 +285,8 @@ static int basicUnitTests(U32 seed, double compressibility, ZSTD_customMem custo const void* cStart = (char*)compressedBuffer + (skippableFrameSize + 8); size_t const gfhError = ZSTD_getFrameHeader(&fhi, cStart, cSize); if (gfhError!=0) goto _output_error; - DISPLAYLEVEL(5, " (windowSize : %u) ", fhi.windowSize); - { size_t const s = ZSTD_estimateDStreamSize(fhi) + DISPLAYLEVEL(5, " (windowSize : %u) ", (U32)fhi.windowSize); + { size_t const s = ZSTD_estimateDStreamSize(fhi.windowSize) /* uses ZSTD_initDStream_usingDict() */ + ZSTD_estimateDDictSize(dictSize, 0); if (ZSTD_isError(s)) goto _output_error; From c7fb884eeaa9f03fa25caa1da7ab8f19c203aa6f Mon Sep 17 00:00:00 2001 From: Yann Collet
These functions make it possible to estimate memory usage of a future {D,C}Ctx, before its creation. - The objective is to guide decision before allocation. - ZSTD_estimateCCtxSize() will consider src size to be arbitrarily "large". - If srcSize is known to be small, ZSTD_estimateCCtxSize_advanced() can provide a tighter estimation. + ZSTD_estimateCCtxSize() will provide a budget large for any compression level up to selected one. + It will also consider src size to be arbitrarily "large", which is worst case. + If srcSize is known to always be small, ZSTD_estimateCCtxSize_advanced() can provide a tighter estimation. ZSTD_estimateCCtxSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel. Note : CCtx estimation is only correct for single-threaded compression
ZSTD_estimateCStreamSize() will consider src size to be arbitrarily "large". - If srcSize is known to be small, ZSTD_estimateCStreamSize_advanced() can provide a tighter estimation. +
ZSTD_estimateCStreamSize() will provide a budget large for any compression level up to selected one.
+ It will also consider src size to be arbitrarily "large", which is worst case.
+ If srcSize is known to always be small, ZSTD_estimateCStreamSize_advanced() can provide a tighter estimation.
ZSTD_estimateCStreamSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel.
Note : CStream estimation is only correct for single-threaded compression.
ZSTD_DStream memory budget depends on window Size.
diff --git a/lib/compress/zstd_compress.c b/lib/compress/zstd_compress.c
index c781232c6..c2adc56ab 100644
--- a/lib/compress/zstd_compress.c
+++ b/lib/compress/zstd_compress.c
@@ -4098,6 +4098,43 @@ static const ZSTD_compressionParameters ZSTD_defaultCParameters[4][ZSTD_MAX_CLEV
},
};
+/* This function just controls
+ * the monotonic memory budget increase of ZSTD_defaultCParameters[0].
+ * Run only once, on first ZSTD_getCParams() usage, when ZSTD_DEBUG is enabled
+ */
+MEM_STATIC void ZSTD_check_compressionLevel_monotonicIncrease_memoryBudget(void)
+{
+# define ZSTD_TABLECOST(h,c) ((1<<(h)) + (1<<(c)))
+# define ZDCP_FIELD(l,field) (ZSTD_defaultCParameters[0][l].field)
+# define ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(l) { \
+ assert(ZDCP_FIELD(l,windowLog) <= ZDCP_FIELD(l+1,windowLog) ); \
+ assert(ZSTD_TABLECOST(ZDCP_FIELD(l,hashLog), ZDCP_FIELD(l,chainLog)) <= ZSTD_TABLECOST(ZDCP_FIELD(l+1,hashLog), ZDCP_FIELD(l+1,chainLog)) ); \
+ }
+
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(1);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(2);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(3);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(4);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(5);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(6);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(7);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(8);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(9);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(10);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(11);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(12);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(13);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(14);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(15);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(16);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(17);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(18);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(19);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(20);
+ ZSTD_CHECK_MONOTONIC_INCREASE_LEVEL(21);
+ assert(ZSTD_maxCLevel()==22);
+}
+
/*! 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 */
@@ -4108,6 +4145,15 @@ ZSTD_compressionParameters ZSTD_getCParams(int compressionLevel, unsigned long l
U32 const tableID = (rSize <= 256 KB) + (rSize <= 128 KB) + (rSize <= 16 KB); /* intentional underflow for srcSizeHint == 0 */
if (compressionLevel <= 0) compressionLevel = ZSTD_CLEVEL_DEFAULT; /* 0 == default; no negative compressionLevel yet */
if (compressionLevel > ZSTD_MAX_CLEVEL) compressionLevel = ZSTD_MAX_CLEVEL;
+
+#if defined(ZSTD_DEBUG) && (ZSTD_DEBUG>=1)
+ static int g_monotonicTest = 1;
+ if (g_monotonicTest) {
+ ZSTD_check_compressionLevel_monotonicIncrease_memoryBudget();
+ g_monotonicTest=0;
+ }
+#endif
+
{ ZSTD_compressionParameters const cp = ZSTD_defaultCParameters[tableID][compressionLevel];
return ZSTD_adjustCParams_internal(cp, srcSizeHint, dictSize); }
}
diff --git a/lib/zstd.h b/lib/zstd.h
index 08358f0df..75bdb4895 100644
--- a/lib/zstd.h
+++ b/lib/zstd.h
@@ -496,9 +496,9 @@ ZSTDLIB_API size_t ZSTD_sizeof_DDict(const ZSTD_DDict* ddict);
/*! ZSTD_estimate*() :
* These functions make it possible to estimate memory usage
* of a future {D,C}Ctx, before its creation.
- * The objective is to guide decision before allocation.
- * ZSTD_estimateCCtxSize() will consider src size to be arbitrarily "large".
- * If srcSize is known to be small, ZSTD_estimateCCtxSize_advanced() can provide a tighter estimation.
+ * ZSTD_estimateCCtxSize() will provide a budget large enough for any compression level up to selected one.
+ * It will also consider src size to be arbitrarily "large", which is worst case.
+ * If srcSize is known to always be small, ZSTD_estimateCCtxSize_advanced() can provide a tighter estimation.
* ZSTD_estimateCCtxSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel.
* Note : CCtx estimation is only correct for single-threaded compression */
ZSTDLIB_API size_t ZSTD_estimateCCtxSize(int compressionLevel);
@@ -506,8 +506,9 @@ ZSTDLIB_API size_t ZSTD_estimateCCtxSize_advanced(ZSTD_compressionParameters cPa
ZSTDLIB_API size_t ZSTD_estimateDCtxSize(void);
/*! ZSTD_estimate?StreamSize() :
- * ZSTD_estimateCStreamSize() will consider src size to be arbitrarily "large".
- * If srcSize is known to be small, ZSTD_estimateCStreamSize_advanced() can provide a tighter estimation.
+ * ZSTD_estimateCStreamSize() will provide a budget large enough for any compression level up to selected one.
+ * It will also consider src size to be arbitrarily "large", which is worst case.
+ * If srcSize is known to always be small, ZSTD_estimateCStreamSize_advanced() can provide a tighter estimation.
* ZSTD_estimateCStreamSize_advanced() can be used in tandem with ZSTD_getCParams() to create cParams from compressionLevel.
* Note : CStream estimation is only correct for single-threaded compression.
* ZSTD_DStream memory budget depends on window Size.
From 1ca76039af730c11fc761961a9ccb7de99d4c057 Mon Sep 17 00:00:00 2001
From: Yann Collet