diff --git a/.travis.yml b/.travis.yml index 4105c1f09..e9f918481 100644 --- a/.travis.yml +++ b/.travis.yml @@ -63,6 +63,15 @@ matrix: - make clang38install - CC=clang-3.8 make clean msan-test-zstd + - name: Trusty (Minimal Decompressor Macros) + script: + - make clean + - make check -j all MOREFLAGS="-Werror -DHUF_FORCE_DECOMPRESS_X1 -DZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT" + - make clean + - make check -j all MOREFLAGS="-Werror -DHUF_FORCE_DECOMPRESS_X2 -DZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG" + - make clean + - make check -j all MOREFLAGS="-Werror -DZSTD_NO_INLINE -DZSTD_STRIP_ERROR_STRINGS" + - name: Trusty (CMake) script: - make cmakebuild diff --git a/lib/Makefile b/lib/Makefile index 3fddf4fcd..386583e72 100644 --- a/lib/Makefile +++ b/lib/Makefile @@ -50,6 +50,12 @@ ZSTD_LIB_COMPRESSION ?= 1 ZSTD_LIB_DECOMPRESSION ?= 1 ZSTD_LIB_DICTBUILDER ?= 1 ZSTD_LIB_DEPRECATED ?= 1 +HUF_FORCE_DECOMPRESS_X1 ?= 0 +HUF_FORCE_DECOMPRESS_X2 ?= 0 +ZSTD_FORCE_DECOMPRESS_SHORT ?= 0 +ZSTD_FORCE_DECOMPRESS_LONG ?= 0 +ZSTD_NO_INLINE ?= 0 +ZSTD_STRIP_ERROR_STRINGS ?= 0 ifeq ($(ZSTD_LIB_COMPRESSION), 0) ZSTD_LIB_DICTBUILDER = 0 @@ -77,6 +83,30 @@ ifneq ($(ZSTD_LIB_DICTBUILDER), 0) ZSTD_FILES += $(ZDICT_FILES) endif +ifneq ($(HUF_FORCE_DECOMPRESS_X1), 0) + CFLAGS += -DHUF_FORCE_DECOMPRESS_X1 +endif + +ifneq ($(HUF_FORCE_DECOMPRESS_X2), 0) + CFLAGS += -DHUF_FORCE_DECOMPRESS_X2 +endif + +ifneq ($(ZSTD_FORCE_DECOMPRESS_SHORT), 0) + CFLAGS += -DZSTD_FORCE_DECOMPRESS_SHORT +endif + +ifneq ($(ZSTD_FORCE_DECOMPRESS_LONG), 0) + CFLAGS += -DZSTD_FORCE_DECOMPRESS_LONG +endif + +ifneq ($(ZSTD_NO_INLINE), 0) + CFLAGS += -DZSTD_NO_INLINE +endif + +ifneq ($(ZSTD_STRIP_ERROR_STRINGS), 0) + CFLAGS += -DZSTD_STRIP_ERROR_STRINGS +endif + ifneq ($(ZSTD_LEGACY_SUPPORT), 0) ifeq ($(shell test $(ZSTD_LEGACY_SUPPORT) -lt 8; echo $$?), 0) ZSTD_FILES += $(shell ls legacy/*.c | $(GREP) 'v0[$(ZSTD_LEGACY_SUPPORT)-7]') diff --git a/lib/README.md b/lib/README.md index 0966c7aef..45fd62551 100644 --- a/lib/README.md +++ b/lib/README.md @@ -66,6 +66,24 @@ It's possible to compile only a limited set of features. and `ZSTD_LIB_DEPRECATED` as 0 to forgo compilation of the corresponding features. This will also disable compilation of all dependencies (eg. `ZSTD_LIB_COMPRESSION=0` will also disable dictBuilder). +- There are some additional macros that can be used to minify the decoder. + + Zstandard often has more than one implementation of a piece of functionality, + where each implementation optimizes for different scenarios. For example, the + Huffman decoder has complementary implementations that decode the stream one + symbol at a time or two symbols at a time. Zstd normally includes both (and + dispatches between them at runtime), but by defining `HUF_FORCE_DECOMPRESS_X1` + or `HUF_FORCE_DECOMPRESS_X2`, you can force the use of one or the other, avoiding + compilation of the other. Similarly, `ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT` + and `ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG` force the compilation and use of + only one or the other of two decompression implementations. The smallest + binary is achieved by using `HUF_FORCE_DECOMPRESS_X1` and + `ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT`. + + For squeezing the last ounce of size out, you can also define + `ZSTD_NO_INLINE`, which disables inlining, and `ZSTD_STRIP_ERROR_STRINGS`, + which removes the error messages that are otherwise returned by + `ZSTD_getErrorName`. #### Multithreading support diff --git a/lib/common/compiler.h b/lib/common/compiler.h index e6267e90b..7f561282c 100644 --- a/lib/common/compiler.h +++ b/lib/common/compiler.h @@ -15,6 +15,8 @@ * Compiler specifics *********************************************************/ /* force inlining */ + +#if !defined(ZSTD_NO_INLINE) #if defined (__GNUC__) || defined(__cplusplus) || defined(__STDC_VERSION__) && __STDC_VERSION__ >= 199901L /* C99 */ # define INLINE_KEYWORD inline #else @@ -29,6 +31,13 @@ # define FORCE_INLINE_ATTR #endif +#else + +#define INLINE_KEYWORD +#define FORCE_INLINE_ATTR + +#endif + /** * FORCE_INLINE_TEMPLATE is used to define C "templates", which take constant * parameters. They must be inlined for the compiler to elimininate the constant diff --git a/lib/common/error_private.c b/lib/common/error_private.c index d39e1a7ad..7c1bb67a2 100644 --- a/lib/common/error_private.c +++ b/lib/common/error_private.c @@ -14,6 +14,10 @@ const char* ERR_getErrorString(ERR_enum code) { +#ifdef ZSTD_STRIP_ERROR_STRINGS + (void)code; + return "Error strings stripped"; +#else static const char* const notErrorCode = "Unspecified error code"; switch( code ) { @@ -46,4 +50,5 @@ const char* ERR_getErrorString(ERR_enum code) case PREFIX(maxCode): default: return notErrorCode; } +#endif } diff --git a/lib/common/huf.h b/lib/common/huf.h index de9464111..34a16f117 100644 --- a/lib/common/huf.h +++ b/lib/common/huf.h @@ -173,15 +173,19 @@ typedef U32 HUF_DTable; * Advanced decompression functions ******************************************/ size_t HUF_decompress4X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< single-symbol decoder */ +#ifndef HUF_FORCE_DECOMPRESS_X1 size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< double-symbols decoder */ +#endif size_t HUF_decompress4X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< decodes RLE and uncompressed */ size_t HUF_decompress4X_hufOnly(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< considers RLE and uncompressed as errors */ size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< considers RLE and uncompressed as errors */ size_t HUF_decompress4X1_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< single-symbol decoder */ size_t HUF_decompress4X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< single-symbol decoder */ +#ifndef HUF_FORCE_DECOMPRESS_X1 size_t HUF_decompress4X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< double-symbols decoder */ size_t HUF_decompress4X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< double-symbols decoder */ +#endif /* **************************************** @@ -277,14 +281,22 @@ U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize); #define HUF_DECOMPRESS_WORKSPACE_SIZE (2 << 10) #define HUF_DECOMPRESS_WORKSPACE_SIZE_U32 (HUF_DECOMPRESS_WORKSPACE_SIZE / sizeof(U32)) +#ifndef HUF_FORCE_DECOMPRESS_X2 size_t HUF_readDTableX1 (HUF_DTable* DTable, const void* src, size_t srcSize); size_t HUF_readDTableX1_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize); +#endif +#ifndef HUF_FORCE_DECOMPRESS_X1 size_t HUF_readDTableX2 (HUF_DTable* DTable, const void* src, size_t srcSize); size_t HUF_readDTableX2_wksp (HUF_DTable* DTable, const void* src, size_t srcSize, void* workSpace, size_t wkspSize); +#endif size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#ifndef HUF_FORCE_DECOMPRESS_X2 size_t HUF_decompress4X1_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#endif +#ifndef HUF_FORCE_DECOMPRESS_X1 size_t HUF_decompress4X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#endif /* ====================== */ @@ -306,24 +318,36 @@ size_t HUF_compress1X_repeat(void* dst, size_t dstSize, HUF_CElt* hufTable, HUF_repeat* repeat, int preferRepeat, int bmi2); size_t HUF_decompress1X1 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /* single-symbol decoder */ +#ifndef HUF_FORCE_DECOMPRESS_X1 size_t HUF_decompress1X2 (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /* double-symbol decoder */ +#endif size_t HUF_decompress1X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); size_t HUF_decompress1X_DCtx_wksp (HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); +#ifndef HUF_FORCE_DECOMPRESS_X2 size_t HUF_decompress1X1_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< single-symbol decoder */ size_t HUF_decompress1X1_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< single-symbol decoder */ +#endif +#ifndef HUF_FORCE_DECOMPRESS_X1 size_t HUF_decompress1X2_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); /**< double-symbols decoder */ size_t HUF_decompress1X2_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize); /**< double-symbols decoder */ +#endif size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); /**< automatic selection of sing or double symbol decoder, based on DTable */ +#ifndef HUF_FORCE_DECOMPRESS_X2 size_t HUF_decompress1X1_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#endif +#ifndef HUF_FORCE_DECOMPRESS_X1 size_t HUF_decompress1X2_usingDTable(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable); +#endif /* BMI2 variants. * If the CPU has BMI2 support, pass bmi2=1, otherwise pass bmi2=0. */ size_t HUF_decompress1X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2); +#ifndef HUF_FORCE_DECOMPRESS_X2 size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2); +#endif size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2); size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2); diff --git a/lib/decompress/huf_decompress.c b/lib/decompress/huf_decompress.c index 83ecaff01..3f8bd2973 100644 --- a/lib/decompress/huf_decompress.c +++ b/lib/decompress/huf_decompress.c @@ -43,6 +43,19 @@ #include "huf.h" #include "error_private.h" +/* ************************************************************** +* Macros +****************************************************************/ + +/* These two optional macros force the use one way or another of the two + * Huffman decompression implementations. You can't force in both directions + * at the same time. + */ +#if defined(HUF_FORCE_DECOMPRESS_X1) && \ + defined(HUF_FORCE_DECOMPRESS_X2) +#error "Cannot force the use of the X1 and X2 decoders at the same time!" +#endif + /* ************************************************************** * Error Management @@ -58,6 +71,51 @@ #define HUF_ALIGN_MASK(x, mask) (((x) + (mask)) & ~(mask)) +/* ************************************************************** +* BMI2 Variant Wrappers +****************************************************************/ +#if DYNAMIC_BMI2 + +#define HUF_DGEN(fn) \ + \ + static size_t fn##_default( \ + void* dst, size_t dstSize, \ + const void* cSrc, size_t cSrcSize, \ + const HUF_DTable* DTable) \ + { \ + return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \ + } \ + \ + static TARGET_ATTRIBUTE("bmi2") size_t fn##_bmi2( \ + void* dst, size_t dstSize, \ + const void* cSrc, size_t cSrcSize, \ + const HUF_DTable* DTable) \ + { \ + return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \ + } \ + \ + static size_t fn(void* dst, size_t dstSize, void const* cSrc, \ + size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \ + { \ + if (bmi2) { \ + return fn##_bmi2(dst, dstSize, cSrc, cSrcSize, DTable); \ + } \ + return fn##_default(dst, dstSize, cSrc, cSrcSize, DTable); \ + } + +#else + +#define HUF_DGEN(fn) \ + static size_t fn(void* dst, size_t dstSize, void const* cSrc, \ + size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \ + { \ + (void)bmi2; \ + return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \ + } + +#endif + + /*-***************************/ /* generic DTableDesc */ /*-***************************/ @@ -71,6 +129,8 @@ static DTableDesc HUF_getDTableDesc(const HUF_DTable* table) } +#ifndef HUF_FORCE_DECOMPRESS_X2 + /*-***************************/ /* single-symbol decoding */ /*-***************************/ @@ -307,46 +367,6 @@ typedef size_t (*HUF_decompress_usingDTable_t)(void *dst, size_t dstSize, const void *cSrc, size_t cSrcSize, const HUF_DTable *DTable); -#if DYNAMIC_BMI2 - -#define HUF_DGEN(fn) \ - \ - static size_t fn##_default( \ - void* dst, size_t dstSize, \ - const void* cSrc, size_t cSrcSize, \ - const HUF_DTable* DTable) \ - { \ - return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \ - } \ - \ - static TARGET_ATTRIBUTE("bmi2") size_t fn##_bmi2( \ - void* dst, size_t dstSize, \ - const void* cSrc, size_t cSrcSize, \ - const HUF_DTable* DTable) \ - { \ - return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \ - } \ - \ - static size_t fn(void* dst, size_t dstSize, void const* cSrc, \ - size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \ - { \ - if (bmi2) { \ - return fn##_bmi2(dst, dstSize, cSrc, cSrcSize, DTable); \ - } \ - return fn##_default(dst, dstSize, cSrc, cSrcSize, DTable); \ - } - -#else - -#define HUF_DGEN(fn) \ - static size_t fn(void* dst, size_t dstSize, void const* cSrc, \ - size_t cSrcSize, HUF_DTable const* DTable, int bmi2) \ - { \ - (void)bmi2; \ - return fn##_body(dst, dstSize, cSrc, cSrcSize, DTable); \ - } - -#endif HUF_DGEN(HUF_decompress1X1_usingDTable_internal) HUF_DGEN(HUF_decompress4X1_usingDTable_internal) @@ -437,6 +457,10 @@ size_t HUF_decompress4X1 (void* dst, size_t dstSize, const void* cSrc, size_t cS return HUF_decompress4X1_DCtx(DTable, dst, dstSize, cSrc, cSrcSize); } +#endif /* HUF_FORCE_DECOMPRESS_X2 */ + + +#ifndef HUF_FORCE_DECOMPRESS_X1 /* *************************/ /* double-symbols decoding */ @@ -911,6 +935,8 @@ size_t HUF_decompress4X2 (void* dst, size_t dstSize, const void* cSrc, size_t cS return HUF_decompress4X2_DCtx(DTable, dst, dstSize, cSrc, cSrcSize); } +#endif /* HUF_FORCE_DECOMPRESS_X1 */ + /* ***********************************/ /* Universal decompression selectors */ @@ -921,8 +947,18 @@ size_t HUF_decompress1X_usingDTable(void* dst, size_t maxDstSize, const HUF_DTable* DTable) { DTableDesc const dtd = HUF_getDTableDesc(DTable); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dtd; + assert(dtd.tableType == 0); + return HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dtd; + assert(dtd.tableType == 1); + return HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#else return dtd.tableType ? HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0) : HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#endif } size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, @@ -930,11 +966,22 @@ size_t HUF_decompress4X_usingDTable(void* dst, size_t maxDstSize, const HUF_DTable* DTable) { DTableDesc const dtd = HUF_getDTableDesc(DTable); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dtd; + assert(dtd.tableType == 0); + return HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dtd; + assert(dtd.tableType == 1); + return HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#else return dtd.tableType ? HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0) : HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, /* bmi2 */ 0); +#endif } +#if !defined(HUF_FORCE_DECOMPRESS_X1) && !defined(HUF_FORCE_DECOMPRESS_X2) typedef struct { U32 tableTime; U32 decode256Time; } algo_time_t; static const algo_time_t algoTime[16 /* Quantization */][3 /* single, double, quad */] = { @@ -956,6 +1003,7 @@ static const algo_time_t algoTime[16 /* Quantization */][3 /* single, double, qu {{1455,128}, {2422,124}, {4174,124}}, /* Q ==14 : 87-93% */ {{ 722,128}, {1891,145}, {1936,146}}, /* Q ==15 : 93-99% */ }; +#endif /** HUF_selectDecoder() : * Tells which decoder is likely to decode faster, @@ -966,6 +1014,15 @@ U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize) { assert(dstSize > 0); assert(dstSize <= 128*1024); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dstSize; + (void)cSrcSize; + return 0; +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dstSize; + (void)cSrcSize; + return 1; +#else /* decoder timing evaluation */ { U32 const Q = (cSrcSize >= dstSize) ? 15 : (U32)(cSrcSize * 16 / dstSize); /* Q < 16 */ U32 const D256 = (U32)(dstSize >> 8); @@ -973,14 +1030,18 @@ U32 HUF_selectDecoder (size_t dstSize, size_t cSrcSize) U32 DTime1 = algoTime[Q][1].tableTime + (algoTime[Q][1].decode256Time * D256); DTime1 += DTime1 >> 3; /* advantage to algorithm using less memory, to reduce cache eviction */ return DTime1 < DTime0; -} } + } +#endif +} typedef size_t (*decompressionAlgo)(void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize); size_t HUF_decompress (void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize) { +#if !defined(HUF_FORCE_DECOMPRESS_X1) && !defined(HUF_FORCE_DECOMPRESS_X2) static const decompressionAlgo decompress[2] = { HUF_decompress4X1, HUF_decompress4X2 }; +#endif /* validation checks */ if (dstSize == 0) return ERROR(dstSize_tooSmall); @@ -989,7 +1050,17 @@ size_t HUF_decompress (void* dst, size_t dstSize, const void* cSrc, size_t cSrcS if (cSrcSize == 1) { memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */ { U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)algoNb; + assert(algoNb == 0); + return HUF_decompress4X1(dst, dstSize, cSrc, cSrcSize); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)algoNb; + assert(algoNb == 1); + return HUF_decompress4X2(dst, dstSize, cSrc, cSrcSize); +#else return decompress[algoNb](dst, dstSize, cSrc, cSrcSize); +#endif } } @@ -1002,8 +1073,18 @@ size_t HUF_decompress4X_DCtx (HUF_DTable* dctx, void* dst, size_t dstSize, const if (cSrcSize == 1) { memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */ { U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)algoNb; + assert(algoNb == 0); + return HUF_decompress4X1_DCtx(dctx, dst, dstSize, cSrc, cSrcSize); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)algoNb; + assert(algoNb == 1); + return HUF_decompress4X2_DCtx(dctx, dst, dstSize, cSrc, cSrcSize); +#else return algoNb ? HUF_decompress4X2_DCtx(dctx, dst, dstSize, cSrc, cSrcSize) : HUF_decompress4X1_DCtx(dctx, dst, dstSize, cSrc, cSrcSize) ; +#endif } } @@ -1025,8 +1106,19 @@ size_t HUF_decompress4X_hufOnly_wksp(HUF_DTable* dctx, void* dst, if (cSrcSize == 0) return ERROR(corruption_detected); { U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); - return algoNb ? HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize): +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)algoNb; + assert(algoNb == 0); + return HUF_decompress4X1_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)algoNb; + assert(algoNb == 1); + return HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize); +#else + return algoNb ? HUF_decompress4X2_DCtx_wksp(dctx, dst, dstSize, cSrc, + cSrcSize, workSpace, wkspSize): HUF_decompress4X1_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize); +#endif } } @@ -1041,10 +1133,22 @@ size_t HUF_decompress1X_DCtx_wksp(HUF_DTable* dctx, void* dst, size_t dstSize, if (cSrcSize == 1) { memset(dst, *(const BYTE*)cSrc, dstSize); return dstSize; } /* RLE */ { U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)algoNb; + assert(algoNb == 0); + return HUF_decompress1X1_DCtx_wksp(dctx, dst, dstSize, cSrc, + cSrcSize, workSpace, wkspSize); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)algoNb; + assert(algoNb == 1); + return HUF_decompress1X2_DCtx_wksp(dctx, dst, dstSize, cSrc, + cSrcSize, workSpace, wkspSize); +#else return algoNb ? HUF_decompress1X2_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize): HUF_decompress1X1_DCtx_wksp(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize); +#endif } } @@ -1060,10 +1164,21 @@ size_t HUF_decompress1X_DCtx(HUF_DTable* dctx, void* dst, size_t dstSize, size_t HUF_decompress1X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2) { DTableDesc const dtd = HUF_getDTableDesc(DTable); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dtd; + assert(dtd.tableType == 0); + return HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dtd; + assert(dtd.tableType == 1); + return HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#else return dtd.tableType ? HUF_decompress1X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2) : HUF_decompress1X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#endif } +#ifndef HUF_FORCE_DECOMPRESS_X2 size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2) { const BYTE* ip = (const BYTE*) cSrc; @@ -1075,12 +1190,23 @@ size_t HUF_decompress1X1_DCtx_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstS return HUF_decompress1X1_usingDTable_internal(dst, dstSize, ip, cSrcSize, dctx, bmi2); } +#endif size_t HUF_decompress4X_usingDTable_bmi2(void* dst, size_t maxDstSize, const void* cSrc, size_t cSrcSize, const HUF_DTable* DTable, int bmi2) { DTableDesc const dtd = HUF_getDTableDesc(DTable); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)dtd; + assert(dtd.tableType == 0); + return HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)dtd; + assert(dtd.tableType == 1); + return HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#else return dtd.tableType ? HUF_decompress4X2_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2) : HUF_decompress4X1_usingDTable_internal(dst, maxDstSize, cSrc, cSrcSize, DTable, bmi2); +#endif } size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t dstSize, const void* cSrc, size_t cSrcSize, void* workSpace, size_t wkspSize, int bmi2) @@ -1090,7 +1216,17 @@ size_t HUF_decompress4X_hufOnly_wksp_bmi2(HUF_DTable* dctx, void* dst, size_t ds if (cSrcSize == 0) return ERROR(corruption_detected); { U32 const algoNb = HUF_selectDecoder(dstSize, cSrcSize); +#if defined(HUF_FORCE_DECOMPRESS_X1) + (void)algoNb; + assert(algoNb == 0); + return HUF_decompress4X1_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2); +#elif defined(HUF_FORCE_DECOMPRESS_X2) + (void)algoNb; + assert(algoNb == 1); + return HUF_decompress4X2_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2); +#else return algoNb ? HUF_decompress4X2_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2) : HUF_decompress4X1_DCtx_wksp_bmi2(dctx, dst, dstSize, cSrc, cSrcSize, workSpace, wkspSize, bmi2); +#endif } } diff --git a/lib/decompress/zstd_decompress.c b/lib/decompress/zstd_decompress.c index 510ce3c65..58cc98a4a 100644 --- a/lib/decompress/zstd_decompress.c +++ b/lib/decompress/zstd_decompress.c @@ -954,9 +954,16 @@ ZSTD_loadDEntropy(ZSTD_entropyDTables_t* entropy, ZSTD_STATIC_ASSERT(sizeof(entropy->LLTable) + sizeof(entropy->OFTable) + sizeof(entropy->MLTable) >= HUF_DECOMPRESS_WORKSPACE_SIZE); { void* const workspace = &entropy->LLTable; /* use fse tables as temporary workspace; implies fse tables are grouped together */ size_t const workspaceSize = sizeof(entropy->LLTable) + sizeof(entropy->OFTable) + sizeof(entropy->MLTable); +#ifdef HUF_FORCE_DECOMPRESS_X1 + /* in minimal huffman, we always use X1 variants */ + size_t const hSize = HUF_readDTableX1_wksp(entropy->hufTable, + dictPtr, dictEnd - dictPtr, + workspace, workspaceSize); +#else size_t const hSize = HUF_readDTableX2_wksp(entropy->hufTable, dictPtr, dictEnd - dictPtr, workspace, workspaceSize); +#endif if (HUF_isError(hSize)) return ERROR(dictionary_corrupted); dictPtr += hSize; } diff --git a/lib/decompress/zstd_decompress_block.c b/lib/decompress/zstd_decompress_block.c index 869bdd9aa..a4b6c4c1a 100644 --- a/lib/decompress/zstd_decompress_block.c +++ b/lib/decompress/zstd_decompress_block.c @@ -27,6 +27,18 @@ #include "zstd_ddict.h" /* ZSTD_DDictDictContent */ #include "zstd_decompress_block.h" +/*_******************************************************* +* Macros +**********************************************************/ + +/* These two optional macros force the use one way or another of the two + * ZSTD_decompressSequences implementations. You can't force in both directions + * at the same time. + */ +#if defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) +#error "Cannot force the use of the short and the long ZSTD_decompressSequences variants!" +#endif /*_******************************************************* @@ -83,6 +95,7 @@ size_t ZSTD_decodeLiteralsBlock(ZSTD_DCtx* dctx, U32 singleStream=0; U32 const lhlCode = (istart[0] >> 2) & 3; U32 const lhc = MEM_readLE32(istart); + size_t hufSuccess; switch(lhlCode) { case 0: case 1: default: /* note : default is impossible, since lhlCode into [0..3] */ @@ -113,16 +126,38 @@ size_t ZSTD_decodeLiteralsBlock(ZSTD_DCtx* dctx, PREFETCH_AREA(dctx->HUFptr, sizeof(dctx->entropy.hufTable)); } - if (HUF_isError((litEncType==set_repeat) ? - ( singleStream ? - HUF_decompress1X_usingDTable_bmi2(dctx->litBuffer, litSize, istart+lhSize, litCSize, dctx->HUFptr, dctx->bmi2) : - HUF_decompress4X_usingDTable_bmi2(dctx->litBuffer, litSize, istart+lhSize, litCSize, dctx->HUFptr, dctx->bmi2) ) : - ( singleStream ? - HUF_decompress1X1_DCtx_wksp_bmi2(dctx->entropy.hufTable, dctx->litBuffer, litSize, istart+lhSize, litCSize, - dctx->workspace, sizeof(dctx->workspace), dctx->bmi2) : - HUF_decompress4X_hufOnly_wksp_bmi2(dctx->entropy.hufTable, dctx->litBuffer, litSize, istart+lhSize, litCSize, - dctx->workspace, sizeof(dctx->workspace), dctx->bmi2)))) - return ERROR(corruption_detected); + if (litEncType==set_repeat) { + if (singleStream) { + hufSuccess = HUF_decompress1X_usingDTable_bmi2( + dctx->litBuffer, litSize, istart+lhSize, litCSize, + dctx->HUFptr, dctx->bmi2); + } else { + hufSuccess = HUF_decompress4X_usingDTable_bmi2( + dctx->litBuffer, litSize, istart+lhSize, litCSize, + dctx->HUFptr, dctx->bmi2); + } + } else { + if (singleStream) { +#if defined(HUF_FORCE_DECOMPRESS_X2) + hufSuccess = HUF_decompress1X_DCtx_wksp( + dctx->entropy.hufTable, dctx->litBuffer, litSize, + istart+lhSize, litCSize, dctx->workspace, + sizeof(dctx->workspace)); +#else + hufSuccess = HUF_decompress1X1_DCtx_wksp_bmi2( + dctx->entropy.hufTable, dctx->litBuffer, litSize, + istart+lhSize, litCSize, dctx->workspace, + sizeof(dctx->workspace), dctx->bmi2); +#endif + } else { + hufSuccess = HUF_decompress4X_hufOnly_wksp_bmi2( + dctx->entropy.hufTable, dctx->litBuffer, litSize, + istart+lhSize, litCSize, dctx->workspace, + sizeof(dctx->workspace), dctx->bmi2); + } + } + + if (HUF_isError(hufSuccess)) return ERROR(corruption_detected); dctx->litPtr = dctx->litBuffer; dctx->litSize = litSize; @@ -775,6 +810,7 @@ ZSTD_updateFseState(ZSTD_fseState* DStatePtr, BIT_DStream_t* bitD) typedef enum { ZSTD_lo_isRegularOffset, ZSTD_lo_isLongOffset=1 } ZSTD_longOffset_e; +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG FORCE_INLINE_TEMPLATE seq_t ZSTD_decodeSequence(seqState_t* seqState, const ZSTD_longOffset_e longOffsets) { @@ -914,9 +950,11 @@ ZSTD_decompressSequences_default(ZSTD_DCtx* dctx, { return ZSTD_decompressSequences_body(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset); } +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG */ +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT FORCE_INLINE_TEMPLATE seq_t ZSTD_decodeSequenceLong(seqState_t* seqState, ZSTD_longOffset_e const longOffsets) { @@ -1081,11 +1119,13 @@ ZSTD_decompressSequencesLong_default(ZSTD_DCtx* dctx, { return ZSTD_decompressSequencesLong_body(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset); } +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT */ #if DYNAMIC_BMI2 +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG static TARGET_ATTRIBUTE("bmi2") size_t ZSTD_decompressSequences_bmi2(ZSTD_DCtx* dctx, void* dst, size_t maxDstSize, @@ -1094,7 +1134,9 @@ ZSTD_decompressSequences_bmi2(ZSTD_DCtx* dctx, { return ZSTD_decompressSequences_body(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset); } +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG */ +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT static TARGET_ATTRIBUTE("bmi2") size_t ZSTD_decompressSequencesLong_bmi2(ZSTD_DCtx* dctx, void* dst, size_t maxDstSize, @@ -1103,8 +1145,9 @@ ZSTD_decompressSequencesLong_bmi2(ZSTD_DCtx* dctx, { return ZSTD_decompressSequencesLong_body(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset); } +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT */ -#endif +#endif /* DYNAMIC_BMI2 */ typedef size_t (*ZSTD_decompressSequences_t)( ZSTD_DCtx* dctx, @@ -1112,6 +1155,7 @@ typedef size_t (*ZSTD_decompressSequences_t)( const void* seqStart, size_t seqSize, int nbSeq, const ZSTD_longOffset_e isLongOffset); +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG static size_t ZSTD_decompressSequences(ZSTD_DCtx* dctx, void* dst, size_t maxDstSize, const void* seqStart, size_t seqSize, int nbSeq, @@ -1125,8 +1169,10 @@ ZSTD_decompressSequences(ZSTD_DCtx* dctx, void* dst, size_t maxDstSize, #endif return ZSTD_decompressSequences_default(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset); } +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG */ +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT /* ZSTD_decompressSequencesLong() : * decompression function triggered when a minimum share of offsets is considered "long", * aka out of cache. @@ -1146,9 +1192,12 @@ ZSTD_decompressSequencesLong(ZSTD_DCtx* dctx, #endif return ZSTD_decompressSequencesLong_default(dctx, dst, maxDstSize, seqStart, seqSize, nbSeq, isLongOffset); } +#endif /* ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT */ +#if !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) /* ZSTD_getLongOffsetsShare() : * condition : offTable must be valid * @return : "share" of long offsets (arbitrarily defined as > (1<<23)) @@ -1173,6 +1222,7 @@ ZSTD_getLongOffsetsShare(const ZSTD_seqSymbol* offTable) return total; } +#endif size_t @@ -1201,13 +1251,23 @@ ZSTD_decompressBlock_internal(ZSTD_DCtx* dctx, } /* Build Decoding Tables */ - { int usePrefetchDecoder = dctx->ddictIsCold; + { + /* These macros control at build-time which decompressor implementation + * we use. If neither is defined, we do some inspection and dispatch at + * runtime. + */ +#if !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) + int usePrefetchDecoder = dctx->ddictIsCold; +#endif int nbSeq; size_t const seqHSize = ZSTD_decodeSeqHeaders(dctx, &nbSeq, ip, srcSize); if (ZSTD_isError(seqHSize)) return seqHSize; ip += seqHSize; srcSize -= seqHSize; +#if !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) if ( !usePrefetchDecoder && (!frame || (dctx->fParams.windowSize > (1<<24))) && (nbSeq>ADVANCED_SEQS) ) { /* could probably use a larger nbSeq limit */ @@ -1215,14 +1275,22 @@ ZSTD_decompressBlock_internal(ZSTD_DCtx* dctx, U32 const minShare = MEM_64bits() ? 7 : 20; /* heuristic values, correspond to 2.73% and 7.81% */ usePrefetchDecoder = (shareLongOffsets >= minShare); } +#endif dctx->ddictIsCold = 0; +#if !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT) && \ + !defined(ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG) if (usePrefetchDecoder) +#endif +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_SHORT return ZSTD_decompressSequencesLong(dctx, dst, dstCapacity, ip, srcSize, nbSeq, isLongOffset); +#endif +#ifndef ZSTD_FORCE_DECOMPRESS_SEQUENCES_LONG /* else */ return ZSTD_decompressSequences(dctx, dst, dstCapacity, ip, srcSize, nbSeq, isLongOffset); +#endif } }