From 2587cf3b9526959202fa0d64d51e18b73fce5edf Mon Sep 17 00:00:00 2001 From: Connor McLaughlin Date: Thu, 12 May 2022 17:48:39 +1000 Subject: [PATCH] 3rdparty: Add lzma/7Z SDK 21.07 --- 3rdparty/lzma/CMakeLists.txt | 77 + 3rdparty/lzma/LICENSE.txt | 5 + 3rdparty/lzma/include/7z.h | 204 ++ 3rdparty/lzma/include/7zAlloc.h | 19 + 3rdparty/lzma/include/7zBuf.h | 35 + 3rdparty/lzma/include/7zCrc.h | 25 + 3rdparty/lzma/include/7zFile.h | 91 + 3rdparty/lzma/include/7zTypes.h | 525 +++++ 3rdparty/lzma/include/7zVersion.h | 27 + 3rdparty/lzma/include/Alloc.h | 58 + 3rdparty/lzma/include/Bcj2.h | 146 ++ 3rdparty/lzma/include/Bra.h | 64 + 3rdparty/lzma/include/Compiler.h | 43 + 3rdparty/lzma/include/CpuArch.h | 442 ++++ 3rdparty/lzma/include/Delta.h | 19 + 3rdparty/lzma/include/LzFind.h | 136 ++ 3rdparty/lzma/include/LzHash.h | 34 + 3rdparty/lzma/include/Lzma2Dec.h | 120 ++ 3rdparty/lzma/include/Lzma2DecMt.h | 79 + 3rdparty/lzma/include/Lzma2Enc.h | 55 + 3rdparty/lzma/include/Lzma86.h | 111 + 3rdparty/lzma/include/LzmaDec.h | 236 +++ 3rdparty/lzma/include/LzmaEnc.h | 78 + 3rdparty/lzma/include/LzmaLib.h | 138 ++ 3rdparty/lzma/include/Ppmd.h | 167 ++ 3rdparty/lzma/include/Ppmd7.h | 181 ++ 3rdparty/lzma/include/Precomp.h | 10 + 3rdparty/lzma/include/RotateDefs.h | 30 + 3rdparty/lzma/include/Sha256.h | 76 + 3rdparty/lzma/include/Xz.h | 517 +++++ 3rdparty/lzma/include/XzCrc64.h | 26 + 3rdparty/lzma/include/XzEnc.h | 60 + 3rdparty/lzma/lzma.vcxproj | 112 + 3rdparty/lzma/lzma.vcxproj.filters | 75 + 3rdparty/lzma/src/7zAlloc.c | 80 + 3rdparty/lzma/src/7zArcIn.c | 1783 ++++++++++++++++ 3rdparty/lzma/src/7zBuf.c | 36 + 3rdparty/lzma/src/7zBuf2.c | 52 + 3rdparty/lzma/src/7zCrc.c | 322 +++ 3rdparty/lzma/src/7zCrcOpt.c | 117 + 3rdparty/lzma/src/7zDec.c | 600 ++++++ 3rdparty/lzma/src/7zFile.c | 442 ++++ 3rdparty/lzma/src/7zStream.c | 176 ++ 3rdparty/lzma/src/Alloc.c | 463 ++++ 3rdparty/lzma/src/Bcj2.c | 257 +++ 3rdparty/lzma/src/Bcj2Enc.c | 311 +++ 3rdparty/lzma/src/Bra.c | 230 ++ 3rdparty/lzma/src/Bra86.c | 82 + 3rdparty/lzma/src/BraIA64.c | 53 + 3rdparty/lzma/src/CpuArch.c | 478 +++++ 3rdparty/lzma/src/Delta.c | 169 ++ 3rdparty/lzma/src/LzFind.c | 1628 ++++++++++++++ 3rdparty/lzma/src/LzFindOpt.c | 578 +++++ 3rdparty/lzma/src/Lzma2Dec.c | 489 +++++ 3rdparty/lzma/src/Lzma2DecMt.c | 1090 ++++++++++ 3rdparty/lzma/src/Lzma2Enc.c | 803 +++++++ 3rdparty/lzma/src/Lzma86Dec.c | 54 + 3rdparty/lzma/src/Lzma86Enc.c | 104 + 3rdparty/lzma/src/LzmaDec.c | 1363 ++++++++++++ 3rdparty/lzma/src/LzmaEnc.c | 3165 ++++++++++++++++++++++++++++ 3rdparty/lzma/src/LzmaLib.c | 40 + 3rdparty/lzma/src/Ppmd7.c | 1104 ++++++++++ 3rdparty/lzma/src/Ppmd7Dec.c | 297 +++ 3rdparty/lzma/src/Ppmd7Enc.c | 323 +++ 3rdparty/lzma/src/Sha256.c | 486 +++++ 3rdparty/lzma/src/Sha256Opt.c | 373 ++++ 3rdparty/lzma/src/Xz.c | 90 + 3rdparty/lzma/src/XzCrc64.c | 86 + 3rdparty/lzma/src/XzCrc64Opt.c | 71 + 3rdparty/lzma/src/XzDec.c | 2837 +++++++++++++++++++++++++ 3rdparty/lzma/src/XzEnc.c | 1330 ++++++++++++ 3rdparty/lzma/src/XzIn.c | 325 +++ PCSX2_qt.sln | 15 + cmake/SearchForStuff.cmake | 1 + 74 files changed, 26324 insertions(+) create mode 100644 3rdparty/lzma/CMakeLists.txt create mode 100644 3rdparty/lzma/LICENSE.txt create mode 100644 3rdparty/lzma/include/7z.h create mode 100644 3rdparty/lzma/include/7zAlloc.h create mode 100644 3rdparty/lzma/include/7zBuf.h create mode 100644 3rdparty/lzma/include/7zCrc.h create mode 100644 3rdparty/lzma/include/7zFile.h create mode 100644 3rdparty/lzma/include/7zTypes.h create mode 100644 3rdparty/lzma/include/7zVersion.h create mode 100644 3rdparty/lzma/include/Alloc.h create mode 100644 3rdparty/lzma/include/Bcj2.h create mode 100644 3rdparty/lzma/include/Bra.h create mode 100644 3rdparty/lzma/include/Compiler.h create mode 100644 3rdparty/lzma/include/CpuArch.h create mode 100644 3rdparty/lzma/include/Delta.h create mode 100644 3rdparty/lzma/include/LzFind.h create mode 100644 3rdparty/lzma/include/LzHash.h create mode 100644 3rdparty/lzma/include/Lzma2Dec.h create mode 100644 3rdparty/lzma/include/Lzma2DecMt.h create mode 100644 3rdparty/lzma/include/Lzma2Enc.h create mode 100644 3rdparty/lzma/include/Lzma86.h create mode 100644 3rdparty/lzma/include/LzmaDec.h create mode 100644 3rdparty/lzma/include/LzmaEnc.h create mode 100644 3rdparty/lzma/include/LzmaLib.h create mode 100644 3rdparty/lzma/include/Ppmd.h create mode 100644 3rdparty/lzma/include/Ppmd7.h create mode 100644 3rdparty/lzma/include/Precomp.h create mode 100644 3rdparty/lzma/include/RotateDefs.h create mode 100644 3rdparty/lzma/include/Sha256.h create mode 100644 3rdparty/lzma/include/Xz.h create mode 100644 3rdparty/lzma/include/XzCrc64.h create mode 100644 3rdparty/lzma/include/XzEnc.h create mode 100644 3rdparty/lzma/lzma.vcxproj create mode 100644 3rdparty/lzma/lzma.vcxproj.filters create mode 100644 3rdparty/lzma/src/7zAlloc.c create mode 100644 3rdparty/lzma/src/7zArcIn.c create mode 100644 3rdparty/lzma/src/7zBuf.c create mode 100644 3rdparty/lzma/src/7zBuf2.c create mode 100644 3rdparty/lzma/src/7zCrc.c create mode 100644 3rdparty/lzma/src/7zCrcOpt.c create mode 100644 3rdparty/lzma/src/7zDec.c create mode 100644 3rdparty/lzma/src/7zFile.c create mode 100644 3rdparty/lzma/src/7zStream.c create mode 100644 3rdparty/lzma/src/Alloc.c create mode 100644 3rdparty/lzma/src/Bcj2.c create mode 100644 3rdparty/lzma/src/Bcj2Enc.c create mode 100644 3rdparty/lzma/src/Bra.c create mode 100644 3rdparty/lzma/src/Bra86.c create mode 100644 3rdparty/lzma/src/BraIA64.c create mode 100644 3rdparty/lzma/src/CpuArch.c create mode 100644 3rdparty/lzma/src/Delta.c create mode 100644 3rdparty/lzma/src/LzFind.c create mode 100644 3rdparty/lzma/src/LzFindOpt.c create mode 100644 3rdparty/lzma/src/Lzma2Dec.c create mode 100644 3rdparty/lzma/src/Lzma2DecMt.c create mode 100644 3rdparty/lzma/src/Lzma2Enc.c create mode 100644 3rdparty/lzma/src/Lzma86Dec.c create mode 100644 3rdparty/lzma/src/Lzma86Enc.c create mode 100644 3rdparty/lzma/src/LzmaDec.c create mode 100644 3rdparty/lzma/src/LzmaEnc.c create mode 100644 3rdparty/lzma/src/LzmaLib.c create mode 100644 3rdparty/lzma/src/Ppmd7.c create mode 100644 3rdparty/lzma/src/Ppmd7Dec.c create mode 100644 3rdparty/lzma/src/Ppmd7Enc.c create mode 100644 3rdparty/lzma/src/Sha256.c create mode 100644 3rdparty/lzma/src/Sha256Opt.c create mode 100644 3rdparty/lzma/src/Xz.c create mode 100644 3rdparty/lzma/src/XzCrc64.c create mode 100644 3rdparty/lzma/src/XzCrc64Opt.c create mode 100644 3rdparty/lzma/src/XzDec.c create mode 100644 3rdparty/lzma/src/XzEnc.c create mode 100644 3rdparty/lzma/src/XzIn.c diff --git a/3rdparty/lzma/CMakeLists.txt b/3rdparty/lzma/CMakeLists.txt new file mode 100644 index 0000000000..cf6361f9b5 --- /dev/null +++ b/3rdparty/lzma/CMakeLists.txt @@ -0,0 +1,77 @@ +add_library(pcsx2-lzma STATIC + include/7z.h + include/7zAlloc.h + include/7zBuf.h + include/7zCrc.h + include/7zFile.h + include/7zTypes.h + include/7zVersion.h + include/Alloc.h + include/Bcj2.h + include/Bra.h + include/Compiler.h + include/CpuArch.h + include/Delta.h + include/LzFind.h + include/LzHash.h + include/Lzma2Dec.h + include/Lzma2DecMt.h + include/Lzma2Enc.h + include/Lzma86.h + include/LzmaDec.h + include/LzmaEnc.h + include/LzmaLib.h + include/Ppmd.h + include/Ppmd7.h + include/Precomp.h + include/RotateDefs.h + include/Sha256.h + include/Xz.h + include/XzCrc64.h + include/XzEnc.h + src/7zAlloc.c + src/7zArcIn.c + src/7zBuf.c + src/7zBuf2.c + src/7zCrc.c + src/7zCrcOpt.c + src/7zDec.c + src/7zFile.c + src/7zStream.c + src/Alloc.c + src/Bcj2.c + src/Bcj2Enc.c + src/Bra.c + src/Bra86.c + src/BraIA64.c + src/CpuArch.c + src/Delta.c + src/LzFind.c + src/LzFindOpt.c + src/Lzma2Dec.c + src/Lzma2DecMt.c + src/Lzma2Enc.c + src/Lzma86Dec.c + src/Lzma86Enc.c + src/LzmaDec.c + src/LzmaEnc.c + src/LzmaLib.c + src/Ppmd7.c + src/Ppmd7Dec.c + src/Ppmd7Enc.c + src/Sha256.c + src/Sha256Opt.c + src/Xz.c + src/XzCrc64.c + src/XzCrc64Opt.c + src/XzDec.c + src/XzEnc.c + src/XzIn.c +) + +target_compile_definitions(pcsx2-lzma PRIVATE _7ZIP_ST) + +target_include_directories(pcsx2-lzma PRIVATE "${CMAKE_CURRENT_SOURCE_DIR}/include") +target_include_directories(pcsx2-lzma INTERFACE "${CMAKE_CURRENT_SOURCE_DIR}/include") + +add_library(LZMA::LZMA ALIAS pcsx2-lzma) \ No newline at end of file diff --git a/3rdparty/lzma/LICENSE.txt b/3rdparty/lzma/LICENSE.txt new file mode 100644 index 0000000000..f5338922a2 --- /dev/null +++ b/3rdparty/lzma/LICENSE.txt @@ -0,0 +1,5 @@ +https://www.7-zip.org/sdk.html + +LZMA SDK is placed in the public domain. + +Anyone is free to copy, modify, publish, use, compile, sell, or distribute the original LZMA SDK code, either in source code form or as a compiled binary, for any purpose, commercial or non-commercial, and by any means. \ No newline at end of file diff --git a/3rdparty/lzma/include/7z.h b/3rdparty/lzma/include/7z.h new file mode 100644 index 0000000000..304f75ffc5 --- /dev/null +++ b/3rdparty/lzma/include/7z.h @@ -0,0 +1,204 @@ +/* 7z.h -- 7z interface +2018-07-02 : Igor Pavlov : Public domain */ + +#ifndef __7Z_H +#define __7Z_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define k7zStartHeaderSize 0x20 +#define k7zSignatureSize 6 + +extern const Byte k7zSignature[k7zSignatureSize]; + +typedef struct +{ + const Byte *Data; + size_t Size; +} CSzData; + +/* CSzCoderInfo & CSzFolder support only default methods */ + +typedef struct +{ + size_t PropsOffset; + UInt32 MethodID; + Byte NumStreams; + Byte PropsSize; +} CSzCoderInfo; + +typedef struct +{ + UInt32 InIndex; + UInt32 OutIndex; +} CSzBond; + +#define SZ_NUM_CODERS_IN_FOLDER_MAX 4 +#define SZ_NUM_BONDS_IN_FOLDER_MAX 3 +#define SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX 4 + +typedef struct +{ + UInt32 NumCoders; + UInt32 NumBonds; + UInt32 NumPackStreams; + UInt32 UnpackStream; + UInt32 PackStreams[SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX]; + CSzBond Bonds[SZ_NUM_BONDS_IN_FOLDER_MAX]; + CSzCoderInfo Coders[SZ_NUM_CODERS_IN_FOLDER_MAX]; +} CSzFolder; + + +SRes SzGetNextFolderItem(CSzFolder *f, CSzData *sd); + +typedef struct +{ + UInt32 Low; + UInt32 High; +} CNtfsFileTime; + +typedef struct +{ + Byte *Defs; /* MSB 0 bit numbering */ + UInt32 *Vals; +} CSzBitUi32s; + +typedef struct +{ + Byte *Defs; /* MSB 0 bit numbering */ + // UInt64 *Vals; + CNtfsFileTime *Vals; +} CSzBitUi64s; + +#define SzBitArray_Check(p, i) (((p)[(i) >> 3] & (0x80 >> ((i) & 7))) != 0) + +#define SzBitWithVals_Check(p, i) ((p)->Defs && ((p)->Defs[(i) >> 3] & (0x80 >> ((i) & 7))) != 0) + +typedef struct +{ + UInt32 NumPackStreams; + UInt32 NumFolders; + + UInt64 *PackPositions; // NumPackStreams + 1 + CSzBitUi32s FolderCRCs; // NumFolders + + size_t *FoCodersOffsets; // NumFolders + 1 + UInt32 *FoStartPackStreamIndex; // NumFolders + 1 + UInt32 *FoToCoderUnpackSizes; // NumFolders + 1 + Byte *FoToMainUnpackSizeIndex; // NumFolders + UInt64 *CoderUnpackSizes; // for all coders in all folders + + Byte *CodersData; + + UInt64 RangeLimit; +} CSzAr; + +UInt64 SzAr_GetFolderUnpackSize(const CSzAr *p, UInt32 folderIndex); + +SRes SzAr_DecodeFolder(const CSzAr *p, UInt32 folderIndex, + ILookInStream *stream, UInt64 startPos, + Byte *outBuffer, size_t outSize, + ISzAllocPtr allocMain); + +typedef struct +{ + CSzAr db; + + UInt64 startPosAfterHeader; + UInt64 dataPos; + + UInt32 NumFiles; + + UInt64 *UnpackPositions; // NumFiles + 1 + // Byte *IsEmptyFiles; + Byte *IsDirs; + CSzBitUi32s CRCs; + + CSzBitUi32s Attribs; + // CSzBitUi32s Parents; + CSzBitUi64s MTime; + CSzBitUi64s CTime; + + UInt32 *FolderToFile; // NumFolders + 1 + UInt32 *FileToFolder; // NumFiles + + size_t *FileNameOffsets; /* in 2-byte steps */ + Byte *FileNames; /* UTF-16-LE */ +} CSzArEx; + +#define SzArEx_IsDir(p, i) (SzBitArray_Check((p)->IsDirs, i)) + +#define SzArEx_GetFileSize(p, i) ((p)->UnpackPositions[(i) + 1] - (p)->UnpackPositions[i]) + +void SzArEx_Init(CSzArEx *p); +void SzArEx_Free(CSzArEx *p, ISzAllocPtr alloc); +UInt64 SzArEx_GetFolderStreamPos(const CSzArEx *p, UInt32 folderIndex, UInt32 indexInFolder); +int SzArEx_GetFolderFullPackSize(const CSzArEx *p, UInt32 folderIndex, UInt64 *resSize); + +/* +if dest == NULL, the return value specifies the required size of the buffer, + in 16-bit characters, including the null-terminating character. +if dest != NULL, the return value specifies the number of 16-bit characters that + are written to the dest, including the null-terminating character. */ + +size_t SzArEx_GetFileNameUtf16(const CSzArEx *p, size_t fileIndex, UInt16 *dest); + +/* +size_t SzArEx_GetFullNameLen(const CSzArEx *p, size_t fileIndex); +UInt16 *SzArEx_GetFullNameUtf16_Back(const CSzArEx *p, size_t fileIndex, UInt16 *dest); +*/ + + + +/* + SzArEx_Extract extracts file from archive + + *outBuffer must be 0 before first call for each new archive. + + Extracting cache: + If you need to decompress more than one file, you can send + these values from previous call: + *blockIndex, + *outBuffer, + *outBufferSize + You can consider "*outBuffer" as cache of solid block. If your archive is solid, + it will increase decompression speed. + + If you use external function, you can declare these 3 cache variables + (blockIndex, outBuffer, outBufferSize) as static in that external function. + + Free *outBuffer and set *outBuffer to 0, if you want to flush cache. +*/ + +SRes SzArEx_Extract( + const CSzArEx *db, + ILookInStream *inStream, + UInt32 fileIndex, /* index of file */ + UInt32 *blockIndex, /* index of solid block */ + Byte **outBuffer, /* pointer to pointer to output buffer (allocated with allocMain) */ + size_t *outBufferSize, /* buffer size for output buffer */ + size_t *offset, /* offset of stream for required file in *outBuffer */ + size_t *outSizeProcessed, /* size of file in *outBuffer */ + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp); + + +/* +SzArEx_Open Errors: +SZ_ERROR_NO_ARCHIVE +SZ_ERROR_ARCHIVE +SZ_ERROR_UNSUPPORTED +SZ_ERROR_MEM +SZ_ERROR_CRC +SZ_ERROR_INPUT_EOF +SZ_ERROR_FAIL +*/ + +SRes SzArEx_Open(CSzArEx *p, ILookInStream *inStream, + ISzAllocPtr allocMain, ISzAllocPtr allocTemp); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/7zAlloc.h b/3rdparty/lzma/include/7zAlloc.h new file mode 100644 index 0000000000..44778f9b2e --- /dev/null +++ b/3rdparty/lzma/include/7zAlloc.h @@ -0,0 +1,19 @@ +/* 7zAlloc.h -- Allocation functions +2017-04-03 : Igor Pavlov : Public domain */ + +#ifndef __7Z_ALLOC_H +#define __7Z_ALLOC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +void *SzAlloc(ISzAllocPtr p, size_t size); +void SzFree(ISzAllocPtr p, void *address); + +void *SzAllocTemp(ISzAllocPtr p, size_t size); +void SzFreeTemp(ISzAllocPtr p, void *address); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/7zBuf.h b/3rdparty/lzma/include/7zBuf.h new file mode 100644 index 0000000000..81d1b5b646 --- /dev/null +++ b/3rdparty/lzma/include/7zBuf.h @@ -0,0 +1,35 @@ +/* 7zBuf.h -- Byte Buffer +2017-04-03 : Igor Pavlov : Public domain */ + +#ifndef __7Z_BUF_H +#define __7Z_BUF_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +typedef struct +{ + Byte *data; + size_t size; +} CBuf; + +void Buf_Init(CBuf *p); +int Buf_Create(CBuf *p, size_t size, ISzAllocPtr alloc); +void Buf_Free(CBuf *p, ISzAllocPtr alloc); + +typedef struct +{ + Byte *data; + size_t size; + size_t pos; +} CDynBuf; + +void DynBuf_Construct(CDynBuf *p); +void DynBuf_SeekToBeg(CDynBuf *p); +int DynBuf_Write(CDynBuf *p, const Byte *buf, size_t size, ISzAllocPtr alloc); +void DynBuf_Free(CDynBuf *p, ISzAllocPtr alloc); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/7zCrc.h b/3rdparty/lzma/include/7zCrc.h new file mode 100644 index 0000000000..8fd5795871 --- /dev/null +++ b/3rdparty/lzma/include/7zCrc.h @@ -0,0 +1,25 @@ +/* 7zCrc.h -- CRC32 calculation +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __7Z_CRC_H +#define __7Z_CRC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +extern UInt32 g_CrcTable[]; + +/* Call CrcGenerateTable one time before other CRC functions */ +void MY_FAST_CALL CrcGenerateTable(void); + +#define CRC_INIT_VAL 0xFFFFFFFF +#define CRC_GET_DIGEST(crc) ((crc) ^ CRC_INIT_VAL) +#define CRC_UPDATE_BYTE(crc, b) (g_CrcTable[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt32 MY_FAST_CALL CrcUpdate(UInt32 crc, const void *data, size_t size); +UInt32 MY_FAST_CALL CrcCalc(const void *data, size_t size); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/7zFile.h b/3rdparty/lzma/include/7zFile.h new file mode 100644 index 0000000000..788abb6b9d --- /dev/null +++ b/3rdparty/lzma/include/7zFile.h @@ -0,0 +1,91 @@ +/* 7zFile.h -- File IO +2021-02-15 : Igor Pavlov : Public domain */ + +#ifndef __7Z_FILE_H +#define __7Z_FILE_H + +#ifdef _WIN32 +#define USE_WINDOWS_FILE +// #include +#endif + +#ifdef USE_WINDOWS_FILE +#include +#else +// note: USE_FOPEN mode is limited to 32-bit file size +// #define USE_FOPEN +// #include +#endif + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* ---------- File ---------- */ + +typedef struct +{ + #ifdef USE_WINDOWS_FILE + HANDLE handle; + #elif defined(USE_FOPEN) + FILE *file; + #else + int fd; + #endif +} CSzFile; + +void File_Construct(CSzFile *p); +#if !defined(UNDER_CE) || !defined(USE_WINDOWS_FILE) +WRes InFile_Open(CSzFile *p, const char *name); +WRes OutFile_Open(CSzFile *p, const char *name); +#endif +#ifdef USE_WINDOWS_FILE +WRes InFile_OpenW(CSzFile *p, const WCHAR *name); +WRes OutFile_OpenW(CSzFile *p, const WCHAR *name); +#endif +WRes File_Close(CSzFile *p); + +/* reads max(*size, remain file's size) bytes */ +WRes File_Read(CSzFile *p, void *data, size_t *size); + +/* writes *size bytes */ +WRes File_Write(CSzFile *p, const void *data, size_t *size); + +WRes File_Seek(CSzFile *p, Int64 *pos, ESzSeek origin); +WRes File_GetLength(CSzFile *p, UInt64 *length); + + +/* ---------- FileInStream ---------- */ + +typedef struct +{ + ISeqInStream vt; + CSzFile file; + WRes wres; +} CFileSeqInStream; + +void FileSeqInStream_CreateVTable(CFileSeqInStream *p); + + +typedef struct +{ + ISeekInStream vt; + CSzFile file; + WRes wres; +} CFileInStream; + +void FileInStream_CreateVTable(CFileInStream *p); + + +typedef struct +{ + ISeqOutStream vt; + CSzFile file; + WRes wres; +} CFileOutStream; + +void FileOutStream_CreateVTable(CFileOutStream *p); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/7zTypes.h b/3rdparty/lzma/include/7zTypes.h new file mode 100644 index 0000000000..fe4fde3ffc --- /dev/null +++ b/3rdparty/lzma/include/7zTypes.h @@ -0,0 +1,525 @@ +/* 7zTypes.h -- Basic types +2021-12-25 : Igor Pavlov : Public domain */ + +#ifndef __7Z_TYPES_H +#define __7Z_TYPES_H + +#ifdef _WIN32 +/* #include */ +#else +#include +#endif + +#include + +#ifndef EXTERN_C_BEGIN +#ifdef __cplusplus +#define EXTERN_C_BEGIN extern "C" { +#define EXTERN_C_END } +#else +#define EXTERN_C_BEGIN +#define EXTERN_C_END +#endif +#endif + +EXTERN_C_BEGIN + +#define SZ_OK 0 + +#define SZ_ERROR_DATA 1 +#define SZ_ERROR_MEM 2 +#define SZ_ERROR_CRC 3 +#define SZ_ERROR_UNSUPPORTED 4 +#define SZ_ERROR_PARAM 5 +#define SZ_ERROR_INPUT_EOF 6 +#define SZ_ERROR_OUTPUT_EOF 7 +#define SZ_ERROR_READ 8 +#define SZ_ERROR_WRITE 9 +#define SZ_ERROR_PROGRESS 10 +#define SZ_ERROR_FAIL 11 +#define SZ_ERROR_THREAD 12 + +#define SZ_ERROR_ARCHIVE 16 +#define SZ_ERROR_NO_ARCHIVE 17 + +typedef int SRes; + + +#ifdef _MSC_VER + #if _MSC_VER > 1200 + #define MY_ALIGN(n) __declspec(align(n)) + #else + #define MY_ALIGN(n) + #endif +#else + #define MY_ALIGN(n) __attribute__ ((aligned(n))) +#endif + + +#ifdef _WIN32 + +/* typedef DWORD WRes; */ +typedef unsigned WRes; +#define MY_SRes_HRESULT_FROM_WRes(x) HRESULT_FROM_WIN32(x) + +// #define MY_HRES_ERROR__INTERNAL_ERROR MY_SRes_HRESULT_FROM_WRes(ERROR_INTERNAL_ERROR) + +#else // _WIN32 + +// #define ENV_HAVE_LSTAT +typedef int WRes; + +// (FACILITY_ERRNO = 0x800) is 7zip's FACILITY constant to represent (errno) errors in HRESULT +#define MY__FACILITY_ERRNO 0x800 +#define MY__FACILITY_WIN32 7 +#define MY__FACILITY__WRes MY__FACILITY_ERRNO + +#define MY_HRESULT_FROM_errno_CONST_ERROR(x) ((HRESULT)( \ + ( (HRESULT)(x) & 0x0000FFFF) \ + | (MY__FACILITY__WRes << 16) \ + | (HRESULT)0x80000000 )) + +#define MY_SRes_HRESULT_FROM_WRes(x) \ + ((HRESULT)(x) <= 0 ? ((HRESULT)(x)) : MY_HRESULT_FROM_errno_CONST_ERROR(x)) + +// we call macro HRESULT_FROM_WIN32 for system errors (WRes) that are (errno) +#define HRESULT_FROM_WIN32(x) MY_SRes_HRESULT_FROM_WRes(x) + +/* +#define ERROR_FILE_NOT_FOUND 2L +#define ERROR_ACCESS_DENIED 5L +#define ERROR_NO_MORE_FILES 18L +#define ERROR_LOCK_VIOLATION 33L +#define ERROR_FILE_EXISTS 80L +#define ERROR_DISK_FULL 112L +#define ERROR_NEGATIVE_SEEK 131L +#define ERROR_ALREADY_EXISTS 183L +#define ERROR_DIRECTORY 267L +#define ERROR_TOO_MANY_POSTS 298L + +#define ERROR_INTERNAL_ERROR 1359L +#define ERROR_INVALID_REPARSE_DATA 4392L +#define ERROR_REPARSE_TAG_INVALID 4393L +#define ERROR_REPARSE_TAG_MISMATCH 4394L +*/ + +// we use errno equivalents for some WIN32 errors: + +#define ERROR_INVALID_PARAMETER EINVAL +#define ERROR_INVALID_FUNCTION EINVAL +#define ERROR_ALREADY_EXISTS EEXIST +#define ERROR_FILE_EXISTS EEXIST +#define ERROR_PATH_NOT_FOUND ENOENT +#define ERROR_FILE_NOT_FOUND ENOENT +#define ERROR_DISK_FULL ENOSPC +// #define ERROR_INVALID_HANDLE EBADF + +// we use FACILITY_WIN32 for errors that has no errno equivalent +// Too many posts were made to a semaphore. +#define ERROR_TOO_MANY_POSTS ((HRESULT)0x8007012AL) +#define ERROR_INVALID_REPARSE_DATA ((HRESULT)0x80071128L) +#define ERROR_REPARSE_TAG_INVALID ((HRESULT)0x80071129L) + +// if (MY__FACILITY__WRes != FACILITY_WIN32), +// we use FACILITY_WIN32 for COM errors: +#define E_OUTOFMEMORY ((HRESULT)0x8007000EL) +#define E_INVALIDARG ((HRESULT)0x80070057L) +#define MY__E_ERROR_NEGATIVE_SEEK ((HRESULT)0x80070083L) + +/* +// we can use FACILITY_ERRNO for some COM errors, that have errno equivalents: +#define E_OUTOFMEMORY MY_HRESULT_FROM_errno_CONST_ERROR(ENOMEM) +#define E_INVALIDARG MY_HRESULT_FROM_errno_CONST_ERROR(EINVAL) +#define MY__E_ERROR_NEGATIVE_SEEK MY_HRESULT_FROM_errno_CONST_ERROR(EINVAL) +*/ + +// gcc / clang : (sizeof(long) == sizeof(void*)) in 32/64 bits +typedef long INT_PTR; +typedef unsigned long UINT_PTR; + +#define TEXT(quote) quote + +#define FILE_ATTRIBUTE_READONLY 0x0001 +#define FILE_ATTRIBUTE_HIDDEN 0x0002 +#define FILE_ATTRIBUTE_SYSTEM 0x0004 +#define FILE_ATTRIBUTE_DIRECTORY 0x0010 +#define FILE_ATTRIBUTE_ARCHIVE 0x0020 +#define FILE_ATTRIBUTE_DEVICE 0x0040 +#define FILE_ATTRIBUTE_NORMAL 0x0080 +#define FILE_ATTRIBUTE_TEMPORARY 0x0100 +#define FILE_ATTRIBUTE_SPARSE_FILE 0x0200 +#define FILE_ATTRIBUTE_REPARSE_POINT 0x0400 +#define FILE_ATTRIBUTE_COMPRESSED 0x0800 +#define FILE_ATTRIBUTE_OFFLINE 0x1000 +#define FILE_ATTRIBUTE_NOT_CONTENT_INDEXED 0x2000 +#define FILE_ATTRIBUTE_ENCRYPTED 0x4000 + +#define FILE_ATTRIBUTE_UNIX_EXTENSION 0x8000 /* trick for Unix */ + +#endif + + +#ifndef RINOK +#define RINOK(x) { int __result__ = (x); if (__result__ != 0) return __result__; } +#endif + +#ifndef RINOK_WRes +#define RINOK_WRes(x) { WRes __result__ = (x); if (__result__ != 0) return __result__; } +#endif + +typedef unsigned char Byte; +typedef short Int16; +typedef unsigned short UInt16; + +#ifdef _LZMA_UINT32_IS_ULONG +typedef long Int32; +typedef unsigned long UInt32; +#else +typedef int Int32; +typedef unsigned int UInt32; +#endif + + +#ifndef _WIN32 + +typedef int INT; +typedef Int32 INT32; +typedef unsigned int UINT; +typedef UInt32 UINT32; +typedef INT32 LONG; // LONG, ULONG and DWORD must be 32-bit for _WIN32 compatibility +typedef UINT32 ULONG; + +#undef DWORD +typedef UINT32 DWORD; + +#define VOID void + +#define HRESULT LONG + +typedef void *LPVOID; +// typedef void VOID; +// typedef ULONG_PTR DWORD_PTR, *PDWORD_PTR; +// gcc / clang on Unix : sizeof(long==sizeof(void*) in 32 or 64 bits) +typedef long INT_PTR; +typedef unsigned long UINT_PTR; +typedef long LONG_PTR; +typedef unsigned long DWORD_PTR; + +typedef size_t SIZE_T; + +#endif // _WIN32 + + +#define MY_HRES_ERROR__INTERNAL_ERROR ((HRESULT)0x8007054FL) + + +#ifdef _SZ_NO_INT_64 + +/* define _SZ_NO_INT_64, if your compiler doesn't support 64-bit integers. + NOTES: Some code will work incorrectly in that case! */ + +typedef long Int64; +typedef unsigned long UInt64; + +#else + +#if defined(_MSC_VER) || defined(__BORLANDC__) +typedef __int64 Int64; +typedef unsigned __int64 UInt64; +#define UINT64_CONST(n) n +#else +typedef long long int Int64; +typedef unsigned long long int UInt64; +#define UINT64_CONST(n) n ## ULL +#endif + +#endif + +#ifdef _LZMA_NO_SYSTEM_SIZE_T +typedef UInt32 SizeT; +#else +typedef size_t SizeT; +#endif + +typedef int BoolInt; +/* typedef BoolInt Bool; */ +#define True 1 +#define False 0 + + +#ifdef _WIN32 +#define MY_STD_CALL __stdcall +#else +#define MY_STD_CALL +#endif + +#ifdef _MSC_VER + +#if _MSC_VER >= 1300 +#define MY_NO_INLINE __declspec(noinline) +#else +#define MY_NO_INLINE +#endif + +#define MY_FORCE_INLINE __forceinline + +#define MY_CDECL __cdecl +#define MY_FAST_CALL __fastcall + +#else // _MSC_VER + +#if (defined(__GNUC__) && (__GNUC__ >= 4)) \ + || (defined(__clang__) && (__clang_major__ >= 4)) \ + || defined(__INTEL_COMPILER) \ + || defined(__xlC__) +#define MY_NO_INLINE __attribute__((noinline)) +// #define MY_FORCE_INLINE __attribute__((always_inline)) inline +#else +#define MY_NO_INLINE +#endif + +#define MY_FORCE_INLINE + + +#define MY_CDECL + +#if defined(_M_IX86) \ + || defined(__i386__) +// #define MY_FAST_CALL __attribute__((fastcall)) +// #define MY_FAST_CALL __attribute__((cdecl)) +#define MY_FAST_CALL +#elif defined(MY_CPU_AMD64) +// #define MY_FAST_CALL __attribute__((ms_abi)) +#define MY_FAST_CALL +#else +#define MY_FAST_CALL +#endif + +#endif // _MSC_VER + + +/* The following interfaces use first parameter as pointer to structure */ + +typedef struct IByteIn IByteIn; +struct IByteIn +{ + Byte (*Read)(const IByteIn *p); /* reads one byte, returns 0 in case of EOF or error */ +}; +#define IByteIn_Read(p) (p)->Read(p) + + +typedef struct IByteOut IByteOut; +struct IByteOut +{ + void (*Write)(const IByteOut *p, Byte b); +}; +#define IByteOut_Write(p, b) (p)->Write(p, b) + + +typedef struct ISeqInStream ISeqInStream; +struct ISeqInStream +{ + SRes (*Read)(const ISeqInStream *p, void *buf, size_t *size); + /* if (input(*size) != 0 && output(*size) == 0) means end_of_stream. + (output(*size) < input(*size)) is allowed */ +}; +#define ISeqInStream_Read(p, buf, size) (p)->Read(p, buf, size) + +/* it can return SZ_ERROR_INPUT_EOF */ +SRes SeqInStream_Read(const ISeqInStream *stream, void *buf, size_t size); +SRes SeqInStream_Read2(const ISeqInStream *stream, void *buf, size_t size, SRes errorType); +SRes SeqInStream_ReadByte(const ISeqInStream *stream, Byte *buf); + + +typedef struct ISeqOutStream ISeqOutStream; +struct ISeqOutStream +{ + size_t (*Write)(const ISeqOutStream *p, const void *buf, size_t size); + /* Returns: result - the number of actually written bytes. + (result < size) means error */ +}; +#define ISeqOutStream_Write(p, buf, size) (p)->Write(p, buf, size) + +typedef enum +{ + SZ_SEEK_SET = 0, + SZ_SEEK_CUR = 1, + SZ_SEEK_END = 2 +} ESzSeek; + + +typedef struct ISeekInStream ISeekInStream; +struct ISeekInStream +{ + SRes (*Read)(const ISeekInStream *p, void *buf, size_t *size); /* same as ISeqInStream::Read */ + SRes (*Seek)(const ISeekInStream *p, Int64 *pos, ESzSeek origin); +}; +#define ISeekInStream_Read(p, buf, size) (p)->Read(p, buf, size) +#define ISeekInStream_Seek(p, pos, origin) (p)->Seek(p, pos, origin) + + +typedef struct ILookInStream ILookInStream; +struct ILookInStream +{ + SRes (*Look)(const ILookInStream *p, const void **buf, size_t *size); + /* if (input(*size) != 0 && output(*size) == 0) means end_of_stream. + (output(*size) > input(*size)) is not allowed + (output(*size) < input(*size)) is allowed */ + SRes (*Skip)(const ILookInStream *p, size_t offset); + /* offset must be <= output(*size) of Look */ + + SRes (*Read)(const ILookInStream *p, void *buf, size_t *size); + /* reads directly (without buffer). It's same as ISeqInStream::Read */ + SRes (*Seek)(const ILookInStream *p, Int64 *pos, ESzSeek origin); +}; + +#define ILookInStream_Look(p, buf, size) (p)->Look(p, buf, size) +#define ILookInStream_Skip(p, offset) (p)->Skip(p, offset) +#define ILookInStream_Read(p, buf, size) (p)->Read(p, buf, size) +#define ILookInStream_Seek(p, pos, origin) (p)->Seek(p, pos, origin) + + +SRes LookInStream_LookRead(const ILookInStream *stream, void *buf, size_t *size); +SRes LookInStream_SeekTo(const ILookInStream *stream, UInt64 offset); + +/* reads via ILookInStream::Read */ +SRes LookInStream_Read2(const ILookInStream *stream, void *buf, size_t size, SRes errorType); +SRes LookInStream_Read(const ILookInStream *stream, void *buf, size_t size); + + + +typedef struct +{ + ILookInStream vt; + const ISeekInStream *realStream; + + size_t pos; + size_t size; /* it's data size */ + + /* the following variables must be set outside */ + Byte *buf; + size_t bufSize; +} CLookToRead2; + +void LookToRead2_CreateVTable(CLookToRead2 *p, int lookahead); + +#define LookToRead2_Init(p) { (p)->pos = (p)->size = 0; } + + +typedef struct +{ + ISeqInStream vt; + const ILookInStream *realStream; +} CSecToLook; + +void SecToLook_CreateVTable(CSecToLook *p); + + + +typedef struct +{ + ISeqInStream vt; + const ILookInStream *realStream; +} CSecToRead; + +void SecToRead_CreateVTable(CSecToRead *p); + + +typedef struct ICompressProgress ICompressProgress; + +struct ICompressProgress +{ + SRes (*Progress)(const ICompressProgress *p, UInt64 inSize, UInt64 outSize); + /* Returns: result. (result != SZ_OK) means break. + Value (UInt64)(Int64)-1 for size means unknown value. */ +}; +#define ICompressProgress_Progress(p, inSize, outSize) (p)->Progress(p, inSize, outSize) + + + +typedef struct ISzAlloc ISzAlloc; +typedef const ISzAlloc * ISzAllocPtr; + +struct ISzAlloc +{ + void *(*Alloc)(ISzAllocPtr p, size_t size); + void (*Free)(ISzAllocPtr p, void *address); /* address can be 0 */ +}; + +#define ISzAlloc_Alloc(p, size) (p)->Alloc(p, size) +#define ISzAlloc_Free(p, a) (p)->Free(p, a) + +/* deprecated */ +#define IAlloc_Alloc(p, size) ISzAlloc_Alloc(p, size) +#define IAlloc_Free(p, a) ISzAlloc_Free(p, a) + + + + + +#ifndef MY_offsetof + #ifdef offsetof + #define MY_offsetof(type, m) offsetof(type, m) + /* + #define MY_offsetof(type, m) FIELD_OFFSET(type, m) + */ + #else + #define MY_offsetof(type, m) ((size_t)&(((type *)0)->m)) + #endif +#endif + + + +#ifndef MY_container_of + +/* +#define MY_container_of(ptr, type, m) container_of(ptr, type, m) +#define MY_container_of(ptr, type, m) CONTAINING_RECORD(ptr, type, m) +#define MY_container_of(ptr, type, m) ((type *)((char *)(ptr) - offsetof(type, m))) +#define MY_container_of(ptr, type, m) (&((type *)0)->m == (ptr), ((type *)(((char *)(ptr)) - MY_offsetof(type, m)))) +*/ + +/* + GCC shows warning: "perhaps the 'offsetof' macro was used incorrectly" + GCC 3.4.4 : classes with constructor + GCC 4.8.1 : classes with non-public variable members" +*/ + +#define MY_container_of(ptr, type, m) ((type *)(void *)((char *)(void *)(1 ? (ptr) : &((type *)0)->m) - MY_offsetof(type, m))) + +#endif + +#define CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) ((type *)(void *)(ptr)) + +/* +#define CONTAINER_FROM_VTBL(ptr, type, m) CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) +*/ +#define CONTAINER_FROM_VTBL(ptr, type, m) MY_container_of(ptr, type, m) + +#define CONTAINER_FROM_VTBL_CLS(ptr, type, m) CONTAINER_FROM_VTBL_SIMPLE(ptr, type, m) +/* +#define CONTAINER_FROM_VTBL_CLS(ptr, type, m) CONTAINER_FROM_VTBL(ptr, type, m) +*/ + + +#define MY_memset_0_ARRAY(a) memset((a), 0, sizeof(a)) + +#ifdef _WIN32 + +#define CHAR_PATH_SEPARATOR '\\' +#define WCHAR_PATH_SEPARATOR L'\\' +#define STRING_PATH_SEPARATOR "\\" +#define WSTRING_PATH_SEPARATOR L"\\" + +#else + +#define CHAR_PATH_SEPARATOR '/' +#define WCHAR_PATH_SEPARATOR L'/' +#define STRING_PATH_SEPARATOR "/" +#define WSTRING_PATH_SEPARATOR L"/" + +#endif + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/7zVersion.h b/3rdparty/lzma/include/7zVersion.h new file mode 100644 index 0000000000..e9363d37be --- /dev/null +++ b/3rdparty/lzma/include/7zVersion.h @@ -0,0 +1,27 @@ +#define MY_VER_MAJOR 21 +#define MY_VER_MINOR 07 +#define MY_VER_BUILD 0 +#define MY_VERSION_NUMBERS "21.07" +#define MY_VERSION MY_VERSION_NUMBERS + +#ifdef MY_CPU_NAME + #define MY_VERSION_CPU MY_VERSION " (" MY_CPU_NAME ")" +#else + #define MY_VERSION_CPU MY_VERSION +#endif + +#define MY_DATE "2021-12-26" +#undef MY_COPYRIGHT +#undef MY_VERSION_COPYRIGHT_DATE +#define MY_AUTHOR_NAME "Igor Pavlov" +#define MY_COPYRIGHT_PD "Igor Pavlov : Public domain" +#define MY_COPYRIGHT_CR "Copyright (c) 1999-2021 Igor Pavlov" + +#ifdef USE_COPYRIGHT_CR + #define MY_COPYRIGHT MY_COPYRIGHT_CR +#else + #define MY_COPYRIGHT MY_COPYRIGHT_PD +#endif + +#define MY_COPYRIGHT_DATE MY_COPYRIGHT " : " MY_DATE +#define MY_VERSION_COPYRIGHT_DATE MY_VERSION_CPU " : " MY_COPYRIGHT " : " MY_DATE diff --git a/3rdparty/lzma/include/Alloc.h b/3rdparty/lzma/include/Alloc.h new file mode 100644 index 0000000000..3be2041eb7 --- /dev/null +++ b/3rdparty/lzma/include/Alloc.h @@ -0,0 +1,58 @@ +/* Alloc.h -- Memory allocation functions +2021-07-13 : Igor Pavlov : Public domain */ + +#ifndef __COMMON_ALLOC_H +#define __COMMON_ALLOC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +void *MyAlloc(size_t size); +void MyFree(void *address); + +#ifdef _WIN32 + +void SetLargePageSize(void); + +void *MidAlloc(size_t size); +void MidFree(void *address); +void *BigAlloc(size_t size); +void BigFree(void *address); + +#else + +#define MidAlloc(size) MyAlloc(size) +#define MidFree(address) MyFree(address) +#define BigAlloc(size) MyAlloc(size) +#define BigFree(address) MyFree(address) + +#endif + +extern const ISzAlloc g_Alloc; + +#ifdef _WIN32 +extern const ISzAlloc g_BigAlloc; +extern const ISzAlloc g_MidAlloc; +#else +#define g_BigAlloc g_AlignedAlloc +#define g_MidAlloc g_AlignedAlloc +#endif + +extern const ISzAlloc g_AlignedAlloc; + + +typedef struct +{ + ISzAlloc vt; + ISzAllocPtr baseAlloc; + unsigned numAlignBits; /* ((1 << numAlignBits) >= sizeof(void *)) */ + size_t offset; /* (offset == (k * sizeof(void *)) && offset < (1 << numAlignBits) */ +} CAlignOffsetAlloc; + +void AlignOffsetAlloc_CreateVTable(CAlignOffsetAlloc *p); + + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Bcj2.h b/3rdparty/lzma/include/Bcj2.h new file mode 100644 index 0000000000..8824080acf --- /dev/null +++ b/3rdparty/lzma/include/Bcj2.h @@ -0,0 +1,146 @@ +/* Bcj2.h -- BCJ2 Converter for x86 code +2014-11-10 : Igor Pavlov : Public domain */ + +#ifndef __BCJ2_H +#define __BCJ2_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define BCJ2_NUM_STREAMS 4 + +enum +{ + BCJ2_STREAM_MAIN, + BCJ2_STREAM_CALL, + BCJ2_STREAM_JUMP, + BCJ2_STREAM_RC +}; + +enum +{ + BCJ2_DEC_STATE_ORIG_0 = BCJ2_NUM_STREAMS, + BCJ2_DEC_STATE_ORIG_1, + BCJ2_DEC_STATE_ORIG_2, + BCJ2_DEC_STATE_ORIG_3, + + BCJ2_DEC_STATE_ORIG, + BCJ2_DEC_STATE_OK +}; + +enum +{ + BCJ2_ENC_STATE_ORIG = BCJ2_NUM_STREAMS, + BCJ2_ENC_STATE_OK +}; + + +#define BCJ2_IS_32BIT_STREAM(s) ((s) == BCJ2_STREAM_CALL || (s) == BCJ2_STREAM_JUMP) + +/* +CBcj2Dec / CBcj2Enc +bufs sizes: + BUF_SIZE(n) = lims[n] - bufs[n] +bufs sizes for BCJ2_STREAM_CALL and BCJ2_STREAM_JUMP must be mutliply of 4: + (BUF_SIZE(BCJ2_STREAM_CALL) & 3) == 0 + (BUF_SIZE(BCJ2_STREAM_JUMP) & 3) == 0 +*/ + +/* +CBcj2Dec: +dest is allowed to overlap with bufs[BCJ2_STREAM_MAIN], with the following conditions: + bufs[BCJ2_STREAM_MAIN] >= dest && + bufs[BCJ2_STREAM_MAIN] - dest >= tempReserv + + BUF_SIZE(BCJ2_STREAM_CALL) + + BUF_SIZE(BCJ2_STREAM_JUMP) + tempReserv = 0 : for first call of Bcj2Dec_Decode + tempReserv = 4 : for any other calls of Bcj2Dec_Decode + overlap with offset = 1 is not allowed +*/ + +typedef struct +{ + const Byte *bufs[BCJ2_NUM_STREAMS]; + const Byte *lims[BCJ2_NUM_STREAMS]; + Byte *dest; + const Byte *destLim; + + unsigned state; /* BCJ2_STREAM_MAIN has more priority than BCJ2_STATE_ORIG */ + + UInt32 ip; + Byte temp[4]; + UInt32 range; + UInt32 code; + UInt16 probs[2 + 256]; +} CBcj2Dec; + +void Bcj2Dec_Init(CBcj2Dec *p); + +/* Returns: SZ_OK or SZ_ERROR_DATA */ +SRes Bcj2Dec_Decode(CBcj2Dec *p); + +#define Bcj2Dec_IsFinished(_p_) ((_p_)->code == 0) + + + +typedef enum +{ + BCJ2_ENC_FINISH_MODE_CONTINUE, + BCJ2_ENC_FINISH_MODE_END_BLOCK, + BCJ2_ENC_FINISH_MODE_END_STREAM +} EBcj2Enc_FinishMode; + +typedef struct +{ + Byte *bufs[BCJ2_NUM_STREAMS]; + const Byte *lims[BCJ2_NUM_STREAMS]; + const Byte *src; + const Byte *srcLim; + + unsigned state; + EBcj2Enc_FinishMode finishMode; + + Byte prevByte; + + Byte cache; + UInt32 range; + UInt64 low; + UInt64 cacheSize; + + UInt32 ip; + + /* 32-bit ralative offset in JUMP/CALL commands is + - (mod 4 GB) in 32-bit mode + - signed Int32 in 64-bit mode + We use (mod 4 GB) check for fileSize. + Use fileSize up to 2 GB, if you want to support 32-bit and 64-bit code conversion. */ + UInt32 fileIp; + UInt32 fileSize; /* (fileSize <= ((UInt32)1 << 31)), 0 means no_limit */ + UInt32 relatLimit; /* (relatLimit <= ((UInt32)1 << 31)), 0 means desable_conversion */ + + UInt32 tempTarget; + unsigned tempPos; + Byte temp[4 * 2]; + + unsigned flushPos; + + UInt16 probs[2 + 256]; +} CBcj2Enc; + +void Bcj2Enc_Init(CBcj2Enc *p); +void Bcj2Enc_Encode(CBcj2Enc *p); + +#define Bcj2Enc_Get_InputData_Size(p) ((SizeT)((p)->srcLim - (p)->src) + (p)->tempPos) +#define Bcj2Enc_IsFinished(p) ((p)->flushPos == 5) + + +#define BCJ2_RELAT_LIMIT_NUM_BITS 26 +#define BCJ2_RELAT_LIMIT ((UInt32)1 << BCJ2_RELAT_LIMIT_NUM_BITS) + +/* limit for CBcj2Enc::fileSize variable */ +#define BCJ2_FileSize_MAX ((UInt32)1 << 31) + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Bra.h b/3rdparty/lzma/include/Bra.h new file mode 100644 index 0000000000..855e37a6b5 --- /dev/null +++ b/3rdparty/lzma/include/Bra.h @@ -0,0 +1,64 @@ +/* Bra.h -- Branch converters for executables +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __BRA_H +#define __BRA_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* +These functions convert relative addresses to absolute addresses +in CALL instructions to increase the compression ratio. + + In: + data - data buffer + size - size of data + ip - current virtual Instruction Pinter (IP) value + state - state variable for x86 converter + encoding - 0 (for decoding), 1 (for encoding) + + Out: + state - state variable for x86 converter + + Returns: + The number of processed bytes. If you call these functions with multiple calls, + you must start next call with first byte after block of processed bytes. + + Type Endian Alignment LookAhead + + x86 little 1 4 + ARMT little 2 2 + ARM little 4 0 + PPC big 4 0 + SPARC big 4 0 + IA64 little 16 0 + + size must be >= Alignment + LookAhead, if it's not last block. + If (size < Alignment + LookAhead), converter returns 0. + + Example: + + UInt32 ip = 0; + for () + { + ; size must be >= Alignment + LookAhead, if it's not last block + SizeT processed = Convert(data, size, ip, 1); + data += processed; + size -= processed; + ip += processed; + } +*/ + +#define x86_Convert_Init(state) { state = 0; } +SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding); +SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT PPC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT SPARC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); +SizeT IA64_Convert(Byte *data, SizeT size, UInt32 ip, int encoding); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Compiler.h b/3rdparty/lzma/include/Compiler.h new file mode 100644 index 0000000000..a9816fa5ad --- /dev/null +++ b/3rdparty/lzma/include/Compiler.h @@ -0,0 +1,43 @@ +/* Compiler.h +2021-01-05 : Igor Pavlov : Public domain */ + +#ifndef __7Z_COMPILER_H +#define __7Z_COMPILER_H + + #ifdef __clang__ + #pragma clang diagnostic ignored "-Wunused-private-field" + #endif + +#ifdef _MSC_VER + + #ifdef UNDER_CE + #define RPC_NO_WINDOWS_H + /* #pragma warning(disable : 4115) // '_RPC_ASYNC_STATE' : named type definition in parentheses */ + #pragma warning(disable : 4201) // nonstandard extension used : nameless struct/union + #pragma warning(disable : 4214) // nonstandard extension used : bit field types other than int + #endif + + #if _MSC_VER >= 1300 + #pragma warning(disable : 4996) // This function or variable may be unsafe + #else + #pragma warning(disable : 4511) // copy constructor could not be generated + #pragma warning(disable : 4512) // assignment operator could not be generated + #pragma warning(disable : 4514) // unreferenced inline function has been removed + #pragma warning(disable : 4702) // unreachable code + #pragma warning(disable : 4710) // not inlined + #pragma warning(disable : 4714) // function marked as __forceinline not inlined + #pragma warning(disable : 4786) // identifier was truncated to '255' characters in the debug information + #endif + + #ifdef __clang__ + #pragma clang diagnostic ignored "-Wdeprecated-declarations" + #pragma clang diagnostic ignored "-Wmicrosoft-exception-spec" + // #pragma clang diagnostic ignored "-Wreserved-id-macro" + #endif + +#endif + +#define UNUSED_VAR(x) (void)x; +/* #define UNUSED_VAR(x) x=x; */ + +#endif diff --git a/3rdparty/lzma/include/CpuArch.h b/3rdparty/lzma/include/CpuArch.h new file mode 100644 index 0000000000..529d3a502b --- /dev/null +++ b/3rdparty/lzma/include/CpuArch.h @@ -0,0 +1,442 @@ +/* CpuArch.h -- CPU specific code +2021-07-13 : Igor Pavlov : Public domain */ + +#ifndef __CPU_ARCH_H +#define __CPU_ARCH_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* +MY_CPU_LE means that CPU is LITTLE ENDIAN. +MY_CPU_BE means that CPU is BIG ENDIAN. +If MY_CPU_LE and MY_CPU_BE are not defined, we don't know about ENDIANNESS of platform. + +MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned memory accesses. + +MY_CPU_64BIT means that processor can work with 64-bit registers. + MY_CPU_64BIT can be used to select fast code branch + MY_CPU_64BIT doesn't mean that (sizeof(void *) == 8) +*/ + +#if defined(_M_X64) \ + || defined(_M_AMD64) \ + || defined(__x86_64__) \ + || defined(__AMD64__) \ + || defined(__amd64__) + #define MY_CPU_AMD64 + #ifdef __ILP32__ + #define MY_CPU_NAME "x32" + #define MY_CPU_SIZEOF_POINTER 4 + #else + #define MY_CPU_NAME "x64" + #define MY_CPU_SIZEOF_POINTER 8 + #endif + #define MY_CPU_64BIT +#endif + + +#if defined(_M_IX86) \ + || defined(__i386__) + #define MY_CPU_X86 + #define MY_CPU_NAME "x86" + /* #define MY_CPU_32BIT */ + #define MY_CPU_SIZEOF_POINTER 4 +#endif + + +#if defined(_M_ARM64) \ + || defined(__AARCH64EL__) \ + || defined(__AARCH64EB__) \ + || defined(__aarch64__) + #define MY_CPU_ARM64 + #define MY_CPU_NAME "arm64" + #define MY_CPU_64BIT +#endif + + +#if defined(_M_ARM) \ + || defined(_M_ARM_NT) \ + || defined(_M_ARMT) \ + || defined(__arm__) \ + || defined(__thumb__) \ + || defined(__ARMEL__) \ + || defined(__ARMEB__) \ + || defined(__THUMBEL__) \ + || defined(__THUMBEB__) + #define MY_CPU_ARM + + #if defined(__thumb__) || defined(__THUMBEL__) || defined(_M_ARMT) + #define MY_CPU_NAME "armt" + #else + #define MY_CPU_NAME "arm" + #endif + /* #define MY_CPU_32BIT */ + #define MY_CPU_SIZEOF_POINTER 4 +#endif + + +#if defined(_M_IA64) \ + || defined(__ia64__) + #define MY_CPU_IA64 + #define MY_CPU_NAME "ia64" + #define MY_CPU_64BIT +#endif + + +#if defined(__mips64) \ + || defined(__mips64__) \ + || (defined(__mips) && (__mips == 64 || __mips == 4 || __mips == 3)) + #define MY_CPU_NAME "mips64" + #define MY_CPU_64BIT +#elif defined(__mips__) + #define MY_CPU_NAME "mips" + /* #define MY_CPU_32BIT */ +#endif + + +#if defined(__ppc64__) \ + || defined(__powerpc64__) \ + || defined(__ppc__) \ + || defined(__powerpc__) \ + || defined(__PPC__) \ + || defined(_POWER) + +#if defined(__ppc64__) \ + || defined(__powerpc64__) \ + || defined(_LP64) \ + || defined(__64BIT__) + #ifdef __ILP32__ + #define MY_CPU_NAME "ppc64-32" + #define MY_CPU_SIZEOF_POINTER 4 + #else + #define MY_CPU_NAME "ppc64" + #define MY_CPU_SIZEOF_POINTER 8 + #endif + #define MY_CPU_64BIT +#else + #define MY_CPU_NAME "ppc" + #define MY_CPU_SIZEOF_POINTER 4 + /* #define MY_CPU_32BIT */ +#endif +#endif + + +#if defined(__sparc64__) + #define MY_CPU_NAME "sparc64" + #define MY_CPU_64BIT +#elif defined(__sparc__) + #define MY_CPU_NAME "sparc" + /* #define MY_CPU_32BIT */ +#endif + + +#if defined(MY_CPU_X86) || defined(MY_CPU_AMD64) +#define MY_CPU_X86_OR_AMD64 +#endif + +#if defined(MY_CPU_ARM) || defined(MY_CPU_ARM64) +#define MY_CPU_ARM_OR_ARM64 +#endif + + +#ifdef _WIN32 + + #ifdef MY_CPU_ARM + #define MY_CPU_ARM_LE + #endif + + #ifdef MY_CPU_ARM64 + #define MY_CPU_ARM64_LE + #endif + + #ifdef _M_IA64 + #define MY_CPU_IA64_LE + #endif + +#endif + + +#if defined(MY_CPU_X86_OR_AMD64) \ + || defined(MY_CPU_ARM_LE) \ + || defined(MY_CPU_ARM64_LE) \ + || defined(MY_CPU_IA64_LE) \ + || defined(__LITTLE_ENDIAN__) \ + || defined(__ARMEL__) \ + || defined(__THUMBEL__) \ + || defined(__AARCH64EL__) \ + || defined(__MIPSEL__) \ + || defined(__MIPSEL) \ + || defined(_MIPSEL) \ + || defined(__BFIN__) \ + || (defined(__BYTE_ORDER__) && (__BYTE_ORDER__ == __ORDER_LITTLE_ENDIAN__)) + #define MY_CPU_LE +#endif + +#if defined(__BIG_ENDIAN__) \ + || defined(__ARMEB__) \ + || defined(__THUMBEB__) \ + || defined(__AARCH64EB__) \ + || defined(__MIPSEB__) \ + || defined(__MIPSEB) \ + || defined(_MIPSEB) \ + || defined(__m68k__) \ + || defined(__s390__) \ + || defined(__s390x__) \ + || defined(__zarch__) \ + || (defined(__BYTE_ORDER__) && (__BYTE_ORDER__ == __ORDER_BIG_ENDIAN__)) + #define MY_CPU_BE +#endif + + +#if defined(MY_CPU_LE) && defined(MY_CPU_BE) + #error Stop_Compiling_Bad_Endian +#endif + + +#if defined(MY_CPU_32BIT) && defined(MY_CPU_64BIT) + #error Stop_Compiling_Bad_32_64_BIT +#endif + +#ifdef __SIZEOF_POINTER__ + #ifdef MY_CPU_SIZEOF_POINTER + #if MY_CPU_SIZEOF_POINTER != __SIZEOF_POINTER__ + #error Stop_Compiling_Bad_MY_CPU_PTR_SIZE + #endif + #else + #define MY_CPU_SIZEOF_POINTER __SIZEOF_POINTER__ + #endif +#endif + +#if defined(MY_CPU_SIZEOF_POINTER) && (MY_CPU_SIZEOF_POINTER == 4) +#if defined (_LP64) + #error Stop_Compiling_Bad_MY_CPU_PTR_SIZE +#endif +#endif + +#ifdef _MSC_VER + #if _MSC_VER >= 1300 + #define MY_CPU_pragma_pack_push_1 __pragma(pack(push, 1)) + #define MY_CPU_pragma_pop __pragma(pack(pop)) + #else + #define MY_CPU_pragma_pack_push_1 + #define MY_CPU_pragma_pop + #endif +#else + #ifdef __xlC__ + #define MY_CPU_pragma_pack_push_1 _Pragma("pack(1)") + #define MY_CPU_pragma_pop _Pragma("pack()") + #else + #define MY_CPU_pragma_pack_push_1 _Pragma("pack(push, 1)") + #define MY_CPU_pragma_pop _Pragma("pack(pop)") + #endif +#endif + + +#ifndef MY_CPU_NAME + #ifdef MY_CPU_LE + #define MY_CPU_NAME "LE" + #elif defined(MY_CPU_BE) + #define MY_CPU_NAME "BE" + #else + /* + #define MY_CPU_NAME "" + */ + #endif +#endif + + + + + +#ifdef MY_CPU_LE + #if defined(MY_CPU_X86_OR_AMD64) \ + || defined(MY_CPU_ARM64) + #define MY_CPU_LE_UNALIGN + #define MY_CPU_LE_UNALIGN_64 + #elif defined(__ARM_FEATURE_UNALIGNED) + /* gcc9 for 32-bit arm can use LDRD instruction that requires 32-bit alignment. + So we can't use unaligned 64-bit operations. */ + #define MY_CPU_LE_UNALIGN + #endif +#endif + + +#ifdef MY_CPU_LE_UNALIGN + +#define GetUi16(p) (*(const UInt16 *)(const void *)(p)) +#define GetUi32(p) (*(const UInt32 *)(const void *)(p)) +#ifdef MY_CPU_LE_UNALIGN_64 +#define GetUi64(p) (*(const UInt64 *)(const void *)(p)) +#endif + +#define SetUi16(p, v) { *(UInt16 *)(void *)(p) = (v); } +#define SetUi32(p, v) { *(UInt32 *)(void *)(p) = (v); } +#ifdef MY_CPU_LE_UNALIGN_64 +#define SetUi64(p, v) { *(UInt64 *)(void *)(p) = (v); } +#endif + +#else + +#define GetUi16(p) ( (UInt16) ( \ + ((const Byte *)(p))[0] | \ + ((UInt16)((const Byte *)(p))[1] << 8) )) + +#define GetUi32(p) ( \ + ((const Byte *)(p))[0] | \ + ((UInt32)((const Byte *)(p))[1] << 8) | \ + ((UInt32)((const Byte *)(p))[2] << 16) | \ + ((UInt32)((const Byte *)(p))[3] << 24)) + +#define SetUi16(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ + _ppp_[0] = (Byte)_vvv_; \ + _ppp_[1] = (Byte)(_vvv_ >> 8); } + +#define SetUi32(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ + _ppp_[0] = (Byte)_vvv_; \ + _ppp_[1] = (Byte)(_vvv_ >> 8); \ + _ppp_[2] = (Byte)(_vvv_ >> 16); \ + _ppp_[3] = (Byte)(_vvv_ >> 24); } + +#endif + + +#ifndef MY_CPU_LE_UNALIGN_64 + +#define GetUi64(p) (GetUi32(p) | ((UInt64)GetUi32(((const Byte *)(p)) + 4) << 32)) + +#define SetUi64(p, v) { Byte *_ppp2_ = (Byte *)(p); UInt64 _vvv2_ = (v); \ + SetUi32(_ppp2_ , (UInt32)_vvv2_); \ + SetUi32(_ppp2_ + 4, (UInt32)(_vvv2_ >> 32)); } + +#endif + + + + +#ifdef __has_builtin + #define MY__has_builtin(x) __has_builtin(x) +#else + #define MY__has_builtin(x) 0 +#endif + +#if defined(MY_CPU_LE_UNALIGN) && /* defined(_WIN64) && */ defined(_MSC_VER) && (_MSC_VER >= 1300) + +/* Note: we use bswap instruction, that is unsupported in 386 cpu */ + +#include + +#pragma intrinsic(_byteswap_ushort) +#pragma intrinsic(_byteswap_ulong) +#pragma intrinsic(_byteswap_uint64) + +/* #define GetBe16(p) _byteswap_ushort(*(const UInt16 *)(const Byte *)(p)) */ +#define GetBe32(p) _byteswap_ulong (*(const UInt32 *)(const void *)(p)) +#define GetBe64(p) _byteswap_uint64(*(const UInt64 *)(const void *)(p)) + +#define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = _byteswap_ulong(v) + +#elif defined(MY_CPU_LE_UNALIGN) && ( \ + (defined(__GNUC__) && (__GNUC__ > 4 || (__GNUC__ == 4 && __GNUC_MINOR__ >= 3))) \ + || (defined(__clang__) && MY__has_builtin(__builtin_bswap16)) ) + +/* #define GetBe16(p) __builtin_bswap16(*(const UInt16 *)(const void *)(p)) */ +#define GetBe32(p) __builtin_bswap32(*(const UInt32 *)(const void *)(p)) +#define GetBe64(p) __builtin_bswap64(*(const UInt64 *)(const void *)(p)) + +#define SetBe32(p, v) (*(UInt32 *)(void *)(p)) = __builtin_bswap32(v) + +#else + +#define GetBe32(p) ( \ + ((UInt32)((const Byte *)(p))[0] << 24) | \ + ((UInt32)((const Byte *)(p))[1] << 16) | \ + ((UInt32)((const Byte *)(p))[2] << 8) | \ + ((const Byte *)(p))[3] ) + +#define GetBe64(p) (((UInt64)GetBe32(p) << 32) | GetBe32(((const Byte *)(p)) + 4)) + +#define SetBe32(p, v) { Byte *_ppp_ = (Byte *)(p); UInt32 _vvv_ = (v); \ + _ppp_[0] = (Byte)(_vvv_ >> 24); \ + _ppp_[1] = (Byte)(_vvv_ >> 16); \ + _ppp_[2] = (Byte)(_vvv_ >> 8); \ + _ppp_[3] = (Byte)_vvv_; } + +#endif + + +#ifndef GetBe16 + +#define GetBe16(p) ( (UInt16) ( \ + ((UInt16)((const Byte *)(p))[0] << 8) | \ + ((const Byte *)(p))[1] )) + +#endif + + + +#ifdef MY_CPU_X86_OR_AMD64 + +typedef struct +{ + UInt32 maxFunc; + UInt32 vendor[3]; + UInt32 ver; + UInt32 b; + UInt32 c; + UInt32 d; +} Cx86cpuid; + +enum +{ + CPU_FIRM_INTEL, + CPU_FIRM_AMD, + CPU_FIRM_VIA +}; + +void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d); + +BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p); +int x86cpuid_GetFirm(const Cx86cpuid *p); + +#define x86cpuid_GetFamily(ver) (((ver >> 16) & 0xFF0) | ((ver >> 8) & 0xF)) +#define x86cpuid_GetModel(ver) (((ver >> 12) & 0xF0) | ((ver >> 4) & 0xF)) +#define x86cpuid_GetStepping(ver) (ver & 0xF) + +BoolInt CPU_Is_InOrder(void); + +BoolInt CPU_IsSupported_AES(void); +BoolInt CPU_IsSupported_AVX2(void); +BoolInt CPU_IsSupported_VAES_AVX2(void); +BoolInt CPU_IsSupported_SSSE3(void); +BoolInt CPU_IsSupported_SSE41(void); +BoolInt CPU_IsSupported_SHA(void); +BoolInt CPU_IsSupported_PageGB(void); + +#elif defined(MY_CPU_ARM_OR_ARM64) + +BoolInt CPU_IsSupported_CRC32(void); +BoolInt CPU_IsSupported_NEON(void); + +#if defined(_WIN32) +BoolInt CPU_IsSupported_CRYPTO(void); +#define CPU_IsSupported_SHA1 CPU_IsSupported_CRYPTO +#define CPU_IsSupported_SHA2 CPU_IsSupported_CRYPTO +#define CPU_IsSupported_AES CPU_IsSupported_CRYPTO +#else +BoolInt CPU_IsSupported_SHA1(void); +BoolInt CPU_IsSupported_SHA2(void); +BoolInt CPU_IsSupported_AES(void); +#endif + +#endif + +#if defined(__APPLE__) +int My_sysctlbyname_Get(const char *name, void *buf, size_t *bufSize); +int My_sysctlbyname_Get_UInt32(const char *name, UInt32 *val); +#endif + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Delta.h b/3rdparty/lzma/include/Delta.h new file mode 100644 index 0000000000..2fa54ad67b --- /dev/null +++ b/3rdparty/lzma/include/Delta.h @@ -0,0 +1,19 @@ +/* Delta.h -- Delta converter +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __DELTA_H +#define __DELTA_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define DELTA_STATE_SIZE 256 + +void Delta_Init(Byte *state); +void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size); +void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/LzFind.h b/3rdparty/lzma/include/LzFind.h new file mode 100644 index 0000000000..eea873ff61 --- /dev/null +++ b/3rdparty/lzma/include/LzFind.h @@ -0,0 +1,136 @@ +/* LzFind.h -- Match finder for LZ algorithms +2021-07-13 : Igor Pavlov : Public domain */ + +#ifndef __LZ_FIND_H +#define __LZ_FIND_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +typedef UInt32 CLzRef; + +typedef struct _CMatchFinder +{ + Byte *buffer; + UInt32 pos; + UInt32 posLimit; + UInt32 streamPos; /* wrap over Zero is allowed (streamPos < pos). Use (UInt32)(streamPos - pos) */ + UInt32 lenLimit; + + UInt32 cyclicBufferPos; + UInt32 cyclicBufferSize; /* it must be = (historySize + 1) */ + + Byte streamEndWasReached; + Byte btMode; + Byte bigHash; + Byte directInput; + + UInt32 matchMaxLen; + CLzRef *hash; + CLzRef *son; + UInt32 hashMask; + UInt32 cutValue; + + Byte *bufferBase; + ISeqInStream *stream; + + UInt32 blockSize; + UInt32 keepSizeBefore; + UInt32 keepSizeAfter; + + UInt32 numHashBytes; + size_t directInputRem; + UInt32 historySize; + UInt32 fixedHashSize; + UInt32 hashSizeSum; + SRes result; + UInt32 crc[256]; + size_t numRefs; + + UInt64 expectedDataSize; +} CMatchFinder; + +#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((const Byte *)(p)->buffer) + +#define Inline_MatchFinder_GetNumAvailableBytes(p) ((UInt32)((p)->streamPos - (p)->pos)) + +/* +#define Inline_MatchFinder_IsFinishedOK(p) \ + ((p)->streamEndWasReached \ + && (p)->streamPos == (p)->pos \ + && (!(p)->directInput || (p)->directInputRem == 0)) +*/ + +int MatchFinder_NeedMove(CMatchFinder *p); +/* Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p); */ +void MatchFinder_MoveBlock(CMatchFinder *p); +void MatchFinder_ReadIfRequired(CMatchFinder *p); + +void MatchFinder_Construct(CMatchFinder *p); + +/* Conditions: + historySize <= 3 GB + keepAddBufferBefore + matchMaxLen + keepAddBufferAfter < 511MB +*/ +int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, + UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, + ISzAllocPtr alloc); +void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc); +void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems); +// void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue); + +/* +#define Inline_MatchFinder_InitPos(p, val) \ + (p)->pos = (val); \ + (p)->streamPos = (val); +*/ + +#define Inline_MatchFinder_ReduceOffsets(p, subValue) \ + (p)->pos -= (subValue); \ + (p)->streamPos -= (subValue); + + +UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *buffer, CLzRef *son, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue, + UInt32 *distances, UInt32 maxLen); + +/* +Conditions: + Mf_GetNumAvailableBytes_Func must be called before each Mf_GetMatchLen_Func. + Mf_GetPointerToCurrentPos_Func's result must be used only before any other function +*/ + +typedef void (*Mf_Init_Func)(void *object); +typedef UInt32 (*Mf_GetNumAvailableBytes_Func)(void *object); +typedef const Byte * (*Mf_GetPointerToCurrentPos_Func)(void *object); +typedef UInt32 * (*Mf_GetMatches_Func)(void *object, UInt32 *distances); +typedef void (*Mf_Skip_Func)(void *object, UInt32); + +typedef struct _IMatchFinder +{ + Mf_Init_Func Init; + Mf_GetNumAvailableBytes_Func GetNumAvailableBytes; + Mf_GetPointerToCurrentPos_Func GetPointerToCurrentPos; + Mf_GetMatches_Func GetMatches; + Mf_Skip_Func Skip; +} IMatchFinder2; + +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder2 *vTable); + +void MatchFinder_Init_LowHash(CMatchFinder *p); +void MatchFinder_Init_HighHash(CMatchFinder *p); +void MatchFinder_Init_4(CMatchFinder *p); +void MatchFinder_Init(CMatchFinder *p); + +UInt32* Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); +UInt32* Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances); + +void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); +void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num); + +void LzFindPrepare(void); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/LzHash.h b/3rdparty/lzma/include/LzHash.h new file mode 100644 index 0000000000..77b898cfab --- /dev/null +++ b/3rdparty/lzma/include/LzHash.h @@ -0,0 +1,34 @@ +/* LzHash.h -- HASH functions for LZ algorithms +2019-10-30 : Igor Pavlov : Public domain */ + +#ifndef __LZ_HASH_H +#define __LZ_HASH_H + +/* + (kHash2Size >= (1 << 8)) : Required + (kHash3Size >= (1 << 16)) : Required +*/ + +#define kHash2Size (1 << 10) +#define kHash3Size (1 << 16) +// #define kHash4Size (1 << 20) + +#define kFix3HashSize (kHash2Size) +#define kFix4HashSize (kHash2Size + kHash3Size) +// #define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size) + +/* + We use up to 3 crc values for hash: + crc0 + crc1 << Shift_1 + crc2 << Shift_2 + (Shift_1 = 5) and (Shift_2 = 10) is good tradeoff. + Small values for Shift are not good for collision rate. + Big value for Shift_2 increases the minimum size + of hash table, that will be slow for small files. +*/ + +#define kLzHash_CrcShift_1 5 +#define kLzHash_CrcShift_2 10 + +#endif diff --git a/3rdparty/lzma/include/Lzma2Dec.h b/3rdparty/lzma/include/Lzma2Dec.h new file mode 100644 index 0000000000..b8ddeac890 --- /dev/null +++ b/3rdparty/lzma/include/Lzma2Dec.h @@ -0,0 +1,120 @@ +/* Lzma2Dec.h -- LZMA2 Decoder +2018-02-19 : Igor Pavlov : Public domain */ + +#ifndef __LZMA2_DEC_H +#define __LZMA2_DEC_H + +#include "LzmaDec.h" + +EXTERN_C_BEGIN + +/* ---------- State Interface ---------- */ + +typedef struct +{ + unsigned state; + Byte control; + Byte needInitLevel; + Byte isExtraMode; + Byte _pad_; + UInt32 packSize; + UInt32 unpackSize; + CLzmaDec decoder; +} CLzma2Dec; + +#define Lzma2Dec_Construct(p) LzmaDec_Construct(&(p)->decoder) +#define Lzma2Dec_FreeProbs(p, alloc) LzmaDec_FreeProbs(&(p)->decoder, alloc) +#define Lzma2Dec_Free(p, alloc) LzmaDec_Free(&(p)->decoder, alloc) + +SRes Lzma2Dec_AllocateProbs(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc); +SRes Lzma2Dec_Allocate(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc); +void Lzma2Dec_Init(CLzma2Dec *p); + +/* +finishMode: + It has meaning only if the decoding reaches output limit (*destLen or dicLimit). + LZMA_FINISH_ANY - use smallest number of input bytes + LZMA_FINISH_END - read EndOfStream marker after decoding + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_NEEDS_MORE_INPUT + SZ_ERROR_DATA - Data error +*/ + +SRes Lzma2Dec_DecodeToDic(CLzma2Dec *p, SizeT dicLimit, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + +SRes Lzma2Dec_DecodeToBuf(CLzma2Dec *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + + +/* ---------- LZMA2 block and chunk parsing ---------- */ + +/* +Lzma2Dec_Parse() parses compressed data stream up to next independent block or next chunk data. +It can return LZMA_STATUS_* code or LZMA2_PARSE_STATUS_* code: + - LZMA2_PARSE_STATUS_NEW_BLOCK - there is new block, and 1 additional byte (control byte of next block header) was read from input. + - LZMA2_PARSE_STATUS_NEW_CHUNK - there is new chunk, and only lzma2 header of new chunk was read. + CLzma2Dec::unpackSize contains unpack size of that chunk +*/ + +typedef enum +{ +/* + LZMA_STATUS_NOT_SPECIFIED // data error + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED // + LZMA_STATUS_NEEDS_MORE_INPUT + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK // unused +*/ + LZMA2_PARSE_STATUS_NEW_BLOCK = LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + 1, + LZMA2_PARSE_STATUS_NEW_CHUNK +} ELzma2ParseStatus; + +ELzma2ParseStatus Lzma2Dec_Parse(CLzma2Dec *p, + SizeT outSize, // output size + const Byte *src, SizeT *srcLen, + int checkFinishBlock // set (checkFinishBlock = 1), if it must read full input data, if decoder.dicPos reaches blockMax position. + ); + +/* +LZMA2 parser doesn't decode LZMA chunks, so we must read + full input LZMA chunk to decode some part of LZMA chunk. + +Lzma2Dec_GetUnpackExtra() returns the value that shows + max possible number of output bytes that can be output by decoder + at current input positon. +*/ + +#define Lzma2Dec_GetUnpackExtra(p) ((p)->isExtraMode ? (p)->unpackSize : 0); + + +/* ---------- One Call Interface ---------- */ + +/* +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - use smallest number of input bytes + LZMA_FINISH_END - read EndOfStream marker after decoding + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). +*/ + +SRes Lzma2Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + Byte prop, ELzmaFinishMode finishMode, ELzmaStatus *status, ISzAllocPtr alloc); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Lzma2DecMt.h b/3rdparty/lzma/include/Lzma2DecMt.h new file mode 100644 index 0000000000..7791c310bc --- /dev/null +++ b/3rdparty/lzma/include/Lzma2DecMt.h @@ -0,0 +1,79 @@ +/* Lzma2DecMt.h -- LZMA2 Decoder Multi-thread +2018-02-17 : Igor Pavlov : Public domain */ + +#ifndef __LZMA2_DEC_MT_H +#define __LZMA2_DEC_MT_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +typedef struct +{ + size_t inBufSize_ST; + size_t outStep_ST; + + #ifndef _7ZIP_ST + unsigned numThreads; + size_t inBufSize_MT; + size_t outBlockMax; + size_t inBlockMax; + #endif +} CLzma2DecMtProps; + +/* init to single-thread mode */ +void Lzma2DecMtProps_Init(CLzma2DecMtProps *p); + + +/* ---------- CLzma2DecMtHandle Interface ---------- */ + +/* Lzma2DecMt_ * functions can return the following exit codes: +SRes: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater in props + SZ_ERROR_WRITE - ISeqOutStream write callback error + // SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output + SZ_ERROR_PROGRESS - some break from progress callback + SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) +*/ + +typedef void * CLzma2DecMtHandle; + +CLzma2DecMtHandle Lzma2DecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid); +void Lzma2DecMt_Destroy(CLzma2DecMtHandle p); + +SRes Lzma2DecMt_Decode(CLzma2DecMtHandle p, + Byte prop, + const CLzma2DecMtProps *props, + ISeqOutStream *outStream, + const UInt64 *outDataSize, // NULL means undefined + int finishMode, // 0 - partial unpacking is allowed, 1 - if lzma2 stream must be finished + // Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + // const Byte *inData, size_t inDataSize, + + // out variables: + UInt64 *inProcessed, + int *isMT, /* out: (*isMT == 0), if single thread decoding was used */ + + // UInt64 *outProcessed, + ICompressProgress *progress); + + +/* ---------- Read from CLzma2DecMtHandle Interface ---------- */ + +SRes Lzma2DecMt_Init(CLzma2DecMtHandle pp, + Byte prop, + const CLzma2DecMtProps *props, + const UInt64 *outDataSize, int finishMode, + ISeqInStream *inStream); + +SRes Lzma2DecMt_Read(CLzma2DecMtHandle pp, + Byte *data, size_t *outSize, + UInt64 *inStreamProcessed); + + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Lzma2Enc.h b/3rdparty/lzma/include/Lzma2Enc.h new file mode 100644 index 0000000000..6a6110ff7e --- /dev/null +++ b/3rdparty/lzma/include/Lzma2Enc.h @@ -0,0 +1,55 @@ +/* Lzma2Enc.h -- LZMA2 Encoder +2017-07-27 : Igor Pavlov : Public domain */ + +#ifndef __LZMA2_ENC_H +#define __LZMA2_ENC_H + +#include "LzmaEnc.h" + +EXTERN_C_BEGIN + +#define LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO 0 +#define LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID ((UInt64)(Int64)-1) + +typedef struct +{ + CLzmaEncProps lzmaProps; + UInt64 blockSize; + int numBlockThreads_Reduced; + int numBlockThreads_Max; + int numTotalThreads; +} CLzma2EncProps; + +void Lzma2EncProps_Init(CLzma2EncProps *p); +void Lzma2EncProps_Normalize(CLzma2EncProps *p); + +/* ---------- CLzmaEnc2Handle Interface ---------- */ + +/* Lzma2Enc_* functions can return the following exit codes: +SRes: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater in props + SZ_ERROR_WRITE - ISeqOutStream write callback error + SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output + SZ_ERROR_PROGRESS - some break from progress callback + SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) +*/ + +typedef void * CLzma2EncHandle; + +CLzma2EncHandle Lzma2Enc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig); +void Lzma2Enc_Destroy(CLzma2EncHandle p); +SRes Lzma2Enc_SetProps(CLzma2EncHandle p, const CLzma2EncProps *props); +void Lzma2Enc_SetDataSize(CLzma2EncHandle p, UInt64 expectedDataSiize); +Byte Lzma2Enc_WriteProperties(CLzma2EncHandle p); +SRes Lzma2Enc_Encode2(CLzma2EncHandle p, + ISeqOutStream *outStream, + Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + const Byte *inData, size_t inDataSize, + ICompressProgress *progress); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Lzma86.h b/3rdparty/lzma/include/Lzma86.h new file mode 100644 index 0000000000..bebed5cb76 --- /dev/null +++ b/3rdparty/lzma/include/Lzma86.h @@ -0,0 +1,111 @@ +/* Lzma86.h -- LZMA + x86 (BCJ) Filter +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __LZMA86_H +#define __LZMA86_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define LZMA86_SIZE_OFFSET (1 + 5) +#define LZMA86_HEADER_SIZE (LZMA86_SIZE_OFFSET + 8) + +/* +It's an example for LZMA + x86 Filter use. +You can use .lzma86 extension, if you write that stream to file. +.lzma86 header adds one additional byte to standard .lzma header. +.lzma86 header (14 bytes): + Offset Size Description + 0 1 = 0 - no filter, pure LZMA + = 1 - x86 filter + LZMA + 1 1 lc, lp and pb in encoded form + 2 4 dictSize (little endian) + 6 8 uncompressed size (little endian) + + +Lzma86_Encode +------------- +level - compression level: 0 <= level <= 9, the default value for "level" is 5. + +dictSize - The dictionary size in bytes. The maximum value is + 128 MB = (1 << 27) bytes for 32-bit version + 1 GB = (1 << 30) bytes for 64-bit version + The default value is 16 MB = (1 << 24) bytes, for level = 5. + It's recommended to use the dictionary that is larger than 4 KB and + that can be calculated as (1 << N) or (3 << N) sizes. + For better compression ratio dictSize must be >= inSize. + +filterMode: + SZ_FILTER_NO - no Filter + SZ_FILTER_YES - x86 Filter + SZ_FILTER_AUTO - it tries both alternatives to select best. + Encoder will use 2 or 3 passes: + 2 passes when FILTER_NO provides better compression. + 3 passes when FILTER_YES provides better compression. + +Lzma86Encode allocates Data with MyAlloc functions. +RAM Requirements for compressing: + RamSize = dictionarySize * 11.5 + 6MB + FilterBlockSize + filterMode FilterBlockSize + SZ_FILTER_NO 0 + SZ_FILTER_YES inSize + SZ_FILTER_AUTO inSize + + +Return code: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater + SZ_ERROR_OUTPUT_EOF - output buffer overflow + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) +*/ + +enum ESzFilterMode +{ + SZ_FILTER_NO, + SZ_FILTER_YES, + SZ_FILTER_AUTO +}; + +SRes Lzma86_Encode(Byte *dest, size_t *destLen, const Byte *src, size_t srcLen, + int level, UInt32 dictSize, int filterMode); + + +/* +Lzma86_GetUnpackSize: + In: + src - input data + srcLen - input data size + Out: + unpackSize - size of uncompressed stream + Return code: + SZ_OK - OK + SZ_ERROR_INPUT_EOF - Error in headers +*/ + +SRes Lzma86_GetUnpackSize(const Byte *src, SizeT srcLen, UInt64 *unpackSize); + +/* +Lzma86_Decode: + In: + dest - output data + destLen - output data size + src - input data + srcLen - input data size + Out: + destLen - processed output size + srcLen - processed input size + Return code: + SZ_OK - OK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - unsupported file + SZ_ERROR_INPUT_EOF - it needs more bytes in input buffer +*/ + +SRes Lzma86_Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/LzmaDec.h b/3rdparty/lzma/include/LzmaDec.h new file mode 100644 index 0000000000..6f1296250c --- /dev/null +++ b/3rdparty/lzma/include/LzmaDec.h @@ -0,0 +1,236 @@ +/* LzmaDec.h -- LZMA Decoder +2020-03-19 : Igor Pavlov : Public domain */ + +#ifndef __LZMA_DEC_H +#define __LZMA_DEC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* #define _LZMA_PROB32 */ +/* _LZMA_PROB32 can increase the speed on some CPUs, + but memory usage for CLzmaDec::probs will be doubled in that case */ + +typedef +#ifdef _LZMA_PROB32 + UInt32 +#else + UInt16 +#endif + CLzmaProb; + + +/* ---------- LZMA Properties ---------- */ + +#define LZMA_PROPS_SIZE 5 + +typedef struct _CLzmaProps +{ + Byte lc; + Byte lp; + Byte pb; + Byte _pad_; + UInt32 dicSize; +} CLzmaProps; + +/* LzmaProps_Decode - decodes properties +Returns: + SZ_OK + SZ_ERROR_UNSUPPORTED - Unsupported properties +*/ + +SRes LzmaProps_Decode(CLzmaProps *p, const Byte *data, unsigned size); + + +/* ---------- LZMA Decoder state ---------- */ + +/* LZMA_REQUIRED_INPUT_MAX = number of required input bytes for worst case. + Num bits = log2((2^11 / 31) ^ 22) + 26 < 134 + 26 = 160; */ + +#define LZMA_REQUIRED_INPUT_MAX 20 + +typedef struct +{ + /* Don't change this structure. ASM code can use it. */ + CLzmaProps prop; + CLzmaProb *probs; + CLzmaProb *probs_1664; + Byte *dic; + SizeT dicBufSize; + SizeT dicPos; + const Byte *buf; + UInt32 range; + UInt32 code; + UInt32 processedPos; + UInt32 checkDicSize; + UInt32 reps[4]; + UInt32 state; + UInt32 remainLen; + + UInt32 numProbs; + unsigned tempBufSize; + Byte tempBuf[LZMA_REQUIRED_INPUT_MAX]; +} CLzmaDec; + +#define LzmaDec_Construct(p) { (p)->dic = NULL; (p)->probs = NULL; } + +void LzmaDec_Init(CLzmaDec *p); + +/* There are two types of LZMA streams: + - Stream with end mark. That end mark adds about 6 bytes to compressed size. + - Stream without end mark. You must know exact uncompressed size to decompress such stream. */ + +typedef enum +{ + LZMA_FINISH_ANY, /* finish at any point */ + LZMA_FINISH_END /* block must be finished at the end */ +} ELzmaFinishMode; + +/* ELzmaFinishMode has meaning only if the decoding reaches output limit !!! + + You must use LZMA_FINISH_END, when you know that current output buffer + covers last bytes of block. In other cases you must use LZMA_FINISH_ANY. + + If LZMA decoder sees end marker before reaching output limit, it returns SZ_OK, + and output value of destLen will be less than output buffer size limit. + You can check status result also. + + You can use multiple checks to test data integrity after full decompression: + 1) Check Result and "status" variable. + 2) Check that output(destLen) = uncompressedSize, if you know real uncompressedSize. + 3) Check that output(srcLen) = compressedSize, if you know real compressedSize. + You must use correct finish mode in that case. */ + +typedef enum +{ + LZMA_STATUS_NOT_SPECIFIED, /* use main error code instead */ + LZMA_STATUS_FINISHED_WITH_MARK, /* stream was finished with end mark. */ + LZMA_STATUS_NOT_FINISHED, /* stream was not finished */ + LZMA_STATUS_NEEDS_MORE_INPUT, /* you must provide more input bytes */ + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK /* there is probability that stream was finished without end mark */ +} ELzmaStatus; + +/* ELzmaStatus is used only as output value for function call */ + + +/* ---------- Interfaces ---------- */ + +/* There are 3 levels of interfaces: + 1) Dictionary Interface + 2) Buffer Interface + 3) One Call Interface + You can select any of these interfaces, but don't mix functions from different + groups for same object. */ + + +/* There are two variants to allocate state for Dictionary Interface: + 1) LzmaDec_Allocate / LzmaDec_Free + 2) LzmaDec_AllocateProbs / LzmaDec_FreeProbs + You can use variant 2, if you set dictionary buffer manually. + For Buffer Interface you must always use variant 1. + +LzmaDec_Allocate* can return: + SZ_OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties +*/ + +SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc); +void LzmaDec_FreeProbs(CLzmaDec *p, ISzAllocPtr alloc); + +SRes LzmaDec_Allocate(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc); +void LzmaDec_Free(CLzmaDec *p, ISzAllocPtr alloc); + +/* ---------- Dictionary Interface ---------- */ + +/* You can use it, if you want to eliminate the overhead for data copying from + dictionary to some other external buffer. + You must work with CLzmaDec variables directly in this interface. + + STEPS: + LzmaDec_Construct() + LzmaDec_Allocate() + for (each new stream) + { + LzmaDec_Init() + while (it needs more decompression) + { + LzmaDec_DecodeToDic() + use data from CLzmaDec::dic and update CLzmaDec::dicPos + } + } + LzmaDec_Free() +*/ + +/* LzmaDec_DecodeToDic + + The decoding to internal dictionary buffer (CLzmaDec::dic). + You must manually update CLzmaDec::dicPos, if it reaches CLzmaDec::dicBufSize !!! + +finishMode: + It has meaning only if the decoding reaches output limit (dicLimit). + LZMA_FINISH_ANY - Decode just dicLimit bytes. + LZMA_FINISH_END - Stream must be finished after dicLimit. + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_NEEDS_MORE_INPUT + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + SZ_ERROR_DATA - Data error + SZ_ERROR_FAIL - Some unexpected error: internal error of code, memory corruption or hardware failure +*/ + +SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + + +/* ---------- Buffer Interface ---------- */ + +/* It's zlib-like interface. + See LzmaDec_DecodeToDic description for information about STEPS and return results, + but you must use LzmaDec_DecodeToBuf instead of LzmaDec_DecodeToDic and you don't need + to work with CLzmaDec variables manually. + +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - Decode just destLen bytes. + LZMA_FINISH_END - Stream must be finished after (*destLen). +*/ + +SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status); + + +/* ---------- One Call Interface ---------- */ + +/* LzmaDecode + +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + LZMA_FINISH_ANY - Decode just destLen bytes. + LZMA_FINISH_END - Stream must be finished after (*destLen). + +Returns: + SZ_OK + status: + LZMA_STATUS_FINISHED_WITH_MARK + LZMA_STATUS_NOT_FINISHED + LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). + SZ_ERROR_FAIL - Some unexpected error: internal error of code, memory corruption or hardware failure +*/ + +SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, + ELzmaStatus *status, ISzAllocPtr alloc); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/LzmaEnc.h b/3rdparty/lzma/include/LzmaEnc.h new file mode 100644 index 0000000000..bc2ed5042b --- /dev/null +++ b/3rdparty/lzma/include/LzmaEnc.h @@ -0,0 +1,78 @@ +/* LzmaEnc.h -- LZMA Encoder +2019-10-30 : Igor Pavlov : Public domain */ + +#ifndef __LZMA_ENC_H +#define __LZMA_ENC_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define LZMA_PROPS_SIZE 5 + +typedef struct _CLzmaEncProps +{ + int level; /* 0 <= level <= 9 */ + UInt32 dictSize; /* (1 << 12) <= dictSize <= (1 << 27) for 32-bit version + (1 << 12) <= dictSize <= (3 << 29) for 64-bit version + default = (1 << 24) */ + int lc; /* 0 <= lc <= 8, default = 3 */ + int lp; /* 0 <= lp <= 4, default = 0 */ + int pb; /* 0 <= pb <= 4, default = 2 */ + int algo; /* 0 - fast, 1 - normal, default = 1 */ + int fb; /* 5 <= fb <= 273, default = 32 */ + int btMode; /* 0 - hashChain Mode, 1 - binTree mode - normal, default = 1 */ + int numHashBytes; /* 2, 3 or 4, default = 4 */ + UInt32 mc; /* 1 <= mc <= (1 << 30), default = 32 */ + unsigned writeEndMark; /* 0 - do not write EOPM, 1 - write EOPM, default = 0 */ + int numThreads; /* 1 or 2, default = 2 */ + + UInt64 reduceSize; /* estimated size of data that will be compressed. default = (UInt64)(Int64)-1. + Encoder uses this value to reduce dictionary size */ + + UInt64 affinity; +} CLzmaEncProps; + +void LzmaEncProps_Init(CLzmaEncProps *p); +void LzmaEncProps_Normalize(CLzmaEncProps *p); +UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2); + + +/* ---------- CLzmaEncHandle Interface ---------- */ + +/* LzmaEnc* functions can return the following exit codes: +SRes: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater in props + SZ_ERROR_WRITE - ISeqOutStream write callback error + SZ_ERROR_OUTPUT_EOF - output buffer overflow - version with (Byte *) output + SZ_ERROR_PROGRESS - some break from progress callback + SZ_ERROR_THREAD - error in multithreading functions (only for Mt version) +*/ + +typedef void * CLzmaEncHandle; + +CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc); +void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAllocPtr alloc, ISzAllocPtr allocBig); + +SRes LzmaEnc_SetProps(CLzmaEncHandle p, const CLzmaEncProps *props); +void LzmaEnc_SetDataSize(CLzmaEncHandle p, UInt64 expectedDataSiize); +SRes LzmaEnc_WriteProperties(CLzmaEncHandle p, Byte *properties, SizeT *size); +unsigned LzmaEnc_IsWriteEndMark(CLzmaEncHandle p); + +SRes LzmaEnc_Encode(CLzmaEncHandle p, ISeqOutStream *outStream, ISeqInStream *inStream, + ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_MemEncode(CLzmaEncHandle p, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + int writeEndMark, ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); + + +/* ---------- One Call Interface ---------- */ + +SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, + ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/LzmaLib.h b/3rdparty/lzma/include/LzmaLib.h new file mode 100644 index 0000000000..c343a8596d --- /dev/null +++ b/3rdparty/lzma/include/LzmaLib.h @@ -0,0 +1,138 @@ +/* LzmaLib.h -- LZMA library interface +2021-04-03 : Igor Pavlov : Public domain */ + +#ifndef __LZMA_LIB_H +#define __LZMA_LIB_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define MY_STDAPI int MY_STD_CALL + +#define LZMA_PROPS_SIZE 5 + +/* +RAM requirements for LZMA: + for compression: (dictSize * 11.5 + 6 MB) + state_size + for decompression: dictSize + state_size + state_size = (4 + (1.5 << (lc + lp))) KB + by default (lc=3, lp=0), state_size = 16 KB. + +LZMA properties (5 bytes) format + Offset Size Description + 0 1 lc, lp and pb in encoded form. + 1 4 dictSize (little endian). +*/ + +/* +LzmaCompress +------------ + +outPropsSize - + In: the pointer to the size of outProps buffer; *outPropsSize = LZMA_PROPS_SIZE = 5. + Out: the pointer to the size of written properties in outProps buffer; *outPropsSize = LZMA_PROPS_SIZE = 5. + + LZMA Encoder will use defult values for any parameter, if it is + -1 for any from: level, loc, lp, pb, fb, numThreads + 0 for dictSize + +level - compression level: 0 <= level <= 9; + + level dictSize algo fb + 0: 64 KB 0 32 + 1: 256 KB 0 32 + 2: 1 MB 0 32 + 3: 4 MB 0 32 + 4: 16 MB 0 32 + 5: 16 MB 1 32 + 6: 32 MB 1 32 + 7: 32 MB 1 64 + 8: 64 MB 1 64 + 9: 64 MB 1 64 + + The default value for "level" is 5. + + algo = 0 means fast method + algo = 1 means normal method + +dictSize - The dictionary size in bytes. The maximum value is + 128 MB = (1 << 27) bytes for 32-bit version + 1 GB = (1 << 30) bytes for 64-bit version + The default value is 16 MB = (1 << 24) bytes. + It's recommended to use the dictionary that is larger than 4 KB and + that can be calculated as (1 << N) or (3 << N) sizes. + +lc - The number of literal context bits (high bits of previous literal). + It can be in the range from 0 to 8. The default value is 3. + Sometimes lc=4 gives the gain for big files. + +lp - The number of literal pos bits (low bits of current position for literals). + It can be in the range from 0 to 4. The default value is 0. + The lp switch is intended for periodical data when the period is equal to 2^lp. + For example, for 32-bit (4 bytes) periodical data you can use lp=2. Often it's + better to set lc=0, if you change lp switch. + +pb - The number of pos bits (low bits of current position). + It can be in the range from 0 to 4. The default value is 2. + The pb switch is intended for periodical data when the period is equal 2^pb. + +fb - Word size (the number of fast bytes). + It can be in the range from 5 to 273. The default value is 32. + Usually, a big number gives a little bit better compression ratio and + slower compression process. + +numThreads - The number of thereads. 1 or 2. The default value is 2. + Fast mode (algo = 0) can use only 1 thread. + +In: + dest - output data buffer + destLen - output data buffer size + src - input data + srcLen - input data size +Out: + destLen - processed output size +Returns: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater + SZ_ERROR_OUTPUT_EOF - output buffer overflow + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) +*/ + +MY_STDAPI LzmaCompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t srcLen, + unsigned char *outProps, size_t *outPropsSize, /* *outPropsSize must be = 5 */ + int level, /* 0 <= level <= 9, default = 5 */ + unsigned dictSize, /* default = (1 << 24) */ + int lc, /* 0 <= lc <= 8, default = 3 */ + int lp, /* 0 <= lp <= 4, default = 0 */ + int pb, /* 0 <= pb <= 4, default = 2 */ + int fb, /* 5 <= fb <= 273, default = 32 */ + int numThreads /* 1 or 2, default = 2 */ + ); + +/* +LzmaUncompress +-------------- +In: + dest - output data buffer + destLen - output data buffer size + src - input data + srcLen - input data size +Out: + destLen - processed output size + srcLen - processed input size +Returns: + SZ_OK - OK + SZ_ERROR_DATA - Data error + SZ_ERROR_MEM - Memory allocation arror + SZ_ERROR_UNSUPPORTED - Unsupported properties + SZ_ERROR_INPUT_EOF - it needs more bytes in input buffer (src) +*/ + +MY_STDAPI LzmaUncompress(unsigned char *dest, size_t *destLen, const unsigned char *src, SizeT *srcLen, + const unsigned char *props, size_t propsSize); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Ppmd.h b/3rdparty/lzma/include/Ppmd.h new file mode 100644 index 0000000000..b198792087 --- /dev/null +++ b/3rdparty/lzma/include/Ppmd.h @@ -0,0 +1,167 @@ +/* Ppmd.h -- PPMD codec common code +2021-04-13 : Igor Pavlov : Public domain +This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ + +#ifndef __PPMD_H +#define __PPMD_H + +#include "CpuArch.h" + +EXTERN_C_BEGIN + +#if defined(MY_CPU_SIZEOF_POINTER) && (MY_CPU_SIZEOF_POINTER == 4) +/* + PPMD code always uses 32-bit internal fields in PPMD structures to store internal references in main block. + if (PPMD_32BIT is defined), the PPMD code stores internal pointers to 32-bit reference fields. + if (PPMD_32BIT is NOT defined), the PPMD code stores internal UInt32 offsets to reference fields. + if (pointer size is 64-bit), then (PPMD_32BIT) mode is not allowed, + if (pointer size is 32-bit), then (PPMD_32BIT) mode is optional, + and it's allowed to disable PPMD_32BIT mode even if pointer is 32-bit. + PPMD code works slightly faster in (PPMD_32BIT) mode. +*/ + #define PPMD_32BIT +#endif + +#define PPMD_INT_BITS 7 +#define PPMD_PERIOD_BITS 7 +#define PPMD_BIN_SCALE (1 << (PPMD_INT_BITS + PPMD_PERIOD_BITS)) + +#define PPMD_GET_MEAN_SPEC(summ, shift, round) (((summ) + (1 << ((shift) - (round)))) >> (shift)) +#define PPMD_GET_MEAN(summ) PPMD_GET_MEAN_SPEC((summ), PPMD_PERIOD_BITS, 2) +#define PPMD_UPDATE_PROB_0(prob) ((prob) + (1 << PPMD_INT_BITS) - PPMD_GET_MEAN(prob)) +#define PPMD_UPDATE_PROB_1(prob) ((prob) - PPMD_GET_MEAN(prob)) + +#define PPMD_N1 4 +#define PPMD_N2 4 +#define PPMD_N3 4 +#define PPMD_N4 ((128 + 3 - 1 * PPMD_N1 - 2 * PPMD_N2 - 3 * PPMD_N3) / 4) +#define PPMD_NUM_INDEXES (PPMD_N1 + PPMD_N2 + PPMD_N3 + PPMD_N4) + +MY_CPU_pragma_pack_push_1 +/* Most compilers works OK here even without #pragma pack(push, 1), but some GCC compilers need it. */ + +/* SEE-contexts for PPM-contexts with masked symbols */ +typedef struct +{ + UInt16 Summ; /* Freq */ + Byte Shift; /* Speed of Freq change; low Shift is for fast change */ + Byte Count; /* Count to next change of Shift */ +} CPpmd_See; + +#define Ppmd_See_Update(p) if ((p)->Shift < PPMD_PERIOD_BITS && --(p)->Count == 0) \ + { (p)->Summ = (UInt16)((p)->Summ << 1); (p)->Count = (Byte)(3 << (p)->Shift++); } + + +typedef struct +{ + Byte Symbol; + Byte Freq; + UInt16 Successor_0; + UInt16 Successor_1; +} CPpmd_State; + +typedef struct CPpmd_State2_ +{ + Byte Symbol; + Byte Freq; +} CPpmd_State2; + +typedef struct CPpmd_State4_ +{ + UInt16 Successor_0; + UInt16 Successor_1; +} CPpmd_State4; + +MY_CPU_pragma_pop + +/* + PPMD code can write full CPpmd_State structure data to CPpmd*_Context + at (byte offset = 2) instead of some fields of original CPpmd*_Context structure. + + If we use pointers to different types, but that point to shared + memory space, we can have aliasing problem (strict aliasing). + + XLC compiler in -O2 mode can change the order of memory write instructions + in relation to read instructions, if we have use pointers to different types. + + To solve that aliasing problem we use combined CPpmd*_Context structure + with unions that contain the fields from both structures: + the original CPpmd*_Context and CPpmd_State. + So we can access the fields from both structures via one pointer, + and the compiler doesn't change the order of write instructions + in relation to read instructions. + + If we don't use memory write instructions to shared memory in + some local code, and we use only reading instructions (read only), + then probably it's safe to use pointers to different types for reading. +*/ + + + +#ifdef PPMD_32BIT + + #define Ppmd_Ref_Type(type) type * + #define Ppmd_GetRef(p, ptr) (ptr) + #define Ppmd_GetPtr(p, ptr) (ptr) + #define Ppmd_GetPtr_Type(p, ptr, note_type) (ptr) + +#else + + #define Ppmd_Ref_Type(type) UInt32 + #define Ppmd_GetRef(p, ptr) ((UInt32)((Byte *)(ptr) - (p)->Base)) + #define Ppmd_GetPtr(p, offs) ((void *)((p)->Base + (offs))) + #define Ppmd_GetPtr_Type(p, offs, type) ((type *)Ppmd_GetPtr(p, offs)) + +#endif // PPMD_32BIT + + +typedef Ppmd_Ref_Type(CPpmd_State) CPpmd_State_Ref; +typedef Ppmd_Ref_Type(void) CPpmd_Void_Ref; +typedef Ppmd_Ref_Type(Byte) CPpmd_Byte_Ref; + + +/* +#ifdef MY_CPU_LE_UNALIGN +// the unaligned 32-bit access latency can be too large, if the data is not in L1 cache. +#define Ppmd_GET_SUCCESSOR(p) ((CPpmd_Void_Ref)*(const UInt32 *)(const void *)&(p)->Successor_0) +#define Ppmd_SET_SUCCESSOR(p, v) *(UInt32 *)(void *)(void *)&(p)->Successor_0 = (UInt32)(v) + +#else +*/ + +/* + We can write 16-bit halves to 32-bit (Successor) field in any selected order. + But the native order is more consistent way. + So we use the native order, if LE/BE order can be detected here at compile time. +*/ + +#ifdef MY_CPU_BE + + #define Ppmd_GET_SUCCESSOR(p) \ + ( (CPpmd_Void_Ref) (((UInt32)(p)->Successor_0 << 16) | (p)->Successor_1) ) + + #define Ppmd_SET_SUCCESSOR(p, v) { \ + (p)->Successor_0 = (UInt16)(((UInt32)(v) >> 16) /* & 0xFFFF */); \ + (p)->Successor_1 = (UInt16)((UInt32)(v) /* & 0xFFFF */); } + +#else + + #define Ppmd_GET_SUCCESSOR(p) \ + ( (CPpmd_Void_Ref) ((p)->Successor_0 | ((UInt32)(p)->Successor_1 << 16)) ) + + #define Ppmd_SET_SUCCESSOR(p, v) { \ + (p)->Successor_0 = (UInt16)((UInt32)(v) /* & 0xFFFF */); \ + (p)->Successor_1 = (UInt16)(((UInt32)(v) >> 16) /* & 0xFFFF */); } + +#endif + +// #endif + + +#define PPMD_SetAllBitsIn256Bytes(p) \ + { size_t z; for (z = 0; z < 256 / sizeof(p[0]); z += 8) { \ + p[z+7] = p[z+6] = p[z+5] = p[z+4] = p[z+3] = p[z+2] = p[z+1] = p[z+0] = ~(size_t)0; }} + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Ppmd7.h b/3rdparty/lzma/include/Ppmd7.h new file mode 100644 index 0000000000..d31809aebd --- /dev/null +++ b/3rdparty/lzma/include/Ppmd7.h @@ -0,0 +1,181 @@ +/* Ppmd7.h -- Ppmd7 (PPMdH) compression codec +2021-04-13 : Igor Pavlov : Public domain +This code is based on: + PPMd var.H (2001): Dmitry Shkarin : Public domain */ + + +#ifndef __PPMD7_H +#define __PPMD7_H + +#include "Ppmd.h" + +EXTERN_C_BEGIN + +#define PPMD7_MIN_ORDER 2 +#define PPMD7_MAX_ORDER 64 + +#define PPMD7_MIN_MEM_SIZE (1 << 11) +#define PPMD7_MAX_MEM_SIZE (0xFFFFFFFF - 12 * 3) + +struct CPpmd7_Context_; + +typedef Ppmd_Ref_Type(struct CPpmd7_Context_) CPpmd7_Context_Ref; + +// MY_CPU_pragma_pack_push_1 + +typedef struct CPpmd7_Context_ +{ + UInt16 NumStats; + + + union + { + UInt16 SummFreq; + CPpmd_State2 State2; + } Union2; + + union + { + CPpmd_State_Ref Stats; + CPpmd_State4 State4; + } Union4; + + CPpmd7_Context_Ref Suffix; +} CPpmd7_Context; + +// MY_CPU_pragma_pop + +#define Ppmd7Context_OneState(p) ((CPpmd_State *)&(p)->Union2) + + + + +typedef struct +{ + UInt32 Range; + UInt32 Code; + UInt32 Low; + IByteIn *Stream; +} CPpmd7_RangeDec; + + +typedef struct +{ + UInt32 Range; + Byte Cache; + // Byte _dummy_[3]; + UInt64 Low; + UInt64 CacheSize; + IByteOut *Stream; +} CPpmd7z_RangeEnc; + + +typedef struct +{ + CPpmd7_Context *MinContext, *MaxContext; + CPpmd_State *FoundState; + unsigned OrderFall, InitEsc, PrevSuccess, MaxOrder, HiBitsFlag; + Int32 RunLength, InitRL; /* must be 32-bit at least */ + + UInt32 Size; + UInt32 GlueCount; + UInt32 AlignOffset; + Byte *Base, *LoUnit, *HiUnit, *Text, *UnitsStart; + + + + + union + { + CPpmd7_RangeDec dec; + CPpmd7z_RangeEnc enc; + } rc; + + Byte Indx2Units[PPMD_NUM_INDEXES + 2]; // +2 for alignment + Byte Units2Indx[128]; + CPpmd_Void_Ref FreeList[PPMD_NUM_INDEXES]; + + Byte NS2BSIndx[256], NS2Indx[256]; + Byte ExpEscape[16]; + CPpmd_See DummySee, See[25][16]; + UInt16 BinSumm[128][64]; + // int LastSymbol; +} CPpmd7; + + +void Ppmd7_Construct(CPpmd7 *p); +BoolInt Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAllocPtr alloc); +void Ppmd7_Free(CPpmd7 *p, ISzAllocPtr alloc); +void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder); +#define Ppmd7_WasAllocated(p) ((p)->Base != NULL) + + +/* ---------- Internal Functions ---------- */ + +#define Ppmd7_GetPtr(p, ptr) Ppmd_GetPtr(p, ptr) +#define Ppmd7_GetContext(p, ptr) Ppmd_GetPtr_Type(p, ptr, CPpmd7_Context) +#define Ppmd7_GetStats(p, ctx) Ppmd_GetPtr_Type(p, (ctx)->Union4.Stats, CPpmd_State) + +void Ppmd7_Update1(CPpmd7 *p); +void Ppmd7_Update1_0(CPpmd7 *p); +void Ppmd7_Update2(CPpmd7 *p); + +#define PPMD7_HiBitsFlag_3(sym) ((((unsigned)sym + 0xC0) >> (8 - 3)) & (1 << 3)) +#define PPMD7_HiBitsFlag_4(sym) ((((unsigned)sym + 0xC0) >> (8 - 4)) & (1 << 4)) +// #define PPMD7_HiBitsFlag_3(sym) ((sym) < 0x40 ? 0 : (1 << 3)) +// #define PPMD7_HiBitsFlag_4(sym) ((sym) < 0x40 ? 0 : (1 << 4)) + +#define Ppmd7_GetBinSumm(p) \ + &p->BinSumm[(size_t)(unsigned)Ppmd7Context_OneState(p->MinContext)->Freq - 1] \ + [ p->PrevSuccess + ((p->RunLength >> 26) & 0x20) \ + + p->NS2BSIndx[(size_t)Ppmd7_GetContext(p, p->MinContext->Suffix)->NumStats - 1] \ + + PPMD7_HiBitsFlag_4(Ppmd7Context_OneState(p->MinContext)->Symbol) \ + + (p->HiBitsFlag = PPMD7_HiBitsFlag_3(p->FoundState->Symbol)) ] + +CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *scale); + + +/* +We support two versions of Ppmd7 (PPMdH) methods that use same CPpmd7 structure: + 1) Ppmd7a_*: original PPMdH + 2) Ppmd7z_*: modified PPMdH with 7z Range Coder +Ppmd7_*: the structures and functions that are common for both versions of PPMd7 (PPMdH) +*/ + +/* ---------- Decode ---------- */ + +#define PPMD7_SYM_END (-1) +#define PPMD7_SYM_ERROR (-2) + +/* +You must set (CPpmd7::rc.dec.Stream) before Ppmd7*_RangeDec_Init() + +Ppmd7*_DecodeSymbol() +out: + >= 0 : decoded byte + -1 : PPMD7_SYM_END : End of payload marker + -2 : PPMD7_SYM_ERROR : Data error +*/ + +/* Ppmd7a_* : original PPMdH */ +BoolInt Ppmd7a_RangeDec_Init(CPpmd7_RangeDec *p); +#define Ppmd7a_RangeDec_IsFinishedOK(p) ((p)->Code == 0) +int Ppmd7a_DecodeSymbol(CPpmd7 *p); + +/* Ppmd7z_* : modified PPMdH with 7z Range Coder */ +BoolInt Ppmd7z_RangeDec_Init(CPpmd7_RangeDec *p); +#define Ppmd7z_RangeDec_IsFinishedOK(p) ((p)->Code == 0) +int Ppmd7z_DecodeSymbol(CPpmd7 *p); +// Byte *Ppmd7z_DecodeSymbols(CPpmd7 *p, Byte *buf, const Byte *lim); + + +/* ---------- Encode ---------- */ + +void Ppmd7z_Init_RangeEnc(CPpmd7 *p); +void Ppmd7z_Flush_RangeEnc(CPpmd7 *p); +// void Ppmd7z_EncodeSymbol(CPpmd7 *p, int symbol); +void Ppmd7z_EncodeSymbols(CPpmd7 *p, const Byte *buf, const Byte *lim); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Precomp.h b/3rdparty/lzma/include/Precomp.h new file mode 100644 index 0000000000..e8ff8b40e8 --- /dev/null +++ b/3rdparty/lzma/include/Precomp.h @@ -0,0 +1,10 @@ +/* Precomp.h -- StdAfx +2013-11-12 : Igor Pavlov : Public domain */ + +#ifndef __7Z_PRECOMP_H +#define __7Z_PRECOMP_H + +#include "Compiler.h" +/* #include "7zTypes.h" */ + +#endif diff --git a/3rdparty/lzma/include/RotateDefs.h b/3rdparty/lzma/include/RotateDefs.h new file mode 100644 index 0000000000..8f01d1a6c5 --- /dev/null +++ b/3rdparty/lzma/include/RotateDefs.h @@ -0,0 +1,30 @@ +/* RotateDefs.h -- Rotate functions +2015-03-25 : Igor Pavlov : Public domain */ + +#ifndef __ROTATE_DEFS_H +#define __ROTATE_DEFS_H + +#ifdef _MSC_VER + +#include + +/* don't use _rotl with MINGW. It can insert slow call to function. */ + +/* #if (_MSC_VER >= 1200) */ +#pragma intrinsic(_rotl) +#pragma intrinsic(_rotr) +/* #endif */ + +#define rotlFixed(x, n) _rotl((x), (n)) +#define rotrFixed(x, n) _rotr((x), (n)) + +#else + +/* new compilers can translate these macros to fast commands. */ + +#define rotlFixed(x, n) (((x) << (n)) | ((x) >> (32 - (n)))) +#define rotrFixed(x, n) (((x) >> (n)) | ((x) << (32 - (n)))) + +#endif + +#endif diff --git a/3rdparty/lzma/include/Sha256.h b/3rdparty/lzma/include/Sha256.h new file mode 100644 index 0000000000..aa38501e57 --- /dev/null +++ b/3rdparty/lzma/include/Sha256.h @@ -0,0 +1,76 @@ +/* Sha256.h -- SHA-256 Hash +2021-01-01 : Igor Pavlov : Public domain */ + +#ifndef __7Z_SHA256_H +#define __7Z_SHA256_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define SHA256_NUM_BLOCK_WORDS 16 +#define SHA256_NUM_DIGEST_WORDS 8 + +#define SHA256_BLOCK_SIZE (SHA256_NUM_BLOCK_WORDS * 4) +#define SHA256_DIGEST_SIZE (SHA256_NUM_DIGEST_WORDS * 4) + +typedef void (MY_FAST_CALL *SHA256_FUNC_UPDATE_BLOCKS)(UInt32 state[8], const Byte *data, size_t numBlocks); + +/* + if (the system supports different SHA256 code implementations) + { + (CSha256::func_UpdateBlocks) will be used + (CSha256::func_UpdateBlocks) can be set by + Sha256_Init() - to default (fastest) + Sha256_SetFunction() - to any algo + } + else + { + (CSha256::func_UpdateBlocks) is ignored. + } +*/ + +typedef struct +{ + SHA256_FUNC_UPDATE_BLOCKS func_UpdateBlocks; + UInt64 count; + UInt64 __pad_2[2]; + UInt32 state[SHA256_NUM_DIGEST_WORDS]; + + Byte buffer[SHA256_BLOCK_SIZE]; +} CSha256; + + +#define SHA256_ALGO_DEFAULT 0 +#define SHA256_ALGO_SW 1 +#define SHA256_ALGO_HW 2 + +/* +Sha256_SetFunction() +return: + 0 - (algo) value is not supported, and func_UpdateBlocks was not changed + 1 - func_UpdateBlocks was set according (algo) value. +*/ + +BoolInt Sha256_SetFunction(CSha256 *p, unsigned algo); + +void Sha256_InitState(CSha256 *p); +void Sha256_Init(CSha256 *p); +void Sha256_Update(CSha256 *p, const Byte *data, size_t size); +void Sha256_Final(CSha256 *p, Byte *digest); + + + + +// void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks); + +/* +call Sha256Prepare() once at program start. +It prepares all supported implementations, and detects the fastest implementation. +*/ + +void Sha256Prepare(void); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/Xz.h b/3rdparty/lzma/include/Xz.h new file mode 100644 index 0000000000..849b944bf5 --- /dev/null +++ b/3rdparty/lzma/include/Xz.h @@ -0,0 +1,517 @@ +/* Xz.h - Xz interface +2021-04-01 : Igor Pavlov : Public domain */ + +#ifndef __XZ_H +#define __XZ_H + +#include "Sha256.h" + +EXTERN_C_BEGIN + +#define XZ_ID_Subblock 1 +#define XZ_ID_Delta 3 +#define XZ_ID_X86 4 +#define XZ_ID_PPC 5 +#define XZ_ID_IA64 6 +#define XZ_ID_ARM 7 +#define XZ_ID_ARMT 8 +#define XZ_ID_SPARC 9 +#define XZ_ID_LZMA2 0x21 + +unsigned Xz_ReadVarInt(const Byte *p, size_t maxSize, UInt64 *value); +unsigned Xz_WriteVarInt(Byte *buf, UInt64 v); + +/* ---------- xz block ---------- */ + +#define XZ_BLOCK_HEADER_SIZE_MAX 1024 + +#define XZ_NUM_FILTERS_MAX 4 +#define XZ_BF_NUM_FILTERS_MASK 3 +#define XZ_BF_PACK_SIZE (1 << 6) +#define XZ_BF_UNPACK_SIZE (1 << 7) + +#define XZ_FILTER_PROPS_SIZE_MAX 20 + +typedef struct +{ + UInt64 id; + UInt32 propsSize; + Byte props[XZ_FILTER_PROPS_SIZE_MAX]; +} CXzFilter; + +typedef struct +{ + UInt64 packSize; + UInt64 unpackSize; + Byte flags; + CXzFilter filters[XZ_NUM_FILTERS_MAX]; +} CXzBlock; + +#define XzBlock_GetNumFilters(p) (((unsigned)(p)->flags & XZ_BF_NUM_FILTERS_MASK) + 1) +#define XzBlock_HasPackSize(p) (((p)->flags & XZ_BF_PACK_SIZE) != 0) +#define XzBlock_HasUnpackSize(p) (((p)->flags & XZ_BF_UNPACK_SIZE) != 0) +#define XzBlock_HasUnsupportedFlags(p) (((p)->flags & ~(XZ_BF_NUM_FILTERS_MASK | XZ_BF_PACK_SIZE | XZ_BF_UNPACK_SIZE)) != 0) + +SRes XzBlock_Parse(CXzBlock *p, const Byte *header); +SRes XzBlock_ReadHeader(CXzBlock *p, ISeqInStream *inStream, BoolInt *isIndex, UInt32 *headerSizeRes); + +/* ---------- xz stream ---------- */ + +#define XZ_SIG_SIZE 6 +#define XZ_FOOTER_SIG_SIZE 2 + +extern const Byte XZ_SIG[XZ_SIG_SIZE]; + +/* +extern const Byte XZ_FOOTER_SIG[XZ_FOOTER_SIG_SIZE]; +*/ + +#define XZ_FOOTER_SIG_0 'Y' +#define XZ_FOOTER_SIG_1 'Z' + +#define XZ_STREAM_FLAGS_SIZE 2 +#define XZ_STREAM_CRC_SIZE 4 + +#define XZ_STREAM_HEADER_SIZE (XZ_SIG_SIZE + XZ_STREAM_FLAGS_SIZE + XZ_STREAM_CRC_SIZE) +#define XZ_STREAM_FOOTER_SIZE (XZ_FOOTER_SIG_SIZE + XZ_STREAM_FLAGS_SIZE + XZ_STREAM_CRC_SIZE + 4) + +#define XZ_CHECK_MASK 0xF +#define XZ_CHECK_NO 0 +#define XZ_CHECK_CRC32 1 +#define XZ_CHECK_CRC64 4 +#define XZ_CHECK_SHA256 10 + +typedef struct +{ + unsigned mode; + UInt32 crc; + UInt64 crc64; + CSha256 sha; +} CXzCheck; + +void XzCheck_Init(CXzCheck *p, unsigned mode); +void XzCheck_Update(CXzCheck *p, const void *data, size_t size); +int XzCheck_Final(CXzCheck *p, Byte *digest); + +typedef UInt16 CXzStreamFlags; + +#define XzFlags_IsSupported(f) ((f) <= XZ_CHECK_MASK) +#define XzFlags_GetCheckType(f) ((f) & XZ_CHECK_MASK) +#define XzFlags_HasDataCrc32(f) (Xz_GetCheckType(f) == XZ_CHECK_CRC32) +unsigned XzFlags_GetCheckSize(CXzStreamFlags f); + +SRes Xz_ParseHeader(CXzStreamFlags *p, const Byte *buf); +SRes Xz_ReadHeader(CXzStreamFlags *p, ISeqInStream *inStream); + +typedef struct +{ + UInt64 unpackSize; + UInt64 totalSize; +} CXzBlockSizes; + +typedef struct +{ + CXzStreamFlags flags; + size_t numBlocks; + CXzBlockSizes *blocks; + UInt64 startOffset; +} CXzStream; + +void Xz_Construct(CXzStream *p); +void Xz_Free(CXzStream *p, ISzAllocPtr alloc); + +#define XZ_SIZE_OVERFLOW ((UInt64)(Int64)-1) + +UInt64 Xz_GetUnpackSize(const CXzStream *p); +UInt64 Xz_GetPackSize(const CXzStream *p); + +typedef struct +{ + size_t num; + size_t numAllocated; + CXzStream *streams; +} CXzs; + +void Xzs_Construct(CXzs *p); +void Xzs_Free(CXzs *p, ISzAllocPtr alloc); +SRes Xzs_ReadBackward(CXzs *p, ILookInStream *inStream, Int64 *startOffset, ICompressProgress *progress, ISzAllocPtr alloc); + +UInt64 Xzs_GetNumBlocks(const CXzs *p); +UInt64 Xzs_GetUnpackSize(const CXzs *p); + + +// ECoderStatus values are identical to ELzmaStatus values of LZMA2 decoder + +typedef enum +{ + CODER_STATUS_NOT_SPECIFIED, /* use main error code instead */ + CODER_STATUS_FINISHED_WITH_MARK, /* stream was finished with end mark. */ + CODER_STATUS_NOT_FINISHED, /* stream was not finished */ + CODER_STATUS_NEEDS_MORE_INPUT /* you must provide more input bytes */ +} ECoderStatus; + + +// ECoderFinishMode values are identical to ELzmaFinishMode + +typedef enum +{ + CODER_FINISH_ANY, /* finish at any point */ + CODER_FINISH_END /* block must be finished at the end */ +} ECoderFinishMode; + + +typedef struct _IStateCoder +{ + void *p; + void (*Free)(void *p, ISzAllocPtr alloc); + SRes (*SetProps)(void *p, const Byte *props, size_t propSize, ISzAllocPtr alloc); + void (*Init)(void *p); + SRes (*Code2)(void *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, + // int *wasFinished, + ECoderStatus *status); + SizeT (*Filter)(void *p, Byte *data, SizeT size); +} IStateCoder; + + + +#define MIXCODER_NUM_FILTERS_MAX 4 + +typedef struct +{ + ISzAllocPtr alloc; + Byte *buf; + unsigned numCoders; + + Byte *outBuf; + size_t outBufSize; + size_t outWritten; // is equal to lzmaDecoder.dicPos (in outBuf mode) + BoolInt wasFinished; + SRes res; + ECoderStatus status; + // BoolInt SingleBufMode; + + int finished[MIXCODER_NUM_FILTERS_MAX - 1]; + size_t pos[MIXCODER_NUM_FILTERS_MAX - 1]; + size_t size[MIXCODER_NUM_FILTERS_MAX - 1]; + UInt64 ids[MIXCODER_NUM_FILTERS_MAX]; + SRes results[MIXCODER_NUM_FILTERS_MAX]; + IStateCoder coders[MIXCODER_NUM_FILTERS_MAX]; +} CMixCoder; + + +typedef enum +{ + XZ_STATE_STREAM_HEADER, + XZ_STATE_STREAM_INDEX, + XZ_STATE_STREAM_INDEX_CRC, + XZ_STATE_STREAM_FOOTER, + XZ_STATE_STREAM_PADDING, + XZ_STATE_BLOCK_HEADER, + XZ_STATE_BLOCK, + XZ_STATE_BLOCK_FOOTER +} EXzState; + + +typedef struct +{ + EXzState state; + UInt32 pos; + unsigned alignPos; + unsigned indexPreSize; + + CXzStreamFlags streamFlags; + + UInt32 blockHeaderSize; + UInt64 packSize; + UInt64 unpackSize; + + UInt64 numBlocks; // number of finished blocks in current stream + UInt64 indexSize; + UInt64 indexPos; + UInt64 padSize; + + UInt64 numStartedStreams; + UInt64 numFinishedStreams; + UInt64 numTotalBlocks; + + UInt32 crc; + CMixCoder decoder; + CXzBlock block; + CXzCheck check; + CSha256 sha; + + BoolInt parseMode; + BoolInt headerParsedOk; + BoolInt decodeToStreamSignature; + unsigned decodeOnlyOneBlock; + + Byte *outBuf; + size_t outBufSize; + size_t outDataWritten; // the size of data in (outBuf) that were fully unpacked + + Byte shaDigest[SHA256_DIGEST_SIZE]; + Byte buf[XZ_BLOCK_HEADER_SIZE_MAX]; +} CXzUnpacker; + +/* alloc : aligned for cache line allocation is better */ +void XzUnpacker_Construct(CXzUnpacker *p, ISzAllocPtr alloc); +void XzUnpacker_Init(CXzUnpacker *p); +void XzUnpacker_SetOutBuf(CXzUnpacker *p, Byte *outBuf, size_t outBufSize); +void XzUnpacker_Free(CXzUnpacker *p); + +/* + XzUnpacker + The sequence for decoding functions: + { + XzUnpacker_Construct() + [Decoding_Calls] + XzUnpacker_Free() + } + + [Decoding_Calls] + + There are 3 types of interfaces for [Decoding_Calls] calls: + + Interface-1 : Partial output buffers: + { + XzUnpacker_Init() + for() + { + XzUnpacker_Code(); + } + XzUnpacker_IsStreamWasFinished() + } + + Interface-2 : Direct output buffer: + Use it, if you know exact size of decoded data, and you need + whole xz unpacked data in one output buffer. + xz unpacker doesn't allocate additional buffer for lzma2 dictionary in that mode. + { + XzUnpacker_Init() + XzUnpacker_SetOutBufMode(); // to set output buffer and size + for() + { + XzUnpacker_Code(); // (dest = NULL) in XzUnpacker_Code() + } + XzUnpacker_IsStreamWasFinished() + } + + Interface-3 : Direct output buffer : One call full decoding + It unpacks whole input buffer to output buffer in one call. + It uses Interface-2 internally. + { + XzUnpacker_CodeFull() + XzUnpacker_IsStreamWasFinished() + } +*/ + +/* +finishMode: + It has meaning only if the decoding reaches output limit (*destLen). + CODER_FINISH_ANY - use smallest number of input bytes + CODER_FINISH_END - read EndOfStream marker after decoding + +Returns: + SZ_OK + status: + CODER_STATUS_NOT_FINISHED, + CODER_STATUS_NEEDS_MORE_INPUT - the decoder can return it in two cases: + 1) it needs more input data to finish current xz stream + 2) xz stream was finished successfully. But the decoder supports multiple + concatented xz streams. So it expects more input data for new xz streams. + Call XzUnpacker_IsStreamWasFinished() to check that latest xz stream was finished successfully. + + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_DATA - Data error + SZ_ERROR_UNSUPPORTED - Unsupported method or method properties + SZ_ERROR_CRC - CRC error + // SZ_ERROR_INPUT_EOF - It needs more bytes in input buffer (src). + + SZ_ERROR_NO_ARCHIVE - the error with xz Stream Header with one of the following reasons: + - xz Stream Signature failure + - CRC32 of xz Stream Header is failed + - The size of Stream padding is not multiple of four bytes. + It's possible to get that error, if xz stream was finished and the stream + contains some another data. In that case you can call XzUnpacker_GetExtraSize() + function to get real size of xz stream. +*/ + + +SRes XzUnpacker_Code(CXzUnpacker *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, int srcFinished, + ECoderFinishMode finishMode, ECoderStatus *status); + +SRes XzUnpacker_CodeFull(CXzUnpacker *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, + ECoderFinishMode finishMode, ECoderStatus *status); + +/* +If you decode full xz stream(s), then you can call XzUnpacker_IsStreamWasFinished() +after successful XzUnpacker_CodeFull() or after last call of XzUnpacker_Code(). +*/ + +BoolInt XzUnpacker_IsStreamWasFinished(const CXzUnpacker *p); + +/* +XzUnpacker_GetExtraSize() returns then number of unconfirmed bytes, + if it's in (XZ_STATE_STREAM_HEADER) state or in (XZ_STATE_STREAM_PADDING) state. +These bytes can be some data after xz archive, or +it can be start of new xz stream. + +Call XzUnpacker_GetExtraSize() after XzUnpacker_Code() function to detect real size of +xz stream in two cases, if XzUnpacker_Code() returns: + res == SZ_OK && status == CODER_STATUS_NEEDS_MORE_INPUT + res == SZ_ERROR_NO_ARCHIVE +*/ + +UInt64 XzUnpacker_GetExtraSize(const CXzUnpacker *p); + + +/* + for random block decoding: + XzUnpacker_Init(); + set CXzUnpacker::streamFlags + XzUnpacker_PrepareToRandomBlockDecoding() + loop + { + XzUnpacker_Code() + XzUnpacker_IsBlockFinished() + } +*/ + +void XzUnpacker_PrepareToRandomBlockDecoding(CXzUnpacker *p); +BoolInt XzUnpacker_IsBlockFinished(const CXzUnpacker *p); + +#define XzUnpacker_GetPackSizeForIndex(p) ((p)->packSize + (p)->blockHeaderSize + XzFlags_GetCheckSize((p)->streamFlags)) + + + + + + +/* ---- Single-Thread and Multi-Thread xz Decoding with Input/Output Streams ---- */ + +/* + if (CXzDecMtProps::numThreads > 1), the decoder can try to use + Multi-Threading. The decoder analyses xz block header, and if + there are pack size and unpack size values stored in xz block header, + the decoder reads compressed data of block to internal buffers, + and then it can start parallel decoding, if there are another blocks. + The decoder can switch back to Single-Thread decoding after some conditions. + + The sequence of calls for xz decoding with in/out Streams: + { + XzDecMt_Create() + XzDecMtProps_Init(XzDecMtProps) to set default values of properties + // then you can change some XzDecMtProps parameters with required values + // here you can set the number of threads and (memUseMax) - the maximum + Memory usage for multithreading decoding. + for() + { + XzDecMt_Decode() // one call per one file + } + XzDecMt_Destroy() + } +*/ + + +typedef struct +{ + size_t inBufSize_ST; // size of input buffer for Single-Thread decoding + size_t outStep_ST; // size of output buffer for Single-Thread decoding + BoolInt ignoreErrors; // if set to 1, the decoder can ignore some errors and it skips broken parts of data. + + #ifndef _7ZIP_ST + unsigned numThreads; // the number of threads for Multi-Thread decoding. if (umThreads == 1) it will use Single-thread decoding + size_t inBufSize_MT; // size of small input data buffers for Multi-Thread decoding. Big number of such small buffers can be created + size_t memUseMax; // the limit of total memory usage for Multi-Thread decoding. + // it's recommended to set (memUseMax) manually to value that is smaller of total size of RAM in computer. + #endif +} CXzDecMtProps; + +void XzDecMtProps_Init(CXzDecMtProps *p); + + +typedef void * CXzDecMtHandle; + +/* + alloc : XzDecMt uses CAlignOffsetAlloc internally for addresses allocated by (alloc). + allocMid : for big allocations, aligned allocation is better +*/ + +CXzDecMtHandle XzDecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid); +void XzDecMt_Destroy(CXzDecMtHandle p); + + +typedef struct +{ + Byte UnpackSize_Defined; + Byte NumStreams_Defined; + Byte NumBlocks_Defined; + + Byte DataAfterEnd; // there are some additional data after good xz streams, and that data is not new xz stream. + Byte DecodingTruncated; // Decoding was Truncated, we need only partial output data + + UInt64 InSize; // pack size processed. That value doesn't include the data after + // end of xz stream, if that data was not correct + UInt64 OutSize; + + UInt64 NumStreams; + UInt64 NumBlocks; + + SRes DecodeRes; // the error code of xz streams data decoding + SRes ReadRes; // error code from ISeqInStream:Read() + SRes ProgressRes; // error code from ICompressProgress:Progress() + + SRes CombinedRes; // Combined result error code that shows main rusult + // = S_OK, if there is no error. + // but check also (DataAfterEnd) that can show additional minor errors. + + SRes CombinedRes_Type; // = SZ_ERROR_READ, if error from ISeqInStream + // = SZ_ERROR_PROGRESS, if error from ICompressProgress + // = SZ_ERROR_WRITE, if error from ISeqOutStream + // = SZ_ERROR_* codes for decoding +} CXzStatInfo; + +void XzStatInfo_Clear(CXzStatInfo *p); + +/* + +XzDecMt_Decode() +SRes: it's combined decoding result. It also is equal to stat->CombinedRes. + + SZ_OK - no error + check also output value in (stat->DataAfterEnd) + that can show additional possible error + + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_NO_ARCHIVE - is not xz archive + SZ_ERROR_ARCHIVE - Headers error + SZ_ERROR_DATA - Data Error + SZ_ERROR_UNSUPPORTED - Unsupported method or method properties + SZ_ERROR_CRC - CRC Error + SZ_ERROR_INPUT_EOF - it needs more input data + SZ_ERROR_WRITE - ISeqOutStream error + (SZ_ERROR_READ) - ISeqInStream errors + (SZ_ERROR_PROGRESS) - ICompressProgress errors + // SZ_ERROR_THREAD - error in multi-threading functions + MY_SRes_HRESULT_FROM_WRes(WRes_error) - error in multi-threading function +*/ + +SRes XzDecMt_Decode(CXzDecMtHandle p, + const CXzDecMtProps *props, + const UInt64 *outDataSize, // NULL means undefined + int finishMode, // 0 - partial unpacking is allowed, 1 - xz stream(s) must be finished + ISeqOutStream *outStream, + // Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + // const Byte *inData, size_t inDataSize, + CXzStatInfo *stat, // out: decoding results and statistics + int *isMT, // out: 0 means that ST (Single-Thread) version was used + // 1 means that MT (Multi-Thread) version was used + ICompressProgress *progress); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/XzCrc64.h b/3rdparty/lzma/include/XzCrc64.h new file mode 100644 index 0000000000..08dbc330c2 --- /dev/null +++ b/3rdparty/lzma/include/XzCrc64.h @@ -0,0 +1,26 @@ +/* XzCrc64.h -- CRC64 calculation +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __XZ_CRC64_H +#define __XZ_CRC64_H + +#include + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +extern UInt64 g_Crc64Table[]; + +void MY_FAST_CALL Crc64GenerateTable(void); + +#define CRC64_INIT_VAL UINT64_CONST(0xFFFFFFFFFFFFFFFF) +#define CRC64_GET_DIGEST(crc) ((crc) ^ CRC64_INIT_VAL) +#define CRC64_UPDATE_BYTE(crc, b) (g_Crc64Table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt64 MY_FAST_CALL Crc64Update(UInt64 crc, const void *data, size_t size); +UInt64 MY_FAST_CALL Crc64Calc(const void *data, size_t size); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/include/XzEnc.h b/3rdparty/lzma/include/XzEnc.h new file mode 100644 index 0000000000..0c29e7e1e1 --- /dev/null +++ b/3rdparty/lzma/include/XzEnc.h @@ -0,0 +1,60 @@ +/* XzEnc.h -- Xz Encode +2017-06-27 : Igor Pavlov : Public domain */ + +#ifndef __XZ_ENC_H +#define __XZ_ENC_H + +#include "Lzma2Enc.h" + +#include "Xz.h" + +EXTERN_C_BEGIN + + +#define XZ_PROPS__BLOCK_SIZE__AUTO LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO +#define XZ_PROPS__BLOCK_SIZE__SOLID LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + + +typedef struct +{ + UInt32 id; + UInt32 delta; + UInt32 ip; + int ipDefined; +} CXzFilterProps; + +void XzFilterProps_Init(CXzFilterProps *p); + + +typedef struct +{ + CLzma2EncProps lzma2Props; + CXzFilterProps filterProps; + unsigned checkId; + UInt64 blockSize; + int numBlockThreads_Reduced; + int numBlockThreads_Max; + int numTotalThreads; + int forceWriteSizesInHeader; + UInt64 reduceSize; +} CXzProps; + +void XzProps_Init(CXzProps *p); + + +typedef void * CXzEncHandle; + +CXzEncHandle XzEnc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig); +void XzEnc_Destroy(CXzEncHandle p); +SRes XzEnc_SetProps(CXzEncHandle p, const CXzProps *props); +void XzEnc_SetDataSize(CXzEncHandle p, UInt64 expectedDataSiize); +SRes XzEnc_Encode(CXzEncHandle p, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress); + +SRes Xz_Encode(ISeqOutStream *outStream, ISeqInStream *inStream, + const CXzProps *props, ICompressProgress *progress); + +SRes Xz_EncodeEmpty(ISeqOutStream *outStream); + +EXTERN_C_END + +#endif diff --git a/3rdparty/lzma/lzma.vcxproj b/3rdparty/lzma/lzma.vcxproj new file mode 100644 index 0000000000..25fde6a13b --- /dev/null +++ b/3rdparty/lzma/lzma.vcxproj @@ -0,0 +1,112 @@ + + + + + + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2} + + + + StaticLibrary + $(DefaultPlatformToolset) + MultiByte + true + true + false + + + + + + + + + + + + + + AllRules.ruleset + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + _7ZIP_ST;%(PreprocessorDefinitions) + TurnOffAllWarnings + $(SolutionDir)3rdparty\lzma\include;%(AdditionalIncludeDirectories) + + + + + \ No newline at end of file diff --git a/3rdparty/lzma/lzma.vcxproj.filters b/3rdparty/lzma/lzma.vcxproj.filters new file mode 100644 index 0000000000..f5ecde7d08 --- /dev/null +++ b/3rdparty/lzma/lzma.vcxproj.filters @@ -0,0 +1,75 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + \ No newline at end of file diff --git a/3rdparty/lzma/src/7zAlloc.c b/3rdparty/lzma/src/7zAlloc.c new file mode 100644 index 0000000000..c924a529fa --- /dev/null +++ b/3rdparty/lzma/src/7zAlloc.c @@ -0,0 +1,80 @@ +/* 7zAlloc.c -- Allocation functions +2017-04-03 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7zAlloc.h" + +/* #define _SZ_ALLOC_DEBUG */ +/* use _SZ_ALLOC_DEBUG to debug alloc/free operations */ + +#ifdef _SZ_ALLOC_DEBUG + +#ifdef _WIN32 +#include +#endif + +#include +int g_allocCount = 0; +int g_allocCountTemp = 0; + +#endif + +void *SzAlloc(ISzAllocPtr p, size_t size) +{ + UNUSED_VAR(p); + if (size == 0) + return 0; + #ifdef _SZ_ALLOC_DEBUG + fprintf(stderr, "\nAlloc %10u bytes; count = %10d", (unsigned)size, g_allocCount); + g_allocCount++; + #endif + return malloc(size); +} + +void SzFree(ISzAllocPtr p, void *address) +{ + UNUSED_VAR(p); + #ifdef _SZ_ALLOC_DEBUG + if (address != 0) + { + g_allocCount--; + fprintf(stderr, "\nFree; count = %10d", g_allocCount); + } + #endif + free(address); +} + +void *SzAllocTemp(ISzAllocPtr p, size_t size) +{ + UNUSED_VAR(p); + if (size == 0) + return 0; + #ifdef _SZ_ALLOC_DEBUG + fprintf(stderr, "\nAlloc_temp %10u bytes; count = %10d", (unsigned)size, g_allocCountTemp); + g_allocCountTemp++; + #ifdef _WIN32 + return HeapAlloc(GetProcessHeap(), 0, size); + #endif + #endif + return malloc(size); +} + +void SzFreeTemp(ISzAllocPtr p, void *address) +{ + UNUSED_VAR(p); + #ifdef _SZ_ALLOC_DEBUG + if (address != 0) + { + g_allocCountTemp--; + fprintf(stderr, "\nFree_temp; count = %10d", g_allocCountTemp); + } + #ifdef _WIN32 + HeapFree(GetProcessHeap(), 0, address); + return; + #endif + #endif + free(address); +} diff --git a/3rdparty/lzma/src/7zArcIn.c b/3rdparty/lzma/src/7zArcIn.c new file mode 100644 index 0000000000..0d9dec41e5 --- /dev/null +++ b/3rdparty/lzma/src/7zArcIn.c @@ -0,0 +1,1783 @@ +/* 7zArcIn.c -- 7z Input functions +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7z.h" +#include "7zBuf.h" +#include "7zCrc.h" +#include "CpuArch.h" + +#define MY_ALLOC(T, p, size, alloc) { \ + if ((p = (T *)ISzAlloc_Alloc(alloc, (size) * sizeof(T))) == NULL) return SZ_ERROR_MEM; } + +#define MY_ALLOC_ZE(T, p, size, alloc) { if ((size) == 0) p = NULL; else MY_ALLOC(T, p, size, alloc) } + +#define MY_ALLOC_AND_CPY(to, size, from, alloc) \ + { MY_ALLOC(Byte, to, size, alloc); memcpy(to, from, size); } + +#define MY_ALLOC_ZE_AND_CPY(to, size, from, alloc) \ + { if ((size) == 0) to = NULL; else { MY_ALLOC_AND_CPY(to, size, from, alloc) } } + +#define k7zMajorVersion 0 + +enum EIdEnum +{ + k7zIdEnd, + k7zIdHeader, + k7zIdArchiveProperties, + k7zIdAdditionalStreamsInfo, + k7zIdMainStreamsInfo, + k7zIdFilesInfo, + k7zIdPackInfo, + k7zIdUnpackInfo, + k7zIdSubStreamsInfo, + k7zIdSize, + k7zIdCRC, + k7zIdFolder, + k7zIdCodersUnpackSize, + k7zIdNumUnpackStream, + k7zIdEmptyStream, + k7zIdEmptyFile, + k7zIdAnti, + k7zIdName, + k7zIdCTime, + k7zIdATime, + k7zIdMTime, + k7zIdWinAttrib, + k7zIdComment, + k7zIdEncodedHeader, + k7zIdStartPos, + k7zIdDummy + // k7zNtSecure, + // k7zParent, + // k7zIsReal +}; + +const Byte k7zSignature[k7zSignatureSize] = {'7', 'z', 0xBC, 0xAF, 0x27, 0x1C}; + +#define SzBitUi32s_Init(p) { (p)->Defs = NULL; (p)->Vals = NULL; } + +static SRes SzBitUi32s_Alloc(CSzBitUi32s *p, size_t num, ISzAllocPtr alloc) +{ + if (num == 0) + { + p->Defs = NULL; + p->Vals = NULL; + } + else + { + MY_ALLOC(Byte, p->Defs, (num + 7) >> 3, alloc); + MY_ALLOC(UInt32, p->Vals, num, alloc); + } + return SZ_OK; +} + +static void SzBitUi32s_Free(CSzBitUi32s *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->Defs); p->Defs = NULL; + ISzAlloc_Free(alloc, p->Vals); p->Vals = NULL; +} + +#define SzBitUi64s_Init(p) { (p)->Defs = NULL; (p)->Vals = NULL; } + +static void SzBitUi64s_Free(CSzBitUi64s *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->Defs); p->Defs = NULL; + ISzAlloc_Free(alloc, p->Vals); p->Vals = NULL; +} + + +static void SzAr_Init(CSzAr *p) +{ + p->NumPackStreams = 0; + p->NumFolders = 0; + + p->PackPositions = NULL; + SzBitUi32s_Init(&p->FolderCRCs); + + p->FoCodersOffsets = NULL; + p->FoStartPackStreamIndex = NULL; + p->FoToCoderUnpackSizes = NULL; + p->FoToMainUnpackSizeIndex = NULL; + p->CoderUnpackSizes = NULL; + + p->CodersData = NULL; + + p->RangeLimit = 0; +} + +static void SzAr_Free(CSzAr *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->PackPositions); + SzBitUi32s_Free(&p->FolderCRCs, alloc); + + ISzAlloc_Free(alloc, p->FoCodersOffsets); + ISzAlloc_Free(alloc, p->FoStartPackStreamIndex); + ISzAlloc_Free(alloc, p->FoToCoderUnpackSizes); + ISzAlloc_Free(alloc, p->FoToMainUnpackSizeIndex); + ISzAlloc_Free(alloc, p->CoderUnpackSizes); + + ISzAlloc_Free(alloc, p->CodersData); + + SzAr_Init(p); +} + + +void SzArEx_Init(CSzArEx *p) +{ + SzAr_Init(&p->db); + + p->NumFiles = 0; + p->dataPos = 0; + + p->UnpackPositions = NULL; + p->IsDirs = NULL; + + p->FolderToFile = NULL; + p->FileToFolder = NULL; + + p->FileNameOffsets = NULL; + p->FileNames = NULL; + + SzBitUi32s_Init(&p->CRCs); + SzBitUi32s_Init(&p->Attribs); + // SzBitUi32s_Init(&p->Parents); + SzBitUi64s_Init(&p->MTime); + SzBitUi64s_Init(&p->CTime); +} + +void SzArEx_Free(CSzArEx *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->UnpackPositions); + ISzAlloc_Free(alloc, p->IsDirs); + + ISzAlloc_Free(alloc, p->FolderToFile); + ISzAlloc_Free(alloc, p->FileToFolder); + + ISzAlloc_Free(alloc, p->FileNameOffsets); + ISzAlloc_Free(alloc, p->FileNames); + + SzBitUi32s_Free(&p->CRCs, alloc); + SzBitUi32s_Free(&p->Attribs, alloc); + // SzBitUi32s_Free(&p->Parents, alloc); + SzBitUi64s_Free(&p->MTime, alloc); + SzBitUi64s_Free(&p->CTime, alloc); + + SzAr_Free(&p->db, alloc); + SzArEx_Init(p); +} + + +static int TestSignatureCandidate(const Byte *testBytes) +{ + unsigned i; + for (i = 0; i < k7zSignatureSize; i++) + if (testBytes[i] != k7zSignature[i]) + return 0; + return 1; +} + +#define SzData_Clear(p) { (p)->Data = NULL; (p)->Size = 0; } + +#define SZ_READ_BYTE_SD(_sd_, dest) if ((_sd_)->Size == 0) return SZ_ERROR_ARCHIVE; (_sd_)->Size--; dest = *(_sd_)->Data++; +#define SZ_READ_BYTE(dest) SZ_READ_BYTE_SD(sd, dest) +#define SZ_READ_BYTE_2(dest) if (sd.Size == 0) return SZ_ERROR_ARCHIVE; sd.Size--; dest = *sd.Data++; + +#define SKIP_DATA(sd, size) { sd->Size -= (size_t)(size); sd->Data += (size_t)(size); } +#define SKIP_DATA2(sd, size) { sd.Size -= (size_t)(size); sd.Data += (size_t)(size); } + +#define SZ_READ_32(dest) if (sd.Size < 4) return SZ_ERROR_ARCHIVE; \ + dest = GetUi32(sd.Data); SKIP_DATA2(sd, 4); + +static MY_NO_INLINE SRes ReadNumber(CSzData *sd, UInt64 *value) +{ + Byte firstByte, mask; + unsigned i; + UInt32 v; + + SZ_READ_BYTE(firstByte); + if ((firstByte & 0x80) == 0) + { + *value = firstByte; + return SZ_OK; + } + SZ_READ_BYTE(v); + if ((firstByte & 0x40) == 0) + { + *value = (((UInt32)firstByte & 0x3F) << 8) | v; + return SZ_OK; + } + SZ_READ_BYTE(mask); + *value = v | ((UInt32)mask << 8); + mask = 0x20; + for (i = 2; i < 8; i++) + { + Byte b; + if ((firstByte & mask) == 0) + { + UInt64 highPart = (unsigned)firstByte & (unsigned)(mask - 1); + *value |= (highPart << (8 * i)); + return SZ_OK; + } + SZ_READ_BYTE(b); + *value |= ((UInt64)b << (8 * i)); + mask >>= 1; + } + return SZ_OK; +} + + +static MY_NO_INLINE SRes SzReadNumber32(CSzData *sd, UInt32 *value) +{ + Byte firstByte; + UInt64 value64; + if (sd->Size == 0) + return SZ_ERROR_ARCHIVE; + firstByte = *sd->Data; + if ((firstByte & 0x80) == 0) + { + *value = firstByte; + sd->Data++; + sd->Size--; + return SZ_OK; + } + RINOK(ReadNumber(sd, &value64)); + if (value64 >= (UInt32)0x80000000 - 1) + return SZ_ERROR_UNSUPPORTED; + if (value64 >= ((UInt64)(1) << ((sizeof(size_t) - 1) * 8 + 4))) + return SZ_ERROR_UNSUPPORTED; + *value = (UInt32)value64; + return SZ_OK; +} + +#define ReadID(sd, value) ReadNumber(sd, value) + +static SRes SkipData(CSzData *sd) +{ + UInt64 size; + RINOK(ReadNumber(sd, &size)); + if (size > sd->Size) + return SZ_ERROR_ARCHIVE; + SKIP_DATA(sd, size); + return SZ_OK; +} + +static SRes WaitId(CSzData *sd, UInt32 id) +{ + for (;;) + { + UInt64 type; + RINOK(ReadID(sd, &type)); + if (type == id) + return SZ_OK; + if (type == k7zIdEnd) + return SZ_ERROR_ARCHIVE; + RINOK(SkipData(sd)); + } +} + +static SRes RememberBitVector(CSzData *sd, UInt32 numItems, const Byte **v) +{ + UInt32 numBytes = (numItems + 7) >> 3; + if (numBytes > sd->Size) + return SZ_ERROR_ARCHIVE; + *v = sd->Data; + SKIP_DATA(sd, numBytes); + return SZ_OK; +} + +static UInt32 CountDefinedBits(const Byte *bits, UInt32 numItems) +{ + Byte b = 0; + unsigned m = 0; + UInt32 sum = 0; + for (; numItems != 0; numItems--) + { + if (m == 0) + { + b = *bits++; + m = 8; + } + m--; + sum += ((b >> m) & 1); + } + return sum; +} + +static MY_NO_INLINE SRes ReadBitVector(CSzData *sd, UInt32 numItems, Byte **v, ISzAllocPtr alloc) +{ + Byte allAreDefined; + Byte *v2; + UInt32 numBytes = (numItems + 7) >> 3; + *v = NULL; + SZ_READ_BYTE(allAreDefined); + if (numBytes == 0) + return SZ_OK; + if (allAreDefined == 0) + { + if (numBytes > sd->Size) + return SZ_ERROR_ARCHIVE; + MY_ALLOC_AND_CPY(*v, numBytes, sd->Data, alloc); + SKIP_DATA(sd, numBytes); + return SZ_OK; + } + MY_ALLOC(Byte, *v, numBytes, alloc); + v2 = *v; + memset(v2, 0xFF, (size_t)numBytes); + { + unsigned numBits = (unsigned)numItems & 7; + if (numBits != 0) + v2[(size_t)numBytes - 1] = (Byte)((((UInt32)1 << numBits) - 1) << (8 - numBits)); + } + return SZ_OK; +} + +static MY_NO_INLINE SRes ReadUi32s(CSzData *sd2, UInt32 numItems, CSzBitUi32s *crcs, ISzAllocPtr alloc) +{ + UInt32 i; + CSzData sd; + UInt32 *vals; + const Byte *defs; + MY_ALLOC_ZE(UInt32, crcs->Vals, numItems, alloc); + sd = *sd2; + defs = crcs->Defs; + vals = crcs->Vals; + for (i = 0; i < numItems; i++) + if (SzBitArray_Check(defs, i)) + { + SZ_READ_32(vals[i]); + } + else + vals[i] = 0; + *sd2 = sd; + return SZ_OK; +} + +static SRes ReadBitUi32s(CSzData *sd, UInt32 numItems, CSzBitUi32s *crcs, ISzAllocPtr alloc) +{ + SzBitUi32s_Free(crcs, alloc); + RINOK(ReadBitVector(sd, numItems, &crcs->Defs, alloc)); + return ReadUi32s(sd, numItems, crcs, alloc); +} + +static SRes SkipBitUi32s(CSzData *sd, UInt32 numItems) +{ + Byte allAreDefined; + UInt32 numDefined = numItems; + SZ_READ_BYTE(allAreDefined); + if (!allAreDefined) + { + size_t numBytes = (numItems + 7) >> 3; + if (numBytes > sd->Size) + return SZ_ERROR_ARCHIVE; + numDefined = CountDefinedBits(sd->Data, numItems); + SKIP_DATA(sd, numBytes); + } + if (numDefined > (sd->Size >> 2)) + return SZ_ERROR_ARCHIVE; + SKIP_DATA(sd, (size_t)numDefined * 4); + return SZ_OK; +} + +static SRes ReadPackInfo(CSzAr *p, CSzData *sd, ISzAllocPtr alloc) +{ + RINOK(SzReadNumber32(sd, &p->NumPackStreams)); + + RINOK(WaitId(sd, k7zIdSize)); + MY_ALLOC(UInt64, p->PackPositions, (size_t)p->NumPackStreams + 1, alloc); + { + UInt64 sum = 0; + UInt32 i; + UInt32 numPackStreams = p->NumPackStreams; + for (i = 0; i < numPackStreams; i++) + { + UInt64 packSize; + p->PackPositions[i] = sum; + RINOK(ReadNumber(sd, &packSize)); + sum += packSize; + if (sum < packSize) + return SZ_ERROR_ARCHIVE; + } + p->PackPositions[i] = sum; + } + + for (;;) + { + UInt64 type; + RINOK(ReadID(sd, &type)); + if (type == k7zIdEnd) + return SZ_OK; + if (type == k7zIdCRC) + { + /* CRC of packed streams is unused now */ + RINOK(SkipBitUi32s(sd, p->NumPackStreams)); + continue; + } + RINOK(SkipData(sd)); + } +} + +/* +static SRes SzReadSwitch(CSzData *sd) +{ + Byte external; + RINOK(SzReadByte(sd, &external)); + return (external == 0) ? SZ_OK: SZ_ERROR_UNSUPPORTED; +} +*/ + +#define k_NumCodersStreams_in_Folder_MAX (SZ_NUM_BONDS_IN_FOLDER_MAX + SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX) + +SRes SzGetNextFolderItem(CSzFolder *f, CSzData *sd) +{ + UInt32 numCoders, i; + UInt32 numInStreams = 0; + const Byte *dataStart = sd->Data; + + f->NumCoders = 0; + f->NumBonds = 0; + f->NumPackStreams = 0; + f->UnpackStream = 0; + + RINOK(SzReadNumber32(sd, &numCoders)); + if (numCoders == 0 || numCoders > SZ_NUM_CODERS_IN_FOLDER_MAX) + return SZ_ERROR_UNSUPPORTED; + + for (i = 0; i < numCoders; i++) + { + Byte mainByte; + CSzCoderInfo *coder = f->Coders + i; + unsigned idSize, j; + UInt64 id; + + SZ_READ_BYTE(mainByte); + if ((mainByte & 0xC0) != 0) + return SZ_ERROR_UNSUPPORTED; + + idSize = (unsigned)(mainByte & 0xF); + if (idSize > sizeof(id)) + return SZ_ERROR_UNSUPPORTED; + if (idSize > sd->Size) + return SZ_ERROR_ARCHIVE; + id = 0; + for (j = 0; j < idSize; j++) + { + id = ((id << 8) | *sd->Data); + sd->Data++; + sd->Size--; + } + if (id > (UInt32)0xFFFFFFFF) + return SZ_ERROR_UNSUPPORTED; + coder->MethodID = (UInt32)id; + + coder->NumStreams = 1; + coder->PropsOffset = 0; + coder->PropsSize = 0; + + if ((mainByte & 0x10) != 0) + { + UInt32 numStreams; + + RINOK(SzReadNumber32(sd, &numStreams)); + if (numStreams > k_NumCodersStreams_in_Folder_MAX) + return SZ_ERROR_UNSUPPORTED; + coder->NumStreams = (Byte)numStreams; + + RINOK(SzReadNumber32(sd, &numStreams)); + if (numStreams != 1) + return SZ_ERROR_UNSUPPORTED; + } + + numInStreams += coder->NumStreams; + + if (numInStreams > k_NumCodersStreams_in_Folder_MAX) + return SZ_ERROR_UNSUPPORTED; + + if ((mainByte & 0x20) != 0) + { + UInt32 propsSize = 0; + RINOK(SzReadNumber32(sd, &propsSize)); + if (propsSize > sd->Size) + return SZ_ERROR_ARCHIVE; + if (propsSize >= 0x80) + return SZ_ERROR_UNSUPPORTED; + coder->PropsOffset = (size_t)(sd->Data - dataStart); + coder->PropsSize = (Byte)propsSize; + sd->Data += (size_t)propsSize; + sd->Size -= (size_t)propsSize; + } + } + + /* + if (numInStreams == 1 && numCoders == 1) + { + f->NumPackStreams = 1; + f->PackStreams[0] = 0; + } + else + */ + { + Byte streamUsed[k_NumCodersStreams_in_Folder_MAX]; + UInt32 numBonds, numPackStreams; + + numBonds = numCoders - 1; + if (numInStreams < numBonds) + return SZ_ERROR_ARCHIVE; + if (numBonds > SZ_NUM_BONDS_IN_FOLDER_MAX) + return SZ_ERROR_UNSUPPORTED; + f->NumBonds = numBonds; + + numPackStreams = numInStreams - numBonds; + if (numPackStreams > SZ_NUM_PACK_STREAMS_IN_FOLDER_MAX) + return SZ_ERROR_UNSUPPORTED; + f->NumPackStreams = numPackStreams; + + for (i = 0; i < numInStreams; i++) + streamUsed[i] = False; + + if (numBonds != 0) + { + Byte coderUsed[SZ_NUM_CODERS_IN_FOLDER_MAX]; + + for (i = 0; i < numCoders; i++) + coderUsed[i] = False; + + for (i = 0; i < numBonds; i++) + { + CSzBond *bp = f->Bonds + i; + + RINOK(SzReadNumber32(sd, &bp->InIndex)); + if (bp->InIndex >= numInStreams || streamUsed[bp->InIndex]) + return SZ_ERROR_ARCHIVE; + streamUsed[bp->InIndex] = True; + + RINOK(SzReadNumber32(sd, &bp->OutIndex)); + if (bp->OutIndex >= numCoders || coderUsed[bp->OutIndex]) + return SZ_ERROR_ARCHIVE; + coderUsed[bp->OutIndex] = True; + } + + for (i = 0; i < numCoders; i++) + if (!coderUsed[i]) + { + f->UnpackStream = i; + break; + } + + if (i == numCoders) + return SZ_ERROR_ARCHIVE; + } + + if (numPackStreams == 1) + { + for (i = 0; i < numInStreams; i++) + if (!streamUsed[i]) + break; + if (i == numInStreams) + return SZ_ERROR_ARCHIVE; + f->PackStreams[0] = i; + } + else + for (i = 0; i < numPackStreams; i++) + { + UInt32 index; + RINOK(SzReadNumber32(sd, &index)); + if (index >= numInStreams || streamUsed[index]) + return SZ_ERROR_ARCHIVE; + streamUsed[index] = True; + f->PackStreams[i] = index; + } + } + + f->NumCoders = numCoders; + + return SZ_OK; +} + + +static MY_NO_INLINE SRes SkipNumbers(CSzData *sd2, UInt32 num) +{ + CSzData sd; + sd = *sd2; + for (; num != 0; num--) + { + Byte firstByte, mask; + unsigned i; + SZ_READ_BYTE_2(firstByte); + if ((firstByte & 0x80) == 0) + continue; + if ((firstByte & 0x40) == 0) + { + if (sd.Size == 0) + return SZ_ERROR_ARCHIVE; + sd.Size--; + sd.Data++; + continue; + } + mask = 0x20; + for (i = 2; i < 8 && (firstByte & mask) != 0; i++) + mask >>= 1; + if (i > sd.Size) + return SZ_ERROR_ARCHIVE; + SKIP_DATA2(sd, i); + } + *sd2 = sd; + return SZ_OK; +} + + +#define k_Scan_NumCoders_MAX 64 +#define k_Scan_NumCodersStreams_in_Folder_MAX 64 + + +static SRes ReadUnpackInfo(CSzAr *p, + CSzData *sd2, + UInt32 numFoldersMax, + const CBuf *tempBufs, UInt32 numTempBufs, + ISzAllocPtr alloc) +{ + CSzData sd; + + UInt32 fo, numFolders, numCodersOutStreams, packStreamIndex; + const Byte *startBufPtr; + Byte external; + + RINOK(WaitId(sd2, k7zIdFolder)); + + RINOK(SzReadNumber32(sd2, &numFolders)); + if (numFolders > numFoldersMax) + return SZ_ERROR_UNSUPPORTED; + p->NumFolders = numFolders; + + SZ_READ_BYTE_SD(sd2, external); + if (external == 0) + sd = *sd2; + else + { + UInt32 index; + RINOK(SzReadNumber32(sd2, &index)); + if (index >= numTempBufs) + return SZ_ERROR_ARCHIVE; + sd.Data = tempBufs[index].data; + sd.Size = tempBufs[index].size; + } + + MY_ALLOC(size_t, p->FoCodersOffsets, (size_t)numFolders + 1, alloc); + MY_ALLOC(UInt32, p->FoStartPackStreamIndex, (size_t)numFolders + 1, alloc); + MY_ALLOC(UInt32, p->FoToCoderUnpackSizes, (size_t)numFolders + 1, alloc); + MY_ALLOC_ZE(Byte, p->FoToMainUnpackSizeIndex, (size_t)numFolders, alloc); + + startBufPtr = sd.Data; + + packStreamIndex = 0; + numCodersOutStreams = 0; + + for (fo = 0; fo < numFolders; fo++) + { + UInt32 numCoders, ci, numInStreams = 0; + + p->FoCodersOffsets[fo] = (size_t)(sd.Data - startBufPtr); + + RINOK(SzReadNumber32(&sd, &numCoders)); + if (numCoders == 0 || numCoders > k_Scan_NumCoders_MAX) + return SZ_ERROR_UNSUPPORTED; + + for (ci = 0; ci < numCoders; ci++) + { + Byte mainByte; + unsigned idSize; + UInt32 coderInStreams; + + SZ_READ_BYTE_2(mainByte); + if ((mainByte & 0xC0) != 0) + return SZ_ERROR_UNSUPPORTED; + idSize = (mainByte & 0xF); + if (idSize > 8) + return SZ_ERROR_UNSUPPORTED; + if (idSize > sd.Size) + return SZ_ERROR_ARCHIVE; + SKIP_DATA2(sd, idSize); + + coderInStreams = 1; + + if ((mainByte & 0x10) != 0) + { + UInt32 coderOutStreams; + RINOK(SzReadNumber32(&sd, &coderInStreams)); + RINOK(SzReadNumber32(&sd, &coderOutStreams)); + if (coderInStreams > k_Scan_NumCodersStreams_in_Folder_MAX || coderOutStreams != 1) + return SZ_ERROR_UNSUPPORTED; + } + + numInStreams += coderInStreams; + + if ((mainByte & 0x20) != 0) + { + UInt32 propsSize; + RINOK(SzReadNumber32(&sd, &propsSize)); + if (propsSize > sd.Size) + return SZ_ERROR_ARCHIVE; + SKIP_DATA2(sd, propsSize); + } + } + + { + UInt32 indexOfMainStream = 0; + UInt32 numPackStreams = 1; + + if (numCoders != 1 || numInStreams != 1) + { + Byte streamUsed[k_Scan_NumCodersStreams_in_Folder_MAX]; + Byte coderUsed[k_Scan_NumCoders_MAX]; + + UInt32 i; + UInt32 numBonds = numCoders - 1; + if (numInStreams < numBonds) + return SZ_ERROR_ARCHIVE; + + if (numInStreams > k_Scan_NumCodersStreams_in_Folder_MAX) + return SZ_ERROR_UNSUPPORTED; + + for (i = 0; i < numInStreams; i++) + streamUsed[i] = False; + for (i = 0; i < numCoders; i++) + coderUsed[i] = False; + + for (i = 0; i < numBonds; i++) + { + UInt32 index; + + RINOK(SzReadNumber32(&sd, &index)); + if (index >= numInStreams || streamUsed[index]) + return SZ_ERROR_ARCHIVE; + streamUsed[index] = True; + + RINOK(SzReadNumber32(&sd, &index)); + if (index >= numCoders || coderUsed[index]) + return SZ_ERROR_ARCHIVE; + coderUsed[index] = True; + } + + numPackStreams = numInStreams - numBonds; + + if (numPackStreams != 1) + for (i = 0; i < numPackStreams; i++) + { + UInt32 index; + RINOK(SzReadNumber32(&sd, &index)); + if (index >= numInStreams || streamUsed[index]) + return SZ_ERROR_ARCHIVE; + streamUsed[index] = True; + } + + for (i = 0; i < numCoders; i++) + if (!coderUsed[i]) + { + indexOfMainStream = i; + break; + } + + if (i == numCoders) + return SZ_ERROR_ARCHIVE; + } + + p->FoStartPackStreamIndex[fo] = packStreamIndex; + p->FoToCoderUnpackSizes[fo] = numCodersOutStreams; + p->FoToMainUnpackSizeIndex[fo] = (Byte)indexOfMainStream; + numCodersOutStreams += numCoders; + if (numCodersOutStreams < numCoders) + return SZ_ERROR_UNSUPPORTED; + if (numPackStreams > p->NumPackStreams - packStreamIndex) + return SZ_ERROR_ARCHIVE; + packStreamIndex += numPackStreams; + } + } + + p->FoToCoderUnpackSizes[fo] = numCodersOutStreams; + + { + const size_t dataSize = (size_t)(sd.Data - startBufPtr); + p->FoStartPackStreamIndex[fo] = packStreamIndex; + p->FoCodersOffsets[fo] = dataSize; + MY_ALLOC_ZE_AND_CPY(p->CodersData, dataSize, startBufPtr, alloc); + } + + if (external != 0) + { + if (sd.Size != 0) + return SZ_ERROR_ARCHIVE; + sd = *sd2; + } + + RINOK(WaitId(&sd, k7zIdCodersUnpackSize)); + + MY_ALLOC_ZE(UInt64, p->CoderUnpackSizes, (size_t)numCodersOutStreams, alloc); + { + UInt32 i; + for (i = 0; i < numCodersOutStreams; i++) + { + RINOK(ReadNumber(&sd, p->CoderUnpackSizes + i)); + } + } + + for (;;) + { + UInt64 type; + RINOK(ReadID(&sd, &type)); + if (type == k7zIdEnd) + { + *sd2 = sd; + return SZ_OK; + } + if (type == k7zIdCRC) + { + RINOK(ReadBitUi32s(&sd, numFolders, &p->FolderCRCs, alloc)); + continue; + } + RINOK(SkipData(&sd)); + } +} + + +UInt64 SzAr_GetFolderUnpackSize(const CSzAr *p, UInt32 folderIndex) +{ + return p->CoderUnpackSizes[p->FoToCoderUnpackSizes[folderIndex] + p->FoToMainUnpackSizeIndex[folderIndex]]; +} + + +typedef struct +{ + UInt32 NumTotalSubStreams; + UInt32 NumSubDigests; + CSzData sdNumSubStreams; + CSzData sdSizes; + CSzData sdCRCs; +} CSubStreamInfo; + + +static SRes ReadSubStreamsInfo(CSzAr *p, CSzData *sd, CSubStreamInfo *ssi) +{ + UInt64 type = 0; + UInt32 numSubDigests = 0; + UInt32 numFolders = p->NumFolders; + UInt32 numUnpackStreams = numFolders; + UInt32 numUnpackSizesInData = 0; + + for (;;) + { + RINOK(ReadID(sd, &type)); + if (type == k7zIdNumUnpackStream) + { + UInt32 i; + ssi->sdNumSubStreams.Data = sd->Data; + numUnpackStreams = 0; + numSubDigests = 0; + for (i = 0; i < numFolders; i++) + { + UInt32 numStreams; + RINOK(SzReadNumber32(sd, &numStreams)); + if (numUnpackStreams > numUnpackStreams + numStreams) + return SZ_ERROR_UNSUPPORTED; + numUnpackStreams += numStreams; + if (numStreams != 0) + numUnpackSizesInData += (numStreams - 1); + if (numStreams != 1 || !SzBitWithVals_Check(&p->FolderCRCs, i)) + numSubDigests += numStreams; + } + ssi->sdNumSubStreams.Size = (size_t)(sd->Data - ssi->sdNumSubStreams.Data); + continue; + } + if (type == k7zIdCRC || type == k7zIdSize || type == k7zIdEnd) + break; + RINOK(SkipData(sd)); + } + + if (!ssi->sdNumSubStreams.Data) + { + numSubDigests = numFolders; + if (p->FolderCRCs.Defs) + numSubDigests = numFolders - CountDefinedBits(p->FolderCRCs.Defs, numFolders); + } + + ssi->NumTotalSubStreams = numUnpackStreams; + ssi->NumSubDigests = numSubDigests; + + if (type == k7zIdSize) + { + ssi->sdSizes.Data = sd->Data; + RINOK(SkipNumbers(sd, numUnpackSizesInData)); + ssi->sdSizes.Size = (size_t)(sd->Data - ssi->sdSizes.Data); + RINOK(ReadID(sd, &type)); + } + + for (;;) + { + if (type == k7zIdEnd) + return SZ_OK; + if (type == k7zIdCRC) + { + ssi->sdCRCs.Data = sd->Data; + RINOK(SkipBitUi32s(sd, numSubDigests)); + ssi->sdCRCs.Size = (size_t)(sd->Data - ssi->sdCRCs.Data); + } + else + { + RINOK(SkipData(sd)); + } + RINOK(ReadID(sd, &type)); + } +} + +static SRes SzReadStreamsInfo(CSzAr *p, + CSzData *sd, + UInt32 numFoldersMax, const CBuf *tempBufs, UInt32 numTempBufs, + UInt64 *dataOffset, + CSubStreamInfo *ssi, + ISzAllocPtr alloc) +{ + UInt64 type; + + SzData_Clear(&ssi->sdSizes); + SzData_Clear(&ssi->sdCRCs); + SzData_Clear(&ssi->sdNumSubStreams); + + *dataOffset = 0; + RINOK(ReadID(sd, &type)); + if (type == k7zIdPackInfo) + { + RINOK(ReadNumber(sd, dataOffset)); + if (*dataOffset > p->RangeLimit) + return SZ_ERROR_ARCHIVE; + RINOK(ReadPackInfo(p, sd, alloc)); + if (p->PackPositions[p->NumPackStreams] > p->RangeLimit - *dataOffset) + return SZ_ERROR_ARCHIVE; + RINOK(ReadID(sd, &type)); + } + if (type == k7zIdUnpackInfo) + { + RINOK(ReadUnpackInfo(p, sd, numFoldersMax, tempBufs, numTempBufs, alloc)); + RINOK(ReadID(sd, &type)); + } + if (type == k7zIdSubStreamsInfo) + { + RINOK(ReadSubStreamsInfo(p, sd, ssi)); + RINOK(ReadID(sd, &type)); + } + else + { + ssi->NumTotalSubStreams = p->NumFolders; + // ssi->NumSubDigests = 0; + } + + return (type == k7zIdEnd ? SZ_OK : SZ_ERROR_UNSUPPORTED); +} + +static SRes SzReadAndDecodePackedStreams( + ILookInStream *inStream, + CSzData *sd, + CBuf *tempBufs, + UInt32 numFoldersMax, + UInt64 baseOffset, + CSzAr *p, + ISzAllocPtr allocTemp) +{ + UInt64 dataStartPos; + UInt32 fo; + CSubStreamInfo ssi; + + RINOK(SzReadStreamsInfo(p, sd, numFoldersMax, NULL, 0, &dataStartPos, &ssi, allocTemp)); + + dataStartPos += baseOffset; + if (p->NumFolders == 0) + return SZ_ERROR_ARCHIVE; + + for (fo = 0; fo < p->NumFolders; fo++) + Buf_Init(tempBufs + fo); + + for (fo = 0; fo < p->NumFolders; fo++) + { + CBuf *tempBuf = tempBufs + fo; + UInt64 unpackSize = SzAr_GetFolderUnpackSize(p, fo); + if ((size_t)unpackSize != unpackSize) + return SZ_ERROR_MEM; + if (!Buf_Create(tempBuf, (size_t)unpackSize, allocTemp)) + return SZ_ERROR_MEM; + } + + for (fo = 0; fo < p->NumFolders; fo++) + { + const CBuf *tempBuf = tempBufs + fo; + RINOK(LookInStream_SeekTo(inStream, dataStartPos)); + RINOK(SzAr_DecodeFolder(p, fo, inStream, dataStartPos, tempBuf->data, tempBuf->size, allocTemp)); + } + + return SZ_OK; +} + +static SRes SzReadFileNames(const Byte *data, size_t size, UInt32 numFiles, size_t *offsets) +{ + size_t pos = 0; + *offsets++ = 0; + if (numFiles == 0) + return (size == 0) ? SZ_OK : SZ_ERROR_ARCHIVE; + if (size < 2) + return SZ_ERROR_ARCHIVE; + if (data[size - 2] != 0 || data[size - 1] != 0) + return SZ_ERROR_ARCHIVE; + do + { + const Byte *p; + if (pos == size) + return SZ_ERROR_ARCHIVE; + for (p = data + pos; + #ifdef _WIN32 + *(const UInt16 *)(const void *)p != 0 + #else + p[0] != 0 || p[1] != 0 + #endif + ; p += 2); + pos = (size_t)(p - data) + 2; + *offsets++ = (pos >> 1); + } + while (--numFiles); + return (pos == size) ? SZ_OK : SZ_ERROR_ARCHIVE; +} + +static MY_NO_INLINE SRes ReadTime(CSzBitUi64s *p, UInt32 num, + CSzData *sd2, + const CBuf *tempBufs, UInt32 numTempBufs, + ISzAllocPtr alloc) +{ + CSzData sd; + UInt32 i; + CNtfsFileTime *vals; + Byte *defs; + Byte external; + + RINOK(ReadBitVector(sd2, num, &p->Defs, alloc)); + + SZ_READ_BYTE_SD(sd2, external); + if (external == 0) + sd = *sd2; + else + { + UInt32 index; + RINOK(SzReadNumber32(sd2, &index)); + if (index >= numTempBufs) + return SZ_ERROR_ARCHIVE; + sd.Data = tempBufs[index].data; + sd.Size = tempBufs[index].size; + } + + MY_ALLOC_ZE(CNtfsFileTime, p->Vals, num, alloc); + vals = p->Vals; + defs = p->Defs; + for (i = 0; i < num; i++) + if (SzBitArray_Check(defs, i)) + { + if (sd.Size < 8) + return SZ_ERROR_ARCHIVE; + vals[i].Low = GetUi32(sd.Data); + vals[i].High = GetUi32(sd.Data + 4); + SKIP_DATA2(sd, 8); + } + else + vals[i].High = vals[i].Low = 0; + + if (external == 0) + *sd2 = sd; + + return SZ_OK; +} + + +#define NUM_ADDITIONAL_STREAMS_MAX 8 + + +static SRes SzReadHeader2( + CSzArEx *p, /* allocMain */ + CSzData *sd, + ILookInStream *inStream, + CBuf *tempBufs, UInt32 *numTempBufs, + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp + ) +{ + CSubStreamInfo ssi; + +{ + UInt64 type; + + SzData_Clear(&ssi.sdSizes); + SzData_Clear(&ssi.sdCRCs); + SzData_Clear(&ssi.sdNumSubStreams); + + ssi.NumSubDigests = 0; + ssi.NumTotalSubStreams = 0; + + RINOK(ReadID(sd, &type)); + + if (type == k7zIdArchiveProperties) + { + for (;;) + { + UInt64 type2; + RINOK(ReadID(sd, &type2)); + if (type2 == k7zIdEnd) + break; + RINOK(SkipData(sd)); + } + RINOK(ReadID(sd, &type)); + } + + if (type == k7zIdAdditionalStreamsInfo) + { + CSzAr tempAr; + SRes res; + + SzAr_Init(&tempAr); + tempAr.RangeLimit = p->db.RangeLimit; + + res = SzReadAndDecodePackedStreams(inStream, sd, tempBufs, NUM_ADDITIONAL_STREAMS_MAX, + p->startPosAfterHeader, &tempAr, allocTemp); + *numTempBufs = tempAr.NumFolders; + SzAr_Free(&tempAr, allocTemp); + + if (res != SZ_OK) + return res; + RINOK(ReadID(sd, &type)); + } + + if (type == k7zIdMainStreamsInfo) + { + RINOK(SzReadStreamsInfo(&p->db, sd, (UInt32)1 << 30, tempBufs, *numTempBufs, + &p->dataPos, &ssi, allocMain)); + p->dataPos += p->startPosAfterHeader; + RINOK(ReadID(sd, &type)); + } + + if (type == k7zIdEnd) + { + return SZ_OK; + } + + if (type != k7zIdFilesInfo) + return SZ_ERROR_ARCHIVE; +} + +{ + UInt32 numFiles = 0; + UInt32 numEmptyStreams = 0; + const Byte *emptyStreams = NULL; + const Byte *emptyFiles = NULL; + + RINOK(SzReadNumber32(sd, &numFiles)); + p->NumFiles = numFiles; + + for (;;) + { + UInt64 type; + UInt64 size; + RINOK(ReadID(sd, &type)); + if (type == k7zIdEnd) + break; + RINOK(ReadNumber(sd, &size)); + if (size > sd->Size) + return SZ_ERROR_ARCHIVE; + + if (type >= ((UInt32)1 << 8)) + { + SKIP_DATA(sd, size); + } + else switch ((unsigned)type) + { + case k7zIdName: + { + size_t namesSize; + const Byte *namesData; + Byte external; + + SZ_READ_BYTE(external); + if (external == 0) + { + namesSize = (size_t)size - 1; + namesData = sd->Data; + } + else + { + UInt32 index; + RINOK(SzReadNumber32(sd, &index)); + if (index >= *numTempBufs) + return SZ_ERROR_ARCHIVE; + namesData = (tempBufs)[index].data; + namesSize = (tempBufs)[index].size; + } + + if ((namesSize & 1) != 0) + return SZ_ERROR_ARCHIVE; + MY_ALLOC(size_t, p->FileNameOffsets, numFiles + 1, allocMain); + MY_ALLOC_ZE_AND_CPY(p->FileNames, namesSize, namesData, allocMain); + RINOK(SzReadFileNames(p->FileNames, namesSize, numFiles, p->FileNameOffsets)) + if (external == 0) + { + SKIP_DATA(sd, namesSize); + } + break; + } + case k7zIdEmptyStream: + { + RINOK(RememberBitVector(sd, numFiles, &emptyStreams)); + numEmptyStreams = CountDefinedBits(emptyStreams, numFiles); + emptyFiles = NULL; + break; + } + case k7zIdEmptyFile: + { + RINOK(RememberBitVector(sd, numEmptyStreams, &emptyFiles)); + break; + } + case k7zIdWinAttrib: + { + Byte external; + CSzData sdSwitch; + CSzData *sdPtr; + SzBitUi32s_Free(&p->Attribs, allocMain); + RINOK(ReadBitVector(sd, numFiles, &p->Attribs.Defs, allocMain)); + + SZ_READ_BYTE(external); + if (external == 0) + sdPtr = sd; + else + { + UInt32 index; + RINOK(SzReadNumber32(sd, &index)); + if (index >= *numTempBufs) + return SZ_ERROR_ARCHIVE; + sdSwitch.Data = (tempBufs)[index].data; + sdSwitch.Size = (tempBufs)[index].size; + sdPtr = &sdSwitch; + } + RINOK(ReadUi32s(sdPtr, numFiles, &p->Attribs, allocMain)); + break; + } + /* + case k7zParent: + { + SzBitUi32s_Free(&p->Parents, allocMain); + RINOK(ReadBitVector(sd, numFiles, &p->Parents.Defs, allocMain)); + RINOK(SzReadSwitch(sd)); + RINOK(ReadUi32s(sd, numFiles, &p->Parents, allocMain)); + break; + } + */ + case k7zIdMTime: RINOK(ReadTime(&p->MTime, numFiles, sd, tempBufs, *numTempBufs, allocMain)); break; + case k7zIdCTime: RINOK(ReadTime(&p->CTime, numFiles, sd, tempBufs, *numTempBufs, allocMain)); break; + default: + { + SKIP_DATA(sd, size); + } + } + } + + if (numFiles - numEmptyStreams != ssi.NumTotalSubStreams) + return SZ_ERROR_ARCHIVE; + + for (;;) + { + UInt64 type; + RINOK(ReadID(sd, &type)); + if (type == k7zIdEnd) + break; + RINOK(SkipData(sd)); + } + + { + UInt32 i; + UInt32 emptyFileIndex = 0; + UInt32 folderIndex = 0; + UInt32 remSubStreams = 0; + UInt32 numSubStreams = 0; + UInt64 unpackPos = 0; + const Byte *digestsDefs = NULL; + const Byte *digestsVals = NULL; + UInt32 digestsValsIndex = 0; + UInt32 digestIndex; + Byte allDigestsDefined = 0; + Byte isDirMask = 0; + Byte crcMask = 0; + Byte mask = 0x80; + + MY_ALLOC(UInt32, p->FolderToFile, p->db.NumFolders + 1, allocMain); + MY_ALLOC_ZE(UInt32, p->FileToFolder, p->NumFiles, allocMain); + MY_ALLOC(UInt64, p->UnpackPositions, p->NumFiles + 1, allocMain); + MY_ALLOC_ZE(Byte, p->IsDirs, (p->NumFiles + 7) >> 3, allocMain); + + RINOK(SzBitUi32s_Alloc(&p->CRCs, p->NumFiles, allocMain)); + + if (ssi.sdCRCs.Size != 0) + { + SZ_READ_BYTE_SD(&ssi.sdCRCs, allDigestsDefined); + if (allDigestsDefined) + digestsVals = ssi.sdCRCs.Data; + else + { + size_t numBytes = (ssi.NumSubDigests + 7) >> 3; + digestsDefs = ssi.sdCRCs.Data; + digestsVals = digestsDefs + numBytes; + } + } + + digestIndex = 0; + + for (i = 0; i < numFiles; i++, mask >>= 1) + { + if (mask == 0) + { + UInt32 byteIndex = (i - 1) >> 3; + p->IsDirs[byteIndex] = isDirMask; + p->CRCs.Defs[byteIndex] = crcMask; + isDirMask = 0; + crcMask = 0; + mask = 0x80; + } + + p->UnpackPositions[i] = unpackPos; + p->CRCs.Vals[i] = 0; + + if (emptyStreams && SzBitArray_Check(emptyStreams, i)) + { + if (emptyFiles) + { + if (!SzBitArray_Check(emptyFiles, emptyFileIndex)) + isDirMask |= mask; + emptyFileIndex++; + } + else + isDirMask |= mask; + if (remSubStreams == 0) + { + p->FileToFolder[i] = (UInt32)-1; + continue; + } + } + + if (remSubStreams == 0) + { + for (;;) + { + if (folderIndex >= p->db.NumFolders) + return SZ_ERROR_ARCHIVE; + p->FolderToFile[folderIndex] = i; + numSubStreams = 1; + if (ssi.sdNumSubStreams.Data) + { + RINOK(SzReadNumber32(&ssi.sdNumSubStreams, &numSubStreams)); + } + remSubStreams = numSubStreams; + if (numSubStreams != 0) + break; + { + UInt64 folderUnpackSize = SzAr_GetFolderUnpackSize(&p->db, folderIndex); + unpackPos += folderUnpackSize; + if (unpackPos < folderUnpackSize) + return SZ_ERROR_ARCHIVE; + } + + folderIndex++; + } + } + + p->FileToFolder[i] = folderIndex; + + if (emptyStreams && SzBitArray_Check(emptyStreams, i)) + continue; + + if (--remSubStreams == 0) + { + UInt64 folderUnpackSize = SzAr_GetFolderUnpackSize(&p->db, folderIndex); + UInt64 startFolderUnpackPos = p->UnpackPositions[p->FolderToFile[folderIndex]]; + if (folderUnpackSize < unpackPos - startFolderUnpackPos) + return SZ_ERROR_ARCHIVE; + unpackPos = startFolderUnpackPos + folderUnpackSize; + if (unpackPos < folderUnpackSize) + return SZ_ERROR_ARCHIVE; + + if (numSubStreams == 1 && SzBitWithVals_Check(&p->db.FolderCRCs, i)) + { + p->CRCs.Vals[i] = p->db.FolderCRCs.Vals[folderIndex]; + crcMask |= mask; + } + else if (allDigestsDefined || (digestsDefs && SzBitArray_Check(digestsDefs, digestIndex))) + { + p->CRCs.Vals[i] = GetUi32(digestsVals + (size_t)digestsValsIndex * 4); + digestsValsIndex++; + crcMask |= mask; + } + + folderIndex++; + } + else + { + UInt64 v; + RINOK(ReadNumber(&ssi.sdSizes, &v)); + unpackPos += v; + if (unpackPos < v) + return SZ_ERROR_ARCHIVE; + if (allDigestsDefined || (digestsDefs && SzBitArray_Check(digestsDefs, digestIndex))) + { + p->CRCs.Vals[i] = GetUi32(digestsVals + (size_t)digestsValsIndex * 4); + digestsValsIndex++; + crcMask |= mask; + } + } + } + + if (mask != 0x80) + { + UInt32 byteIndex = (i - 1) >> 3; + p->IsDirs[byteIndex] = isDirMask; + p->CRCs.Defs[byteIndex] = crcMask; + } + + p->UnpackPositions[i] = unpackPos; + + if (remSubStreams != 0) + return SZ_ERROR_ARCHIVE; + + for (;;) + { + p->FolderToFile[folderIndex] = i; + if (folderIndex >= p->db.NumFolders) + break; + if (!ssi.sdNumSubStreams.Data) + return SZ_ERROR_ARCHIVE; + RINOK(SzReadNumber32(&ssi.sdNumSubStreams, &numSubStreams)); + if (numSubStreams != 0) + return SZ_ERROR_ARCHIVE; + /* + { + UInt64 folderUnpackSize = SzAr_GetFolderUnpackSize(&p->db, folderIndex); + unpackPos += folderUnpackSize; + if (unpackPos < folderUnpackSize) + return SZ_ERROR_ARCHIVE; + } + */ + folderIndex++; + } + + if (ssi.sdNumSubStreams.Data && ssi.sdNumSubStreams.Size != 0) + return SZ_ERROR_ARCHIVE; + } +} + return SZ_OK; +} + + +static SRes SzReadHeader( + CSzArEx *p, + CSzData *sd, + ILookInStream *inStream, + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp) +{ + UInt32 i; + UInt32 numTempBufs = 0; + SRes res; + CBuf tempBufs[NUM_ADDITIONAL_STREAMS_MAX]; + + for (i = 0; i < NUM_ADDITIONAL_STREAMS_MAX; i++) + Buf_Init(tempBufs + i); + + res = SzReadHeader2(p, sd, inStream, + tempBufs, &numTempBufs, + allocMain, allocTemp); + + for (i = 0; i < NUM_ADDITIONAL_STREAMS_MAX; i++) + Buf_Free(tempBufs + i, allocTemp); + + RINOK(res); + + if (sd->Size != 0) + return SZ_ERROR_FAIL; + + return res; +} + +static SRes SzArEx_Open2( + CSzArEx *p, + ILookInStream *inStream, + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp) +{ + Byte header[k7zStartHeaderSize]; + Int64 startArcPos; + UInt64 nextHeaderOffset, nextHeaderSize; + size_t nextHeaderSizeT; + UInt32 nextHeaderCRC; + CBuf buf; + SRes res; + + startArcPos = 0; + RINOK(ILookInStream_Seek(inStream, &startArcPos, SZ_SEEK_CUR)); + + RINOK(LookInStream_Read2(inStream, header, k7zStartHeaderSize, SZ_ERROR_NO_ARCHIVE)); + + if (!TestSignatureCandidate(header)) + return SZ_ERROR_NO_ARCHIVE; + if (header[6] != k7zMajorVersion) + return SZ_ERROR_UNSUPPORTED; + + nextHeaderOffset = GetUi64(header + 12); + nextHeaderSize = GetUi64(header + 20); + nextHeaderCRC = GetUi32(header + 28); + + p->startPosAfterHeader = (UInt64)startArcPos + k7zStartHeaderSize; + + if (CrcCalc(header + 12, 20) != GetUi32(header + 8)) + return SZ_ERROR_CRC; + + p->db.RangeLimit = nextHeaderOffset; + + nextHeaderSizeT = (size_t)nextHeaderSize; + if (nextHeaderSizeT != nextHeaderSize) + return SZ_ERROR_MEM; + if (nextHeaderSizeT == 0) + return SZ_OK; + if (nextHeaderOffset > nextHeaderOffset + nextHeaderSize || + nextHeaderOffset > nextHeaderOffset + nextHeaderSize + k7zStartHeaderSize) + return SZ_ERROR_NO_ARCHIVE; + + { + Int64 pos = 0; + RINOK(ILookInStream_Seek(inStream, &pos, SZ_SEEK_END)); + if ((UInt64)pos < (UInt64)startArcPos + nextHeaderOffset || + (UInt64)pos < (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset || + (UInt64)pos < (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset + nextHeaderSize) + return SZ_ERROR_INPUT_EOF; + } + + RINOK(LookInStream_SeekTo(inStream, (UInt64)startArcPos + k7zStartHeaderSize + nextHeaderOffset)); + + if (!Buf_Create(&buf, nextHeaderSizeT, allocTemp)) + return SZ_ERROR_MEM; + + res = LookInStream_Read(inStream, buf.data, nextHeaderSizeT); + + if (res == SZ_OK) + { + res = SZ_ERROR_ARCHIVE; + if (CrcCalc(buf.data, nextHeaderSizeT) == nextHeaderCRC) + { + CSzData sd; + UInt64 type; + sd.Data = buf.data; + sd.Size = buf.size; + + res = ReadID(&sd, &type); + + if (res == SZ_OK && type == k7zIdEncodedHeader) + { + CSzAr tempAr; + CBuf tempBuf; + Buf_Init(&tempBuf); + + SzAr_Init(&tempAr); + tempAr.RangeLimit = p->db.RangeLimit; + + res = SzReadAndDecodePackedStreams(inStream, &sd, &tempBuf, 1, p->startPosAfterHeader, &tempAr, allocTemp); + SzAr_Free(&tempAr, allocTemp); + + if (res != SZ_OK) + { + Buf_Free(&tempBuf, allocTemp); + } + else + { + Buf_Free(&buf, allocTemp); + buf.data = tempBuf.data; + buf.size = tempBuf.size; + sd.Data = buf.data; + sd.Size = buf.size; + res = ReadID(&sd, &type); + } + } + + if (res == SZ_OK) + { + if (type == k7zIdHeader) + { + /* + CSzData sd2; + unsigned ttt; + for (ttt = 0; ttt < 40000; ttt++) + { + SzArEx_Free(p, allocMain); + sd2 = sd; + res = SzReadHeader(p, &sd2, inStream, allocMain, allocTemp); + if (res != SZ_OK) + break; + } + */ + res = SzReadHeader(p, &sd, inStream, allocMain, allocTemp); + } + else + res = SZ_ERROR_UNSUPPORTED; + } + } + } + + Buf_Free(&buf, allocTemp); + return res; +} + + +SRes SzArEx_Open(CSzArEx *p, ILookInStream *inStream, + ISzAllocPtr allocMain, ISzAllocPtr allocTemp) +{ + SRes res = SzArEx_Open2(p, inStream, allocMain, allocTemp); + if (res != SZ_OK) + SzArEx_Free(p, allocMain); + return res; +} + + +SRes SzArEx_Extract( + const CSzArEx *p, + ILookInStream *inStream, + UInt32 fileIndex, + UInt32 *blockIndex, + Byte **tempBuf, + size_t *outBufferSize, + size_t *offset, + size_t *outSizeProcessed, + ISzAllocPtr allocMain, + ISzAllocPtr allocTemp) +{ + UInt32 folderIndex = p->FileToFolder[fileIndex]; + SRes res = SZ_OK; + + *offset = 0; + *outSizeProcessed = 0; + + if (folderIndex == (UInt32)-1) + { + ISzAlloc_Free(allocMain, *tempBuf); + *blockIndex = folderIndex; + *tempBuf = NULL; + *outBufferSize = 0; + return SZ_OK; + } + + if (*tempBuf == NULL || *blockIndex != folderIndex) + { + UInt64 unpackSizeSpec = SzAr_GetFolderUnpackSize(&p->db, folderIndex); + /* + UInt64 unpackSizeSpec = + p->UnpackPositions[p->FolderToFile[(size_t)folderIndex + 1]] - + p->UnpackPositions[p->FolderToFile[folderIndex]]; + */ + size_t unpackSize = (size_t)unpackSizeSpec; + + if (unpackSize != unpackSizeSpec) + return SZ_ERROR_MEM; + *blockIndex = folderIndex; + ISzAlloc_Free(allocMain, *tempBuf); + *tempBuf = NULL; + + if (res == SZ_OK) + { + *outBufferSize = unpackSize; + if (unpackSize != 0) + { + *tempBuf = (Byte *)ISzAlloc_Alloc(allocMain, unpackSize); + if (*tempBuf == NULL) + res = SZ_ERROR_MEM; + } + + if (res == SZ_OK) + { + res = SzAr_DecodeFolder(&p->db, folderIndex, + inStream, p->dataPos, *tempBuf, unpackSize, allocTemp); + } + } + } + + if (res == SZ_OK) + { + UInt64 unpackPos = p->UnpackPositions[fileIndex]; + *offset = (size_t)(unpackPos - p->UnpackPositions[p->FolderToFile[folderIndex]]); + *outSizeProcessed = (size_t)(p->UnpackPositions[(size_t)fileIndex + 1] - unpackPos); + if (*offset + *outSizeProcessed > *outBufferSize) + return SZ_ERROR_FAIL; + if (SzBitWithVals_Check(&p->CRCs, fileIndex)) + if (CrcCalc(*tempBuf + *offset, *outSizeProcessed) != p->CRCs.Vals[fileIndex]) + res = SZ_ERROR_CRC; + } + + return res; +} + + +size_t SzArEx_GetFileNameUtf16(const CSzArEx *p, size_t fileIndex, UInt16 *dest) +{ + size_t offs = p->FileNameOffsets[fileIndex]; + size_t len = p->FileNameOffsets[fileIndex + 1] - offs; + if (dest != 0) + { + size_t i; + const Byte *src = p->FileNames + offs * 2; + for (i = 0; i < len; i++) + dest[i] = GetUi16(src + i * 2); + } + return len; +} + +/* +size_t SzArEx_GetFullNameLen(const CSzArEx *p, size_t fileIndex) +{ + size_t len; + if (!p->FileNameOffsets) + return 1; + len = 0; + for (;;) + { + UInt32 parent = (UInt32)(Int32)-1; + len += p->FileNameOffsets[fileIndex + 1] - p->FileNameOffsets[fileIndex]; + if SzBitWithVals_Check(&p->Parents, fileIndex) + parent = p->Parents.Vals[fileIndex]; + if (parent == (UInt32)(Int32)-1) + return len; + fileIndex = parent; + } +} + +UInt16 *SzArEx_GetFullNameUtf16_Back(const CSzArEx *p, size_t fileIndex, UInt16 *dest) +{ + BoolInt needSlash; + if (!p->FileNameOffsets) + { + *(--dest) = 0; + return dest; + } + needSlash = False; + for (;;) + { + UInt32 parent = (UInt32)(Int32)-1; + size_t curLen = p->FileNameOffsets[fileIndex + 1] - p->FileNameOffsets[fileIndex]; + SzArEx_GetFileNameUtf16(p, fileIndex, dest - curLen); + if (needSlash) + *(dest - 1) = '/'; + needSlash = True; + dest -= curLen; + + if SzBitWithVals_Check(&p->Parents, fileIndex) + parent = p->Parents.Vals[fileIndex]; + if (parent == (UInt32)(Int32)-1) + return dest; + fileIndex = parent; + } +} +*/ diff --git a/3rdparty/lzma/src/7zBuf.c b/3rdparty/lzma/src/7zBuf.c new file mode 100644 index 0000000000..8865c32a81 --- /dev/null +++ b/3rdparty/lzma/src/7zBuf.c @@ -0,0 +1,36 @@ +/* 7zBuf.c -- Byte Buffer +2017-04-03 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "7zBuf.h" + +void Buf_Init(CBuf *p) +{ + p->data = 0; + p->size = 0; +} + +int Buf_Create(CBuf *p, size_t size, ISzAllocPtr alloc) +{ + p->size = 0; + if (size == 0) + { + p->data = 0; + return 1; + } + p->data = (Byte *)ISzAlloc_Alloc(alloc, size); + if (p->data) + { + p->size = size; + return 1; + } + return 0; +} + +void Buf_Free(CBuf *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->data); + p->data = 0; + p->size = 0; +} diff --git a/3rdparty/lzma/src/7zBuf2.c b/3rdparty/lzma/src/7zBuf2.c new file mode 100644 index 0000000000..2083474165 --- /dev/null +++ b/3rdparty/lzma/src/7zBuf2.c @@ -0,0 +1,52 @@ +/* 7zBuf2.c -- Byte Buffer +2017-04-03 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7zBuf.h" + +void DynBuf_Construct(CDynBuf *p) +{ + p->data = 0; + p->size = 0; + p->pos = 0; +} + +void DynBuf_SeekToBeg(CDynBuf *p) +{ + p->pos = 0; +} + +int DynBuf_Write(CDynBuf *p, const Byte *buf, size_t size, ISzAllocPtr alloc) +{ + if (size > p->size - p->pos) + { + size_t newSize = p->pos + size; + Byte *data; + newSize += newSize / 4; + data = (Byte *)ISzAlloc_Alloc(alloc, newSize); + if (!data) + return 0; + p->size = newSize; + if (p->pos != 0) + memcpy(data, p->data, p->pos); + ISzAlloc_Free(alloc, p->data); + p->data = data; + } + if (size != 0) + { + memcpy(p->data + p->pos, buf, size); + p->pos += size; + } + return 1; +} + +void DynBuf_Free(CDynBuf *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->data); + p->data = 0; + p->size = 0; + p->pos = 0; +} diff --git a/3rdparty/lzma/src/7zCrc.c b/3rdparty/lzma/src/7zCrc.c new file mode 100644 index 0000000000..f186324ddc --- /dev/null +++ b/3rdparty/lzma/src/7zCrc.c @@ -0,0 +1,322 @@ +/* 7zCrc.c -- CRC32 init +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "7zCrc.h" +#include "CpuArch.h" + +#define kCrcPoly 0xEDB88320 + +#ifdef MY_CPU_LE + #define CRC_NUM_TABLES 8 +#else + #define CRC_NUM_TABLES 9 + + #define CRC_UINT32_SWAP(v) ((v >> 24) | ((v >> 8) & 0xFF00) | ((v << 8) & 0xFF0000) | (v << 24)) + + UInt32 MY_FAST_CALL CrcUpdateT1_BeT4(UInt32 v, const void *data, size_t size, const UInt32 *table); + UInt32 MY_FAST_CALL CrcUpdateT1_BeT8(UInt32 v, const void *data, size_t size, const UInt32 *table); +#endif + +#ifndef MY_CPU_BE + UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table); + UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table); +#endif + +typedef UInt32 (MY_FAST_CALL *CRC_FUNC)(UInt32 v, const void *data, size_t size, const UInt32 *table); + +extern +CRC_FUNC g_CrcUpdateT4; +CRC_FUNC g_CrcUpdateT4; +extern +CRC_FUNC g_CrcUpdateT8; +CRC_FUNC g_CrcUpdateT8; +extern +CRC_FUNC g_CrcUpdateT0_32; +CRC_FUNC g_CrcUpdateT0_32; +extern +CRC_FUNC g_CrcUpdateT0_64; +CRC_FUNC g_CrcUpdateT0_64; +extern +CRC_FUNC g_CrcUpdate; +CRC_FUNC g_CrcUpdate; + +UInt32 g_CrcTable[256 * CRC_NUM_TABLES]; + +UInt32 MY_FAST_CALL CrcUpdate(UInt32 v, const void *data, size_t size) +{ + return g_CrcUpdate(v, data, size, g_CrcTable); +} + +UInt32 MY_FAST_CALL CrcCalc(const void *data, size_t size) +{ + return g_CrcUpdate(CRC_INIT_VAL, data, size, g_CrcTable) ^ CRC_INIT_VAL; +} + +#define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table); +UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + const Byte *pEnd = p + size; + for (; p != pEnd; p++) + v = CRC_UPDATE_BYTE_2(v, *p); + return v; +} + + +/* ---------- hardware CRC ---------- */ + +#ifdef MY_CPU_LE + +#if defined(MY_CPU_ARM_OR_ARM64) + +// #pragma message("ARM*") + + #if defined(_MSC_VER) + #if defined(MY_CPU_ARM64) + #if (_MSC_VER >= 1910) + #define USE_ARM64_CRC + #endif + #endif + #elif (defined(__clang__) && (__clang_major__ >= 3)) \ + || (defined(__GNUC__) && (__GNUC__ > 4)) + #if !defined(__ARM_FEATURE_CRC32) + #define __ARM_FEATURE_CRC32 1 + #if (!defined(__clang__) || (__clang_major__ > 3)) // fix these numbers + #define ATTRIB_CRC __attribute__((__target__("arch=armv8-a+crc"))) + #endif + #endif + #if defined(__ARM_FEATURE_CRC32) + #define USE_ARM64_CRC + #include + #endif + #endif + +#else + +// no hardware CRC + +// #define USE_CRC_EMU + +#ifdef USE_CRC_EMU + +#pragma message("ARM64 CRC emulation") + +MY_FORCE_INLINE +UInt32 __crc32b(UInt32 v, UInt32 data) +{ + const UInt32 *table = g_CrcTable; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); + return v; +} + +MY_FORCE_INLINE +UInt32 __crc32w(UInt32 v, UInt32 data) +{ + const UInt32 *table = g_CrcTable; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + return v; +} + +MY_FORCE_INLINE +UInt32 __crc32d(UInt32 v, UInt64 data) +{ + const UInt32 *table = g_CrcTable; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + v = CRC_UPDATE_BYTE_2(v, (Byte)data); data >>= 8; + return v; +} + +#endif // USE_CRC_EMU + +#endif // defined(MY_CPU_ARM64) && defined(MY_CPU_LE) + + + +#if defined(USE_ARM64_CRC) || defined(USE_CRC_EMU) + +#define T0_32_UNROLL_BYTES (4 * 4) +#define T0_64_UNROLL_BYTES (4 * 8) + +#ifndef ATTRIB_CRC +#define ATTRIB_CRC +#endif +// #pragma message("USE ARM HW CRC") + +ATTRIB_CRC +UInt32 MY_FAST_CALL CrcUpdateT0_32(UInt32 v, const void *data, size_t size, const UInt32 *table); +ATTRIB_CRC +UInt32 MY_FAST_CALL CrcUpdateT0_32(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + UNUSED_VAR(table); + + for (; size != 0 && ((unsigned)(ptrdiff_t)p & (T0_32_UNROLL_BYTES - 1)) != 0; size--) + v = __crc32b(v, *p++); + + if (size >= T0_32_UNROLL_BYTES) + { + const Byte *lim = p + size; + size &= (T0_32_UNROLL_BYTES - 1); + lim -= size; + do + { + v = __crc32w(v, *(const UInt32 *)(const void *)(p)); + v = __crc32w(v, *(const UInt32 *)(const void *)(p + 4)); p += 2 * 4; + v = __crc32w(v, *(const UInt32 *)(const void *)(p)); + v = __crc32w(v, *(const UInt32 *)(const void *)(p + 4)); p += 2 * 4; + } + while (p != lim); + } + + for (; size != 0; size--) + v = __crc32b(v, *p++); + + return v; +} + +ATTRIB_CRC +UInt32 MY_FAST_CALL CrcUpdateT0_64(UInt32 v, const void *data, size_t size, const UInt32 *table); +ATTRIB_CRC +UInt32 MY_FAST_CALL CrcUpdateT0_64(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + UNUSED_VAR(table); + + for (; size != 0 && ((unsigned)(ptrdiff_t)p & (T0_64_UNROLL_BYTES - 1)) != 0; size--) + v = __crc32b(v, *p++); + + if (size >= T0_64_UNROLL_BYTES) + { + const Byte *lim = p + size; + size &= (T0_64_UNROLL_BYTES - 1); + lim -= size; + do + { + v = __crc32d(v, *(const UInt64 *)(const void *)(p)); + v = __crc32d(v, *(const UInt64 *)(const void *)(p + 8)); p += 2 * 8; + v = __crc32d(v, *(const UInt64 *)(const void *)(p)); + v = __crc32d(v, *(const UInt64 *)(const void *)(p + 8)); p += 2 * 8; + } + while (p != lim); + } + + for (; size != 0; size--) + v = __crc32b(v, *p++); + + return v; +} + +#endif // defined(USE_ARM64_CRC) || defined(USE_CRC_EMU) + +#endif // MY_CPU_LE + + + + +void MY_FAST_CALL CrcGenerateTable() +{ + UInt32 i; + for (i = 0; i < 256; i++) + { + UInt32 r = i; + unsigned j; + for (j = 0; j < 8; j++) + r = (r >> 1) ^ (kCrcPoly & ((UInt32)0 - (r & 1))); + g_CrcTable[i] = r; + } + for (i = 256; i < 256 * CRC_NUM_TABLES; i++) + { + UInt32 r = g_CrcTable[(size_t)i - 256]; + g_CrcTable[i] = g_CrcTable[r & 0xFF] ^ (r >> 8); + } + + #if CRC_NUM_TABLES < 4 + + g_CrcUpdate = CrcUpdateT1; + + #else + + #ifdef MY_CPU_LE + + g_CrcUpdateT4 = CrcUpdateT4; + g_CrcUpdate = CrcUpdateT4; + + #if CRC_NUM_TABLES >= 8 + g_CrcUpdateT8 = CrcUpdateT8; + + #ifdef MY_CPU_X86_OR_AMD64 + if (!CPU_Is_InOrder()) + #endif + g_CrcUpdate = CrcUpdateT8; + #endif + + #else + { + #ifndef MY_CPU_BE + UInt32 k = 0x01020304; + const Byte *p = (const Byte *)&k; + if (p[0] == 4 && p[1] == 3) + { + g_CrcUpdateT4 = CrcUpdateT4; + g_CrcUpdate = CrcUpdateT4; + #if CRC_NUM_TABLES >= 8 + g_CrcUpdateT8 = CrcUpdateT8; + g_CrcUpdate = CrcUpdateT8; + #endif + } + else if (p[0] != 1 || p[1] != 2) + g_CrcUpdate = CrcUpdateT1; + else + #endif + { + for (i = 256 * CRC_NUM_TABLES - 1; i >= 256; i--) + { + UInt32 x = g_CrcTable[(size_t)i - 256]; + g_CrcTable[i] = CRC_UINT32_SWAP(x); + } + g_CrcUpdateT4 = CrcUpdateT1_BeT4; + g_CrcUpdate = CrcUpdateT1_BeT4; + #if CRC_NUM_TABLES >= 8 + g_CrcUpdateT8 = CrcUpdateT1_BeT8; + g_CrcUpdate = CrcUpdateT1_BeT8; + #endif + } + } + #endif + #endif + + #ifdef MY_CPU_LE + #ifdef USE_ARM64_CRC + if (CPU_IsSupported_CRC32()) + { + g_CrcUpdateT0_32 = CrcUpdateT0_32; + g_CrcUpdateT0_64 = CrcUpdateT0_64; + g_CrcUpdate = + #if defined(MY_CPU_ARM) + CrcUpdateT0_32; + #else + CrcUpdateT0_64; + #endif + } + #endif + + #ifdef USE_CRC_EMU + g_CrcUpdateT0_32 = CrcUpdateT0_32; + g_CrcUpdateT0_64 = CrcUpdateT0_64; + g_CrcUpdate = CrcUpdateT0_64; + #endif + #endif +} diff --git a/3rdparty/lzma/src/7zCrcOpt.c b/3rdparty/lzma/src/7zCrcOpt.c new file mode 100644 index 0000000000..69fad9ca2b --- /dev/null +++ b/3rdparty/lzma/src/7zCrcOpt.c @@ -0,0 +1,117 @@ +/* 7zCrcOpt.c -- CRC32 calculation +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" + +#ifndef MY_CPU_BE + +#define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table); +UInt32 MY_FAST_CALL CrcUpdateT4(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) + v = CRC_UPDATE_BYTE_2(v, *p); + for (; size >= 4; size -= 4, p += 4) + { + v ^= *(const UInt32 *)(const void *)p; + v = + (table + 0x300)[((v ) & 0xFF)] + ^ (table + 0x200)[((v >> 8) & 0xFF)] + ^ (table + 0x100)[((v >> 16) & 0xFF)] + ^ (table + 0x000)[((v >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC_UPDATE_BYTE_2(v, *p); + return v; +} + +UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table); +UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 7) != 0; size--, p++) + v = CRC_UPDATE_BYTE_2(v, *p); + for (; size >= 8; size -= 8, p += 8) + { + UInt32 d; + v ^= *(const UInt32 *)(const void *)p; + v = + (table + 0x700)[((v ) & 0xFF)] + ^ (table + 0x600)[((v >> 8) & 0xFF)] + ^ (table + 0x500)[((v >> 16) & 0xFF)] + ^ (table + 0x400)[((v >> 24))]; + d = *((const UInt32 *)(const void *)p + 1); + v ^= + (table + 0x300)[((d ) & 0xFF)] + ^ (table + 0x200)[((d >> 8) & 0xFF)] + ^ (table + 0x100)[((d >> 16) & 0xFF)] + ^ (table + 0x000)[((d >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC_UPDATE_BYTE_2(v, *p); + return v; +} + +#endif + + +#ifndef MY_CPU_LE + +#define CRC_UINT32_SWAP(v) ((v >> 24) | ((v >> 8) & 0xFF00) | ((v << 8) & 0xFF0000) | (v << 24)) + +#define CRC_UPDATE_BYTE_2_BE(crc, b) (table[(((crc) >> 24) ^ (b))] ^ ((crc) << 8)) + +UInt32 MY_FAST_CALL CrcUpdateT1_BeT4(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + table += 0x100; + v = CRC_UINT32_SWAP(v); + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) + v = CRC_UPDATE_BYTE_2_BE(v, *p); + for (; size >= 4; size -= 4, p += 4) + { + v ^= *(const UInt32 *)(const void *)p; + v = + (table + 0x000)[((v ) & 0xFF)] + ^ (table + 0x100)[((v >> 8) & 0xFF)] + ^ (table + 0x200)[((v >> 16) & 0xFF)] + ^ (table + 0x300)[((v >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC_UPDATE_BYTE_2_BE(v, *p); + return CRC_UINT32_SWAP(v); +} + +UInt32 MY_FAST_CALL CrcUpdateT1_BeT8(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + table += 0x100; + v = CRC_UINT32_SWAP(v); + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 7) != 0; size--, p++) + v = CRC_UPDATE_BYTE_2_BE(v, *p); + for (; size >= 8; size -= 8, p += 8) + { + UInt32 d; + v ^= *(const UInt32 *)(const void *)p; + v = + (table + 0x400)[((v ) & 0xFF)] + ^ (table + 0x500)[((v >> 8) & 0xFF)] + ^ (table + 0x600)[((v >> 16) & 0xFF)] + ^ (table + 0x700)[((v >> 24))]; + d = *((const UInt32 *)(const void *)p + 1); + v ^= + (table + 0x000)[((d ) & 0xFF)] + ^ (table + 0x100)[((d >> 8) & 0xFF)] + ^ (table + 0x200)[((d >> 16) & 0xFF)] + ^ (table + 0x300)[((d >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC_UPDATE_BYTE_2_BE(v, *p); + return CRC_UINT32_SWAP(v); +} + +#endif diff --git a/3rdparty/lzma/src/7zDec.c b/3rdparty/lzma/src/7zDec.c new file mode 100644 index 0000000000..fbfd016e1e --- /dev/null +++ b/3rdparty/lzma/src/7zDec.c @@ -0,0 +1,600 @@ +/* 7zDec.c -- Decoding from 7z folder +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #define _7ZIP_PPMD_SUPPPORT */ + +#include "7z.h" +#include "7zCrc.h" + +#include "Bcj2.h" +#include "Bra.h" +#include "CpuArch.h" +#include "Delta.h" +#include "LzmaDec.h" +#include "Lzma2Dec.h" +#ifdef _7ZIP_PPMD_SUPPPORT +#include "Ppmd7.h" +#endif + +#define k_Copy 0 +#ifndef _7Z_NO_METHOD_LZMA2 +#define k_LZMA2 0x21 +#endif +#define k_LZMA 0x30101 +#define k_BCJ2 0x303011B +#ifndef _7Z_NO_METHODS_FILTERS +#define k_Delta 3 +#define k_BCJ 0x3030103 +#define k_PPC 0x3030205 +#define k_IA64 0x3030401 +#define k_ARM 0x3030501 +#define k_ARMT 0x3030701 +#define k_SPARC 0x3030805 +#endif + +#ifdef _7ZIP_PPMD_SUPPPORT + +#define k_PPMD 0x30401 + +typedef struct +{ + IByteIn vt; + const Byte *cur; + const Byte *end; + const Byte *begin; + UInt64 processed; + BoolInt extra; + SRes res; + const ILookInStream *inStream; +} CByteInToLook; + +static Byte ReadByte(const IByteIn *pp) +{ + CByteInToLook *p = CONTAINER_FROM_VTBL(pp, CByteInToLook, vt); + if (p->cur != p->end) + return *p->cur++; + if (p->res == SZ_OK) + { + size_t size = (size_t)(p->cur - p->begin); + p->processed += size; + p->res = ILookInStream_Skip(p->inStream, size); + size = (1 << 25); + p->res = ILookInStream_Look(p->inStream, (const void **)&p->begin, &size); + p->cur = p->begin; + p->end = p->begin + size; + if (size != 0) + return *p->cur++;; + } + p->extra = True; + return 0; +} + +static SRes SzDecodePpmd(const Byte *props, unsigned propsSize, UInt64 inSize, const ILookInStream *inStream, + Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain) +{ + CPpmd7 ppmd; + CByteInToLook s; + SRes res = SZ_OK; + + s.vt.Read = ReadByte; + s.inStream = inStream; + s.begin = s.end = s.cur = NULL; + s.extra = False; + s.res = SZ_OK; + s.processed = 0; + + if (propsSize != 5) + return SZ_ERROR_UNSUPPORTED; + + { + unsigned order = props[0]; + UInt32 memSize = GetUi32(props + 1); + if (order < PPMD7_MIN_ORDER || + order > PPMD7_MAX_ORDER || + memSize < PPMD7_MIN_MEM_SIZE || + memSize > PPMD7_MAX_MEM_SIZE) + return SZ_ERROR_UNSUPPORTED; + Ppmd7_Construct(&ppmd); + if (!Ppmd7_Alloc(&ppmd, memSize, allocMain)) + return SZ_ERROR_MEM; + Ppmd7_Init(&ppmd, order); + } + { + ppmd.rc.dec.Stream = &s.vt; + if (!Ppmd7z_RangeDec_Init(&ppmd.rc.dec)) + res = SZ_ERROR_DATA; + else if (!s.extra) + { + Byte *buf = outBuffer; + const Byte *lim = buf + outSize; + for (; buf != lim; buf++) + { + int sym = Ppmd7z_DecodeSymbol(&ppmd); + if (s.extra || sym < 0) + break; + *buf = (Byte)sym; + } + if (buf != lim) + res = SZ_ERROR_DATA; + else if (!Ppmd7z_RangeDec_IsFinishedOK(&ppmd.rc.dec)) + { + /* if (Ppmd7z_DecodeSymbol(&ppmd) != PPMD7_SYM_END || !Ppmd7z_RangeDec_IsFinishedOK(&ppmd.rc.dec)) */ + res = SZ_ERROR_DATA; + } + } + if (s.extra) + res = (s.res != SZ_OK ? s.res : SZ_ERROR_DATA); + else if (s.processed + (size_t)(s.cur - s.begin) != inSize) + res = SZ_ERROR_DATA; + } + Ppmd7_Free(&ppmd, allocMain); + return res; +} + +#endif + + +static SRes SzDecodeLzma(const Byte *props, unsigned propsSize, UInt64 inSize, ILookInStream *inStream, + Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain) +{ + CLzmaDec state; + SRes res = SZ_OK; + + LzmaDec_Construct(&state); + RINOK(LzmaDec_AllocateProbs(&state, props, propsSize, allocMain)); + state.dic = outBuffer; + state.dicBufSize = outSize; + LzmaDec_Init(&state); + + for (;;) + { + const void *inBuf = NULL; + size_t lookahead = (1 << 18); + if (lookahead > inSize) + lookahead = (size_t)inSize; + res = ILookInStream_Look(inStream, &inBuf, &lookahead); + if (res != SZ_OK) + break; + + { + SizeT inProcessed = (SizeT)lookahead, dicPos = state.dicPos; + ELzmaStatus status; + res = LzmaDec_DecodeToDic(&state, outSize, (const Byte *)inBuf, &inProcessed, LZMA_FINISH_END, &status); + lookahead -= inProcessed; + inSize -= inProcessed; + if (res != SZ_OK) + break; + + if (status == LZMA_STATUS_FINISHED_WITH_MARK) + { + if (outSize != state.dicPos || inSize != 0) + res = SZ_ERROR_DATA; + break; + } + + if (outSize == state.dicPos && inSize == 0 && status == LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK) + break; + + if (inProcessed == 0 && dicPos == state.dicPos) + { + res = SZ_ERROR_DATA; + break; + } + + res = ILookInStream_Skip(inStream, inProcessed); + if (res != SZ_OK) + break; + } + } + + LzmaDec_FreeProbs(&state, allocMain); + return res; +} + + +#ifndef _7Z_NO_METHOD_LZMA2 + +static SRes SzDecodeLzma2(const Byte *props, unsigned propsSize, UInt64 inSize, ILookInStream *inStream, + Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain) +{ + CLzma2Dec state; + SRes res = SZ_OK; + + Lzma2Dec_Construct(&state); + if (propsSize != 1) + return SZ_ERROR_DATA; + RINOK(Lzma2Dec_AllocateProbs(&state, props[0], allocMain)); + state.decoder.dic = outBuffer; + state.decoder.dicBufSize = outSize; + Lzma2Dec_Init(&state); + + for (;;) + { + const void *inBuf = NULL; + size_t lookahead = (1 << 18); + if (lookahead > inSize) + lookahead = (size_t)inSize; + res = ILookInStream_Look(inStream, &inBuf, &lookahead); + if (res != SZ_OK) + break; + + { + SizeT inProcessed = (SizeT)lookahead, dicPos = state.decoder.dicPos; + ELzmaStatus status; + res = Lzma2Dec_DecodeToDic(&state, outSize, (const Byte *)inBuf, &inProcessed, LZMA_FINISH_END, &status); + lookahead -= inProcessed; + inSize -= inProcessed; + if (res != SZ_OK) + break; + + if (status == LZMA_STATUS_FINISHED_WITH_MARK) + { + if (outSize != state.decoder.dicPos || inSize != 0) + res = SZ_ERROR_DATA; + break; + } + + if (inProcessed == 0 && dicPos == state.decoder.dicPos) + { + res = SZ_ERROR_DATA; + break; + } + + res = ILookInStream_Skip(inStream, inProcessed); + if (res != SZ_OK) + break; + } + } + + Lzma2Dec_FreeProbs(&state, allocMain); + return res; +} + +#endif + + +static SRes SzDecodeCopy(UInt64 inSize, ILookInStream *inStream, Byte *outBuffer) +{ + while (inSize > 0) + { + const void *inBuf; + size_t curSize = (1 << 18); + if (curSize > inSize) + curSize = (size_t)inSize; + RINOK(ILookInStream_Look(inStream, &inBuf, &curSize)); + if (curSize == 0) + return SZ_ERROR_INPUT_EOF; + memcpy(outBuffer, inBuf, curSize); + outBuffer += curSize; + inSize -= curSize; + RINOK(ILookInStream_Skip(inStream, curSize)); + } + return SZ_OK; +} + +static BoolInt IS_MAIN_METHOD(UInt32 m) +{ + switch (m) + { + case k_Copy: + case k_LZMA: + #ifndef _7Z_NO_METHOD_LZMA2 + case k_LZMA2: + #endif + #ifdef _7ZIP_PPMD_SUPPPORT + case k_PPMD: + #endif + return True; + } + return False; +} + +static BoolInt IS_SUPPORTED_CODER(const CSzCoderInfo *c) +{ + return + c->NumStreams == 1 + /* && c->MethodID <= (UInt32)0xFFFFFFFF */ + && IS_MAIN_METHOD((UInt32)c->MethodID); +} + +#define IS_BCJ2(c) ((c)->MethodID == k_BCJ2 && (c)->NumStreams == 4) + +static SRes CheckSupportedFolder(const CSzFolder *f) +{ + if (f->NumCoders < 1 || f->NumCoders > 4) + return SZ_ERROR_UNSUPPORTED; + if (!IS_SUPPORTED_CODER(&f->Coders[0])) + return SZ_ERROR_UNSUPPORTED; + if (f->NumCoders == 1) + { + if (f->NumPackStreams != 1 || f->PackStreams[0] != 0 || f->NumBonds != 0) + return SZ_ERROR_UNSUPPORTED; + return SZ_OK; + } + + + #ifndef _7Z_NO_METHODS_FILTERS + + if (f->NumCoders == 2) + { + const CSzCoderInfo *c = &f->Coders[1]; + if ( + /* c->MethodID > (UInt32)0xFFFFFFFF || */ + c->NumStreams != 1 + || f->NumPackStreams != 1 + || f->PackStreams[0] != 0 + || f->NumBonds != 1 + || f->Bonds[0].InIndex != 1 + || f->Bonds[0].OutIndex != 0) + return SZ_ERROR_UNSUPPORTED; + switch ((UInt32)c->MethodID) + { + case k_Delta: + case k_BCJ: + case k_PPC: + case k_IA64: + case k_SPARC: + case k_ARM: + case k_ARMT: + break; + default: + return SZ_ERROR_UNSUPPORTED; + } + return SZ_OK; + } + + #endif + + + if (f->NumCoders == 4) + { + if (!IS_SUPPORTED_CODER(&f->Coders[1]) + || !IS_SUPPORTED_CODER(&f->Coders[2]) + || !IS_BCJ2(&f->Coders[3])) + return SZ_ERROR_UNSUPPORTED; + if (f->NumPackStreams != 4 + || f->PackStreams[0] != 2 + || f->PackStreams[1] != 6 + || f->PackStreams[2] != 1 + || f->PackStreams[3] != 0 + || f->NumBonds != 3 + || f->Bonds[0].InIndex != 5 || f->Bonds[0].OutIndex != 0 + || f->Bonds[1].InIndex != 4 || f->Bonds[1].OutIndex != 1 + || f->Bonds[2].InIndex != 3 || f->Bonds[2].OutIndex != 2) + return SZ_ERROR_UNSUPPORTED; + return SZ_OK; + } + + return SZ_ERROR_UNSUPPORTED; +} + +#ifndef _7Z_NO_METHODS_FILTERS +#define CASE_BRA_CONV(isa) case k_ ## isa: isa ## _Convert(outBuffer, outSize, 0, 0); break; +#endif + +static SRes SzFolder_Decode2(const CSzFolder *folder, + const Byte *propsData, + const UInt64 *unpackSizes, + const UInt64 *packPositions, + ILookInStream *inStream, UInt64 startPos, + Byte *outBuffer, SizeT outSize, ISzAllocPtr allocMain, + Byte *tempBuf[]) +{ + UInt32 ci; + SizeT tempSizes[3] = { 0, 0, 0}; + SizeT tempSize3 = 0; + Byte *tempBuf3 = 0; + + RINOK(CheckSupportedFolder(folder)); + + for (ci = 0; ci < folder->NumCoders; ci++) + { + const CSzCoderInfo *coder = &folder->Coders[ci]; + + if (IS_MAIN_METHOD((UInt32)coder->MethodID)) + { + UInt32 si = 0; + UInt64 offset; + UInt64 inSize; + Byte *outBufCur = outBuffer; + SizeT outSizeCur = outSize; + if (folder->NumCoders == 4) + { + UInt32 indices[] = { 3, 2, 0 }; + UInt64 unpackSize = unpackSizes[ci]; + si = indices[ci]; + if (ci < 2) + { + Byte *temp; + outSizeCur = (SizeT)unpackSize; + if (outSizeCur != unpackSize) + return SZ_ERROR_MEM; + temp = (Byte *)ISzAlloc_Alloc(allocMain, outSizeCur); + if (!temp && outSizeCur != 0) + return SZ_ERROR_MEM; + outBufCur = tempBuf[1 - ci] = temp; + tempSizes[1 - ci] = outSizeCur; + } + else if (ci == 2) + { + if (unpackSize > outSize) /* check it */ + return SZ_ERROR_PARAM; + tempBuf3 = outBufCur = outBuffer + (outSize - (size_t)unpackSize); + tempSize3 = outSizeCur = (SizeT)unpackSize; + } + else + return SZ_ERROR_UNSUPPORTED; + } + offset = packPositions[si]; + inSize = packPositions[(size_t)si + 1] - offset; + RINOK(LookInStream_SeekTo(inStream, startPos + offset)); + + if (coder->MethodID == k_Copy) + { + if (inSize != outSizeCur) /* check it */ + return SZ_ERROR_DATA; + RINOK(SzDecodeCopy(inSize, inStream, outBufCur)); + } + else if (coder->MethodID == k_LZMA) + { + RINOK(SzDecodeLzma(propsData + coder->PropsOffset, coder->PropsSize, inSize, inStream, outBufCur, outSizeCur, allocMain)); + } + #ifndef _7Z_NO_METHOD_LZMA2 + else if (coder->MethodID == k_LZMA2) + { + RINOK(SzDecodeLzma2(propsData + coder->PropsOffset, coder->PropsSize, inSize, inStream, outBufCur, outSizeCur, allocMain)); + } + #endif + #ifdef _7ZIP_PPMD_SUPPPORT + else if (coder->MethodID == k_PPMD) + { + RINOK(SzDecodePpmd(propsData + coder->PropsOffset, coder->PropsSize, inSize, inStream, outBufCur, outSizeCur, allocMain)); + } + #endif + else + return SZ_ERROR_UNSUPPORTED; + } + else if (coder->MethodID == k_BCJ2) + { + UInt64 offset = packPositions[1]; + UInt64 s3Size = packPositions[2] - offset; + + if (ci != 3) + return SZ_ERROR_UNSUPPORTED; + + tempSizes[2] = (SizeT)s3Size; + if (tempSizes[2] != s3Size) + return SZ_ERROR_MEM; + tempBuf[2] = (Byte *)ISzAlloc_Alloc(allocMain, tempSizes[2]); + if (!tempBuf[2] && tempSizes[2] != 0) + return SZ_ERROR_MEM; + + RINOK(LookInStream_SeekTo(inStream, startPos + offset)); + RINOK(SzDecodeCopy(s3Size, inStream, tempBuf[2])); + + if ((tempSizes[0] & 3) != 0 || + (tempSizes[1] & 3) != 0 || + tempSize3 + tempSizes[0] + tempSizes[1] != outSize) + return SZ_ERROR_DATA; + + { + CBcj2Dec p; + + p.bufs[0] = tempBuf3; p.lims[0] = tempBuf3 + tempSize3; + p.bufs[1] = tempBuf[0]; p.lims[1] = tempBuf[0] + tempSizes[0]; + p.bufs[2] = tempBuf[1]; p.lims[2] = tempBuf[1] + tempSizes[1]; + p.bufs[3] = tempBuf[2]; p.lims[3] = tempBuf[2] + tempSizes[2]; + + p.dest = outBuffer; + p.destLim = outBuffer + outSize; + + Bcj2Dec_Init(&p); + RINOK(Bcj2Dec_Decode(&p)); + + { + unsigned i; + for (i = 0; i < 4; i++) + if (p.bufs[i] != p.lims[i]) + return SZ_ERROR_DATA; + + if (!Bcj2Dec_IsFinished(&p)) + return SZ_ERROR_DATA; + + if (p.dest != p.destLim + || p.state != BCJ2_STREAM_MAIN) + return SZ_ERROR_DATA; + } + } + } + #ifndef _7Z_NO_METHODS_FILTERS + else if (ci == 1) + { + if (coder->MethodID == k_Delta) + { + if (coder->PropsSize != 1) + return SZ_ERROR_UNSUPPORTED; + { + Byte state[DELTA_STATE_SIZE]; + Delta_Init(state); + Delta_Decode(state, (unsigned)(propsData[coder->PropsOffset]) + 1, outBuffer, outSize); + } + } + else + { + if (coder->PropsSize != 0) + return SZ_ERROR_UNSUPPORTED; + switch (coder->MethodID) + { + case k_BCJ: + { + UInt32 state; + x86_Convert_Init(state); + x86_Convert(outBuffer, outSize, 0, &state, 0); + break; + } + CASE_BRA_CONV(PPC) + CASE_BRA_CONV(IA64) + CASE_BRA_CONV(SPARC) + CASE_BRA_CONV(ARM) + CASE_BRA_CONV(ARMT) + default: + return SZ_ERROR_UNSUPPORTED; + } + } + } + #endif + else + return SZ_ERROR_UNSUPPORTED; + } + + return SZ_OK; +} + + +SRes SzAr_DecodeFolder(const CSzAr *p, UInt32 folderIndex, + ILookInStream *inStream, UInt64 startPos, + Byte *outBuffer, size_t outSize, + ISzAllocPtr allocMain) +{ + SRes res; + CSzFolder folder; + CSzData sd; + + const Byte *data = p->CodersData + p->FoCodersOffsets[folderIndex]; + sd.Data = data; + sd.Size = p->FoCodersOffsets[(size_t)folderIndex + 1] - p->FoCodersOffsets[folderIndex]; + + res = SzGetNextFolderItem(&folder, &sd); + + if (res != SZ_OK) + return res; + + if (sd.Size != 0 + || folder.UnpackStream != p->FoToMainUnpackSizeIndex[folderIndex] + || outSize != SzAr_GetFolderUnpackSize(p, folderIndex)) + return SZ_ERROR_FAIL; + { + unsigned i; + Byte *tempBuf[3] = { 0, 0, 0}; + + res = SzFolder_Decode2(&folder, data, + &p->CoderUnpackSizes[p->FoToCoderUnpackSizes[folderIndex]], + p->PackPositions + p->FoStartPackStreamIndex[folderIndex], + inStream, startPos, + outBuffer, (SizeT)outSize, allocMain, tempBuf); + + for (i = 0; i < 3; i++) + ISzAlloc_Free(allocMain, tempBuf[i]); + + if (res == SZ_OK) + if (SzBitWithVals_Check(&p->FolderCRCs, folderIndex)) + if (CrcCalc(outBuffer, outSize) != p->FolderCRCs.Vals[folderIndex]) + res = SZ_ERROR_CRC; + + return res; + } +} diff --git a/3rdparty/lzma/src/7zFile.c b/3rdparty/lzma/src/7zFile.c new file mode 100644 index 0000000000..13d2efa47f --- /dev/null +++ b/3rdparty/lzma/src/7zFile.c @@ -0,0 +1,442 @@ +/* 7zFile.c -- File IO +2021-04-29 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "7zFile.h" + +#ifndef USE_WINDOWS_FILE + + #include + + #ifndef USE_FOPEN + #include + #include + #ifdef _WIN32 + #include + typedef int ssize_t; + typedef int off_t; + #else + #include + #endif + #endif + +#else + +/* + ReadFile and WriteFile functions in Windows have BUG: + If you Read or Write 64MB or more (probably min_failure_size = 64MB - 32KB + 1) + from/to Network file, it returns ERROR_NO_SYSTEM_RESOURCES + (Insufficient system resources exist to complete the requested service). + Probably in some version of Windows there are problems with other sizes: + for 32 MB (maybe also for 16 MB). + And message can be "Network connection was lost" +*/ + +#endif + +#define kChunkSizeMax (1 << 22) + +void File_Construct(CSzFile *p) +{ + #ifdef USE_WINDOWS_FILE + p->handle = INVALID_HANDLE_VALUE; + #elif defined(USE_FOPEN) + p->file = NULL; + #else + p->fd = -1; + #endif +} + +#if !defined(UNDER_CE) || !defined(USE_WINDOWS_FILE) + +static WRes File_Open(CSzFile *p, const char *name, int writeMode) +{ + #ifdef USE_WINDOWS_FILE + + p->handle = CreateFileA(name, + writeMode ? GENERIC_WRITE : GENERIC_READ, + FILE_SHARE_READ, NULL, + writeMode ? CREATE_ALWAYS : OPEN_EXISTING, + FILE_ATTRIBUTE_NORMAL, NULL); + return (p->handle != INVALID_HANDLE_VALUE) ? 0 : GetLastError(); + + #elif defined(USE_FOPEN) + + p->file = fopen(name, writeMode ? "wb+" : "rb"); + return (p->file != 0) ? 0 : + #ifdef UNDER_CE + 2; /* ENOENT */ + #else + errno; + #endif + + #else + + int flags = (writeMode ? (O_CREAT | O_EXCL | O_WRONLY) : O_RDONLY); + #ifdef O_BINARY + flags |= O_BINARY; + #endif + p->fd = open(name, flags, 0666); + return (p->fd != -1) ? 0 : errno; + + #endif +} + +WRes InFile_Open(CSzFile *p, const char *name) { return File_Open(p, name, 0); } + +WRes OutFile_Open(CSzFile *p, const char *name) +{ + #if defined(USE_WINDOWS_FILE) || defined(USE_FOPEN) + return File_Open(p, name, 1); + #else + p->fd = creat(name, 0666); + return (p->fd != -1) ? 0 : errno; + #endif +} + +#endif + + +#ifdef USE_WINDOWS_FILE +static WRes File_OpenW(CSzFile *p, const WCHAR *name, int writeMode) +{ + p->handle = CreateFileW(name, + writeMode ? GENERIC_WRITE : GENERIC_READ, + FILE_SHARE_READ, NULL, + writeMode ? CREATE_ALWAYS : OPEN_EXISTING, + FILE_ATTRIBUTE_NORMAL, NULL); + return (p->handle != INVALID_HANDLE_VALUE) ? 0 : GetLastError(); +} +WRes InFile_OpenW(CSzFile *p, const WCHAR *name) { return File_OpenW(p, name, 0); } +WRes OutFile_OpenW(CSzFile *p, const WCHAR *name) { return File_OpenW(p, name, 1); } +#endif + +WRes File_Close(CSzFile *p) +{ + #ifdef USE_WINDOWS_FILE + + if (p->handle != INVALID_HANDLE_VALUE) + { + if (!CloseHandle(p->handle)) + return GetLastError(); + p->handle = INVALID_HANDLE_VALUE; + } + + #elif defined(USE_FOPEN) + + if (p->file != NULL) + { + int res = fclose(p->file); + if (res != 0) + { + if (res == EOF) + return errno; + return res; + } + p->file = NULL; + } + + #else + + if (p->fd != -1) + { + if (close(p->fd) != 0) + return errno; + p->fd = -1; + } + + #endif + + return 0; +} + + +WRes File_Read(CSzFile *p, void *data, size_t *size) +{ + size_t originalSize = *size; + *size = 0; + if (originalSize == 0) + return 0; + + #ifdef USE_WINDOWS_FILE + + do + { + const DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize; + DWORD processed = 0; + const BOOL res = ReadFile(p->handle, data, curSize, &processed, NULL); + data = (void *)((Byte *)data + processed); + originalSize -= processed; + *size += processed; + if (!res) + return GetLastError(); + // debug : we can break here for partial reading mode + if (processed == 0) + break; + } + while (originalSize > 0); + + #elif defined(USE_FOPEN) + + do + { + const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; + const size_t processed = fread(data, 1, curSize, p->file); + data = (void *)((Byte *)data + (size_t)processed); + originalSize -= processed; + *size += processed; + if (processed != curSize) + return ferror(p->file); + // debug : we can break here for partial reading mode + if (processed == 0) + break; + } + while (originalSize > 0); + + #else + + do + { + const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; + const ssize_t processed = read(p->fd, data, curSize); + if (processed == -1) + return errno; + if (processed == 0) + break; + data = (void *)((Byte *)data + (size_t)processed); + originalSize -= (size_t)processed; + *size += (size_t)processed; + // debug : we can break here for partial reading mode + // break; + } + while (originalSize > 0); + + #endif + + return 0; +} + + +WRes File_Write(CSzFile *p, const void *data, size_t *size) +{ + size_t originalSize = *size; + *size = 0; + if (originalSize == 0) + return 0; + + #ifdef USE_WINDOWS_FILE + + do + { + const DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize; + DWORD processed = 0; + const BOOL res = WriteFile(p->handle, data, curSize, &processed, NULL); + data = (const void *)((const Byte *)data + processed); + originalSize -= processed; + *size += processed; + if (!res) + return GetLastError(); + if (processed == 0) + break; + } + while (originalSize > 0); + + #elif defined(USE_FOPEN) + + do + { + const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; + const size_t processed = fwrite(data, 1, curSize, p->file); + data = (void *)((Byte *)data + (size_t)processed); + originalSize -= processed; + *size += processed; + if (processed != curSize) + return ferror(p->file); + if (processed == 0) + break; + } + while (originalSize > 0); + + #else + + do + { + const size_t curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : originalSize; + const ssize_t processed = write(p->fd, data, curSize); + if (processed == -1) + return errno; + if (processed == 0) + break; + data = (void *)((Byte *)data + (size_t)processed); + originalSize -= (size_t)processed; + *size += (size_t)processed; + } + while (originalSize > 0); + + #endif + + return 0; +} + + +WRes File_Seek(CSzFile *p, Int64 *pos, ESzSeek origin) +{ + #ifdef USE_WINDOWS_FILE + + DWORD moveMethod; + UInt32 low = (UInt32)*pos; + LONG high = (LONG)((UInt64)*pos >> 16 >> 16); /* for case when UInt64 is 32-bit only */ + switch (origin) + { + case SZ_SEEK_SET: moveMethod = FILE_BEGIN; break; + case SZ_SEEK_CUR: moveMethod = FILE_CURRENT; break; + case SZ_SEEK_END: moveMethod = FILE_END; break; + default: return ERROR_INVALID_PARAMETER; + } + low = SetFilePointer(p->handle, (LONG)low, &high, moveMethod); + if (low == (UInt32)0xFFFFFFFF) + { + WRes res = GetLastError(); + if (res != NO_ERROR) + return res; + } + *pos = ((Int64)high << 32) | low; + return 0; + + #else + + int moveMethod; // = origin; + + switch (origin) + { + case SZ_SEEK_SET: moveMethod = SEEK_SET; break; + case SZ_SEEK_CUR: moveMethod = SEEK_CUR; break; + case SZ_SEEK_END: moveMethod = SEEK_END; break; + default: return EINVAL; + } + + #if defined(USE_FOPEN) + { + int res = fseek(p->file, (long)*pos, moveMethod); + if (res == -1) + return errno; + *pos = ftell(p->file); + if (*pos == -1) + return errno; + return 0; + } + #else + { + off_t res = lseek(p->fd, (off_t)*pos, moveMethod); + if (res == -1) + return errno; + *pos = res; + return 0; + } + + #endif // USE_FOPEN + #endif // USE_WINDOWS_FILE +} + + +WRes File_GetLength(CSzFile *p, UInt64 *length) +{ + #ifdef USE_WINDOWS_FILE + + DWORD sizeHigh; + DWORD sizeLow = GetFileSize(p->handle, &sizeHigh); + if (sizeLow == 0xFFFFFFFF) + { + DWORD res = GetLastError(); + if (res != NO_ERROR) + return res; + } + *length = (((UInt64)sizeHigh) << 32) + sizeLow; + return 0; + + #elif defined(USE_FOPEN) + + long pos = ftell(p->file); + int res = fseek(p->file, 0, SEEK_END); + *length = ftell(p->file); + fseek(p->file, pos, SEEK_SET); + return res; + + #else + + off_t pos; + *length = 0; + pos = lseek(p->fd, 0, SEEK_CUR); + if (pos != -1) + { + const off_t len2 = lseek(p->fd, 0, SEEK_END); + const off_t res2 = lseek(p->fd, pos, SEEK_SET); + if (len2 != -1) + { + *length = (UInt64)len2; + if (res2 != -1) + return 0; + } + } + return errno; + + #endif +} + + +/* ---------- FileSeqInStream ---------- */ + +static SRes FileSeqInStream_Read(const ISeqInStream *pp, void *buf, size_t *size) +{ + CFileSeqInStream *p = CONTAINER_FROM_VTBL(pp, CFileSeqInStream, vt); + WRes wres = File_Read(&p->file, buf, size); + p->wres = wres; + return (wres == 0) ? SZ_OK : SZ_ERROR_READ; +} + +void FileSeqInStream_CreateVTable(CFileSeqInStream *p) +{ + p->vt.Read = FileSeqInStream_Read; +} + + +/* ---------- FileInStream ---------- */ + +static SRes FileInStream_Read(const ISeekInStream *pp, void *buf, size_t *size) +{ + CFileInStream *p = CONTAINER_FROM_VTBL(pp, CFileInStream, vt); + WRes wres = File_Read(&p->file, buf, size); + p->wres = wres; + return (wres == 0) ? SZ_OK : SZ_ERROR_READ; +} + +static SRes FileInStream_Seek(const ISeekInStream *pp, Int64 *pos, ESzSeek origin) +{ + CFileInStream *p = CONTAINER_FROM_VTBL(pp, CFileInStream, vt); + WRes wres = File_Seek(&p->file, pos, origin); + p->wres = wres; + return (wres == 0) ? SZ_OK : SZ_ERROR_READ; +} + +void FileInStream_CreateVTable(CFileInStream *p) +{ + p->vt.Read = FileInStream_Read; + p->vt.Seek = FileInStream_Seek; +} + + +/* ---------- FileOutStream ---------- */ + +static size_t FileOutStream_Write(const ISeqOutStream *pp, const void *data, size_t size) +{ + CFileOutStream *p = CONTAINER_FROM_VTBL(pp, CFileOutStream, vt); + WRes wres = File_Write(&p->file, data, &size); + p->wres = wres; + return size; +} + +void FileOutStream_CreateVTable(CFileOutStream *p) +{ + p->vt.Write = FileOutStream_Write; +} diff --git a/3rdparty/lzma/src/7zStream.c b/3rdparty/lzma/src/7zStream.c new file mode 100644 index 0000000000..28a14604fb --- /dev/null +++ b/3rdparty/lzma/src/7zStream.c @@ -0,0 +1,176 @@ +/* 7zStream.c -- 7z Stream functions +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7zTypes.h" + +SRes SeqInStream_Read2(const ISeqInStream *stream, void *buf, size_t size, SRes errorType) +{ + while (size != 0) + { + size_t processed = size; + RINOK(ISeqInStream_Read(stream, buf, &processed)); + if (processed == 0) + return errorType; + buf = (void *)((Byte *)buf + processed); + size -= processed; + } + return SZ_OK; +} + +SRes SeqInStream_Read(const ISeqInStream *stream, void *buf, size_t size) +{ + return SeqInStream_Read2(stream, buf, size, SZ_ERROR_INPUT_EOF); +} + +SRes SeqInStream_ReadByte(const ISeqInStream *stream, Byte *buf) +{ + size_t processed = 1; + RINOK(ISeqInStream_Read(stream, buf, &processed)); + return (processed == 1) ? SZ_OK : SZ_ERROR_INPUT_EOF; +} + + + +SRes LookInStream_SeekTo(const ILookInStream *stream, UInt64 offset) +{ + Int64 t = (Int64)offset; + return ILookInStream_Seek(stream, &t, SZ_SEEK_SET); +} + +SRes LookInStream_LookRead(const ILookInStream *stream, void *buf, size_t *size) +{ + const void *lookBuf; + if (*size == 0) + return SZ_OK; + RINOK(ILookInStream_Look(stream, &lookBuf, size)); + memcpy(buf, lookBuf, *size); + return ILookInStream_Skip(stream, *size); +} + +SRes LookInStream_Read2(const ILookInStream *stream, void *buf, size_t size, SRes errorType) +{ + while (size != 0) + { + size_t processed = size; + RINOK(ILookInStream_Read(stream, buf, &processed)); + if (processed == 0) + return errorType; + buf = (void *)((Byte *)buf + processed); + size -= processed; + } + return SZ_OK; +} + +SRes LookInStream_Read(const ILookInStream *stream, void *buf, size_t size) +{ + return LookInStream_Read2(stream, buf, size, SZ_ERROR_INPUT_EOF); +} + + + +#define GET_LookToRead2 CLookToRead2 *p = CONTAINER_FROM_VTBL(pp, CLookToRead2, vt); + +static SRes LookToRead2_Look_Lookahead(const ILookInStream *pp, const void **buf, size_t *size) +{ + SRes res = SZ_OK; + GET_LookToRead2 + size_t size2 = p->size - p->pos; + if (size2 == 0 && *size != 0) + { + p->pos = 0; + p->size = 0; + size2 = p->bufSize; + res = ISeekInStream_Read(p->realStream, p->buf, &size2); + p->size = size2; + } + if (*size > size2) + *size = size2; + *buf = p->buf + p->pos; + return res; +} + +static SRes LookToRead2_Look_Exact(const ILookInStream *pp, const void **buf, size_t *size) +{ + SRes res = SZ_OK; + GET_LookToRead2 + size_t size2 = p->size - p->pos; + if (size2 == 0 && *size != 0) + { + p->pos = 0; + p->size = 0; + if (*size > p->bufSize) + *size = p->bufSize; + res = ISeekInStream_Read(p->realStream, p->buf, size); + size2 = p->size = *size; + } + if (*size > size2) + *size = size2; + *buf = p->buf + p->pos; + return res; +} + +static SRes LookToRead2_Skip(const ILookInStream *pp, size_t offset) +{ + GET_LookToRead2 + p->pos += offset; + return SZ_OK; +} + +static SRes LookToRead2_Read(const ILookInStream *pp, void *buf, size_t *size) +{ + GET_LookToRead2 + size_t rem = p->size - p->pos; + if (rem == 0) + return ISeekInStream_Read(p->realStream, buf, size); + if (rem > *size) + rem = *size; + memcpy(buf, p->buf + p->pos, rem); + p->pos += rem; + *size = rem; + return SZ_OK; +} + +static SRes LookToRead2_Seek(const ILookInStream *pp, Int64 *pos, ESzSeek origin) +{ + GET_LookToRead2 + p->pos = p->size = 0; + return ISeekInStream_Seek(p->realStream, pos, origin); +} + +void LookToRead2_CreateVTable(CLookToRead2 *p, int lookahead) +{ + p->vt.Look = lookahead ? + LookToRead2_Look_Lookahead : + LookToRead2_Look_Exact; + p->vt.Skip = LookToRead2_Skip; + p->vt.Read = LookToRead2_Read; + p->vt.Seek = LookToRead2_Seek; +} + + + +static SRes SecToLook_Read(const ISeqInStream *pp, void *buf, size_t *size) +{ + CSecToLook *p = CONTAINER_FROM_VTBL(pp, CSecToLook, vt); + return LookInStream_LookRead(p->realStream, buf, size); +} + +void SecToLook_CreateVTable(CSecToLook *p) +{ + p->vt.Read = SecToLook_Read; +} + +static SRes SecToRead_Read(const ISeqInStream *pp, void *buf, size_t *size) +{ + CSecToRead *p = CONTAINER_FROM_VTBL(pp, CSecToRead, vt); + return ILookInStream_Read(p->realStream, buf, size); +} + +void SecToRead_CreateVTable(CSecToRead *p) +{ + p->vt.Read = SecToRead_Read; +} diff --git a/3rdparty/lzma/src/Alloc.c b/3rdparty/lzma/src/Alloc.c new file mode 100644 index 0000000000..d1af76c5c3 --- /dev/null +++ b/3rdparty/lzma/src/Alloc.c @@ -0,0 +1,463 @@ +/* Alloc.c -- Memory allocation functions +2021-07-13 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#ifdef _WIN32 +#include +#endif +#include + +#include "Alloc.h" + +/* #define _SZ_ALLOC_DEBUG */ + +/* use _SZ_ALLOC_DEBUG to debug alloc/free operations */ +#ifdef _SZ_ALLOC_DEBUG + +#include +int g_allocCount = 0; +int g_allocCountMid = 0; +int g_allocCountBig = 0; + + +#define CONVERT_INT_TO_STR(charType, tempSize) \ + unsigned char temp[tempSize]; unsigned i = 0; \ + while (val >= 10) { temp[i++] = (unsigned char)('0' + (unsigned)(val % 10)); val /= 10; } \ + *s++ = (charType)('0' + (unsigned)val); \ + while (i != 0) { i--; *s++ = temp[i]; } \ + *s = 0; + +static void ConvertUInt64ToString(UInt64 val, char *s) +{ + CONVERT_INT_TO_STR(char, 24); +} + +#define GET_HEX_CHAR(t) ((char)(((t < 10) ? ('0' + t) : ('A' + (t - 10))))) + +static void ConvertUInt64ToHex(UInt64 val, char *s) +{ + UInt64 v = val; + unsigned i; + for (i = 1;; i++) + { + v >>= 4; + if (v == 0) + break; + } + s[i] = 0; + do + { + unsigned t = (unsigned)(val & 0xF); + val >>= 4; + s[--i] = GET_HEX_CHAR(t); + } + while (i); +} + +#define DEBUG_OUT_STREAM stderr + +static void Print(const char *s) +{ + fputs(s, DEBUG_OUT_STREAM); +} + +static void PrintAligned(const char *s, size_t align) +{ + size_t len = strlen(s); + for(;;) + { + fputc(' ', DEBUG_OUT_STREAM); + if (len >= align) + break; + ++len; + } + Print(s); +} + +static void PrintLn() +{ + Print("\n"); +} + +static void PrintHex(UInt64 v, size_t align) +{ + char s[32]; + ConvertUInt64ToHex(v, s); + PrintAligned(s, align); +} + +static void PrintDec(UInt64 v, size_t align) +{ + char s[32]; + ConvertUInt64ToString(v, s); + PrintAligned(s, align); +} + +static void PrintAddr(void *p) +{ + PrintHex((UInt64)(size_t)(ptrdiff_t)p, 12); +} + + +#define PRINT_ALLOC(name, cnt, size, ptr) \ + Print(name " "); \ + PrintDec(cnt++, 10); \ + PrintHex(size, 10); \ + PrintAddr(ptr); \ + PrintLn(); + +#define PRINT_FREE(name, cnt, ptr) if (ptr) { \ + Print(name " "); \ + PrintDec(--cnt, 10); \ + PrintAddr(ptr); \ + PrintLn(); } + +#else + +#define PRINT_ALLOC(name, cnt, size, ptr) +#define PRINT_FREE(name, cnt, ptr) +#define Print(s) +#define PrintLn() +#define PrintHex(v, align) +#define PrintAddr(p) + +#endif + + + +void *MyAlloc(size_t size) +{ + if (size == 0) + return NULL; + PRINT_ALLOC("Alloc ", g_allocCount, size, NULL); + #ifdef _SZ_ALLOC_DEBUG + { + void *p = malloc(size); + // PRINT_ALLOC("Alloc ", g_allocCount, size, p); + return p; + } + #else + return malloc(size); + #endif +} + +void MyFree(void *address) +{ + PRINT_FREE("Free ", g_allocCount, address); + + free(address); +} + +#ifdef _WIN32 + +void *MidAlloc(size_t size) +{ + if (size == 0) + return NULL; + + PRINT_ALLOC("Alloc-Mid", g_allocCountMid, size, NULL); + + return VirtualAlloc(NULL, size, MEM_COMMIT, PAGE_READWRITE); +} + +void MidFree(void *address) +{ + PRINT_FREE("Free-Mid", g_allocCountMid, address); + + if (!address) + return; + VirtualFree(address, 0, MEM_RELEASE); +} + +#ifdef _7ZIP_LARGE_PAGES + +#ifdef MEM_LARGE_PAGES + #define MY__MEM_LARGE_PAGES MEM_LARGE_PAGES +#else + #define MY__MEM_LARGE_PAGES 0x20000000 +#endif + +extern +SIZE_T g_LargePageSize; +SIZE_T g_LargePageSize = 0; +typedef SIZE_T (WINAPI *GetLargePageMinimumP)(VOID); + +#endif // _7ZIP_LARGE_PAGES + +void SetLargePageSize() +{ + #ifdef _7ZIP_LARGE_PAGES + SIZE_T size; + GetLargePageMinimumP largePageMinimum = (GetLargePageMinimumP) + GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "GetLargePageMinimum"); + if (!largePageMinimum) + return; + size = largePageMinimum(); + if (size == 0 || (size & (size - 1)) != 0) + return; + g_LargePageSize = size; + #endif +} + + +void *BigAlloc(size_t size) +{ + if (size == 0) + return NULL; + + PRINT_ALLOC("Alloc-Big", g_allocCountBig, size, NULL); + + #ifdef _7ZIP_LARGE_PAGES + { + SIZE_T ps = g_LargePageSize; + if (ps != 0 && ps <= (1 << 30) && size > (ps / 2)) + { + size_t size2; + ps--; + size2 = (size + ps) & ~ps; + if (size2 >= size) + { + void *res = VirtualAlloc(NULL, size2, MEM_COMMIT | MY__MEM_LARGE_PAGES, PAGE_READWRITE); + if (res) + return res; + } + } + } + #endif + + return VirtualAlloc(NULL, size, MEM_COMMIT, PAGE_READWRITE); +} + +void BigFree(void *address) +{ + PRINT_FREE("Free-Big", g_allocCountBig, address); + + if (!address) + return; + VirtualFree(address, 0, MEM_RELEASE); +} + +#endif + + +static void *SzAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MyAlloc(size); } +static void SzFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MyFree(address); } +const ISzAlloc g_Alloc = { SzAlloc, SzFree }; + +#ifdef _WIN32 +static void *SzMidAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return MidAlloc(size); } +static void SzMidFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); MidFree(address); } +static void *SzBigAlloc(ISzAllocPtr p, size_t size) { UNUSED_VAR(p); return BigAlloc(size); } +static void SzBigFree(ISzAllocPtr p, void *address) { UNUSED_VAR(p); BigFree(address); } +const ISzAlloc g_MidAlloc = { SzMidAlloc, SzMidFree }; +const ISzAlloc g_BigAlloc = { SzBigAlloc, SzBigFree }; +#endif + +/* + uintptr_t : C99 (optional) + : unsupported in VS6 +*/ + +#ifdef _WIN32 + typedef UINT_PTR UIntPtr; +#else + /* + typedef uintptr_t UIntPtr; + */ + typedef ptrdiff_t UIntPtr; +#endif + + +#define ADJUST_ALLOC_SIZE 0 +/* +#define ADJUST_ALLOC_SIZE (sizeof(void *) - 1) +*/ +/* + Use (ADJUST_ALLOC_SIZE = (sizeof(void *) - 1)), if + MyAlloc() can return address that is NOT multiple of sizeof(void *). +*/ + + +/* +#define MY_ALIGN_PTR_DOWN(p, align) ((void *)((char *)(p) - ((size_t)(UIntPtr)(p) & ((align) - 1)))) +*/ +#define MY_ALIGN_PTR_DOWN(p, align) ((void *)((((UIntPtr)(p)) & ~((UIntPtr)(align) - 1)))) + + +#if !defined(_WIN32) && defined(_POSIX_C_SOURCE) && (_POSIX_C_SOURCE >= 200112L) + #define USE_posix_memalign +#endif + +#ifndef USE_posix_memalign +#define MY_ALIGN_PTR_UP_PLUS(p, align) MY_ALIGN_PTR_DOWN(((char *)(p) + (align) + ADJUST_ALLOC_SIZE), align) +#endif + +/* + This posix_memalign() is for test purposes only. + We also need special Free() function instead of free(), + if this posix_memalign() is used. +*/ + +/* +static int posix_memalign(void **ptr, size_t align, size_t size) +{ + size_t newSize = size + align; + void *p; + void *pAligned; + *ptr = NULL; + if (newSize < size) + return 12; // ENOMEM + p = MyAlloc(newSize); + if (!p) + return 12; // ENOMEM + pAligned = MY_ALIGN_PTR_UP_PLUS(p, align); + ((void **)pAligned)[-1] = p; + *ptr = pAligned; + return 0; +} +*/ + +/* + ALLOC_ALIGN_SIZE >= sizeof(void *) + ALLOC_ALIGN_SIZE >= cache_line_size +*/ + +#define ALLOC_ALIGN_SIZE ((size_t)1 << 7) + +static void *SzAlignedAlloc(ISzAllocPtr pp, size_t size) +{ + #ifndef USE_posix_memalign + + void *p; + void *pAligned; + size_t newSize; + UNUSED_VAR(pp); + + /* also we can allocate additional dummy ALLOC_ALIGN_SIZE bytes after aligned + block to prevent cache line sharing with another allocated blocks */ + + newSize = size + ALLOC_ALIGN_SIZE * 1 + ADJUST_ALLOC_SIZE; + if (newSize < size) + return NULL; + + p = MyAlloc(newSize); + + if (!p) + return NULL; + pAligned = MY_ALIGN_PTR_UP_PLUS(p, ALLOC_ALIGN_SIZE); + + Print(" size="); PrintHex(size, 8); + Print(" a_size="); PrintHex(newSize, 8); + Print(" ptr="); PrintAddr(p); + Print(" a_ptr="); PrintAddr(pAligned); + PrintLn(); + + ((void **)pAligned)[-1] = p; + + return pAligned; + + #else + + void *p; + UNUSED_VAR(pp); + if (posix_memalign(&p, ALLOC_ALIGN_SIZE, size)) + return NULL; + + Print(" posix_memalign="); PrintAddr(p); + PrintLn(); + + return p; + + #endif +} + + +static void SzAlignedFree(ISzAllocPtr pp, void *address) +{ + UNUSED_VAR(pp); + #ifndef USE_posix_memalign + if (address) + MyFree(((void **)address)[-1]); + #else + free(address); + #endif +} + + +const ISzAlloc g_AlignedAlloc = { SzAlignedAlloc, SzAlignedFree }; + + + +#define MY_ALIGN_PTR_DOWN_1(p) MY_ALIGN_PTR_DOWN(p, sizeof(void *)) + +/* we align ptr to support cases where CAlignOffsetAlloc::offset is not multiply of sizeof(void *) */ +#define REAL_BLOCK_PTR_VAR(p) ((void **)MY_ALIGN_PTR_DOWN_1(p))[-1] +/* +#define REAL_BLOCK_PTR_VAR(p) ((void **)(p))[-1] +*/ + +static void *AlignOffsetAlloc_Alloc(ISzAllocPtr pp, size_t size) +{ + CAlignOffsetAlloc *p = CONTAINER_FROM_VTBL(pp, CAlignOffsetAlloc, vt); + void *adr; + void *pAligned; + size_t newSize; + size_t extra; + size_t alignSize = (size_t)1 << p->numAlignBits; + + if (alignSize < sizeof(void *)) + alignSize = sizeof(void *); + + if (p->offset >= alignSize) + return NULL; + + /* also we can allocate additional dummy ALLOC_ALIGN_SIZE bytes after aligned + block to prevent cache line sharing with another allocated blocks */ + extra = p->offset & (sizeof(void *) - 1); + newSize = size + alignSize + extra + ADJUST_ALLOC_SIZE; + if (newSize < size) + return NULL; + + adr = ISzAlloc_Alloc(p->baseAlloc, newSize); + + if (!adr) + return NULL; + + pAligned = (char *)MY_ALIGN_PTR_DOWN((char *)adr + + alignSize - p->offset + extra + ADJUST_ALLOC_SIZE, alignSize) + p->offset; + + PrintLn(); + Print("- Aligned: "); + Print(" size="); PrintHex(size, 8); + Print(" a_size="); PrintHex(newSize, 8); + Print(" ptr="); PrintAddr(adr); + Print(" a_ptr="); PrintAddr(pAligned); + PrintLn(); + + REAL_BLOCK_PTR_VAR(pAligned) = adr; + + return pAligned; +} + + +static void AlignOffsetAlloc_Free(ISzAllocPtr pp, void *address) +{ + if (address) + { + CAlignOffsetAlloc *p = CONTAINER_FROM_VTBL(pp, CAlignOffsetAlloc, vt); + PrintLn(); + Print("- Aligned Free: "); + PrintLn(); + ISzAlloc_Free(p->baseAlloc, REAL_BLOCK_PTR_VAR(address)); + } +} + + +void AlignOffsetAlloc_CreateVTable(CAlignOffsetAlloc *p) +{ + p->vt.Alloc = AlignOffsetAlloc_Alloc; + p->vt.Free = AlignOffsetAlloc_Free; +} diff --git a/3rdparty/lzma/src/Bcj2.c b/3rdparty/lzma/src/Bcj2.c new file mode 100644 index 0000000000..c7b956708f --- /dev/null +++ b/3rdparty/lzma/src/Bcj2.c @@ -0,0 +1,257 @@ +/* Bcj2.c -- BCJ2 Decoder (Converter for x86 code) +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Bcj2.h" +#include "CpuArch.h" + +#define CProb UInt16 + +#define kTopValue ((UInt32)1 << 24) +#define kNumModelBits 11 +#define kBitModelTotal (1 << kNumModelBits) +#define kNumMoveBits 5 + +#define _IF_BIT_0 ttt = *prob; bound = (p->range >> kNumModelBits) * ttt; if (p->code < bound) +#define _UPDATE_0 p->range = bound; *prob = (CProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); +#define _UPDATE_1 p->range -= bound; p->code -= bound; *prob = (CProb)(ttt - (ttt >> kNumMoveBits)); + +void Bcj2Dec_Init(CBcj2Dec *p) +{ + unsigned i; + + p->state = BCJ2_DEC_STATE_OK; + p->ip = 0; + p->temp[3] = 0; + p->range = 0; + p->code = 0; + for (i = 0; i < sizeof(p->probs) / sizeof(p->probs[0]); i++) + p->probs[i] = kBitModelTotal >> 1; +} + +SRes Bcj2Dec_Decode(CBcj2Dec *p) +{ + if (p->range <= 5) + { + p->state = BCJ2_DEC_STATE_OK; + for (; p->range != 5; p->range++) + { + if (p->range == 1 && p->code != 0) + return SZ_ERROR_DATA; + + if (p->bufs[BCJ2_STREAM_RC] == p->lims[BCJ2_STREAM_RC]) + { + p->state = BCJ2_STREAM_RC; + return SZ_OK; + } + + p->code = (p->code << 8) | *(p->bufs[BCJ2_STREAM_RC])++; + } + + if (p->code == 0xFFFFFFFF) + return SZ_ERROR_DATA; + + p->range = 0xFFFFFFFF; + } + else if (p->state >= BCJ2_DEC_STATE_ORIG_0) + { + while (p->state <= BCJ2_DEC_STATE_ORIG_3) + { + Byte *dest = p->dest; + if (dest == p->destLim) + return SZ_OK; + *dest = p->temp[(size_t)p->state - BCJ2_DEC_STATE_ORIG_0]; + p->state++; + p->dest = dest + 1; + } + } + + /* + if (BCJ2_IS_32BIT_STREAM(p->state)) + { + const Byte *cur = p->bufs[p->state]; + if (cur == p->lims[p->state]) + return SZ_OK; + p->bufs[p->state] = cur + 4; + + { + UInt32 val; + Byte *dest; + SizeT rem; + + p->ip += 4; + val = GetBe32(cur) - p->ip; + dest = p->dest; + rem = p->destLim - dest; + if (rem < 4) + { + SizeT i; + SetUi32(p->temp, val); + for (i = 0; i < rem; i++) + dest[i] = p->temp[i]; + p->dest = dest + rem; + p->state = BCJ2_DEC_STATE_ORIG_0 + (unsigned)rem; + return SZ_OK; + } + SetUi32(dest, val); + p->temp[3] = (Byte)(val >> 24); + p->dest = dest + 4; + p->state = BCJ2_DEC_STATE_OK; + } + } + */ + + for (;;) + { + if (BCJ2_IS_32BIT_STREAM(p->state)) + p->state = BCJ2_DEC_STATE_OK; + else + { + if (p->range < kTopValue) + { + if (p->bufs[BCJ2_STREAM_RC] == p->lims[BCJ2_STREAM_RC]) + { + p->state = BCJ2_STREAM_RC; + return SZ_OK; + } + p->range <<= 8; + p->code = (p->code << 8) | *(p->bufs[BCJ2_STREAM_RC])++; + } + + { + const Byte *src = p->bufs[BCJ2_STREAM_MAIN]; + const Byte *srcLim; + Byte *dest; + SizeT num = (SizeT)(p->lims[BCJ2_STREAM_MAIN] - src); + + if (num == 0) + { + p->state = BCJ2_STREAM_MAIN; + return SZ_OK; + } + + dest = p->dest; + if (num > (SizeT)(p->destLim - dest)) + { + num = (SizeT)(p->destLim - dest); + if (num == 0) + { + p->state = BCJ2_DEC_STATE_ORIG; + return SZ_OK; + } + } + + srcLim = src + num; + + if (p->temp[3] == 0x0F && (src[0] & 0xF0) == 0x80) + *dest = src[0]; + else for (;;) + { + Byte b = *src; + *dest = b; + if (b != 0x0F) + { + if ((b & 0xFE) == 0xE8) + break; + dest++; + if (++src != srcLim) + continue; + break; + } + dest++; + if (++src == srcLim) + break; + if ((*src & 0xF0) != 0x80) + continue; + *dest = *src; + break; + } + + num = (SizeT)(src - p->bufs[BCJ2_STREAM_MAIN]); + + if (src == srcLim) + { + p->temp[3] = src[-1]; + p->bufs[BCJ2_STREAM_MAIN] = src; + p->ip += (UInt32)num; + p->dest += num; + p->state = + p->bufs[BCJ2_STREAM_MAIN] == + p->lims[BCJ2_STREAM_MAIN] ? + (unsigned)BCJ2_STREAM_MAIN : + (unsigned)BCJ2_DEC_STATE_ORIG; + return SZ_OK; + } + + { + UInt32 bound, ttt; + CProb *prob; + Byte b = src[0]; + Byte prev = (Byte)(num == 0 ? p->temp[3] : src[-1]); + + p->temp[3] = b; + p->bufs[BCJ2_STREAM_MAIN] = src + 1; + num++; + p->ip += (UInt32)num; + p->dest += num; + + prob = p->probs + (unsigned)(b == 0xE8 ? 2 + (unsigned)prev : (b == 0xE9 ? 1 : 0)); + + _IF_BIT_0 + { + _UPDATE_0 + continue; + } + _UPDATE_1 + + } + } + } + + { + UInt32 val; + unsigned cj = (p->temp[3] == 0xE8) ? BCJ2_STREAM_CALL : BCJ2_STREAM_JUMP; + const Byte *cur = p->bufs[cj]; + Byte *dest; + SizeT rem; + + if (cur == p->lims[cj]) + { + p->state = cj; + break; + } + + val = GetBe32(cur); + p->bufs[cj] = cur + 4; + + p->ip += 4; + val -= p->ip; + dest = p->dest; + rem = (SizeT)(p->destLim - dest); + + if (rem < 4) + { + p->temp[0] = (Byte)val; if (rem > 0) dest[0] = (Byte)val; val >>= 8; + p->temp[1] = (Byte)val; if (rem > 1) dest[1] = (Byte)val; val >>= 8; + p->temp[2] = (Byte)val; if (rem > 2) dest[2] = (Byte)val; val >>= 8; + p->temp[3] = (Byte)val; + p->dest = dest + rem; + p->state = BCJ2_DEC_STATE_ORIG_0 + (unsigned)rem; + break; + } + + SetUi32(dest, val); + p->temp[3] = (Byte)(val >> 24); + p->dest = dest + 4; + } + } + + if (p->range < kTopValue && p->bufs[BCJ2_STREAM_RC] != p->lims[BCJ2_STREAM_RC]) + { + p->range <<= 8; + p->code = (p->code << 8) | *(p->bufs[BCJ2_STREAM_RC])++; + } + + return SZ_OK; +} diff --git a/3rdparty/lzma/src/Bcj2Enc.c b/3rdparty/lzma/src/Bcj2Enc.c new file mode 100644 index 0000000000..682362a15d --- /dev/null +++ b/3rdparty/lzma/src/Bcj2Enc.c @@ -0,0 +1,311 @@ +/* Bcj2Enc.c -- BCJ2 Encoder (Converter for x86 code) +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +/* #define SHOW_STAT */ + +#ifdef SHOW_STAT +#include +#define PRF(x) x +#else +#define PRF(x) +#endif + +#include + +#include "Bcj2.h" +#include "CpuArch.h" + +#define CProb UInt16 + +#define kTopValue ((UInt32)1 << 24) +#define kNumModelBits 11 +#define kBitModelTotal (1 << kNumModelBits) +#define kNumMoveBits 5 + +void Bcj2Enc_Init(CBcj2Enc *p) +{ + unsigned i; + + p->state = BCJ2_ENC_STATE_OK; + p->finishMode = BCJ2_ENC_FINISH_MODE_CONTINUE; + + p->prevByte = 0; + + p->cache = 0; + p->range = 0xFFFFFFFF; + p->low = 0; + p->cacheSize = 1; + + p->ip = 0; + + p->fileIp = 0; + p->fileSize = 0; + p->relatLimit = BCJ2_RELAT_LIMIT; + + p->tempPos = 0; + + p->flushPos = 0; + + for (i = 0; i < sizeof(p->probs) / sizeof(p->probs[0]); i++) + p->probs[i] = kBitModelTotal >> 1; +} + +static BoolInt MY_FAST_CALL RangeEnc_ShiftLow(CBcj2Enc *p) +{ + if ((UInt32)p->low < (UInt32)0xFF000000 || (UInt32)(p->low >> 32) != 0) + { + Byte *buf = p->bufs[BCJ2_STREAM_RC]; + do + { + if (buf == p->lims[BCJ2_STREAM_RC]) + { + p->state = BCJ2_STREAM_RC; + p->bufs[BCJ2_STREAM_RC] = buf; + return True; + } + *buf++ = (Byte)(p->cache + (Byte)(p->low >> 32)); + p->cache = 0xFF; + } + while (--p->cacheSize); + p->bufs[BCJ2_STREAM_RC] = buf; + p->cache = (Byte)((UInt32)p->low >> 24); + } + p->cacheSize++; + p->low = (UInt32)p->low << 8; + return False; +} + +static void Bcj2Enc_Encode_2(CBcj2Enc *p) +{ + if (BCJ2_IS_32BIT_STREAM(p->state)) + { + Byte *cur = p->bufs[p->state]; + if (cur == p->lims[p->state]) + return; + SetBe32(cur, p->tempTarget); + p->bufs[p->state] = cur + 4; + } + + p->state = BCJ2_ENC_STATE_ORIG; + + for (;;) + { + if (p->range < kTopValue) + { + if (RangeEnc_ShiftLow(p)) + return; + p->range <<= 8; + } + + { + { + const Byte *src = p->src; + const Byte *srcLim; + Byte *dest; + SizeT num = (SizeT)(p->srcLim - src); + + if (p->finishMode == BCJ2_ENC_FINISH_MODE_CONTINUE) + { + if (num <= 4) + return; + num -= 4; + } + else if (num == 0) + break; + + dest = p->bufs[BCJ2_STREAM_MAIN]; + if (num > (SizeT)(p->lims[BCJ2_STREAM_MAIN] - dest)) + { + num = (SizeT)(p->lims[BCJ2_STREAM_MAIN] - dest); + if (num == 0) + { + p->state = BCJ2_STREAM_MAIN; + return; + } + } + + srcLim = src + num; + + if (p->prevByte == 0x0F && (src[0] & 0xF0) == 0x80) + *dest = src[0]; + else for (;;) + { + Byte b = *src; + *dest = b; + if (b != 0x0F) + { + if ((b & 0xFE) == 0xE8) + break; + dest++; + if (++src != srcLim) + continue; + break; + } + dest++; + if (++src == srcLim) + break; + if ((*src & 0xF0) != 0x80) + continue; + *dest = *src; + break; + } + + num = (SizeT)(src - p->src); + + if (src == srcLim) + { + p->prevByte = src[-1]; + p->bufs[BCJ2_STREAM_MAIN] = dest; + p->src = src; + p->ip += (UInt32)num; + continue; + } + + { + Byte context = (Byte)(num == 0 ? p->prevByte : src[-1]); + BoolInt needConvert; + + p->bufs[BCJ2_STREAM_MAIN] = dest + 1; + p->ip += (UInt32)num + 1; + src++; + + needConvert = False; + + if ((SizeT)(p->srcLim - src) >= 4) + { + UInt32 relatVal = GetUi32(src); + if ((p->fileSize == 0 || (UInt32)(p->ip + 4 + relatVal - p->fileIp) < p->fileSize) + && ((relatVal + p->relatLimit) >> 1) < p->relatLimit) + needConvert = True; + } + + { + UInt32 bound; + unsigned ttt; + Byte b = src[-1]; + CProb *prob = p->probs + (unsigned)(b == 0xE8 ? 2 + (unsigned)context : (b == 0xE9 ? 1 : 0)); + + ttt = *prob; + bound = (p->range >> kNumModelBits) * ttt; + + if (!needConvert) + { + p->range = bound; + *prob = (CProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); + p->src = src; + p->prevByte = b; + continue; + } + + p->low += bound; + p->range -= bound; + *prob = (CProb)(ttt - (ttt >> kNumMoveBits)); + + { + UInt32 relatVal = GetUi32(src); + UInt32 absVal; + p->ip += 4; + absVal = p->ip + relatVal; + p->prevByte = src[3]; + src += 4; + p->src = src; + { + unsigned cj = (b == 0xE8) ? BCJ2_STREAM_CALL : BCJ2_STREAM_JUMP; + Byte *cur = p->bufs[cj]; + if (cur == p->lims[cj]) + { + p->state = cj; + p->tempTarget = absVal; + return; + } + SetBe32(cur, absVal); + p->bufs[cj] = cur + 4; + } + } + } + } + } + } + } + + if (p->finishMode != BCJ2_ENC_FINISH_MODE_END_STREAM) + return; + + for (; p->flushPos < 5; p->flushPos++) + if (RangeEnc_ShiftLow(p)) + return; + p->state = BCJ2_ENC_STATE_OK; +} + + +void Bcj2Enc_Encode(CBcj2Enc *p) +{ + PRF(printf("\n")); + PRF(printf("---- ip = %8d tempPos = %8d src = %8d\n", p->ip, p->tempPos, p->srcLim - p->src)); + + if (p->tempPos != 0) + { + unsigned extra = 0; + + for (;;) + { + const Byte *src = p->src; + const Byte *srcLim = p->srcLim; + EBcj2Enc_FinishMode finishMode = p->finishMode; + + p->src = p->temp; + p->srcLim = p->temp + p->tempPos; + if (src != srcLim) + p->finishMode = BCJ2_ENC_FINISH_MODE_CONTINUE; + + PRF(printf(" ip = %8d tempPos = %8d src = %8d\n", p->ip, p->tempPos, p->srcLim - p->src)); + + Bcj2Enc_Encode_2(p); + + { + unsigned num = (unsigned)(p->src - p->temp); + unsigned tempPos = p->tempPos - num; + unsigned i; + p->tempPos = tempPos; + for (i = 0; i < tempPos; i++) + p->temp[i] = p->temp[(size_t)i + num]; + + p->src = src; + p->srcLim = srcLim; + p->finishMode = finishMode; + + if (p->state != BCJ2_ENC_STATE_ORIG || src == srcLim) + return; + + if (extra >= tempPos) + { + p->src = src - tempPos; + p->tempPos = 0; + break; + } + + p->temp[tempPos] = src[0]; + p->tempPos = tempPos + 1; + p->src = src + 1; + extra++; + } + } + } + + PRF(printf("++++ ip = %8d tempPos = %8d src = %8d\n", p->ip, p->tempPos, p->srcLim - p->src)); + + Bcj2Enc_Encode_2(p); + + if (p->state == BCJ2_ENC_STATE_ORIG) + { + const Byte *src = p->src; + unsigned rem = (unsigned)(p->srcLim - src); + unsigned i; + for (i = 0; i < rem; i++) + p->temp[i] = src[i]; + p->tempPos = rem; + p->src = src + rem; + } +} diff --git a/3rdparty/lzma/src/Bra.c b/3rdparty/lzma/src/Bra.c new file mode 100644 index 0000000000..3b854d9cad --- /dev/null +++ b/3rdparty/lzma/src/Bra.c @@ -0,0 +1,230 @@ +/* Bra.c -- Converters for RISC code +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" +#include "Bra.h" + +SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + Byte *p; + const Byte *lim; + size &= ~(size_t)3; + ip += 4; + p = data; + lim = data + size; + + if (encoding) + + for (;;) + { + for (;;) + { + if (p >= lim) + return (SizeT)(p - data); + p += 4; + if (p[-1] == 0xEB) + break; + } + { + UInt32 v = GetUi32(p - 4); + v <<= 2; + v += ip + (UInt32)(p - data); + v >>= 2; + v &= 0x00FFFFFF; + v |= 0xEB000000; + SetUi32(p - 4, v); + } + } + + for (;;) + { + for (;;) + { + if (p >= lim) + return (SizeT)(p - data); + p += 4; + if (p[-1] == 0xEB) + break; + } + { + UInt32 v = GetUi32(p - 4); + v <<= 2; + v -= ip + (UInt32)(p - data); + v >>= 2; + v &= 0x00FFFFFF; + v |= 0xEB000000; + SetUi32(p - 4, v); + } + } +} + + +SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + Byte *p; + const Byte *lim; + size &= ~(size_t)1; + p = data; + lim = data + size - 4; + + if (encoding) + + for (;;) + { + UInt32 b1; + for (;;) + { + UInt32 b3; + if (p > lim) + return (SizeT)(p - data); + b1 = p[1]; + b3 = p[3]; + p += 2; + b1 ^= 8; + if ((b3 & b1) >= 0xF8) + break; + } + { + UInt32 v = + ((UInt32)b1 << 19) + + (((UInt32)p[1] & 0x7) << 8) + + (((UInt32)p[-2] << 11)) + + (p[0]); + + p += 2; + { + UInt32 cur = (ip + (UInt32)(p - data)) >> 1; + v += cur; + } + + p[-4] = (Byte)(v >> 11); + p[-3] = (Byte)(0xF0 | ((v >> 19) & 0x7)); + p[-2] = (Byte)v; + p[-1] = (Byte)(0xF8 | (v >> 8)); + } + } + + for (;;) + { + UInt32 b1; + for (;;) + { + UInt32 b3; + if (p > lim) + return (SizeT)(p - data); + b1 = p[1]; + b3 = p[3]; + p += 2; + b1 ^= 8; + if ((b3 & b1) >= 0xF8) + break; + } + { + UInt32 v = + ((UInt32)b1 << 19) + + (((UInt32)p[1] & 0x7) << 8) + + (((UInt32)p[-2] << 11)) + + (p[0]); + + p += 2; + { + UInt32 cur = (ip + (UInt32)(p - data)) >> 1; + v -= cur; + } + + /* + SetUi16(p - 4, (UInt16)(((v >> 11) & 0x7FF) | 0xF000)); + SetUi16(p - 2, (UInt16)(v | 0xF800)); + */ + + p[-4] = (Byte)(v >> 11); + p[-3] = (Byte)(0xF0 | ((v >> 19) & 0x7)); + p[-2] = (Byte)v; + p[-1] = (Byte)(0xF8 | (v >> 8)); + } + } +} + + +SizeT PPC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + Byte *p; + const Byte *lim; + size &= ~(size_t)3; + ip -= 4; + p = data; + lim = data + size; + + for (;;) + { + for (;;) + { + if (p >= lim) + return (SizeT)(p - data); + p += 4; + /* if ((v & 0xFC000003) == 0x48000001) */ + if ((p[-4] & 0xFC) == 0x48 && (p[-1] & 3) == 1) + break; + } + { + UInt32 v = GetBe32(p - 4); + if (encoding) + v += ip + (UInt32)(p - data); + else + v -= ip + (UInt32)(p - data); + v &= 0x03FFFFFF; + v |= 0x48000000; + SetBe32(p - 4, v); + } + } +} + + +SizeT SPARC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + Byte *p; + const Byte *lim; + size &= ~(size_t)3; + ip -= 4; + p = data; + lim = data + size; + + for (;;) + { + for (;;) + { + if (p >= lim) + return (SizeT)(p - data); + /* + v = GetBe32(p); + p += 4; + m = v + ((UInt32)5 << 29); + m ^= (UInt32)7 << 29; + m += (UInt32)1 << 22; + if ((m & ((UInt32)0x1FF << 23)) == 0) + break; + */ + p += 4; + if ((p[-4] == 0x40 && (p[-3] & 0xC0) == 0) || + (p[-4] == 0x7F && (p[-3] >= 0xC0))) + break; + } + { + UInt32 v = GetBe32(p - 4); + v <<= 2; + if (encoding) + v += ip + (UInt32)(p - data); + else + v -= ip + (UInt32)(p - data); + + v &= 0x01FFFFFF; + v -= (UInt32)1 << 24; + v ^= 0xFF000000; + v >>= 2; + v |= 0x40000000; + SetBe32(p - 4, v); + } + } +} diff --git a/3rdparty/lzma/src/Bra86.c b/3rdparty/lzma/src/Bra86.c new file mode 100644 index 0000000000..10a0fbd161 --- /dev/null +++ b/3rdparty/lzma/src/Bra86.c @@ -0,0 +1,82 @@ +/* Bra86.c -- Converter for x86 code (BCJ) +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Bra.h" + +#define Test86MSByte(b) ((((b) + 1) & 0xFE) == 0) + +SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding) +{ + SizeT pos = 0; + UInt32 mask = *state & 7; + if (size < 5) + return 0; + size -= 4; + ip += 5; + + for (;;) + { + Byte *p = data + pos; + const Byte *limit = data + size; + for (; p < limit; p++) + if ((*p & 0xFE) == 0xE8) + break; + + { + SizeT d = (SizeT)(p - data) - pos; + pos = (SizeT)(p - data); + if (p >= limit) + { + *state = (d > 2 ? 0 : mask >> (unsigned)d); + return pos; + } + if (d > 2) + mask = 0; + else + { + mask >>= (unsigned)d; + if (mask != 0 && (mask > 4 || mask == 3 || Test86MSByte(p[(size_t)(mask >> 1) + 1]))) + { + mask = (mask >> 1) | 4; + pos++; + continue; + } + } + } + + if (Test86MSByte(p[4])) + { + UInt32 v = ((UInt32)p[4] << 24) | ((UInt32)p[3] << 16) | ((UInt32)p[2] << 8) | ((UInt32)p[1]); + UInt32 cur = ip + (UInt32)pos; + pos += 5; + if (encoding) + v += cur; + else + v -= cur; + if (mask != 0) + { + unsigned sh = (mask & 6) << 2; + if (Test86MSByte((Byte)(v >> sh))) + { + v ^= (((UInt32)0x100 << sh) - 1); + if (encoding) + v += cur; + else + v -= cur; + } + mask = 0; + } + p[1] = (Byte)v; + p[2] = (Byte)(v >> 8); + p[3] = (Byte)(v >> 16); + p[4] = (Byte)(0 - ((v >> 24) & 1)); + } + else + { + mask = (mask >> 1) | 4; + pos++; + } + } +} diff --git a/3rdparty/lzma/src/BraIA64.c b/3rdparty/lzma/src/BraIA64.c new file mode 100644 index 0000000000..d1dbc62c55 --- /dev/null +++ b/3rdparty/lzma/src/BraIA64.c @@ -0,0 +1,53 @@ +/* BraIA64.c -- Converter for IA-64 code +2017-01-26 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" +#include "Bra.h" + +SizeT IA64_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + SizeT i; + if (size < 16) + return 0; + size -= 16; + i = 0; + do + { + unsigned m = ((UInt32)0x334B0000 >> (data[i] & 0x1E)) & 3; + if (m) + { + m++; + do + { + Byte *p = data + (i + (size_t)m * 5 - 8); + if (((p[3] >> m) & 15) == 5 + && (((p[-1] | ((UInt32)p[0] << 8)) >> m) & 0x70) == 0) + { + unsigned raw = GetUi32(p); + unsigned v = raw >> m; + v = (v & 0xFFFFF) | ((v & (1 << 23)) >> 3); + + v <<= 4; + if (encoding) + v += ip + (UInt32)i; + else + v -= ip + (UInt32)i; + v >>= 4; + + v &= 0x1FFFFF; + v += 0x700000; + v &= 0x8FFFFF; + raw &= ~((UInt32)0x8FFFFF << m); + raw |= (v << m); + SetUi32(p, raw); + } + } + while (++m <= 4); + } + i += 16; + } + while (i <= size); + return i; +} diff --git a/3rdparty/lzma/src/CpuArch.c b/3rdparty/lzma/src/CpuArch.c new file mode 100644 index 0000000000..fa9afe3970 --- /dev/null +++ b/3rdparty/lzma/src/CpuArch.c @@ -0,0 +1,478 @@ +/* CpuArch.c -- CPU specific code +2021-07-13 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" + +#ifdef MY_CPU_X86_OR_AMD64 + +#if (defined(_MSC_VER) && !defined(MY_CPU_AMD64)) || defined(__GNUC__) +#define USE_ASM +#endif + +#if !defined(USE_ASM) && _MSC_VER >= 1500 +#include +#endif + +#if defined(USE_ASM) && !defined(MY_CPU_AMD64) +static UInt32 CheckFlag(UInt32 flag) +{ + #ifdef _MSC_VER + __asm pushfd; + __asm pop EAX; + __asm mov EDX, EAX; + __asm xor EAX, flag; + __asm push EAX; + __asm popfd; + __asm pushfd; + __asm pop EAX; + __asm xor EAX, EDX; + __asm push EDX; + __asm popfd; + __asm and flag, EAX; + #else + __asm__ __volatile__ ( + "pushf\n\t" + "pop %%EAX\n\t" + "movl %%EAX,%%EDX\n\t" + "xorl %0,%%EAX\n\t" + "push %%EAX\n\t" + "popf\n\t" + "pushf\n\t" + "pop %%EAX\n\t" + "xorl %%EDX,%%EAX\n\t" + "push %%EDX\n\t" + "popf\n\t" + "andl %%EAX, %0\n\t": + "=c" (flag) : "c" (flag) : + "%eax", "%edx"); + #endif + return flag; +} +#define CHECK_CPUID_IS_SUPPORTED if (CheckFlag(1 << 18) == 0 || CheckFlag(1 << 21) == 0) return False; +#else +#define CHECK_CPUID_IS_SUPPORTED +#endif + +#ifndef USE_ASM + #ifdef _MSC_VER + #if _MSC_VER >= 1600 + #define MY__cpuidex __cpuidex + #else + +/* + __cpuid (function == 4) requires subfunction number in ECX. + MSDN: The __cpuid intrinsic clears the ECX register before calling the cpuid instruction. + __cpuid() in new MSVC clears ECX. + __cpuid() in old MSVC (14.00) doesn't clear ECX + We still can use __cpuid for low (function) values that don't require ECX, + but __cpuid() in old MSVC will be incorrect for some function values: (function == 4). + So here we use the hack for old MSVC to send (subFunction) in ECX register to cpuid instruction, + where ECX value is first parameter for FAST_CALL / NO_INLINE function, + So the caller of MY__cpuidex_HACK() sets ECX as subFunction, and + old MSVC for __cpuid() doesn't change ECX and cpuid instruction gets (subFunction) value. + + DON'T remove MY_NO_INLINE and MY_FAST_CALL for MY__cpuidex_HACK() !!! +*/ + +static +MY_NO_INLINE +void MY_FAST_CALL MY__cpuidex_HACK(UInt32 subFunction, int *CPUInfo, UInt32 function) +{ + UNUSED_VAR(subFunction); + __cpuid(CPUInfo, function); +} + + #define MY__cpuidex(info, func, func2) MY__cpuidex_HACK(func2, info, func) + #pragma message("======== MY__cpuidex_HACK WAS USED ========") + #endif + #else + #define MY__cpuidex(info, func, func2) __cpuid(info, func) + #pragma message("======== (INCORRECT ?) cpuid WAS USED ========") + #endif +#endif + + + + +void MyCPUID(UInt32 function, UInt32 *a, UInt32 *b, UInt32 *c, UInt32 *d) +{ + #ifdef USE_ASM + + #ifdef _MSC_VER + + UInt32 a2, b2, c2, d2; + __asm xor EBX, EBX; + __asm xor ECX, ECX; + __asm xor EDX, EDX; + __asm mov EAX, function; + __asm cpuid; + __asm mov a2, EAX; + __asm mov b2, EBX; + __asm mov c2, ECX; + __asm mov d2, EDX; + + *a = a2; + *b = b2; + *c = c2; + *d = d2; + + #else + + __asm__ __volatile__ ( + #if defined(MY_CPU_AMD64) && defined(__PIC__) + "mov %%rbx, %%rdi;" + "cpuid;" + "xchg %%rbx, %%rdi;" + : "=a" (*a) , + "=D" (*b) , + #elif defined(MY_CPU_X86) && defined(__PIC__) + "mov %%ebx, %%edi;" + "cpuid;" + "xchgl %%ebx, %%edi;" + : "=a" (*a) , + "=D" (*b) , + #else + "cpuid" + : "=a" (*a) , + "=b" (*b) , + #endif + "=c" (*c) , + "=d" (*d) + : "0" (function), "c"(0) ) ; + + #endif + + #else + + int CPUInfo[4]; + + MY__cpuidex(CPUInfo, (int)function, 0); + + *a = (UInt32)CPUInfo[0]; + *b = (UInt32)CPUInfo[1]; + *c = (UInt32)CPUInfo[2]; + *d = (UInt32)CPUInfo[3]; + + #endif +} + +BoolInt x86cpuid_CheckAndRead(Cx86cpuid *p) +{ + CHECK_CPUID_IS_SUPPORTED + MyCPUID(0, &p->maxFunc, &p->vendor[0], &p->vendor[2], &p->vendor[1]); + MyCPUID(1, &p->ver, &p->b, &p->c, &p->d); + return True; +} + +static const UInt32 kVendors[][3] = +{ + { 0x756E6547, 0x49656E69, 0x6C65746E}, + { 0x68747541, 0x69746E65, 0x444D4163}, + { 0x746E6543, 0x48727561, 0x736C7561} +}; + +int x86cpuid_GetFirm(const Cx86cpuid *p) +{ + unsigned i; + for (i = 0; i < sizeof(kVendors) / sizeof(kVendors[i]); i++) + { + const UInt32 *v = kVendors[i]; + if (v[0] == p->vendor[0] && + v[1] == p->vendor[1] && + v[2] == p->vendor[2]) + return (int)i; + } + return -1; +} + +BoolInt CPU_Is_InOrder() +{ + Cx86cpuid p; + int firm; + UInt32 family, model; + if (!x86cpuid_CheckAndRead(&p)) + return True; + + family = x86cpuid_GetFamily(p.ver); + model = x86cpuid_GetModel(p.ver); + + firm = x86cpuid_GetFirm(&p); + + switch (firm) + { + case CPU_FIRM_INTEL: return (family < 6 || (family == 6 && ( + /* In-Order Atom CPU */ + model == 0x1C /* 45 nm, N4xx, D4xx, N5xx, D5xx, 230, 330 */ + || model == 0x26 /* 45 nm, Z6xx */ + || model == 0x27 /* 32 nm, Z2460 */ + || model == 0x35 /* 32 nm, Z2760 */ + || model == 0x36 /* 32 nm, N2xxx, D2xxx */ + ))); + case CPU_FIRM_AMD: return (family < 5 || (family == 5 && (model < 6 || model == 0xA))); + case CPU_FIRM_VIA: return (family < 6 || (family == 6 && model < 0xF)); + } + return True; +} + +#if !defined(MY_CPU_AMD64) && defined(_WIN32) +#include +static BoolInt CPU_Sys_Is_SSE_Supported() +{ + OSVERSIONINFO vi; + vi.dwOSVersionInfoSize = sizeof(vi); + if (!GetVersionEx(&vi)) + return False; + return (vi.dwMajorVersion >= 5); +} +#define CHECK_SYS_SSE_SUPPORT if (!CPU_Sys_Is_SSE_Supported()) return False; +#else +#define CHECK_SYS_SSE_SUPPORT +#endif + + +static UInt32 X86_CPUID_ECX_Get_Flags() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + if (!x86cpuid_CheckAndRead(&p)) + return 0; + return p.c; +} + +BoolInt CPU_IsSupported_AES() +{ + return (X86_CPUID_ECX_Get_Flags() >> 25) & 1; +} + +BoolInt CPU_IsSupported_SSSE3() +{ + return (X86_CPUID_ECX_Get_Flags() >> 9) & 1; +} + +BoolInt CPU_IsSupported_SSE41() +{ + return (X86_CPUID_ECX_Get_Flags() >> 19) & 1; +} + +BoolInt CPU_IsSupported_SHA() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + if (!x86cpuid_CheckAndRead(&p)) + return False; + + if (p.maxFunc < 7) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); + return (d[1] >> 29) & 1; + } +} + +// #include + +#ifdef _WIN32 +#include +#endif + +BoolInt CPU_IsSupported_AVX2() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + + #ifdef _WIN32 + #define MY__PF_XSAVE_ENABLED 17 + if (!IsProcessorFeaturePresent(MY__PF_XSAVE_ENABLED)) + return False; + #endif + + if (!x86cpuid_CheckAndRead(&p)) + return False; + if (p.maxFunc < 7) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); + // printf("\ncpuid(7): ebx=%8x ecx=%8x\n", d[1], d[2]); + return 1 + & (d[1] >> 5); // avx2 + } +} + +BoolInt CPU_IsSupported_VAES_AVX2() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + + #ifdef _WIN32 + #define MY__PF_XSAVE_ENABLED 17 + if (!IsProcessorFeaturePresent(MY__PF_XSAVE_ENABLED)) + return False; + #endif + + if (!x86cpuid_CheckAndRead(&p)) + return False; + if (p.maxFunc < 7) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(7, &d[0], &d[1], &d[2], &d[3]); + // printf("\ncpuid(7): ebx=%8x ecx=%8x\n", d[1], d[2]); + return 1 + & (d[1] >> 5) // avx2 + // & (d[1] >> 31) // avx512vl + & (d[2] >> 9); // vaes // VEX-256/EVEX + } +} + +BoolInt CPU_IsSupported_PageGB() +{ + Cx86cpuid cpuid; + if (!x86cpuid_CheckAndRead(&cpuid)) + return False; + { + UInt32 d[4] = { 0 }; + MyCPUID(0x80000000, &d[0], &d[1], &d[2], &d[3]); + if (d[0] < 0x80000001) + return False; + } + { + UInt32 d[4] = { 0 }; + MyCPUID(0x80000001, &d[0], &d[1], &d[2], &d[3]); + return (d[3] >> 26) & 1; + } +} + + +#elif defined(MY_CPU_ARM_OR_ARM64) + +#ifdef _WIN32 + +#include + +BoolInt CPU_IsSupported_CRC32() { return IsProcessorFeaturePresent(PF_ARM_V8_CRC32_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } +BoolInt CPU_IsSupported_CRYPTO() { return IsProcessorFeaturePresent(PF_ARM_V8_CRYPTO_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } +BoolInt CPU_IsSupported_NEON() { return IsProcessorFeaturePresent(PF_ARM_NEON_INSTRUCTIONS_AVAILABLE) ? 1 : 0; } + +#else + +#if defined(__APPLE__) + +/* +#include +#include +static void Print_sysctlbyname(const char *name) +{ + size_t bufSize = 256; + char buf[256]; + int res = sysctlbyname(name, &buf, &bufSize, NULL, 0); + { + int i; + printf("\nres = %d : %s : '%s' : bufSize = %d, numeric", res, name, buf, (unsigned)bufSize); + for (i = 0; i < 20; i++) + printf(" %2x", (unsigned)(Byte)buf[i]); + + } +} +*/ + +static BoolInt My_sysctlbyname_Get_BoolInt(const char *name) +{ + UInt32 val = 0; + if (My_sysctlbyname_Get_UInt32(name, &val) == 0 && val == 1) + return 1; + return 0; +} + + /* + Print_sysctlbyname("hw.pagesize"); + Print_sysctlbyname("machdep.cpu.brand_string"); + */ + +BoolInt CPU_IsSupported_CRC32(void) +{ + return My_sysctlbyname_Get_BoolInt("hw.optional.armv8_crc32"); +} + +BoolInt CPU_IsSupported_NEON(void) +{ + return My_sysctlbyname_Get_BoolInt("hw.optional.neon"); +} + +#ifdef MY_CPU_ARM64 +#define APPLE_CRYPTO_SUPPORT_VAL 1 +#else +#define APPLE_CRYPTO_SUPPORT_VAL 0 +#endif + +BoolInt CPU_IsSupported_SHA1(void) { return APPLE_CRYPTO_SUPPORT_VAL; } +BoolInt CPU_IsSupported_SHA2(void) { return APPLE_CRYPTO_SUPPORT_VAL; } +BoolInt CPU_IsSupported_AES (void) { return APPLE_CRYPTO_SUPPORT_VAL; } + + +#else // __APPLE__ + +#include + +#define USE_HWCAP + +#ifdef USE_HWCAP + +#include + + #define MY_HWCAP_CHECK_FUNC_2(name1, name2) \ + BoolInt CPU_IsSupported_ ## name1() { return (getauxval(AT_HWCAP) & (HWCAP_ ## name2)) ? 1 : 0; } + +#ifdef MY_CPU_ARM64 + #define MY_HWCAP_CHECK_FUNC(name) \ + MY_HWCAP_CHECK_FUNC_2(name, name) + MY_HWCAP_CHECK_FUNC_2(NEON, ASIMD) +// MY_HWCAP_CHECK_FUNC (ASIMD) +#elif defined(MY_CPU_ARM) + #define MY_HWCAP_CHECK_FUNC(name) \ + BoolInt CPU_IsSupported_ ## name() { return (getauxval(AT_HWCAP2) & (HWCAP2_ ## name)) ? 1 : 0; } + MY_HWCAP_CHECK_FUNC_2(NEON, NEON) +#endif + +#else // USE_HWCAP + + #define MY_HWCAP_CHECK_FUNC(name) \ + BoolInt CPU_IsSupported_ ## name() { return 0; } + MY_HWCAP_CHECK_FUNC(NEON) + +#endif // USE_HWCAP + +MY_HWCAP_CHECK_FUNC (CRC32) +MY_HWCAP_CHECK_FUNC (SHA1) +MY_HWCAP_CHECK_FUNC (SHA2) +MY_HWCAP_CHECK_FUNC (AES) + +#endif // __APPLE__ +#endif // _WIN32 + +#endif // MY_CPU_ARM_OR_ARM64 + + + +#ifdef __APPLE__ + +#include + +int My_sysctlbyname_Get(const char *name, void *buf, size_t *bufSize) +{ + return sysctlbyname(name, buf, bufSize, NULL, 0); +} + +int My_sysctlbyname_Get_UInt32(const char *name, UInt32 *val) +{ + size_t bufSize = sizeof(*val); + int res = My_sysctlbyname_Get(name, val, &bufSize); + if (res == 0 && bufSize != sizeof(*val)) + return EFAULT; + return res; +} + +#endif diff --git a/3rdparty/lzma/src/Delta.c b/3rdparty/lzma/src/Delta.c new file mode 100644 index 0000000000..c4a4499fe2 --- /dev/null +++ b/3rdparty/lzma/src/Delta.c @@ -0,0 +1,169 @@ +/* Delta.c -- Delta converter +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Delta.h" + +void Delta_Init(Byte *state) +{ + unsigned i; + for (i = 0; i < DELTA_STATE_SIZE; i++) + state[i] = 0; +} + + +void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size) +{ + Byte temp[DELTA_STATE_SIZE]; + + if (size == 0) + return; + + { + unsigned i = 0; + do + temp[i] = state[i]; + while (++i != delta); + } + + if (size <= delta) + { + unsigned i = 0, k; + do + { + Byte b = *data; + *data++ = (Byte)(b - temp[i]); + temp[i] = b; + } + while (++i != size); + + k = 0; + + do + { + if (i == delta) + i = 0; + state[k] = temp[i++]; + } + while (++k != delta); + + return; + } + + { + Byte *p = data + size - delta; + { + unsigned i = 0; + do + state[i] = *p++; + while (++i != delta); + } + { + const Byte *lim = data + delta; + ptrdiff_t dif = -(ptrdiff_t)delta; + + if (((ptrdiff_t)size + dif) & 1) + { + --p; *p = (Byte)(*p - p[dif]); + } + + while (p != lim) + { + --p; *p = (Byte)(*p - p[dif]); + --p; *p = (Byte)(*p - p[dif]); + } + + dif = -dif; + + do + { + --p; *p = (Byte)(*p - temp[--dif]); + } + while (dif != 0); + } + } +} + + +void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size) +{ + unsigned i; + const Byte *lim; + + if (size == 0) + return; + + i = 0; + lim = data + size; + + if (size <= delta) + { + do + *data = (Byte)(*data + state[i++]); + while (++data != lim); + + for (; delta != i; state++, delta--) + *state = state[i]; + data -= i; + } + else + { + /* + #define B(n) b ## n + #define I(n) Byte B(n) = state[n]; + #define U(n) { B(n) = (Byte)((B(n)) + *data++); data[-1] = (B(n)); } + #define F(n) if (data != lim) { U(n) } + + if (delta == 1) + { + I(0) + if ((lim - data) & 1) { U(0) } + while (data != lim) { U(0) U(0) } + data -= 1; + } + else if (delta == 2) + { + I(0) I(1) + lim -= 1; while (data < lim) { U(0) U(1) } + lim += 1; F(0) + data -= 2; + } + else if (delta == 3) + { + I(0) I(1) I(2) + lim -= 2; while (data < lim) { U(0) U(1) U(2) } + lim += 2; F(0) F(1) + data -= 3; + } + else if (delta == 4) + { + I(0) I(1) I(2) I(3) + lim -= 3; while (data < lim) { U(0) U(1) U(2) U(3) } + lim += 3; F(0) F(1) F(2) + data -= 4; + } + else + */ + { + do + { + *data = (Byte)(*data + state[i++]); + data++; + } + while (i != delta); + + { + ptrdiff_t dif = -(ptrdiff_t)delta; + do + *data = (Byte)(*data + data[dif]); + while (++data != lim); + data += dif; + } + } + } + + do + *state++ = *data; + while (++data != lim); +} diff --git a/3rdparty/lzma/src/LzFind.c b/3rdparty/lzma/src/LzFind.c new file mode 100644 index 0000000000..1b73c28484 --- /dev/null +++ b/3rdparty/lzma/src/LzFind.c @@ -0,0 +1,1628 @@ +/* LzFind.c -- Match finder for LZ algorithms +2021-11-29 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include +// #include + +#include "CpuArch.h" +#include "LzFind.h" +#include "LzHash.h" + +#define kBlockMoveAlign (1 << 7) // alignment for memmove() +#define kBlockSizeAlign (1 << 16) // alignment for block allocation +#define kBlockSizeReserveMin (1 << 24) // it's 1/256 from 4 GB dictinary + +#define kEmptyHashValue 0 + +#define kMaxValForNormalize ((UInt32)0) +// #define kMaxValForNormalize ((UInt32)(1 << 20) + 0xFFF) // for debug + +// #define kNormalizeAlign (1 << 7) // alignment for speculated accesses + +#define GET_AVAIL_BYTES(p) \ + Inline_MatchFinder_GetNumAvailableBytes(p) + + +// #define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size) +#define kFix5HashSize kFix4HashSize + +/* + HASH2_CALC: + if (hv) match, then cur[0] and cur[1] also match +*/ +#define HASH2_CALC hv = GetUi16(cur); + +// (crc[0 ... 255] & 0xFF) provides one-to-one correspondence to [0 ... 255] + +/* + HASH3_CALC: + if (cur[0]) and (h2) match, then cur[1] also match + if (cur[0]) and (hv) match, then cur[1] and cur[2] also match +*/ +#define HASH3_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + hv = (temp ^ ((UInt32)cur[2] << 8)) & p->hashMask; } + +#define HASH4_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + temp ^= ((UInt32)cur[2] << 8); \ + h3 = temp & (kHash3Size - 1); \ + hv = (temp ^ (p->crc[cur[3]] << kLzHash_CrcShift_1)) & p->hashMask; } + +#define HASH5_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + temp ^= ((UInt32)cur[2] << 8); \ + h3 = temp & (kHash3Size - 1); \ + temp ^= (p->crc[cur[3]] << kLzHash_CrcShift_1); \ + /* h4 = temp & p->hash4Mask; */ /* (kHash4Size - 1); */ \ + hv = (temp ^ (p->crc[cur[4]] << kLzHash_CrcShift_2)) & p->hashMask; } + +#define HASH_ZIP_CALC hv = ((cur[2] | ((UInt32)cur[0] << 8)) ^ p->crc[cur[1]]) & 0xFFFF; + + +static void LzInWindow_Free(CMatchFinder *p, ISzAllocPtr alloc) +{ + if (!p->directInput) + { + ISzAlloc_Free(alloc, p->bufferBase); + p->bufferBase = NULL; + } +} + + +static int LzInWindow_Create2(CMatchFinder *p, UInt32 blockSize, ISzAllocPtr alloc) +{ + if (blockSize == 0) + return 0; + if (!p->bufferBase || p->blockSize != blockSize) + { + // size_t blockSizeT; + LzInWindow_Free(p, alloc); + p->blockSize = blockSize; + // blockSizeT = blockSize; + + // printf("\nblockSize = 0x%x\n", blockSize); + /* + #if defined _WIN64 + // we can allocate 4GiB, but still use UInt32 for (p->blockSize) + // we use UInt32 type for (p->blockSize), because + // we don't want to wrap over 4 GiB, + // when we use (p->streamPos - p->pos) that is UInt32. + if (blockSize >= (UInt32)0 - (UInt32)kBlockSizeAlign) + { + blockSizeT = ((size_t)1 << 32); + printf("\nchanged to blockSizeT = 4GiB\n"); + } + #endif + */ + + p->bufferBase = (Byte *)ISzAlloc_Alloc(alloc, blockSize); + // printf("\nbufferBase = %p\n", p->bufferBase); + // return 0; // for debug + } + return (p->bufferBase != NULL); +} + +static const Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p) { return p->buffer; } + +static UInt32 MatchFinder_GetNumAvailableBytes(CMatchFinder *p) { return GET_AVAIL_BYTES(p); } + + +MY_NO_INLINE +static void MatchFinder_ReadBlock(CMatchFinder *p) +{ + if (p->streamEndWasReached || p->result != SZ_OK) + return; + + /* We use (p->streamPos - p->pos) value. + (p->streamPos < p->pos) is allowed. */ + + if (p->directInput) + { + UInt32 curSize = 0xFFFFFFFF - GET_AVAIL_BYTES(p); + if (curSize > p->directInputRem) + curSize = (UInt32)p->directInputRem; + p->directInputRem -= curSize; + p->streamPos += curSize; + if (p->directInputRem == 0) + p->streamEndWasReached = 1; + return; + } + + for (;;) + { + Byte *dest = p->buffer + GET_AVAIL_BYTES(p); + size_t size = (size_t)(p->bufferBase + p->blockSize - dest); + if (size == 0) + { + /* we call ReadBlock() after NeedMove() and MoveBlock(). + NeedMove() and MoveBlock() povide more than (keepSizeAfter) + to the end of (blockSize). + So we don't execute this branch in normal code flow. + We can go here, if we will call ReadBlock() before NeedMove(), MoveBlock(). + */ + // p->result = SZ_ERROR_FAIL; // we can show error here + return; + } + + // #define kRead 3 + // if (size > kRead) size = kRead; // for debug + + p->result = ISeqInStream_Read(p->stream, dest, &size); + if (p->result != SZ_OK) + return; + if (size == 0) + { + p->streamEndWasReached = 1; + return; + } + p->streamPos += (UInt32)size; + if (GET_AVAIL_BYTES(p) > p->keepSizeAfter) + return; + /* here and in another (p->keepSizeAfter) checks we keep on 1 byte more than was requested by Create() function + (GET_AVAIL_BYTES(p) >= p->keepSizeAfter) - minimal required size */ + } + + // on exit: (p->result != SZ_OK || p->streamEndWasReached || GET_AVAIL_BYTES(p) > p->keepSizeAfter) +} + + + +MY_NO_INLINE +void MatchFinder_MoveBlock(CMatchFinder *p) +{ + const size_t offset = (size_t)(p->buffer - p->bufferBase) - p->keepSizeBefore; + const size_t keepBefore = (offset & (kBlockMoveAlign - 1)) + p->keepSizeBefore; + p->buffer = p->bufferBase + keepBefore; + memmove(p->bufferBase, + p->bufferBase + (offset & ~((size_t)kBlockMoveAlign - 1)), + keepBefore + (size_t)GET_AVAIL_BYTES(p)); +} + +/* We call MoveBlock() before ReadBlock(). + So MoveBlock() can be wasteful operation, if the whole input data + can fit in current block even without calling MoveBlock(). + in important case where (dataSize <= historySize) + condition (p->blockSize > dataSize + p->keepSizeAfter) is met + So there is no MoveBlock() in that case case. +*/ + +int MatchFinder_NeedMove(CMatchFinder *p) +{ + if (p->directInput) + return 0; + if (p->streamEndWasReached || p->result != SZ_OK) + return 0; + return ((size_t)(p->bufferBase + p->blockSize - p->buffer) <= p->keepSizeAfter); +} + +void MatchFinder_ReadIfRequired(CMatchFinder *p) +{ + if (p->keepSizeAfter >= GET_AVAIL_BYTES(p)) + MatchFinder_ReadBlock(p); +} + + + +static void MatchFinder_SetDefaultSettings(CMatchFinder *p) +{ + p->cutValue = 32; + p->btMode = 1; + p->numHashBytes = 4; + p->bigHash = 0; +} + +#define kCrcPoly 0xEDB88320 + +void MatchFinder_Construct(CMatchFinder *p) +{ + unsigned i; + p->bufferBase = NULL; + p->directInput = 0; + p->hash = NULL; + p->expectedDataSize = (UInt64)(Int64)-1; + MatchFinder_SetDefaultSettings(p); + + for (i = 0; i < 256; i++) + { + UInt32 r = (UInt32)i; + unsigned j; + for (j = 0; j < 8; j++) + r = (r >> 1) ^ (kCrcPoly & ((UInt32)0 - (r & 1))); + p->crc[i] = r; + } +} + +static void MatchFinder_FreeThisClassMemory(CMatchFinder *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->hash); + p->hash = NULL; +} + +void MatchFinder_Free(CMatchFinder *p, ISzAllocPtr alloc) +{ + MatchFinder_FreeThisClassMemory(p, alloc); + LzInWindow_Free(p, alloc); +} + +static CLzRef* AllocRefs(size_t num, ISzAllocPtr alloc) +{ + size_t sizeInBytes = (size_t)num * sizeof(CLzRef); + if (sizeInBytes / sizeof(CLzRef) != num) + return NULL; + return (CLzRef *)ISzAlloc_Alloc(alloc, sizeInBytes); +} + +#if (kBlockSizeReserveMin < kBlockSizeAlign * 2) + #error Stop_Compiling_Bad_Reserve +#endif + + + +static UInt32 GetBlockSize(CMatchFinder *p, UInt32 historySize) +{ + UInt32 blockSize = (p->keepSizeBefore + p->keepSizeAfter); + /* + if (historySize > kMaxHistorySize) + return 0; + */ + // printf("\nhistorySize == 0x%x\n", historySize); + + if (p->keepSizeBefore < historySize || blockSize < p->keepSizeBefore) // if 32-bit overflow + return 0; + + { + const UInt32 kBlockSizeMax = (UInt32)0 - (UInt32)kBlockSizeAlign; + const UInt32 rem = kBlockSizeMax - blockSize; + const UInt32 reserve = (blockSize >> (blockSize < ((UInt32)1 << 30) ? 1 : 2)) + + (1 << 12) + kBlockMoveAlign + kBlockSizeAlign; // do not overflow 32-bit here + if (blockSize >= kBlockSizeMax + || rem < kBlockSizeReserveMin) // we reject settings that will be slow + return 0; + if (reserve >= rem) + blockSize = kBlockSizeMax; + else + { + blockSize += reserve; + blockSize &= ~(UInt32)(kBlockSizeAlign - 1); + } + } + // printf("\n LzFind_blockSize = %x\n", blockSize); + // printf("\n LzFind_blockSize = %d\n", blockSize >> 20); + return blockSize; +} + + +int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, + UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, + ISzAllocPtr alloc) +{ + /* we need one additional byte in (p->keepSizeBefore), + since we use MoveBlock() after (p->pos++) and before dictionary using */ + // keepAddBufferBefore = (UInt32)0xFFFFFFFF - (1 << 22); // for debug + p->keepSizeBefore = historySize + keepAddBufferBefore + 1; + + keepAddBufferAfter += matchMaxLen; + /* we need (p->keepSizeAfter >= p->numHashBytes) */ + if (keepAddBufferAfter < p->numHashBytes) + keepAddBufferAfter = p->numHashBytes; + // keepAddBufferAfter -= 2; // for debug + p->keepSizeAfter = keepAddBufferAfter; + + if (p->directInput) + p->blockSize = 0; + if (p->directInput || LzInWindow_Create2(p, GetBlockSize(p, historySize), alloc)) + { + const UInt32 newCyclicBufferSize = historySize + 1; // do not change it + UInt32 hs; + p->matchMaxLen = matchMaxLen; + { + // UInt32 hs4; + p->fixedHashSize = 0; + hs = (1 << 16) - 1; + if (p->numHashBytes != 2) + { + hs = historySize; + if (hs > p->expectedDataSize) + hs = (UInt32)p->expectedDataSize; + if (hs != 0) + hs--; + hs |= (hs >> 1); + hs |= (hs >> 2); + hs |= (hs >> 4); + hs |= (hs >> 8); + // we propagated 16 bits in (hs). Low 16 bits must be set later + hs >>= 1; + if (hs >= (1 << 24)) + { + if (p->numHashBytes == 3) + hs = (1 << 24) - 1; + else + hs >>= 1; + /* if (bigHash) mode, GetHeads4b() in LzFindMt.c needs (hs >= ((1 << 24) - 1))) */ + } + + // hs = ((UInt32)1 << 25) - 1; // for test + + // (hash_size >= (1 << 16)) : Required for (numHashBytes > 2) + hs |= (1 << 16) - 1; /* don't change it! */ + + // bt5: we adjust the size with recommended minimum size + if (p->numHashBytes >= 5) + hs |= (256 << kLzHash_CrcShift_2) - 1; + } + p->hashMask = hs; + hs++; + + /* + hs4 = (1 << 20); + if (hs4 > hs) + hs4 = hs; + // hs4 = (1 << 16); // for test + p->hash4Mask = hs4 - 1; + */ + + if (p->numHashBytes > 2) p->fixedHashSize += kHash2Size; + if (p->numHashBytes > 3) p->fixedHashSize += kHash3Size; + // if (p->numHashBytes > 4) p->fixedHashSize += hs4; // kHash4Size; + hs += p->fixedHashSize; + } + + { + size_t newSize; + size_t numSons; + p->historySize = historySize; + p->hashSizeSum = hs; + p->cyclicBufferSize = newCyclicBufferSize; // it must be = (historySize + 1) + + numSons = newCyclicBufferSize; + if (p->btMode) + numSons <<= 1; + newSize = hs + numSons; + + // aligned size is not required here, but it can be better for some loops + #define NUM_REFS_ALIGN_MASK 0xF + newSize = (newSize + NUM_REFS_ALIGN_MASK) & ~(size_t)NUM_REFS_ALIGN_MASK; + + if (p->hash && p->numRefs == newSize) + return 1; + + MatchFinder_FreeThisClassMemory(p, alloc); + p->numRefs = newSize; + p->hash = AllocRefs(newSize, alloc); + + if (p->hash) + { + p->son = p->hash + p->hashSizeSum; + return 1; + } + } + } + + MatchFinder_Free(p, alloc); + return 0; +} + + +static void MatchFinder_SetLimits(CMatchFinder *p) +{ + UInt32 k; + UInt32 n = kMaxValForNormalize - p->pos; + if (n == 0) + n = (UInt32)(Int32)-1; // we allow (pos == 0) at start even with (kMaxValForNormalize == 0) + + k = p->cyclicBufferSize - p->cyclicBufferPos; + if (k < n) + n = k; + + k = GET_AVAIL_BYTES(p); + { + const UInt32 ksa = p->keepSizeAfter; + UInt32 mm = p->matchMaxLen; + if (k > ksa) + k -= ksa; // we must limit exactly to keepSizeAfter for ReadBlock + else if (k >= mm) + { + // the limitation for (p->lenLimit) update + k -= mm; // optimization : to reduce the number of checks + k++; + // k = 1; // non-optimized version : for debug + } + else + { + mm = k; + if (k != 0) + k = 1; + } + p->lenLimit = mm; + } + if (k < n) + n = k; + + p->posLimit = p->pos + n; +} + + +void MatchFinder_Init_LowHash(CMatchFinder *p) +{ + size_t i; + CLzRef *items = p->hash; + const size_t numItems = p->fixedHashSize; + for (i = 0; i < numItems; i++) + items[i] = kEmptyHashValue; +} + + +void MatchFinder_Init_HighHash(CMatchFinder *p) +{ + size_t i; + CLzRef *items = p->hash + p->fixedHashSize; + const size_t numItems = (size_t)p->hashMask + 1; + for (i = 0; i < numItems; i++) + items[i] = kEmptyHashValue; +} + + +void MatchFinder_Init_4(CMatchFinder *p) +{ + p->buffer = p->bufferBase; + { + /* kEmptyHashValue = 0 (Zero) is used in hash tables as NO-VALUE marker. + the code in CMatchFinderMt expects (pos = 1) */ + p->pos = + p->streamPos = + 1; // it's smallest optimal value. do not change it + // 0; // for debug + } + p->result = SZ_OK; + p->streamEndWasReached = 0; +} + + +// (CYC_TO_POS_OFFSET == 0) is expected by some optimized code +#define CYC_TO_POS_OFFSET 0 +// #define CYC_TO_POS_OFFSET 1 // for debug + +void MatchFinder_Init(CMatchFinder *p) +{ + MatchFinder_Init_HighHash(p); + MatchFinder_Init_LowHash(p); + MatchFinder_Init_4(p); + // if (readData) + MatchFinder_ReadBlock(p); + + /* if we init (cyclicBufferPos = pos), then we can use one variable + instead of both (cyclicBufferPos) and (pos) : only before (cyclicBufferPos) wrapping */ + p->cyclicBufferPos = (p->pos - CYC_TO_POS_OFFSET); // init with relation to (pos) + // p->cyclicBufferPos = 0; // smallest value + // p->son[0] = p->son[1] = 0; // unused: we can init skipped record for speculated accesses. + MatchFinder_SetLimits(p); +} + + + +#ifdef MY_CPU_X86_OR_AMD64 + #if defined(__clang__) && (__clang_major__ >= 8) \ + || defined(__GNUC__) && (__GNUC__ >= 8) \ + || defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 1900) + #define USE_SATUR_SUB_128 + #define USE_AVX2 + #define ATTRIB_SSE41 __attribute__((__target__("sse4.1"))) + #define ATTRIB_AVX2 __attribute__((__target__("avx2"))) + #elif defined(_MSC_VER) + #if (_MSC_VER >= 1600) + #define USE_SATUR_SUB_128 + #if (_MSC_VER >= 1900) + #define USE_AVX2 + #include // avx + #endif + #endif + #endif + +// #elif defined(MY_CPU_ARM_OR_ARM64) +#elif defined(MY_CPU_ARM64) + + #if defined(__clang__) && (__clang_major__ >= 8) \ + || defined(__GNUC__) && (__GNUC__ >= 8) + #define USE_SATUR_SUB_128 + #ifdef MY_CPU_ARM64 + // #define ATTRIB_SSE41 __attribute__((__target__(""))) + #else + // #define ATTRIB_SSE41 __attribute__((__target__("fpu=crypto-neon-fp-armv8"))) + #endif + + #elif defined(_MSC_VER) + #if (_MSC_VER >= 1910) + #define USE_SATUR_SUB_128 + #endif + #endif + + #if defined(_MSC_VER) && defined(MY_CPU_ARM64) + #include + #else + #include + #endif + +#endif + +/* +#ifndef ATTRIB_SSE41 + #define ATTRIB_SSE41 +#endif +#ifndef ATTRIB_AVX2 + #define ATTRIB_AVX2 +#endif +*/ + +#ifdef USE_SATUR_SUB_128 + +// #define _SHOW_HW_STATUS + +#ifdef _SHOW_HW_STATUS +#include +#define _PRF(x) x +_PRF(;) +#else +#define _PRF(x) +#endif + +#ifdef MY_CPU_ARM_OR_ARM64 + +#ifdef MY_CPU_ARM64 +// #define FORCE_SATUR_SUB_128 +#endif + +typedef uint32x4_t v128; +#define SASUB_128(i) \ + *(v128 *)(void *)(items + (i) * 4) = \ + vsubq_u32(vmaxq_u32(*(const v128 *)(const void *)(items + (i) * 4), sub2), sub2); + +#else + +#include // sse4.1 + +typedef __m128i v128; +#define SASUB_128(i) \ + *(v128 *)(void *)(items + (i) * 4) = \ + _mm_sub_epi32(_mm_max_epu32(*(const v128 *)(const void *)(items + (i) * 4), sub2), sub2); // SSE 4.1 + +#endif + + + +MY_NO_INLINE +static +#ifdef ATTRIB_SSE41 +ATTRIB_SSE41 +#endif +void +MY_FAST_CALL +LzFind_SaturSub_128(UInt32 subValue, CLzRef *items, const CLzRef *lim) +{ + v128 sub2 = + #ifdef MY_CPU_ARM_OR_ARM64 + vdupq_n_u32(subValue); + #else + _mm_set_epi32((Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue); + #endif + do + { + SASUB_128(0) + SASUB_128(1) + SASUB_128(2) + SASUB_128(3) + items += 4 * 4; + } + while (items != lim); +} + + + +#ifdef USE_AVX2 + +#include // avx + +#define SASUB_256(i) *(__m256i *)(void *)(items + (i) * 8) = _mm256_sub_epi32(_mm256_max_epu32(*(const __m256i *)(const void *)(items + (i) * 8), sub2), sub2); // AVX2 + +MY_NO_INLINE +static +#ifdef ATTRIB_AVX2 +ATTRIB_AVX2 +#endif +void +MY_FAST_CALL +LzFind_SaturSub_256(UInt32 subValue, CLzRef *items, const CLzRef *lim) +{ + __m256i sub2 = _mm256_set_epi32( + (Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue, + (Int32)subValue, (Int32)subValue, (Int32)subValue, (Int32)subValue); + do + { + SASUB_256(0) + SASUB_256(1) + items += 2 * 8; + } + while (items != lim); +} +#endif // USE_AVX2 + +#ifndef FORCE_SATUR_SUB_128 +typedef void (MY_FAST_CALL *LZFIND_SATUR_SUB_CODE_FUNC)( + UInt32 subValue, CLzRef *items, const CLzRef *lim); +static LZFIND_SATUR_SUB_CODE_FUNC g_LzFind_SaturSub; +#endif // FORCE_SATUR_SUB_128 + +#endif // USE_SATUR_SUB_128 + + +// kEmptyHashValue must be zero +// #define SASUB_32(i) v = items[i]; m = v - subValue; if (v < subValue) m = kEmptyHashValue; items[i] = m; +#define SASUB_32(i) v = items[i]; if (v < subValue) v = subValue; items[i] = v - subValue; + +#ifdef FORCE_SATUR_SUB_128 + +#define DEFAULT_SaturSub LzFind_SaturSub_128 + +#else + +#define DEFAULT_SaturSub LzFind_SaturSub_32 + +MY_NO_INLINE +static +void +MY_FAST_CALL +LzFind_SaturSub_32(UInt32 subValue, CLzRef *items, const CLzRef *lim) +{ + do + { + UInt32 v; + SASUB_32(0) + SASUB_32(1) + SASUB_32(2) + SASUB_32(3) + SASUB_32(4) + SASUB_32(5) + SASUB_32(6) + SASUB_32(7) + items += 8; + } + while (items != lim); +} + +#endif + + +MY_NO_INLINE +void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems) +{ + #define K_NORM_ALIGN_BLOCK_SIZE (1 << 6) + + CLzRef *lim; + + for (; numItems != 0 && ((unsigned)(ptrdiff_t)items & (K_NORM_ALIGN_BLOCK_SIZE - 1)) != 0; numItems--) + { + UInt32 v; + SASUB_32(0); + items++; + } + + { + #define K_NORM_ALIGN_MASK (K_NORM_ALIGN_BLOCK_SIZE / 4 - 1) + lim = items + (numItems & ~(size_t)K_NORM_ALIGN_MASK); + numItems &= K_NORM_ALIGN_MASK; + if (items != lim) + { + #if defined(USE_SATUR_SUB_128) && !defined(FORCE_SATUR_SUB_128) + if (g_LzFind_SaturSub) + g_LzFind_SaturSub(subValue, items, lim); + else + #endif + DEFAULT_SaturSub(subValue, items, lim); + } + items = lim; + } + + + for (; numItems != 0; numItems--) + { + UInt32 v; + SASUB_32(0); + items++; + } +} + + + +// call MatchFinder_CheckLimits() only after (p->pos++) update + +MY_NO_INLINE +static void MatchFinder_CheckLimits(CMatchFinder *p) +{ + if (// !p->streamEndWasReached && p->result == SZ_OK && + p->keepSizeAfter == GET_AVAIL_BYTES(p)) + { + // we try to read only in exact state (p->keepSizeAfter == GET_AVAIL_BYTES(p)) + if (MatchFinder_NeedMove(p)) + MatchFinder_MoveBlock(p); + MatchFinder_ReadBlock(p); + } + + if (p->pos == kMaxValForNormalize) + if (GET_AVAIL_BYTES(p) >= p->numHashBytes) // optional optimization for last bytes of data. + /* + if we disable normalization for last bytes of data, and + if (data_size == 4 GiB), we don't call wastfull normalization, + but (pos) will be wrapped over Zero (0) in that case. + And we cannot resume later to normal operation + */ + { + // MatchFinder_Normalize(p); + /* after normalization we need (p->pos >= p->historySize + 1); */ + /* we can reduce subValue to aligned value, if want to keep alignment + of (p->pos) and (p->buffer) for speculated accesses. */ + const UInt32 subValue = (p->pos - p->historySize - 1) /* & ~(UInt32)(kNormalizeAlign - 1) */; + // const UInt32 subValue = (1 << 15); // for debug + // printf("\nMatchFinder_Normalize() subValue == 0x%x\n", subValue); + size_t numSonRefs = p->cyclicBufferSize; + if (p->btMode) + numSonRefs <<= 1; + Inline_MatchFinder_ReduceOffsets(p, subValue); + MatchFinder_Normalize3(subValue, p->hash, (size_t)p->hashSizeSum + numSonRefs); + } + + if (p->cyclicBufferPos == p->cyclicBufferSize) + p->cyclicBufferPos = 0; + + MatchFinder_SetLimits(p); +} + + +/* + (lenLimit > maxLen) +*/ +MY_FORCE_INLINE +static UInt32 * Hc_GetMatchesSpec(size_t lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *d, unsigned maxLen) +{ + /* + son[_cyclicBufferPos] = curMatch; + for (;;) + { + UInt32 delta = pos - curMatch; + if (cutValue-- == 0 || delta >= _cyclicBufferSize) + return d; + { + const Byte *pb = cur - delta; + curMatch = son[_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)]; + if (pb[maxLen] == cur[maxLen] && *pb == *cur) + { + UInt32 len = 0; + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + maxLen = len; + *d++ = len; + *d++ = delta - 1; + if (len == lenLimit) + return d; + } + } + } + } + */ + + const Byte *lim = cur + lenLimit; + son[_cyclicBufferPos] = curMatch; + + do + { + UInt32 delta; + + if (curMatch == 0) + break; + // if (curMatch2 >= curMatch) return NULL; + delta = pos - curMatch; + if (delta >= _cyclicBufferSize) + break; + { + ptrdiff_t diff; + curMatch = son[_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)]; + diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + if (cur[maxLen] == cur[(ptrdiff_t)maxLen + diff]) + { + const Byte *c = cur; + while (*c == c[diff]) + { + if (++c == lim) + { + d[0] = (UInt32)(lim - cur); + d[1] = delta - 1; + return d + 2; + } + } + { + const unsigned len = (unsigned)(c - cur); + if (maxLen < len) + { + maxLen = len; + d[0] = (UInt32)len; + d[1] = delta - 1; + d += 2; + } + } + } + } + } + while (--cutValue); + + return d; +} + + +MY_FORCE_INLINE +UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *d, UInt32 maxLen) +{ + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + unsigned len0 = 0, len1 = 0; + + UInt32 cmCheck; + + // if (curMatch >= pos) { *ptr0 = *ptr1 = kEmptyHashValue; return NULL; } + + cmCheck = (UInt32)(pos - _cyclicBufferSize); + if ((UInt32)pos <= _cyclicBufferSize) + cmCheck = 0; + + if (cmCheck < curMatch) + do + { + const UInt32 delta = pos - curMatch; + { + CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); + const Byte *pb = cur - delta; + unsigned len = (len0 < len1 ? len0 : len1); + const UInt32 pair0 = pair[0]; + if (pb[len] == cur[len]) + { + if (++len != lenLimit && pb[len] == cur[len]) + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + maxLen = (UInt32)len; + *d++ = (UInt32)len; + *d++ = delta - 1; + if (len == lenLimit) + { + *ptr1 = pair0; + *ptr0 = pair[1]; + return d; + } + } + } + if (pb[len] < cur[len]) + { + *ptr1 = curMatch; + // const UInt32 curMatch2 = pair[1]; + // if (curMatch2 >= curMatch) { *ptr0 = *ptr1 = kEmptyHashValue; return NULL; } + // curMatch = curMatch2; + curMatch = pair[1]; + ptr1 = pair + 1; + len1 = len; + } + else + { + *ptr0 = curMatch; + curMatch = pair[0]; + ptr0 = pair; + len0 = len; + } + } + } + while(--cutValue && cmCheck < curMatch); + + *ptr0 = *ptr1 = kEmptyHashValue; + return d; +} + + +static void SkipMatchesSpec(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue) +{ + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + unsigned len0 = 0, len1 = 0; + + UInt32 cmCheck; + + cmCheck = (UInt32)(pos - _cyclicBufferSize); + if ((UInt32)pos <= _cyclicBufferSize) + cmCheck = 0; + + if (// curMatch >= pos || // failure + cmCheck < curMatch) + do + { + const UInt32 delta = pos - curMatch; + { + CLzRef *pair = son + ((size_t)(_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); + const Byte *pb = cur - delta; + unsigned len = (len0 < len1 ? len0 : len1); + if (pb[len] == cur[len]) + { + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + { + if (len == lenLimit) + { + *ptr1 = pair[0]; + *ptr0 = pair[1]; + return; + } + } + } + if (pb[len] < cur[len]) + { + *ptr1 = curMatch; + curMatch = pair[1]; + ptr1 = pair + 1; + len1 = len; + } + else + { + *ptr0 = curMatch; + curMatch = pair[0]; + ptr0 = pair; + len0 = len; + } + } + } + while(--cutValue && cmCheck < curMatch); + + *ptr0 = *ptr1 = kEmptyHashValue; + return; +} + + +#define MOVE_POS \ + ++p->cyclicBufferPos; \ + p->buffer++; \ + { const UInt32 pos1 = p->pos + 1; p->pos = pos1; if (pos1 == p->posLimit) MatchFinder_CheckLimits(p); } + +#define MOVE_POS_RET MOVE_POS return distances; + +MY_NO_INLINE +static void MatchFinder_MovePos(CMatchFinder *p) +{ + /* we go here at the end of stream data, when (avail < num_hash_bytes) + We don't update sons[cyclicBufferPos << btMode]. + So (sons) record will contain junk. And we cannot resume match searching + to normal operation, even if we will provide more input data in buffer. + p->sons[p->cyclicBufferPos << p->btMode] = 0; // kEmptyHashValue + if (p->btMode) + p->sons[(p->cyclicBufferPos << p->btMode) + 1] = 0; // kEmptyHashValue + */ + MOVE_POS; +} + +#define GET_MATCHES_HEADER2(minLen, ret_op) \ + unsigned lenLimit; UInt32 hv; Byte *cur; UInt32 curMatch; \ + lenLimit = (unsigned)p->lenLimit; { if (lenLimit < minLen) { MatchFinder_MovePos(p); ret_op; }} \ + cur = p->buffer; + +#define GET_MATCHES_HEADER(minLen) GET_MATCHES_HEADER2(minLen, return distances) +#define SKIP_HEADER(minLen) do { GET_MATCHES_HEADER2(minLen, continue) + +#define MF_PARAMS(p) lenLimit, curMatch, p->pos, p->buffer, p->son, p->cyclicBufferPos, p->cyclicBufferSize, p->cutValue + +#define SKIP_FOOTER SkipMatchesSpec(MF_PARAMS(p)); MOVE_POS; } while (--num); + +#define GET_MATCHES_FOOTER_BASE(_maxLen_, func) \ + distances = func(MF_PARAMS(p), \ + distances, (UInt32)_maxLen_); MOVE_POS_RET; + +#define GET_MATCHES_FOOTER_BT(_maxLen_) \ + GET_MATCHES_FOOTER_BASE(_maxLen_, GetMatchesSpec1) + +#define GET_MATCHES_FOOTER_HC(_maxLen_) \ + GET_MATCHES_FOOTER_BASE(_maxLen_, Hc_GetMatchesSpec) + + + +#define UPDATE_maxLen { \ + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)d2; \ + const Byte *c = cur + maxLen; \ + const Byte *lim = cur + lenLimit; \ + for (; c != lim; c++) if (*(c + diff) != *c) break; \ + maxLen = (unsigned)(c - cur); } + +static UInt32* Bt2_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + GET_MATCHES_HEADER(2) + HASH2_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + GET_MATCHES_FOOTER_BT(1) +} + +UInt32* Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + GET_MATCHES_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + GET_MATCHES_FOOTER_BT(2) +} + + +#define SET_mmm \ + mmm = p->cyclicBufferSize; \ + if (pos < mmm) \ + mmm = pos; + + +static UInt32* Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, d2, pos; + unsigned maxLen; + UInt32 *hash; + GET_MATCHES_HEADER(3) + + HASH3_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash[h2]; + + curMatch = (hash + kFix3HashSize)[hv]; + + hash[h2] = pos; + (hash + kFix3HashSize)[hv] = pos; + + SET_mmm + + maxLen = 2; + + if (d2 < mmm && *(cur - d2) == *cur) + { + UPDATE_maxLen + distances[0] = (UInt32)maxLen; + distances[1] = d2 - 1; + distances += 2; + if (maxLen == lenLimit) + { + SkipMatchesSpec(MF_PARAMS(p)); + MOVE_POS_RET; + } + } + + GET_MATCHES_FOOTER_BT(maxLen) +} + + +static UInt32* Bt4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, h3, d2, d3, pos; + unsigned maxLen; + UInt32 *hash; + GET_MATCHES_HEADER(4) + + HASH4_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + curMatch = (hash + kFix4HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + (hash + kFix4HashSize)[hv] = pos; + + SET_mmm + + maxLen = 3; + + for (;;) + { + if (d2 < mmm && *(cur - d2) == *cur) + { + distances[0] = 2; + distances[1] = d2 - 1; + distances += 2; + if (*(cur - d2 + 2) == cur[2]) + { + // distances[-2] = 3; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + d2 = d3; + distances[1] = d3 - 1; + distances += 2; + } + else + break; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + d2 = d3; + distances[1] = d3 - 1; + distances += 2; + } + else + break; + + UPDATE_maxLen + distances[-2] = (UInt32)maxLen; + if (maxLen == lenLimit) + { + SkipMatchesSpec(MF_PARAMS(p)); + MOVE_POS_RET + } + break; + } + + GET_MATCHES_FOOTER_BT(maxLen) +} + + +static UInt32* Bt5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, h3, d2, d3, maxLen, pos; + UInt32 *hash; + GET_MATCHES_HEADER(5) + + HASH5_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + // d4 = pos - (hash + kFix4HashSize)[h4]; + + curMatch = (hash + kFix5HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + // (hash + kFix4HashSize)[h4] = pos; + (hash + kFix5HashSize)[hv] = pos; + + SET_mmm + + maxLen = 4; + + for (;;) + { + if (d2 < mmm && *(cur - d2) == *cur) + { + distances[0] = 2; + distances[1] = d2 - 1; + distances += 2; + if (*(cur - d2 + 2) == cur[2]) + { + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + distances[1] = d3 - 1; + distances += 2; + d2 = d3; + } + else + break; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + distances[1] = d3 - 1; + distances += 2; + d2 = d3; + } + else + break; + + distances[-2] = 3; + if (*(cur - d2 + 3) != cur[3]) + break; + UPDATE_maxLen + distances[-2] = (UInt32)maxLen; + if (maxLen == lenLimit) + { + SkipMatchesSpec(MF_PARAMS(p)); + MOVE_POS_RET; + } + break; + } + + GET_MATCHES_FOOTER_BT(maxLen) +} + + +static UInt32* Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, h3, d2, d3, pos; + unsigned maxLen; + UInt32 *hash; + GET_MATCHES_HEADER(4) + + HASH4_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + curMatch = (hash + kFix4HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + (hash + kFix4HashSize)[hv] = pos; + + SET_mmm + + maxLen = 3; + + for (;;) + { + if (d2 < mmm && *(cur - d2) == *cur) + { + distances[0] = 2; + distances[1] = d2 - 1; + distances += 2; + if (*(cur - d2 + 2) == cur[2]) + { + // distances[-2] = 3; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + d2 = d3; + distances[1] = d3 - 1; + distances += 2; + } + else + break; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + d2 = d3; + distances[1] = d3 - 1; + distances += 2; + } + else + break; + + UPDATE_maxLen + distances[-2] = (UInt32)maxLen; + if (maxLen == lenLimit) + { + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS_RET; + } + break; + } + + GET_MATCHES_FOOTER_HC(maxLen); +} + + +static UInt32 * Hc5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 mmm; + UInt32 h2, h3, d2, d3, maxLen, pos; + UInt32 *hash; + GET_MATCHES_HEADER(5) + + HASH5_CALC; + + hash = p->hash; + pos = p->pos; + + d2 = pos - hash [h2]; + d3 = pos - (hash + kFix3HashSize)[h3]; + // d4 = pos - (hash + kFix4HashSize)[h4]; + + curMatch = (hash + kFix5HashSize)[hv]; + + hash [h2] = pos; + (hash + kFix3HashSize)[h3] = pos; + // (hash + kFix4HashSize)[h4] = pos; + (hash + kFix5HashSize)[hv] = pos; + + SET_mmm + + maxLen = 4; + + for (;;) + { + if (d2 < mmm && *(cur - d2) == *cur) + { + distances[0] = 2; + distances[1] = d2 - 1; + distances += 2; + if (*(cur - d2 + 2) == cur[2]) + { + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + distances[1] = d3 - 1; + distances += 2; + d2 = d3; + } + else + break; + } + else if (d3 < mmm && *(cur - d3) == *cur) + { + distances[1] = d3 - 1; + distances += 2; + d2 = d3; + } + else + break; + + distances[-2] = 3; + if (*(cur - d2 + 3) != cur[3]) + break; + UPDATE_maxLen + distances[-2] = maxLen; + if (maxLen == lenLimit) + { + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS_RET; + } + break; + } + + GET_MATCHES_FOOTER_HC(maxLen); +} + + +UInt32* Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + GET_MATCHES_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + GET_MATCHES_FOOTER_HC(2) +} + + +static void Bt2_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(2) + { + HASH2_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + } + SKIP_FOOTER +} + +void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(3) + { + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + } + SKIP_FOOTER +} + +static void Bt3_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(3) + { + UInt32 h2; + UInt32 *hash; + HASH3_CALC; + hash = p->hash; + curMatch = (hash + kFix3HashSize)[hv]; + hash[h2] = + (hash + kFix3HashSize)[hv] = p->pos; + } + SKIP_FOOTER +} + +static void Bt4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(4) + { + UInt32 h2, h3; + UInt32 *hash; + HASH4_CALC; + hash = p->hash; + curMatch = (hash + kFix4HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + (hash + kFix4HashSize)[hv] = p->pos; + } + SKIP_FOOTER +} + +static void Bt5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + SKIP_HEADER(5) + { + UInt32 h2, h3; + UInt32 *hash; + HASH5_CALC; + hash = p->hash; + curMatch = (hash + kFix5HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + // (hash + kFix4HashSize)[h4] = + (hash + kFix5HashSize)[hv] = p->pos; + } + SKIP_FOOTER +} + + +#define HC_SKIP_HEADER(minLen) \ + do { if (p->lenLimit < minLen) { MatchFinder_MovePos(p); num--; continue; } { \ + Byte *cur; \ + UInt32 *hash; \ + UInt32 *son; \ + UInt32 pos = p->pos; \ + UInt32 num2 = num; \ + /* (p->pos == p->posLimit) is not allowed here !!! */ \ + { const UInt32 rem = p->posLimit - pos; if (num2 > rem) num2 = rem; } \ + num -= num2; \ + { const UInt32 cycPos = p->cyclicBufferPos; \ + son = p->son + cycPos; \ + p->cyclicBufferPos = cycPos + num2; } \ + cur = p->buffer; \ + hash = p->hash; \ + do { \ + UInt32 curMatch; \ + UInt32 hv; + + +#define HC_SKIP_FOOTER \ + cur++; pos++; *son++ = curMatch; \ + } while (--num2); \ + p->buffer = cur; \ + p->pos = pos; \ + if (pos == p->posLimit) MatchFinder_CheckLimits(p); \ + }} while(num); \ + + +static void Hc4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + HC_SKIP_HEADER(4) + + UInt32 h2, h3; + HASH4_CALC; + curMatch = (hash + kFix4HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + (hash + kFix4HashSize)[hv] = pos; + + HC_SKIP_FOOTER +} + + +static void Hc5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + HC_SKIP_HEADER(5) + + UInt32 h2, h3; + HASH5_CALC + curMatch = (hash + kFix5HashSize)[hv]; + hash [h2] = + (hash + kFix3HashSize)[h3] = + // (hash + kFix4HashSize)[h4] = + (hash + kFix5HashSize)[hv] = pos; + + HC_SKIP_FOOTER +} + + +void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + HC_SKIP_HEADER(3) + + HASH_ZIP_CALC; + curMatch = hash[hv]; + hash[hv] = pos; + + HC_SKIP_FOOTER +} + + +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder2 *vTable) +{ + vTable->Init = (Mf_Init_Func)MatchFinder_Init; + vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinder_GetNumAvailableBytes; + vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinder_GetPointerToCurrentPos; + if (!p->btMode) + { + if (p->numHashBytes <= 4) + { + vTable->GetMatches = (Mf_GetMatches_Func)Hc4_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Hc4_MatchFinder_Skip; + } + else + { + vTable->GetMatches = (Mf_GetMatches_Func)Hc5_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Hc5_MatchFinder_Skip; + } + } + else if (p->numHashBytes == 2) + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt2_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt2_MatchFinder_Skip; + } + else if (p->numHashBytes == 3) + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt3_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt3_MatchFinder_Skip; + } + else if (p->numHashBytes == 4) + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt4_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt4_MatchFinder_Skip; + } + else + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt5_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt5_MatchFinder_Skip; + } +} + + + +void LzFindPrepare() +{ + #ifndef FORCE_SATUR_SUB_128 + #ifdef USE_SATUR_SUB_128 + LZFIND_SATUR_SUB_CODE_FUNC f = NULL; + #ifdef MY_CPU_ARM_OR_ARM64 + { + if (CPU_IsSupported_NEON()) + { + // #pragma message ("=== LzFind NEON") + _PRF(printf("\n=== LzFind NEON\n")); + f = LzFind_SaturSub_128; + } + // f = 0; // for debug + } + #else // MY_CPU_ARM_OR_ARM64 + if (CPU_IsSupported_SSE41()) + { + // #pragma message ("=== LzFind SSE41") + _PRF(printf("\n=== LzFind SSE41\n")); + f = LzFind_SaturSub_128; + + #ifdef USE_AVX2 + if (CPU_IsSupported_AVX2()) + { + // #pragma message ("=== LzFind AVX2") + _PRF(printf("\n=== LzFind AVX2\n")); + f = LzFind_SaturSub_256; + } + #endif + } + #endif // MY_CPU_ARM_OR_ARM64 + g_LzFind_SaturSub = f; + #endif // USE_SATUR_SUB_128 + #endif // FORCE_SATUR_SUB_128 +} diff --git a/3rdparty/lzma/src/LzFindOpt.c b/3rdparty/lzma/src/LzFindOpt.c new file mode 100644 index 0000000000..8ff006e074 --- /dev/null +++ b/3rdparty/lzma/src/LzFindOpt.c @@ -0,0 +1,578 @@ +/* LzFindOpt.c -- multithreaded Match finder for LZ algorithms +2021-07-13 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" +#include "LzFind.h" + +// #include "LzFindMt.h" + +// #define LOG_ITERS + +// #define LOG_THREAD + +#ifdef LOG_THREAD +#include +#define PRF(x) x +#else +// #define PRF(x) +#endif + +#ifdef LOG_ITERS +#include +UInt64 g_NumIters_Tree; +UInt64 g_NumIters_Loop; +UInt64 g_NumIters_Bytes; +#define LOG_ITER(x) x +#else +#define LOG_ITER(x) +#endif + +// ---------- BT THREAD ---------- + +#define USE_SON_PREFETCH +#define USE_LONG_MATCH_OPT + +#define kEmptyHashValue 0 + +// #define CYC_TO_POS_OFFSET 0 + +// #define CYC_TO_POS_OFFSET 1 // for debug + +/* +MY_NO_INLINE +UInt32 * MY_FAST_CALL GetMatchesSpecN_1(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, UInt32 *posRes) +{ + do + { + UInt32 delta; + if (hash == size) + break; + delta = *hash++; + + if (delta == 0 || delta > (UInt32)pos) + return NULL; + + lenLimit++; + + if (delta == (UInt32)pos) + { + CLzRef *ptr1 = son + ((size_t)pos << 1) - CYC_TO_POS_OFFSET * 2; + *d++ = 0; + ptr1[0] = kEmptyHashValue; + ptr1[1] = kEmptyHashValue; + } +else +{ + UInt32 *_distances = ++d; + + CLzRef *ptr0 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2 + 1; + CLzRef *ptr1 = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; + + const Byte *len0 = cur, *len1 = cur; + UInt32 cutValue = _cutValue; + const Byte *maxLen = cur + _maxLen; + + for (LOG_ITER(g_NumIters_Tree++);;) + { + LOG_ITER(g_NumIters_Loop++); + { + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + CLzRef *pair = son + ((size_t)(((ptrdiff_t)pos - CYC_TO_POS_OFFSET) + diff) << 1); + const Byte *len = (len0 < len1 ? len0 : len1); + + #ifdef USE_SON_PREFETCH + const UInt32 pair0 = *pair; + #endif + + if (len[diff] == len[0]) + { + if (++len != lenLimit && len[diff] == len[0]) + while (++len != lenLimit) + { + LOG_ITER(g_NumIters_Bytes++); + if (len[diff] != len[0]) + break; + } + if (maxLen < len) + { + maxLen = len; + *d++ = (UInt32)(len - cur); + *d++ = delta - 1; + + if (len == lenLimit) + { + const UInt32 pair1 = pair[1]; + *ptr1 = + #ifdef USE_SON_PREFETCH + pair0; + #else + pair[0]; + #endif + *ptr0 = pair1; + + _distances[-1] = (UInt32)(d - _distances); + + #ifdef USE_LONG_MATCH_OPT + + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + + { + for (;;) + { + hash++; + pos++; + cur++; + lenLimit++; + { + CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; + #if 0 + *(UInt64 *)(void *)ptr = ((const UInt64 *)(const void *)ptr)[diff]; + #else + const UInt32 p0 = ptr[0 + (diff * 2)]; + const UInt32 p1 = ptr[1 + (diff * 2)]; + ptr[0] = p0; + ptr[1] = p1; + // ptr[0] = ptr[0 + (diff * 2)]; + // ptr[1] = ptr[1 + (diff * 2)]; + #endif + } + // PrintSon(son + 2, pos - 1); + // printf("\npos = %x delta = %x\n", pos, delta); + len++; + *d++ = 2; + *d++ = (UInt32)(len - cur); + *d++ = delta - 1; + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + } + } + #endif + + break; + } + } + } + + { + const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff); + if (len[diff] < len[0]) + { + delta = pair[1]; + if (delta >= curMatch) + return NULL; + *ptr1 = curMatch; + ptr1 = pair + 1; + len1 = len; + } + else + { + delta = *pair; + if (delta >= curMatch) + return NULL; + *ptr0 = curMatch; + ptr0 = pair; + len0 = len; + } + + delta = (UInt32)pos - delta; + + if (--cutValue == 0 || delta >= pos) + { + *ptr0 = *ptr1 = kEmptyHashValue; + _distances[-1] = (UInt32)(d - _distances); + break; + } + } + } + } // for (tree iterations) +} + pos++; + cur++; + } + while (d < limit); + *posRes = (UInt32)pos; + return d; +} +*/ + +/* define cbs if you use 2 functions. + GetMatchesSpecN_1() : (pos < _cyclicBufferSize) + GetMatchesSpecN_2() : (pos >= _cyclicBufferSize) + + do not define cbs if you use 1 function: + GetMatchesSpecN_2() +*/ + +// #define cbs _cyclicBufferSize + +/* + we use size_t for (pos) and (_cyclicBufferPos_ instead of UInt32 + to eliminate "movsx" BUG in old MSVC x64 compiler. +*/ + +UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes); + +MY_NO_INLINE +UInt32 * MY_FAST_CALL GetMatchesSpecN_2(const Byte *lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, size_t _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes) +{ + do // while (hash != size) + { + UInt32 delta; + + #ifndef cbs + UInt32 cbs; + #endif + + if (hash == size) + break; + + delta = *hash++; + + if (delta == 0) + return NULL; + + lenLimit++; + + #ifndef cbs + cbs = _cyclicBufferSize; + if ((UInt32)pos < cbs) + { + if (delta > (UInt32)pos) + return NULL; + cbs = (UInt32)pos; + } + #endif + + if (delta >= cbs) + { + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + *d++ = 0; + ptr1[0] = kEmptyHashValue; + ptr1[1] = kEmptyHashValue; + } +else +{ + UInt32 *_distances = ++d; + + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + + UInt32 cutValue = _cutValue; + const Byte *len0 = cur, *len1 = cur; + const Byte *maxLen = cur + _maxLen; + + // if (cutValue == 0) { *ptr0 = *ptr1 = kEmptyHashValue; } else + for (LOG_ITER(g_NumIters_Tree++);;) + { + LOG_ITER(g_NumIters_Loop++); + { + // SPEC code + CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - (ptrdiff_t)delta + + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0) + ) << 1); + + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + const Byte *len = (len0 < len1 ? len0 : len1); + + #ifdef USE_SON_PREFETCH + const UInt32 pair0 = *pair; + #endif + + if (len[diff] == len[0]) + { + if (++len != lenLimit && len[diff] == len[0]) + while (++len != lenLimit) + { + LOG_ITER(g_NumIters_Bytes++); + if (len[diff] != len[0]) + break; + } + if (maxLen < len) + { + maxLen = len; + *d++ = (UInt32)(len - cur); + *d++ = delta - 1; + + if (len == lenLimit) + { + const UInt32 pair1 = pair[1]; + *ptr1 = + #ifdef USE_SON_PREFETCH + pair0; + #else + pair[0]; + #endif + *ptr0 = pair1; + + _distances[-1] = (UInt32)(d - _distances); + + #ifdef USE_LONG_MATCH_OPT + + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + + { + for (;;) + { + *d++ = 2; + *d++ = (UInt32)(lenLimit - cur); + *d++ = delta - 1; + cur++; + lenLimit++; + // SPEC + _cyclicBufferPos++; + { + // SPEC code + CLzRef *dest = son + ((size_t)(_cyclicBufferPos) << 1); + const CLzRef *src = dest + ((diff + + (ptrdiff_t)(UInt32)((_cyclicBufferPos < delta) ? cbs : 0)) << 1); + // CLzRef *ptr = son + ((size_t)(pos) << 1) - CYC_TO_POS_OFFSET * 2; + #if 0 + *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src); + #else + const UInt32 p0 = src[0]; + const UInt32 p1 = src[1]; + dest[0] = p0; + dest[1] = p1; + #endif + } + pos++; + hash++; + if (hash == size || *hash != delta || lenLimit[diff] != lenLimit[0] || d >= limit) + break; + } // for() end for long matches + } + #endif + + break; // break from TREE iterations + } + } + } + { + const UInt32 curMatch = (UInt32)pos - delta; // (UInt32)(pos + diff); + if (len[diff] < len[0]) + { + delta = pair[1]; + *ptr1 = curMatch; + ptr1 = pair + 1; + len1 = len; + if (delta >= curMatch) + return NULL; + } + else + { + delta = *pair; + *ptr0 = curMatch; + ptr0 = pair; + len0 = len; + if (delta >= curMatch) + return NULL; + } + delta = (UInt32)pos - delta; + + if (--cutValue == 0 || delta >= cbs) + { + *ptr0 = *ptr1 = kEmptyHashValue; + _distances[-1] = (UInt32)(d - _distances); + break; + } + } + } + } // for (tree iterations) +} + pos++; + _cyclicBufferPos++; + cur++; + } + while (d < limit); + *posRes = (UInt32)pos; + return d; +} + + + +/* +typedef UInt32 uint32plus; // size_t + +UInt32 * MY_FAST_CALL GetMatchesSpecN_3(uint32plus lenLimit, size_t pos, const Byte *cur, CLzRef *son, + UInt32 _cutValue, UInt32 *d, uint32plus _maxLen, const UInt32 *hash, const UInt32 *limit, const UInt32 *size, + size_t _cyclicBufferPos, UInt32 _cyclicBufferSize, + UInt32 *posRes) +{ + do // while (hash != size) + { + UInt32 delta; + + #ifndef cbs + UInt32 cbs; + #endif + + if (hash == size) + break; + + delta = *hash++; + + if (delta == 0) + return NULL; + + #ifndef cbs + cbs = _cyclicBufferSize; + if ((UInt32)pos < cbs) + { + if (delta > (UInt32)pos) + return NULL; + cbs = (UInt32)pos; + } + #endif + + if (delta >= cbs) + { + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + *d++ = 0; + ptr1[0] = kEmptyHashValue; + ptr1[1] = kEmptyHashValue; + } +else +{ + CLzRef *ptr0 = son + ((size_t)_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + ((size_t)_cyclicBufferPos << 1); + UInt32 *_distances = ++d; + uint32plus len0 = 0, len1 = 0; + UInt32 cutValue = _cutValue; + uint32plus maxLen = _maxLen; + // lenLimit++; // const Byte *lenLimit = cur + _lenLimit; + + for (LOG_ITER(g_NumIters_Tree++);;) + { + LOG_ITER(g_NumIters_Loop++); + { + // const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + CLzRef *pair = son + ((size_t)((ptrdiff_t)_cyclicBufferPos - delta + + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0) + ) << 1); + const Byte *pb = cur - delta; + uint32plus len = (len0 < len1 ? len0 : len1); + + #ifdef USE_SON_PREFETCH + const UInt32 pair0 = *pair; + #endif + + if (pb[len] == cur[len]) + { + if (++len != lenLimit && pb[len] == cur[len]) + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + maxLen = len; + *d++ = (UInt32)len; + *d++ = delta - 1; + if (len == lenLimit) + { + { + const UInt32 pair1 = pair[1]; + *ptr0 = pair1; + *ptr1 = + #ifdef USE_SON_PREFETCH + pair0; + #else + pair[0]; + #endif + } + + _distances[-1] = (UInt32)(d - _distances); + + #ifdef USE_LONG_MATCH_OPT + + if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit) + break; + + { + const ptrdiff_t diff = (ptrdiff_t)0 - (ptrdiff_t)delta; + for (;;) + { + *d++ = 2; + *d++ = (UInt32)lenLimit; + *d++ = delta - 1; + _cyclicBufferPos++; + { + CLzRef *dest = son + ((size_t)_cyclicBufferPos << 1); + const CLzRef *src = dest + ((diff + + (ptrdiff_t)(UInt32)(_cyclicBufferPos < delta ? cbs : 0)) << 1); + #if 0 + *(UInt64 *)(void *)dest = *((const UInt64 *)(const void *)src); + #else + const UInt32 p0 = src[0]; + const UInt32 p1 = src[1]; + dest[0] = p0; + dest[1] = p1; + #endif + } + hash++; + pos++; + cur++; + pb++; + if (hash == size || *hash != delta || pb[lenLimit] != cur[lenLimit] || d >= limit) + break; + } + } + #endif + + break; + } + } + } + { + const UInt32 curMatch = (UInt32)pos - delta; + if (pb[len] < cur[len]) + { + delta = pair[1]; + *ptr1 = curMatch; + ptr1 = pair + 1; + len1 = len; + } + else + { + delta = *pair; + *ptr0 = curMatch; + ptr0 = pair; + len0 = len; + } + + { + if (delta >= curMatch) + return NULL; + delta = (UInt32)pos - delta; + if (delta >= cbs + // delta >= _cyclicBufferSize || delta >= pos + || --cutValue == 0) + { + *ptr0 = *ptr1 = kEmptyHashValue; + _distances[-1] = (UInt32)(d - _distances); + break; + } + } + } + } + } // for (tree iterations) +} + pos++; + _cyclicBufferPos++; + cur++; + } + while (d < limit); + *posRes = (UInt32)pos; + return d; +} +*/ diff --git a/3rdparty/lzma/src/Lzma2Dec.c b/3rdparty/lzma/src/Lzma2Dec.c new file mode 100644 index 0000000000..ac970a843d --- /dev/null +++ b/3rdparty/lzma/src/Lzma2Dec.c @@ -0,0 +1,489 @@ +/* Lzma2Dec.c -- LZMA2 Decoder +2021-02-09 : Igor Pavlov : Public domain */ + +/* #define SHOW_DEBUG_INFO */ + +#include "Precomp.h" + +#ifdef SHOW_DEBUG_INFO +#include +#endif + +#include + +#include "Lzma2Dec.h" + +/* +00000000 - End of data +00000001 U U - Uncompressed, reset dic, need reset state and set new prop +00000010 U U - Uncompressed, no reset +100uuuuu U U P P - LZMA, no reset +101uuuuu U U P P - LZMA, reset state +110uuuuu U U P P S - LZMA, reset state + set new prop +111uuuuu U U P P S - LZMA, reset state + set new prop, reset dic + + u, U - Unpack Size + P - Pack Size + S - Props +*/ + +#define LZMA2_CONTROL_COPY_RESET_DIC 1 + +#define LZMA2_IS_UNCOMPRESSED_STATE(p) (((p)->control & (1 << 7)) == 0) + +#define LZMA2_LCLP_MAX 4 +#define LZMA2_DIC_SIZE_FROM_PROP(p) (((UInt32)2 | ((p) & 1)) << ((p) / 2 + 11)) + +#ifdef SHOW_DEBUG_INFO +#define PRF(x) x +#else +#define PRF(x) +#endif + +typedef enum +{ + LZMA2_STATE_CONTROL, + LZMA2_STATE_UNPACK0, + LZMA2_STATE_UNPACK1, + LZMA2_STATE_PACK0, + LZMA2_STATE_PACK1, + LZMA2_STATE_PROP, + LZMA2_STATE_DATA, + LZMA2_STATE_DATA_CONT, + LZMA2_STATE_FINISHED, + LZMA2_STATE_ERROR +} ELzma2State; + +static SRes Lzma2Dec_GetOldProps(Byte prop, Byte *props) +{ + UInt32 dicSize; + if (prop > 40) + return SZ_ERROR_UNSUPPORTED; + dicSize = (prop == 40) ? 0xFFFFFFFF : LZMA2_DIC_SIZE_FROM_PROP(prop); + props[0] = (Byte)LZMA2_LCLP_MAX; + props[1] = (Byte)(dicSize); + props[2] = (Byte)(dicSize >> 8); + props[3] = (Byte)(dicSize >> 16); + props[4] = (Byte)(dicSize >> 24); + return SZ_OK; +} + +SRes Lzma2Dec_AllocateProbs(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc) +{ + Byte props[LZMA_PROPS_SIZE]; + RINOK(Lzma2Dec_GetOldProps(prop, props)); + return LzmaDec_AllocateProbs(&p->decoder, props, LZMA_PROPS_SIZE, alloc); +} + +SRes Lzma2Dec_Allocate(CLzma2Dec *p, Byte prop, ISzAllocPtr alloc) +{ + Byte props[LZMA_PROPS_SIZE]; + RINOK(Lzma2Dec_GetOldProps(prop, props)); + return LzmaDec_Allocate(&p->decoder, props, LZMA_PROPS_SIZE, alloc); +} + +void Lzma2Dec_Init(CLzma2Dec *p) +{ + p->state = LZMA2_STATE_CONTROL; + p->needInitLevel = 0xE0; + p->isExtraMode = False; + p->unpackSize = 0; + + // p->decoder.dicPos = 0; // we can use it instead of full init + LzmaDec_Init(&p->decoder); +} + +// ELzma2State +static unsigned Lzma2Dec_UpdateState(CLzma2Dec *p, Byte b) +{ + switch (p->state) + { + case LZMA2_STATE_CONTROL: + p->isExtraMode = False; + p->control = b; + PRF(printf("\n %8X", (unsigned)p->decoder.dicPos)); + PRF(printf(" %02X", (unsigned)b)); + if (b == 0) + return LZMA2_STATE_FINISHED; + if (LZMA2_IS_UNCOMPRESSED_STATE(p)) + { + if (b == LZMA2_CONTROL_COPY_RESET_DIC) + p->needInitLevel = 0xC0; + else if (b > 2 || p->needInitLevel == 0xE0) + return LZMA2_STATE_ERROR; + } + else + { + if (b < p->needInitLevel) + return LZMA2_STATE_ERROR; + p->needInitLevel = 0; + p->unpackSize = (UInt32)(b & 0x1F) << 16; + } + return LZMA2_STATE_UNPACK0; + + case LZMA2_STATE_UNPACK0: + p->unpackSize |= (UInt32)b << 8; + return LZMA2_STATE_UNPACK1; + + case LZMA2_STATE_UNPACK1: + p->unpackSize |= (UInt32)b; + p->unpackSize++; + PRF(printf(" %7u", (unsigned)p->unpackSize)); + return LZMA2_IS_UNCOMPRESSED_STATE(p) ? LZMA2_STATE_DATA : LZMA2_STATE_PACK0; + + case LZMA2_STATE_PACK0: + p->packSize = (UInt32)b << 8; + return LZMA2_STATE_PACK1; + + case LZMA2_STATE_PACK1: + p->packSize |= (UInt32)b; + p->packSize++; + // if (p->packSize < 5) return LZMA2_STATE_ERROR; + PRF(printf(" %5u", (unsigned)p->packSize)); + return (p->control & 0x40) ? LZMA2_STATE_PROP : LZMA2_STATE_DATA; + + case LZMA2_STATE_PROP: + { + unsigned lc, lp; + if (b >= (9 * 5 * 5)) + return LZMA2_STATE_ERROR; + lc = b % 9; + b /= 9; + p->decoder.prop.pb = (Byte)(b / 5); + lp = b % 5; + if (lc + lp > LZMA2_LCLP_MAX) + return LZMA2_STATE_ERROR; + p->decoder.prop.lc = (Byte)lc; + p->decoder.prop.lp = (Byte)lp; + return LZMA2_STATE_DATA; + } + } + return LZMA2_STATE_ERROR; +} + +static void LzmaDec_UpdateWithUncompressed(CLzmaDec *p, const Byte *src, SizeT size) +{ + memcpy(p->dic + p->dicPos, src, size); + p->dicPos += size; + if (p->checkDicSize == 0 && p->prop.dicSize - p->processedPos <= size) + p->checkDicSize = p->prop.dicSize; + p->processedPos += (UInt32)size; +} + +void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState); + + +SRes Lzma2Dec_DecodeToDic(CLzma2Dec *p, SizeT dicLimit, + const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT inSize = *srcLen; + *srcLen = 0; + *status = LZMA_STATUS_NOT_SPECIFIED; + + while (p->state != LZMA2_STATE_ERROR) + { + SizeT dicPos; + + if (p->state == LZMA2_STATE_FINISHED) + { + *status = LZMA_STATUS_FINISHED_WITH_MARK; + return SZ_OK; + } + + dicPos = p->decoder.dicPos; + + if (dicPos == dicLimit && finishMode == LZMA_FINISH_ANY) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_OK; + } + + if (p->state != LZMA2_STATE_DATA && p->state != LZMA2_STATE_DATA_CONT) + { + if (*srcLen == inSize) + { + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + (*srcLen)++; + p->state = Lzma2Dec_UpdateState(p, *src++); + if (dicPos == dicLimit && p->state != LZMA2_STATE_FINISHED) + break; + continue; + } + + { + SizeT inCur = inSize - *srcLen; + SizeT outCur = dicLimit - dicPos; + ELzmaFinishMode curFinishMode = LZMA_FINISH_ANY; + + if (outCur >= p->unpackSize) + { + outCur = (SizeT)p->unpackSize; + curFinishMode = LZMA_FINISH_END; + } + + if (LZMA2_IS_UNCOMPRESSED_STATE(p)) + { + if (inCur == 0) + { + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + + if (p->state == LZMA2_STATE_DATA) + { + BoolInt initDic = (p->control == LZMA2_CONTROL_COPY_RESET_DIC); + LzmaDec_InitDicAndState(&p->decoder, initDic, False); + } + + if (inCur > outCur) + inCur = outCur; + if (inCur == 0) + break; + + LzmaDec_UpdateWithUncompressed(&p->decoder, src, inCur); + + src += inCur; + *srcLen += inCur; + p->unpackSize -= (UInt32)inCur; + p->state = (p->unpackSize == 0) ? LZMA2_STATE_CONTROL : LZMA2_STATE_DATA_CONT; + } + else + { + SRes res; + + if (p->state == LZMA2_STATE_DATA) + { + BoolInt initDic = (p->control >= 0xE0); + BoolInt initState = (p->control >= 0xA0); + LzmaDec_InitDicAndState(&p->decoder, initDic, initState); + p->state = LZMA2_STATE_DATA_CONT; + } + + if (inCur > p->packSize) + inCur = (SizeT)p->packSize; + + res = LzmaDec_DecodeToDic(&p->decoder, dicPos + outCur, src, &inCur, curFinishMode, status); + + src += inCur; + *srcLen += inCur; + p->packSize -= (UInt32)inCur; + outCur = p->decoder.dicPos - dicPos; + p->unpackSize -= (UInt32)outCur; + + if (res != 0) + break; + + if (*status == LZMA_STATUS_NEEDS_MORE_INPUT) + { + if (p->packSize == 0) + break; + return SZ_OK; + } + + if (inCur == 0 && outCur == 0) + { + if (*status != LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + || p->unpackSize != 0 + || p->packSize != 0) + break; + p->state = LZMA2_STATE_CONTROL; + } + + *status = LZMA_STATUS_NOT_SPECIFIED; + } + } + } + + *status = LZMA_STATUS_NOT_SPECIFIED; + p->state = LZMA2_STATE_ERROR; + return SZ_ERROR_DATA; +} + + + + +ELzma2ParseStatus Lzma2Dec_Parse(CLzma2Dec *p, + SizeT outSize, + const Byte *src, SizeT *srcLen, + int checkFinishBlock) +{ + SizeT inSize = *srcLen; + *srcLen = 0; + + while (p->state != LZMA2_STATE_ERROR) + { + if (p->state == LZMA2_STATE_FINISHED) + return (ELzma2ParseStatus)LZMA_STATUS_FINISHED_WITH_MARK; + + if (outSize == 0 && !checkFinishBlock) + return (ELzma2ParseStatus)LZMA_STATUS_NOT_FINISHED; + + if (p->state != LZMA2_STATE_DATA && p->state != LZMA2_STATE_DATA_CONT) + { + if (*srcLen == inSize) + return (ELzma2ParseStatus)LZMA_STATUS_NEEDS_MORE_INPUT; + (*srcLen)++; + + p->state = Lzma2Dec_UpdateState(p, *src++); + + if (p->state == LZMA2_STATE_UNPACK0) + { + // if (p->decoder.dicPos != 0) + if (p->control == LZMA2_CONTROL_COPY_RESET_DIC || p->control >= 0xE0) + return LZMA2_PARSE_STATUS_NEW_BLOCK; + // if (outSize == 0) return LZMA_STATUS_NOT_FINISHED; + } + + // The following code can be commented. + // It's not big problem, if we read additional input bytes. + // It will be stopped later in LZMA2_STATE_DATA / LZMA2_STATE_DATA_CONT state. + + if (outSize == 0 && p->state != LZMA2_STATE_FINISHED) + { + // checkFinishBlock is true. So we expect that block must be finished, + // We can return LZMA_STATUS_NOT_SPECIFIED or LZMA_STATUS_NOT_FINISHED here + // break; + return (ELzma2ParseStatus)LZMA_STATUS_NOT_FINISHED; + } + + if (p->state == LZMA2_STATE_DATA) + return LZMA2_PARSE_STATUS_NEW_CHUNK; + + continue; + } + + if (outSize == 0) + return (ELzma2ParseStatus)LZMA_STATUS_NOT_FINISHED; + + { + SizeT inCur = inSize - *srcLen; + + if (LZMA2_IS_UNCOMPRESSED_STATE(p)) + { + if (inCur == 0) + return (ELzma2ParseStatus)LZMA_STATUS_NEEDS_MORE_INPUT; + if (inCur > p->unpackSize) + inCur = p->unpackSize; + if (inCur > outSize) + inCur = outSize; + p->decoder.dicPos += inCur; + src += inCur; + *srcLen += inCur; + outSize -= inCur; + p->unpackSize -= (UInt32)inCur; + p->state = (p->unpackSize == 0) ? LZMA2_STATE_CONTROL : LZMA2_STATE_DATA_CONT; + } + else + { + p->isExtraMode = True; + + if (inCur == 0) + { + if (p->packSize != 0) + return (ELzma2ParseStatus)LZMA_STATUS_NEEDS_MORE_INPUT; + } + else if (p->state == LZMA2_STATE_DATA) + { + p->state = LZMA2_STATE_DATA_CONT; + if (*src != 0) + { + // first byte of lzma chunk must be Zero + *srcLen += 1; + p->packSize--; + break; + } + } + + if (inCur > p->packSize) + inCur = (SizeT)p->packSize; + + src += inCur; + *srcLen += inCur; + p->packSize -= (UInt32)inCur; + + if (p->packSize == 0) + { + SizeT rem = outSize; + if (rem > p->unpackSize) + rem = p->unpackSize; + p->decoder.dicPos += rem; + p->unpackSize -= (UInt32)rem; + outSize -= rem; + if (p->unpackSize == 0) + p->state = LZMA2_STATE_CONTROL; + } + } + } + } + + p->state = LZMA2_STATE_ERROR; + return (ELzma2ParseStatus)LZMA_STATUS_NOT_SPECIFIED; +} + + + + +SRes Lzma2Dec_DecodeToBuf(CLzma2Dec *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT outSize = *destLen, inSize = *srcLen; + *srcLen = *destLen = 0; + + for (;;) + { + SizeT inCur = inSize, outCur, dicPos; + ELzmaFinishMode curFinishMode; + SRes res; + + if (p->decoder.dicPos == p->decoder.dicBufSize) + p->decoder.dicPos = 0; + dicPos = p->decoder.dicPos; + curFinishMode = LZMA_FINISH_ANY; + outCur = p->decoder.dicBufSize - dicPos; + + if (outCur >= outSize) + { + outCur = outSize; + curFinishMode = finishMode; + } + + res = Lzma2Dec_DecodeToDic(p, dicPos + outCur, src, &inCur, curFinishMode, status); + + src += inCur; + inSize -= inCur; + *srcLen += inCur; + outCur = p->decoder.dicPos - dicPos; + memcpy(dest, p->decoder.dic + dicPos, outCur); + dest += outCur; + outSize -= outCur; + *destLen += outCur; + if (res != 0) + return res; + if (outCur == 0 || outSize == 0) + return SZ_OK; + } +} + + +SRes Lzma2Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + Byte prop, ELzmaFinishMode finishMode, ELzmaStatus *status, ISzAllocPtr alloc) +{ + CLzma2Dec p; + SRes res; + SizeT outSize = *destLen, inSize = *srcLen; + *destLen = *srcLen = 0; + *status = LZMA_STATUS_NOT_SPECIFIED; + Lzma2Dec_Construct(&p); + RINOK(Lzma2Dec_AllocateProbs(&p, prop, alloc)); + p.decoder.dic = dest; + p.decoder.dicBufSize = outSize; + Lzma2Dec_Init(&p); + *srcLen = inSize; + res = Lzma2Dec_DecodeToDic(&p, outSize, src, srcLen, finishMode, status); + *destLen = p.decoder.dicPos; + if (res == SZ_OK && *status == LZMA_STATUS_NEEDS_MORE_INPUT) + res = SZ_ERROR_INPUT_EOF; + Lzma2Dec_FreeProbs(&p, alloc); + return res; +} diff --git a/3rdparty/lzma/src/Lzma2DecMt.c b/3rdparty/lzma/src/Lzma2DecMt.c new file mode 100644 index 0000000000..9f1dc52baa --- /dev/null +++ b/3rdparty/lzma/src/Lzma2DecMt.c @@ -0,0 +1,1090 @@ +/* Lzma2DecMt.c -- LZMA2 Decoder Multi-thread +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +// #define SHOW_DEBUG_INFO + +// #define _7ZIP_ST + +#ifdef SHOW_DEBUG_INFO +#include +#endif + +#ifndef _7ZIP_ST +#ifdef SHOW_DEBUG_INFO +#define PRF(x) x +#else +#define PRF(x) +#endif +#define PRF_STR(s) PRF(printf("\n" s "\n")) +#define PRF_STR_INT_2(s, d1, d2) PRF(printf("\n" s " %d %d\n", (unsigned)d1, (unsigned)d2)) +#endif + +#include "Alloc.h" + +#include "Lzma2Dec.h" +#include "Lzma2DecMt.h" + +#ifndef _7ZIP_ST +#include "MtDec.h" + +#define LZMA2DECMT_OUT_BLOCK_MAX_DEFAULT (1 << 28) +#endif + + +void Lzma2DecMtProps_Init(CLzma2DecMtProps *p) +{ + p->inBufSize_ST = 1 << 20; + p->outStep_ST = 1 << 20; + + #ifndef _7ZIP_ST + p->numThreads = 1; + p->inBufSize_MT = 1 << 18; + p->outBlockMax = LZMA2DECMT_OUT_BLOCK_MAX_DEFAULT; + p->inBlockMax = p->outBlockMax + p->outBlockMax / 16; + #endif +} + + + +#ifndef _7ZIP_ST + +/* ---------- CLzma2DecMtThread ---------- */ + +typedef struct +{ + CLzma2Dec dec; + Byte dec_created; + Byte needInit; + + Byte *outBuf; + size_t outBufSize; + + EMtDecParseState state; + ELzma2ParseStatus parseStatus; + + size_t inPreSize; + size_t outPreSize; + + size_t inCodeSize; + size_t outCodeSize; + SRes codeRes; + + CAlignOffsetAlloc alloc; + + Byte mtPad[1 << 7]; +} CLzma2DecMtThread; + +#endif + + +/* ---------- CLzma2DecMt ---------- */ + +typedef struct +{ + // ISzAllocPtr alloc; + ISzAllocPtr allocMid; + + CAlignOffsetAlloc alignOffsetAlloc; + CLzma2DecMtProps props; + Byte prop; + + ISeqInStream *inStream; + ISeqOutStream *outStream; + ICompressProgress *progress; + + BoolInt finishMode; + BoolInt outSize_Defined; + UInt64 outSize; + + UInt64 outProcessed; + UInt64 inProcessed; + BoolInt readWasFinished; + SRes readRes; + + Byte *inBuf; + size_t inBufSize; + Byte dec_created; + CLzma2Dec dec; + + size_t inPos; + size_t inLim; + + #ifndef _7ZIP_ST + UInt64 outProcessed_Parse; + BoolInt mtc_WasConstructed; + CMtDec mtc; + CLzma2DecMtThread coders[MTDEC__THREADS_MAX]; + #endif + +} CLzma2DecMt; + + + +CLzma2DecMtHandle Lzma2DecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid) +{ + CLzma2DecMt *p = (CLzma2DecMt *)ISzAlloc_Alloc(alloc, sizeof(CLzma2DecMt)); + if (!p) + return NULL; + + // p->alloc = alloc; + p->allocMid = allocMid; + + AlignOffsetAlloc_CreateVTable(&p->alignOffsetAlloc); + p->alignOffsetAlloc.numAlignBits = 7; + p->alignOffsetAlloc.offset = 0; + p->alignOffsetAlloc.baseAlloc = alloc; + + p->inBuf = NULL; + p->inBufSize = 0; + p->dec_created = False; + + // Lzma2DecMtProps_Init(&p->props); + + #ifndef _7ZIP_ST + p->mtc_WasConstructed = False; + { + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CLzma2DecMtThread *t = &p->coders[i]; + t->dec_created = False; + t->outBuf = NULL; + t->outBufSize = 0; + } + } + #endif + + return p; +} + + +#ifndef _7ZIP_ST + +static void Lzma2DecMt_FreeOutBufs(CLzma2DecMt *p) +{ + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CLzma2DecMtThread *t = &p->coders[i]; + if (t->outBuf) + { + ISzAlloc_Free(p->allocMid, t->outBuf); + t->outBuf = NULL; + t->outBufSize = 0; + } + } +} + +#endif + + +static void Lzma2DecMt_FreeSt(CLzma2DecMt *p) +{ + if (p->dec_created) + { + Lzma2Dec_Free(&p->dec, &p->alignOffsetAlloc.vt); + p->dec_created = False; + } + if (p->inBuf) + { + ISzAlloc_Free(p->allocMid, p->inBuf); + p->inBuf = NULL; + } + p->inBufSize = 0; +} + + +void Lzma2DecMt_Destroy(CLzma2DecMtHandle pp) +{ + CLzma2DecMt *p = (CLzma2DecMt *)pp; + + Lzma2DecMt_FreeSt(p); + + #ifndef _7ZIP_ST + + if (p->mtc_WasConstructed) + { + MtDec_Destruct(&p->mtc); + p->mtc_WasConstructed = False; + } + { + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CLzma2DecMtThread *t = &p->coders[i]; + if (t->dec_created) + { + // we don't need to free dict here + Lzma2Dec_FreeProbs(&t->dec, &t->alloc.vt); // p->alloc !!! + t->dec_created = False; + } + } + } + Lzma2DecMt_FreeOutBufs(p); + + #endif + + ISzAlloc_Free(p->alignOffsetAlloc.baseAlloc, pp); +} + + + +#ifndef _7ZIP_ST + +static void Lzma2DecMt_MtCallback_Parse(void *obj, unsigned coderIndex, CMtDecCallbackInfo *cc) +{ + CLzma2DecMt *me = (CLzma2DecMt *)obj; + CLzma2DecMtThread *t = &me->coders[coderIndex]; + + PRF_STR_INT_2("Parse", coderIndex, cc->srcSize); + + cc->state = MTDEC_PARSE_CONTINUE; + + if (cc->startCall) + { + if (!t->dec_created) + { + Lzma2Dec_Construct(&t->dec); + t->dec_created = True; + AlignOffsetAlloc_CreateVTable(&t->alloc); + { + /* (1 << 12) is expected size of one way in data cache. + We optimize alignment for cache line size of 128 bytes and smaller */ + const unsigned kNumAlignBits = 12; + const unsigned kNumCacheLineBits = 7; /* <= kNumAlignBits */ + t->alloc.numAlignBits = kNumAlignBits; + t->alloc.offset = ((UInt32)coderIndex * (((unsigned)1 << 11) + (1 << 8) + (1 << 6))) & (((unsigned)1 << kNumAlignBits) - ((unsigned)1 << kNumCacheLineBits)); + t->alloc.baseAlloc = me->alignOffsetAlloc.baseAlloc; + } + } + Lzma2Dec_Init(&t->dec); + + t->inPreSize = 0; + t->outPreSize = 0; + // t->blockWasFinished = False; + // t->finishedWithMark = False; + t->parseStatus = (ELzma2ParseStatus)LZMA_STATUS_NOT_SPECIFIED; + t->state = MTDEC_PARSE_CONTINUE; + + t->inCodeSize = 0; + t->outCodeSize = 0; + t->codeRes = SZ_OK; + + // (cc->srcSize == 0) is allowed + } + + { + ELzma2ParseStatus status; + BoolInt overflow; + UInt32 unpackRem = 0; + + int checkFinishBlock = True; + size_t limit = me->props.outBlockMax; + if (me->outSize_Defined) + { + UInt64 rem = me->outSize - me->outProcessed_Parse; + if (limit >= rem) + { + limit = (size_t)rem; + if (!me->finishMode) + checkFinishBlock = False; + } + } + + // checkFinishBlock = False, if we want to decode partial data + // that must be finished at position <= outBlockMax. + + { + const SizeT srcOrig = cc->srcSize; + SizeT srcSize_Point = 0; + SizeT dicPos_Point = 0; + + cc->srcSize = 0; + overflow = False; + + for (;;) + { + SizeT srcCur = srcOrig - cc->srcSize; + + status = Lzma2Dec_Parse(&t->dec, + limit - t->dec.decoder.dicPos, + cc->src + cc->srcSize, &srcCur, + checkFinishBlock); + + cc->srcSize += srcCur; + + if (status == LZMA2_PARSE_STATUS_NEW_CHUNK) + { + if (t->dec.unpackSize > me->props.outBlockMax - t->dec.decoder.dicPos) + { + overflow = True; + break; + } + continue; + } + + if (status == LZMA2_PARSE_STATUS_NEW_BLOCK) + { + if (t->dec.decoder.dicPos == 0) + continue; + // we decode small blocks in one thread + if (t->dec.decoder.dicPos >= (1 << 14)) + break; + dicPos_Point = t->dec.decoder.dicPos; + srcSize_Point = cc->srcSize; + continue; + } + + if ((int)status == LZMA_STATUS_NOT_FINISHED && checkFinishBlock + // && limit == t->dec.decoder.dicPos + // && limit == me->props.outBlockMax + ) + { + overflow = True; + break; + } + + unpackRem = Lzma2Dec_GetUnpackExtra(&t->dec); + break; + } + + if (dicPos_Point != 0 + && (int)status != LZMA2_PARSE_STATUS_NEW_BLOCK + && (int)status != LZMA_STATUS_FINISHED_WITH_MARK + && (int)status != LZMA_STATUS_NOT_SPECIFIED) + { + // we revert to latest newBlock state + status = LZMA2_PARSE_STATUS_NEW_BLOCK; + unpackRem = 0; + t->dec.decoder.dicPos = dicPos_Point; + cc->srcSize = srcSize_Point; + overflow = False; + } + } + + t->inPreSize += cc->srcSize; + t->parseStatus = status; + + if (overflow) + cc->state = MTDEC_PARSE_OVERFLOW; + else + { + size_t dicPos = t->dec.decoder.dicPos; + + if ((int)status != LZMA_STATUS_NEEDS_MORE_INPUT) + { + if (status == LZMA2_PARSE_STATUS_NEW_BLOCK) + { + cc->state = MTDEC_PARSE_NEW; + cc->srcSize--; // we don't need control byte of next block + t->inPreSize--; + } + else + { + cc->state = MTDEC_PARSE_END; + if ((int)status != LZMA_STATUS_FINISHED_WITH_MARK) + { + // (status == LZMA_STATUS_NOT_SPECIFIED) + // (status == LZMA_STATUS_NOT_FINISHED) + if (unpackRem != 0) + { + /* we also reserve space for max possible number of output bytes of current LZMA chunk */ + SizeT rem = limit - dicPos; + if (rem > unpackRem) + rem = unpackRem; + dicPos += rem; + } + } + } + + me->outProcessed_Parse += dicPos; + } + + cc->outPos = dicPos; + t->outPreSize = (size_t)dicPos; + } + + t->state = cc->state; + return; + } +} + + +static SRes Lzma2DecMt_MtCallback_PreCode(void *pp, unsigned coderIndex) +{ + CLzma2DecMt *me = (CLzma2DecMt *)pp; + CLzma2DecMtThread *t = &me->coders[coderIndex]; + Byte *dest = t->outBuf; + + if (t->inPreSize == 0) + { + t->codeRes = SZ_ERROR_DATA; + return t->codeRes; + } + + if (!dest || t->outBufSize < t->outPreSize) + { + if (dest) + { + ISzAlloc_Free(me->allocMid, dest); + t->outBuf = NULL; + t->outBufSize = 0; + } + + dest = (Byte *)ISzAlloc_Alloc(me->allocMid, t->outPreSize + // + (1 << 28) + ); + // Sleep(200); + if (!dest) + return SZ_ERROR_MEM; + t->outBuf = dest; + t->outBufSize = t->outPreSize; + } + + t->dec.decoder.dic = dest; + t->dec.decoder.dicBufSize = t->outPreSize; + + t->needInit = True; + + return Lzma2Dec_AllocateProbs(&t->dec, me->prop, &t->alloc.vt); // alloc.vt +} + + +static SRes Lzma2DecMt_MtCallback_Code(void *pp, unsigned coderIndex, + const Byte *src, size_t srcSize, int srcFinished, + // int finished, int blockFinished, + UInt64 *inCodePos, UInt64 *outCodePos, int *stop) +{ + CLzma2DecMt *me = (CLzma2DecMt *)pp; + CLzma2DecMtThread *t = &me->coders[coderIndex]; + + UNUSED_VAR(srcFinished) + + PRF_STR_INT_2("Code", coderIndex, srcSize); + + *inCodePos = t->inCodeSize; + *outCodePos = 0; + *stop = True; + + if (t->needInit) + { + Lzma2Dec_Init(&t->dec); + t->needInit = False; + } + + { + ELzmaStatus status; + size_t srcProcessed = srcSize; + BoolInt blockWasFinished = + ((int)t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK + || t->parseStatus == LZMA2_PARSE_STATUS_NEW_BLOCK); + + SRes res = Lzma2Dec_DecodeToDic(&t->dec, + t->outPreSize, + src, &srcProcessed, + blockWasFinished ? LZMA_FINISH_END : LZMA_FINISH_ANY, + &status); + + t->codeRes = res; + + t->inCodeSize += srcProcessed; + *inCodePos = t->inCodeSize; + t->outCodeSize = t->dec.decoder.dicPos; + *outCodePos = t->dec.decoder.dicPos; + + if (res != SZ_OK) + return res; + + if (srcProcessed == srcSize) + *stop = False; + + if (blockWasFinished) + { + if (srcSize != srcProcessed) + return SZ_ERROR_FAIL; + + if (t->inPreSize == t->inCodeSize) + { + if (t->outPreSize != t->outCodeSize) + return SZ_ERROR_FAIL; + *stop = True; + } + } + else + { + if (t->outPreSize == t->outCodeSize) + *stop = True; + } + + return SZ_OK; + } +} + + +#define LZMA2DECMT_STREAM_WRITE_STEP (1 << 24) + +static SRes Lzma2DecMt_MtCallback_Write(void *pp, unsigned coderIndex, + BoolInt needWriteToStream, + const Byte *src, size_t srcSize, BoolInt isCross, + BoolInt *needContinue, BoolInt *canRecode) +{ + CLzma2DecMt *me = (CLzma2DecMt *)pp; + const CLzma2DecMtThread *t = &me->coders[coderIndex]; + size_t size = t->outCodeSize; + const Byte *data = t->outBuf; + BoolInt needContinue2 = True; + + UNUSED_VAR(src) + UNUSED_VAR(srcSize) + UNUSED_VAR(isCross) + + PRF_STR_INT_2("Write", coderIndex, srcSize); + + *needContinue = False; + *canRecode = True; + + if ( + // t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK + t->state == MTDEC_PARSE_OVERFLOW + || t->state == MTDEC_PARSE_END) + needContinue2 = False; + + + if (!needWriteToStream) + return SZ_OK; + + me->mtc.inProcessed += t->inCodeSize; + + if (t->codeRes == SZ_OK) + if ((int)t->parseStatus == LZMA_STATUS_FINISHED_WITH_MARK + || t->parseStatus == LZMA2_PARSE_STATUS_NEW_BLOCK) + if (t->outPreSize != t->outCodeSize + || t->inPreSize != t->inCodeSize) + return SZ_ERROR_FAIL; + + *canRecode = False; + + if (me->outStream) + { + for (;;) + { + size_t cur = size; + size_t written; + if (cur > LZMA2DECMT_STREAM_WRITE_STEP) + cur = LZMA2DECMT_STREAM_WRITE_STEP; + + written = ISeqOutStream_Write(me->outStream, data, cur); + + me->outProcessed += written; + // me->mtc.writtenTotal += written; + if (written != cur) + return SZ_ERROR_WRITE; + data += cur; + size -= cur; + if (size == 0) + { + *needContinue = needContinue2; + return SZ_OK; + } + RINOK(MtProgress_ProgressAdd(&me->mtc.mtProgress, 0, 0)); + } + } + + return SZ_ERROR_FAIL; + /* + if (size > me->outBufSize) + return SZ_ERROR_OUTPUT_EOF; + memcpy(me->outBuf, data, size); + me->outBufSize -= size; + me->outBuf += size; + *needContinue = needContinue2; + return SZ_OK; + */ +} + +#endif + + +static SRes Lzma2Dec_Prepare_ST(CLzma2DecMt *p) +{ + if (!p->dec_created) + { + Lzma2Dec_Construct(&p->dec); + p->dec_created = True; + } + + RINOK(Lzma2Dec_Allocate(&p->dec, p->prop, &p->alignOffsetAlloc.vt)); + + if (!p->inBuf || p->inBufSize != p->props.inBufSize_ST) + { + ISzAlloc_Free(p->allocMid, p->inBuf); + p->inBufSize = 0; + p->inBuf = (Byte *)ISzAlloc_Alloc(p->allocMid, p->props.inBufSize_ST); + if (!p->inBuf) + return SZ_ERROR_MEM; + p->inBufSize = p->props.inBufSize_ST; + } + + Lzma2Dec_Init(&p->dec); + + return SZ_OK; +} + + +static SRes Lzma2Dec_Decode_ST(CLzma2DecMt *p + #ifndef _7ZIP_ST + , BoolInt tMode + #endif + ) +{ + SizeT wrPos; + size_t inPos, inLim; + const Byte *inData; + UInt64 inPrev, outPrev; + + CLzma2Dec *dec; + + #ifndef _7ZIP_ST + if (tMode) + { + Lzma2DecMt_FreeOutBufs(p); + tMode = MtDec_PrepareRead(&p->mtc); + } + #endif + + RINOK(Lzma2Dec_Prepare_ST(p)); + + dec = &p->dec; + + inPrev = p->inProcessed; + outPrev = p->outProcessed; + + inPos = 0; + inLim = 0; + inData = NULL; + wrPos = dec->decoder.dicPos; + + for (;;) + { + SizeT dicPos; + SizeT size; + ELzmaFinishMode finishMode; + SizeT inProcessed; + ELzmaStatus status; + SRes res; + + SizeT outProcessed; + BoolInt outFinished; + BoolInt needStop; + + if (inPos == inLim) + { + #ifndef _7ZIP_ST + if (tMode) + { + inData = MtDec_Read(&p->mtc, &inLim); + inPos = 0; + if (inData) + continue; + tMode = False; + inLim = 0; + } + #endif + + if (!p->readWasFinished) + { + inPos = 0; + inLim = p->inBufSize; + inData = p->inBuf; + p->readRes = ISeqInStream_Read(p->inStream, (void *)(p->inBuf), &inLim); + // p->readProcessed += inLim; + // inLim -= 5; p->readWasFinished = True; // for test + if (inLim == 0 || p->readRes != SZ_OK) + p->readWasFinished = True; + } + } + + dicPos = dec->decoder.dicPos; + { + SizeT next = dec->decoder.dicBufSize; + if (next - wrPos > p->props.outStep_ST) + next = wrPos + p->props.outStep_ST; + size = next - dicPos; + } + + finishMode = LZMA_FINISH_ANY; + if (p->outSize_Defined) + { + const UInt64 rem = p->outSize - p->outProcessed; + if (size >= rem) + { + size = (SizeT)rem; + if (p->finishMode) + finishMode = LZMA_FINISH_END; + } + } + + inProcessed = inLim - inPos; + + res = Lzma2Dec_DecodeToDic(dec, dicPos + size, inData + inPos, &inProcessed, finishMode, &status); + + inPos += inProcessed; + p->inProcessed += inProcessed; + outProcessed = dec->decoder.dicPos - dicPos; + p->outProcessed += outProcessed; + + outFinished = (p->outSize_Defined && p->outSize <= p->outProcessed); + + needStop = (res != SZ_OK + || (inProcessed == 0 && outProcessed == 0) + || status == LZMA_STATUS_FINISHED_WITH_MARK + || (!p->finishMode && outFinished)); + + if (needStop || outProcessed >= size) + { + SRes res2; + { + size_t writeSize = dec->decoder.dicPos - wrPos; + size_t written = ISeqOutStream_Write(p->outStream, dec->decoder.dic + wrPos, writeSize); + res2 = (written == writeSize) ? SZ_OK : SZ_ERROR_WRITE; + } + + if (dec->decoder.dicPos == dec->decoder.dicBufSize) + dec->decoder.dicPos = 0; + wrPos = dec->decoder.dicPos; + + RINOK(res2); + + if (needStop) + { + if (res != SZ_OK) + return res; + + if (status == LZMA_STATUS_FINISHED_WITH_MARK) + { + if (p->finishMode) + { + if (p->outSize_Defined && p->outSize != p->outProcessed) + return SZ_ERROR_DATA; + } + return SZ_OK; + } + + if (!p->finishMode && outFinished) + return SZ_OK; + + if (status == LZMA_STATUS_NEEDS_MORE_INPUT) + return SZ_ERROR_INPUT_EOF; + + return SZ_ERROR_DATA; + } + } + + if (p->progress) + { + UInt64 inDelta = p->inProcessed - inPrev; + UInt64 outDelta = p->outProcessed - outPrev; + if (inDelta >= (1 << 22) || outDelta >= (1 << 22)) + { + RINOK(ICompressProgress_Progress(p->progress, p->inProcessed, p->outProcessed)); + inPrev = p->inProcessed; + outPrev = p->outProcessed; + } + } + } +} + + + +SRes Lzma2DecMt_Decode(CLzma2DecMtHandle pp, + Byte prop, + const CLzma2DecMtProps *props, + ISeqOutStream *outStream, const UInt64 *outDataSize, int finishMode, + // Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + // const Byte *inData, size_t inDataSize, + UInt64 *inProcessed, + // UInt64 *outProcessed, + int *isMT, + ICompressProgress *progress) +{ + CLzma2DecMt *p = (CLzma2DecMt *)pp; + #ifndef _7ZIP_ST + BoolInt tMode; + #endif + + *inProcessed = 0; + + if (prop > 40) + return SZ_ERROR_UNSUPPORTED; + + p->prop = prop; + p->props = *props; + + p->inStream = inStream; + p->outStream = outStream; + p->progress = progress; + + p->outSize = 0; + p->outSize_Defined = False; + if (outDataSize) + { + p->outSize_Defined = True; + p->outSize = *outDataSize; + } + p->finishMode = finishMode; + + p->outProcessed = 0; + p->inProcessed = 0; + + p->readWasFinished = False; + p->readRes = SZ_OK; + + *isMT = False; + + + #ifndef _7ZIP_ST + + tMode = False; + + // p->mtc.parseRes = SZ_OK; + + // p->mtc.numFilledThreads = 0; + // p->mtc.crossStart = 0; + // p->mtc.crossEnd = 0; + // p->mtc.allocError_for_Read_BlockIndex = 0; + // p->mtc.isAllocError = False; + + if (p->props.numThreads > 1) + { + IMtDecCallback2 vt; + + Lzma2DecMt_FreeSt(p); + + p->outProcessed_Parse = 0; + + if (!p->mtc_WasConstructed) + { + p->mtc_WasConstructed = True; + MtDec_Construct(&p->mtc); + } + + p->mtc.progress = progress; + p->mtc.inStream = inStream; + + // p->outBuf = NULL; + // p->outBufSize = 0; + /* + if (!outStream) + { + // p->outBuf = outBuf; + // p->outBufSize = *outBufSize; + // *outBufSize = 0; + return SZ_ERROR_PARAM; + } + */ + + // p->mtc.inBlockMax = p->props.inBlockMax; + p->mtc.alloc = &p->alignOffsetAlloc.vt; + // p->alignOffsetAlloc.baseAlloc; + // p->mtc.inData = inData; + // p->mtc.inDataSize = inDataSize; + p->mtc.mtCallback = &vt; + p->mtc.mtCallbackObject = p; + + p->mtc.inBufSize = p->props.inBufSize_MT; + + p->mtc.numThreadsMax = p->props.numThreads; + + *isMT = True; + + vt.Parse = Lzma2DecMt_MtCallback_Parse; + vt.PreCode = Lzma2DecMt_MtCallback_PreCode; + vt.Code = Lzma2DecMt_MtCallback_Code; + vt.Write = Lzma2DecMt_MtCallback_Write; + + { + BoolInt needContinue = False; + + SRes res = MtDec_Code(&p->mtc); + + /* + if (!outStream) + *outBufSize = p->outBuf - outBuf; + */ + + *inProcessed = p->mtc.inProcessed; + + needContinue = False; + + if (res == SZ_OK) + { + if (p->mtc.mtProgress.res != SZ_OK) + res = p->mtc.mtProgress.res; + else + needContinue = p->mtc.needContinue; + } + + if (!needContinue) + { + if (res == SZ_OK) + return p->mtc.readRes; + return res; + } + + tMode = True; + p->readRes = p->mtc.readRes; + p->readWasFinished = p->mtc.readWasFinished; + p->inProcessed = p->mtc.inProcessed; + + PRF_STR("----- decoding ST -----"); + } + } + + #endif + + + *isMT = False; + + { + SRes res = Lzma2Dec_Decode_ST(p + #ifndef _7ZIP_ST + , tMode + #endif + ); + + *inProcessed = p->inProcessed; + + // res = SZ_OK; // for test + if (res == SZ_ERROR_INPUT_EOF) + { + if (p->readRes != SZ_OK) + res = p->readRes; + } + else if (res == SZ_OK && p->readRes != SZ_OK) + res = p->readRes; + + /* + #ifndef _7ZIP_ST + if (res == SZ_OK && tMode && p->mtc.parseRes != SZ_OK) + res = p->mtc.parseRes; + #endif + */ + + return res; + } +} + + +/* ---------- Read from CLzma2DecMtHandle Interface ---------- */ + +SRes Lzma2DecMt_Init(CLzma2DecMtHandle pp, + Byte prop, + const CLzma2DecMtProps *props, + const UInt64 *outDataSize, int finishMode, + ISeqInStream *inStream) +{ + CLzma2DecMt *p = (CLzma2DecMt *)pp; + + if (prop > 40) + return SZ_ERROR_UNSUPPORTED; + + p->prop = prop; + p->props = *props; + + p->inStream = inStream; + + p->outSize = 0; + p->outSize_Defined = False; + if (outDataSize) + { + p->outSize_Defined = True; + p->outSize = *outDataSize; + } + p->finishMode = finishMode; + + p->outProcessed = 0; + p->inProcessed = 0; + + p->inPos = 0; + p->inLim = 0; + + return Lzma2Dec_Prepare_ST(p); +} + + +SRes Lzma2DecMt_Read(CLzma2DecMtHandle pp, + Byte *data, size_t *outSize, + UInt64 *inStreamProcessed) +{ + CLzma2DecMt *p = (CLzma2DecMt *)pp; + ELzmaFinishMode finishMode; + SRes readRes; + size_t size = *outSize; + + *outSize = 0; + *inStreamProcessed = 0; + + finishMode = LZMA_FINISH_ANY; + if (p->outSize_Defined) + { + const UInt64 rem = p->outSize - p->outProcessed; + if (size >= rem) + { + size = (size_t)rem; + if (p->finishMode) + finishMode = LZMA_FINISH_END; + } + } + + readRes = SZ_OK; + + for (;;) + { + SizeT inCur; + SizeT outCur; + ELzmaStatus status; + SRes res; + + if (p->inPos == p->inLim && readRes == SZ_OK) + { + p->inPos = 0; + p->inLim = p->props.inBufSize_ST; + readRes = ISeqInStream_Read(p->inStream, p->inBuf, &p->inLim); + } + + inCur = p->inLim - p->inPos; + outCur = size; + + res = Lzma2Dec_DecodeToBuf(&p->dec, data, &outCur, + p->inBuf + p->inPos, &inCur, finishMode, &status); + + p->inPos += inCur; + p->inProcessed += inCur; + *inStreamProcessed += inCur; + p->outProcessed += outCur; + *outSize += outCur; + size -= outCur; + data += outCur; + + if (res != 0) + return res; + + /* + if (status == LZMA_STATUS_FINISHED_WITH_MARK) + return readRes; + + if (size == 0 && status != LZMA_STATUS_NEEDS_MORE_INPUT) + { + if (p->finishMode && p->outSize_Defined && p->outProcessed >= p->outSize) + return SZ_ERROR_DATA; + return readRes; + } + */ + + if (inCur == 0 && outCur == 0) + return readRes; + } +} diff --git a/3rdparty/lzma/src/Lzma2Enc.c b/3rdparty/lzma/src/Lzma2Enc.c new file mode 100644 index 0000000000..e61a5dfe0b --- /dev/null +++ b/3rdparty/lzma/src/Lzma2Enc.c @@ -0,0 +1,803 @@ +/* Lzma2Enc.c -- LZMA2 Encoder +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #define _7ZIP_ST */ + +#include "Lzma2Enc.h" + +#ifndef _7ZIP_ST +#include "MtCoder.h" +#else +#define MTCODER__THREADS_MAX 1 +#endif + +#define LZMA2_CONTROL_LZMA (1 << 7) +#define LZMA2_CONTROL_COPY_NO_RESET 2 +#define LZMA2_CONTROL_COPY_RESET_DIC 1 +#define LZMA2_CONTROL_EOF 0 + +#define LZMA2_LCLP_MAX 4 + +#define LZMA2_DIC_SIZE_FROM_PROP(p) (((UInt32)2 | ((p) & 1)) << ((p) / 2 + 11)) + +#define LZMA2_PACK_SIZE_MAX (1 << 16) +#define LZMA2_COPY_CHUNK_SIZE LZMA2_PACK_SIZE_MAX +#define LZMA2_UNPACK_SIZE_MAX (1 << 21) +#define LZMA2_KEEP_WINDOW_SIZE LZMA2_UNPACK_SIZE_MAX + +#define LZMA2_CHUNK_SIZE_COMPRESSED_MAX ((1 << 16) + 16) + + +#define PRF(x) /* x */ + + +/* ---------- CLimitedSeqInStream ---------- */ + +typedef struct +{ + ISeqInStream vt; + ISeqInStream *realStream; + UInt64 limit; + UInt64 processed; + int finished; +} CLimitedSeqInStream; + +static void LimitedSeqInStream_Init(CLimitedSeqInStream *p) +{ + p->limit = (UInt64)(Int64)-1; + p->processed = 0; + p->finished = 0; +} + +static SRes LimitedSeqInStream_Read(const ISeqInStream *pp, void *data, size_t *size) +{ + CLimitedSeqInStream *p = CONTAINER_FROM_VTBL(pp, CLimitedSeqInStream, vt); + size_t size2 = *size; + SRes res = SZ_OK; + + if (p->limit != (UInt64)(Int64)-1) + { + UInt64 rem = p->limit - p->processed; + if (size2 > rem) + size2 = (size_t)rem; + } + if (size2 != 0) + { + res = ISeqInStream_Read(p->realStream, data, &size2); + p->finished = (size2 == 0 ? 1 : 0); + p->processed += size2; + } + *size = size2; + return res; +} + + +/* ---------- CLzma2EncInt ---------- */ + +typedef struct +{ + CLzmaEncHandle enc; + Byte propsAreSet; + Byte propsByte; + Byte needInitState; + Byte needInitProp; + UInt64 srcPos; +} CLzma2EncInt; + + +static SRes Lzma2EncInt_InitStream(CLzma2EncInt *p, const CLzma2EncProps *props) +{ + if (!p->propsAreSet) + { + SizeT propsSize = LZMA_PROPS_SIZE; + Byte propsEncoded[LZMA_PROPS_SIZE]; + RINOK(LzmaEnc_SetProps(p->enc, &props->lzmaProps)); + RINOK(LzmaEnc_WriteProperties(p->enc, propsEncoded, &propsSize)); + p->propsByte = propsEncoded[0]; + p->propsAreSet = True; + } + return SZ_OK; +} + +static void Lzma2EncInt_InitBlock(CLzma2EncInt *p) +{ + p->srcPos = 0; + p->needInitState = True; + p->needInitProp = True; +} + + +SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, ISeqInStream *inStream, UInt32 keepWindowSize, + ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, + UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, + Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize); +const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp); +void LzmaEnc_Finish(CLzmaEncHandle pp); +void LzmaEnc_SaveState(CLzmaEncHandle pp); +void LzmaEnc_RestoreState(CLzmaEncHandle pp); + +/* +UInt32 LzmaEnc_GetNumAvailableBytes(CLzmaEncHandle pp); +*/ + +static SRes Lzma2EncInt_EncodeSubblock(CLzma2EncInt *p, Byte *outBuf, + size_t *packSizeRes, ISeqOutStream *outStream) +{ + size_t packSizeLimit = *packSizeRes; + size_t packSize = packSizeLimit; + UInt32 unpackSize = LZMA2_UNPACK_SIZE_MAX; + unsigned lzHeaderSize = 5 + (p->needInitProp ? 1 : 0); + BoolInt useCopyBlock; + SRes res; + + *packSizeRes = 0; + if (packSize < lzHeaderSize) + return SZ_ERROR_OUTPUT_EOF; + packSize -= lzHeaderSize; + + LzmaEnc_SaveState(p->enc); + res = LzmaEnc_CodeOneMemBlock(p->enc, p->needInitState, + outBuf + lzHeaderSize, &packSize, LZMA2_PACK_SIZE_MAX, &unpackSize); + + PRF(printf("\npackSize = %7d unpackSize = %7d ", packSize, unpackSize)); + + if (unpackSize == 0) + return res; + + if (res == SZ_OK) + useCopyBlock = (packSize + 2 >= unpackSize || packSize > (1 << 16)); + else + { + if (res != SZ_ERROR_OUTPUT_EOF) + return res; + res = SZ_OK; + useCopyBlock = True; + } + + if (useCopyBlock) + { + size_t destPos = 0; + PRF(printf("################# COPY ")); + + while (unpackSize > 0) + { + UInt32 u = (unpackSize < LZMA2_COPY_CHUNK_SIZE) ? unpackSize : LZMA2_COPY_CHUNK_SIZE; + if (packSizeLimit - destPos < u + 3) + return SZ_ERROR_OUTPUT_EOF; + outBuf[destPos++] = (Byte)(p->srcPos == 0 ? LZMA2_CONTROL_COPY_RESET_DIC : LZMA2_CONTROL_COPY_NO_RESET); + outBuf[destPos++] = (Byte)((u - 1) >> 8); + outBuf[destPos++] = (Byte)(u - 1); + memcpy(outBuf + destPos, LzmaEnc_GetCurBuf(p->enc) - unpackSize, u); + unpackSize -= u; + destPos += u; + p->srcPos += u; + + if (outStream) + { + *packSizeRes += destPos; + if (ISeqOutStream_Write(outStream, outBuf, destPos) != destPos) + return SZ_ERROR_WRITE; + destPos = 0; + } + else + *packSizeRes = destPos; + /* needInitState = True; */ + } + + LzmaEnc_RestoreState(p->enc); + return SZ_OK; + } + + { + size_t destPos = 0; + UInt32 u = unpackSize - 1; + UInt32 pm = (UInt32)(packSize - 1); + unsigned mode = (p->srcPos == 0) ? 3 : (p->needInitState ? (p->needInitProp ? 2 : 1) : 0); + + PRF(printf(" ")); + + outBuf[destPos++] = (Byte)(LZMA2_CONTROL_LZMA | (mode << 5) | ((u >> 16) & 0x1F)); + outBuf[destPos++] = (Byte)(u >> 8); + outBuf[destPos++] = (Byte)u; + outBuf[destPos++] = (Byte)(pm >> 8); + outBuf[destPos++] = (Byte)pm; + + if (p->needInitProp) + outBuf[destPos++] = p->propsByte; + + p->needInitProp = False; + p->needInitState = False; + destPos += packSize; + p->srcPos += unpackSize; + + if (outStream) + if (ISeqOutStream_Write(outStream, outBuf, destPos) != destPos) + return SZ_ERROR_WRITE; + + *packSizeRes = destPos; + return SZ_OK; + } +} + + +/* ---------- Lzma2 Props ---------- */ + +void Lzma2EncProps_Init(CLzma2EncProps *p) +{ + LzmaEncProps_Init(&p->lzmaProps); + p->blockSize = LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO; + p->numBlockThreads_Reduced = -1; + p->numBlockThreads_Max = -1; + p->numTotalThreads = -1; +} + +void Lzma2EncProps_Normalize(CLzma2EncProps *p) +{ + UInt64 fileSize; + int t1, t1n, t2, t2r, t3; + { + CLzmaEncProps lzmaProps = p->lzmaProps; + LzmaEncProps_Normalize(&lzmaProps); + t1n = lzmaProps.numThreads; + } + + t1 = p->lzmaProps.numThreads; + t2 = p->numBlockThreads_Max; + t3 = p->numTotalThreads; + + if (t2 > MTCODER__THREADS_MAX) + t2 = MTCODER__THREADS_MAX; + + if (t3 <= 0) + { + if (t2 <= 0) + t2 = 1; + t3 = t1n * t2; + } + else if (t2 <= 0) + { + t2 = t3 / t1n; + if (t2 == 0) + { + t1 = 1; + t2 = t3; + } + if (t2 > MTCODER__THREADS_MAX) + t2 = MTCODER__THREADS_MAX; + } + else if (t1 <= 0) + { + t1 = t3 / t2; + if (t1 == 0) + t1 = 1; + } + else + t3 = t1n * t2; + + p->lzmaProps.numThreads = t1; + + t2r = t2; + + fileSize = p->lzmaProps.reduceSize; + + if ( p->blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + && p->blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO + && (p->blockSize < fileSize || fileSize == (UInt64)(Int64)-1)) + p->lzmaProps.reduceSize = p->blockSize; + + LzmaEncProps_Normalize(&p->lzmaProps); + + p->lzmaProps.reduceSize = fileSize; + + t1 = p->lzmaProps.numThreads; + + if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) + { + t2r = t2 = 1; + t3 = t1; + } + else if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO && t2 <= 1) + { + /* if there is no block multi-threading, we use SOLID block */ + p->blockSize = LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID; + } + else + { + if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) + { + const UInt32 kMinSize = (UInt32)1 << 20; + const UInt32 kMaxSize = (UInt32)1 << 28; + const UInt32 dictSize = p->lzmaProps.dictSize; + UInt64 blockSize = (UInt64)dictSize << 2; + if (blockSize < kMinSize) blockSize = kMinSize; + if (blockSize > kMaxSize) blockSize = kMaxSize; + if (blockSize < dictSize) blockSize = dictSize; + blockSize += (kMinSize - 1); + blockSize &= ~(UInt64)(kMinSize - 1); + p->blockSize = blockSize; + } + + if (t2 > 1 && fileSize != (UInt64)(Int64)-1) + { + UInt64 numBlocks = fileSize / p->blockSize; + if (numBlocks * p->blockSize != fileSize) + numBlocks++; + if (numBlocks < (unsigned)t2) + { + t2r = (int)numBlocks; + if (t2r == 0) + t2r = 1; + t3 = t1 * t2r; + } + } + } + + p->numBlockThreads_Max = t2; + p->numBlockThreads_Reduced = t2r; + p->numTotalThreads = t3; +} + + +static SRes Progress(ICompressProgress *p, UInt64 inSize, UInt64 outSize) +{ + return (p && ICompressProgress_Progress(p, inSize, outSize) != SZ_OK) ? SZ_ERROR_PROGRESS : SZ_OK; +} + + +/* ---------- Lzma2 ---------- */ + +typedef struct +{ + Byte propEncoded; + CLzma2EncProps props; + UInt64 expectedDataSize; + + Byte *tempBufLzma; + + ISzAllocPtr alloc; + ISzAllocPtr allocBig; + + CLzma2EncInt coders[MTCODER__THREADS_MAX]; + + #ifndef _7ZIP_ST + + ISeqOutStream *outStream; + Byte *outBuf; + size_t outBuf_Rem; /* remainder in outBuf */ + + size_t outBufSize; /* size of allocated outBufs[i] */ + size_t outBufsDataSizes[MTCODER__BLOCKS_MAX]; + BoolInt mtCoder_WasConstructed; + CMtCoder mtCoder; + Byte *outBufs[MTCODER__BLOCKS_MAX]; + + #endif + +} CLzma2Enc; + + + +CLzma2EncHandle Lzma2Enc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzma2Enc *p = (CLzma2Enc *)ISzAlloc_Alloc(alloc, sizeof(CLzma2Enc)); + if (!p) + return NULL; + Lzma2EncProps_Init(&p->props); + Lzma2EncProps_Normalize(&p->props); + p->expectedDataSize = (UInt64)(Int64)-1; + p->tempBufLzma = NULL; + p->alloc = alloc; + p->allocBig = allocBig; + { + unsigned i; + for (i = 0; i < MTCODER__THREADS_MAX; i++) + p->coders[i].enc = NULL; + } + + #ifndef _7ZIP_ST + p->mtCoder_WasConstructed = False; + { + unsigned i; + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + p->outBufs[i] = NULL; + p->outBufSize = 0; + } + #endif + + return p; +} + + +#ifndef _7ZIP_ST + +static void Lzma2Enc_FreeOutBufs(CLzma2Enc *p) +{ + unsigned i; + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + if (p->outBufs[i]) + { + ISzAlloc_Free(p->alloc, p->outBufs[i]); + p->outBufs[i] = NULL; + } + p->outBufSize = 0; +} + +#endif + + +void Lzma2Enc_Destroy(CLzma2EncHandle pp) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + unsigned i; + for (i = 0; i < MTCODER__THREADS_MAX; i++) + { + CLzma2EncInt *t = &p->coders[i]; + if (t->enc) + { + LzmaEnc_Destroy(t->enc, p->alloc, p->allocBig); + t->enc = NULL; + } + } + + + #ifndef _7ZIP_ST + if (p->mtCoder_WasConstructed) + { + MtCoder_Destruct(&p->mtCoder); + p->mtCoder_WasConstructed = False; + } + Lzma2Enc_FreeOutBufs(p); + #endif + + ISzAlloc_Free(p->alloc, p->tempBufLzma); + p->tempBufLzma = NULL; + + ISzAlloc_Free(p->alloc, pp); +} + + +SRes Lzma2Enc_SetProps(CLzma2EncHandle pp, const CLzma2EncProps *props) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + CLzmaEncProps lzmaProps = props->lzmaProps; + LzmaEncProps_Normalize(&lzmaProps); + if (lzmaProps.lc + lzmaProps.lp > LZMA2_LCLP_MAX) + return SZ_ERROR_PARAM; + p->props = *props; + Lzma2EncProps_Normalize(&p->props); + return SZ_OK; +} + + +void Lzma2Enc_SetDataSize(CLzmaEncHandle pp, UInt64 expectedDataSiize) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + p->expectedDataSize = expectedDataSiize; +} + + +Byte Lzma2Enc_WriteProperties(CLzma2EncHandle pp) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + unsigned i; + UInt32 dicSize = LzmaEncProps_GetDictSize(&p->props.lzmaProps); + for (i = 0; i < 40; i++) + if (dicSize <= LZMA2_DIC_SIZE_FROM_PROP(i)) + break; + return (Byte)i; +} + + +static SRes Lzma2Enc_EncodeMt1( + CLzma2Enc *me, + CLzma2EncInt *p, + ISeqOutStream *outStream, + Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + const Byte *inData, size_t inDataSize, + int finished, + ICompressProgress *progress) +{ + UInt64 unpackTotal = 0; + UInt64 packTotal = 0; + size_t outLim = 0; + CLimitedSeqInStream limitedInStream; + + if (outBuf) + { + outLim = *outBufSize; + *outBufSize = 0; + } + + if (!p->enc) + { + p->propsAreSet = False; + p->enc = LzmaEnc_Create(me->alloc); + if (!p->enc) + return SZ_ERROR_MEM; + } + + limitedInStream.realStream = inStream; + if (inStream) + { + limitedInStream.vt.Read = LimitedSeqInStream_Read; + } + + if (!outBuf) + { + // outStream version works only in one thread. So we use CLzma2Enc::tempBufLzma + if (!me->tempBufLzma) + { + me->tempBufLzma = (Byte *)ISzAlloc_Alloc(me->alloc, LZMA2_CHUNK_SIZE_COMPRESSED_MAX); + if (!me->tempBufLzma) + return SZ_ERROR_MEM; + } + } + + RINOK(Lzma2EncInt_InitStream(p, &me->props)); + + for (;;) + { + SRes res = SZ_OK; + size_t inSizeCur = 0; + + Lzma2EncInt_InitBlock(p); + + LimitedSeqInStream_Init(&limitedInStream); + limitedInStream.limit = me->props.blockSize; + + if (inStream) + { + UInt64 expected = (UInt64)(Int64)-1; + // inStream version works only in one thread. So we use CLzma2Enc::expectedDataSize + if (me->expectedDataSize != (UInt64)(Int64)-1 + && me->expectedDataSize >= unpackTotal) + expected = me->expectedDataSize - unpackTotal; + if (me->props.blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + && expected > me->props.blockSize) + expected = (size_t)me->props.blockSize; + + LzmaEnc_SetDataSize(p->enc, expected); + + RINOK(LzmaEnc_PrepareForLzma2(p->enc, + &limitedInStream.vt, + LZMA2_KEEP_WINDOW_SIZE, + me->alloc, + me->allocBig)); + } + else + { + inSizeCur = inDataSize - (size_t)unpackTotal; + if (me->props.blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + && inSizeCur > me->props.blockSize) + inSizeCur = (size_t)me->props.blockSize; + + // LzmaEnc_SetDataSize(p->enc, inSizeCur); + + RINOK(LzmaEnc_MemPrepare(p->enc, + inData + (size_t)unpackTotal, inSizeCur, + LZMA2_KEEP_WINDOW_SIZE, + me->alloc, + me->allocBig)); + } + + for (;;) + { + size_t packSize = LZMA2_CHUNK_SIZE_COMPRESSED_MAX; + if (outBuf) + packSize = outLim - (size_t)packTotal; + + res = Lzma2EncInt_EncodeSubblock(p, + outBuf ? outBuf + (size_t)packTotal : me->tempBufLzma, &packSize, + outBuf ? NULL : outStream); + + if (res != SZ_OK) + break; + + packTotal += packSize; + if (outBuf) + *outBufSize = (size_t)packTotal; + + res = Progress(progress, unpackTotal + p->srcPos, packTotal); + if (res != SZ_OK) + break; + + /* + if (LzmaEnc_GetNumAvailableBytes(p->enc) == 0) + break; + */ + + if (packSize == 0) + break; + } + + LzmaEnc_Finish(p->enc); + + unpackTotal += p->srcPos; + + RINOK(res); + + if (p->srcPos != (inStream ? limitedInStream.processed : inSizeCur)) + return SZ_ERROR_FAIL; + + if (inStream ? limitedInStream.finished : (unpackTotal == inDataSize)) + { + if (finished) + { + if (outBuf) + { + const size_t destPos = *outBufSize; + if (destPos >= outLim) + return SZ_ERROR_OUTPUT_EOF; + outBuf[destPos] = LZMA2_CONTROL_EOF; // 0 + *outBufSize = destPos + 1; + } + else + { + const Byte b = LZMA2_CONTROL_EOF; // 0; + if (ISeqOutStream_Write(outStream, &b, 1) != 1) + return SZ_ERROR_WRITE; + } + } + return SZ_OK; + } + } +} + + + +#ifndef _7ZIP_ST + +static SRes Lzma2Enc_MtCallback_Code(void *pp, unsigned coderIndex, unsigned outBufIndex, + const Byte *src, size_t srcSize, int finished) +{ + CLzma2Enc *me = (CLzma2Enc *)pp; + size_t destSize = me->outBufSize; + SRes res; + CMtProgressThunk progressThunk; + + Byte *dest = me->outBufs[outBufIndex]; + + me->outBufsDataSizes[outBufIndex] = 0; + + if (!dest) + { + dest = (Byte *)ISzAlloc_Alloc(me->alloc, me->outBufSize); + if (!dest) + return SZ_ERROR_MEM; + me->outBufs[outBufIndex] = dest; + } + + MtProgressThunk_CreateVTable(&progressThunk); + progressThunk.mtProgress = &me->mtCoder.mtProgress; + progressThunk.inSize = 0; + progressThunk.outSize = 0; + + res = Lzma2Enc_EncodeMt1(me, + &me->coders[coderIndex], + NULL, dest, &destSize, + NULL, src, srcSize, + finished, + &progressThunk.vt); + + me->outBufsDataSizes[outBufIndex] = destSize; + + return res; +} + + +static SRes Lzma2Enc_MtCallback_Write(void *pp, unsigned outBufIndex) +{ + CLzma2Enc *me = (CLzma2Enc *)pp; + size_t size = me->outBufsDataSizes[outBufIndex]; + const Byte *data = me->outBufs[outBufIndex]; + + if (me->outStream) + return ISeqOutStream_Write(me->outStream, data, size) == size ? SZ_OK : SZ_ERROR_WRITE; + + if (size > me->outBuf_Rem) + return SZ_ERROR_OUTPUT_EOF; + memcpy(me->outBuf, data, size); + me->outBuf_Rem -= size; + me->outBuf += size; + return SZ_OK; +} + +#endif + + + +SRes Lzma2Enc_Encode2(CLzma2EncHandle pp, + ISeqOutStream *outStream, + Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + const Byte *inData, size_t inDataSize, + ICompressProgress *progress) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + + if (inStream && inData) + return SZ_ERROR_PARAM; + + if (outStream && outBuf) + return SZ_ERROR_PARAM; + + { + unsigned i; + for (i = 0; i < MTCODER__THREADS_MAX; i++) + p->coders[i].propsAreSet = False; + } + + #ifndef _7ZIP_ST + + if (p->props.numBlockThreads_Reduced > 1) + { + IMtCoderCallback2 vt; + + if (!p->mtCoder_WasConstructed) + { + p->mtCoder_WasConstructed = True; + MtCoder_Construct(&p->mtCoder); + } + + vt.Code = Lzma2Enc_MtCallback_Code; + vt.Write = Lzma2Enc_MtCallback_Write; + + p->outStream = outStream; + p->outBuf = NULL; + p->outBuf_Rem = 0; + if (!outStream) + { + p->outBuf = outBuf; + p->outBuf_Rem = *outBufSize; + *outBufSize = 0; + } + + p->mtCoder.allocBig = p->allocBig; + p->mtCoder.progress = progress; + p->mtCoder.inStream = inStream; + p->mtCoder.inData = inData; + p->mtCoder.inDataSize = inDataSize; + p->mtCoder.mtCallback = &vt; + p->mtCoder.mtCallbackObject = p; + + p->mtCoder.blockSize = (size_t)p->props.blockSize; + if (p->mtCoder.blockSize != p->props.blockSize) + return SZ_ERROR_PARAM; /* SZ_ERROR_MEM */ + + { + size_t destBlockSize = p->mtCoder.blockSize + (p->mtCoder.blockSize >> 10) + 16; + if (destBlockSize < p->mtCoder.blockSize) + return SZ_ERROR_PARAM; + if (p->outBufSize != destBlockSize) + Lzma2Enc_FreeOutBufs(p); + p->outBufSize = destBlockSize; + } + + p->mtCoder.numThreadsMax = (unsigned)p->props.numBlockThreads_Max; + p->mtCoder.expectedDataSize = p->expectedDataSize; + + { + SRes res = MtCoder_Code(&p->mtCoder); + if (!outStream) + *outBufSize = (size_t)(p->outBuf - outBuf); + return res; + } + } + + #endif + + + return Lzma2Enc_EncodeMt1(p, + &p->coders[0], + outStream, outBuf, outBufSize, + inStream, inData, inDataSize, + True, /* finished */ + progress); +} diff --git a/3rdparty/lzma/src/Lzma86Dec.c b/3rdparty/lzma/src/Lzma86Dec.c new file mode 100644 index 0000000000..21031745ce --- /dev/null +++ b/3rdparty/lzma/src/Lzma86Dec.c @@ -0,0 +1,54 @@ +/* Lzma86Dec.c -- LZMA + x86 (BCJ) Filter Decoder +2016-05-16 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "Lzma86.h" + +#include "Alloc.h" +#include "Bra.h" +#include "LzmaDec.h" + +SRes Lzma86_GetUnpackSize(const Byte *src, SizeT srcLen, UInt64 *unpackSize) +{ + unsigned i; + if (srcLen < LZMA86_HEADER_SIZE) + return SZ_ERROR_INPUT_EOF; + *unpackSize = 0; + for (i = 0; i < sizeof(UInt64); i++) + *unpackSize += ((UInt64)src[LZMA86_SIZE_OFFSET + i]) << (8 * i); + return SZ_OK; +} + +SRes Lzma86_Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen) +{ + SRes res; + int useFilter; + SizeT inSizePure; + ELzmaStatus status; + + if (*srcLen < LZMA86_HEADER_SIZE) + return SZ_ERROR_INPUT_EOF; + + useFilter = src[0]; + + if (useFilter > 1) + { + *destLen = 0; + return SZ_ERROR_UNSUPPORTED; + } + + inSizePure = *srcLen - LZMA86_HEADER_SIZE; + res = LzmaDecode(dest, destLen, src + LZMA86_HEADER_SIZE, &inSizePure, + src + 1, LZMA_PROPS_SIZE, LZMA_FINISH_ANY, &status, &g_Alloc); + *srcLen = inSizePure + LZMA86_HEADER_SIZE; + if (res != SZ_OK) + return res; + if (useFilter == 1) + { + UInt32 x86State; + x86_Convert_Init(x86State); + x86_Convert(dest, *destLen, 0, &x86State, 0); + } + return SZ_OK; +} diff --git a/3rdparty/lzma/src/Lzma86Enc.c b/3rdparty/lzma/src/Lzma86Enc.c new file mode 100644 index 0000000000..14fcd65c67 --- /dev/null +++ b/3rdparty/lzma/src/Lzma86Enc.c @@ -0,0 +1,104 @@ +/* Lzma86Enc.c -- LZMA + x86 (BCJ) Filter Encoder +2018-07-04 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "Lzma86.h" + +#include "Alloc.h" +#include "Bra.h" +#include "LzmaEnc.h" + +int Lzma86_Encode(Byte *dest, size_t *destLen, const Byte *src, size_t srcLen, + int level, UInt32 dictSize, int filterMode) +{ + size_t outSize2 = *destLen; + Byte *filteredStream; + BoolInt useFilter; + int mainResult = SZ_ERROR_OUTPUT_EOF; + CLzmaEncProps props; + LzmaEncProps_Init(&props); + props.level = level; + props.dictSize = dictSize; + + *destLen = 0; + if (outSize2 < LZMA86_HEADER_SIZE) + return SZ_ERROR_OUTPUT_EOF; + + { + int i; + UInt64 t = srcLen; + for (i = 0; i < 8; i++, t >>= 8) + dest[LZMA86_SIZE_OFFSET + i] = (Byte)t; + } + + filteredStream = 0; + useFilter = (filterMode != SZ_FILTER_NO); + if (useFilter) + { + if (srcLen != 0) + { + filteredStream = (Byte *)MyAlloc(srcLen); + if (filteredStream == 0) + return SZ_ERROR_MEM; + memcpy(filteredStream, src, srcLen); + } + { + UInt32 x86State; + x86_Convert_Init(x86State); + x86_Convert(filteredStream, srcLen, 0, &x86State, 1); + } + } + + { + size_t minSize = 0; + BoolInt bestIsFiltered = False; + + /* passes for SZ_FILTER_AUTO: + 0 - BCJ + LZMA + 1 - LZMA + 2 - BCJ + LZMA agaian, if pass 0 (BCJ + LZMA) is better. + */ + int numPasses = (filterMode == SZ_FILTER_AUTO) ? 3 : 1; + + int i; + for (i = 0; i < numPasses; i++) + { + size_t outSizeProcessed = outSize2 - LZMA86_HEADER_SIZE; + size_t outPropsSize = 5; + SRes curRes; + BoolInt curModeIsFiltered = (numPasses > 1 && i == numPasses - 1); + if (curModeIsFiltered && !bestIsFiltered) + break; + if (useFilter && i == 0) + curModeIsFiltered = True; + + curRes = LzmaEncode(dest + LZMA86_HEADER_SIZE, &outSizeProcessed, + curModeIsFiltered ? filteredStream : src, srcLen, + &props, dest + 1, &outPropsSize, 0, + NULL, &g_Alloc, &g_Alloc); + + if (curRes != SZ_ERROR_OUTPUT_EOF) + { + if (curRes != SZ_OK) + { + mainResult = curRes; + break; + } + if (outSizeProcessed <= minSize || mainResult != SZ_OK) + { + minSize = outSizeProcessed; + bestIsFiltered = curModeIsFiltered; + mainResult = SZ_OK; + } + } + } + dest[0] = (Byte)(bestIsFiltered ? 1 : 0); + *destLen = LZMA86_HEADER_SIZE + minSize; + } + if (useFilter) + MyFree(filteredStream); + return mainResult; +} diff --git a/3rdparty/lzma/src/LzmaDec.c b/3rdparty/lzma/src/LzmaDec.c new file mode 100644 index 0000000000..d6742e5af8 --- /dev/null +++ b/3rdparty/lzma/src/LzmaDec.c @@ -0,0 +1,1363 @@ +/* LzmaDec.c -- LZMA Decoder +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #include "CpuArch.h" */ +#include "LzmaDec.h" + +#define kNumTopBits 24 +#define kTopValue ((UInt32)1 << kNumTopBits) + +#define kNumBitModelTotalBits 11 +#define kBitModelTotal (1 << kNumBitModelTotalBits) + +#define RC_INIT_SIZE 5 + +#ifndef _LZMA_DEC_OPT + +#define kNumMoveBits 5 +#define NORMALIZE if (range < kTopValue) { range <<= 8; code = (code << 8) | (*buf++); } + +#define IF_BIT_0(p) ttt = *(p); NORMALIZE; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound) +#define UPDATE_0(p) range = bound; *(p) = (CLzmaProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); +#define UPDATE_1(p) range -= bound; code -= bound; *(p) = (CLzmaProb)(ttt - (ttt >> kNumMoveBits)); +#define GET_BIT2(p, i, A0, A1) IF_BIT_0(p) \ + { UPDATE_0(p); i = (i + i); A0; } else \ + { UPDATE_1(p); i = (i + i) + 1; A1; } + +#define TREE_GET_BIT(probs, i) { GET_BIT2(probs + i, i, ;, ;); } + +#define REV_BIT(p, i, A0, A1) IF_BIT_0(p + i) \ + { UPDATE_0(p + i); A0; } else \ + { UPDATE_1(p + i); A1; } +#define REV_BIT_VAR( p, i, m) REV_BIT(p, i, i += m; m += m, m += m; i += m; ) +#define REV_BIT_CONST(p, i, m) REV_BIT(p, i, i += m; , i += m * 2; ) +#define REV_BIT_LAST( p, i, m) REV_BIT(p, i, i -= m , ; ) + +#define TREE_DECODE(probs, limit, i) \ + { i = 1; do { TREE_GET_BIT(probs, i); } while (i < limit); i -= limit; } + +/* #define _LZMA_SIZE_OPT */ + +#ifdef _LZMA_SIZE_OPT +#define TREE_6_DECODE(probs, i) TREE_DECODE(probs, (1 << 6), i) +#else +#define TREE_6_DECODE(probs, i) \ + { i = 1; \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + TREE_GET_BIT(probs, i); \ + i -= 0x40; } +#endif + +#define NORMAL_LITER_DEC TREE_GET_BIT(prob, symbol) +#define MATCHED_LITER_DEC \ + matchByte += matchByte; \ + bit = offs; \ + offs &= matchByte; \ + probLit = prob + (offs + bit + symbol); \ + GET_BIT2(probLit, symbol, offs ^= bit; , ;) + +#endif // _LZMA_DEC_OPT + + +#define NORMALIZE_CHECK if (range < kTopValue) { if (buf >= bufLimit) return DUMMY_INPUT_EOF; range <<= 8; code = (code << 8) | (*buf++); } + +#define IF_BIT_0_CHECK(p) ttt = *(p); NORMALIZE_CHECK; bound = (range >> kNumBitModelTotalBits) * (UInt32)ttt; if (code < bound) +#define UPDATE_0_CHECK range = bound; +#define UPDATE_1_CHECK range -= bound; code -= bound; +#define GET_BIT2_CHECK(p, i, A0, A1) IF_BIT_0_CHECK(p) \ + { UPDATE_0_CHECK; i = (i + i); A0; } else \ + { UPDATE_1_CHECK; i = (i + i) + 1; A1; } +#define GET_BIT_CHECK(p, i) GET_BIT2_CHECK(p, i, ; , ;) +#define TREE_DECODE_CHECK(probs, limit, i) \ + { i = 1; do { GET_BIT_CHECK(probs + i, i) } while (i < limit); i -= limit; } + + +#define REV_BIT_CHECK(p, i, m) IF_BIT_0_CHECK(p + i) \ + { UPDATE_0_CHECK; i += m; m += m; } else \ + { UPDATE_1_CHECK; m += m; i += m; } + + +#define kNumPosBitsMax 4 +#define kNumPosStatesMax (1 << kNumPosBitsMax) + +#define kLenNumLowBits 3 +#define kLenNumLowSymbols (1 << kLenNumLowBits) +#define kLenNumHighBits 8 +#define kLenNumHighSymbols (1 << kLenNumHighBits) + +#define LenLow 0 +#define LenHigh (LenLow + 2 * (kNumPosStatesMax << kLenNumLowBits)) +#define kNumLenProbs (LenHigh + kLenNumHighSymbols) + +#define LenChoice LenLow +#define LenChoice2 (LenLow + (1 << kLenNumLowBits)) + +#define kNumStates 12 +#define kNumStates2 16 +#define kNumLitStates 7 + +#define kStartPosModelIndex 4 +#define kEndPosModelIndex 14 +#define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) + +#define kNumPosSlotBits 6 +#define kNumLenToPosStates 4 + +#define kNumAlignBits 4 +#define kAlignTableSize (1 << kNumAlignBits) + +#define kMatchMinLen 2 +#define kMatchSpecLenStart (kMatchMinLen + kLenNumLowSymbols * 2 + kLenNumHighSymbols) + +#define kMatchSpecLen_Error_Data (1 << 9) +#define kMatchSpecLen_Error_Fail (kMatchSpecLen_Error_Data - 1) + +/* External ASM code needs same CLzmaProb array layout. So don't change it. */ + +/* (probs_1664) is faster and better for code size at some platforms */ +/* +#ifdef MY_CPU_X86_OR_AMD64 +*/ +#define kStartOffset 1664 +#define GET_PROBS p->probs_1664 +/* +#define GET_PROBS p->probs + kStartOffset +#else +#define kStartOffset 0 +#define GET_PROBS p->probs +#endif +*/ + +#define SpecPos (-kStartOffset) +#define IsRep0Long (SpecPos + kNumFullDistances) +#define RepLenCoder (IsRep0Long + (kNumStates2 << kNumPosBitsMax)) +#define LenCoder (RepLenCoder + kNumLenProbs) +#define IsMatch (LenCoder + kNumLenProbs) +#define Align (IsMatch + (kNumStates2 << kNumPosBitsMax)) +#define IsRep (Align + kAlignTableSize) +#define IsRepG0 (IsRep + kNumStates) +#define IsRepG1 (IsRepG0 + kNumStates) +#define IsRepG2 (IsRepG1 + kNumStates) +#define PosSlot (IsRepG2 + kNumStates) +#define Literal (PosSlot + (kNumLenToPosStates << kNumPosSlotBits)) +#define NUM_BASE_PROBS (Literal + kStartOffset) + +#if Align != 0 && kStartOffset != 0 + #error Stop_Compiling_Bad_LZMA_kAlign +#endif + +#if NUM_BASE_PROBS != 1984 + #error Stop_Compiling_Bad_LZMA_PROBS +#endif + + +#define LZMA_LIT_SIZE 0x300 + +#define LzmaProps_GetNumProbs(p) (NUM_BASE_PROBS + ((UInt32)LZMA_LIT_SIZE << ((p)->lc + (p)->lp))) + + +#define CALC_POS_STATE(processedPos, pbMask) (((processedPos) & (pbMask)) << 4) +#define COMBINED_PS_STATE (posState + state) +#define GET_LEN_STATE (posState) + +#define LZMA_DIC_MIN (1 << 12) + +/* +p->remainLen : shows status of LZMA decoder: + < kMatchSpecLenStart : the number of bytes to be copied with (p->rep0) offset + = kMatchSpecLenStart : the LZMA stream was finished with end mark + = kMatchSpecLenStart + 1 : need init range coder + = kMatchSpecLenStart + 2 : need init range coder and state + = kMatchSpecLen_Error_Fail : Internal Code Failure + = kMatchSpecLen_Error_Data + [0 ... 273] : LZMA Data Error +*/ + +/* ---------- LZMA_DECODE_REAL ---------- */ +/* +LzmaDec_DecodeReal_3() can be implemented in external ASM file. +3 - is the code compatibility version of that function for check at link time. +*/ + +#define LZMA_DECODE_REAL LzmaDec_DecodeReal_3 + +/* +LZMA_DECODE_REAL() +In: + RangeCoder is normalized + if (p->dicPos == limit) + { + LzmaDec_TryDummy() was called before to exclude LITERAL and MATCH-REP cases. + So first symbol can be only MATCH-NON-REP. And if that MATCH-NON-REP symbol + is not END_OF_PAYALOAD_MARKER, then the function doesn't write any byte to dictionary, + the function returns SZ_OK, and the caller can use (p->remainLen) and (p->reps[0]) later. + } + +Processing: + The first LZMA symbol will be decoded in any case. + All main checks for limits are at the end of main loop, + It decodes additional LZMA-symbols while (p->buf < bufLimit && dicPos < limit), + RangeCoder is still without last normalization when (p->buf < bufLimit) is being checked. + But if (p->buf < bufLimit), the caller provided at least (LZMA_REQUIRED_INPUT_MAX + 1) bytes for + next iteration before limit (bufLimit + LZMA_REQUIRED_INPUT_MAX), + that is enough for worst case LZMA symbol with one additional RangeCoder normalization for one bit. + So that function never reads bufLimit [LZMA_REQUIRED_INPUT_MAX] byte. + +Out: + RangeCoder is normalized + Result: + SZ_OK - OK + p->remainLen: + < kMatchSpecLenStart : the number of bytes to be copied with (p->reps[0]) offset + = kMatchSpecLenStart : the LZMA stream was finished with end mark + + SZ_ERROR_DATA - error, when the MATCH-Symbol refers out of dictionary + p->remainLen : undefined + p->reps[*] : undefined +*/ + + +#ifdef _LZMA_DEC_OPT + +int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit); + +#else + +static +int MY_FAST_CALL LZMA_DECODE_REAL(CLzmaDec *p, SizeT limit, const Byte *bufLimit) +{ + CLzmaProb *probs = GET_PROBS; + unsigned state = (unsigned)p->state; + UInt32 rep0 = p->reps[0], rep1 = p->reps[1], rep2 = p->reps[2], rep3 = p->reps[3]; + unsigned pbMask = ((unsigned)1 << (p->prop.pb)) - 1; + unsigned lc = p->prop.lc; + unsigned lpMask = ((unsigned)0x100 << p->prop.lp) - ((unsigned)0x100 >> lc); + + Byte *dic = p->dic; + SizeT dicBufSize = p->dicBufSize; + SizeT dicPos = p->dicPos; + + UInt32 processedPos = p->processedPos; + UInt32 checkDicSize = p->checkDicSize; + unsigned len = 0; + + const Byte *buf = p->buf; + UInt32 range = p->range; + UInt32 code = p->code; + + do + { + CLzmaProb *prob; + UInt32 bound; + unsigned ttt; + unsigned posState = CALC_POS_STATE(processedPos, pbMask); + + prob = probs + IsMatch + COMBINED_PS_STATE; + IF_BIT_0(prob) + { + unsigned symbol; + UPDATE_0(prob); + prob = probs + Literal; + if (processedPos != 0 || checkDicSize != 0) + prob += (UInt32)3 * ((((processedPos << 8) + dic[(dicPos == 0 ? dicBufSize : dicPos) - 1]) & lpMask) << lc); + processedPos++; + + if (state < kNumLitStates) + { + state -= (state < 4) ? state : 3; + symbol = 1; + #ifdef _LZMA_SIZE_OPT + do { NORMAL_LITER_DEC } while (symbol < 0x100); + #else + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + NORMAL_LITER_DEC + #endif + } + else + { + unsigned matchByte = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + unsigned offs = 0x100; + state -= (state < 10) ? 3 : 6; + symbol = 1; + #ifdef _LZMA_SIZE_OPT + do + { + unsigned bit; + CLzmaProb *probLit; + MATCHED_LITER_DEC + } + while (symbol < 0x100); + #else + { + unsigned bit; + CLzmaProb *probLit; + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + MATCHED_LITER_DEC + } + #endif + } + + dic[dicPos++] = (Byte)symbol; + continue; + } + + { + UPDATE_1(prob); + prob = probs + IsRep + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + state += kNumStates; + prob = probs + LenCoder; + } + else + { + UPDATE_1(prob); + prob = probs + IsRepG0 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + prob = probs + IsRep0Long + COMBINED_PS_STATE; + IF_BIT_0(prob) + { + UPDATE_0(prob); + + // that case was checked before with kBadRepCode + // if (checkDicSize == 0 && processedPos == 0) { len = kMatchSpecLen_Error_Data + 1; break; } + // The caller doesn't allow (dicPos == limit) case here + // so we don't need the following check: + // if (dicPos == limit) { state = state < kNumLitStates ? 9 : 11; len = 1; break; } + + dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + dicPos++; + processedPos++; + state = state < kNumLitStates ? 9 : 11; + continue; + } + UPDATE_1(prob); + } + else + { + UInt32 distance; + UPDATE_1(prob); + prob = probs + IsRepG1 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + distance = rep1; + } + else + { + UPDATE_1(prob); + prob = probs + IsRepG2 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + distance = rep2; + } + else + { + UPDATE_1(prob); + distance = rep3; + rep3 = rep2; + } + rep2 = rep1; + } + rep1 = rep0; + rep0 = distance; + } + state = state < kNumLitStates ? 8 : 11; + prob = probs + RepLenCoder; + } + + #ifdef _LZMA_SIZE_OPT + { + unsigned lim, offset; + CLzmaProb *probLen = prob + LenChoice; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE; + offset = 0; + lim = (1 << kLenNumLowBits); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenChoice2; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); + offset = kLenNumLowSymbols; + lim = (1 << kLenNumLowBits); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenHigh; + offset = kLenNumLowSymbols * 2; + lim = (1 << kLenNumHighBits); + } + } + TREE_DECODE(probLen, lim, len); + len += offset; + } + #else + { + CLzmaProb *probLen = prob + LenChoice; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE; + len = 1; + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + len -= 8; + } + else + { + UPDATE_1(probLen); + probLen = prob + LenChoice2; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); + len = 1; + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + TREE_GET_BIT(probLen, len); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenHigh; + TREE_DECODE(probLen, (1 << kLenNumHighBits), len); + len += kLenNumLowSymbols * 2; + } + } + } + #endif + + if (state >= kNumStates) + { + UInt32 distance; + prob = probs + PosSlot + + ((len < kNumLenToPosStates ? len : kNumLenToPosStates - 1) << kNumPosSlotBits); + TREE_6_DECODE(prob, distance); + if (distance >= kStartPosModelIndex) + { + unsigned posSlot = (unsigned)distance; + unsigned numDirectBits = (unsigned)(((distance >> 1) - 1)); + distance = (2 | (distance & 1)); + if (posSlot < kEndPosModelIndex) + { + distance <<= numDirectBits; + prob = probs + SpecPos; + { + UInt32 m = 1; + distance++; + do + { + REV_BIT_VAR(prob, distance, m); + } + while (--numDirectBits); + distance -= m; + } + } + else + { + numDirectBits -= kNumAlignBits; + do + { + NORMALIZE + range >>= 1; + + { + UInt32 t; + code -= range; + t = (0 - ((UInt32)code >> 31)); /* (UInt32)((Int32)code >> 31) */ + distance = (distance << 1) + (t + 1); + code += range & t; + } + /* + distance <<= 1; + if (code >= range) + { + code -= range; + distance |= 1; + } + */ + } + while (--numDirectBits); + prob = probs + Align; + distance <<= kNumAlignBits; + { + unsigned i = 1; + REV_BIT_CONST(prob, i, 1); + REV_BIT_CONST(prob, i, 2); + REV_BIT_CONST(prob, i, 4); + REV_BIT_LAST (prob, i, 8); + distance |= i; + } + if (distance == (UInt32)0xFFFFFFFF) + { + len = kMatchSpecLenStart; + state -= kNumStates; + break; + } + } + } + + rep3 = rep2; + rep2 = rep1; + rep1 = rep0; + rep0 = distance + 1; + state = (state < kNumStates + kNumLitStates) ? kNumLitStates : kNumLitStates + 3; + if (distance >= (checkDicSize == 0 ? processedPos: checkDicSize)) + { + len += kMatchSpecLen_Error_Data + kMatchMinLen; + // len = kMatchSpecLen_Error_Data; + // len += kMatchMinLen; + break; + } + } + + len += kMatchMinLen; + + { + SizeT rem; + unsigned curLen; + SizeT pos; + + if ((rem = limit - dicPos) == 0) + { + /* + We stop decoding and return SZ_OK, and we can resume decoding later. + Any error conditions can be tested later in caller code. + For more strict mode we can stop decoding with error + // len += kMatchSpecLen_Error_Data; + */ + break; + } + + curLen = ((rem < len) ? (unsigned)rem : len); + pos = dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0); + + processedPos += (UInt32)curLen; + + len -= curLen; + if (curLen <= dicBufSize - pos) + { + Byte *dest = dic + dicPos; + ptrdiff_t src = (ptrdiff_t)pos - (ptrdiff_t)dicPos; + const Byte *lim = dest + curLen; + dicPos += (SizeT)curLen; + do + *(dest) = (Byte)*(dest + src); + while (++dest != lim); + } + else + { + do + { + dic[dicPos++] = dic[pos]; + if (++pos == dicBufSize) + pos = 0; + } + while (--curLen != 0); + } + } + } + } + while (dicPos < limit && buf < bufLimit); + + NORMALIZE; + + p->buf = buf; + p->range = range; + p->code = code; + p->remainLen = (UInt32)len; // & (kMatchSpecLen_Error_Data - 1); // we can write real length for error matches too. + p->dicPos = dicPos; + p->processedPos = processedPos; + p->reps[0] = rep0; + p->reps[1] = rep1; + p->reps[2] = rep2; + p->reps[3] = rep3; + p->state = (UInt32)state; + if (len >= kMatchSpecLen_Error_Data) + return SZ_ERROR_DATA; + return SZ_OK; +} +#endif + + + +static void MY_FAST_CALL LzmaDec_WriteRem(CLzmaDec *p, SizeT limit) +{ + unsigned len = (unsigned)p->remainLen; + if (len == 0 /* || len >= kMatchSpecLenStart */) + return; + { + SizeT dicPos = p->dicPos; + Byte *dic; + SizeT dicBufSize; + SizeT rep0; /* we use SizeT to avoid the BUG of VC14 for AMD64 */ + { + SizeT rem = limit - dicPos; + if (rem < len) + { + len = (unsigned)(rem); + if (len == 0) + return; + } + } + + if (p->checkDicSize == 0 && p->prop.dicSize - p->processedPos <= len) + p->checkDicSize = p->prop.dicSize; + + p->processedPos += (UInt32)len; + p->remainLen -= (UInt32)len; + dic = p->dic; + rep0 = p->reps[0]; + dicBufSize = p->dicBufSize; + do + { + dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + dicPos++; + } + while (--len); + p->dicPos = dicPos; + } +} + + +/* +At staring of new stream we have one of the following symbols: + - Literal - is allowed + - Non-Rep-Match - is allowed only if it's end marker symbol + - Rep-Match - is not allowed +We use early check of (RangeCoder:Code) over kBadRepCode to simplify main decoding code +*/ + +#define kRange0 0xFFFFFFFF +#define kBound0 ((kRange0 >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1)) +#define kBadRepCode (kBound0 + (((kRange0 - kBound0) >> kNumBitModelTotalBits) << (kNumBitModelTotalBits - 1))) +#if kBadRepCode != (0xC0000000 - 0x400) + #error Stop_Compiling_Bad_LZMA_Check +#endif + + +/* +LzmaDec_DecodeReal2(): + It calls LZMA_DECODE_REAL() and it adjusts limit according (p->checkDicSize). + +We correct (p->checkDicSize) after LZMA_DECODE_REAL() and in LzmaDec_WriteRem(), +and we support the following state of (p->checkDicSize): + if (total_processed < p->prop.dicSize) then + { + (total_processed == p->processedPos) + (p->checkDicSize == 0) + } + else + (p->checkDicSize == p->prop.dicSize) +*/ + +static int MY_FAST_CALL LzmaDec_DecodeReal2(CLzmaDec *p, SizeT limit, const Byte *bufLimit) +{ + if (p->checkDicSize == 0) + { + UInt32 rem = p->prop.dicSize - p->processedPos; + if (limit - p->dicPos > rem) + limit = p->dicPos + rem; + } + { + int res = LZMA_DECODE_REAL(p, limit, bufLimit); + if (p->checkDicSize == 0 && p->processedPos >= p->prop.dicSize) + p->checkDicSize = p->prop.dicSize; + return res; + } +} + + + +typedef enum +{ + DUMMY_INPUT_EOF, /* need more input data */ + DUMMY_LIT, + DUMMY_MATCH, + DUMMY_REP +} ELzmaDummy; + + +#define IS_DUMMY_END_MARKER_POSSIBLE(dummyRes) ((dummyRes) == DUMMY_MATCH) + +static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, const Byte **bufOut) +{ + UInt32 range = p->range; + UInt32 code = p->code; + const Byte *bufLimit = *bufOut; + const CLzmaProb *probs = GET_PROBS; + unsigned state = (unsigned)p->state; + ELzmaDummy res; + + for (;;) + { + const CLzmaProb *prob; + UInt32 bound; + unsigned ttt; + unsigned posState = CALC_POS_STATE(p->processedPos, ((unsigned)1 << p->prop.pb) - 1); + + prob = probs + IsMatch + COMBINED_PS_STATE; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK + + prob = probs + Literal; + if (p->checkDicSize != 0 || p->processedPos != 0) + prob += ((UInt32)LZMA_LIT_SIZE * + ((((p->processedPos) & (((unsigned)1 << (p->prop.lp)) - 1)) << p->prop.lc) + + ((unsigned)p->dic[(p->dicPos == 0 ? p->dicBufSize : p->dicPos) - 1] >> (8 - p->prop.lc)))); + + if (state < kNumLitStates) + { + unsigned symbol = 1; + do { GET_BIT_CHECK(prob + symbol, symbol) } while (symbol < 0x100); + } + else + { + unsigned matchByte = p->dic[p->dicPos - p->reps[0] + + (p->dicPos < p->reps[0] ? p->dicBufSize : 0)]; + unsigned offs = 0x100; + unsigned symbol = 1; + do + { + unsigned bit; + const CLzmaProb *probLit; + matchByte += matchByte; + bit = offs; + offs &= matchByte; + probLit = prob + (offs + bit + symbol); + GET_BIT2_CHECK(probLit, symbol, offs ^= bit; , ; ) + } + while (symbol < 0x100); + } + res = DUMMY_LIT; + } + else + { + unsigned len; + UPDATE_1_CHECK; + + prob = probs + IsRep + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + state = 0; + prob = probs + LenCoder; + res = DUMMY_MATCH; + } + else + { + UPDATE_1_CHECK; + res = DUMMY_REP; + prob = probs + IsRepG0 + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + prob = probs + IsRep0Long + COMBINED_PS_STATE; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + break; + } + else + { + UPDATE_1_CHECK; + } + } + else + { + UPDATE_1_CHECK; + prob = probs + IsRepG1 + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + } + else + { + UPDATE_1_CHECK; + prob = probs + IsRepG2 + state; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + } + else + { + UPDATE_1_CHECK; + } + } + } + state = kNumStates; + prob = probs + RepLenCoder; + } + { + unsigned limit, offset; + const CLzmaProb *probLen = prob + LenChoice; + IF_BIT_0_CHECK(probLen) + { + UPDATE_0_CHECK; + probLen = prob + LenLow + GET_LEN_STATE; + offset = 0; + limit = 1 << kLenNumLowBits; + } + else + { + UPDATE_1_CHECK; + probLen = prob + LenChoice2; + IF_BIT_0_CHECK(probLen) + { + UPDATE_0_CHECK; + probLen = prob + LenLow + GET_LEN_STATE + (1 << kLenNumLowBits); + offset = kLenNumLowSymbols; + limit = 1 << kLenNumLowBits; + } + else + { + UPDATE_1_CHECK; + probLen = prob + LenHigh; + offset = kLenNumLowSymbols * 2; + limit = 1 << kLenNumHighBits; + } + } + TREE_DECODE_CHECK(probLen, limit, len); + len += offset; + } + + if (state < 4) + { + unsigned posSlot; + prob = probs + PosSlot + + ((len < kNumLenToPosStates - 1 ? len : kNumLenToPosStates - 1) << + kNumPosSlotBits); + TREE_DECODE_CHECK(prob, 1 << kNumPosSlotBits, posSlot); + if (posSlot >= kStartPosModelIndex) + { + unsigned numDirectBits = ((posSlot >> 1) - 1); + + if (posSlot < kEndPosModelIndex) + { + prob = probs + SpecPos + ((2 | (posSlot & 1)) << numDirectBits); + } + else + { + numDirectBits -= kNumAlignBits; + do + { + NORMALIZE_CHECK + range >>= 1; + code -= range & (((code - range) >> 31) - 1); + /* if (code >= range) code -= range; */ + } + while (--numDirectBits); + prob = probs + Align; + numDirectBits = kNumAlignBits; + } + { + unsigned i = 1; + unsigned m = 1; + do + { + REV_BIT_CHECK(prob, i, m); + } + while (--numDirectBits); + } + } + } + } + break; + } + NORMALIZE_CHECK; + + *bufOut = buf; + return res; +} + +void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState); +void LzmaDec_InitDicAndState(CLzmaDec *p, BoolInt initDic, BoolInt initState) +{ + p->remainLen = kMatchSpecLenStart + 1; + p->tempBufSize = 0; + + if (initDic) + { + p->processedPos = 0; + p->checkDicSize = 0; + p->remainLen = kMatchSpecLenStart + 2; + } + if (initState) + p->remainLen = kMatchSpecLenStart + 2; +} + +void LzmaDec_Init(CLzmaDec *p) +{ + p->dicPos = 0; + LzmaDec_InitDicAndState(p, True, True); +} + + +/* +LZMA supports optional end_marker. +So the decoder can lookahead for one additional LZMA-Symbol to check end_marker. +That additional LZMA-Symbol can require up to LZMA_REQUIRED_INPUT_MAX bytes in input stream. +When the decoder reaches dicLimit, it looks (finishMode) parameter: + if (finishMode == LZMA_FINISH_ANY), the decoder doesn't lookahead + if (finishMode != LZMA_FINISH_ANY), the decoder lookahead, if end_marker is possible for current position + +When the decoder lookahead, and the lookahead symbol is not end_marker, we have two ways: + 1) Strict mode (default) : the decoder returns SZ_ERROR_DATA. + 2) The relaxed mode (alternative mode) : we could return SZ_OK, and the caller + must check (status) value. The caller can show the error, + if the end of stream is expected, and the (status) is noit + LZMA_STATUS_FINISHED_WITH_MARK or LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK. +*/ + + +#define RETURN__NOT_FINISHED__FOR_FINISH \ + *status = LZMA_STATUS_NOT_FINISHED; \ + return SZ_ERROR_DATA; // for strict mode + // return SZ_OK; // for relaxed mode + + +SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, const Byte *src, SizeT *srcLen, + ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT inSize = *srcLen; + (*srcLen) = 0; + *status = LZMA_STATUS_NOT_SPECIFIED; + + if (p->remainLen > kMatchSpecLenStart) + { + if (p->remainLen > kMatchSpecLenStart + 2) + return p->remainLen == kMatchSpecLen_Error_Fail ? SZ_ERROR_FAIL : SZ_ERROR_DATA; + + for (; inSize > 0 && p->tempBufSize < RC_INIT_SIZE; (*srcLen)++, inSize--) + p->tempBuf[p->tempBufSize++] = *src++; + if (p->tempBufSize != 0 && p->tempBuf[0] != 0) + return SZ_ERROR_DATA; + if (p->tempBufSize < RC_INIT_SIZE) + { + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + p->code = + ((UInt32)p->tempBuf[1] << 24) + | ((UInt32)p->tempBuf[2] << 16) + | ((UInt32)p->tempBuf[3] << 8) + | ((UInt32)p->tempBuf[4]); + + if (p->checkDicSize == 0 + && p->processedPos == 0 + && p->code >= kBadRepCode) + return SZ_ERROR_DATA; + + p->range = 0xFFFFFFFF; + p->tempBufSize = 0; + + if (p->remainLen > kMatchSpecLenStart + 1) + { + SizeT numProbs = LzmaProps_GetNumProbs(&p->prop); + SizeT i; + CLzmaProb *probs = p->probs; + for (i = 0; i < numProbs; i++) + probs[i] = kBitModelTotal >> 1; + p->reps[0] = p->reps[1] = p->reps[2] = p->reps[3] = 1; + p->state = 0; + } + + p->remainLen = 0; + } + + for (;;) + { + if (p->remainLen == kMatchSpecLenStart) + { + if (p->code != 0) + return SZ_ERROR_DATA; + *status = LZMA_STATUS_FINISHED_WITH_MARK; + return SZ_OK; + } + + LzmaDec_WriteRem(p, dicLimit); + + { + // (p->remainLen == 0 || p->dicPos == dicLimit) + + int checkEndMarkNow = 0; + + if (p->dicPos >= dicLimit) + { + if (p->remainLen == 0 && p->code == 0) + { + *status = LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK; + return SZ_OK; + } + if (finishMode == LZMA_FINISH_ANY) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_OK; + } + if (p->remainLen != 0) + { + RETURN__NOT_FINISHED__FOR_FINISH; + } + checkEndMarkNow = 1; + } + + // (p->remainLen == 0) + + if (p->tempBufSize == 0) + { + const Byte *bufLimit; + int dummyProcessed = -1; + + if (inSize < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) + { + const Byte *bufOut = src + inSize; + + ELzmaDummy dummyRes = LzmaDec_TryDummy(p, src, &bufOut); + + if (dummyRes == DUMMY_INPUT_EOF) + { + size_t i; + if (inSize >= LZMA_REQUIRED_INPUT_MAX) + break; + (*srcLen) += inSize; + p->tempBufSize = (unsigned)inSize; + for (i = 0; i < inSize; i++) + p->tempBuf[i] = src[i]; + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + + dummyProcessed = (int)(bufOut - src); + if ((unsigned)dummyProcessed > LZMA_REQUIRED_INPUT_MAX) + break; + + if (checkEndMarkNow && !IS_DUMMY_END_MARKER_POSSIBLE(dummyRes)) + { + unsigned i; + (*srcLen) += (unsigned)dummyProcessed; + p->tempBufSize = (unsigned)dummyProcessed; + for (i = 0; i < (unsigned)dummyProcessed; i++) + p->tempBuf[i] = src[i]; + // p->remainLen = kMatchSpecLen_Error_Data; + RETURN__NOT_FINISHED__FOR_FINISH; + } + + bufLimit = src; + // we will decode only one iteration + } + else + bufLimit = src + inSize - LZMA_REQUIRED_INPUT_MAX; + + p->buf = src; + + { + int res = LzmaDec_DecodeReal2(p, dicLimit, bufLimit); + + SizeT processed = (SizeT)(p->buf - src); + + if (dummyProcessed < 0) + { + if (processed > inSize) + break; + } + else if ((unsigned)dummyProcessed != processed) + break; + + src += processed; + inSize -= processed; + (*srcLen) += processed; + + if (res != SZ_OK) + { + p->remainLen = kMatchSpecLen_Error_Data; + return SZ_ERROR_DATA; + } + } + continue; + } + + { + // we have some data in (p->tempBuf) + // in strict mode: tempBufSize is not enough for one Symbol decoding. + // in relaxed mode: tempBufSize not larger than required for one Symbol decoding. + + unsigned rem = p->tempBufSize; + unsigned ahead = 0; + int dummyProcessed = -1; + + while (rem < LZMA_REQUIRED_INPUT_MAX && ahead < inSize) + p->tempBuf[rem++] = src[ahead++]; + + // ahead - the size of new data copied from (src) to (p->tempBuf) + // rem - the size of temp buffer including new data from (src) + + if (rem < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) + { + const Byte *bufOut = p->tempBuf + rem; + + ELzmaDummy dummyRes = LzmaDec_TryDummy(p, p->tempBuf, &bufOut); + + if (dummyRes == DUMMY_INPUT_EOF) + { + if (rem >= LZMA_REQUIRED_INPUT_MAX) + break; + p->tempBufSize = rem; + (*srcLen) += (SizeT)ahead; + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + + dummyProcessed = (int)(bufOut - p->tempBuf); + + if ((unsigned)dummyProcessed < p->tempBufSize) + break; + + if (checkEndMarkNow && !IS_DUMMY_END_MARKER_POSSIBLE(dummyRes)) + { + (*srcLen) += (unsigned)dummyProcessed - p->tempBufSize; + p->tempBufSize = (unsigned)dummyProcessed; + // p->remainLen = kMatchSpecLen_Error_Data; + RETURN__NOT_FINISHED__FOR_FINISH; + } + } + + p->buf = p->tempBuf; + + { + // we decode one symbol from (p->tempBuf) here, so the (bufLimit) is equal to (p->buf) + int res = LzmaDec_DecodeReal2(p, dicLimit, p->buf); + + SizeT processed = (SizeT)(p->buf - p->tempBuf); + rem = p->tempBufSize; + + if (dummyProcessed < 0) + { + if (processed > LZMA_REQUIRED_INPUT_MAX) + break; + if (processed < rem) + break; + } + else if ((unsigned)dummyProcessed != processed) + break; + + processed -= rem; + + src += processed; + inSize -= processed; + (*srcLen) += processed; + p->tempBufSize = 0; + + if (res != SZ_OK) + { + p->remainLen = kMatchSpecLen_Error_Data; + return SZ_ERROR_DATA; + } + } + } + } + } + + /* Some unexpected error: internal error of code, memory corruption or hardware failure */ + p->remainLen = kMatchSpecLen_Error_Fail; + return SZ_ERROR_FAIL; +} + + + +SRes LzmaDec_DecodeToBuf(CLzmaDec *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT outSize = *destLen; + SizeT inSize = *srcLen; + *srcLen = *destLen = 0; + for (;;) + { + SizeT inSizeCur = inSize, outSizeCur, dicPos; + ELzmaFinishMode curFinishMode; + SRes res; + if (p->dicPos == p->dicBufSize) + p->dicPos = 0; + dicPos = p->dicPos; + if (outSize > p->dicBufSize - dicPos) + { + outSizeCur = p->dicBufSize; + curFinishMode = LZMA_FINISH_ANY; + } + else + { + outSizeCur = dicPos + outSize; + curFinishMode = finishMode; + } + + res = LzmaDec_DecodeToDic(p, outSizeCur, src, &inSizeCur, curFinishMode, status); + src += inSizeCur; + inSize -= inSizeCur; + *srcLen += inSizeCur; + outSizeCur = p->dicPos - dicPos; + memcpy(dest, p->dic + dicPos, outSizeCur); + dest += outSizeCur; + outSize -= outSizeCur; + *destLen += outSizeCur; + if (res != 0) + return res; + if (outSizeCur == 0 || outSize == 0) + return SZ_OK; + } +} + +void LzmaDec_FreeProbs(CLzmaDec *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->probs); + p->probs = NULL; +} + +static void LzmaDec_FreeDict(CLzmaDec *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->dic); + p->dic = NULL; +} + +void LzmaDec_Free(CLzmaDec *p, ISzAllocPtr alloc) +{ + LzmaDec_FreeProbs(p, alloc); + LzmaDec_FreeDict(p, alloc); +} + +SRes LzmaProps_Decode(CLzmaProps *p, const Byte *data, unsigned size) +{ + UInt32 dicSize; + Byte d; + + if (size < LZMA_PROPS_SIZE) + return SZ_ERROR_UNSUPPORTED; + else + dicSize = data[1] | ((UInt32)data[2] << 8) | ((UInt32)data[3] << 16) | ((UInt32)data[4] << 24); + + if (dicSize < LZMA_DIC_MIN) + dicSize = LZMA_DIC_MIN; + p->dicSize = dicSize; + + d = data[0]; + if (d >= (9 * 5 * 5)) + return SZ_ERROR_UNSUPPORTED; + + p->lc = (Byte)(d % 9); + d /= 9; + p->pb = (Byte)(d / 5); + p->lp = (Byte)(d % 5); + + return SZ_OK; +} + +static SRes LzmaDec_AllocateProbs2(CLzmaDec *p, const CLzmaProps *propNew, ISzAllocPtr alloc) +{ + UInt32 numProbs = LzmaProps_GetNumProbs(propNew); + if (!p->probs || numProbs != p->numProbs) + { + LzmaDec_FreeProbs(p, alloc); + p->probs = (CLzmaProb *)ISzAlloc_Alloc(alloc, numProbs * sizeof(CLzmaProb)); + if (!p->probs) + return SZ_ERROR_MEM; + p->probs_1664 = p->probs + 1664; + p->numProbs = numProbs; + } + return SZ_OK; +} + +SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc) +{ + CLzmaProps propNew; + RINOK(LzmaProps_Decode(&propNew, props, propsSize)); + RINOK(LzmaDec_AllocateProbs2(p, &propNew, alloc)); + p->prop = propNew; + return SZ_OK; +} + +SRes LzmaDec_Allocate(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAllocPtr alloc) +{ + CLzmaProps propNew; + SizeT dicBufSize; + RINOK(LzmaProps_Decode(&propNew, props, propsSize)); + RINOK(LzmaDec_AllocateProbs2(p, &propNew, alloc)); + + { + UInt32 dictSize = propNew.dicSize; + SizeT mask = ((UInt32)1 << 12) - 1; + if (dictSize >= ((UInt32)1 << 30)) mask = ((UInt32)1 << 22) - 1; + else if (dictSize >= ((UInt32)1 << 22)) mask = ((UInt32)1 << 20) - 1;; + dicBufSize = ((SizeT)dictSize + mask) & ~mask; + if (dicBufSize < dictSize) + dicBufSize = dictSize; + } + + if (!p->dic || dicBufSize != p->dicBufSize) + { + LzmaDec_FreeDict(p, alloc); + p->dic = (Byte *)ISzAlloc_Alloc(alloc, dicBufSize); + if (!p->dic) + { + LzmaDec_FreeProbs(p, alloc); + return SZ_ERROR_MEM; + } + } + p->dicBufSize = dicBufSize; + p->prop = propNew; + return SZ_OK; +} + +SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, + ELzmaStatus *status, ISzAllocPtr alloc) +{ + CLzmaDec p; + SRes res; + SizeT outSize = *destLen, inSize = *srcLen; + *destLen = *srcLen = 0; + *status = LZMA_STATUS_NOT_SPECIFIED; + if (inSize < RC_INIT_SIZE) + return SZ_ERROR_INPUT_EOF; + LzmaDec_Construct(&p); + RINOK(LzmaDec_AllocateProbs(&p, propData, propSize, alloc)); + p.dic = dest; + p.dicBufSize = outSize; + LzmaDec_Init(&p); + *srcLen = inSize; + res = LzmaDec_DecodeToDic(&p, outSize, src, srcLen, finishMode, status); + *destLen = p.dicPos; + if (res == SZ_OK && *status == LZMA_STATUS_NEEDS_MORE_INPUT) + res = SZ_ERROR_INPUT_EOF; + LzmaDec_FreeProbs(&p, alloc); + return res; +} diff --git a/3rdparty/lzma/src/LzmaEnc.c b/3rdparty/lzma/src/LzmaEnc.c new file mode 100644 index 0000000000..b04a7b7b71 --- /dev/null +++ b/3rdparty/lzma/src/LzmaEnc.c @@ -0,0 +1,3165 @@ +/* LzmaEnc.c -- LZMA Encoder +2021-11-18: Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +/* #define SHOW_STAT */ +/* #define SHOW_STAT2 */ + +#if defined(SHOW_STAT) || defined(SHOW_STAT2) +#include +#endif + +#include "CpuArch.h" +#include "LzmaEnc.h" + +#include "LzFind.h" +#ifndef _7ZIP_ST +#include "LzFindMt.h" +#endif + +/* the following LzmaEnc_* declarations is internal LZMA interface for LZMA2 encoder */ + +SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, ISeqInStream *inStream, UInt32 keepWindowSize, + ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, + UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig); +SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, + Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize); +const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp); +void LzmaEnc_Finish(CLzmaEncHandle pp); +void LzmaEnc_SaveState(CLzmaEncHandle pp); +void LzmaEnc_RestoreState(CLzmaEncHandle pp); + +#ifdef SHOW_STAT +static unsigned g_STAT_OFFSET = 0; +#endif + +/* for good normalization speed we still reserve 256 MB before 4 GB range */ +#define kLzmaMaxHistorySize ((UInt32)15 << 28) + +#define kNumTopBits 24 +#define kTopValue ((UInt32)1 << kNumTopBits) + +#define kNumBitModelTotalBits 11 +#define kBitModelTotal (1 << kNumBitModelTotalBits) +#define kNumMoveBits 5 +#define kProbInitValue (kBitModelTotal >> 1) + +#define kNumMoveReducingBits 4 +#define kNumBitPriceShiftBits 4 +// #define kBitPrice (1 << kNumBitPriceShiftBits) + +#define REP_LEN_COUNT 64 + +void LzmaEncProps_Init(CLzmaEncProps *p) +{ + p->level = 5; + p->dictSize = p->mc = 0; + p->reduceSize = (UInt64)(Int64)-1; + p->lc = p->lp = p->pb = p->algo = p->fb = p->btMode = p->numHashBytes = p->numThreads = -1; + p->writeEndMark = 0; + p->affinity = 0; +} + +void LzmaEncProps_Normalize(CLzmaEncProps *p) +{ + int level = p->level; + if (level < 0) level = 5; + p->level = level; + + if (p->dictSize == 0) + p->dictSize = + ( level <= 3 ? ((UInt32)1 << (level * 2 + 16)) : + ( level <= 6 ? ((UInt32)1 << (level + 19)) : + ( level <= 7 ? ((UInt32)1 << 25) : ((UInt32)1 << 26) + ))); + + if (p->dictSize > p->reduceSize) + { + UInt32 v = (UInt32)p->reduceSize; + const UInt32 kReduceMin = ((UInt32)1 << 12); + if (v < kReduceMin) + v = kReduceMin; + if (p->dictSize > v) + p->dictSize = v; + } + + if (p->lc < 0) p->lc = 3; + if (p->lp < 0) p->lp = 0; + if (p->pb < 0) p->pb = 2; + + if (p->algo < 0) p->algo = (level < 5 ? 0 : 1); + if (p->fb < 0) p->fb = (level < 7 ? 32 : 64); + if (p->btMode < 0) p->btMode = (p->algo == 0 ? 0 : 1); + if (p->numHashBytes < 0) p->numHashBytes = (p->btMode ? 4 : 5); + if (p->mc == 0) p->mc = (16 + ((unsigned)p->fb >> 1)) >> (p->btMode ? 0 : 1); + + if (p->numThreads < 0) + p->numThreads = + #ifndef _7ZIP_ST + ((p->btMode && p->algo) ? 2 : 1); + #else + 1; + #endif +} + +UInt32 LzmaEncProps_GetDictSize(const CLzmaEncProps *props2) +{ + CLzmaEncProps props = *props2; + LzmaEncProps_Normalize(&props); + return props.dictSize; +} + + +/* +x86/x64: + +BSR: + IF (SRC == 0) ZF = 1, DEST is undefined; + AMD : DEST is unchanged; + IF (SRC != 0) ZF = 0; DEST is index of top non-zero bit + BSR is slow in some processors + +LZCNT: + IF (SRC == 0) CF = 1, DEST is size_in_bits_of_register(src) (32 or 64) + IF (SRC != 0) CF = 0, DEST = num_lead_zero_bits + IF (DEST == 0) ZF = 1; + +LZCNT works only in new processors starting from Haswell. +if LZCNT is not supported by processor, then it's executed as BSR. +LZCNT can be faster than BSR, if supported. +*/ + +// #define LZMA_LOG_BSR + +#if defined(MY_CPU_ARM_OR_ARM64) /* || defined(MY_CPU_X86_OR_AMD64) */ + + #if (defined(__clang__) && (__clang_major__ >= 6)) \ + || (defined(__GNUC__) && (__GNUC__ >= 6)) + #define LZMA_LOG_BSR + #elif defined(_MSC_VER) && (_MSC_VER >= 1300) + // #if defined(MY_CPU_ARM_OR_ARM64) + #define LZMA_LOG_BSR + // #endif + #endif +#endif + +// #include + +#ifdef LZMA_LOG_BSR + +#if defined(__clang__) \ + || defined(__GNUC__) + +/* + C code: : (30 - __builtin_clz(x)) + gcc9/gcc10 for x64 /x86 : 30 - (bsr(x) xor 31) + clang10 for x64 : 31 + (bsr(x) xor -32) +*/ + + #define MY_clz(x) ((unsigned)__builtin_clz(x)) + // __lzcnt32 + // __builtin_ia32_lzcnt_u32 + +#else // #if defined(_MSC_VER) + + #ifdef MY_CPU_ARM_OR_ARM64 + + #define MY_clz _CountLeadingZeros + + #else // if defined(MY_CPU_X86_OR_AMD64) + + // #define MY_clz __lzcnt // we can use lzcnt (unsupported by old CPU) + // _BitScanReverse code is not optimal for some MSVC compilers + #define BSR2_RET(pos, res) { unsigned long zz; _BitScanReverse(&zz, (pos)); zz--; \ + res = (zz + zz) + (pos >> zz); } + + #endif // MY_CPU_X86_OR_AMD64 + +#endif // _MSC_VER + + +#ifndef BSR2_RET + + #define BSR2_RET(pos, res) { unsigned zz = 30 - MY_clz(pos); \ + res = (zz + zz) + (pos >> zz); } + +#endif + + +unsigned GetPosSlot1(UInt32 pos); +unsigned GetPosSlot1(UInt32 pos) +{ + unsigned res; + BSR2_RET(pos, res); + return res; +} +#define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } +#define GetPosSlot(pos, res) { if (pos < 2) res = pos; else BSR2_RET(pos, res); } + + +#else // ! LZMA_LOG_BSR + +#define kNumLogBits (11 + sizeof(size_t) / 8 * 3) + +#define kDicLogSizeMaxCompress ((kNumLogBits - 1) * 2 + 7) + +static void LzmaEnc_FastPosInit(Byte *g_FastPos) +{ + unsigned slot; + g_FastPos[0] = 0; + g_FastPos[1] = 1; + g_FastPos += 2; + + for (slot = 2; slot < kNumLogBits * 2; slot++) + { + size_t k = ((size_t)1 << ((slot >> 1) - 1)); + size_t j; + for (j = 0; j < k; j++) + g_FastPos[j] = (Byte)slot; + g_FastPos += k; + } +} + +/* we can use ((limit - pos) >> 31) only if (pos < ((UInt32)1 << 31)) */ +/* +#define BSR2_RET(pos, res) { unsigned zz = 6 + ((kNumLogBits - 1) & \ + (0 - (((((UInt32)1 << (kNumLogBits + 6)) - 1) - pos) >> 31))); \ + res = p->g_FastPos[pos >> zz] + (zz * 2); } +*/ + +/* +#define BSR2_RET(pos, res) { unsigned zz = 6 + ((kNumLogBits - 1) & \ + (0 - (((((UInt32)1 << (kNumLogBits)) - 1) - (pos >> 6)) >> 31))); \ + res = p->g_FastPos[pos >> zz] + (zz * 2); } +*/ + +#define BSR2_RET(pos, res) { unsigned zz = (pos < (1 << (kNumLogBits + 6))) ? 6 : 6 + kNumLogBits - 1; \ + res = p->g_FastPos[pos >> zz] + (zz * 2); } + +/* +#define BSR2_RET(pos, res) { res = (pos < (1 << (kNumLogBits + 6))) ? \ + p->g_FastPos[pos >> 6] + 12 : \ + p->g_FastPos[pos >> (6 + kNumLogBits - 1)] + (6 + (kNumLogBits - 1)) * 2; } +*/ + +#define GetPosSlot1(pos) p->g_FastPos[pos] +#define GetPosSlot2(pos, res) { BSR2_RET(pos, res); } +#define GetPosSlot(pos, res) { if (pos < kNumFullDistances) res = p->g_FastPos[pos & (kNumFullDistances - 1)]; else BSR2_RET(pos, res); } + +#endif // LZMA_LOG_BSR + + +#define LZMA_NUM_REPS 4 + +typedef UInt16 CState; +typedef UInt16 CExtra; + +typedef struct +{ + UInt32 price; + CState state; + CExtra extra; + // 0 : normal + // 1 : LIT : MATCH + // > 1 : MATCH (extra-1) : LIT : REP0 (len) + UInt32 len; + UInt32 dist; + UInt32 reps[LZMA_NUM_REPS]; +} COptimal; + + +// 18.06 +#define kNumOpts (1 << 11) +#define kPackReserve (kNumOpts * 8) +// #define kNumOpts (1 << 12) +// #define kPackReserve (1 + kNumOpts * 2) + +#define kNumLenToPosStates 4 +#define kNumPosSlotBits 6 +// #define kDicLogSizeMin 0 +#define kDicLogSizeMax 32 +#define kDistTableSizeMax (kDicLogSizeMax * 2) + +#define kNumAlignBits 4 +#define kAlignTableSize (1 << kNumAlignBits) +#define kAlignMask (kAlignTableSize - 1) + +#define kStartPosModelIndex 4 +#define kEndPosModelIndex 14 +#define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) + +typedef +#ifdef _LZMA_PROB32 + UInt32 +#else + UInt16 +#endif + CLzmaProb; + +#define LZMA_PB_MAX 4 +#define LZMA_LC_MAX 8 +#define LZMA_LP_MAX 4 + +#define LZMA_NUM_PB_STATES_MAX (1 << LZMA_PB_MAX) + +#define kLenNumLowBits 3 +#define kLenNumLowSymbols (1 << kLenNumLowBits) +#define kLenNumHighBits 8 +#define kLenNumHighSymbols (1 << kLenNumHighBits) +#define kLenNumSymbolsTotal (kLenNumLowSymbols * 2 + kLenNumHighSymbols) + +#define LZMA_MATCH_LEN_MIN 2 +#define LZMA_MATCH_LEN_MAX (LZMA_MATCH_LEN_MIN + kLenNumSymbolsTotal - 1) + +#define kNumStates 12 + + +typedef struct +{ + CLzmaProb low[LZMA_NUM_PB_STATES_MAX << (kLenNumLowBits + 1)]; + CLzmaProb high[kLenNumHighSymbols]; +} CLenEnc; + + +typedef struct +{ + unsigned tableSize; + UInt32 prices[LZMA_NUM_PB_STATES_MAX][kLenNumSymbolsTotal]; + // UInt32 prices1[LZMA_NUM_PB_STATES_MAX][kLenNumLowSymbols * 2]; + // UInt32 prices2[kLenNumSymbolsTotal]; +} CLenPriceEnc; + +#define GET_PRICE_LEN(p, posState, len) \ + ((p)->prices[posState][(size_t)(len) - LZMA_MATCH_LEN_MIN]) + +/* +#define GET_PRICE_LEN(p, posState, len) \ + ((p)->prices2[(size_t)(len) - 2] + ((p)->prices1[posState][((len) - 2) & (kLenNumLowSymbols * 2 - 1)] & (((len) - 2 - kLenNumLowSymbols * 2) >> 9))) +*/ + +typedef struct +{ + UInt32 range; + unsigned cache; + UInt64 low; + UInt64 cacheSize; + Byte *buf; + Byte *bufLim; + Byte *bufBase; + ISeqOutStream *outStream; + UInt64 processed; + SRes res; +} CRangeEnc; + + +typedef struct +{ + CLzmaProb *litProbs; + + unsigned state; + UInt32 reps[LZMA_NUM_REPS]; + + CLzmaProb posAlignEncoder[1 << kNumAlignBits]; + CLzmaProb isRep[kNumStates]; + CLzmaProb isRepG0[kNumStates]; + CLzmaProb isRepG1[kNumStates]; + CLzmaProb isRepG2[kNumStates]; + CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; + + CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; + CLzmaProb posEncoders[kNumFullDistances]; + + CLenEnc lenProbs; + CLenEnc repLenProbs; + +} CSaveState; + + +typedef UInt32 CProbPrice; + + +typedef struct +{ + void *matchFinderObj; + IMatchFinder2 matchFinder; + + unsigned optCur; + unsigned optEnd; + + unsigned longestMatchLen; + unsigned numPairs; + UInt32 numAvail; + + unsigned state; + unsigned numFastBytes; + unsigned additionalOffset; + UInt32 reps[LZMA_NUM_REPS]; + unsigned lpMask, pbMask; + CLzmaProb *litProbs; + CRangeEnc rc; + + UInt32 backRes; + + unsigned lc, lp, pb; + unsigned lclp; + + BoolInt fastMode; + BoolInt writeEndMark; + BoolInt finished; + BoolInt multiThread; + BoolInt needInit; + // BoolInt _maxMode; + + UInt64 nowPos64; + + unsigned matchPriceCount; + // unsigned alignPriceCount; + int repLenEncCounter; + + unsigned distTableSize; + + UInt32 dictSize; + SRes result; + + #ifndef _7ZIP_ST + BoolInt mtMode; + // begin of CMatchFinderMt is used in LZ thread + CMatchFinderMt matchFinderMt; + // end of CMatchFinderMt is used in BT and HASH threads + // #else + // CMatchFinder matchFinderBase; + #endif + CMatchFinder matchFinderBase; + + + // we suppose that we have 8-bytes alignment after CMatchFinder + + #ifndef _7ZIP_ST + Byte pad[128]; + #endif + + // LZ thread + CProbPrice ProbPrices[kBitModelTotal >> kNumMoveReducingBits]; + + // we want {len , dist} pairs to be 8-bytes aligned in matches array + UInt32 matches[LZMA_MATCH_LEN_MAX * 2 + 2]; + + // we want 8-bytes alignment here + UInt32 alignPrices[kAlignTableSize]; + UInt32 posSlotPrices[kNumLenToPosStates][kDistTableSizeMax]; + UInt32 distancesPrices[kNumLenToPosStates][kNumFullDistances]; + + CLzmaProb posAlignEncoder[1 << kNumAlignBits]; + CLzmaProb isRep[kNumStates]; + CLzmaProb isRepG0[kNumStates]; + CLzmaProb isRepG1[kNumStates]; + CLzmaProb isRepG2[kNumStates]; + CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; + CLzmaProb posEncoders[kNumFullDistances]; + + CLenEnc lenProbs; + CLenEnc repLenProbs; + + #ifndef LZMA_LOG_BSR + Byte g_FastPos[1 << kNumLogBits]; + #endif + + CLenPriceEnc lenEnc; + CLenPriceEnc repLenEnc; + + COptimal opt[kNumOpts]; + + CSaveState saveState; + + // BoolInt mf_Failure; + #ifndef _7ZIP_ST + Byte pad2[128]; + #endif +} CLzmaEnc; + + +#define MFB (p->matchFinderBase) +/* +#ifndef _7ZIP_ST +#define MFB (p->matchFinderMt.MatchFinder) +#endif +*/ + +#define COPY_ARR(dest, src, arr) memcpy(dest->arr, src->arr, sizeof(src->arr)); + +void LzmaEnc_SaveState(CLzmaEncHandle pp) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + CSaveState *dest = &p->saveState; + + dest->state = p->state; + + dest->lenProbs = p->lenProbs; + dest->repLenProbs = p->repLenProbs; + + COPY_ARR(dest, p, reps); + + COPY_ARR(dest, p, posAlignEncoder); + COPY_ARR(dest, p, isRep); + COPY_ARR(dest, p, isRepG0); + COPY_ARR(dest, p, isRepG1); + COPY_ARR(dest, p, isRepG2); + COPY_ARR(dest, p, isMatch); + COPY_ARR(dest, p, isRep0Long); + COPY_ARR(dest, p, posSlotEncoder); + COPY_ARR(dest, p, posEncoders); + + memcpy(dest->litProbs, p->litProbs, ((UInt32)0x300 << p->lclp) * sizeof(CLzmaProb)); +} + + +void LzmaEnc_RestoreState(CLzmaEncHandle pp) +{ + CLzmaEnc *dest = (CLzmaEnc *)pp; + const CSaveState *p = &dest->saveState; + + dest->state = p->state; + + dest->lenProbs = p->lenProbs; + dest->repLenProbs = p->repLenProbs; + + COPY_ARR(dest, p, reps); + + COPY_ARR(dest, p, posAlignEncoder); + COPY_ARR(dest, p, isRep); + COPY_ARR(dest, p, isRepG0); + COPY_ARR(dest, p, isRepG1); + COPY_ARR(dest, p, isRepG2); + COPY_ARR(dest, p, isMatch); + COPY_ARR(dest, p, isRep0Long); + COPY_ARR(dest, p, posSlotEncoder); + COPY_ARR(dest, p, posEncoders); + + memcpy(dest->litProbs, p->litProbs, ((UInt32)0x300 << dest->lclp) * sizeof(CLzmaProb)); +} + + + +SRes LzmaEnc_SetProps(CLzmaEncHandle pp, const CLzmaEncProps *props2) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + CLzmaEncProps props = *props2; + LzmaEncProps_Normalize(&props); + + if (props.lc > LZMA_LC_MAX + || props.lp > LZMA_LP_MAX + || props.pb > LZMA_PB_MAX) + return SZ_ERROR_PARAM; + + + if (props.dictSize > kLzmaMaxHistorySize) + props.dictSize = kLzmaMaxHistorySize; + + #ifndef LZMA_LOG_BSR + { + const UInt64 dict64 = props.dictSize; + if (dict64 > ((UInt64)1 << kDicLogSizeMaxCompress)) + return SZ_ERROR_PARAM; + } + #endif + + p->dictSize = props.dictSize; + { + unsigned fb = (unsigned)props.fb; + if (fb < 5) + fb = 5; + if (fb > LZMA_MATCH_LEN_MAX) + fb = LZMA_MATCH_LEN_MAX; + p->numFastBytes = fb; + } + p->lc = (unsigned)props.lc; + p->lp = (unsigned)props.lp; + p->pb = (unsigned)props.pb; + p->fastMode = (props.algo == 0); + // p->_maxMode = True; + MFB.btMode = (Byte)(props.btMode ? 1 : 0); + { + unsigned numHashBytes = 4; + if (props.btMode) + { + if (props.numHashBytes < 2) numHashBytes = 2; + else if (props.numHashBytes < 4) numHashBytes = (unsigned)props.numHashBytes; + } + if (props.numHashBytes >= 5) numHashBytes = 5; + + MFB.numHashBytes = numHashBytes; + } + + MFB.cutValue = props.mc; + + p->writeEndMark = (BoolInt)props.writeEndMark; + + #ifndef _7ZIP_ST + /* + if (newMultiThread != _multiThread) + { + ReleaseMatchFinder(); + _multiThread = newMultiThread; + } + */ + p->multiThread = (props.numThreads > 1); + p->matchFinderMt.btSync.affinity = + p->matchFinderMt.hashSync.affinity = props.affinity; + #endif + + return SZ_OK; +} + + +void LzmaEnc_SetDataSize(CLzmaEncHandle pp, UInt64 expectedDataSiize) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + MFB.expectedDataSize = expectedDataSiize; +} + + +#define kState_Start 0 +#define kState_LitAfterMatch 4 +#define kState_LitAfterRep 5 +#define kState_MatchAfterLit 7 +#define kState_RepAfterLit 8 + +static const Byte kLiteralNextStates[kNumStates] = {0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 4, 5}; +static const Byte kMatchNextStates[kNumStates] = {7, 7, 7, 7, 7, 7, 7, 10, 10, 10, 10, 10}; +static const Byte kRepNextStates[kNumStates] = {8, 8, 8, 8, 8, 8, 8, 11, 11, 11, 11, 11}; +static const Byte kShortRepNextStates[kNumStates]= {9, 9, 9, 9, 9, 9, 9, 11, 11, 11, 11, 11}; + +#define IsLitState(s) ((s) < 7) +#define GetLenToPosState2(len) (((len) < kNumLenToPosStates - 1) ? (len) : kNumLenToPosStates - 1) +#define GetLenToPosState(len) (((len) < kNumLenToPosStates + 1) ? (len) - 2 : kNumLenToPosStates - 1) + +#define kInfinityPrice (1 << 30) + +static void RangeEnc_Construct(CRangeEnc *p) +{ + p->outStream = NULL; + p->bufBase = NULL; +} + +#define RangeEnc_GetProcessed(p) ( (p)->processed + (size_t)((p)->buf - (p)->bufBase) + (p)->cacheSize) +#define RangeEnc_GetProcessed_sizet(p) ((size_t)(p)->processed + (size_t)((p)->buf - (p)->bufBase) + (size_t)(p)->cacheSize) + +#define RC_BUF_SIZE (1 << 16) + +static int RangeEnc_Alloc(CRangeEnc *p, ISzAllocPtr alloc) +{ + if (!p->bufBase) + { + p->bufBase = (Byte *)ISzAlloc_Alloc(alloc, RC_BUF_SIZE); + if (!p->bufBase) + return 0; + p->bufLim = p->bufBase + RC_BUF_SIZE; + } + return 1; +} + +static void RangeEnc_Free(CRangeEnc *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->bufBase); + p->bufBase = NULL; +} + +static void RangeEnc_Init(CRangeEnc *p) +{ + p->range = 0xFFFFFFFF; + p->cache = 0; + p->low = 0; + p->cacheSize = 0; + + p->buf = p->bufBase; + + p->processed = 0; + p->res = SZ_OK; +} + +MY_NO_INLINE static void RangeEnc_FlushStream(CRangeEnc *p) +{ + const size_t num = (size_t)(p->buf - p->bufBase); + if (p->res == SZ_OK) + { + if (num != ISeqOutStream_Write(p->outStream, p->bufBase, num)) + p->res = SZ_ERROR_WRITE; + } + p->processed += num; + p->buf = p->bufBase; +} + +MY_NO_INLINE static void MY_FAST_CALL RangeEnc_ShiftLow(CRangeEnc *p) +{ + UInt32 low = (UInt32)p->low; + unsigned high = (unsigned)(p->low >> 32); + p->low = (UInt32)(low << 8); + if (low < (UInt32)0xFF000000 || high != 0) + { + { + Byte *buf = p->buf; + *buf++ = (Byte)(p->cache + high); + p->cache = (unsigned)(low >> 24); + p->buf = buf; + if (buf == p->bufLim) + RangeEnc_FlushStream(p); + if (p->cacheSize == 0) + return; + } + high += 0xFF; + for (;;) + { + Byte *buf = p->buf; + *buf++ = (Byte)(high); + p->buf = buf; + if (buf == p->bufLim) + RangeEnc_FlushStream(p); + if (--p->cacheSize == 0) + return; + } + } + p->cacheSize++; +} + +static void RangeEnc_FlushData(CRangeEnc *p) +{ + int i; + for (i = 0; i < 5; i++) + RangeEnc_ShiftLow(p); +} + +#define RC_NORM(p) if (range < kTopValue) { range <<= 8; RangeEnc_ShiftLow(p); } + +#define RC_BIT_PRE(p, prob) \ + ttt = *(prob); \ + newBound = (range >> kNumBitModelTotalBits) * ttt; + +// #define _LZMA_ENC_USE_BRANCH + +#ifdef _LZMA_ENC_USE_BRANCH + +#define RC_BIT(p, prob, bit) { \ + RC_BIT_PRE(p, prob) \ + if (bit == 0) { range = newBound; ttt += (kBitModelTotal - ttt) >> kNumMoveBits; } \ + else { (p)->low += newBound; range -= newBound; ttt -= ttt >> kNumMoveBits; } \ + *(prob) = (CLzmaProb)ttt; \ + RC_NORM(p) \ + } + +#else + +#define RC_BIT(p, prob, bit) { \ + UInt32 mask; \ + RC_BIT_PRE(p, prob) \ + mask = 0 - (UInt32)bit; \ + range &= mask; \ + mask &= newBound; \ + range -= mask; \ + (p)->low += mask; \ + mask = (UInt32)bit - 1; \ + range += newBound & mask; \ + mask &= (kBitModelTotal - ((1 << kNumMoveBits) - 1)); \ + mask += ((1 << kNumMoveBits) - 1); \ + ttt += (UInt32)((Int32)(mask - ttt) >> kNumMoveBits); \ + *(prob) = (CLzmaProb)ttt; \ + RC_NORM(p) \ + } + +#endif + + + + +#define RC_BIT_0_BASE(p, prob) \ + range = newBound; *(prob) = (CLzmaProb)(ttt + ((kBitModelTotal - ttt) >> kNumMoveBits)); + +#define RC_BIT_1_BASE(p, prob) \ + range -= newBound; (p)->low += newBound; *(prob) = (CLzmaProb)(ttt - (ttt >> kNumMoveBits)); \ + +#define RC_BIT_0(p, prob) \ + RC_BIT_0_BASE(p, prob) \ + RC_NORM(p) + +#define RC_BIT_1(p, prob) \ + RC_BIT_1_BASE(p, prob) \ + RC_NORM(p) + +static void RangeEnc_EncodeBit_0(CRangeEnc *p, CLzmaProb *prob) +{ + UInt32 range, ttt, newBound; + range = p->range; + RC_BIT_PRE(p, prob) + RC_BIT_0(p, prob) + p->range = range; +} + +static void LitEnc_Encode(CRangeEnc *p, CLzmaProb *probs, UInt32 sym) +{ + UInt32 range = p->range; + sym |= 0x100; + do + { + UInt32 ttt, newBound; + // RangeEnc_EncodeBit(p, probs + (sym >> 8), (sym >> 7) & 1); + CLzmaProb *prob = probs + (sym >> 8); + UInt32 bit = (sym >> 7) & 1; + sym <<= 1; + RC_BIT(p, prob, bit); + } + while (sym < 0x10000); + p->range = range; +} + +static void LitEnc_EncodeMatched(CRangeEnc *p, CLzmaProb *probs, UInt32 sym, UInt32 matchByte) +{ + UInt32 range = p->range; + UInt32 offs = 0x100; + sym |= 0x100; + do + { + UInt32 ttt, newBound; + CLzmaProb *prob; + UInt32 bit; + matchByte <<= 1; + // RangeEnc_EncodeBit(p, probs + (offs + (matchByte & offs) + (sym >> 8)), (sym >> 7) & 1); + prob = probs + (offs + (matchByte & offs) + (sym >> 8)); + bit = (sym >> 7) & 1; + sym <<= 1; + offs &= ~(matchByte ^ sym); + RC_BIT(p, prob, bit); + } + while (sym < 0x10000); + p->range = range; +} + + + +static void LzmaEnc_InitPriceTables(CProbPrice *ProbPrices) +{ + UInt32 i; + for (i = 0; i < (kBitModelTotal >> kNumMoveReducingBits); i++) + { + const unsigned kCyclesBits = kNumBitPriceShiftBits; + UInt32 w = (i << kNumMoveReducingBits) + (1 << (kNumMoveReducingBits - 1)); + unsigned bitCount = 0; + unsigned j; + for (j = 0; j < kCyclesBits; j++) + { + w = w * w; + bitCount <<= 1; + while (w >= ((UInt32)1 << 16)) + { + w >>= 1; + bitCount++; + } + } + ProbPrices[i] = (CProbPrice)(((unsigned)kNumBitModelTotalBits << kCyclesBits) - 15 - bitCount); + // printf("\n%3d: %5d", i, ProbPrices[i]); + } +} + + +#define GET_PRICE(prob, bit) \ + p->ProbPrices[((prob) ^ (unsigned)(((-(int)(bit))) & (kBitModelTotal - 1))) >> kNumMoveReducingBits]; + +#define GET_PRICEa(prob, bit) \ + ProbPrices[((prob) ^ (unsigned)((-((int)(bit))) & (kBitModelTotal - 1))) >> kNumMoveReducingBits]; + +#define GET_PRICE_0(prob) p->ProbPrices[(prob) >> kNumMoveReducingBits] +#define GET_PRICE_1(prob) p->ProbPrices[((prob) ^ (kBitModelTotal - 1)) >> kNumMoveReducingBits] + +#define GET_PRICEa_0(prob) ProbPrices[(prob) >> kNumMoveReducingBits] +#define GET_PRICEa_1(prob) ProbPrices[((prob) ^ (kBitModelTotal - 1)) >> kNumMoveReducingBits] + + +static UInt32 LitEnc_GetPrice(const CLzmaProb *probs, UInt32 sym, const CProbPrice *ProbPrices) +{ + UInt32 price = 0; + sym |= 0x100; + do + { + unsigned bit = sym & 1; + sym >>= 1; + price += GET_PRICEa(probs[sym], bit); + } + while (sym >= 2); + return price; +} + + +static UInt32 LitEnc_Matched_GetPrice(const CLzmaProb *probs, UInt32 sym, UInt32 matchByte, const CProbPrice *ProbPrices) +{ + UInt32 price = 0; + UInt32 offs = 0x100; + sym |= 0x100; + do + { + matchByte <<= 1; + price += GET_PRICEa(probs[offs + (matchByte & offs) + (sym >> 8)], (sym >> 7) & 1); + sym <<= 1; + offs &= ~(matchByte ^ sym); + } + while (sym < 0x10000); + return price; +} + + +static void RcTree_ReverseEncode(CRangeEnc *rc, CLzmaProb *probs, unsigned numBits, unsigned sym) +{ + UInt32 range = rc->range; + unsigned m = 1; + do + { + UInt32 ttt, newBound; + unsigned bit = sym & 1; + // RangeEnc_EncodeBit(rc, probs + m, bit); + sym >>= 1; + RC_BIT(rc, probs + m, bit); + m = (m << 1) | bit; + } + while (--numBits); + rc->range = range; +} + + + +static void LenEnc_Init(CLenEnc *p) +{ + unsigned i; + for (i = 0; i < (LZMA_NUM_PB_STATES_MAX << (kLenNumLowBits + 1)); i++) + p->low[i] = kProbInitValue; + for (i = 0; i < kLenNumHighSymbols; i++) + p->high[i] = kProbInitValue; +} + +static void LenEnc_Encode(CLenEnc *p, CRangeEnc *rc, unsigned sym, unsigned posState) +{ + UInt32 range, ttt, newBound; + CLzmaProb *probs = p->low; + range = rc->range; + RC_BIT_PRE(rc, probs); + if (sym >= kLenNumLowSymbols) + { + RC_BIT_1(rc, probs); + probs += kLenNumLowSymbols; + RC_BIT_PRE(rc, probs); + if (sym >= kLenNumLowSymbols * 2) + { + RC_BIT_1(rc, probs); + rc->range = range; + // RcTree_Encode(rc, p->high, kLenNumHighBits, sym - kLenNumLowSymbols * 2); + LitEnc_Encode(rc, p->high, sym - kLenNumLowSymbols * 2); + return; + } + sym -= kLenNumLowSymbols; + } + + // RcTree_Encode(rc, probs + (posState << kLenNumLowBits), kLenNumLowBits, sym); + { + unsigned m; + unsigned bit; + RC_BIT_0(rc, probs); + probs += (posState << (1 + kLenNumLowBits)); + bit = (sym >> 2) ; RC_BIT(rc, probs + 1, bit); m = (1 << 1) + bit; + bit = (sym >> 1) & 1; RC_BIT(rc, probs + m, bit); m = (m << 1) + bit; + bit = sym & 1; RC_BIT(rc, probs + m, bit); + rc->range = range; + } +} + +static void SetPrices_3(const CLzmaProb *probs, UInt32 startPrice, UInt32 *prices, const CProbPrice *ProbPrices) +{ + unsigned i; + for (i = 0; i < 8; i += 2) + { + UInt32 price = startPrice; + UInt32 prob; + price += GET_PRICEa(probs[1 ], (i >> 2)); + price += GET_PRICEa(probs[2 + (i >> 2)], (i >> 1) & 1); + prob = probs[4 + (i >> 1)]; + prices[i ] = price + GET_PRICEa_0(prob); + prices[i + 1] = price + GET_PRICEa_1(prob); + } +} + + +MY_NO_INLINE static void MY_FAST_CALL LenPriceEnc_UpdateTables( + CLenPriceEnc *p, + unsigned numPosStates, + const CLenEnc *enc, + const CProbPrice *ProbPrices) +{ + UInt32 b; + + { + unsigned prob = enc->low[0]; + UInt32 a, c; + unsigned posState; + b = GET_PRICEa_1(prob); + a = GET_PRICEa_0(prob); + c = b + GET_PRICEa_0(enc->low[kLenNumLowSymbols]); + for (posState = 0; posState < numPosStates; posState++) + { + UInt32 *prices = p->prices[posState]; + const CLzmaProb *probs = enc->low + (posState << (1 + kLenNumLowBits)); + SetPrices_3(probs, a, prices, ProbPrices); + SetPrices_3(probs + kLenNumLowSymbols, c, prices + kLenNumLowSymbols, ProbPrices); + } + } + + /* + { + unsigned i; + UInt32 b; + a = GET_PRICEa_0(enc->low[0]); + for (i = 0; i < kLenNumLowSymbols; i++) + p->prices2[i] = a; + a = GET_PRICEa_1(enc->low[0]); + b = a + GET_PRICEa_0(enc->low[kLenNumLowSymbols]); + for (i = kLenNumLowSymbols; i < kLenNumLowSymbols * 2; i++) + p->prices2[i] = b; + a += GET_PRICEa_1(enc->low[kLenNumLowSymbols]); + } + */ + + // p->counter = numSymbols; + // p->counter = 64; + + { + unsigned i = p->tableSize; + + if (i > kLenNumLowSymbols * 2) + { + const CLzmaProb *probs = enc->high; + UInt32 *prices = p->prices[0] + kLenNumLowSymbols * 2; + i -= kLenNumLowSymbols * 2 - 1; + i >>= 1; + b += GET_PRICEa_1(enc->low[kLenNumLowSymbols]); + do + { + /* + p->prices2[i] = a + + // RcTree_GetPrice(enc->high, kLenNumHighBits, i - kLenNumLowSymbols * 2, ProbPrices); + LitEnc_GetPrice(probs, i - kLenNumLowSymbols * 2, ProbPrices); + */ + // UInt32 price = a + RcTree_GetPrice(probs, kLenNumHighBits - 1, sym, ProbPrices); + unsigned sym = --i + (1 << (kLenNumHighBits - 1)); + UInt32 price = b; + do + { + unsigned bit = sym & 1; + sym >>= 1; + price += GET_PRICEa(probs[sym], bit); + } + while (sym >= 2); + + { + unsigned prob = probs[(size_t)i + (1 << (kLenNumHighBits - 1))]; + prices[(size_t)i * 2 ] = price + GET_PRICEa_0(prob); + prices[(size_t)i * 2 + 1] = price + GET_PRICEa_1(prob); + } + } + while (i); + + { + unsigned posState; + size_t num = (p->tableSize - kLenNumLowSymbols * 2) * sizeof(p->prices[0][0]); + for (posState = 1; posState < numPosStates; posState++) + memcpy(p->prices[posState] + kLenNumLowSymbols * 2, p->prices[0] + kLenNumLowSymbols * 2, num); + } + } + } +} + +/* + #ifdef SHOW_STAT + g_STAT_OFFSET += num; + printf("\n MovePos %u", num); + #endif +*/ + +#define MOVE_POS(p, num) { \ + p->additionalOffset += (num); \ + p->matchFinder.Skip(p->matchFinderObj, (UInt32)(num)); } + + +static unsigned ReadMatchDistances(CLzmaEnc *p, unsigned *numPairsRes) +{ + unsigned numPairs; + + p->additionalOffset++; + p->numAvail = p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); + { + const UInt32 *d = p->matchFinder.GetMatches(p->matchFinderObj, p->matches); + // if (!d) { p->mf_Failure = True; *numPairsRes = 0; return 0; } + numPairs = (unsigned)(d - p->matches); + } + *numPairsRes = numPairs; + + #ifdef SHOW_STAT + printf("\n i = %u numPairs = %u ", g_STAT_OFFSET, numPairs / 2); + g_STAT_OFFSET++; + { + unsigned i; + for (i = 0; i < numPairs; i += 2) + printf("%2u %6u | ", p->matches[i], p->matches[i + 1]); + } + #endif + + if (numPairs == 0) + return 0; + { + const unsigned len = p->matches[(size_t)numPairs - 2]; + if (len != p->numFastBytes) + return len; + { + UInt32 numAvail = p->numAvail; + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + { + const Byte *p1 = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + const Byte *p2 = p1 + len; + const ptrdiff_t dif = (ptrdiff_t)-1 - (ptrdiff_t)p->matches[(size_t)numPairs - 1]; + const Byte *lim = p1 + numAvail; + for (; p2 != lim && *p2 == p2[dif]; p2++) + {} + return (unsigned)(p2 - p1); + } + } + } +} + +#define MARK_LIT ((UInt32)(Int32)-1) + +#define MakeAs_Lit(p) { (p)->dist = MARK_LIT; (p)->extra = 0; } +#define MakeAs_ShortRep(p) { (p)->dist = 0; (p)->extra = 0; } +#define IsShortRep(p) ((p)->dist == 0) + + +#define GetPrice_ShortRep(p, state, posState) \ + ( GET_PRICE_0(p->isRepG0[state]) + GET_PRICE_0(p->isRep0Long[state][posState])) + +#define GetPrice_Rep_0(p, state, posState) ( \ + GET_PRICE_1(p->isMatch[state][posState]) \ + + GET_PRICE_1(p->isRep0Long[state][posState])) \ + + GET_PRICE_1(p->isRep[state]) \ + + GET_PRICE_0(p->isRepG0[state]) + +MY_FORCE_INLINE +static UInt32 GetPrice_PureRep(const CLzmaEnc *p, unsigned repIndex, size_t state, size_t posState) +{ + UInt32 price; + UInt32 prob = p->isRepG0[state]; + if (repIndex == 0) + { + price = GET_PRICE_0(prob); + price += GET_PRICE_1(p->isRep0Long[state][posState]); + } + else + { + price = GET_PRICE_1(prob); + prob = p->isRepG1[state]; + if (repIndex == 1) + price += GET_PRICE_0(prob); + else + { + price += GET_PRICE_1(prob); + price += GET_PRICE(p->isRepG2[state], repIndex - 2); + } + } + return price; +} + + +static unsigned Backward(CLzmaEnc *p, unsigned cur) +{ + unsigned wr = cur + 1; + p->optEnd = wr; + + for (;;) + { + UInt32 dist = p->opt[cur].dist; + unsigned len = (unsigned)p->opt[cur].len; + unsigned extra = (unsigned)p->opt[cur].extra; + cur -= len; + + if (extra) + { + wr--; + p->opt[wr].len = (UInt32)len; + cur -= extra; + len = extra; + if (extra == 1) + { + p->opt[wr].dist = dist; + dist = MARK_LIT; + } + else + { + p->opt[wr].dist = 0; + len--; + wr--; + p->opt[wr].dist = MARK_LIT; + p->opt[wr].len = 1; + } + } + + if (cur == 0) + { + p->backRes = dist; + p->optCur = wr; + return len; + } + + wr--; + p->opt[wr].dist = dist; + p->opt[wr].len = (UInt32)len; + } +} + + + +#define LIT_PROBS(pos, prevByte) \ + (p->litProbs + (UInt32)3 * (((((pos) << 8) + (prevByte)) & p->lpMask) << p->lc)) + + +static unsigned GetOptimum(CLzmaEnc *p, UInt32 position) +{ + unsigned last, cur; + UInt32 reps[LZMA_NUM_REPS]; + unsigned repLens[LZMA_NUM_REPS]; + UInt32 *matches; + + { + UInt32 numAvail; + unsigned numPairs, mainLen, repMaxIndex, i, posState; + UInt32 matchPrice, repMatchPrice; + const Byte *data; + Byte curByte, matchByte; + + p->optCur = p->optEnd = 0; + + if (p->additionalOffset == 0) + mainLen = ReadMatchDistances(p, &numPairs); + else + { + mainLen = p->longestMatchLen; + numPairs = p->numPairs; + } + + numAvail = p->numAvail; + if (numAvail < 2) + { + p->backRes = MARK_LIT; + return 1; + } + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + repMaxIndex = 0; + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned len; + const Byte *data2; + reps[i] = p->reps[i]; + data2 = data - reps[i]; + if (data[0] != data2[0] || data[1] != data2[1]) + { + repLens[i] = 0; + continue; + } + for (len = 2; len < numAvail && data[len] == data2[len]; len++) + {} + repLens[i] = len; + if (len > repLens[repMaxIndex]) + repMaxIndex = i; + if (len == LZMA_MATCH_LEN_MAX) // 21.03 : optimization + break; + } + + if (repLens[repMaxIndex] >= p->numFastBytes) + { + unsigned len; + p->backRes = (UInt32)repMaxIndex; + len = repLens[repMaxIndex]; + MOVE_POS(p, len - 1) + return len; + } + + matches = p->matches; + #define MATCHES matches + // #define MATCHES p->matches + + if (mainLen >= p->numFastBytes) + { + p->backRes = MATCHES[(size_t)numPairs - 1] + LZMA_NUM_REPS; + MOVE_POS(p, mainLen - 1) + return mainLen; + } + + curByte = *data; + matchByte = *(data - reps[0]); + + last = repLens[repMaxIndex]; + if (last <= mainLen) + last = mainLen; + + if (last < 2 && curByte != matchByte) + { + p->backRes = MARK_LIT; + return 1; + } + + p->opt[0].state = (CState)p->state; + + posState = (position & p->pbMask); + + { + const CLzmaProb *probs = LIT_PROBS(position, *(data - 1)); + p->opt[1].price = GET_PRICE_0(p->isMatch[p->state][posState]) + + (!IsLitState(p->state) ? + LitEnc_Matched_GetPrice(probs, curByte, matchByte, p->ProbPrices) : + LitEnc_GetPrice(probs, curByte, p->ProbPrices)); + } + + MakeAs_Lit(&p->opt[1]); + + matchPrice = GET_PRICE_1(p->isMatch[p->state][posState]); + repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[p->state]); + + // 18.06 + if (matchByte == curByte && repLens[0] == 0) + { + UInt32 shortRepPrice = repMatchPrice + GetPrice_ShortRep(p, p->state, posState); + if (shortRepPrice < p->opt[1].price) + { + p->opt[1].price = shortRepPrice; + MakeAs_ShortRep(&p->opt[1]); + } + if (last < 2) + { + p->backRes = p->opt[1].dist; + return 1; + } + } + + p->opt[1].len = 1; + + p->opt[0].reps[0] = reps[0]; + p->opt[0].reps[1] = reps[1]; + p->opt[0].reps[2] = reps[2]; + p->opt[0].reps[3] = reps[3]; + + // ---------- REP ---------- + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned repLen = repLens[i]; + UInt32 price; + if (repLen < 2) + continue; + price = repMatchPrice + GetPrice_PureRep(p, i, p->state, posState); + do + { + UInt32 price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState, repLen); + COptimal *opt = &p->opt[repLen]; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)repLen; + opt->dist = (UInt32)i; + opt->extra = 0; + } + } + while (--repLen >= 2); + } + + + // ---------- MATCH ---------- + { + unsigned len = repLens[0] + 1; + if (len <= mainLen) + { + unsigned offs = 0; + UInt32 normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[p->state]); + + if (len < 2) + len = 2; + else + while (len > MATCHES[offs]) + offs += 2; + + for (; ; len++) + { + COptimal *opt; + UInt32 dist = MATCHES[(size_t)offs + 1]; + UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len); + unsigned lenToPosState = GetLenToPosState(len); + + if (dist < kNumFullDistances) + price += p->distancesPrices[lenToPosState][dist & (kNumFullDistances - 1)]; + else + { + unsigned slot; + GetPosSlot2(dist, slot); + price += p->alignPrices[dist & kAlignMask]; + price += p->posSlotPrices[lenToPosState][slot]; + } + + opt = &p->opt[len]; + + if (price < opt->price) + { + opt->price = price; + opt->len = (UInt32)len; + opt->dist = dist + LZMA_NUM_REPS; + opt->extra = 0; + } + + if (len == MATCHES[offs]) + { + offs += 2; + if (offs == numPairs) + break; + } + } + } + } + + + cur = 0; + + #ifdef SHOW_STAT2 + /* if (position >= 0) */ + { + unsigned i; + printf("\n pos = %4X", position); + for (i = cur; i <= last; i++) + printf("\nprice[%4X] = %u", position - cur + i, p->opt[i].price); + } + #endif + } + + + + // ---------- Optimal Parsing ---------- + + for (;;) + { + unsigned numAvail; + UInt32 numAvailFull; + unsigned newLen, numPairs, prev, state, posState, startLen; + UInt32 litPrice, matchPrice, repMatchPrice; + BoolInt nextIsLit; + Byte curByte, matchByte; + const Byte *data; + COptimal *curOpt, *nextOpt; + + if (++cur == last) + break; + + // 18.06 + if (cur >= kNumOpts - 64) + { + unsigned j, best; + UInt32 price = p->opt[cur].price; + best = cur; + for (j = cur + 1; j <= last; j++) + { + UInt32 price2 = p->opt[j].price; + if (price >= price2) + { + price = price2; + best = j; + } + } + { + unsigned delta = best - cur; + if (delta != 0) + { + MOVE_POS(p, delta); + } + } + cur = best; + break; + } + + newLen = ReadMatchDistances(p, &numPairs); + + if (newLen >= p->numFastBytes) + { + p->numPairs = numPairs; + p->longestMatchLen = newLen; + break; + } + + curOpt = &p->opt[cur]; + + position++; + + // we need that check here, if skip_items in p->opt are possible + /* + if (curOpt->price >= kInfinityPrice) + continue; + */ + + prev = cur - curOpt->len; + + if (curOpt->len == 1) + { + state = (unsigned)p->opt[prev].state; + if (IsShortRep(curOpt)) + state = kShortRepNextStates[state]; + else + state = kLiteralNextStates[state]; + } + else + { + const COptimal *prevOpt; + UInt32 b0; + UInt32 dist = curOpt->dist; + + if (curOpt->extra) + { + prev -= (unsigned)curOpt->extra; + state = kState_RepAfterLit; + if (curOpt->extra == 1) + state = (dist < LZMA_NUM_REPS ? kState_RepAfterLit : kState_MatchAfterLit); + } + else + { + state = (unsigned)p->opt[prev].state; + if (dist < LZMA_NUM_REPS) + state = kRepNextStates[state]; + else + state = kMatchNextStates[state]; + } + + prevOpt = &p->opt[prev]; + b0 = prevOpt->reps[0]; + + if (dist < LZMA_NUM_REPS) + { + if (dist == 0) + { + reps[0] = b0; + reps[1] = prevOpt->reps[1]; + reps[2] = prevOpt->reps[2]; + reps[3] = prevOpt->reps[3]; + } + else + { + reps[1] = b0; + b0 = prevOpt->reps[1]; + if (dist == 1) + { + reps[0] = b0; + reps[2] = prevOpt->reps[2]; + reps[3] = prevOpt->reps[3]; + } + else + { + reps[2] = b0; + reps[0] = prevOpt->reps[dist]; + reps[3] = prevOpt->reps[dist ^ 1]; + } + } + } + else + { + reps[0] = (dist - LZMA_NUM_REPS + 1); + reps[1] = b0; + reps[2] = prevOpt->reps[1]; + reps[3] = prevOpt->reps[2]; + } + } + + curOpt->state = (CState)state; + curOpt->reps[0] = reps[0]; + curOpt->reps[1] = reps[1]; + curOpt->reps[2] = reps[2]; + curOpt->reps[3] = reps[3]; + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + curByte = *data; + matchByte = *(data - reps[0]); + + posState = (position & p->pbMask); + + /* + The order of Price checks: + < LIT + <= SHORT_REP + < LIT : REP_0 + < REP [ : LIT : REP_0 ] + < MATCH [ : LIT : REP_0 ] + */ + + { + UInt32 curPrice = curOpt->price; + unsigned prob = p->isMatch[state][posState]; + matchPrice = curPrice + GET_PRICE_1(prob); + litPrice = curPrice + GET_PRICE_0(prob); + } + + nextOpt = &p->opt[(size_t)cur + 1]; + nextIsLit = False; + + // here we can allow skip_items in p->opt, if we don't check (nextOpt->price < kInfinityPrice) + // 18.new.06 + if ((nextOpt->price < kInfinityPrice + // && !IsLitState(state) + && matchByte == curByte) + || litPrice > nextOpt->price + ) + litPrice = 0; + else + { + const CLzmaProb *probs = LIT_PROBS(position, *(data - 1)); + litPrice += (!IsLitState(state) ? + LitEnc_Matched_GetPrice(probs, curByte, matchByte, p->ProbPrices) : + LitEnc_GetPrice(probs, curByte, p->ProbPrices)); + + if (litPrice < nextOpt->price) + { + nextOpt->price = litPrice; + nextOpt->len = 1; + MakeAs_Lit(nextOpt); + nextIsLit = True; + } + } + + repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[state]); + + numAvailFull = p->numAvail; + { + unsigned temp = kNumOpts - 1 - cur; + if (numAvailFull > temp) + numAvailFull = (UInt32)temp; + } + + // 18.06 + // ---------- SHORT_REP ---------- + if (IsLitState(state)) // 18.new + if (matchByte == curByte) + if (repMatchPrice < nextOpt->price) // 18.new + // if (numAvailFull < 2 || data[1] != *(data - reps[0] + 1)) + if ( + // nextOpt->price >= kInfinityPrice || + nextOpt->len < 2 // we can check nextOpt->len, if skip items are not allowed in p->opt + || (nextOpt->dist != 0 + // && nextOpt->extra <= 1 // 17.old + ) + ) + { + UInt32 shortRepPrice = repMatchPrice + GetPrice_ShortRep(p, state, posState); + // if (shortRepPrice <= nextOpt->price) // 17.old + if (shortRepPrice < nextOpt->price) // 18.new + { + nextOpt->price = shortRepPrice; + nextOpt->len = 1; + MakeAs_ShortRep(nextOpt); + nextIsLit = False; + } + } + + if (numAvailFull < 2) + continue; + numAvail = (numAvailFull <= p->numFastBytes ? numAvailFull : p->numFastBytes); + + // numAvail <= p->numFastBytes + + // ---------- LIT : REP_0 ---------- + + if (!nextIsLit + && litPrice != 0 // 18.new + && matchByte != curByte + && numAvailFull > 2) + { + const Byte *data2 = data - reps[0]; + if (data[1] == data2[1] && data[2] == data2[2]) + { + unsigned len; + unsigned limit = p->numFastBytes + 1; + if (limit > numAvailFull) + limit = numAvailFull; + for (len = 3; len < limit && data[len] == data2[len]; len++) + {} + + { + unsigned state2 = kLiteralNextStates[state]; + unsigned posState2 = (position + 1) & p->pbMask; + UInt32 price = litPrice + GetPrice_Rep_0(p, state2, posState2); + { + unsigned offset = cur + len; + + if (last < offset) + last = offset; + + // do + { + UInt32 price2; + COptimal *opt; + len--; + // price2 = price + GetPrice_Len_Rep_0(p, len, state2, posState2); + price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len); + + opt = &p->opt[offset]; + // offset--; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len; + opt->dist = 0; + opt->extra = 1; + } + } + // while (len >= 3); + } + } + } + } + + startLen = 2; /* speed optimization */ + + { + // ---------- REP ---------- + unsigned repIndex = 0; // 17.old + // unsigned repIndex = IsLitState(state) ? 0 : 1; // 18.notused + for (; repIndex < LZMA_NUM_REPS; repIndex++) + { + unsigned len; + UInt32 price; + const Byte *data2 = data - reps[repIndex]; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + + for (len = 2; len < numAvail && data[len] == data2[len]; len++) + {} + + // if (len < startLen) continue; // 18.new: speed optimization + + { + unsigned offset = cur + len; + if (last < offset) + last = offset; + } + { + unsigned len2 = len; + price = repMatchPrice + GetPrice_PureRep(p, repIndex, state, posState); + do + { + UInt32 price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState, len2); + COptimal *opt = &p->opt[cur + len2]; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len2; + opt->dist = (UInt32)repIndex; + opt->extra = 0; + } + } + while (--len2 >= 2); + } + + if (repIndex == 0) startLen = len + 1; // 17.old + // startLen = len + 1; // 18.new + + /* if (_maxMode) */ + { + // ---------- REP : LIT : REP_0 ---------- + // numFastBytes + 1 + numFastBytes + + unsigned len2 = len + 1; + unsigned limit = len2 + p->numFastBytes; + if (limit > numAvailFull) + limit = numAvailFull; + + len2 += 2; + if (len2 <= limit) + if (data[len2 - 2] == data2[len2 - 2]) + if (data[len2 - 1] == data2[len2 - 1]) + { + unsigned state2 = kRepNextStates[state]; + unsigned posState2 = (position + len) & p->pbMask; + price += GET_PRICE_LEN(&p->repLenEnc, posState, len) + + GET_PRICE_0(p->isMatch[state2][posState2]) + + LitEnc_Matched_GetPrice(LIT_PROBS(position + len, data[(size_t)len - 1]), + data[len], data2[len], p->ProbPrices); + + // state2 = kLiteralNextStates[state2]; + state2 = kState_LitAfterRep; + posState2 = (posState2 + 1) & p->pbMask; + + + price += GetPrice_Rep_0(p, state2, posState2); + + for (; len2 < limit && data[len2] == data2[len2]; len2++) + {} + + len2 -= len; + // if (len2 >= 3) + { + { + unsigned offset = cur + len + len2; + + if (last < offset) + last = offset; + // do + { + UInt32 price2; + COptimal *opt; + len2--; + // price2 = price + GetPrice_Len_Rep_0(p, len2, state2, posState2); + price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len2); + + opt = &p->opt[offset]; + // offset--; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len2; + opt->extra = (CExtra)(len + 1); + opt->dist = (UInt32)repIndex; + } + } + // while (len2 >= 3); + } + } + } + } + } + } + + + // ---------- MATCH ---------- + /* for (unsigned len = 2; len <= newLen; len++) */ + if (newLen > numAvail) + { + newLen = numAvail; + for (numPairs = 0; newLen > MATCHES[numPairs]; numPairs += 2); + MATCHES[numPairs] = (UInt32)newLen; + numPairs += 2; + } + + // startLen = 2; /* speed optimization */ + + if (newLen >= startLen) + { + UInt32 normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[state]); + UInt32 dist; + unsigned offs, posSlot, len; + + { + unsigned offset = cur + newLen; + if (last < offset) + last = offset; + } + + offs = 0; + while (startLen > MATCHES[offs]) + offs += 2; + dist = MATCHES[(size_t)offs + 1]; + + // if (dist >= kNumFullDistances) + GetPosSlot2(dist, posSlot); + + for (len = /*2*/ startLen; ; len++) + { + UInt32 price = normalMatchPrice + GET_PRICE_LEN(&p->lenEnc, posState, len); + { + COptimal *opt; + unsigned lenNorm = len - 2; + lenNorm = GetLenToPosState2(lenNorm); + if (dist < kNumFullDistances) + price += p->distancesPrices[lenNorm][dist & (kNumFullDistances - 1)]; + else + price += p->posSlotPrices[lenNorm][posSlot] + p->alignPrices[dist & kAlignMask]; + + opt = &p->opt[cur + len]; + if (price < opt->price) + { + opt->price = price; + opt->len = (UInt32)len; + opt->dist = dist + LZMA_NUM_REPS; + opt->extra = 0; + } + } + + if (len == MATCHES[offs]) + { + // if (p->_maxMode) { + // MATCH : LIT : REP_0 + + const Byte *data2 = data - dist - 1; + unsigned len2 = len + 1; + unsigned limit = len2 + p->numFastBytes; + if (limit > numAvailFull) + limit = numAvailFull; + + len2 += 2; + if (len2 <= limit) + if (data[len2 - 2] == data2[len2 - 2]) + if (data[len2 - 1] == data2[len2 - 1]) + { + for (; len2 < limit && data[len2] == data2[len2]; len2++) + {} + + len2 -= len; + + // if (len2 >= 3) + { + unsigned state2 = kMatchNextStates[state]; + unsigned posState2 = (position + len) & p->pbMask; + unsigned offset; + price += GET_PRICE_0(p->isMatch[state2][posState2]); + price += LitEnc_Matched_GetPrice(LIT_PROBS(position + len, data[(size_t)len - 1]), + data[len], data2[len], p->ProbPrices); + + // state2 = kLiteralNextStates[state2]; + state2 = kState_LitAfterMatch; + + posState2 = (posState2 + 1) & p->pbMask; + price += GetPrice_Rep_0(p, state2, posState2); + + offset = cur + len + len2; + + if (last < offset) + last = offset; + // do + { + UInt32 price2; + COptimal *opt; + len2--; + // price2 = price + GetPrice_Len_Rep_0(p, len2, state2, posState2); + price2 = price + GET_PRICE_LEN(&p->repLenEnc, posState2, len2); + opt = &p->opt[offset]; + // offset--; + if (price2 < opt->price) + { + opt->price = price2; + opt->len = (UInt32)len2; + opt->extra = (CExtra)(len + 1); + opt->dist = dist + LZMA_NUM_REPS; + } + } + // while (len2 >= 3); + } + + } + + offs += 2; + if (offs == numPairs) + break; + dist = MATCHES[(size_t)offs + 1]; + // if (dist >= kNumFullDistances) + GetPosSlot2(dist, posSlot); + } + } + } + } + + do + p->opt[last].price = kInfinityPrice; + while (--last); + + return Backward(p, cur); +} + + + +#define ChangePair(smallDist, bigDist) (((bigDist) >> 7) > (smallDist)) + + + +static unsigned GetOptimumFast(CLzmaEnc *p) +{ + UInt32 numAvail, mainDist; + unsigned mainLen, numPairs, repIndex, repLen, i; + const Byte *data; + + if (p->additionalOffset == 0) + mainLen = ReadMatchDistances(p, &numPairs); + else + { + mainLen = p->longestMatchLen; + numPairs = p->numPairs; + } + + numAvail = p->numAvail; + p->backRes = MARK_LIT; + if (numAvail < 2) + return 1; + // if (mainLen < 2 && p->state == 0) return 1; // 18.06.notused + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + repLen = repIndex = 0; + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned len; + const Byte *data2 = data - p->reps[i]; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + for (len = 2; len < numAvail && data[len] == data2[len]; len++) + {} + if (len >= p->numFastBytes) + { + p->backRes = (UInt32)i; + MOVE_POS(p, len - 1) + return len; + } + if (len > repLen) + { + repIndex = i; + repLen = len; + } + } + + if (mainLen >= p->numFastBytes) + { + p->backRes = p->matches[(size_t)numPairs - 1] + LZMA_NUM_REPS; + MOVE_POS(p, mainLen - 1) + return mainLen; + } + + mainDist = 0; /* for GCC */ + + if (mainLen >= 2) + { + mainDist = p->matches[(size_t)numPairs - 1]; + while (numPairs > 2) + { + UInt32 dist2; + if (mainLen != p->matches[(size_t)numPairs - 4] + 1) + break; + dist2 = p->matches[(size_t)numPairs - 3]; + if (!ChangePair(dist2, mainDist)) + break; + numPairs -= 2; + mainLen--; + mainDist = dist2; + } + if (mainLen == 2 && mainDist >= 0x80) + mainLen = 1; + } + + if (repLen >= 2) + if ( repLen + 1 >= mainLen + || (repLen + 2 >= mainLen && mainDist >= (1 << 9)) + || (repLen + 3 >= mainLen && mainDist >= (1 << 15))) + { + p->backRes = (UInt32)repIndex; + MOVE_POS(p, repLen - 1) + return repLen; + } + + if (mainLen < 2 || numAvail <= 2) + return 1; + + { + unsigned len1 = ReadMatchDistances(p, &p->numPairs); + p->longestMatchLen = len1; + + if (len1 >= 2) + { + UInt32 newDist = p->matches[(size_t)p->numPairs - 1]; + if ( (len1 >= mainLen && newDist < mainDist) + || (len1 == mainLen + 1 && !ChangePair(mainDist, newDist)) + || (len1 > mainLen + 1) + || (len1 + 1 >= mainLen && mainLen >= 3 && ChangePair(newDist, mainDist))) + return 1; + } + } + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + unsigned len, limit; + const Byte *data2 = data - p->reps[i]; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + limit = mainLen - 1; + for (len = 2;; len++) + { + if (len >= limit) + return 1; + if (data[len] != data2[len]) + break; + } + } + + p->backRes = mainDist + LZMA_NUM_REPS; + if (mainLen != 2) + { + MOVE_POS(p, mainLen - 2) + } + return mainLen; +} + + + + +static void WriteEndMarker(CLzmaEnc *p, unsigned posState) +{ + UInt32 range; + range = p->rc.range; + { + UInt32 ttt, newBound; + CLzmaProb *prob = &p->isMatch[p->state][posState]; + RC_BIT_PRE(&p->rc, prob) + RC_BIT_1(&p->rc, prob) + prob = &p->isRep[p->state]; + RC_BIT_PRE(&p->rc, prob) + RC_BIT_0(&p->rc, prob) + } + p->state = kMatchNextStates[p->state]; + + p->rc.range = range; + LenEnc_Encode(&p->lenProbs, &p->rc, 0, posState); + range = p->rc.range; + + { + // RcTree_Encode_PosSlot(&p->rc, p->posSlotEncoder[0], (1 << kNumPosSlotBits) - 1); + CLzmaProb *probs = p->posSlotEncoder[0]; + unsigned m = 1; + do + { + UInt32 ttt, newBound; + RC_BIT_PRE(p, probs + m) + RC_BIT_1(&p->rc, probs + m); + m = (m << 1) + 1; + } + while (m < (1 << kNumPosSlotBits)); + } + { + // RangeEnc_EncodeDirectBits(&p->rc, ((UInt32)1 << (30 - kNumAlignBits)) - 1, 30 - kNumAlignBits); UInt32 range = p->range; + unsigned numBits = 30 - kNumAlignBits; + do + { + range >>= 1; + p->rc.low += range; + RC_NORM(&p->rc) + } + while (--numBits); + } + + { + // RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, kAlignMask); + CLzmaProb *probs = p->posAlignEncoder; + unsigned m = 1; + do + { + UInt32 ttt, newBound; + RC_BIT_PRE(p, probs + m) + RC_BIT_1(&p->rc, probs + m); + m = (m << 1) + 1; + } + while (m < kAlignTableSize); + } + p->rc.range = range; +} + + +static SRes CheckErrors(CLzmaEnc *p) +{ + if (p->result != SZ_OK) + return p->result; + if (p->rc.res != SZ_OK) + p->result = SZ_ERROR_WRITE; + + #ifndef _7ZIP_ST + if ( + // p->mf_Failure || + (p->mtMode && + ( // p->matchFinderMt.failure_LZ_LZ || + p->matchFinderMt.failure_LZ_BT)) + ) + { + p->result = MY_HRES_ERROR__INTERNAL_ERROR; + // printf("\nCheckErrors p->matchFinderMt.failureLZ\n"); + } + #endif + + if (MFB.result != SZ_OK) + p->result = SZ_ERROR_READ; + + if (p->result != SZ_OK) + p->finished = True; + return p->result; +} + + +MY_NO_INLINE static SRes Flush(CLzmaEnc *p, UInt32 nowPos) +{ + /* ReleaseMFStream(); */ + p->finished = True; + if (p->writeEndMark) + WriteEndMarker(p, nowPos & p->pbMask); + RangeEnc_FlushData(&p->rc); + RangeEnc_FlushStream(&p->rc); + return CheckErrors(p); +} + + +MY_NO_INLINE static void FillAlignPrices(CLzmaEnc *p) +{ + unsigned i; + const CProbPrice *ProbPrices = p->ProbPrices; + const CLzmaProb *probs = p->posAlignEncoder; + // p->alignPriceCount = 0; + for (i = 0; i < kAlignTableSize / 2; i++) + { + UInt32 price = 0; + unsigned sym = i; + unsigned m = 1; + unsigned bit; + UInt32 prob; + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[m], bit); m = (m << 1) + bit; + prob = probs[m]; + p->alignPrices[i ] = price + GET_PRICEa_0(prob); + p->alignPrices[i + 8] = price + GET_PRICEa_1(prob); + // p->alignPrices[i] = RcTree_ReverseGetPrice(p->posAlignEncoder, kNumAlignBits, i, p->ProbPrices); + } +} + + +MY_NO_INLINE static void FillDistancesPrices(CLzmaEnc *p) +{ + // int y; for (y = 0; y < 100; y++) { + + UInt32 tempPrices[kNumFullDistances]; + unsigned i, lps; + + const CProbPrice *ProbPrices = p->ProbPrices; + p->matchPriceCount = 0; + + for (i = kStartPosModelIndex / 2; i < kNumFullDistances / 2; i++) + { + unsigned posSlot = GetPosSlot1(i); + unsigned footerBits = (posSlot >> 1) - 1; + unsigned base = ((2 | (posSlot & 1)) << footerBits); + const CLzmaProb *probs = p->posEncoders + (size_t)base * 2; + // tempPrices[i] = RcTree_ReverseGetPrice(p->posEncoders + base, footerBits, i - base, p->ProbPrices); + UInt32 price = 0; + unsigned m = 1; + unsigned sym = i; + unsigned offset = (unsigned)1 << footerBits; + base += i; + + if (footerBits) + do + { + unsigned bit = sym & 1; + sym >>= 1; + price += GET_PRICEa(probs[m], bit); + m = (m << 1) + bit; + } + while (--footerBits); + + { + unsigned prob = probs[m]; + tempPrices[base ] = price + GET_PRICEa_0(prob); + tempPrices[base + offset] = price + GET_PRICEa_1(prob); + } + } + + for (lps = 0; lps < kNumLenToPosStates; lps++) + { + unsigned slot; + unsigned distTableSize2 = (p->distTableSize + 1) >> 1; + UInt32 *posSlotPrices = p->posSlotPrices[lps]; + const CLzmaProb *probs = p->posSlotEncoder[lps]; + + for (slot = 0; slot < distTableSize2; slot++) + { + // posSlotPrices[slot] = RcTree_GetPrice(encoder, kNumPosSlotBits, slot, p->ProbPrices); + UInt32 price; + unsigned bit; + unsigned sym = slot + (1 << (kNumPosSlotBits - 1)); + unsigned prob; + bit = sym & 1; sym >>= 1; price = GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + bit = sym & 1; sym >>= 1; price += GET_PRICEa(probs[sym], bit); + prob = probs[(size_t)slot + (1 << (kNumPosSlotBits - 1))]; + posSlotPrices[(size_t)slot * 2 ] = price + GET_PRICEa_0(prob); + posSlotPrices[(size_t)slot * 2 + 1] = price + GET_PRICEa_1(prob); + } + + { + UInt32 delta = ((UInt32)((kEndPosModelIndex / 2 - 1) - kNumAlignBits) << kNumBitPriceShiftBits); + for (slot = kEndPosModelIndex / 2; slot < distTableSize2; slot++) + { + posSlotPrices[(size_t)slot * 2 ] += delta; + posSlotPrices[(size_t)slot * 2 + 1] += delta; + delta += ((UInt32)1 << kNumBitPriceShiftBits); + } + } + + { + UInt32 *dp = p->distancesPrices[lps]; + + dp[0] = posSlotPrices[0]; + dp[1] = posSlotPrices[1]; + dp[2] = posSlotPrices[2]; + dp[3] = posSlotPrices[3]; + + for (i = 4; i < kNumFullDistances; i += 2) + { + UInt32 slotPrice = posSlotPrices[GetPosSlot1(i)]; + dp[i ] = slotPrice + tempPrices[i]; + dp[i + 1] = slotPrice + tempPrices[i + 1]; + } + } + } + // } +} + + + +static void LzmaEnc_Construct(CLzmaEnc *p) +{ + RangeEnc_Construct(&p->rc); + MatchFinder_Construct(&MFB); + + #ifndef _7ZIP_ST + p->matchFinderMt.MatchFinder = &MFB; + MatchFinderMt_Construct(&p->matchFinderMt); + #endif + + { + CLzmaEncProps props; + LzmaEncProps_Init(&props); + LzmaEnc_SetProps(p, &props); + } + + #ifndef LZMA_LOG_BSR + LzmaEnc_FastPosInit(p->g_FastPos); + #endif + + LzmaEnc_InitPriceTables(p->ProbPrices); + p->litProbs = NULL; + p->saveState.litProbs = NULL; +} + +CLzmaEncHandle LzmaEnc_Create(ISzAllocPtr alloc) +{ + void *p; + p = ISzAlloc_Alloc(alloc, sizeof(CLzmaEnc)); + if (p) + LzmaEnc_Construct((CLzmaEnc *)p); + return p; +} + +static void LzmaEnc_FreeLits(CLzmaEnc *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->litProbs); + ISzAlloc_Free(alloc, p->saveState.litProbs); + p->litProbs = NULL; + p->saveState.litProbs = NULL; +} + +static void LzmaEnc_Destruct(CLzmaEnc *p, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + #ifndef _7ZIP_ST + MatchFinderMt_Destruct(&p->matchFinderMt, allocBig); + #endif + + MatchFinder_Free(&MFB, allocBig); + LzmaEnc_FreeLits(p, alloc); + RangeEnc_Free(&p->rc, alloc); +} + +void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + LzmaEnc_Destruct((CLzmaEnc *)p, alloc, allocBig); + ISzAlloc_Free(alloc, p); +} + + +MY_NO_INLINE +static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, UInt32 maxPackSize, UInt32 maxUnpackSize) +{ + UInt32 nowPos32, startPos32; + if (p->needInit) + { + #ifndef _7ZIP_ST + if (p->mtMode) + { + RINOK(MatchFinderMt_InitMt(&p->matchFinderMt)); + } + #endif + p->matchFinder.Init(p->matchFinderObj); + p->needInit = 0; + } + + if (p->finished) + return p->result; + RINOK(CheckErrors(p)); + + nowPos32 = (UInt32)p->nowPos64; + startPos32 = nowPos32; + + if (p->nowPos64 == 0) + { + unsigned numPairs; + Byte curByte; + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) + return Flush(p, nowPos32); + ReadMatchDistances(p, &numPairs); + RangeEnc_EncodeBit_0(&p->rc, &p->isMatch[kState_Start][0]); + // p->state = kLiteralNextStates[p->state]; + curByte = *(p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset); + LitEnc_Encode(&p->rc, p->litProbs, curByte); + p->additionalOffset--; + nowPos32++; + } + + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) != 0) + + for (;;) + { + UInt32 dist; + unsigned len, posState; + UInt32 range, ttt, newBound; + CLzmaProb *probs; + + if (p->fastMode) + len = GetOptimumFast(p); + else + { + unsigned oci = p->optCur; + if (p->optEnd == oci) + len = GetOptimum(p, nowPos32); + else + { + const COptimal *opt = &p->opt[oci]; + len = opt->len; + p->backRes = opt->dist; + p->optCur = oci + 1; + } + } + + posState = (unsigned)nowPos32 & p->pbMask; + range = p->rc.range; + probs = &p->isMatch[p->state][posState]; + + RC_BIT_PRE(&p->rc, probs) + + dist = p->backRes; + + #ifdef SHOW_STAT2 + printf("\n pos = %6X, len = %3u pos = %6u", nowPos32, len, dist); + #endif + + if (dist == MARK_LIT) + { + Byte curByte; + const Byte *data; + unsigned state; + + RC_BIT_0(&p->rc, probs); + p->rc.range = range; + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset; + probs = LIT_PROBS(nowPos32, *(data - 1)); + curByte = *data; + state = p->state; + p->state = kLiteralNextStates[state]; + if (IsLitState(state)) + LitEnc_Encode(&p->rc, probs, curByte); + else + LitEnc_EncodeMatched(&p->rc, probs, curByte, *(data - p->reps[0])); + } + else + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRep[p->state]; + RC_BIT_PRE(&p->rc, probs) + + if (dist < LZMA_NUM_REPS) + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRepG0[p->state]; + RC_BIT_PRE(&p->rc, probs) + if (dist == 0) + { + RC_BIT_0(&p->rc, probs); + probs = &p->isRep0Long[p->state][posState]; + RC_BIT_PRE(&p->rc, probs) + if (len != 1) + { + RC_BIT_1_BASE(&p->rc, probs); + } + else + { + RC_BIT_0_BASE(&p->rc, probs); + p->state = kShortRepNextStates[p->state]; + } + } + else + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRepG1[p->state]; + RC_BIT_PRE(&p->rc, probs) + if (dist == 1) + { + RC_BIT_0_BASE(&p->rc, probs); + dist = p->reps[1]; + } + else + { + RC_BIT_1(&p->rc, probs); + probs = &p->isRepG2[p->state]; + RC_BIT_PRE(&p->rc, probs) + if (dist == 2) + { + RC_BIT_0_BASE(&p->rc, probs); + dist = p->reps[2]; + } + else + { + RC_BIT_1_BASE(&p->rc, probs); + dist = p->reps[3]; + p->reps[3] = p->reps[2]; + } + p->reps[2] = p->reps[1]; + } + p->reps[1] = p->reps[0]; + p->reps[0] = dist; + } + + RC_NORM(&p->rc) + + p->rc.range = range; + + if (len != 1) + { + LenEnc_Encode(&p->repLenProbs, &p->rc, len - LZMA_MATCH_LEN_MIN, posState); + --p->repLenEncCounter; + p->state = kRepNextStates[p->state]; + } + } + else + { + unsigned posSlot; + RC_BIT_0(&p->rc, probs); + p->rc.range = range; + p->state = kMatchNextStates[p->state]; + + LenEnc_Encode(&p->lenProbs, &p->rc, len - LZMA_MATCH_LEN_MIN, posState); + // --p->lenEnc.counter; + + dist -= LZMA_NUM_REPS; + p->reps[3] = p->reps[2]; + p->reps[2] = p->reps[1]; + p->reps[1] = p->reps[0]; + p->reps[0] = dist + 1; + + p->matchPriceCount++; + GetPosSlot(dist, posSlot); + // RcTree_Encode_PosSlot(&p->rc, p->posSlotEncoder[GetLenToPosState(len)], posSlot); + { + UInt32 sym = (UInt32)posSlot + (1 << kNumPosSlotBits); + range = p->rc.range; + probs = p->posSlotEncoder[GetLenToPosState(len)]; + do + { + CLzmaProb *prob = probs + (sym >> kNumPosSlotBits); + UInt32 bit = (sym >> (kNumPosSlotBits - 1)) & 1; + sym <<= 1; + RC_BIT(&p->rc, prob, bit); + } + while (sym < (1 << kNumPosSlotBits * 2)); + p->rc.range = range; + } + + if (dist >= kStartPosModelIndex) + { + unsigned footerBits = ((posSlot >> 1) - 1); + + if (dist < kNumFullDistances) + { + unsigned base = ((2 | (posSlot & 1)) << footerBits); + RcTree_ReverseEncode(&p->rc, p->posEncoders + base, footerBits, (unsigned)(dist /* - base */)); + } + else + { + UInt32 pos2 = (dist | 0xF) << (32 - footerBits); + range = p->rc.range; + // RangeEnc_EncodeDirectBits(&p->rc, posReduced >> kNumAlignBits, footerBits - kNumAlignBits); + /* + do + { + range >>= 1; + p->rc.low += range & (0 - ((dist >> --footerBits) & 1)); + RC_NORM(&p->rc) + } + while (footerBits > kNumAlignBits); + */ + do + { + range >>= 1; + p->rc.low += range & (0 - (pos2 >> 31)); + pos2 += pos2; + RC_NORM(&p->rc) + } + while (pos2 != 0xF0000000); + + + // RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, posReduced & kAlignMask); + + { + unsigned m = 1; + unsigned bit; + bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; + bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; + bit = dist & 1; dist >>= 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); m = (m << 1) + bit; + bit = dist & 1; RC_BIT(&p->rc, p->posAlignEncoder + m, bit); + p->rc.range = range; + // p->alignPriceCount++; + } + } + } + } + } + + nowPos32 += (UInt32)len; + p->additionalOffset -= len; + + if (p->additionalOffset == 0) + { + UInt32 processed; + + if (!p->fastMode) + { + /* + if (p->alignPriceCount >= 16) // kAlignTableSize + FillAlignPrices(p); + if (p->matchPriceCount >= 128) + FillDistancesPrices(p); + if (p->lenEnc.counter <= 0) + LenPriceEnc_UpdateTables(&p->lenEnc, 1 << p->pb, &p->lenProbs, p->ProbPrices); + */ + if (p->matchPriceCount >= 64) + { + FillAlignPrices(p); + // { int y; for (y = 0; y < 100; y++) { + FillDistancesPrices(p); + // }} + LenPriceEnc_UpdateTables(&p->lenEnc, (unsigned)1 << p->pb, &p->lenProbs, p->ProbPrices); + } + if (p->repLenEncCounter <= 0) + { + p->repLenEncCounter = REP_LEN_COUNT; + LenPriceEnc_UpdateTables(&p->repLenEnc, (unsigned)1 << p->pb, &p->repLenProbs, p->ProbPrices); + } + } + + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) + break; + processed = nowPos32 - startPos32; + + if (maxPackSize) + { + if (processed + kNumOpts + 300 >= maxUnpackSize + || RangeEnc_GetProcessed_sizet(&p->rc) + kPackReserve >= maxPackSize) + break; + } + else if (processed >= (1 << 17)) + { + p->nowPos64 += nowPos32 - startPos32; + return CheckErrors(p); + } + } + } + + p->nowPos64 += nowPos32 - startPos32; + return Flush(p, nowPos32); +} + + + +#define kBigHashDicLimit ((UInt32)1 << 24) + +static SRes LzmaEnc_Alloc(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + UInt32 beforeSize = kNumOpts; + UInt32 dictSize; + + if (!RangeEnc_Alloc(&p->rc, alloc)) + return SZ_ERROR_MEM; + + #ifndef _7ZIP_ST + p->mtMode = (p->multiThread && !p->fastMode && (MFB.btMode != 0)); + #endif + + { + unsigned lclp = p->lc + p->lp; + if (!p->litProbs || !p->saveState.litProbs || p->lclp != lclp) + { + LzmaEnc_FreeLits(p, alloc); + p->litProbs = (CLzmaProb *)ISzAlloc_Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); + p->saveState.litProbs = (CLzmaProb *)ISzAlloc_Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); + if (!p->litProbs || !p->saveState.litProbs) + { + LzmaEnc_FreeLits(p, alloc); + return SZ_ERROR_MEM; + } + p->lclp = lclp; + } + } + + MFB.bigHash = (Byte)(p->dictSize > kBigHashDicLimit ? 1 : 0); + + + dictSize = p->dictSize; + if (dictSize == ((UInt32)2 << 30) || + dictSize == ((UInt32)3 << 30)) + { + /* 21.03 : here we reduce the dictionary for 2 reasons: + 1) we don't want 32-bit back_distance matches in decoder for 2 GB dictionary. + 2) we want to elimate useless last MatchFinder_Normalize3() for corner cases, + where data size is aligned for 1 GB: 5/6/8 GB. + That reducing must be >= 1 for such corner cases. */ + dictSize -= 1; + } + + if (beforeSize + dictSize < keepWindowSize) + beforeSize = keepWindowSize - dictSize; + + /* in worst case we can look ahead for + max(LZMA_MATCH_LEN_MAX, numFastBytes + 1 + numFastBytes) bytes. + we send larger value for (keepAfter) to MantchFinder_Create(): + (numFastBytes + LZMA_MATCH_LEN_MAX + 1) + */ + + #ifndef _7ZIP_ST + if (p->mtMode) + { + RINOK(MatchFinderMt_Create(&p->matchFinderMt, dictSize, beforeSize, + p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 18.04 */ + , allocBig)); + p->matchFinderObj = &p->matchFinderMt; + MFB.bigHash = (Byte)( + (p->dictSize > kBigHashDicLimit && MFB.hashMask >= 0xFFFFFF) ? 1 : 0); + MatchFinderMt_CreateVTable(&p->matchFinderMt, &p->matchFinder); + } + else + #endif + { + if (!MatchFinder_Create(&MFB, dictSize, beforeSize, + p->numFastBytes, LZMA_MATCH_LEN_MAX + 1 /* 21.03 */ + , allocBig)) + return SZ_ERROR_MEM; + p->matchFinderObj = &MFB; + MatchFinder_CreateVTable(&MFB, &p->matchFinder); + } + + return SZ_OK; +} + +static void LzmaEnc_Init(CLzmaEnc *p) +{ + unsigned i; + p->state = 0; + p->reps[0] = + p->reps[1] = + p->reps[2] = + p->reps[3] = 1; + + RangeEnc_Init(&p->rc); + + for (i = 0; i < (1 << kNumAlignBits); i++) + p->posAlignEncoder[i] = kProbInitValue; + + for (i = 0; i < kNumStates; i++) + { + unsigned j; + for (j = 0; j < LZMA_NUM_PB_STATES_MAX; j++) + { + p->isMatch[i][j] = kProbInitValue; + p->isRep0Long[i][j] = kProbInitValue; + } + p->isRep[i] = kProbInitValue; + p->isRepG0[i] = kProbInitValue; + p->isRepG1[i] = kProbInitValue; + p->isRepG2[i] = kProbInitValue; + } + + { + for (i = 0; i < kNumLenToPosStates; i++) + { + CLzmaProb *probs = p->posSlotEncoder[i]; + unsigned j; + for (j = 0; j < (1 << kNumPosSlotBits); j++) + probs[j] = kProbInitValue; + } + } + { + for (i = 0; i < kNumFullDistances; i++) + p->posEncoders[i] = kProbInitValue; + } + + { + UInt32 num = (UInt32)0x300 << (p->lp + p->lc); + UInt32 k; + CLzmaProb *probs = p->litProbs; + for (k = 0; k < num; k++) + probs[k] = kProbInitValue; + } + + + LenEnc_Init(&p->lenProbs); + LenEnc_Init(&p->repLenProbs); + + p->optEnd = 0; + p->optCur = 0; + + { + for (i = 0; i < kNumOpts; i++) + p->opt[i].price = kInfinityPrice; + } + + p->additionalOffset = 0; + + p->pbMask = ((unsigned)1 << p->pb) - 1; + p->lpMask = ((UInt32)0x100 << p->lp) - ((unsigned)0x100 >> p->lc); + + // p->mf_Failure = False; +} + + +static void LzmaEnc_InitPrices(CLzmaEnc *p) +{ + if (!p->fastMode) + { + FillDistancesPrices(p); + FillAlignPrices(p); + } + + p->lenEnc.tableSize = + p->repLenEnc.tableSize = + p->numFastBytes + 1 - LZMA_MATCH_LEN_MIN; + + p->repLenEncCounter = REP_LEN_COUNT; + + LenPriceEnc_UpdateTables(&p->lenEnc, (unsigned)1 << p->pb, &p->lenProbs, p->ProbPrices); + LenPriceEnc_UpdateTables(&p->repLenEnc, (unsigned)1 << p->pb, &p->repLenProbs, p->ProbPrices); +} + +static SRes LzmaEnc_AllocAndInit(CLzmaEnc *p, UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + unsigned i; + for (i = kEndPosModelIndex / 2; i < kDicLogSizeMax; i++) + if (p->dictSize <= ((UInt32)1 << i)) + break; + p->distTableSize = i * 2; + + p->finished = False; + p->result = SZ_OK; + RINOK(LzmaEnc_Alloc(p, keepWindowSize, alloc, allocBig)); + LzmaEnc_Init(p); + LzmaEnc_InitPrices(p); + p->nowPos64 = 0; + return SZ_OK; +} + +static SRes LzmaEnc_Prepare(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, + ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + MFB.stream = inStream; + p->needInit = 1; + p->rc.outStream = outStream; + return LzmaEnc_AllocAndInit(p, 0, alloc, allocBig); +} + +SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, + ISeqInStream *inStream, UInt32 keepWindowSize, + ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + MFB.stream = inStream; + p->needInit = 1; + return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); +} + +static void LzmaEnc_SetInputBuf(CLzmaEnc *p, const Byte *src, SizeT srcLen) +{ + MFB.directInput = 1; + MFB.bufferBase = (Byte *)src; + MFB.directInputRem = srcLen; +} + +SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, + UInt32 keepWindowSize, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + LzmaEnc_SetInputBuf(p, src, srcLen); + p->needInit = 1; + + LzmaEnc_SetDataSize(pp, srcLen); + return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); +} + +void LzmaEnc_Finish(CLzmaEncHandle pp) +{ + #ifndef _7ZIP_ST + CLzmaEnc *p = (CLzmaEnc *)pp; + if (p->mtMode) + MatchFinderMt_ReleaseStream(&p->matchFinderMt); + #else + UNUSED_VAR(pp); + #endif +} + + +typedef struct +{ + ISeqOutStream vt; + Byte *data; + SizeT rem; + BoolInt overflow; +} CLzmaEnc_SeqOutStreamBuf; + +static size_t SeqOutStreamBuf_Write(const ISeqOutStream *pp, const void *data, size_t size) +{ + CLzmaEnc_SeqOutStreamBuf *p = CONTAINER_FROM_VTBL(pp, CLzmaEnc_SeqOutStreamBuf, vt); + if (p->rem < size) + { + size = p->rem; + p->overflow = True; + } + if (size != 0) + { + memcpy(p->data, data, size); + p->rem -= size; + p->data += size; + } + return size; +} + + +/* +UInt32 LzmaEnc_GetNumAvailableBytes(CLzmaEncHandle pp) +{ + const CLzmaEnc *p = (CLzmaEnc *)pp; + return p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); +} +*/ + +const Byte *LzmaEnc_GetCurBuf(CLzmaEncHandle pp) +{ + const CLzmaEnc *p = (CLzmaEnc *)pp; + return p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset; +} + + +SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, BoolInt reInit, + Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + UInt64 nowPos64; + SRes res; + CLzmaEnc_SeqOutStreamBuf outStream; + + outStream.vt.Write = SeqOutStreamBuf_Write; + outStream.data = dest; + outStream.rem = *destLen; + outStream.overflow = False; + + p->writeEndMark = False; + p->finished = False; + p->result = SZ_OK; + + if (reInit) + LzmaEnc_Init(p); + LzmaEnc_InitPrices(p); + + nowPos64 = p->nowPos64; + RangeEnc_Init(&p->rc); + p->rc.outStream = &outStream.vt; + + if (desiredPackSize == 0) + return SZ_ERROR_OUTPUT_EOF; + + res = LzmaEnc_CodeOneBlock(p, desiredPackSize, *unpackSize); + + *unpackSize = (UInt32)(p->nowPos64 - nowPos64); + *destLen -= outStream.rem; + if (outStream.overflow) + return SZ_ERROR_OUTPUT_EOF; + + return res; +} + + +MY_NO_INLINE +static SRes LzmaEnc_Encode2(CLzmaEnc *p, ICompressProgress *progress) +{ + SRes res = SZ_OK; + + #ifndef _7ZIP_ST + Byte allocaDummy[0x300]; + allocaDummy[0] = 0; + allocaDummy[1] = allocaDummy[0]; + #endif + + for (;;) + { + res = LzmaEnc_CodeOneBlock(p, 0, 0); + if (res != SZ_OK || p->finished) + break; + if (progress) + { + res = ICompressProgress_Progress(progress, p->nowPos64, RangeEnc_GetProcessed(&p->rc)); + if (res != SZ_OK) + { + res = SZ_ERROR_PROGRESS; + break; + } + } + } + + LzmaEnc_Finish(p); + + /* + if (res == SZ_OK && !Inline_MatchFinder_IsFinishedOK(&MFB)) + res = SZ_ERROR_FAIL; + } + */ + + return res; +} + + +SRes LzmaEnc_Encode(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress, + ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + RINOK(LzmaEnc_Prepare(pp, outStream, inStream, alloc, allocBig)); + return LzmaEnc_Encode2((CLzmaEnc *)pp, progress); +} + + +SRes LzmaEnc_WriteProperties(CLzmaEncHandle pp, Byte *props, SizeT *size) +{ + if (*size < LZMA_PROPS_SIZE) + return SZ_ERROR_PARAM; + *size = LZMA_PROPS_SIZE; + { + const CLzmaEnc *p = (const CLzmaEnc *)pp; + const UInt32 dictSize = p->dictSize; + UInt32 v; + props[0] = (Byte)((p->pb * 5 + p->lp) * 9 + p->lc); + + // we write aligned dictionary value to properties for lzma decoder + if (dictSize >= ((UInt32)1 << 21)) + { + const UInt32 kDictMask = ((UInt32)1 << 20) - 1; + v = (dictSize + kDictMask) & ~kDictMask; + if (v < dictSize) + v = dictSize; + } + else + { + unsigned i = 11 * 2; + do + { + v = (UInt32)(2 + (i & 1)) << (i >> 1); + i++; + } + while (v < dictSize); + } + + SetUi32(props + 1, v); + return SZ_OK; + } +} + + +unsigned LzmaEnc_IsWriteEndMark(CLzmaEncHandle pp) +{ + return (unsigned)((CLzmaEnc *)pp)->writeEndMark; +} + + +SRes LzmaEnc_MemEncode(CLzmaEncHandle pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + int writeEndMark, ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + SRes res; + CLzmaEnc *p = (CLzmaEnc *)pp; + + CLzmaEnc_SeqOutStreamBuf outStream; + + outStream.vt.Write = SeqOutStreamBuf_Write; + outStream.data = dest; + outStream.rem = *destLen; + outStream.overflow = False; + + p->writeEndMark = writeEndMark; + p->rc.outStream = &outStream.vt; + + res = LzmaEnc_MemPrepare(pp, src, srcLen, 0, alloc, allocBig); + + if (res == SZ_OK) + { + res = LzmaEnc_Encode2(p, progress); + if (res == SZ_OK && p->nowPos64 != srcLen) + res = SZ_ERROR_FAIL; + } + + *destLen -= outStream.rem; + if (outStream.overflow) + return SZ_ERROR_OUTPUT_EOF; + return res; +} + + +SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, + ICompressProgress *progress, ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)LzmaEnc_Create(alloc); + SRes res; + if (!p) + return SZ_ERROR_MEM; + + res = LzmaEnc_SetProps(p, props); + if (res == SZ_OK) + { + res = LzmaEnc_WriteProperties(p, propsEncoded, propsSize); + if (res == SZ_OK) + res = LzmaEnc_MemEncode(p, dest, destLen, src, srcLen, + writeEndMark, progress, alloc, allocBig); + } + + LzmaEnc_Destroy(p, alloc, allocBig); + return res; +} + + +/* +#ifndef _7ZIP_ST +void LzmaEnc_GetLzThreads(CLzmaEncHandle pp, HANDLE lz_threads[2]) +{ + const CLzmaEnc *p = (CLzmaEnc *)pp; + lz_threads[0] = p->matchFinderMt.hashSync.thread; + lz_threads[1] = p->matchFinderMt.btSync.thread; +} +#endif +*/ diff --git a/3rdparty/lzma/src/LzmaLib.c b/3rdparty/lzma/src/LzmaLib.c new file mode 100644 index 0000000000..706e9e58cd --- /dev/null +++ b/3rdparty/lzma/src/LzmaLib.c @@ -0,0 +1,40 @@ +/* LzmaLib.c -- LZMA library wrapper +2015-06-13 : Igor Pavlov : Public domain */ + +#include "Alloc.h" +#include "LzmaDec.h" +#include "LzmaEnc.h" +#include "LzmaLib.h" + +MY_STDAPI LzmaCompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t srcLen, + unsigned char *outProps, size_t *outPropsSize, + int level, /* 0 <= level <= 9, default = 5 */ + unsigned dictSize, /* use (1 << N) or (3 << N). 4 KB < dictSize <= 128 MB */ + int lc, /* 0 <= lc <= 8, default = 3 */ + int lp, /* 0 <= lp <= 4, default = 0 */ + int pb, /* 0 <= pb <= 4, default = 2 */ + int fb, /* 5 <= fb <= 273, default = 32 */ + int numThreads /* 1 or 2, default = 2 */ +) +{ + CLzmaEncProps props; + LzmaEncProps_Init(&props); + props.level = level; + props.dictSize = dictSize; + props.lc = lc; + props.lp = lp; + props.pb = pb; + props.fb = fb; + props.numThreads = numThreads; + + return LzmaEncode(dest, destLen, src, srcLen, &props, outProps, outPropsSize, 0, + NULL, &g_Alloc, &g_Alloc); +} + + +MY_STDAPI LzmaUncompress(unsigned char *dest, size_t *destLen, const unsigned char *src, size_t *srcLen, + const unsigned char *props, size_t propsSize) +{ + ELzmaStatus status; + return LzmaDecode(dest, destLen, src, srcLen, props, (unsigned)propsSize, LZMA_FINISH_ANY, &status, &g_Alloc); +} diff --git a/3rdparty/lzma/src/Ppmd7.c b/3rdparty/lzma/src/Ppmd7.c new file mode 100644 index 0000000000..cf401cb37c --- /dev/null +++ b/3rdparty/lzma/src/Ppmd7.c @@ -0,0 +1,1104 @@ +/* Ppmd7.c -- PPMdH codec +2021-04-13 : Igor Pavlov : Public domain +This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ + +#include "Precomp.h" + +#include + +#include "Ppmd7.h" + +/* define PPMD7_ORDER_0_SUPPPORT to suport order-0 mode, unsupported by orignal PPMd var.H. code */ +// #define PPMD7_ORDER_0_SUPPPORT + +MY_ALIGN(16) +static const Byte PPMD7_kExpEscape[16] = { 25, 14, 9, 7, 5, 5, 4, 4, 4, 3, 3, 3, 2, 2, 2, 2 }; +MY_ALIGN(16) +static const UInt16 kInitBinEsc[] = { 0x3CDD, 0x1F3F, 0x59BF, 0x48F3, 0x64A1, 0x5ABC, 0x6632, 0x6051}; + +#define MAX_FREQ 124 +#define UNIT_SIZE 12 + +#define U2B(nu) ((UInt32)(nu) * UNIT_SIZE) +#define U2I(nu) (p->Units2Indx[(size_t)(nu) - 1]) +#define I2U(indx) ((unsigned)p->Indx2Units[indx]) +#define I2U_UInt16(indx) ((UInt16)p->Indx2Units[indx]) + +#define REF(ptr) Ppmd_GetRef(p, ptr) + +#define STATS_REF(ptr) ((CPpmd_State_Ref)REF(ptr)) + +#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref)) +#define STATS(ctx) Ppmd7_GetStats(p, ctx) +#define ONE_STATE(ctx) Ppmd7Context_OneState(ctx) +#define SUFFIX(ctx) CTX((ctx)->Suffix) + +typedef CPpmd7_Context * CTX_PTR; + +struct CPpmd7_Node_; + +typedef Ppmd_Ref_Type(struct CPpmd7_Node_) CPpmd7_Node_Ref; + +typedef struct CPpmd7_Node_ +{ + UInt16 Stamp; /* must be at offset 0 as CPpmd7_Context::NumStats. Stamp=0 means free */ + UInt16 NU; + CPpmd7_Node_Ref Next; /* must be at offset >= 4 */ + CPpmd7_Node_Ref Prev; +} CPpmd7_Node; + +#define NODE(r) Ppmd_GetPtr_Type(p, r, CPpmd7_Node) + +void Ppmd7_Construct(CPpmd7 *p) +{ + unsigned i, k, m; + + p->Base = NULL; + + for (i = 0, k = 0; i < PPMD_NUM_INDEXES; i++) + { + unsigned step = (i >= 12 ? 4 : (i >> 2) + 1); + do { p->Units2Indx[k++] = (Byte)i; } while (--step); + p->Indx2Units[i] = (Byte)k; + } + + p->NS2BSIndx[0] = (0 << 1); + p->NS2BSIndx[1] = (1 << 1); + memset(p->NS2BSIndx + 2, (2 << 1), 9); + memset(p->NS2BSIndx + 11, (3 << 1), 256 - 11); + + for (i = 0; i < 3; i++) + p->NS2Indx[i] = (Byte)i; + + for (m = i, k = 1; i < 256; i++) + { + p->NS2Indx[i] = (Byte)m; + if (--k == 0) + k = (++m) - 2; + } + + memcpy(p->ExpEscape, PPMD7_kExpEscape, 16); +} + + +void Ppmd7_Free(CPpmd7 *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->Base); + p->Size = 0; + p->Base = NULL; +} + + +BoolInt Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAllocPtr alloc) +{ + if (!p->Base || p->Size != size) + { + Ppmd7_Free(p, alloc); + p->AlignOffset = (4 - size) & 3; + if ((p->Base = (Byte *)ISzAlloc_Alloc(alloc, p->AlignOffset + size)) == NULL) + return False; + p->Size = size; + } + return True; +} + + + +// ---------- Internal Memory Allocator ---------- + +/* We can use CPpmd7_Node in list of free units (as in Ppmd8) + But we still need one additional list walk pass in GlueFreeBlocks(). + So we use simple CPpmd_Void_Ref instead of CPpmd7_Node in InsertNode() / RemoveNode() +*/ + +#define EMPTY_NODE 0 + + +static void InsertNode(CPpmd7 *p, void *node, unsigned indx) +{ + *((CPpmd_Void_Ref *)node) = p->FreeList[indx]; + // ((CPpmd7_Node *)node)->Next = (CPpmd7_Node_Ref)p->FreeList[indx]; + + p->FreeList[indx] = REF(node); + +} + + +static void *RemoveNode(CPpmd7 *p, unsigned indx) +{ + CPpmd_Void_Ref *node = (CPpmd_Void_Ref *)Ppmd7_GetPtr(p, p->FreeList[indx]); + p->FreeList[indx] = *node; + // CPpmd7_Node *node = NODE((CPpmd7_Node_Ref)p->FreeList[indx]); + // p->FreeList[indx] = node->Next; + return node; +} + + +static void SplitBlock(CPpmd7 *p, void *ptr, unsigned oldIndx, unsigned newIndx) +{ + unsigned i, nu = I2U(oldIndx) - I2U(newIndx); + ptr = (Byte *)ptr + U2B(I2U(newIndx)); + if (I2U(i = U2I(nu)) != nu) + { + unsigned k = I2U(--i); + InsertNode(p, ((Byte *)ptr) + U2B(k), nu - k - 1); + } + InsertNode(p, ptr, i); +} + + +/* we use CPpmd7_Node_Union union to solve XLC -O2 strict pointer aliasing problem */ + +typedef union _CPpmd7_Node_Union +{ + CPpmd7_Node Node; + CPpmd7_Node_Ref NextRef; +} CPpmd7_Node_Union; + +/* Original PPmdH (Ppmd7) code uses doubly linked list in GlueFreeBlocks() + we use single linked list similar to Ppmd8 code */ + + +static void GlueFreeBlocks(CPpmd7 *p) +{ + /* + we use first UInt16 field of 12-bytes UNITs as record type stamp + CPpmd_State { Byte Symbol; Byte Freq; : Freq != 0 + CPpmd7_Context { UInt16 NumStats; : NumStats != 0 + CPpmd7_Node { UInt16 Stamp : Stamp == 0 for free record + : Stamp == 1 for head record and guard + Last 12-bytes UNIT in array is always contains 12-bytes order-0 CPpmd7_Context record. + */ + CPpmd7_Node_Ref head, n = 0; + + p->GlueCount = 255; + + + /* we set guard NODE at LoUnit */ + if (p->LoUnit != p->HiUnit) + ((CPpmd7_Node *)(void *)p->LoUnit)->Stamp = 1; + + { + /* Create list of free blocks. + We still need one additional list walk pass before Glue. */ + unsigned i; + for (i = 0; i < PPMD_NUM_INDEXES; i++) + { + const UInt16 nu = I2U_UInt16(i); + CPpmd7_Node_Ref next = (CPpmd7_Node_Ref)p->FreeList[i]; + p->FreeList[i] = 0; + while (next != 0) + { + /* Don't change the order of the following commands: */ + CPpmd7_Node_Union *un = (CPpmd7_Node_Union *)NODE(next); + const CPpmd7_Node_Ref tmp = next; + next = un->NextRef; + un->Node.Stamp = EMPTY_NODE; + un->Node.NU = nu; + un->Node.Next = n; + n = tmp; + } + } + } + + head = n; + /* Glue and Fill must walk the list in same direction */ + { + /* Glue free blocks */ + CPpmd7_Node_Ref *prev = &head; + while (n) + { + CPpmd7_Node *node = NODE(n); + UInt32 nu = node->NU; + n = node->Next; + if (nu == 0) + { + *prev = n; + continue; + } + prev = &node->Next; + for (;;) + { + CPpmd7_Node *node2 = node + nu; + nu += node2->NU; + if (node2->Stamp != EMPTY_NODE || nu >= 0x10000) + break; + node->NU = (UInt16)nu; + node2->NU = 0; + } + } + } + + /* Fill lists of free blocks */ + for (n = head; n != 0;) + { + CPpmd7_Node *node = NODE(n); + UInt32 nu = node->NU; + unsigned i; + n = node->Next; + if (nu == 0) + continue; + for (; nu > 128; nu -= 128, node += 128) + InsertNode(p, node, PPMD_NUM_INDEXES - 1); + if (I2U(i = U2I(nu)) != nu) + { + unsigned k = I2U(--i); + InsertNode(p, node + k, (unsigned)nu - k - 1); + } + InsertNode(p, node, i); + } +} + + +MY_NO_INLINE +static void *AllocUnitsRare(CPpmd7 *p, unsigned indx) +{ + unsigned i; + + if (p->GlueCount == 0) + { + GlueFreeBlocks(p); + if (p->FreeList[indx] != 0) + return RemoveNode(p, indx); + } + + i = indx; + + do + { + if (++i == PPMD_NUM_INDEXES) + { + UInt32 numBytes = U2B(I2U(indx)); + Byte *us = p->UnitsStart; + p->GlueCount--; + return ((UInt32)(us - p->Text) > numBytes) ? (p->UnitsStart = us - numBytes) : NULL; + } + } + while (p->FreeList[i] == 0); + + { + void *block = RemoveNode(p, i); + SplitBlock(p, block, i, indx); + return block; + } +} + + +static void *AllocUnits(CPpmd7 *p, unsigned indx) +{ + if (p->FreeList[indx] != 0) + return RemoveNode(p, indx); + { + UInt32 numBytes = U2B(I2U(indx)); + Byte *lo = p->LoUnit; + if ((UInt32)(p->HiUnit - lo) >= numBytes) + { + p->LoUnit = lo + numBytes; + return lo; + } + } + return AllocUnitsRare(p, indx); +} + + +#define MyMem12Cpy(dest, src, num) \ + { UInt32 *d = (UInt32 *)dest; const UInt32 *z = (const UInt32 *)src; UInt32 n = num; \ + do { d[0] = z[0]; d[1] = z[1]; d[2] = z[2]; z += 3; d += 3; } while (--n); } + + +/* +static void *ShrinkUnits(CPpmd7 *p, void *oldPtr, unsigned oldNU, unsigned newNU) +{ + unsigned i0 = U2I(oldNU); + unsigned i1 = U2I(newNU); + if (i0 == i1) + return oldPtr; + if (p->FreeList[i1] != 0) + { + void *ptr = RemoveNode(p, i1); + MyMem12Cpy(ptr, oldPtr, newNU); + InsertNode(p, oldPtr, i0); + return ptr; + } + SplitBlock(p, oldPtr, i0, i1); + return oldPtr; +} +*/ + + +#define SUCCESSOR(p) Ppmd_GET_SUCCESSOR(p) +static void SetSuccessor(CPpmd_State *p, CPpmd_Void_Ref v) +{ + Ppmd_SET_SUCCESSOR(p, v); +} + + + +MY_NO_INLINE +static +void RestartModel(CPpmd7 *p) +{ + unsigned i, k; + + memset(p->FreeList, 0, sizeof(p->FreeList)); + + p->Text = p->Base + p->AlignOffset; + p->HiUnit = p->Text + p->Size; + p->LoUnit = p->UnitsStart = p->HiUnit - p->Size / 8 / UNIT_SIZE * 7 * UNIT_SIZE; + p->GlueCount = 0; + + p->OrderFall = p->MaxOrder; + p->RunLength = p->InitRL = -(Int32)((p->MaxOrder < 12) ? p->MaxOrder : 12) - 1; + p->PrevSuccess = 0; + + { + CPpmd7_Context *mc = (CTX_PTR)(void *)(p->HiUnit -= UNIT_SIZE); /* AllocContext(p); */ + CPpmd_State *s = (CPpmd_State *)p->LoUnit; /* AllocUnits(p, PPMD_NUM_INDEXES - 1); */ + + p->LoUnit += U2B(256 / 2); + p->MaxContext = p->MinContext = mc; + p->FoundState = s; + + mc->NumStats = 256; + mc->Union2.SummFreq = 256 + 1; + mc->Union4.Stats = REF(s); + mc->Suffix = 0; + + for (i = 0; i < 256; i++, s++) + { + s->Symbol = (Byte)i; + s->Freq = 1; + SetSuccessor(s, 0); + } + + #ifdef PPMD7_ORDER_0_SUPPPORT + if (p->MaxOrder == 0) + { + CPpmd_Void_Ref r = REF(mc); + s = p->FoundState; + for (i = 0; i < 256; i++, s++) + SetSuccessor(s, r); + return; + } + #endif + } + + for (i = 0; i < 128; i++) + + + + for (k = 0; k < 8; k++) + { + unsigned m; + UInt16 *dest = p->BinSumm[i] + k; + UInt16 val = (UInt16)(PPMD_BIN_SCALE - kInitBinEsc[k] / (i + 2)); + for (m = 0; m < 64; m += 8) + dest[m] = val; + } + + + for (i = 0; i < 25; i++) + { + + CPpmd_See *s = p->See[i]; + + + + unsigned summ = ((5 * i + 10) << (PPMD_PERIOD_BITS - 4)); + for (k = 0; k < 16; k++, s++) + { + s->Summ = (UInt16)summ; + s->Shift = (PPMD_PERIOD_BITS - 4); + s->Count = 4; + } + } + + p->DummySee.Summ = 0; /* unused */ + p->DummySee.Shift = PPMD_PERIOD_BITS; + p->DummySee.Count = 64; /* unused */ +} + + +void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder) +{ + p->MaxOrder = maxOrder; + + RestartModel(p); +} + + + +/* + CreateSuccessors() + It's called when (FoundState->Successor) is RAW-Successor, + that is the link to position in Raw text. + So we create Context records and write the links to + FoundState->Successor and to identical RAW-Successors in suffix + contexts of MinContex. + + The function returns: + if (OrderFall == 0) then MinContext is already at MAX order, + { return pointer to new or existing context of same MAX order } + else + { return pointer to new real context that will be (Order+1) in comparison with MinContext + + also it can return pointer to real context of same order, +*/ + +MY_NO_INLINE +static CTX_PTR CreateSuccessors(CPpmd7 *p) +{ + CTX_PTR c = p->MinContext; + CPpmd_Byte_Ref upBranch = (CPpmd_Byte_Ref)SUCCESSOR(p->FoundState); + Byte newSym, newFreq; + unsigned numPs = 0; + CPpmd_State *ps[PPMD7_MAX_ORDER]; + + if (p->OrderFall != 0) + ps[numPs++] = p->FoundState; + + while (c->Suffix) + { + CPpmd_Void_Ref successor; + CPpmd_State *s; + c = SUFFIX(c); + + + if (c->NumStats != 1) + { + Byte sym = p->FoundState->Symbol; + for (s = STATS(c); s->Symbol != sym; s++); + + } + else + { + s = ONE_STATE(c); + + } + successor = SUCCESSOR(s); + if (successor != upBranch) + { + // (c) is real record Context here, + c = CTX(successor); + if (numPs == 0) + { + // (c) is real record MAX Order Context here, + // So we don't need to create any new contexts. + return c; + } + break; + } + ps[numPs++] = s; + } + + // All created contexts will have single-symbol with new RAW-Successor + // All new RAW-Successors will point to next position in RAW text + // after FoundState->Successor + + newSym = *(const Byte *)Ppmd7_GetPtr(p, upBranch); + upBranch++; + + + if (c->NumStats == 1) + newFreq = ONE_STATE(c)->Freq; + else + { + UInt32 cf, s0; + CPpmd_State *s; + for (s = STATS(c); s->Symbol != newSym; s++); + cf = (UInt32)s->Freq - 1; + s0 = (UInt32)c->Union2.SummFreq - c->NumStats - cf; + /* + cf - is frequency of symbol that will be Successor in new context records. + s0 - is commulative frequency sum of another symbols from parent context. + max(newFreq)= (s->Freq + 1), when (s0 == 1) + we have requirement (Ppmd7Context_OneState()->Freq <= 128) in BinSumm[] + so (s->Freq < 128) - is requirement for multi-symbol contexts + */ + newFreq = (Byte)(1 + ((2 * cf <= s0) ? (5 * cf > s0) : (2 * cf + s0 - 1) / (2 * s0) + 1)); + } + + // Create new single-symbol contexts from low order to high order in loop + + do + { + CTX_PTR c1; + /* = AllocContext(p); */ + if (p->HiUnit != p->LoUnit) + c1 = (CTX_PTR)(void *)(p->HiUnit -= UNIT_SIZE); + else if (p->FreeList[0] != 0) + c1 = (CTX_PTR)RemoveNode(p, 0); + else + { + c1 = (CTX_PTR)AllocUnitsRare(p, 0); + if (!c1) + return NULL; + } + + c1->NumStats = 1; + ONE_STATE(c1)->Symbol = newSym; + ONE_STATE(c1)->Freq = newFreq; + SetSuccessor(ONE_STATE(c1), upBranch); + c1->Suffix = REF(c); + SetSuccessor(ps[--numPs], REF(c1)); + c = c1; + } + while (numPs != 0); + + return c; +} + + + +#define SwapStates(s) \ + { CPpmd_State tmp = s[0]; s[0] = s[-1]; s[-1] = tmp; } + + +void Ppmd7_UpdateModel(CPpmd7 *p); +MY_NO_INLINE +void Ppmd7_UpdateModel(CPpmd7 *p) +{ + CPpmd_Void_Ref maxSuccessor, minSuccessor; + CTX_PTR c, mc; + unsigned s0, ns; + + + + if (p->FoundState->Freq < MAX_FREQ / 4 && p->MinContext->Suffix != 0) + { + /* Update Freqs in Suffix Context */ + + c = SUFFIX(p->MinContext); + + if (c->NumStats == 1) + { + CPpmd_State *s = ONE_STATE(c); + if (s->Freq < 32) + s->Freq++; + } + else + { + CPpmd_State *s = STATS(c); + Byte sym = p->FoundState->Symbol; + + if (s->Symbol != sym) + { + do + { + // s++; if (s->Symbol == sym) break; + s++; + } + while (s->Symbol != sym); + + if (s[0].Freq >= s[-1].Freq) + { + SwapStates(s); + s--; + } + } + + if (s->Freq < MAX_FREQ - 9) + { + s->Freq = (Byte)(s->Freq + 2); + c->Union2.SummFreq = (UInt16)(c->Union2.SummFreq + 2); + } + } + } + + + if (p->OrderFall == 0) + { + /* MAX ORDER context */ + /* (FoundState->Successor) is RAW-Successor. */ + p->MaxContext = p->MinContext = CreateSuccessors(p); + if (!p->MinContext) + { + RestartModel(p); + return; + } + SetSuccessor(p->FoundState, REF(p->MinContext)); + return; + } + + + /* NON-MAX ORDER context */ + + { + Byte *text = p->Text; + *text++ = p->FoundState->Symbol; + p->Text = text; + if (text >= p->UnitsStart) + { + RestartModel(p); + return; + } + maxSuccessor = REF(text); + } + + minSuccessor = SUCCESSOR(p->FoundState); + + if (minSuccessor) + { + // there is Successor for FoundState in MinContext. + // So the next context will be one order higher than MinContext. + + if (minSuccessor <= maxSuccessor) + { + // minSuccessor is RAW-Successor. So we will create real contexts records: + CTX_PTR cs = CreateSuccessors(p); + if (!cs) + { + RestartModel(p); + return; + } + minSuccessor = REF(cs); + } + + // minSuccessor now is real Context pointer that points to existing (Order+1) context + + if (--p->OrderFall == 0) + { + /* + if we move to MaxOrder context, then minSuccessor will be common Succesor for both: + MinContext that is (MaxOrder - 1) + MaxContext that is (MaxOrder) + so we don't need new RAW-Successor, and we can use real minSuccessor + as succssors for both MinContext and MaxContext. + */ + maxSuccessor = minSuccessor; + + /* + if (MaxContext != MinContext) + { + there was order fall from MaxOrder and we don't need current symbol + to transfer some RAW-Succesors to real contexts. + So we roll back pointer in raw data for one position. + } + */ + p->Text -= (p->MaxContext != p->MinContext); + } + } + else + { + /* + FoundState has NULL-Successor here. + And only root 0-order context can contain NULL-Successors. + We change Successor in FoundState to RAW-Successor, + And next context will be same 0-order root Context. + */ + SetSuccessor(p->FoundState, maxSuccessor); + minSuccessor = REF(p->MinContext); + } + + mc = p->MinContext; + c = p->MaxContext; + + p->MaxContext = p->MinContext = CTX(minSuccessor); + + if (c == mc) + return; + + // s0 : is pure Escape Freq + s0 = mc->Union2.SummFreq - (ns = mc->NumStats) - ((unsigned)p->FoundState->Freq - 1); + + do + { + unsigned ns1; + UInt32 sum; + + if ((ns1 = c->NumStats) != 1) + { + if ((ns1 & 1) == 0) + { + /* Expand for one UNIT */ + unsigned oldNU = ns1 >> 1; + unsigned i = U2I(oldNU); + if (i != U2I((size_t)oldNU + 1)) + { + void *ptr = AllocUnits(p, i + 1); + void *oldPtr; + if (!ptr) + { + RestartModel(p); + return; + } + oldPtr = STATS(c); + MyMem12Cpy(ptr, oldPtr, oldNU); + InsertNode(p, oldPtr, i); + c->Union4.Stats = STATS_REF(ptr); + } + } + sum = c->Union2.SummFreq; + /* max increase of Escape_Freq is 3 here. + total increase of Union2.SummFreq for all symbols is less than 256 here */ + sum += (UInt32)(2 * ns1 < ns) + 2 * ((unsigned)(4 * ns1 <= ns) & (sum <= 8 * ns1)); + /* original PPMdH uses 16-bit variable for (sum) here. + But (sum < 0x9000). So we don't truncate (sum) to 16-bit */ + // sum = (UInt16)sum; + } + else + { + // instead of One-symbol context we create 2-symbol context + CPpmd_State *s = (CPpmd_State*)AllocUnits(p, 0); + if (!s) + { + RestartModel(p); + return; + } + { + unsigned freq = c->Union2.State2.Freq; + // s = *ONE_STATE(c); + s->Symbol = c->Union2.State2.Symbol; + s->Successor_0 = c->Union4.State4.Successor_0; + s->Successor_1 = c->Union4.State4.Successor_1; + // SetSuccessor(s, c->Union4.Stats); // call it only for debug purposes to check the order of + // (Successor_0 and Successor_1) in LE/BE. + c->Union4.Stats = REF(s); + if (freq < MAX_FREQ / 4 - 1) + freq <<= 1; + else + freq = MAX_FREQ - 4; + // (max(s->freq) == 120), when we convert from 1-symbol into 2-symbol context + s->Freq = (Byte)freq; + // max(InitEsc = PPMD7_kExpEscape[*]) is 25. So the max(escapeFreq) is 26 here + sum = freq + p->InitEsc + (ns > 3); + } + } + + { + CPpmd_State *s = STATS(c) + ns1; + UInt32 cf = 2 * (sum + 6) * (UInt32)p->FoundState->Freq; + UInt32 sf = (UInt32)s0 + sum; + s->Symbol = p->FoundState->Symbol; + c->NumStats = (UInt16)(ns1 + 1); + SetSuccessor(s, maxSuccessor); + + if (cf < 6 * sf) + { + cf = (UInt32)1 + (cf > sf) + (cf >= 4 * sf); + sum += 3; + /* It can add (0, 1, 2) to Escape_Freq */ + } + else + { + cf = (UInt32)4 + (cf >= 9 * sf) + (cf >= 12 * sf) + (cf >= 15 * sf); + sum += cf; + } + + c->Union2.SummFreq = (UInt16)sum; + s->Freq = (Byte)cf; + } + c = SUFFIX(c); + } + while (c != mc); +} + + + +MY_NO_INLINE +static void Rescale(CPpmd7 *p) +{ + unsigned i, adder, sumFreq, escFreq; + CPpmd_State *stats = STATS(p->MinContext); + CPpmd_State *s = p->FoundState; + + /* Sort the list by Freq */ + if (s != stats) + { + CPpmd_State tmp = *s; + do + s[0] = s[-1]; + while (--s != stats); + *s = tmp; + } + + sumFreq = s->Freq; + escFreq = p->MinContext->Union2.SummFreq - sumFreq; + + /* + if (p->OrderFall == 0), adder = 0 : it's allowed to remove symbol from MAX Order context + if (p->OrderFall != 0), adder = 1 : it's NOT allowed to remove symbol from NON-MAX Order context + */ + + adder = (p->OrderFall != 0); + + #ifdef PPMD7_ORDER_0_SUPPPORT + adder |= (p->MaxOrder == 0); // we don't remove symbols from order-0 context + #endif + + sumFreq = (sumFreq + 4 + adder) >> 1; + i = (unsigned)p->MinContext->NumStats - 1; + s->Freq = (Byte)sumFreq; + + do + { + unsigned freq = (++s)->Freq; + escFreq -= freq; + freq = (freq + adder) >> 1; + sumFreq += freq; + s->Freq = (Byte)freq; + if (freq > s[-1].Freq) + { + CPpmd_State tmp = *s; + CPpmd_State *s1 = s; + do + { + s1[0] = s1[-1]; + } + while (--s1 != stats && freq > s1[-1].Freq); + *s1 = tmp; + } + } + while (--i); + + if (s->Freq == 0) + { + /* Remove all items with Freq == 0 */ + CPpmd7_Context *mc; + unsigned numStats, numStatsNew, n0, n1; + + i = 0; do { i++; } while ((--s)->Freq == 0); + + /* We increase (escFreq) for the number of removed symbols. + So we will have (0.5) increase for Escape_Freq in avarage per + removed symbol after Escape_Freq halving */ + escFreq += i; + mc = p->MinContext; + numStats = mc->NumStats; + numStatsNew = numStats - i; + mc->NumStats = (UInt16)(numStatsNew); + n0 = (numStats + 1) >> 1; + + if (numStatsNew == 1) + { + /* Create Single-Symbol context */ + unsigned freq = stats->Freq; + + do + { + escFreq >>= 1; + freq = (freq + 1) >> 1; + } + while (escFreq > 1); + + s = ONE_STATE(mc); + *s = *stats; + s->Freq = (Byte)freq; // (freq <= 260 / 4) + p->FoundState = s; + InsertNode(p, stats, U2I(n0)); + return; + } + + n1 = (numStatsNew + 1) >> 1; + if (n0 != n1) + { + // p->MinContext->Union4.Stats = STATS_REF(ShrinkUnits(p, stats, n0, n1)); + unsigned i0 = U2I(n0); + unsigned i1 = U2I(n1); + if (i0 != i1) + { + if (p->FreeList[i1] != 0) + { + void *ptr = RemoveNode(p, i1); + p->MinContext->Union4.Stats = STATS_REF(ptr); + MyMem12Cpy(ptr, (const void *)stats, n1); + InsertNode(p, stats, i0); + } + else + SplitBlock(p, stats, i0, i1); + } + } + } + { + CPpmd7_Context *mc = p->MinContext; + mc->Union2.SummFreq = (UInt16)(sumFreq + escFreq - (escFreq >> 1)); + // Escape_Freq halving here + p->FoundState = STATS(mc); + } +} + + +CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *escFreq) +{ + CPpmd_See *see; + const CPpmd7_Context *mc = p->MinContext; + unsigned numStats = mc->NumStats; + if (numStats != 256) + { + unsigned nonMasked = numStats - numMasked; + see = p->See[(unsigned)p->NS2Indx[(size_t)nonMasked - 1]] + + (nonMasked < (unsigned)SUFFIX(mc)->NumStats - numStats) + + 2 * (unsigned)(mc->Union2.SummFreq < 11 * numStats) + + 4 * (unsigned)(numMasked > nonMasked) + + p->HiBitsFlag; + { + // if (see->Summ) field is larger than 16-bit, we need only low 16 bits of Summ + unsigned summ = (UInt16)see->Summ; // & 0xFFFF + unsigned r = (summ >> see->Shift); + see->Summ = (UInt16)(summ - r); + *escFreq = r + (r == 0); + } + } + else + { + see = &p->DummySee; + *escFreq = 1; + } + return see; +} + + +static void NextContext(CPpmd7 *p) +{ + CTX_PTR c = CTX(SUCCESSOR(p->FoundState)); + if (p->OrderFall == 0 && (const Byte *)c > p->Text) + p->MaxContext = p->MinContext = c; + else + Ppmd7_UpdateModel(p); +} + + +void Ppmd7_Update1(CPpmd7 *p) +{ + CPpmd_State *s = p->FoundState; + unsigned freq = s->Freq; + freq += 4; + p->MinContext->Union2.SummFreq = (UInt16)(p->MinContext->Union2.SummFreq + 4); + s->Freq = (Byte)freq; + if (freq > s[-1].Freq) + { + SwapStates(s); + p->FoundState = --s; + if (freq > MAX_FREQ) + Rescale(p); + } + NextContext(p); +} + + +void Ppmd7_Update1_0(CPpmd7 *p) +{ + CPpmd_State *s = p->FoundState; + CPpmd7_Context *mc = p->MinContext; + unsigned freq = s->Freq; + unsigned summFreq = mc->Union2.SummFreq; + p->PrevSuccess = (2 * freq > summFreq); + p->RunLength += (int)p->PrevSuccess; + mc->Union2.SummFreq = (UInt16)(summFreq + 4); + freq += 4; + s->Freq = (Byte)freq; + if (freq > MAX_FREQ) + Rescale(p); + NextContext(p); +} + + +/* +void Ppmd7_UpdateBin(CPpmd7 *p) +{ + unsigned freq = p->FoundState->Freq; + p->FoundState->Freq = (Byte)(freq + (freq < 128)); + p->PrevSuccess = 1; + p->RunLength++; + NextContext(p); +} +*/ + +void Ppmd7_Update2(CPpmd7 *p) +{ + CPpmd_State *s = p->FoundState; + unsigned freq = s->Freq; + freq += 4; + p->RunLength = p->InitRL; + p->MinContext->Union2.SummFreq = (UInt16)(p->MinContext->Union2.SummFreq + 4); + s->Freq = (Byte)freq; + if (freq > MAX_FREQ) + Rescale(p); + Ppmd7_UpdateModel(p); +} + + + +/* +PPMd Memory Map: +{ + [ 0 ] contains subset of original raw text, that is required to create context + records, Some symbols are not written, when max order context was reached + [ Text ] free area + [ UnitsStart ] CPpmd_State vectors and CPpmd7_Context records + [ LoUnit ] free area for CPpmd_State and CPpmd7_Context items +[ HiUnit ] CPpmd7_Context records + [ Size ] end of array +} + +These addresses don't cross at any time. +And the following condtions is true for addresses: + (0 <= Text < UnitsStart <= LoUnit <= HiUnit <= Size) + +Raw text is BYTE--aligned. +the data in block [ UnitsStart ... Size ] contains 12-bytes aligned UNITs. + +Last UNIT of array at offset (Size - 12) is root order-0 CPpmd7_Context record. +The code can free UNITs memory blocks that were allocated to store CPpmd_State vectors. +The code doesn't free UNITs allocated for CPpmd7_Context records. + +The code calls RestartModel(), when there is no free memory for allocation. +And RestartModel() changes the state to orignal start state, with full free block. + + +The code allocates UNITs with the following order: + +Allocation of 1 UNIT for Context record + - from free space (HiUnit) down to (LoUnit) + - from FreeList[0] + - AllocUnitsRare() + +AllocUnits() for CPpmd_State vectors: + - from FreeList[i] + - from free space (LoUnit) up to (HiUnit) + - AllocUnitsRare() + +AllocUnitsRare() + - if (GlueCount == 0) + { Glue lists, GlueCount = 255, allocate from FreeList[i]] } + - loop for all higher sized FreeList[...] lists + - from (UnitsStart - Text), GlueCount-- + - ERROR + + +Each Record with Context contains the CPpmd_State vector, where each +CPpmd_State contains the link to Successor. +There are 3 types of Successor: + 1) NULL-Successor - NULL pointer. NULL-Successor links can be stored + only in 0-order Root Context Record. + We use 0 value as NULL-Successor + 2) RAW-Successor - the link to position in raw text, + that "RAW-Successor" is being created after first + occurrence of new symbol for some existing context record. + (RAW-Successor > 0). + 3) RECORD-Successor - the link to CPpmd7_Context record of (Order+1), + that record is being created when we go via RAW-Successor again. + +For any successors at any time: the following condtions are true for Successor links: +(NULL-Successor < RAW-Successor < UnitsStart <= RECORD-Successor) + + +---------- Symbol Frequency, SummFreq and Range in Range_Coder ---------- + +CPpmd7_Context::SummFreq = Sum(Stats[].Freq) + Escape_Freq + +The PPMd code tries to fulfill the condition: + (SummFreq <= (256 * 128 = RC::kBot)) + +We have (Sum(Stats[].Freq) <= 256 * 124), because of (MAX_FREQ = 124) +So (4 = 128 - 124) is average reserve for Escape_Freq for each symbol. +If (CPpmd_State::Freq) is not aligned for 4, the reserve can be 5, 6 or 7. +SummFreq and Escape_Freq can be changed in Rescale() and *Update*() functions. +Rescale() can remove symbols only from max-order contexts. So Escape_Freq can increase after multiple calls of Rescale() for +max-order context. + +When the PPMd code still break (Total <= RC::Range) condition in range coder, +we have two ways to resolve that problem: + 1) we can report error, if we want to keep compatibility with original PPMd code that has no fix for such cases. + 2) we can reduce (Total) value to (RC::Range) by reducing (Escape_Freq) part of (Total) value. +*/ diff --git a/3rdparty/lzma/src/Ppmd7Dec.c b/3rdparty/lzma/src/Ppmd7Dec.c new file mode 100644 index 0000000000..55d74ff9dc --- /dev/null +++ b/3rdparty/lzma/src/Ppmd7Dec.c @@ -0,0 +1,297 @@ +/* Ppmd7Dec.c -- Ppmd7z (PPMdH with 7z Range Coder) Decoder +2021-04-13 : Igor Pavlov : Public domain +This code is based on: + PPMd var.H (2001): Dmitry Shkarin : Public domain */ + + +#include "Precomp.h" + +#include "Ppmd7.h" + +#define kTopValue (1 << 24) + + +#define READ_BYTE(p) IByteIn_Read((p)->Stream) + +BoolInt Ppmd7z_RangeDec_Init(CPpmd7_RangeDec *p) +{ + unsigned i; + p->Code = 0; + p->Range = 0xFFFFFFFF; + if (READ_BYTE(p) != 0) + return False; + for (i = 0; i < 4; i++) + p->Code = (p->Code << 8) | READ_BYTE(p); + return (p->Code < 0xFFFFFFFF); +} + +#define RC_NORM_BASE(p) if ((p)->Range < kTopValue) \ + { (p)->Code = ((p)->Code << 8) | READ_BYTE(p); (p)->Range <<= 8; + +#define RC_NORM_1(p) RC_NORM_BASE(p) } +#define RC_NORM(p) RC_NORM_BASE(p) RC_NORM_BASE(p) }} + +// we must use only one type of Normalization from two: LOCAL or REMOTE +#define RC_NORM_LOCAL(p) // RC_NORM(p) +#define RC_NORM_REMOTE(p) RC_NORM(p) + +#define R (&p->rc.dec) + +MY_FORCE_INLINE +// MY_NO_INLINE +static void RangeDec_Decode(CPpmd7 *p, UInt32 start, UInt32 size) +{ + + + R->Code -= start * R->Range; + R->Range *= size; + RC_NORM_LOCAL(R) +} + +#define RC_Decode(start, size) RangeDec_Decode(p, start, size); +#define RC_DecodeFinal(start, size) RC_Decode(start, size) RC_NORM_REMOTE(R) +#define RC_GetThreshold(total) (R->Code / (R->Range /= (total))) + + +#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref)) +typedef CPpmd7_Context * CTX_PTR; +#define SUCCESSOR(p) Ppmd_GET_SUCCESSOR(p) +void Ppmd7_UpdateModel(CPpmd7 *p); + +#define MASK(sym) ((unsigned char *)charMask)[sym] +// MY_FORCE_INLINE +// static +int Ppmd7z_DecodeSymbol(CPpmd7 *p) +{ + size_t charMask[256 / sizeof(size_t)]; + + if (p->MinContext->NumStats != 1) + { + CPpmd_State *s = Ppmd7_GetStats(p, p->MinContext); + unsigned i; + UInt32 count, hiCnt; + UInt32 summFreq = p->MinContext->Union2.SummFreq; + + + + + count = RC_GetThreshold(summFreq); + hiCnt = count; + + if ((Int32)(count -= s->Freq) < 0) + { + Byte sym; + RC_DecodeFinal(0, s->Freq); + p->FoundState = s; + sym = s->Symbol; + Ppmd7_Update1_0(p); + return sym; + } + + p->PrevSuccess = 0; + i = (unsigned)p->MinContext->NumStats - 1; + + do + { + if ((Int32)(count -= (++s)->Freq) < 0) + { + Byte sym; + RC_DecodeFinal((hiCnt - count) - s->Freq, s->Freq); + p->FoundState = s; + sym = s->Symbol; + Ppmd7_Update1(p); + return sym; + } + } + while (--i); + + if (hiCnt >= summFreq) + return PPMD7_SYM_ERROR; + + hiCnt -= count; + RC_Decode(hiCnt, summFreq - hiCnt); + + p->HiBitsFlag = PPMD7_HiBitsFlag_3(p->FoundState->Symbol); + PPMD_SetAllBitsIn256Bytes(charMask); + // i = p->MinContext->NumStats - 1; + // do { MASK((--s)->Symbol) = 0; } while (--i); + { + CPpmd_State *s2 = Ppmd7_GetStats(p, p->MinContext); + MASK(s->Symbol) = 0; + do + { + unsigned sym0 = s2[0].Symbol; + unsigned sym1 = s2[1].Symbol; + s2 += 2; + MASK(sym0) = 0; + MASK(sym1) = 0; + } + while (s2 < s); + } + } + else + { + CPpmd_State *s = Ppmd7Context_OneState(p->MinContext); + UInt16 *prob = Ppmd7_GetBinSumm(p); + UInt32 pr = *prob; + UInt32 size0 = (R->Range >> 14) * pr; + pr = PPMD_UPDATE_PROB_1(pr); + + if (R->Code < size0) + { + Byte sym; + *prob = (UInt16)(pr + (1 << PPMD_INT_BITS)); + + // RangeDec_DecodeBit0(size0); + R->Range = size0; + RC_NORM_1(R) + /* we can use single byte normalization here because of + (min(BinSumm[][]) = 95) > (1 << (14 - 8)) */ + + // sym = (p->FoundState = Ppmd7Context_OneState(p->MinContext))->Symbol; + // Ppmd7_UpdateBin(p); + { + unsigned freq = s->Freq; + CTX_PTR c = CTX(SUCCESSOR(s)); + sym = s->Symbol; + p->FoundState = s; + p->PrevSuccess = 1; + p->RunLength++; + s->Freq = (Byte)(freq + (freq < 128)); + // NextContext(p); + if (p->OrderFall == 0 && (const Byte *)c > p->Text) + p->MaxContext = p->MinContext = c; + else + Ppmd7_UpdateModel(p); + } + return sym; + } + + *prob = (UInt16)pr; + p->InitEsc = p->ExpEscape[pr >> 10]; + + // RangeDec_DecodeBit1(size0); + + R->Code -= size0; + R->Range -= size0; + RC_NORM_LOCAL(R) + + PPMD_SetAllBitsIn256Bytes(charMask); + MASK(Ppmd7Context_OneState(p->MinContext)->Symbol) = 0; + p->PrevSuccess = 0; + } + + for (;;) + { + CPpmd_State *s, *s2; + UInt32 freqSum, count, hiCnt; + + CPpmd_See *see; + CPpmd7_Context *mc; + unsigned numMasked; + RC_NORM_REMOTE(R) + mc = p->MinContext; + numMasked = mc->NumStats; + + do + { + p->OrderFall++; + if (!mc->Suffix) + return PPMD7_SYM_END; + mc = Ppmd7_GetContext(p, mc->Suffix); + } + while (mc->NumStats == numMasked); + + s = Ppmd7_GetStats(p, mc); + + { + unsigned num = mc->NumStats; + unsigned num2 = num / 2; + + num &= 1; + hiCnt = (s->Freq & (unsigned)(MASK(s->Symbol))) & (0 - (UInt32)num); + s += num; + p->MinContext = mc; + + do + { + unsigned sym0 = s[0].Symbol; + unsigned sym1 = s[1].Symbol; + s += 2; + hiCnt += (s[-2].Freq & (unsigned)(MASK(sym0))); + hiCnt += (s[-1].Freq & (unsigned)(MASK(sym1))); + } + while (--num2); + } + + see = Ppmd7_MakeEscFreq(p, numMasked, &freqSum); + freqSum += hiCnt; + + + + + count = RC_GetThreshold(freqSum); + + if (count < hiCnt) + { + Byte sym; + + s = Ppmd7_GetStats(p, p->MinContext); + hiCnt = count; + // count -= s->Freq & (unsigned)(MASK(s->Symbol)); + // if ((Int32)count >= 0) + { + for (;;) + { + count -= s->Freq & (unsigned)(MASK((s)->Symbol)); s++; if ((Int32)count < 0) break; + // count -= s->Freq & (unsigned)(MASK((s)->Symbol)); s++; if ((Int32)count < 0) break; + }; + } + s--; + RC_DecodeFinal((hiCnt - count) - s->Freq, s->Freq); + + // new (see->Summ) value can overflow over 16-bits in some rare cases + Ppmd_See_Update(see); + p->FoundState = s; + sym = s->Symbol; + Ppmd7_Update2(p); + return sym; + } + + if (count >= freqSum) + return PPMD7_SYM_ERROR; + + RC_Decode(hiCnt, freqSum - hiCnt); + + // We increase (see->Summ) for sum of Freqs of all non_Masked symbols. + // new (see->Summ) value can overflow over 16-bits in some rare cases + see->Summ = (UInt16)(see->Summ + freqSum); + + s = Ppmd7_GetStats(p, p->MinContext); + s2 = s + p->MinContext->NumStats; + do + { + MASK(s->Symbol) = 0; + s++; + } + while (s != s2); + } +} + +/* +Byte *Ppmd7z_DecodeSymbols(CPpmd7 *p, Byte *buf, const Byte *lim) +{ + int sym = 0; + if (buf != lim) + do + { + sym = Ppmd7z_DecodeSymbol(p); + if (sym < 0) + break; + *buf = (Byte)sym; + } + while (++buf < lim); + p->LastSymbol = sym; + return buf; +} +*/ diff --git a/3rdparty/lzma/src/Ppmd7Enc.c b/3rdparty/lzma/src/Ppmd7Enc.c new file mode 100644 index 0000000000..62139c5b66 --- /dev/null +++ b/3rdparty/lzma/src/Ppmd7Enc.c @@ -0,0 +1,323 @@ +/* Ppmd7Enc.c -- Ppmd7z (PPMdH with 7z Range Coder) Encoder +2021-04-13 : Igor Pavlov : Public domain +This code is based on: + PPMd var.H (2001): Dmitry Shkarin : Public domain */ + + +#include "Precomp.h" + +#include "Ppmd7.h" + +#define kTopValue (1 << 24) + +#define R (&p->rc.enc) + +void Ppmd7z_Init_RangeEnc(CPpmd7 *p) +{ + R->Low = 0; + R->Range = 0xFFFFFFFF; + R->Cache = 0; + R->CacheSize = 1; +} + +MY_NO_INLINE +static void RangeEnc_ShiftLow(CPpmd7 *p) +{ + if ((UInt32)R->Low < (UInt32)0xFF000000 || (unsigned)(R->Low >> 32) != 0) + { + Byte temp = R->Cache; + do + { + IByteOut_Write(R->Stream, (Byte)(temp + (Byte)(R->Low >> 32))); + temp = 0xFF; + } + while (--R->CacheSize != 0); + R->Cache = (Byte)((UInt32)R->Low >> 24); + } + R->CacheSize++; + R->Low = (UInt32)((UInt32)R->Low << 8); +} + +#define RC_NORM_BASE(p) if (R->Range < kTopValue) { R->Range <<= 8; RangeEnc_ShiftLow(p); +#define RC_NORM_1(p) RC_NORM_BASE(p) } +#define RC_NORM(p) RC_NORM_BASE(p) RC_NORM_BASE(p) }} + +// we must use only one type of Normalization from two: LOCAL or REMOTE +#define RC_NORM_LOCAL(p) // RC_NORM(p) +#define RC_NORM_REMOTE(p) RC_NORM(p) + +/* +#define RangeEnc_Encode(p, start, _size_) \ + { UInt32 size = _size_; \ + R->Low += start * R->Range; \ + R->Range *= size; \ + RC_NORM_LOCAL(p); } +*/ + +MY_FORCE_INLINE +// MY_NO_INLINE +static void RangeEnc_Encode(CPpmd7 *p, UInt32 start, UInt32 size) +{ + R->Low += start * R->Range; + R->Range *= size; + RC_NORM_LOCAL(p); +} + +void Ppmd7z_Flush_RangeEnc(CPpmd7 *p) +{ + unsigned i; + for (i = 0; i < 5; i++) + RangeEnc_ShiftLow(p); +} + + + +#define RC_Encode(start, size) RangeEnc_Encode(p, start, size); +#define RC_EncodeFinal(start, size) RC_Encode(start, size); RC_NORM_REMOTE(p); + +#define CTX(ref) ((CPpmd7_Context *)Ppmd7_GetContext(p, ref)) +#define SUFFIX(ctx) CTX((ctx)->Suffix) +typedef CPpmd7_Context * CTX_PTR; +#define SUCCESSOR(p) Ppmd_GET_SUCCESSOR(p) + +void Ppmd7_UpdateModel(CPpmd7 *p); + +#define MASK(sym) ((unsigned char *)charMask)[sym] + +MY_FORCE_INLINE +static +void Ppmd7z_EncodeSymbol(CPpmd7 *p, int symbol) +{ + size_t charMask[256 / sizeof(size_t)]; + + if (p->MinContext->NumStats != 1) + { + CPpmd_State *s = Ppmd7_GetStats(p, p->MinContext); + UInt32 sum; + unsigned i; + + + + + R->Range /= p->MinContext->Union2.SummFreq; + + if (s->Symbol == symbol) + { + // R->Range /= p->MinContext->Union2.SummFreq; + RC_EncodeFinal(0, s->Freq); + p->FoundState = s; + Ppmd7_Update1_0(p); + return; + } + p->PrevSuccess = 0; + sum = s->Freq; + i = (unsigned)p->MinContext->NumStats - 1; + do + { + if ((++s)->Symbol == symbol) + { + // R->Range /= p->MinContext->Union2.SummFreq; + RC_EncodeFinal(sum, s->Freq); + p->FoundState = s; + Ppmd7_Update1(p); + return; + } + sum += s->Freq; + } + while (--i); + + // R->Range /= p->MinContext->Union2.SummFreq; + RC_Encode(sum, p->MinContext->Union2.SummFreq - sum); + + p->HiBitsFlag = PPMD7_HiBitsFlag_3(p->FoundState->Symbol); + PPMD_SetAllBitsIn256Bytes(charMask); + // MASK(s->Symbol) = 0; + // i = p->MinContext->NumStats - 1; + // do { MASK((--s)->Symbol) = 0; } while (--i); + { + CPpmd_State *s2 = Ppmd7_GetStats(p, p->MinContext); + MASK(s->Symbol) = 0; + do + { + unsigned sym0 = s2[0].Symbol; + unsigned sym1 = s2[1].Symbol; + s2 += 2; + MASK(sym0) = 0; + MASK(sym1) = 0; + } + while (s2 < s); + } + } + else + { + UInt16 *prob = Ppmd7_GetBinSumm(p); + CPpmd_State *s = Ppmd7Context_OneState(p->MinContext); + UInt32 pr = *prob; + UInt32 bound = (R->Range >> 14) * pr; + pr = PPMD_UPDATE_PROB_1(pr); + if (s->Symbol == symbol) + { + *prob = (UInt16)(pr + (1 << PPMD_INT_BITS)); + // RangeEnc_EncodeBit_0(p, bound); + R->Range = bound; + RC_NORM_1(p); + + // p->FoundState = s; + // Ppmd7_UpdateBin(p); + { + unsigned freq = s->Freq; + CTX_PTR c = CTX(SUCCESSOR(s)); + p->FoundState = s; + p->PrevSuccess = 1; + p->RunLength++; + s->Freq = (Byte)(freq + (freq < 128)); + // NextContext(p); + if (p->OrderFall == 0 && (const Byte *)c > p->Text) + p->MaxContext = p->MinContext = c; + else + Ppmd7_UpdateModel(p); + } + return; + } + + *prob = (UInt16)pr; + p->InitEsc = p->ExpEscape[pr >> 10]; + // RangeEnc_EncodeBit_1(p, bound); + R->Low += bound; + R->Range -= bound; + RC_NORM_LOCAL(p) + + PPMD_SetAllBitsIn256Bytes(charMask); + MASK(s->Symbol) = 0; + p->PrevSuccess = 0; + } + + for (;;) + { + CPpmd_See *see; + CPpmd_State *s; + UInt32 sum, escFreq; + CPpmd7_Context *mc; + unsigned i, numMasked; + + RC_NORM_REMOTE(p) + + mc = p->MinContext; + numMasked = mc->NumStats; + + do + { + p->OrderFall++; + if (!mc->Suffix) + return; /* EndMarker (symbol = -1) */ + mc = Ppmd7_GetContext(p, mc->Suffix); + i = mc->NumStats; + } + while (i == numMasked); + + p->MinContext = mc; + + // see = Ppmd7_MakeEscFreq(p, numMasked, &escFreq); + { + if (i != 256) + { + unsigned nonMasked = i - numMasked; + see = p->See[(unsigned)p->NS2Indx[(size_t)nonMasked - 1]] + + p->HiBitsFlag + + (nonMasked < (unsigned)SUFFIX(mc)->NumStats - i) + + 2 * (unsigned)(mc->Union2.SummFreq < 11 * i) + + 4 * (unsigned)(numMasked > nonMasked); + { + // if (see->Summ) field is larger than 16-bit, we need only low 16 bits of Summ + unsigned summ = (UInt16)see->Summ; // & 0xFFFF + unsigned r = (summ >> see->Shift); + see->Summ = (UInt16)(summ - r); + escFreq = r + (r == 0); + } + } + else + { + see = &p->DummySee; + escFreq = 1; + } + } + + s = Ppmd7_GetStats(p, mc); + sum = 0; + // i = mc->NumStats; + + do + { + unsigned cur = s->Symbol; + if ((int)cur == symbol) + { + UInt32 low = sum; + UInt32 freq = s->Freq; + unsigned num2; + + Ppmd_See_Update(see); + p->FoundState = s; + sum += escFreq; + + num2 = i / 2; + i &= 1; + sum += freq & (0 - (UInt32)i); + if (num2 != 0) + { + s += i; + for (;;) + { + unsigned sym0 = s[0].Symbol; + unsigned sym1 = s[1].Symbol; + s += 2; + sum += (s[-2].Freq & (unsigned)(MASK(sym0))); + sum += (s[-1].Freq & (unsigned)(MASK(sym1))); + if (--num2 == 0) + break; + } + } + + + R->Range /= sum; + RC_EncodeFinal(low, freq); + Ppmd7_Update2(p); + return; + } + sum += (s->Freq & (unsigned)(MASK(cur))); + s++; + } + while (--i); + + { + UInt32 total = sum + escFreq; + see->Summ = (UInt16)(see->Summ + total); + + R->Range /= total; + RC_Encode(sum, escFreq); + } + + { + CPpmd_State *s2 = Ppmd7_GetStats(p, p->MinContext); + s--; + MASK(s->Symbol) = 0; + do + { + unsigned sym0 = s2[0].Symbol; + unsigned sym1 = s2[1].Symbol; + s2 += 2; + MASK(sym0) = 0; + MASK(sym1) = 0; + } + while (s2 < s); + } + } +} + + +void Ppmd7z_EncodeSymbols(CPpmd7 *p, const Byte *buf, const Byte *lim) +{ + for (; buf < lim; buf++) + { + Ppmd7z_EncodeSymbol(p, *buf); + } +} diff --git a/3rdparty/lzma/src/Sha256.c b/3rdparty/lzma/src/Sha256.c new file mode 100644 index 0000000000..8b3983ea73 --- /dev/null +++ b/3rdparty/lzma/src/Sha256.c @@ -0,0 +1,486 @@ +/* Sha256.c -- SHA-256 Hash +2021-04-01 : Igor Pavlov : Public domain +This code is based on public domain code from Wei Dai's Crypto++ library. */ + +#include "Precomp.h" + +#include + +#include "CpuArch.h" +#include "RotateDefs.h" +#include "Sha256.h" + +#if defined(_MSC_VER) && (_MSC_VER < 1900) +// #define USE_MY_MM +#endif + +#ifdef MY_CPU_X86_OR_AMD64 + #ifdef _MSC_VER + #if _MSC_VER >= 1200 + #define _SHA_SUPPORTED + #endif + #elif defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define _SHA_SUPPORTED + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 8) // fix that check + #define _SHA_SUPPORTED + #endif + #elif defined(__INTEL_COMPILER) + #if (__INTEL_COMPILER >= 1800) // fix that check + #define _SHA_SUPPORTED + #endif + #endif +#elif defined(MY_CPU_ARM_OR_ARM64) + #ifdef _MSC_VER + #if _MSC_VER >= 1910 + #define _SHA_SUPPORTED + #endif + #elif defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define _SHA_SUPPORTED + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 6) // fix that check + #define _SHA_SUPPORTED + #endif + #endif +#endif + +void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks); + +#ifdef _SHA_SUPPORTED + void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); + + static SHA256_FUNC_UPDATE_BLOCKS g_FUNC_UPDATE_BLOCKS = Sha256_UpdateBlocks; + static SHA256_FUNC_UPDATE_BLOCKS g_FUNC_UPDATE_BLOCKS_HW; + + #define UPDATE_BLOCKS(p) p->func_UpdateBlocks +#else + #define UPDATE_BLOCKS(p) Sha256_UpdateBlocks +#endif + + +BoolInt Sha256_SetFunction(CSha256 *p, unsigned algo) +{ + SHA256_FUNC_UPDATE_BLOCKS func = Sha256_UpdateBlocks; + + #ifdef _SHA_SUPPORTED + if (algo != SHA256_ALGO_SW) + { + if (algo == SHA256_ALGO_DEFAULT) + func = g_FUNC_UPDATE_BLOCKS; + else + { + if (algo != SHA256_ALGO_HW) + return False; + func = g_FUNC_UPDATE_BLOCKS_HW; + if (!func) + return False; + } + } + #else + if (algo > 1) + return False; + #endif + + p->func_UpdateBlocks = func; + return True; +} + + +/* define it for speed optimization */ + +#ifdef _SFX + #define STEP_PRE 1 + #define STEP_MAIN 1 +#else + #define STEP_PRE 2 + #define STEP_MAIN 4 + // #define _SHA256_UNROLL +#endif + +#if STEP_MAIN != 16 + #define _SHA256_BIG_W +#endif + + + + +void Sha256_InitState(CSha256 *p) +{ + p->count = 0; + p->state[0] = 0x6a09e667; + p->state[1] = 0xbb67ae85; + p->state[2] = 0x3c6ef372; + p->state[3] = 0xa54ff53a; + p->state[4] = 0x510e527f; + p->state[5] = 0x9b05688c; + p->state[6] = 0x1f83d9ab; + p->state[7] = 0x5be0cd19; +} + +void Sha256_Init(CSha256 *p) +{ + p->func_UpdateBlocks = + #ifdef _SHA_SUPPORTED + g_FUNC_UPDATE_BLOCKS; + #else + NULL; + #endif + Sha256_InitState(p); +} + +#define S0(x) (rotrFixed(x, 2) ^ rotrFixed(x,13) ^ rotrFixed(x, 22)) +#define S1(x) (rotrFixed(x, 6) ^ rotrFixed(x,11) ^ rotrFixed(x, 25)) +#define s0(x) (rotrFixed(x, 7) ^ rotrFixed(x,18) ^ (x >> 3)) +#define s1(x) (rotrFixed(x,17) ^ rotrFixed(x,19) ^ (x >> 10)) + +#define Ch(x,y,z) (z^(x&(y^z))) +#define Maj(x,y,z) ((x&y)|(z&(x|y))) + + +#define W_PRE(i) (W[(i) + (size_t)(j)] = GetBe32(data + ((size_t)(j) + i) * 4)) + +#define blk2_main(j, i) s1(w(j, (i)-2)) + w(j, (i)-7) + s0(w(j, (i)-15)) + +#ifdef _SHA256_BIG_W + // we use +i instead of +(i) to change the order to solve CLANG compiler warning for signed/unsigned. + #define w(j, i) W[(size_t)(j) + i] + #define blk2(j, i) (w(j, i) = w(j, (i)-16) + blk2_main(j, i)) +#else + #if STEP_MAIN == 16 + #define w(j, i) W[(i) & 15] + #else + #define w(j, i) W[((size_t)(j) + (i)) & 15] + #endif + #define blk2(j, i) (w(j, i) += blk2_main(j, i)) +#endif + +#define W_MAIN(i) blk2(j, i) + + +#define T1(wx, i) \ + tmp = h + S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + wx(i); \ + h = g; \ + g = f; \ + f = e; \ + e = d + tmp; \ + tmp += S0(a) + Maj(a, b, c); \ + d = c; \ + c = b; \ + b = a; \ + a = tmp; \ + +#define R1_PRE(i) T1( W_PRE, i) +#define R1_MAIN(i) T1( W_MAIN, i) + +#if (!defined(_SHA256_UNROLL) || STEP_MAIN < 8) && (STEP_MAIN >= 4) +#define R2_MAIN(i) \ + R1_MAIN(i) \ + R1_MAIN(i + 1) \ + +#endif + + + +#if defined(_SHA256_UNROLL) && STEP_MAIN >= 8 + +#define T4( a,b,c,d,e,f,g,h, wx, i) \ + h += S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + wx(i); \ + tmp = h; \ + h += d; \ + d = tmp + S0(a) + Maj(a, b, c); \ + +#define R4( wx, i) \ + T4 ( a,b,c,d,e,f,g,h, wx, (i )); \ + T4 ( d,a,b,c,h,e,f,g, wx, (i+1)); \ + T4 ( c,d,a,b,g,h,e,f, wx, (i+2)); \ + T4 ( b,c,d,a,f,g,h,e, wx, (i+3)); \ + +#define R4_PRE(i) R4( W_PRE, i) +#define R4_MAIN(i) R4( W_MAIN, i) + + +#define T8( a,b,c,d,e,f,g,h, wx, i) \ + h += S1(e) + Ch(e,f,g) + K[(i)+(size_t)(j)] + wx(i); \ + d += h; \ + h += S0(a) + Maj(a, b, c); \ + +#define R8( wx, i) \ + T8 ( a,b,c,d,e,f,g,h, wx, i ); \ + T8 ( h,a,b,c,d,e,f,g, wx, i+1); \ + T8 ( g,h,a,b,c,d,e,f, wx, i+2); \ + T8 ( f,g,h,a,b,c,d,e, wx, i+3); \ + T8 ( e,f,g,h,a,b,c,d, wx, i+4); \ + T8 ( d,e,f,g,h,a,b,c, wx, i+5); \ + T8 ( c,d,e,f,g,h,a,b, wx, i+6); \ + T8 ( b,c,d,e,f,g,h,a, wx, i+7); \ + +#define R8_PRE(i) R8( W_PRE, i) +#define R8_MAIN(i) R8( W_MAIN, i) + +#endif + +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); + +// static +extern MY_ALIGN(64) +const UInt32 SHA256_K_ARRAY[64]; + +MY_ALIGN(64) +const UInt32 SHA256_K_ARRAY[64] = { + 0x428a2f98, 0x71374491, 0xb5c0fbcf, 0xe9b5dba5, + 0x3956c25b, 0x59f111f1, 0x923f82a4, 0xab1c5ed5, + 0xd807aa98, 0x12835b01, 0x243185be, 0x550c7dc3, + 0x72be5d74, 0x80deb1fe, 0x9bdc06a7, 0xc19bf174, + 0xe49b69c1, 0xefbe4786, 0x0fc19dc6, 0x240ca1cc, + 0x2de92c6f, 0x4a7484aa, 0x5cb0a9dc, 0x76f988da, + 0x983e5152, 0xa831c66d, 0xb00327c8, 0xbf597fc7, + 0xc6e00bf3, 0xd5a79147, 0x06ca6351, 0x14292967, + 0x27b70a85, 0x2e1b2138, 0x4d2c6dfc, 0x53380d13, + 0x650a7354, 0x766a0abb, 0x81c2c92e, 0x92722c85, + 0xa2bfe8a1, 0xa81a664b, 0xc24b8b70, 0xc76c51a3, + 0xd192e819, 0xd6990624, 0xf40e3585, 0x106aa070, + 0x19a4c116, 0x1e376c08, 0x2748774c, 0x34b0bcb5, + 0x391c0cb3, 0x4ed8aa4a, 0x5b9cca4f, 0x682e6ff3, + 0x748f82ee, 0x78a5636f, 0x84c87814, 0x8cc70208, + 0x90befffa, 0xa4506ceb, 0xbef9a3f7, 0xc67178f2 +}; + +#define K SHA256_K_ARRAY + + +MY_NO_INLINE +void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks) +{ + UInt32 W + #ifdef _SHA256_BIG_W + [64]; + #else + [16]; + #endif + + unsigned j; + + UInt32 a,b,c,d,e,f,g,h; + + #if !defined(_SHA256_UNROLL) || (STEP_MAIN <= 4) || (STEP_PRE <= 4) + UInt32 tmp; + #endif + + a = state[0]; + b = state[1]; + c = state[2]; + d = state[3]; + e = state[4]; + f = state[5]; + g = state[6]; + h = state[7]; + + while (numBlocks) + { + + for (j = 0; j < 16; j += STEP_PRE) + { + #if STEP_PRE > 4 + + #if STEP_PRE < 8 + R4_PRE(0); + #else + R8_PRE(0); + #if STEP_PRE == 16 + R8_PRE(8); + #endif + #endif + + #else + + R1_PRE(0); + #if STEP_PRE >= 2 + R1_PRE(1); + #if STEP_PRE >= 4 + R1_PRE(2); + R1_PRE(3); + #endif + #endif + + #endif + } + + for (j = 16; j < 64; j += STEP_MAIN) + { + #if defined(_SHA256_UNROLL) && STEP_MAIN >= 8 + + #if STEP_MAIN < 8 + R4_MAIN(0); + #else + R8_MAIN(0); + #if STEP_MAIN == 16 + R8_MAIN(8); + #endif + #endif + + #else + + R1_MAIN(0); + #if STEP_MAIN >= 2 + R1_MAIN(1); + #if STEP_MAIN >= 4 + R2_MAIN(2); + #if STEP_MAIN >= 8 + R2_MAIN(4); + R2_MAIN(6); + #if STEP_MAIN >= 16 + R2_MAIN(8); + R2_MAIN(10); + R2_MAIN(12); + R2_MAIN(14); + #endif + #endif + #endif + #endif + #endif + } + + a += state[0]; state[0] = a; + b += state[1]; state[1] = b; + c += state[2]; state[2] = c; + d += state[3]; state[3] = d; + e += state[4]; state[4] = e; + f += state[5]; state[5] = f; + g += state[6]; state[6] = g; + h += state[7]; state[7] = h; + + data += 64; + numBlocks--; + } + + /* Wipe variables */ + /* memset(W, 0, sizeof(W)); */ +} + +#undef S0 +#undef S1 +#undef s0 +#undef s1 +#undef K + +#define Sha256_UpdateBlock(p) UPDATE_BLOCKS(p)(p->state, p->buffer, 1) + +void Sha256_Update(CSha256 *p, const Byte *data, size_t size) +{ + if (size == 0) + return; + + { + unsigned pos = (unsigned)p->count & 0x3F; + unsigned num; + + p->count += size; + + num = 64 - pos; + if (num > size) + { + memcpy(p->buffer + pos, data, size); + return; + } + + if (pos != 0) + { + size -= num; + memcpy(p->buffer + pos, data, num); + data += num; + Sha256_UpdateBlock(p); + } + } + { + size_t numBlocks = size >> 6; + UPDATE_BLOCKS(p)(p->state, data, numBlocks); + size &= 0x3F; + if (size == 0) + return; + data += (numBlocks << 6); + memcpy(p->buffer, data, size); + } +} + + +void Sha256_Final(CSha256 *p, Byte *digest) +{ + unsigned pos = (unsigned)p->count & 0x3F; + unsigned i; + + p->buffer[pos++] = 0x80; + + if (pos > (64 - 8)) + { + while (pos != 64) { p->buffer[pos++] = 0; } + // memset(&p->buf.buffer[pos], 0, 64 - pos); + Sha256_UpdateBlock(p); + pos = 0; + } + + /* + if (pos & 3) + { + p->buffer[pos] = 0; + p->buffer[pos + 1] = 0; + p->buffer[pos + 2] = 0; + pos += 3; + pos &= ~3; + } + { + for (; pos < 64 - 8; pos += 4) + *(UInt32 *)(&p->buffer[pos]) = 0; + } + */ + + memset(&p->buffer[pos], 0, (64 - 8) - pos); + + { + UInt64 numBits = (p->count << 3); + SetBe32(p->buffer + 64 - 8, (UInt32)(numBits >> 32)); + SetBe32(p->buffer + 64 - 4, (UInt32)(numBits)); + } + + Sha256_UpdateBlock(p); + + for (i = 0; i < 8; i += 2) + { + UInt32 v0 = p->state[i]; + UInt32 v1 = p->state[(size_t)i + 1]; + SetBe32(digest , v0); + SetBe32(digest + 4, v1); + digest += 8; + } + + Sha256_InitState(p); +} + + +void Sha256Prepare() +{ + #ifdef _SHA_SUPPORTED + SHA256_FUNC_UPDATE_BLOCKS f, f_hw; + f = Sha256_UpdateBlocks; + f_hw = NULL; + #ifdef MY_CPU_X86_OR_AMD64 + #ifndef USE_MY_MM + if (CPU_IsSupported_SHA() + && CPU_IsSupported_SSSE3() + // && CPU_IsSupported_SSE41() + ) + #endif + #else + if (CPU_IsSupported_SHA2()) + #endif + { + // printf("\n========== HW SHA256 ======== \n"); + f = f_hw = Sha256_UpdateBlocks_HW; + } + g_FUNC_UPDATE_BLOCKS = f; + g_FUNC_UPDATE_BLOCKS_HW = f_hw; + #endif +} diff --git a/3rdparty/lzma/src/Sha256Opt.c b/3rdparty/lzma/src/Sha256Opt.c new file mode 100644 index 0000000000..decc1382cd --- /dev/null +++ b/3rdparty/lzma/src/Sha256Opt.c @@ -0,0 +1,373 @@ +/* Sha256Opt.c -- SHA-256 optimized code for SHA-256 hardware instructions +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#if defined(_MSC_VER) +#if (_MSC_VER < 1900) && (_MSC_VER >= 1200) +// #define USE_MY_MM +#endif +#endif + +#include "CpuArch.h" + +#ifdef MY_CPU_X86_OR_AMD64 + #if defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define USE_HW_SHA + #ifndef __SHA__ + #define ATTRIB_SHA __attribute__((__target__("sha,ssse3"))) + #if defined(_MSC_VER) + // SSSE3: for clang-cl: + #include + #define __SHA__ + #endif + #endif + + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 8) // fix that check + #define USE_HW_SHA + #ifndef __SHA__ + #define ATTRIB_SHA __attribute__((__target__("sha,ssse3"))) + // #pragma GCC target("sha,ssse3") + #endif + #endif + #elif defined(__INTEL_COMPILER) + #if (__INTEL_COMPILER >= 1800) // fix that check + #define USE_HW_SHA + #endif + #elif defined(_MSC_VER) + #ifdef USE_MY_MM + #define USE_VER_MIN 1300 + #else + #define USE_VER_MIN 1910 + #endif + #if _MSC_VER >= USE_VER_MIN + #define USE_HW_SHA + #endif + #endif +// #endif // MY_CPU_X86_OR_AMD64 + +#ifdef USE_HW_SHA + +// #pragma message("Sha256 HW") +// #include + +#if !defined(_MSC_VER) || (_MSC_VER >= 1900) +#include +#else +#include + +#if defined(_MSC_VER) && (_MSC_VER >= 1600) +// #include +#endif + +#ifdef USE_MY_MM +#include "My_mm.h" +#endif + +#endif + +/* +SHA256 uses: +SSE2: + _mm_loadu_si128 + _mm_storeu_si128 + _mm_set_epi32 + _mm_add_epi32 + _mm_shuffle_epi32 / pshufd + + + +SSSE3: + _mm_shuffle_epi8 / pshufb + _mm_alignr_epi8 +SHA: + _mm_sha256* +*/ + +// K array must be aligned for 16-bytes at least. +// The compiler can look align attribute and selects +// movdqu - for code without align attribute +// movdqa - for code with align attribute +extern +MY_ALIGN(64) +const UInt32 SHA256_K_ARRAY[64]; + +#define K SHA256_K_ARRAY + + +#define ADD_EPI32(dest, src) dest = _mm_add_epi32(dest, src); +#define SHA256_MSG1(dest, src) dest = _mm_sha256msg1_epu32(dest, src); +#define SHA25G_MSG2(dest, src) dest = _mm_sha256msg2_epu32(dest, src); + + +#define LOAD_SHUFFLE(m, k) \ + m = _mm_loadu_si128((const __m128i *)(const void *)(data + (k) * 16)); \ + m = _mm_shuffle_epi8(m, mask); \ + +#define SM1(g0, g1, g2, g3) \ + SHA256_MSG1(g3, g0); \ + +#define SM2(g0, g1, g2, g3) \ + tmp = _mm_alignr_epi8(g1, g0, 4); \ + ADD_EPI32(g2, tmp); \ + SHA25G_MSG2(g2, g1); \ + +// #define LS0(k, g0, g1, g2, g3) LOAD_SHUFFLE(g0, k) +// #define LS1(k, g0, g1, g2, g3) LOAD_SHUFFLE(g1, k+1) + + +#define NNN(g0, g1, g2, g3) + + +#define RND2(t0, t1) \ + t0 = _mm_sha256rnds2_epu32(t0, t1, msg); + +#define RND2_0(m, k) \ + msg = _mm_add_epi32(m, *(const __m128i *) (const void *) &K[(k) * 4]); \ + RND2(state0, state1); \ + msg = _mm_shuffle_epi32(msg, 0x0E); \ + + +#define RND2_1 \ + RND2(state1, state0); \ + + +// We use scheme with 3 rounds ahead for SHA256_MSG1 / 2 rounds ahead for SHA256_MSG2 + +#define R4(k, g0, g1, g2, g3, OP0, OP1) \ + RND2_0(g0, k); \ + OP0(g0, g1, g2, g3); \ + RND2_1; \ + OP1(g0, g1, g2, g3); \ + +#define R16(k, OP0, OP1, OP2, OP3, OP4, OP5, OP6, OP7) \ + R4 ( (k)*4+0, m0, m1, m2, m3, OP0, OP1 ) \ + R4 ( (k)*4+1, m1, m2, m3, m0, OP2, OP3 ) \ + R4 ( (k)*4+2, m2, m3, m0, m1, OP4, OP5 ) \ + R4 ( (k)*4+3, m3, m0, m1, m2, OP6, OP7 ) \ + +#define PREPARE_STATE \ + tmp = _mm_shuffle_epi32(state0, 0x1B); /* abcd */ \ + state0 = _mm_shuffle_epi32(state1, 0x1B); /* efgh */ \ + state1 = state0; \ + state0 = _mm_unpacklo_epi64(state0, tmp); /* cdgh */ \ + state1 = _mm_unpackhi_epi64(state1, tmp); /* abef */ \ + + +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); +#ifdef ATTRIB_SHA +ATTRIB_SHA +#endif +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks) +{ + const __m128i mask = _mm_set_epi32(0x0c0d0e0f, 0x08090a0b, 0x04050607, 0x00010203); + __m128i tmp; + __m128i state0, state1; + + if (numBlocks == 0) + return; + + state0 = _mm_loadu_si128((const __m128i *) (const void *) &state[0]); + state1 = _mm_loadu_si128((const __m128i *) (const void *) &state[4]); + + PREPARE_STATE + + do + { + __m128i state0_save, state1_save; + __m128i m0, m1, m2, m3; + __m128i msg; + // #define msg tmp + + state0_save = state0; + state1_save = state1; + + LOAD_SHUFFLE (m0, 0) + LOAD_SHUFFLE (m1, 1) + LOAD_SHUFFLE (m2, 2) + LOAD_SHUFFLE (m3, 3) + + + + R16 ( 0, NNN, NNN, SM1, NNN, SM1, SM2, SM1, SM2 ); + R16 ( 1, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); + R16 ( 2, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); + R16 ( 3, SM1, SM2, NNN, SM2, NNN, NNN, NNN, NNN ); + + ADD_EPI32(state0, state0_save); + ADD_EPI32(state1, state1_save); + + data += 64; + } + while (--numBlocks); + + PREPARE_STATE + + _mm_storeu_si128((__m128i *) (void *) &state[0], state0); + _mm_storeu_si128((__m128i *) (void *) &state[4], state1); +} + +#endif // USE_HW_SHA + +#elif defined(MY_CPU_ARM_OR_ARM64) + + #if defined(__clang__) + #if (__clang_major__ >= 8) // fix that check + #define USE_HW_SHA + #endif + #elif defined(__GNUC__) + #if (__GNUC__ >= 6) // fix that check + #define USE_HW_SHA + #endif + #elif defined(_MSC_VER) + #if _MSC_VER >= 1910 + #define USE_HW_SHA + #endif + #endif + +#ifdef USE_HW_SHA + +// #pragma message("=== Sha256 HW === ") + +#if defined(__clang__) || defined(__GNUC__) + #ifdef MY_CPU_ARM64 + #define ATTRIB_SHA __attribute__((__target__("+crypto"))) + #else + #define ATTRIB_SHA __attribute__((__target__("fpu=crypto-neon-fp-armv8"))) + #endif +#else + // _MSC_VER + // for arm32 + #define _ARM_USE_NEW_NEON_INTRINSICS +#endif + +#if defined(_MSC_VER) && defined(MY_CPU_ARM64) +#include +#else +#include +#endif + +typedef uint32x4_t v128; +// typedef __n128 v128; // MSVC + +#ifdef MY_CPU_BE + #define MY_rev32_for_LE(x) +#else + #define MY_rev32_for_LE(x) x = vreinterpretq_u32_u8(vrev32q_u8(vreinterpretq_u8_u32(x))) +#endif + +#define LOAD_128(_p) (*(const v128 *)(const void *)(_p)) +#define STORE_128(_p, _v) *(v128 *)(void *)(_p) = (_v) + +#define LOAD_SHUFFLE(m, k) \ + m = LOAD_128((data + (k) * 16)); \ + MY_rev32_for_LE(m); \ + +// K array must be aligned for 16-bytes at least. +extern +MY_ALIGN(64) +const UInt32 SHA256_K_ARRAY[64]; + +#define K SHA256_K_ARRAY + + +#define SHA256_SU0(dest, src) dest = vsha256su0q_u32(dest, src); +#define SHA25G_SU1(dest, src2, src3) dest = vsha256su1q_u32(dest, src2, src3); + +#define SM1(g0, g1, g2, g3) SHA256_SU0(g3, g0) +#define SM2(g0, g1, g2, g3) SHA25G_SU1(g2, g0, g1) +#define NNN(g0, g1, g2, g3) + + +#define R4(k, g0, g1, g2, g3, OP0, OP1) \ + msg = vaddq_u32(g0, *(const v128 *) (const void *) &K[(k) * 4]); \ + tmp = state0; \ + state0 = vsha256hq_u32( state0, state1, msg ); \ + state1 = vsha256h2q_u32( state1, tmp, msg ); \ + OP0(g0, g1, g2, g3); \ + OP1(g0, g1, g2, g3); \ + + +#define R16(k, OP0, OP1, OP2, OP3, OP4, OP5, OP6, OP7) \ + R4 ( (k)*4+0, m0, m1, m2, m3, OP0, OP1 ) \ + R4 ( (k)*4+1, m1, m2, m3, m0, OP2, OP3 ) \ + R4 ( (k)*4+2, m2, m3, m0, m1, OP4, OP5 ) \ + R4 ( (k)*4+3, m3, m0, m1, m2, OP6, OP7 ) \ + + +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); +#ifdef ATTRIB_SHA +ATTRIB_SHA +#endif +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks) +{ + v128 state0, state1; + + if (numBlocks == 0) + return; + + state0 = LOAD_128(&state[0]); + state1 = LOAD_128(&state[4]); + + do + { + v128 state0_save, state1_save; + v128 m0, m1, m2, m3; + v128 msg, tmp; + + state0_save = state0; + state1_save = state1; + + LOAD_SHUFFLE (m0, 0) + LOAD_SHUFFLE (m1, 1) + LOAD_SHUFFLE (m2, 2) + LOAD_SHUFFLE (m3, 3) + + R16 ( 0, NNN, NNN, SM1, NNN, SM1, SM2, SM1, SM2 ); + R16 ( 1, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); + R16 ( 2, SM1, SM2, SM1, SM2, SM1, SM2, SM1, SM2 ); + R16 ( 3, SM1, SM2, NNN, SM2, NNN, NNN, NNN, NNN ); + + state0 = vaddq_u32(state0, state0_save); + state1 = vaddq_u32(state1, state1_save); + + data += 64; + } + while (--numBlocks); + + STORE_128(&state[0], state0); + STORE_128(&state[4], state1); +} + +#endif // USE_HW_SHA + +#endif // MY_CPU_ARM_OR_ARM64 + + +#ifndef USE_HW_SHA + +// #error Stop_Compiling_UNSUPPORTED_SHA +// #include + +// #include "Sha256.h" +void MY_FAST_CALL Sha256_UpdateBlocks(UInt32 state[8], const Byte *data, size_t numBlocks); + +#pragma message("Sha256 HW-SW stub was used") + +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks); +void MY_FAST_CALL Sha256_UpdateBlocks_HW(UInt32 state[8], const Byte *data, size_t numBlocks) +{ + Sha256_UpdateBlocks(state, data, numBlocks); + /* + UNUSED_VAR(state); + UNUSED_VAR(data); + UNUSED_VAR(numBlocks); + exit(1); + return; + */ +} + +#endif diff --git a/3rdparty/lzma/src/Xz.c b/3rdparty/lzma/src/Xz.c new file mode 100644 index 0000000000..7c53b60073 --- /dev/null +++ b/3rdparty/lzma/src/Xz.c @@ -0,0 +1,90 @@ +/* Xz.c - Xz +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "7zCrc.h" +#include "CpuArch.h" +#include "Xz.h" +#include "XzCrc64.h" + +const Byte XZ_SIG[XZ_SIG_SIZE] = { 0xFD, '7', 'z', 'X', 'Z', 0 }; +/* const Byte XZ_FOOTER_SIG[XZ_FOOTER_SIG_SIZE] = { 'Y', 'Z' }; */ + +unsigned Xz_WriteVarInt(Byte *buf, UInt64 v) +{ + unsigned i = 0; + do + { + buf[i++] = (Byte)((v & 0x7F) | 0x80); + v >>= 7; + } + while (v != 0); + buf[(size_t)i - 1] &= 0x7F; + return i; +} + +void Xz_Construct(CXzStream *p) +{ + p->numBlocks = 0; + p->blocks = NULL; + p->flags = 0; +} + +void Xz_Free(CXzStream *p, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, p->blocks); + p->numBlocks = 0; + p->blocks = NULL; +} + +unsigned XzFlags_GetCheckSize(CXzStreamFlags f) +{ + unsigned t = XzFlags_GetCheckType(f); + return (t == 0) ? 0 : ((unsigned)4 << ((t - 1) / 3)); +} + +void XzCheck_Init(CXzCheck *p, unsigned mode) +{ + p->mode = mode; + switch (mode) + { + case XZ_CHECK_CRC32: p->crc = CRC_INIT_VAL; break; + case XZ_CHECK_CRC64: p->crc64 = CRC64_INIT_VAL; break; + case XZ_CHECK_SHA256: Sha256_Init(&p->sha); break; + } +} + +void XzCheck_Update(CXzCheck *p, const void *data, size_t size) +{ + switch (p->mode) + { + case XZ_CHECK_CRC32: p->crc = CrcUpdate(p->crc, data, size); break; + case XZ_CHECK_CRC64: p->crc64 = Crc64Update(p->crc64, data, size); break; + case XZ_CHECK_SHA256: Sha256_Update(&p->sha, (const Byte *)data, size); break; + } +} + +int XzCheck_Final(CXzCheck *p, Byte *digest) +{ + switch (p->mode) + { + case XZ_CHECK_CRC32: + SetUi32(digest, CRC_GET_DIGEST(p->crc)); + break; + case XZ_CHECK_CRC64: + { + int i; + UInt64 v = CRC64_GET_DIGEST(p->crc64); + for (i = 0; i < 8; i++, v >>= 8) + digest[i] = (Byte)(v & 0xFF); + break; + } + case XZ_CHECK_SHA256: + Sha256_Final(&p->sha, digest); + break; + default: + return 0; + } + return 1; +} diff --git a/3rdparty/lzma/src/XzCrc64.c b/3rdparty/lzma/src/XzCrc64.c new file mode 100644 index 0000000000..b6d02cbebd --- /dev/null +++ b/3rdparty/lzma/src/XzCrc64.c @@ -0,0 +1,86 @@ +/* XzCrc64.c -- CRC64 calculation +2017-05-23 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "XzCrc64.h" +#include "CpuArch.h" + +#define kCrc64Poly UINT64_CONST(0xC96C5795D7870F42) + +#ifdef MY_CPU_LE + #define CRC64_NUM_TABLES 4 +#else + #define CRC64_NUM_TABLES 5 + #define CRC_UINT64_SWAP(v) \ + ((v >> 56) \ + | ((v >> 40) & ((UInt64)0xFF << 8)) \ + | ((v >> 24) & ((UInt64)0xFF << 16)) \ + | ((v >> 8) & ((UInt64)0xFF << 24)) \ + | ((v << 8) & ((UInt64)0xFF << 32)) \ + | ((v << 24) & ((UInt64)0xFF << 40)) \ + | ((v << 40) & ((UInt64)0xFF << 48)) \ + | ((v << 56))) + + UInt64 MY_FAST_CALL XzCrc64UpdateT1_BeT4(UInt64 v, const void *data, size_t size, const UInt64 *table); +#endif + +#ifndef MY_CPU_BE + UInt64 MY_FAST_CALL XzCrc64UpdateT4(UInt64 v, const void *data, size_t size, const UInt64 *table); +#endif + +typedef UInt64 (MY_FAST_CALL *CRC64_FUNC)(UInt64 v, const void *data, size_t size, const UInt64 *table); + +static CRC64_FUNC g_Crc64Update; +UInt64 g_Crc64Table[256 * CRC64_NUM_TABLES]; + +UInt64 MY_FAST_CALL Crc64Update(UInt64 v, const void *data, size_t size) +{ + return g_Crc64Update(v, data, size, g_Crc64Table); +} + +UInt64 MY_FAST_CALL Crc64Calc(const void *data, size_t size) +{ + return g_Crc64Update(CRC64_INIT_VAL, data, size, g_Crc64Table) ^ CRC64_INIT_VAL; +} + +void MY_FAST_CALL Crc64GenerateTable() +{ + UInt32 i; + for (i = 0; i < 256; i++) + { + UInt64 r = i; + unsigned j; + for (j = 0; j < 8; j++) + r = (r >> 1) ^ (kCrc64Poly & ((UInt64)0 - (r & 1))); + g_Crc64Table[i] = r; + } + for (i = 256; i < 256 * CRC64_NUM_TABLES; i++) + { + UInt64 r = g_Crc64Table[(size_t)i - 256]; + g_Crc64Table[i] = g_Crc64Table[r & 0xFF] ^ (r >> 8); + } + + #ifdef MY_CPU_LE + + g_Crc64Update = XzCrc64UpdateT4; + + #else + { + #ifndef MY_CPU_BE + UInt32 k = 1; + if (*(const Byte *)&k == 1) + g_Crc64Update = XzCrc64UpdateT4; + else + #endif + { + for (i = 256 * CRC64_NUM_TABLES - 1; i >= 256; i--) + { + UInt64 x = g_Crc64Table[(size_t)i - 256]; + g_Crc64Table[i] = CRC_UINT64_SWAP(x); + } + g_Crc64Update = XzCrc64UpdateT1_BeT4; + } + } + #endif +} diff --git a/3rdparty/lzma/src/XzCrc64Opt.c b/3rdparty/lzma/src/XzCrc64Opt.c new file mode 100644 index 0000000000..93a9ffff5f --- /dev/null +++ b/3rdparty/lzma/src/XzCrc64Opt.c @@ -0,0 +1,71 @@ +/* XzCrc64Opt.c -- CRC64 calculation +2021-02-09 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include "CpuArch.h" + +#ifndef MY_CPU_BE + +#define CRC64_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +UInt64 MY_FAST_CALL XzCrc64UpdateT4(UInt64 v, const void *data, size_t size, const UInt64 *table); +UInt64 MY_FAST_CALL XzCrc64UpdateT4(UInt64 v, const void *data, size_t size, const UInt64 *table) +{ + const Byte *p = (const Byte *)data; + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) + v = CRC64_UPDATE_BYTE_2(v, *p); + for (; size >= 4; size -= 4, p += 4) + { + UInt32 d = (UInt32)v ^ *(const UInt32 *)(const void *)p; + v = (v >> 32) + ^ (table + 0x300)[((d ) & 0xFF)] + ^ (table + 0x200)[((d >> 8) & 0xFF)] + ^ (table + 0x100)[((d >> 16) & 0xFF)] + ^ (table + 0x000)[((d >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC64_UPDATE_BYTE_2(v, *p); + return v; +} + +#endif + + +#ifndef MY_CPU_LE + +#define CRC_UINT64_SWAP(v) \ + ((v >> 56) \ + | ((v >> 40) & ((UInt64)0xFF << 8)) \ + | ((v >> 24) & ((UInt64)0xFF << 16)) \ + | ((v >> 8) & ((UInt64)0xFF << 24)) \ + | ((v << 8) & ((UInt64)0xFF << 32)) \ + | ((v << 24) & ((UInt64)0xFF << 40)) \ + | ((v << 40) & ((UInt64)0xFF << 48)) \ + | ((v << 56))) + +#define CRC64_UPDATE_BYTE_2_BE(crc, b) (table[(Byte)((crc) >> 56) ^ (b)] ^ ((crc) << 8)) + +UInt64 MY_FAST_CALL XzCrc64UpdateT1_BeT4(UInt64 v, const void *data, size_t size, const UInt64 *table); +UInt64 MY_FAST_CALL XzCrc64UpdateT1_BeT4(UInt64 v, const void *data, size_t size, const UInt64 *table) +{ + const Byte *p = (const Byte *)data; + table += 0x100; + v = CRC_UINT64_SWAP(v); + for (; size > 0 && ((unsigned)(ptrdiff_t)p & 3) != 0; size--, p++) + v = CRC64_UPDATE_BYTE_2_BE(v, *p); + for (; size >= 4; size -= 4, p += 4) + { + UInt32 d = (UInt32)(v >> 32) ^ *(const UInt32 *)(const void *)p; + v = (v << 32) + ^ (table + 0x000)[((d ) & 0xFF)] + ^ (table + 0x100)[((d >> 8) & 0xFF)] + ^ (table + 0x200)[((d >> 16) & 0xFF)] + ^ (table + 0x300)[((d >> 24))]; + } + for (; size > 0; size--, p++) + v = CRC64_UPDATE_BYTE_2_BE(v, *p); + return CRC_UINT64_SWAP(v); +} + +#endif diff --git a/3rdparty/lzma/src/XzDec.c b/3rdparty/lzma/src/XzDec.c new file mode 100644 index 0000000000..3f96a37f94 --- /dev/null +++ b/3rdparty/lzma/src/XzDec.c @@ -0,0 +1,2837 @@ +/* XzDec.c -- Xz Decode +2021-09-04 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +// #include + +// #define XZ_DUMP + +/* #define XZ_DUMP */ + +#ifdef XZ_DUMP +#include +#endif + +// #define SHOW_DEBUG_INFO + +#ifdef SHOW_DEBUG_INFO +#include +#endif + +#ifdef SHOW_DEBUG_INFO +#define PRF(x) x +#else +#define PRF(x) +#endif + +#define PRF_STR(s) PRF(printf("\n" s "\n")) +#define PRF_STR_INT(s, d) PRF(printf("\n" s " %d\n", (unsigned)d)) + +#include +#include + +#include "7zCrc.h" +#include "Alloc.h" +#include "Bra.h" +#include "CpuArch.h" +#include "Delta.h" +#include "Lzma2Dec.h" + +// #define USE_SUBBLOCK + +#ifdef USE_SUBBLOCK +#include "Bcj3Dec.c" +#include "SbDec.h" +#endif + +#include "Xz.h" + +#define XZ_CHECK_SIZE_MAX 64 + +#define CODER_BUF_SIZE ((size_t)1 << 17) + +unsigned Xz_ReadVarInt(const Byte *p, size_t maxSize, UInt64 *value) +{ + unsigned i, limit; + *value = 0; + limit = (maxSize > 9) ? 9 : (unsigned)maxSize; + + for (i = 0; i < limit;) + { + Byte b = p[i]; + *value |= (UInt64)(b & 0x7F) << (7 * i++); + if ((b & 0x80) == 0) + return (b == 0 && i != 1) ? 0 : i; + } + return 0; +} + +/* ---------- BraState ---------- */ + +#define BRA_BUF_SIZE (1 << 14) + +typedef struct +{ + size_t bufPos; + size_t bufConv; + size_t bufTotal; + + int encodeMode; + + UInt32 methodId; + UInt32 delta; + UInt32 ip; + UInt32 x86State; + Byte deltaState[DELTA_STATE_SIZE]; + + Byte buf[BRA_BUF_SIZE]; +} CBraState; + +static void BraState_Free(void *pp, ISzAllocPtr alloc) +{ + ISzAlloc_Free(alloc, pp); +} + +static SRes BraState_SetProps(void *pp, const Byte *props, size_t propSize, ISzAllocPtr alloc) +{ + CBraState *p = ((CBraState *)pp); + UNUSED_VAR(alloc); + p->ip = 0; + if (p->methodId == XZ_ID_Delta) + { + if (propSize != 1) + return SZ_ERROR_UNSUPPORTED; + p->delta = (unsigned)props[0] + 1; + } + else + { + if (propSize == 4) + { + UInt32 v = GetUi32(props); + switch (p->methodId) + { + case XZ_ID_PPC: + case XZ_ID_ARM: + case XZ_ID_SPARC: + if ((v & 3) != 0) + return SZ_ERROR_UNSUPPORTED; + break; + case XZ_ID_ARMT: + if ((v & 1) != 0) + return SZ_ERROR_UNSUPPORTED; + break; + case XZ_ID_IA64: + if ((v & 0xF) != 0) + return SZ_ERROR_UNSUPPORTED; + break; + } + p->ip = v; + } + else if (propSize != 0) + return SZ_ERROR_UNSUPPORTED; + } + return SZ_OK; +} + +static void BraState_Init(void *pp) +{ + CBraState *p = ((CBraState *)pp); + p->bufPos = p->bufConv = p->bufTotal = 0; + x86_Convert_Init(p->x86State); + if (p->methodId == XZ_ID_Delta) + Delta_Init(p->deltaState); +} + + +#define CASE_BRA_CONV(isa) case XZ_ID_ ## isa: size = isa ## _Convert(data, size, p->ip, p->encodeMode); break; + +static SizeT BraState_Filter(void *pp, Byte *data, SizeT size) +{ + CBraState *p = ((CBraState *)pp); + switch (p->methodId) + { + case XZ_ID_Delta: + if (p->encodeMode) + Delta_Encode(p->deltaState, p->delta, data, size); + else + Delta_Decode(p->deltaState, p->delta, data, size); + break; + case XZ_ID_X86: + size = x86_Convert(data, size, p->ip, &p->x86State, p->encodeMode); + break; + CASE_BRA_CONV(PPC) + CASE_BRA_CONV(IA64) + CASE_BRA_CONV(ARM) + CASE_BRA_CONV(ARMT) + CASE_BRA_CONV(SPARC) + } + p->ip += (UInt32)size; + return size; +} + + +static SRes BraState_Code2(void *pp, + Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, int srcWasFinished, + ECoderFinishMode finishMode, + // int *wasFinished + ECoderStatus *status) +{ + CBraState *p = ((CBraState *)pp); + SizeT destRem = *destLen; + SizeT srcRem = *srcLen; + UNUSED_VAR(finishMode); + + *destLen = 0; + *srcLen = 0; + // *wasFinished = False; + *status = CODER_STATUS_NOT_FINISHED; + + while (destRem > 0) + { + if (p->bufPos != p->bufConv) + { + size_t size = p->bufConv - p->bufPos; + if (size > destRem) + size = destRem; + memcpy(dest, p->buf + p->bufPos, size); + p->bufPos += size; + *destLen += size; + dest += size; + destRem -= size; + continue; + } + + p->bufTotal -= p->bufPos; + memmove(p->buf, p->buf + p->bufPos, p->bufTotal); + p->bufPos = 0; + p->bufConv = 0; + { + size_t size = BRA_BUF_SIZE - p->bufTotal; + if (size > srcRem) + size = srcRem; + memcpy(p->buf + p->bufTotal, src, size); + *srcLen += size; + src += size; + srcRem -= size; + p->bufTotal += size; + } + if (p->bufTotal == 0) + break; + + p->bufConv = BraState_Filter(pp, p->buf, p->bufTotal); + + if (p->bufConv == 0) + { + if (!srcWasFinished) + break; + p->bufConv = p->bufTotal; + } + } + + if (p->bufTotal == p->bufPos && srcRem == 0 && srcWasFinished) + { + *status = CODER_STATUS_FINISHED_WITH_MARK; + // *wasFinished = 1; + } + + return SZ_OK; +} + + +SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAllocPtr alloc); +SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAllocPtr alloc) +{ + CBraState *decoder; + if (id < XZ_ID_Delta || id > XZ_ID_SPARC) + return SZ_ERROR_UNSUPPORTED; + decoder = (CBraState *)p->p; + if (!decoder) + { + decoder = (CBraState *)ISzAlloc_Alloc(alloc, sizeof(CBraState)); + if (!decoder) + return SZ_ERROR_MEM; + p->p = decoder; + p->Free = BraState_Free; + p->SetProps = BraState_SetProps; + p->Init = BraState_Init; + p->Code2 = BraState_Code2; + p->Filter = BraState_Filter; + } + decoder->methodId = (UInt32)id; + decoder->encodeMode = encodeMode; + return SZ_OK; +} + + + +/* ---------- SbState ---------- */ + +#ifdef USE_SUBBLOCK + +static void SbState_Free(void *pp, ISzAllocPtr alloc) +{ + CSbDec *p = (CSbDec *)pp; + SbDec_Free(p); + ISzAlloc_Free(alloc, pp); +} + +static SRes SbState_SetProps(void *pp, const Byte *props, size_t propSize, ISzAllocPtr alloc) +{ + UNUSED_VAR(pp); + UNUSED_VAR(props); + UNUSED_VAR(alloc); + return (propSize == 0) ? SZ_OK : SZ_ERROR_UNSUPPORTED; +} + +static void SbState_Init(void *pp) +{ + SbDec_Init((CSbDec *)pp); +} + +static SRes SbState_Code2(void *pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, + // int *wasFinished + ECoderStatus *status) +{ + CSbDec *p = (CSbDec *)pp; + SRes res; + UNUSED_VAR(srcWasFinished); + p->dest = dest; + p->destLen = *destLen; + p->src = src; + p->srcLen = *srcLen; + p->finish = finishMode; /* change it */ + res = SbDec_Decode((CSbDec *)pp); + *destLen -= p->destLen; + *srcLen -= p->srcLen; + // *wasFinished = (*destLen == 0 && *srcLen == 0); /* change it */ + *status = (*destLen == 0 && *srcLen == 0) ? + CODER_STATUS_FINISHED_WITH_MARK : + CODER_STATUS_NOT_FINISHED; + return res; +} + +static SRes SbState_SetFromMethod(IStateCoder *p, ISzAllocPtr alloc) +{ + CSbDec *decoder = (CSbDec *)p->p; + if (!decoder) + { + decoder = (CSbDec *)ISzAlloc_Alloc(alloc, sizeof(CSbDec)); + if (!decoder) + return SZ_ERROR_MEM; + p->p = decoder; + p->Free = SbState_Free; + p->SetProps = SbState_SetProps; + p->Init = SbState_Init; + p->Code2 = SbState_Code2; + p->Filter = NULL; + } + SbDec_Construct(decoder); + SbDec_SetAlloc(decoder, alloc); + return SZ_OK; +} + +#endif + + + +/* ---------- Lzma2 ---------- */ + +typedef struct +{ + CLzma2Dec decoder; + BoolInt outBufMode; +} CLzma2Dec_Spec; + + +static void Lzma2State_Free(void *pp, ISzAllocPtr alloc) +{ + CLzma2Dec_Spec *p = (CLzma2Dec_Spec *)pp; + if (p->outBufMode) + Lzma2Dec_FreeProbs(&p->decoder, alloc); + else + Lzma2Dec_Free(&p->decoder, alloc); + ISzAlloc_Free(alloc, pp); +} + +static SRes Lzma2State_SetProps(void *pp, const Byte *props, size_t propSize, ISzAllocPtr alloc) +{ + if (propSize != 1) + return SZ_ERROR_UNSUPPORTED; + { + CLzma2Dec_Spec *p = (CLzma2Dec_Spec *)pp; + if (p->outBufMode) + return Lzma2Dec_AllocateProbs(&p->decoder, props[0], alloc); + else + return Lzma2Dec_Allocate(&p->decoder, props[0], alloc); + } +} + +static void Lzma2State_Init(void *pp) +{ + Lzma2Dec_Init(&((CLzma2Dec_Spec *)pp)->decoder); +} + + +/* + if (outBufMode), then (dest) is not used. Use NULL. + Data is unpacked to (spec->decoder.decoder.dic) output buffer. +*/ + +static SRes Lzma2State_Code2(void *pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, + // int *wasFinished, + ECoderStatus *status) +{ + CLzma2Dec_Spec *spec = (CLzma2Dec_Spec *)pp; + ELzmaStatus status2; + /* ELzmaFinishMode fm = (finishMode == LZMA_FINISH_ANY) ? LZMA_FINISH_ANY : LZMA_FINISH_END; */ + SRes res; + UNUSED_VAR(srcWasFinished); + if (spec->outBufMode) + { + SizeT dicPos = spec->decoder.decoder.dicPos; + SizeT dicLimit = dicPos + *destLen; + res = Lzma2Dec_DecodeToDic(&spec->decoder, dicLimit, src, srcLen, (ELzmaFinishMode)finishMode, &status2); + *destLen = spec->decoder.decoder.dicPos - dicPos; + } + else + res = Lzma2Dec_DecodeToBuf(&spec->decoder, dest, destLen, src, srcLen, (ELzmaFinishMode)finishMode, &status2); + // *wasFinished = (status2 == LZMA_STATUS_FINISHED_WITH_MARK); + // ECoderStatus values are identical to ELzmaStatus values of LZMA2 decoder + *status = (ECoderStatus)status2; + return res; +} + + +static SRes Lzma2State_SetFromMethod(IStateCoder *p, Byte *outBuf, size_t outBufSize, ISzAllocPtr alloc) +{ + CLzma2Dec_Spec *spec = (CLzma2Dec_Spec *)p->p; + if (!spec) + { + spec = (CLzma2Dec_Spec *)ISzAlloc_Alloc(alloc, sizeof(CLzma2Dec_Spec)); + if (!spec) + return SZ_ERROR_MEM; + p->p = spec; + p->Free = Lzma2State_Free; + p->SetProps = Lzma2State_SetProps; + p->Init = Lzma2State_Init; + p->Code2 = Lzma2State_Code2; + p->Filter = NULL; + Lzma2Dec_Construct(&spec->decoder); + } + spec->outBufMode = False; + if (outBuf) + { + spec->outBufMode = True; + spec->decoder.decoder.dic = outBuf; + spec->decoder.decoder.dicBufSize = outBufSize; + } + return SZ_OK; +} + + +static SRes Lzma2State_ResetOutBuf(IStateCoder *p, Byte *outBuf, size_t outBufSize) +{ + CLzma2Dec_Spec *spec = (CLzma2Dec_Spec *)p->p; + if ((spec->outBufMode && !outBuf) || (!spec->outBufMode && outBuf)) + return SZ_ERROR_FAIL; + if (outBuf) + { + spec->decoder.decoder.dic = outBuf; + spec->decoder.decoder.dicBufSize = outBufSize; + } + return SZ_OK; +} + + + +static void MixCoder_Construct(CMixCoder *p, ISzAllocPtr alloc) +{ + unsigned i; + p->alloc = alloc; + p->buf = NULL; + p->numCoders = 0; + + p->outBufSize = 0; + p->outBuf = NULL; + // p->SingleBufMode = False; + + for (i = 0; i < MIXCODER_NUM_FILTERS_MAX; i++) + p->coders[i].p = NULL; +} + + +static void MixCoder_Free(CMixCoder *p) +{ + unsigned i; + p->numCoders = 0; + for (i = 0; i < MIXCODER_NUM_FILTERS_MAX; i++) + { + IStateCoder *sc = &p->coders[i]; + if (sc->p) + { + sc->Free(sc->p, p->alloc); + sc->p = NULL; + } + } + if (p->buf) + { + ISzAlloc_Free(p->alloc, p->buf); + p->buf = NULL; /* 9.31: the BUG was fixed */ + } +} + +static void MixCoder_Init(CMixCoder *p) +{ + unsigned i; + for (i = 0; i < MIXCODER_NUM_FILTERS_MAX - 1; i++) + { + p->size[i] = 0; + p->pos[i] = 0; + p->finished[i] = 0; + } + for (i = 0; i < p->numCoders; i++) + { + IStateCoder *coder = &p->coders[i]; + coder->Init(coder->p); + p->results[i] = SZ_OK; + } + p->outWritten = 0; + p->wasFinished = False; + p->res = SZ_OK; + p->status = CODER_STATUS_NOT_SPECIFIED; +} + + +static SRes MixCoder_SetFromMethod(CMixCoder *p, unsigned coderIndex, UInt64 methodId, Byte *outBuf, size_t outBufSize) +{ + IStateCoder *sc = &p->coders[coderIndex]; + p->ids[coderIndex] = methodId; + switch (methodId) + { + case XZ_ID_LZMA2: return Lzma2State_SetFromMethod(sc, outBuf, outBufSize, p->alloc); + #ifdef USE_SUBBLOCK + case XZ_ID_Subblock: return SbState_SetFromMethod(sc, p->alloc); + #endif + } + if (coderIndex == 0) + return SZ_ERROR_UNSUPPORTED; + return BraState_SetFromMethod(sc, methodId, 0, p->alloc); +} + + +static SRes MixCoder_ResetFromMethod(CMixCoder *p, unsigned coderIndex, UInt64 methodId, Byte *outBuf, size_t outBufSize) +{ + IStateCoder *sc = &p->coders[coderIndex]; + switch (methodId) + { + case XZ_ID_LZMA2: return Lzma2State_ResetOutBuf(sc, outBuf, outBufSize); + } + return SZ_ERROR_UNSUPPORTED; +} + + + +/* + if (destFinish) - then unpack data block is finished at (*destLen) position, + and we can return data that were not processed by filter + +output (status) can be : + CODER_STATUS_NOT_FINISHED + CODER_STATUS_FINISHED_WITH_MARK + CODER_STATUS_NEEDS_MORE_INPUT - not implemented still +*/ + +static SRes MixCoder_Code(CMixCoder *p, + Byte *dest, SizeT *destLen, int destFinish, + const Byte *src, SizeT *srcLen, int srcWasFinished, + ECoderFinishMode finishMode) +{ + SizeT destLenOrig = *destLen; + SizeT srcLenOrig = *srcLen; + + *destLen = 0; + *srcLen = 0; + + if (p->wasFinished) + return p->res; + + p->status = CODER_STATUS_NOT_FINISHED; + + // if (p->SingleBufMode) + if (p->outBuf) + { + SRes res; + SizeT destLen2, srcLen2; + int wasFinished; + + PRF_STR("------- MixCoder Single ----------"); + + srcLen2 = srcLenOrig; + destLen2 = destLenOrig; + + { + IStateCoder *coder = &p->coders[0]; + res = coder->Code2(coder->p, NULL, &destLen2, src, &srcLen2, srcWasFinished, finishMode, + // &wasFinished, + &p->status); + wasFinished = (p->status == CODER_STATUS_FINISHED_WITH_MARK); + } + + p->res = res; + + /* + if (wasFinished) + p->status = CODER_STATUS_FINISHED_WITH_MARK; + else + { + if (res == SZ_OK) + if (destLen2 != destLenOrig) + p->status = CODER_STATUS_NEEDS_MORE_INPUT; + } + */ + + + *srcLen = srcLen2; + src += srcLen2; + p->outWritten += destLen2; + + if (res != SZ_OK || srcWasFinished || wasFinished) + p->wasFinished = True; + + if (p->numCoders == 1) + *destLen = destLen2; + else if (p->wasFinished) + { + unsigned i; + size_t processed = p->outWritten; + + for (i = 1; i < p->numCoders; i++) + { + IStateCoder *coder = &p->coders[i]; + processed = coder->Filter(coder->p, p->outBuf, processed); + if (wasFinished || (destFinish && p->outWritten == destLenOrig)) + processed = p->outWritten; + PRF_STR_INT("filter", i); + } + *destLen = processed; + } + return res; + } + + PRF_STR("standard mix"); + + if (p->numCoders != 1) + { + if (!p->buf) + { + p->buf = (Byte *)ISzAlloc_Alloc(p->alloc, CODER_BUF_SIZE * (MIXCODER_NUM_FILTERS_MAX - 1)); + if (!p->buf) + return SZ_ERROR_MEM; + } + + finishMode = CODER_FINISH_ANY; + } + + for (;;) + { + BoolInt processed = False; + BoolInt allFinished = True; + SRes resMain = SZ_OK; + unsigned i; + + p->status = CODER_STATUS_NOT_FINISHED; + /* + if (p->numCoders == 1 && *destLen == destLenOrig && finishMode == LZMA_FINISH_ANY) + break; + */ + + for (i = 0; i < p->numCoders; i++) + { + SRes res; + IStateCoder *coder = &p->coders[i]; + Byte *dest2; + SizeT destLen2, srcLen2; // destLen2_Orig; + const Byte *src2; + int srcFinished2; + int encodingWasFinished; + ECoderStatus status2; + + if (i == 0) + { + src2 = src; + srcLen2 = srcLenOrig - *srcLen; + srcFinished2 = srcWasFinished; + } + else + { + size_t k = i - 1; + src2 = p->buf + (CODER_BUF_SIZE * k) + p->pos[k]; + srcLen2 = p->size[k] - p->pos[k]; + srcFinished2 = p->finished[k]; + } + + if (i == p->numCoders - 1) + { + dest2 = dest; + destLen2 = destLenOrig - *destLen; + } + else + { + if (p->pos[i] != p->size[i]) + continue; + dest2 = p->buf + (CODER_BUF_SIZE * i); + destLen2 = CODER_BUF_SIZE; + } + + // destLen2_Orig = destLen2; + + if (p->results[i] != SZ_OK) + { + if (resMain == SZ_OK) + resMain = p->results[i]; + continue; + } + + res = coder->Code2(coder->p, + dest2, &destLen2, + src2, &srcLen2, srcFinished2, + finishMode, + // &encodingWasFinished, + &status2); + + if (res != SZ_OK) + { + p->results[i] = res; + if (resMain == SZ_OK) + resMain = res; + } + + encodingWasFinished = (status2 == CODER_STATUS_FINISHED_WITH_MARK); + + if (!encodingWasFinished) + { + allFinished = False; + if (p->numCoders == 1 && res == SZ_OK) + p->status = status2; + } + + if (i == 0) + { + *srcLen += srcLen2; + src += srcLen2; + } + else + p->pos[(size_t)i - 1] += srcLen2; + + if (i == p->numCoders - 1) + { + *destLen += destLen2; + dest += destLen2; + } + else + { + p->size[i] = destLen2; + p->pos[i] = 0; + p->finished[i] = encodingWasFinished; + } + + if (destLen2 != 0 || srcLen2 != 0) + processed = True; + } + + if (!processed) + { + if (allFinished) + p->status = CODER_STATUS_FINISHED_WITH_MARK; + return resMain; + } + } +} + + +SRes Xz_ParseHeader(CXzStreamFlags *p, const Byte *buf) +{ + *p = (CXzStreamFlags)GetBe16(buf + XZ_SIG_SIZE); + if (CrcCalc(buf + XZ_SIG_SIZE, XZ_STREAM_FLAGS_SIZE) != + GetUi32(buf + XZ_SIG_SIZE + XZ_STREAM_FLAGS_SIZE)) + return SZ_ERROR_NO_ARCHIVE; + return XzFlags_IsSupported(*p) ? SZ_OK : SZ_ERROR_UNSUPPORTED; +} + +static BoolInt Xz_CheckFooter(CXzStreamFlags flags, UInt64 indexSize, const Byte *buf) +{ + return indexSize == (((UInt64)GetUi32(buf + 4) + 1) << 2) + && GetUi32(buf) == CrcCalc(buf + 4, 6) + && flags == GetBe16(buf + 8) + && buf[10] == XZ_FOOTER_SIG_0 + && buf[11] == XZ_FOOTER_SIG_1; +} + +#define READ_VARINT_AND_CHECK(buf, pos, size, res) \ + { unsigned s = Xz_ReadVarInt(buf + pos, size - pos, res); \ + if (s == 0) return SZ_ERROR_ARCHIVE; \ + pos += s; } + + +static BoolInt XzBlock_AreSupportedFilters(const CXzBlock *p) +{ + unsigned numFilters = XzBlock_GetNumFilters(p) - 1; + unsigned i; + { + const CXzFilter *f = &p->filters[numFilters]; + if (f->id != XZ_ID_LZMA2 || f->propsSize != 1 || f->props[0] > 40) + return False; + } + + for (i = 0; i < numFilters; i++) + { + const CXzFilter *f = &p->filters[i]; + if (f->id == XZ_ID_Delta) + { + if (f->propsSize != 1) + return False; + } + else if (f->id < XZ_ID_Delta + || f->id > XZ_ID_SPARC + || (f->propsSize != 0 && f->propsSize != 4)) + return False; + } + return True; +} + + +SRes XzBlock_Parse(CXzBlock *p, const Byte *header) +{ + unsigned pos; + unsigned numFilters, i; + unsigned headerSize = (unsigned)header[0] << 2; + + /* (headerSize != 0) : another code checks */ + + if (CrcCalc(header, headerSize) != GetUi32(header + headerSize)) + return SZ_ERROR_ARCHIVE; + + pos = 1; + p->flags = header[pos++]; + + p->packSize = (UInt64)(Int64)-1; + if (XzBlock_HasPackSize(p)) + { + READ_VARINT_AND_CHECK(header, pos, headerSize, &p->packSize); + if (p->packSize == 0 || p->packSize + headerSize >= (UInt64)1 << 63) + return SZ_ERROR_ARCHIVE; + } + + p->unpackSize = (UInt64)(Int64)-1; + if (XzBlock_HasUnpackSize(p)) + READ_VARINT_AND_CHECK(header, pos, headerSize, &p->unpackSize); + + numFilters = XzBlock_GetNumFilters(p); + for (i = 0; i < numFilters; i++) + { + CXzFilter *filter = p->filters + i; + UInt64 size; + READ_VARINT_AND_CHECK(header, pos, headerSize, &filter->id); + READ_VARINT_AND_CHECK(header, pos, headerSize, &size); + if (size > headerSize - pos || size > XZ_FILTER_PROPS_SIZE_MAX) + return SZ_ERROR_ARCHIVE; + filter->propsSize = (UInt32)size; + memcpy(filter->props, header + pos, (size_t)size); + pos += (unsigned)size; + + #ifdef XZ_DUMP + printf("\nf[%u] = %2X: ", i, (unsigned)filter->id); + { + unsigned i; + for (i = 0; i < size; i++) + printf(" %2X", filter->props[i]); + } + #endif + } + + if (XzBlock_HasUnsupportedFlags(p)) + return SZ_ERROR_UNSUPPORTED; + + while (pos < headerSize) + if (header[pos++] != 0) + return SZ_ERROR_ARCHIVE; + return SZ_OK; +} + + + + +static SRes XzDecMix_Init(CMixCoder *p, const CXzBlock *block, Byte *outBuf, size_t outBufSize) +{ + unsigned i; + BoolInt needReInit = True; + unsigned numFilters = XzBlock_GetNumFilters(block); + + if (numFilters == p->numCoders && ((p->outBuf && outBuf) || (!p->outBuf && !outBuf))) + { + needReInit = False; + for (i = 0; i < numFilters; i++) + if (p->ids[i] != block->filters[numFilters - 1 - i].id) + { + needReInit = True; + break; + } + } + + // p->SingleBufMode = (outBuf != NULL); + p->outBuf = outBuf; + p->outBufSize = outBufSize; + + // p->SingleBufMode = False; + // outBuf = NULL; + + if (needReInit) + { + MixCoder_Free(p); + for (i = 0; i < numFilters; i++) + { + RINOK(MixCoder_SetFromMethod(p, i, block->filters[numFilters - 1 - i].id, outBuf, outBufSize)); + } + p->numCoders = numFilters; + } + else + { + RINOK(MixCoder_ResetFromMethod(p, 0, block->filters[numFilters - 1].id, outBuf, outBufSize)); + } + + for (i = 0; i < numFilters; i++) + { + const CXzFilter *f = &block->filters[numFilters - 1 - i]; + IStateCoder *sc = &p->coders[i]; + RINOK(sc->SetProps(sc->p, f->props, f->propsSize, p->alloc)); + } + + MixCoder_Init(p); + return SZ_OK; +} + + + +void XzUnpacker_Init(CXzUnpacker *p) +{ + p->state = XZ_STATE_STREAM_HEADER; + p->pos = 0; + p->numStartedStreams = 0; + p->numFinishedStreams = 0; + p->numTotalBlocks = 0; + p->padSize = 0; + p->decodeOnlyOneBlock = 0; + + p->parseMode = False; + p->decodeToStreamSignature = False; + + // p->outBuf = NULL; + // p->outBufSize = 0; + p->outDataWritten = 0; +} + + +void XzUnpacker_SetOutBuf(CXzUnpacker *p, Byte *outBuf, size_t outBufSize) +{ + p->outBuf = outBuf; + p->outBufSize = outBufSize; +} + + +void XzUnpacker_Construct(CXzUnpacker *p, ISzAllocPtr alloc) +{ + MixCoder_Construct(&p->decoder, alloc); + p->outBuf = NULL; + p->outBufSize = 0; + XzUnpacker_Init(p); +} + + +void XzUnpacker_Free(CXzUnpacker *p) +{ + MixCoder_Free(&p->decoder); +} + + +void XzUnpacker_PrepareToRandomBlockDecoding(CXzUnpacker *p) +{ + p->indexSize = 0; + p->numBlocks = 0; + Sha256_Init(&p->sha); + p->state = XZ_STATE_BLOCK_HEADER; + p->pos = 0; + p->decodeOnlyOneBlock = 1; +} + + +static void XzUnpacker_UpdateIndex(CXzUnpacker *p, UInt64 packSize, UInt64 unpackSize) +{ + Byte temp[32]; + unsigned num = Xz_WriteVarInt(temp, packSize); + num += Xz_WriteVarInt(temp + num, unpackSize); + Sha256_Update(&p->sha, temp, num); + p->indexSize += num; + p->numBlocks++; +} + + + +SRes XzUnpacker_Code(CXzUnpacker *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, int srcFinished, + ECoderFinishMode finishMode, ECoderStatus *status) +{ + SizeT destLenOrig = *destLen; + SizeT srcLenOrig = *srcLen; + *destLen = 0; + *srcLen = 0; + *status = CODER_STATUS_NOT_SPECIFIED; + + for (;;) + { + SizeT srcRem; + + if (p->state == XZ_STATE_BLOCK) + { + SizeT destLen2 = destLenOrig - *destLen; + SizeT srcLen2 = srcLenOrig - *srcLen; + SRes res; + + ECoderFinishMode finishMode2 = finishMode; + BoolInt srcFinished2 = srcFinished; + BoolInt destFinish = False; + + if (p->block.packSize != (UInt64)(Int64)-1) + { + UInt64 rem = p->block.packSize - p->packSize; + if (srcLen2 >= rem) + { + srcFinished2 = True; + srcLen2 = (SizeT)rem; + } + if (rem == 0 && p->block.unpackSize == p->unpackSize) + return SZ_ERROR_DATA; + } + + if (p->block.unpackSize != (UInt64)(Int64)-1) + { + UInt64 rem = p->block.unpackSize - p->unpackSize; + if (destLen2 >= rem) + { + destFinish = True; + finishMode2 = CODER_FINISH_END; + destLen2 = (SizeT)rem; + } + } + + /* + if (srcLen2 == 0 && destLen2 == 0) + { + *status = CODER_STATUS_NOT_FINISHED; + return SZ_OK; + } + */ + + { + res = MixCoder_Code(&p->decoder, + (p->outBuf ? NULL : dest), &destLen2, destFinish, + src, &srcLen2, srcFinished2, + finishMode2); + + *status = p->decoder.status; + XzCheck_Update(&p->check, (p->outBuf ? p->outBuf + p->outDataWritten : dest), destLen2); + if (!p->outBuf) + dest += destLen2; + p->outDataWritten += destLen2; + } + + (*srcLen) += srcLen2; + src += srcLen2; + p->packSize += srcLen2; + (*destLen) += destLen2; + p->unpackSize += destLen2; + + RINOK(res); + + if (*status != CODER_STATUS_FINISHED_WITH_MARK) + { + if (p->block.packSize == p->packSize + && *status == CODER_STATUS_NEEDS_MORE_INPUT) + { + PRF_STR("CODER_STATUS_NEEDS_MORE_INPUT"); + *status = CODER_STATUS_NOT_SPECIFIED; + return SZ_ERROR_DATA; + } + + return SZ_OK; + } + { + XzUnpacker_UpdateIndex(p, XzUnpacker_GetPackSizeForIndex(p), p->unpackSize); + p->state = XZ_STATE_BLOCK_FOOTER; + p->pos = 0; + p->alignPos = 0; + *status = CODER_STATUS_NOT_SPECIFIED; + + if ((p->block.packSize != (UInt64)(Int64)-1 && p->block.packSize != p->packSize) + || (p->block.unpackSize != (UInt64)(Int64)-1 && p->block.unpackSize != p->unpackSize)) + { + PRF_STR("ERROR: block.size mismatch"); + return SZ_ERROR_DATA; + } + } + // continue; + } + + srcRem = srcLenOrig - *srcLen; + + // XZ_STATE_BLOCK_FOOTER can transit to XZ_STATE_BLOCK_HEADER without input bytes + if (srcRem == 0 && p->state != XZ_STATE_BLOCK_FOOTER) + { + *status = CODER_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + + switch (p->state) + { + case XZ_STATE_STREAM_HEADER: + { + if (p->pos < XZ_STREAM_HEADER_SIZE) + { + if (p->pos < XZ_SIG_SIZE && *src != XZ_SIG[p->pos]) + return SZ_ERROR_NO_ARCHIVE; + if (p->decodeToStreamSignature) + return SZ_OK; + p->buf[p->pos++] = *src++; + (*srcLen)++; + } + else + { + RINOK(Xz_ParseHeader(&p->streamFlags, p->buf)); + p->numStartedStreams++; + p->indexSize = 0; + p->numBlocks = 0; + Sha256_Init(&p->sha); + p->state = XZ_STATE_BLOCK_HEADER; + p->pos = 0; + } + break; + } + + case XZ_STATE_BLOCK_HEADER: + { + if (p->pos == 0) + { + p->buf[p->pos++] = *src++; + (*srcLen)++; + if (p->buf[0] == 0) + { + if (p->decodeOnlyOneBlock) + return SZ_ERROR_DATA; + p->indexPreSize = 1 + Xz_WriteVarInt(p->buf + 1, p->numBlocks); + p->indexPos = p->indexPreSize; + p->indexSize += p->indexPreSize; + Sha256_Final(&p->sha, p->shaDigest); + Sha256_Init(&p->sha); + p->crc = CrcUpdate(CRC_INIT_VAL, p->buf, p->indexPreSize); + p->state = XZ_STATE_STREAM_INDEX; + break; + } + p->blockHeaderSize = ((UInt32)p->buf[0] << 2) + 4; + break; + } + + if (p->pos != p->blockHeaderSize) + { + UInt32 cur = p->blockHeaderSize - p->pos; + if (cur > srcRem) + cur = (UInt32)srcRem; + memcpy(p->buf + p->pos, src, cur); + p->pos += cur; + (*srcLen) += cur; + src += cur; + } + else + { + RINOK(XzBlock_Parse(&p->block, p->buf)); + if (!XzBlock_AreSupportedFilters(&p->block)) + return SZ_ERROR_UNSUPPORTED; + p->numTotalBlocks++; + p->state = XZ_STATE_BLOCK; + p->packSize = 0; + p->unpackSize = 0; + XzCheck_Init(&p->check, XzFlags_GetCheckType(p->streamFlags)); + if (p->parseMode) + { + p->headerParsedOk = True; + return SZ_OK; + } + RINOK(XzDecMix_Init(&p->decoder, &p->block, p->outBuf, p->outBufSize)); + } + break; + } + + case XZ_STATE_BLOCK_FOOTER: + { + if ((((unsigned)p->packSize + p->alignPos) & 3) != 0) + { + if (srcRem == 0) + { + *status = CODER_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + (*srcLen)++; + p->alignPos++; + if (*src++ != 0) + return SZ_ERROR_CRC; + } + else + { + UInt32 checkSize = XzFlags_GetCheckSize(p->streamFlags); + UInt32 cur = checkSize - p->pos; + if (cur != 0) + { + if (srcRem == 0) + { + *status = CODER_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + if (cur > srcRem) + cur = (UInt32)srcRem; + memcpy(p->buf + p->pos, src, cur); + p->pos += cur; + (*srcLen) += cur; + src += cur; + if (checkSize != p->pos) + break; + } + { + Byte digest[XZ_CHECK_SIZE_MAX]; + p->state = XZ_STATE_BLOCK_HEADER; + p->pos = 0; + if (XzCheck_Final(&p->check, digest) && memcmp(digest, p->buf, checkSize) != 0) + return SZ_ERROR_CRC; + if (p->decodeOnlyOneBlock) + { + *status = CODER_STATUS_FINISHED_WITH_MARK; + return SZ_OK; + } + } + } + break; + } + + case XZ_STATE_STREAM_INDEX: + { + if (p->pos < p->indexPreSize) + { + (*srcLen)++; + if (*src++ != p->buf[p->pos++]) + return SZ_ERROR_CRC; + } + else + { + if (p->indexPos < p->indexSize) + { + UInt64 cur = p->indexSize - p->indexPos; + if (srcRem > cur) + srcRem = (SizeT)cur; + p->crc = CrcUpdate(p->crc, src, srcRem); + Sha256_Update(&p->sha, src, srcRem); + (*srcLen) += srcRem; + src += srcRem; + p->indexPos += srcRem; + } + else if ((p->indexPos & 3) != 0) + { + Byte b = *src++; + p->crc = CRC_UPDATE_BYTE(p->crc, b); + (*srcLen)++; + p->indexPos++; + p->indexSize++; + if (b != 0) + return SZ_ERROR_CRC; + } + else + { + Byte digest[SHA256_DIGEST_SIZE]; + p->state = XZ_STATE_STREAM_INDEX_CRC; + p->indexSize += 4; + p->pos = 0; + Sha256_Final(&p->sha, digest); + if (memcmp(digest, p->shaDigest, SHA256_DIGEST_SIZE) != 0) + return SZ_ERROR_CRC; + } + } + break; + } + + case XZ_STATE_STREAM_INDEX_CRC: + { + if (p->pos < 4) + { + (*srcLen)++; + p->buf[p->pos++] = *src++; + } + else + { + const Byte *ptr = p->buf; + p->state = XZ_STATE_STREAM_FOOTER; + p->pos = 0; + if (CRC_GET_DIGEST(p->crc) != GetUi32(ptr)) + return SZ_ERROR_CRC; + } + break; + } + + case XZ_STATE_STREAM_FOOTER: + { + UInt32 cur = XZ_STREAM_FOOTER_SIZE - p->pos; + if (cur > srcRem) + cur = (UInt32)srcRem; + memcpy(p->buf + p->pos, src, cur); + p->pos += cur; + (*srcLen) += cur; + src += cur; + if (p->pos == XZ_STREAM_FOOTER_SIZE) + { + p->state = XZ_STATE_STREAM_PADDING; + p->numFinishedStreams++; + p->padSize = 0; + if (!Xz_CheckFooter(p->streamFlags, p->indexSize, p->buf)) + return SZ_ERROR_CRC; + } + break; + } + + case XZ_STATE_STREAM_PADDING: + { + if (*src != 0) + { + if (((UInt32)p->padSize & 3) != 0) + return SZ_ERROR_NO_ARCHIVE; + p->pos = 0; + p->state = XZ_STATE_STREAM_HEADER; + } + else + { + (*srcLen)++; + src++; + p->padSize++; + } + break; + } + + case XZ_STATE_BLOCK: break; /* to disable GCC warning */ + } + } + /* + if (p->state == XZ_STATE_FINISHED) + *status = CODER_STATUS_FINISHED_WITH_MARK; + return SZ_OK; + */ +} + + +SRes XzUnpacker_CodeFull(CXzUnpacker *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, + ECoderFinishMode finishMode, ECoderStatus *status) +{ + XzUnpacker_Init(p); + XzUnpacker_SetOutBuf(p, dest, *destLen); + + return XzUnpacker_Code(p, + NULL, destLen, + src, srcLen, True, + finishMode, status); +} + + +BoolInt XzUnpacker_IsBlockFinished(const CXzUnpacker *p) +{ + return (p->state == XZ_STATE_BLOCK_HEADER) && (p->pos == 0); +} + +BoolInt XzUnpacker_IsStreamWasFinished(const CXzUnpacker *p) +{ + return (p->state == XZ_STATE_STREAM_PADDING) && (((UInt32)p->padSize & 3) == 0); +} + +UInt64 XzUnpacker_GetExtraSize(const CXzUnpacker *p) +{ + UInt64 num = 0; + if (p->state == XZ_STATE_STREAM_PADDING) + num = p->padSize; + else if (p->state == XZ_STATE_STREAM_HEADER) + num = p->padSize + p->pos; + return num; +} + + + + + + + + + + + + + + + + + + + + + +#ifndef _7ZIP_ST +#include "MtDec.h" +#endif + + +void XzDecMtProps_Init(CXzDecMtProps *p) +{ + p->inBufSize_ST = 1 << 18; + p->outStep_ST = 1 << 20; + p->ignoreErrors = False; + + #ifndef _7ZIP_ST + p->numThreads = 1; + p->inBufSize_MT = 1 << 18; + p->memUseMax = sizeof(size_t) << 28; + #endif +} + + + +#ifndef _7ZIP_ST + +/* ---------- CXzDecMtThread ---------- */ + +typedef struct +{ + Byte *outBuf; + size_t outBufSize; + size_t outPreSize; + size_t inPreSize; + size_t inPreHeaderSize; + size_t blockPackSize_for_Index; // including block header and checksum. + size_t blockPackTotal; // including stream header, block header and checksum. + size_t inCodeSize; + size_t outCodeSize; + ECoderStatus status; + SRes codeRes; + BoolInt skipMode; + // BoolInt finishedWithMark; + EMtDecParseState parseState; + BoolInt parsing_Truncated; + BoolInt atBlockHeader; + CXzStreamFlags streamFlags; + // UInt64 numFinishedStreams + UInt64 numStreams; + UInt64 numTotalBlocks; + UInt64 numBlocks; + + BoolInt dec_created; + CXzUnpacker dec; + + Byte mtPad[1 << 7]; +} CXzDecMtThread; + +#endif + + +/* ---------- CXzDecMt ---------- */ + +typedef struct +{ + CAlignOffsetAlloc alignOffsetAlloc; + ISzAllocPtr allocMid; + + CXzDecMtProps props; + size_t unpackBlockMaxSize; + + ISeqInStream *inStream; + ISeqOutStream *outStream; + ICompressProgress *progress; + + BoolInt finishMode; + BoolInt outSize_Defined; + UInt64 outSize; + + UInt64 outProcessed; + UInt64 inProcessed; + UInt64 readProcessed; + BoolInt readWasFinished; + SRes readRes; + SRes writeRes; + + Byte *outBuf; + size_t outBufSize; + Byte *inBuf; + size_t inBufSize; + + CXzUnpacker dec; + + ECoderStatus status; + SRes codeRes; + + #ifndef _7ZIP_ST + BoolInt mainDecoderWasCalled; + // int statErrorDefined; + int finishedDecoderIndex; + + // global values that are used in Parse stage + CXzStreamFlags streamFlags; + // UInt64 numFinishedStreams + UInt64 numStreams; + UInt64 numTotalBlocks; + UInt64 numBlocks; + + // UInt64 numBadBlocks; + SRes mainErrorCode; // it's set to error code, if the size Code() output doesn't patch the size from Parsing stage + // it can be = SZ_ERROR_INPUT_EOF + // it can be = SZ_ERROR_DATA, in some another cases + BoolInt isBlockHeaderState_Parse; + BoolInt isBlockHeaderState_Write; + UInt64 outProcessed_Parse; + BoolInt parsing_Truncated; + + BoolInt mtc_WasConstructed; + CMtDec mtc; + CXzDecMtThread coders[MTDEC__THREADS_MAX]; + #endif + +} CXzDecMt; + + + +CXzDecMtHandle XzDecMt_Create(ISzAllocPtr alloc, ISzAllocPtr allocMid) +{ + CXzDecMt *p = (CXzDecMt *)ISzAlloc_Alloc(alloc, sizeof(CXzDecMt)); + if (!p) + return NULL; + + AlignOffsetAlloc_CreateVTable(&p->alignOffsetAlloc); + p->alignOffsetAlloc.baseAlloc = alloc; + p->alignOffsetAlloc.numAlignBits = 7; + p->alignOffsetAlloc.offset = 0; + + p->allocMid = allocMid; + + p->outBuf = NULL; + p->outBufSize = 0; + p->inBuf = NULL; + p->inBufSize = 0; + + XzUnpacker_Construct(&p->dec, &p->alignOffsetAlloc.vt); + + p->unpackBlockMaxSize = 0; + + XzDecMtProps_Init(&p->props); + + #ifndef _7ZIP_ST + p->mtc_WasConstructed = False; + { + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CXzDecMtThread *coder = &p->coders[i]; + coder->dec_created = False; + coder->outBuf = NULL; + coder->outBufSize = 0; + } + } + #endif + + return p; +} + + +#ifndef _7ZIP_ST + +static void XzDecMt_FreeOutBufs(CXzDecMt *p) +{ + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CXzDecMtThread *coder = &p->coders[i]; + if (coder->outBuf) + { + ISzAlloc_Free(p->allocMid, coder->outBuf); + coder->outBuf = NULL; + coder->outBufSize = 0; + } + } + p->unpackBlockMaxSize = 0; +} + +#endif + + + +static void XzDecMt_FreeSt(CXzDecMt *p) +{ + XzUnpacker_Free(&p->dec); + + if (p->outBuf) + { + ISzAlloc_Free(p->allocMid, p->outBuf); + p->outBuf = NULL; + } + p->outBufSize = 0; + + if (p->inBuf) + { + ISzAlloc_Free(p->allocMid, p->inBuf); + p->inBuf = NULL; + } + p->inBufSize = 0; +} + + +void XzDecMt_Destroy(CXzDecMtHandle pp) +{ + CXzDecMt *p = (CXzDecMt *)pp; + + XzDecMt_FreeSt(p); + + #ifndef _7ZIP_ST + + if (p->mtc_WasConstructed) + { + MtDec_Destruct(&p->mtc); + p->mtc_WasConstructed = False; + } + { + unsigned i; + for (i = 0; i < MTDEC__THREADS_MAX; i++) + { + CXzDecMtThread *t = &p->coders[i]; + if (t->dec_created) + { + // we don't need to free dict here + XzUnpacker_Free(&t->dec); + t->dec_created = False; + } + } + } + XzDecMt_FreeOutBufs(p); + + #endif + + ISzAlloc_Free(p->alignOffsetAlloc.baseAlloc, pp); +} + + + +#ifndef _7ZIP_ST + +static void XzDecMt_Callback_Parse(void *obj, unsigned coderIndex, CMtDecCallbackInfo *cc) +{ + CXzDecMt *me = (CXzDecMt *)obj; + CXzDecMtThread *coder = &me->coders[coderIndex]; + size_t srcSize = cc->srcSize; + + cc->srcSize = 0; + cc->outPos = 0; + cc->state = MTDEC_PARSE_CONTINUE; + + cc->canCreateNewThread = True; + + if (cc->startCall) + { + coder->outPreSize = 0; + coder->inPreSize = 0; + coder->inPreHeaderSize = 0; + coder->parseState = MTDEC_PARSE_CONTINUE; + coder->parsing_Truncated = False; + coder->skipMode = False; + coder->codeRes = SZ_OK; + coder->status = CODER_STATUS_NOT_SPECIFIED; + coder->inCodeSize = 0; + coder->outCodeSize = 0; + + coder->numStreams = me->numStreams; + coder->numTotalBlocks = me->numTotalBlocks; + coder->numBlocks = me->numBlocks; + + if (!coder->dec_created) + { + XzUnpacker_Construct(&coder->dec, &me->alignOffsetAlloc.vt); + coder->dec_created = True; + } + + XzUnpacker_Init(&coder->dec); + + if (me->isBlockHeaderState_Parse) + { + coder->dec.streamFlags = me->streamFlags; + coder->atBlockHeader = True; + XzUnpacker_PrepareToRandomBlockDecoding(&coder->dec); + } + else + { + coder->atBlockHeader = False; + me->isBlockHeaderState_Parse = True; + } + + coder->dec.numStartedStreams = me->numStreams; + coder->dec.numTotalBlocks = me->numTotalBlocks; + coder->dec.numBlocks = me->numBlocks; + } + + while (!coder->skipMode) + { + ECoderStatus status; + SRes res; + size_t srcSize2 = srcSize; + size_t destSize = (size_t)0 - 1; + + coder->dec.parseMode = True; + coder->dec.headerParsedOk = False; + + PRF_STR_INT("Parse", srcSize2); + + res = XzUnpacker_Code(&coder->dec, + NULL, &destSize, + cc->src, &srcSize2, cc->srcFinished, + CODER_FINISH_END, &status); + + // PRF(printf(" res = %d, srcSize2 = %d", res, (unsigned)srcSize2)); + + coder->codeRes = res; + coder->status = status; + cc->srcSize += srcSize2; + srcSize -= srcSize2; + coder->inPreHeaderSize += srcSize2; + coder->inPreSize = coder->inPreHeaderSize; + + if (res != SZ_OK) + { + cc->state = + coder->parseState = MTDEC_PARSE_END; + /* + if (res == SZ_ERROR_MEM) + return res; + return SZ_OK; + */ + return; // res; + } + + if (coder->dec.headerParsedOk) + { + const CXzBlock *block = &coder->dec.block; + if (XzBlock_HasUnpackSize(block) + // && block->unpackSize <= me->props.outBlockMax + && XzBlock_HasPackSize(block)) + { + { + if (block->unpackSize * 2 * me->mtc.numStartedThreads > me->props.memUseMax) + { + cc->state = MTDEC_PARSE_OVERFLOW; + return; // SZ_OK; + } + } + { + UInt64 packSize = block->packSize; + UInt64 packSizeAligned = packSize + ((0 - (unsigned)packSize) & 3); + UInt32 checkSize = XzFlags_GetCheckSize(coder->dec.streamFlags); + UInt64 blockPackSum = coder->inPreSize + packSizeAligned + checkSize; + // if (blockPackSum <= me->props.inBlockMax) + // unpackBlockMaxSize + { + coder->blockPackSize_for_Index = (size_t)(coder->dec.blockHeaderSize + packSize + checkSize); + coder->blockPackTotal = (size_t)blockPackSum; + coder->outPreSize = (size_t)block->unpackSize; + coder->streamFlags = coder->dec.streamFlags; + me->streamFlags = coder->dec.streamFlags; + coder->skipMode = True; + break; + } + } + } + } + else + // if (coder->inPreSize <= me->props.inBlockMax) + { + if (!cc->srcFinished) + return; // SZ_OK; + cc->state = + coder->parseState = MTDEC_PARSE_END; + return; // SZ_OK; + } + cc->state = MTDEC_PARSE_OVERFLOW; + return; // SZ_OK; + } + + // ---------- skipMode ---------- + { + UInt64 rem = coder->blockPackTotal - coder->inPreSize; + size_t cur = srcSize; + if (cur > rem) + cur = (size_t)rem; + cc->srcSize += cur; + coder->inPreSize += cur; + srcSize -= cur; + + if (coder->inPreSize == coder->blockPackTotal) + { + if (srcSize == 0) + { + if (!cc->srcFinished) + return; // SZ_OK; + cc->state = MTDEC_PARSE_END; + } + else if ((cc->src)[cc->srcSize] == 0) // we check control byte of next block + cc->state = MTDEC_PARSE_END; + else + { + cc->state = MTDEC_PARSE_NEW; + + { + size_t blockMax = me->unpackBlockMaxSize; + if (blockMax < coder->outPreSize) + blockMax = coder->outPreSize; + { + UInt64 required = (UInt64)blockMax * (me->mtc.numStartedThreads + 1) * 2; + if (me->props.memUseMax < required) + cc->canCreateNewThread = False; + } + } + + if (me->outSize_Defined) + { + // next block can be zero size + const UInt64 rem2 = me->outSize - me->outProcessed_Parse; + if (rem2 < coder->outPreSize) + { + coder->parsing_Truncated = True; + cc->state = MTDEC_PARSE_END; + } + me->outProcessed_Parse += coder->outPreSize; + } + } + } + else if (cc->srcFinished) + cc->state = MTDEC_PARSE_END; + else + return; // SZ_OK; + + coder->parseState = cc->state; + cc->outPos = coder->outPreSize; + + me->numStreams = coder->dec.numStartedStreams; + me->numTotalBlocks = coder->dec.numTotalBlocks; + me->numBlocks = coder->dec.numBlocks + 1; + return; // SZ_OK; + } +} + + +static SRes XzDecMt_Callback_PreCode(void *pp, unsigned coderIndex) +{ + CXzDecMt *me = (CXzDecMt *)pp; + CXzDecMtThread *coder = &me->coders[coderIndex]; + Byte *dest; + + if (!coder->dec.headerParsedOk) + return SZ_OK; + + dest = coder->outBuf; + + if (!dest || coder->outBufSize < coder->outPreSize) + { + if (dest) + { + ISzAlloc_Free(me->allocMid, dest); + coder->outBuf = NULL; + coder->outBufSize = 0; + } + { + size_t outPreSize = coder->outPreSize; + if (outPreSize == 0) + outPreSize = 1; + dest = (Byte *)ISzAlloc_Alloc(me->allocMid, outPreSize); + } + if (!dest) + return SZ_ERROR_MEM; + coder->outBuf = dest; + coder->outBufSize = coder->outPreSize; + + if (coder->outBufSize > me->unpackBlockMaxSize) + me->unpackBlockMaxSize = coder->outBufSize; + } + + // return SZ_ERROR_MEM; + + XzUnpacker_SetOutBuf(&coder->dec, coder->outBuf, coder->outBufSize); + + { + SRes res = XzDecMix_Init(&coder->dec.decoder, &coder->dec.block, coder->outBuf, coder->outBufSize); + // res = SZ_ERROR_UNSUPPORTED; // to test + coder->codeRes = res; + if (res != SZ_OK) + { + // if (res == SZ_ERROR_MEM) return res; + if (me->props.ignoreErrors && res != SZ_ERROR_MEM) + return SZ_OK; + return res; + } + } + + return SZ_OK; +} + + +static SRes XzDecMt_Callback_Code(void *pp, unsigned coderIndex, + const Byte *src, size_t srcSize, int srcFinished, + // int finished, int blockFinished, + UInt64 *inCodePos, UInt64 *outCodePos, int *stop) +{ + CXzDecMt *me = (CXzDecMt *)pp; + CXzDecMtThread *coder = &me->coders[coderIndex]; + + *inCodePos = coder->inCodeSize; + *outCodePos = coder->outCodeSize; + *stop = True; + + if (srcSize > coder->inPreSize - coder->inCodeSize) + return SZ_ERROR_FAIL; + + if (coder->inCodeSize < coder->inPreHeaderSize) + { + size_t step = coder->inPreHeaderSize - coder->inCodeSize; + if (step > srcSize) + step = srcSize; + src += step; + srcSize -= step; + coder->inCodeSize += step; + *inCodePos = coder->inCodeSize; + if (coder->inCodeSize < coder->inPreHeaderSize) + { + *stop = False; + return SZ_OK; + } + } + + if (!coder->dec.headerParsedOk) + return SZ_OK; + if (!coder->outBuf) + return SZ_OK; + + if (coder->codeRes == SZ_OK) + { + ECoderStatus status; + SRes res; + size_t srcProcessed = srcSize; + size_t outSizeCur = coder->outPreSize - coder->dec.outDataWritten; + + // PRF(printf("\nCallback_Code: Code %d %d\n", (unsigned)srcSize, (unsigned)outSizeCur)); + + res = XzUnpacker_Code(&coder->dec, + NULL, &outSizeCur, + src, &srcProcessed, srcFinished, + // coder->finishedWithMark ? CODER_FINISH_END : CODER_FINISH_ANY, + CODER_FINISH_END, + &status); + + // PRF(printf(" res = %d, srcSize2 = %d, outSizeCur = %d", res, (unsigned)srcProcessed, (unsigned)outSizeCur)); + + coder->codeRes = res; + coder->status = status; + coder->inCodeSize += srcProcessed; + coder->outCodeSize = coder->dec.outDataWritten; + *inCodePos = coder->inCodeSize; + *outCodePos = coder->outCodeSize; + + if (res == SZ_OK) + { + if (srcProcessed == srcSize) + *stop = False; + return SZ_OK; + } + } + + if (me->props.ignoreErrors && coder->codeRes != SZ_ERROR_MEM) + { + *inCodePos = coder->inPreSize; + *outCodePos = coder->outPreSize; + return SZ_OK; + } + return coder->codeRes; +} + + +#define XZDECMT_STREAM_WRITE_STEP (1 << 24) + +static SRes XzDecMt_Callback_Write(void *pp, unsigned coderIndex, + BoolInt needWriteToStream, + const Byte *src, size_t srcSize, BoolInt isCross, + // int srcFinished, + BoolInt *needContinue, + BoolInt *canRecode) +{ + CXzDecMt *me = (CXzDecMt *)pp; + const CXzDecMtThread *coder = &me->coders[coderIndex]; + + // PRF(printf("\nWrite processed = %d srcSize = %d\n", (unsigned)me->mtc.inProcessed, (unsigned)srcSize)); + + *needContinue = False; + *canRecode = True; + + if (!needWriteToStream) + return SZ_OK; + + if (!coder->dec.headerParsedOk || !coder->outBuf) + { + if (me->finishedDecoderIndex < 0) + me->finishedDecoderIndex = (int)coderIndex; + return SZ_OK; + } + + if (me->finishedDecoderIndex >= 0) + return SZ_OK; + + me->mtc.inProcessed += coder->inCodeSize; + + *canRecode = False; + + { + SRes res; + size_t size = coder->outCodeSize; + Byte *data = coder->outBuf; + + // we use in me->dec: sha, numBlocks, indexSize + + if (!me->isBlockHeaderState_Write) + { + XzUnpacker_PrepareToRandomBlockDecoding(&me->dec); + me->dec.decodeOnlyOneBlock = False; + me->dec.numStartedStreams = coder->dec.numStartedStreams; + me->dec.streamFlags = coder->streamFlags; + + me->isBlockHeaderState_Write = True; + } + + me->dec.numTotalBlocks = coder->dec.numTotalBlocks; + XzUnpacker_UpdateIndex(&me->dec, coder->blockPackSize_for_Index, coder->outPreSize); + + if (coder->outPreSize != size) + { + if (me->props.ignoreErrors) + { + memset(data + size, 0, coder->outPreSize - size); + size = coder->outPreSize; + } + // me->numBadBlocks++; + if (me->mainErrorCode == SZ_OK) + { + if ((int)coder->status == LZMA_STATUS_NEEDS_MORE_INPUT) + me->mainErrorCode = SZ_ERROR_INPUT_EOF; + else + me->mainErrorCode = SZ_ERROR_DATA; + } + } + + if (me->writeRes != SZ_OK) + return me->writeRes; + + res = SZ_OK; + { + if (me->outSize_Defined) + { + const UInt64 rem = me->outSize - me->outProcessed; + if (size > rem) + size = (SizeT)rem; + } + + for (;;) + { + size_t cur = size; + size_t written; + if (cur > XZDECMT_STREAM_WRITE_STEP) + cur = XZDECMT_STREAM_WRITE_STEP; + + written = ISeqOutStream_Write(me->outStream, data, cur); + + // PRF(printf("\nWritten ask = %d written = %d\n", (unsigned)cur, (unsigned)written)); + + me->outProcessed += written; + if (written != cur) + { + me->writeRes = SZ_ERROR_WRITE; + res = me->writeRes; + break; + } + data += cur; + size -= cur; + // PRF_STR_INT("Written size =", size); + if (size == 0) + break; + res = MtProgress_ProgressAdd(&me->mtc.mtProgress, 0, 0); + if (res != SZ_OK) + break; + } + } + + if (coder->codeRes != SZ_OK) + if (!me->props.ignoreErrors) + { + me->finishedDecoderIndex = (int)coderIndex; + return res; + } + + RINOK(res); + + if (coder->inPreSize != coder->inCodeSize + || coder->blockPackTotal != coder->inCodeSize) + { + me->finishedDecoderIndex = (int)coderIndex; + return SZ_OK; + } + + if (coder->parseState != MTDEC_PARSE_END) + { + *needContinue = True; + return SZ_OK; + } + } + + // (coder->state == MTDEC_PARSE_END) means that there are no other working threads + // so we can use mtc variables without lock + + PRF_STR_INT("Write MTDEC_PARSE_END", me->mtc.inProcessed); + + me->mtc.mtProgress.totalInSize = me->mtc.inProcessed; + { + CXzUnpacker *dec = &me->dec; + + PRF_STR_INT("PostSingle", srcSize); + + { + size_t srcProcessed = srcSize; + ECoderStatus status; + size_t outSizeCur = 0; + SRes res; + + // dec->decodeOnlyOneBlock = False; + dec->decodeToStreamSignature = True; + + me->mainDecoderWasCalled = True; + + if (coder->parsing_Truncated) + { + me->parsing_Truncated = True; + return SZ_OK; + } + + /* + We have processed all xz-blocks of stream, + And xz unpacker is at XZ_STATE_BLOCK_HEADER state, where + (src) is a pointer to xz-Index structure. + We finish reading of current xz-Stream, including Zero padding after xz-Stream. + We exit, if we reach extra byte (first byte of new-Stream or another data). + But we don't update input stream pointer for that new extra byte. + If extra byte is not correct first byte of xz-signature, + we have SZ_ERROR_NO_ARCHIVE error here. + */ + + res = XzUnpacker_Code(dec, + NULL, &outSizeCur, + src, &srcProcessed, + me->mtc.readWasFinished, // srcFinished + CODER_FINISH_END, // CODER_FINISH_ANY, + &status); + + // res = SZ_ERROR_ARCHIVE; // for failure test + + me->status = status; + me->codeRes = res; + + if (isCross) + me->mtc.crossStart += srcProcessed; + + me->mtc.inProcessed += srcProcessed; + me->mtc.mtProgress.totalInSize = me->mtc.inProcessed; + + srcSize -= srcProcessed; + src += srcProcessed; + + if (res != SZ_OK) + { + return SZ_OK; + // return res; + } + + if (dec->state == XZ_STATE_STREAM_HEADER) + { + *needContinue = True; + me->isBlockHeaderState_Parse = False; + me->isBlockHeaderState_Write = False; + + if (!isCross) + { + Byte *crossBuf = MtDec_GetCrossBuff(&me->mtc); + if (!crossBuf) + return SZ_ERROR_MEM; + if (srcSize != 0) + memcpy(crossBuf, src, srcSize); + me->mtc.crossStart = 0; + me->mtc.crossEnd = srcSize; + } + + PRF_STR_INT("XZ_STATE_STREAM_HEADER crossEnd = ", (unsigned)me->mtc.crossEnd); + + return SZ_OK; + } + + if (status != CODER_STATUS_NEEDS_MORE_INPUT || srcSize != 0) + { + return SZ_ERROR_FAIL; + } + + if (me->mtc.readWasFinished) + { + return SZ_OK; + } + } + + { + size_t inPos; + size_t inLim; + // const Byte *inData; + UInt64 inProgressPrev = me->mtc.inProcessed; + + // XzDecMt_Prepare_InBuf_ST(p); + Byte *crossBuf = MtDec_GetCrossBuff(&me->mtc); + if (!crossBuf) + return SZ_ERROR_MEM; + + inPos = 0; + inLim = 0; + + // inData = crossBuf; + + for (;;) + { + SizeT inProcessed; + SizeT outProcessed; + ECoderStatus status; + SRes res; + + if (inPos == inLim) + { + if (!me->mtc.readWasFinished) + { + inPos = 0; + inLim = me->mtc.inBufSize; + me->mtc.readRes = ISeqInStream_Read(me->inStream, (void *)crossBuf, &inLim); + me->mtc.readProcessed += inLim; + if (inLim == 0 || me->mtc.readRes != SZ_OK) + me->mtc.readWasFinished = True; + } + } + + inProcessed = inLim - inPos; + outProcessed = 0; + + res = XzUnpacker_Code(dec, + NULL, &outProcessed, + crossBuf + inPos, &inProcessed, + (inProcessed == 0), // srcFinished + CODER_FINISH_END, &status); + + me->codeRes = res; + me->status = status; + inPos += inProcessed; + me->mtc.inProcessed += inProcessed; + me->mtc.mtProgress.totalInSize = me->mtc.inProcessed; + + if (res != SZ_OK) + { + return SZ_OK; + // return res; + } + + if (dec->state == XZ_STATE_STREAM_HEADER) + { + *needContinue = True; + me->mtc.crossStart = inPos; + me->mtc.crossEnd = inLim; + me->isBlockHeaderState_Parse = False; + me->isBlockHeaderState_Write = False; + return SZ_OK; + } + + if (status != CODER_STATUS_NEEDS_MORE_INPUT) + return SZ_ERROR_FAIL; + + if (me->mtc.progress) + { + UInt64 inDelta = me->mtc.inProcessed - inProgressPrev; + if (inDelta >= (1 << 22)) + { + RINOK(MtProgress_Progress_ST(&me->mtc.mtProgress)); + inProgressPrev = me->mtc.inProcessed; + } + } + if (me->mtc.readWasFinished) + return SZ_OK; + } + } + } +} + + +#endif + + + +void XzStatInfo_Clear(CXzStatInfo *p) +{ + p->InSize = 0; + p->OutSize = 0; + + p->NumStreams = 0; + p->NumBlocks = 0; + + p->UnpackSize_Defined = False; + + p->NumStreams_Defined = False; + p->NumBlocks_Defined = False; + + p->DataAfterEnd = False; + p->DecodingTruncated = False; + + p->DecodeRes = SZ_OK; + p->ReadRes = SZ_OK; + p->ProgressRes = SZ_OK; + + p->CombinedRes = SZ_OK; + p->CombinedRes_Type = SZ_OK; +} + + + +/* + XzDecMt_Decode_ST() can return SZ_OK or the following errors + - SZ_ERROR_MEM for memory allocation error + - error from XzUnpacker_Code() function + - SZ_ERROR_WRITE for ISeqOutStream::Write(). stat->CombinedRes_Type = SZ_ERROR_WRITE in that case + - ICompressProgress::Progress() error, stat->CombinedRes_Type = SZ_ERROR_PROGRESS. + But XzDecMt_Decode_ST() doesn't return ISeqInStream::Read() errors. + ISeqInStream::Read() result is set to p->readRes. + also it can set stat->CombinedRes_Type to SZ_ERROR_WRITE or SZ_ERROR_PROGRESS. +*/ + +static SRes XzDecMt_Decode_ST(CXzDecMt *p + #ifndef _7ZIP_ST + , BoolInt tMode + #endif + , CXzStatInfo *stat) +{ + size_t outPos; + size_t inPos, inLim; + const Byte *inData; + UInt64 inPrev, outPrev; + + CXzUnpacker *dec; + + #ifndef _7ZIP_ST + if (tMode) + { + XzDecMt_FreeOutBufs(p); + tMode = MtDec_PrepareRead(&p->mtc); + } + #endif + + if (!p->outBuf || p->outBufSize != p->props.outStep_ST) + { + ISzAlloc_Free(p->allocMid, p->outBuf); + p->outBufSize = 0; + p->outBuf = (Byte *)ISzAlloc_Alloc(p->allocMid, p->props.outStep_ST); + if (!p->outBuf) + return SZ_ERROR_MEM; + p->outBufSize = p->props.outStep_ST; + } + + if (!p->inBuf || p->inBufSize != p->props.inBufSize_ST) + { + ISzAlloc_Free(p->allocMid, p->inBuf); + p->inBufSize = 0; + p->inBuf = (Byte *)ISzAlloc_Alloc(p->allocMid, p->props.inBufSize_ST); + if (!p->inBuf) + return SZ_ERROR_MEM; + p->inBufSize = p->props.inBufSize_ST; + } + + dec = &p->dec; + dec->decodeToStreamSignature = False; + // dec->decodeOnlyOneBlock = False; + + XzUnpacker_SetOutBuf(dec, NULL, 0); + + inPrev = p->inProcessed; + outPrev = p->outProcessed; + + inPos = 0; + inLim = 0; + inData = NULL; + outPos = 0; + + for (;;) + { + SizeT outSize; + BoolInt finished; + ECoderFinishMode finishMode; + SizeT inProcessed; + ECoderStatus status; + SRes res; + + SizeT outProcessed; + + + + if (inPos == inLim) + { + #ifndef _7ZIP_ST + if (tMode) + { + inData = MtDec_Read(&p->mtc, &inLim); + inPos = 0; + if (inData) + continue; + tMode = False; + inLim = 0; + } + #endif + + if (!p->readWasFinished) + { + inPos = 0; + inLim = p->inBufSize; + inData = p->inBuf; + p->readRes = ISeqInStream_Read(p->inStream, (void *)p->inBuf, &inLim); + p->readProcessed += inLim; + if (inLim == 0 || p->readRes != SZ_OK) + p->readWasFinished = True; + } + } + + outSize = p->props.outStep_ST - outPos; + + finishMode = CODER_FINISH_ANY; + if (p->outSize_Defined) + { + const UInt64 rem = p->outSize - p->outProcessed; + if (outSize >= rem) + { + outSize = (SizeT)rem; + if (p->finishMode) + finishMode = CODER_FINISH_END; + } + } + + inProcessed = inLim - inPos; + outProcessed = outSize; + + res = XzUnpacker_Code(dec, p->outBuf + outPos, &outProcessed, + inData + inPos, &inProcessed, + (inPos == inLim), // srcFinished + finishMode, &status); + + p->codeRes = res; + p->status = status; + + inPos += inProcessed; + outPos += outProcessed; + p->inProcessed += inProcessed; + p->outProcessed += outProcessed; + + finished = ((inProcessed == 0 && outProcessed == 0) || res != SZ_OK); + + if (finished || outProcessed >= outSize) + if (outPos != 0) + { + const size_t written = ISeqOutStream_Write(p->outStream, p->outBuf, outPos); + // p->outProcessed += written; // 21.01: BUG fixed + if (written != outPos) + { + stat->CombinedRes_Type = SZ_ERROR_WRITE; + return SZ_ERROR_WRITE; + } + outPos = 0; + } + + if (p->progress && res == SZ_OK) + { + if (p->inProcessed - inPrev >= (1 << 22) || + p->outProcessed - outPrev >= (1 << 22)) + { + res = ICompressProgress_Progress(p->progress, p->inProcessed, p->outProcessed); + if (res != SZ_OK) + { + stat->CombinedRes_Type = SZ_ERROR_PROGRESS; + stat->ProgressRes = res; + return res; + } + inPrev = p->inProcessed; + outPrev = p->outProcessed; + } + } + + if (finished) + { + // p->codeRes is preliminary error from XzUnpacker_Code. + // and it can be corrected later as final result + // so we return SZ_OK here instead of (res); + return SZ_OK; + // return res; + } + } +} + + + +/* +XzStatInfo_SetStat() transforms + CXzUnpacker return code and status to combined CXzStatInfo results. + it can convert SZ_OK to SZ_ERROR_INPUT_EOF + it can convert SZ_ERROR_NO_ARCHIVE to SZ_OK and (DataAfterEnd = 1) +*/ + +static void XzStatInfo_SetStat(const CXzUnpacker *dec, + int finishMode, + // UInt64 readProcessed, + UInt64 inProcessed, + SRes res, // it's result from CXzUnpacker unpacker + ECoderStatus status, + BoolInt decodingTruncated, + CXzStatInfo *stat) +{ + UInt64 extraSize; + + stat->DecodingTruncated = (Byte)(decodingTruncated ? 1 : 0); + stat->InSize = inProcessed; + stat->NumStreams = dec->numStartedStreams; + stat->NumBlocks = dec->numTotalBlocks; + + stat->UnpackSize_Defined = True; + stat->NumStreams_Defined = True; + stat->NumBlocks_Defined = True; + + extraSize = XzUnpacker_GetExtraSize(dec); + + if (res == SZ_OK) + { + if (status == CODER_STATUS_NEEDS_MORE_INPUT) + { + // CODER_STATUS_NEEDS_MORE_INPUT is expected status for correct xz streams + // any extra data is part of correct data + extraSize = 0; + // if xz stream was not finished, then we need more data + if (!XzUnpacker_IsStreamWasFinished(dec)) + res = SZ_ERROR_INPUT_EOF; + } + else + { + // CODER_STATUS_FINISHED_WITH_MARK is not possible for multi stream xz decoding + // so he we have (status == CODER_STATUS_NOT_FINISHED) + // if (status != CODER_STATUS_FINISHED_WITH_MARK) + if (!decodingTruncated || finishMode) + res = SZ_ERROR_DATA; + } + } + else if (res == SZ_ERROR_NO_ARCHIVE) + { + /* + SZ_ERROR_NO_ARCHIVE is possible for 2 states: + XZ_STATE_STREAM_HEADER - if bad signature or bad CRC + XZ_STATE_STREAM_PADDING - if non-zero padding data + extraSize and inProcessed don't include "bad" byte + */ + // if (inProcessed == extraSize), there was no any good xz stream header, and we keep error + if (inProcessed != extraSize) // if there were good xz streams before error + { + // if (extraSize != 0 || readProcessed != inProcessed) + { + // he we suppose that all xz streams were finsihed OK, and we have + // some extra data after all streams + stat->DataAfterEnd = True; + res = SZ_OK; + } + } + } + + if (stat->DecodeRes == SZ_OK) + stat->DecodeRes = res; + + stat->InSize -= extraSize; +} + + + +SRes XzDecMt_Decode(CXzDecMtHandle pp, + const CXzDecMtProps *props, + const UInt64 *outDataSize, int finishMode, + ISeqOutStream *outStream, + // Byte *outBuf, size_t *outBufSize, + ISeqInStream *inStream, + // const Byte *inData, size_t inDataSize, + CXzStatInfo *stat, + int *isMT, + ICompressProgress *progress) +{ + CXzDecMt *p = (CXzDecMt *)pp; + #ifndef _7ZIP_ST + BoolInt tMode; + #endif + + XzStatInfo_Clear(stat); + + p->props = *props; + + p->inStream = inStream; + p->outStream = outStream; + p->progress = progress; + // p->stat = stat; + + p->outSize = 0; + p->outSize_Defined = False; + if (outDataSize) + { + p->outSize_Defined = True; + p->outSize = *outDataSize; + } + + p->finishMode = finishMode; + + // p->outSize = 457; p->outSize_Defined = True; p->finishMode = False; // for test + + p->writeRes = SZ_OK; + p->outProcessed = 0; + p->inProcessed = 0; + p->readProcessed = 0; + p->readWasFinished = False; + p->readRes = SZ_OK; + + p->codeRes = SZ_OK; + p->status = CODER_STATUS_NOT_SPECIFIED; + + XzUnpacker_Init(&p->dec); + + *isMT = False; + + /* + p->outBuf = NULL; + p->outBufSize = 0; + if (!outStream) + { + p->outBuf = outBuf; + p->outBufSize = *outBufSize; + *outBufSize = 0; + } + */ + + + #ifndef _7ZIP_ST + + p->isBlockHeaderState_Parse = False; + p->isBlockHeaderState_Write = False; + // p->numBadBlocks = 0; + p->mainErrorCode = SZ_OK; + p->mainDecoderWasCalled = False; + + tMode = False; + + if (p->props.numThreads > 1) + { + IMtDecCallback2 vt; + BoolInt needContinue; + SRes res; + // we just free ST buffers here + // but we still keep state variables, that was set in XzUnpacker_Init() + XzDecMt_FreeSt(p); + + p->outProcessed_Parse = 0; + p->parsing_Truncated = False; + + p->numStreams = 0; + p->numTotalBlocks = 0; + p->numBlocks = 0; + p->finishedDecoderIndex = -1; + + if (!p->mtc_WasConstructed) + { + p->mtc_WasConstructed = True; + MtDec_Construct(&p->mtc); + } + + p->mtc.mtCallback = &vt; + p->mtc.mtCallbackObject = p; + + p->mtc.progress = progress; + p->mtc.inStream = inStream; + p->mtc.alloc = &p->alignOffsetAlloc.vt; + // p->mtc.inData = inData; + // p->mtc.inDataSize = inDataSize; + p->mtc.inBufSize = p->props.inBufSize_MT; + // p->mtc.inBlockMax = p->props.inBlockMax; + p->mtc.numThreadsMax = p->props.numThreads; + + *isMT = True; + + vt.Parse = XzDecMt_Callback_Parse; + vt.PreCode = XzDecMt_Callback_PreCode; + vt.Code = XzDecMt_Callback_Code; + vt.Write = XzDecMt_Callback_Write; + + + res = MtDec_Code(&p->mtc); + + + stat->InSize = p->mtc.inProcessed; + + p->inProcessed = p->mtc.inProcessed; + p->readRes = p->mtc.readRes; + p->readWasFinished = p->mtc.readWasFinished; + p->readProcessed = p->mtc.readProcessed; + + tMode = True; + needContinue = False; + + if (res == SZ_OK) + { + if (p->mtc.mtProgress.res != SZ_OK) + { + res = p->mtc.mtProgress.res; + stat->ProgressRes = res; + stat->CombinedRes_Type = SZ_ERROR_PROGRESS; + } + else + needContinue = p->mtc.needContinue; + } + + if (!needContinue) + { + { + SRes codeRes; + BoolInt truncated = False; + ECoderStatus status; + const CXzUnpacker *dec; + + stat->OutSize = p->outProcessed; + + if (p->finishedDecoderIndex >= 0) + { + const CXzDecMtThread *coder = &p->coders[(unsigned)p->finishedDecoderIndex]; + codeRes = coder->codeRes; + dec = &coder->dec; + status = coder->status; + } + else if (p->mainDecoderWasCalled) + { + codeRes = p->codeRes; + dec = &p->dec; + status = p->status; + truncated = p->parsing_Truncated; + } + else + return SZ_ERROR_FAIL; + + if (p->mainErrorCode != SZ_OK) + stat->DecodeRes = p->mainErrorCode; + + XzStatInfo_SetStat(dec, p->finishMode, + // p->mtc.readProcessed, + p->mtc.inProcessed, + codeRes, status, + truncated, + stat); + } + + if (res == SZ_OK) + { + stat->ReadRes = p->mtc.readRes; + + if (p->writeRes != SZ_OK) + { + res = p->writeRes; + stat->CombinedRes_Type = SZ_ERROR_WRITE; + } + else if (p->mtc.readRes != SZ_OK + // && p->mtc.inProcessed == p->mtc.readProcessed + && stat->DecodeRes == SZ_ERROR_INPUT_EOF) + { + res = p->mtc.readRes; + stat->CombinedRes_Type = SZ_ERROR_READ; + } + else if (stat->DecodeRes != SZ_OK) + res = stat->DecodeRes; + } + + stat->CombinedRes = res; + if (stat->CombinedRes_Type == SZ_OK) + stat->CombinedRes_Type = res; + return res; + } + + PRF_STR("----- decoding ST -----"); + } + + #endif + + + *isMT = False; + + { + SRes res = XzDecMt_Decode_ST(p + #ifndef _7ZIP_ST + , tMode + #endif + , stat + ); + + #ifndef _7ZIP_ST + // we must set error code from MT decoding at first + if (p->mainErrorCode != SZ_OK) + stat->DecodeRes = p->mainErrorCode; + #endif + + XzStatInfo_SetStat(&p->dec, + p->finishMode, + // p->readProcessed, + p->inProcessed, + p->codeRes, p->status, + False, // truncated + stat); + + stat->ReadRes = p->readRes; + + if (res == SZ_OK) + { + if (p->readRes != SZ_OK + // && p->inProcessed == p->readProcessed + && stat->DecodeRes == SZ_ERROR_INPUT_EOF) + { + // we set read error as combined error, only if that error was the reason + // of decoding problem + res = p->readRes; + stat->CombinedRes_Type = SZ_ERROR_READ; + } + else if (stat->DecodeRes != SZ_OK) + res = stat->DecodeRes; + } + + stat->CombinedRes = res; + if (stat->CombinedRes_Type == SZ_OK) + stat->CombinedRes_Type = res; + return res; + } +} diff --git a/3rdparty/lzma/src/XzEnc.c b/3rdparty/lzma/src/XzEnc.c new file mode 100644 index 0000000000..be174cccf5 --- /dev/null +++ b/3rdparty/lzma/src/XzEnc.c @@ -0,0 +1,1330 @@ +/* XzEnc.c -- Xz Encode +2021-04-01 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include +#include + +#include "7zCrc.h" +#include "Bra.h" +#include "CpuArch.h" + +#ifdef USE_SUBBLOCK +#include "Bcj3Enc.c" +#include "SbFind.c" +#include "SbEnc.c" +#endif + +#include "XzEnc.h" + +// #define _7ZIP_ST + +#ifndef _7ZIP_ST +#include "MtCoder.h" +#else +#define MTCODER__THREADS_MAX 1 +#define MTCODER__BLOCKS_MAX 1 +#endif + +#define XZ_GET_PAD_SIZE(dataSize) ((4 - ((unsigned)(dataSize) & 3)) & 3) + +/* max pack size for LZMA2 block + check-64bytrs: */ +#define XZ_GET_MAX_BLOCK_PACK_SIZE(unpackSize) ((unpackSize) + ((unpackSize) >> 10) + 16 + 64) + +#define XZ_GET_ESTIMATED_BLOCK_TOTAL_PACK_SIZE(unpackSize) (XZ_BLOCK_HEADER_SIZE_MAX + XZ_GET_MAX_BLOCK_PACK_SIZE(unpackSize)) + + +#define XzBlock_ClearFlags(p) (p)->flags = 0; +#define XzBlock_SetNumFilters(p, n) (p)->flags = (Byte)((p)->flags | ((n) - 1)); +#define XzBlock_SetHasPackSize(p) (p)->flags |= XZ_BF_PACK_SIZE; +#define XzBlock_SetHasUnpackSize(p) (p)->flags |= XZ_BF_UNPACK_SIZE; + + +static SRes WriteBytes(ISeqOutStream *s, const void *buf, size_t size) +{ + return (ISeqOutStream_Write(s, buf, size) == size) ? SZ_OK : SZ_ERROR_WRITE; +} + +static SRes WriteBytesUpdateCrc(ISeqOutStream *s, const void *buf, size_t size, UInt32 *crc) +{ + *crc = CrcUpdate(*crc, buf, size); + return WriteBytes(s, buf, size); +} + + +static SRes Xz_WriteHeader(CXzStreamFlags f, ISeqOutStream *s) +{ + UInt32 crc; + Byte header[XZ_STREAM_HEADER_SIZE]; + memcpy(header, XZ_SIG, XZ_SIG_SIZE); + header[XZ_SIG_SIZE] = (Byte)(f >> 8); + header[XZ_SIG_SIZE + 1] = (Byte)(f & 0xFF); + crc = CrcCalc(header + XZ_SIG_SIZE, XZ_STREAM_FLAGS_SIZE); + SetUi32(header + XZ_SIG_SIZE + XZ_STREAM_FLAGS_SIZE, crc); + return WriteBytes(s, header, XZ_STREAM_HEADER_SIZE); +} + + +static SRes XzBlock_WriteHeader(const CXzBlock *p, ISeqOutStream *s) +{ + Byte header[XZ_BLOCK_HEADER_SIZE_MAX]; + + unsigned pos = 1; + unsigned numFilters, i; + header[pos++] = p->flags; + + if (XzBlock_HasPackSize(p)) pos += Xz_WriteVarInt(header + pos, p->packSize); + if (XzBlock_HasUnpackSize(p)) pos += Xz_WriteVarInt(header + pos, p->unpackSize); + numFilters = XzBlock_GetNumFilters(p); + + for (i = 0; i < numFilters; i++) + { + const CXzFilter *f = &p->filters[i]; + pos += Xz_WriteVarInt(header + pos, f->id); + pos += Xz_WriteVarInt(header + pos, f->propsSize); + memcpy(header + pos, f->props, f->propsSize); + pos += f->propsSize; + } + + while ((pos & 3) != 0) + header[pos++] = 0; + + header[0] = (Byte)(pos >> 2); + SetUi32(header + pos, CrcCalc(header, pos)); + return WriteBytes(s, header, pos + 4); +} + + + + +typedef struct +{ + size_t numBlocks; + size_t size; + size_t allocated; + Byte *blocks; +} CXzEncIndex; + + +static void XzEncIndex_Construct(CXzEncIndex *p) +{ + p->numBlocks = 0; + p->size = 0; + p->allocated = 0; + p->blocks = NULL; +} + +static void XzEncIndex_Init(CXzEncIndex *p) +{ + p->numBlocks = 0; + p->size = 0; +} + +static void XzEncIndex_Free(CXzEncIndex *p, ISzAllocPtr alloc) +{ + if (p->blocks) + { + ISzAlloc_Free(alloc, p->blocks); + p->blocks = NULL; + } + p->numBlocks = 0; + p->size = 0; + p->allocated = 0; +} + + +static SRes XzEncIndex_ReAlloc(CXzEncIndex *p, size_t newSize, ISzAllocPtr alloc) +{ + Byte *blocks = (Byte *)ISzAlloc_Alloc(alloc, newSize); + if (!blocks) + return SZ_ERROR_MEM; + if (p->size != 0) + memcpy(blocks, p->blocks, p->size); + if (p->blocks) + ISzAlloc_Free(alloc, p->blocks); + p->blocks = blocks; + p->allocated = newSize; + return SZ_OK; +} + + +static SRes XzEncIndex_PreAlloc(CXzEncIndex *p, UInt64 numBlocks, UInt64 unpackSize, UInt64 totalSize, ISzAllocPtr alloc) +{ + UInt64 pos; + { + Byte buf[32]; + unsigned pos2 = Xz_WriteVarInt(buf, totalSize); + pos2 += Xz_WriteVarInt(buf + pos2, unpackSize); + pos = numBlocks * pos2; + } + + if (pos <= p->allocated - p->size) + return SZ_OK; + { + UInt64 newSize64 = p->size + pos; + size_t newSize = (size_t)newSize64; + if (newSize != newSize64) + return SZ_ERROR_MEM; + return XzEncIndex_ReAlloc(p, newSize, alloc); + } +} + + +static SRes XzEncIndex_AddIndexRecord(CXzEncIndex *p, UInt64 unpackSize, UInt64 totalSize, ISzAllocPtr alloc) +{ + Byte buf[32]; + unsigned pos = Xz_WriteVarInt(buf, totalSize); + pos += Xz_WriteVarInt(buf + pos, unpackSize); + + if (pos > p->allocated - p->size) + { + size_t newSize = p->allocated * 2 + 16 * 2; + if (newSize < p->size + pos) + return SZ_ERROR_MEM; + RINOK(XzEncIndex_ReAlloc(p, newSize, alloc)); + } + memcpy(p->blocks + p->size, buf, pos); + p->size += pos; + p->numBlocks++; + return SZ_OK; +} + + +static SRes XzEncIndex_WriteFooter(const CXzEncIndex *p, CXzStreamFlags flags, ISeqOutStream *s) +{ + Byte buf[32]; + UInt64 globalPos; + UInt32 crc = CRC_INIT_VAL; + unsigned pos = 1 + Xz_WriteVarInt(buf + 1, p->numBlocks); + + globalPos = pos; + buf[0] = 0; + RINOK(WriteBytesUpdateCrc(s, buf, pos, &crc)); + RINOK(WriteBytesUpdateCrc(s, p->blocks, p->size, &crc)); + globalPos += p->size; + + pos = XZ_GET_PAD_SIZE(globalPos); + buf[1] = 0; + buf[2] = 0; + buf[3] = 0; + globalPos += pos; + + crc = CrcUpdate(crc, buf + 4 - pos, pos); + SetUi32(buf + 4, CRC_GET_DIGEST(crc)); + + SetUi32(buf + 8 + 4, (UInt32)(globalPos >> 2)); + buf[8 + 8] = (Byte)(flags >> 8); + buf[8 + 9] = (Byte)(flags & 0xFF); + SetUi32(buf + 8, CrcCalc(buf + 8 + 4, 6)); + buf[8 + 10] = XZ_FOOTER_SIG_0; + buf[8 + 11] = XZ_FOOTER_SIG_1; + + return WriteBytes(s, buf + 4 - pos, pos + 4 + 12); +} + + + +/* ---------- CSeqCheckInStream ---------- */ + +typedef struct +{ + ISeqInStream vt; + ISeqInStream *realStream; + const Byte *data; + UInt64 limit; + UInt64 processed; + int realStreamFinished; + CXzCheck check; +} CSeqCheckInStream; + +static void SeqCheckInStream_Init(CSeqCheckInStream *p, unsigned checkMode) +{ + p->limit = (UInt64)(Int64)-1; + p->processed = 0; + p->realStreamFinished = 0; + XzCheck_Init(&p->check, checkMode); +} + +static void SeqCheckInStream_GetDigest(CSeqCheckInStream *p, Byte *digest) +{ + XzCheck_Final(&p->check, digest); +} + +static SRes SeqCheckInStream_Read(const ISeqInStream *pp, void *data, size_t *size) +{ + CSeqCheckInStream *p = CONTAINER_FROM_VTBL(pp, CSeqCheckInStream, vt); + size_t size2 = *size; + SRes res = SZ_OK; + + if (p->limit != (UInt64)(Int64)-1) + { + UInt64 rem = p->limit - p->processed; + if (size2 > rem) + size2 = (size_t)rem; + } + if (size2 != 0) + { + if (p->realStream) + { + res = ISeqInStream_Read(p->realStream, data, &size2); + p->realStreamFinished = (size2 == 0) ? 1 : 0; + } + else + memcpy(data, p->data + (size_t)p->processed, size2); + XzCheck_Update(&p->check, data, size2); + p->processed += size2; + } + *size = size2; + return res; +} + + +/* ---------- CSeqSizeOutStream ---------- */ + +typedef struct +{ + ISeqOutStream vt; + ISeqOutStream *realStream; + Byte *outBuf; + size_t outBufLimit; + UInt64 processed; +} CSeqSizeOutStream; + +static size_t SeqSizeOutStream_Write(const ISeqOutStream *pp, const void *data, size_t size) +{ + CSeqSizeOutStream *p = CONTAINER_FROM_VTBL(pp, CSeqSizeOutStream, vt); + if (p->realStream) + size = ISeqOutStream_Write(p->realStream, data, size); + else + { + if (size > p->outBufLimit - (size_t)p->processed) + return 0; + memcpy(p->outBuf + (size_t)p->processed, data, size); + } + p->processed += size; + return size; +} + + +/* ---------- CSeqInFilter ---------- */ + +#define FILTER_BUF_SIZE (1 << 20) + +typedef struct +{ + ISeqInStream p; + ISeqInStream *realStream; + IStateCoder StateCoder; + Byte *buf; + size_t curPos; + size_t endPos; + int srcWasFinished; +} CSeqInFilter; + + +SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAllocPtr alloc); + +static SRes SeqInFilter_Init(CSeqInFilter *p, const CXzFilter *props, ISzAllocPtr alloc) +{ + if (!p->buf) + { + p->buf = (Byte *)ISzAlloc_Alloc(alloc, FILTER_BUF_SIZE); + if (!p->buf) + return SZ_ERROR_MEM; + } + p->curPos = p->endPos = 0; + p->srcWasFinished = 0; + RINOK(BraState_SetFromMethod(&p->StateCoder, props->id, 1, alloc)); + RINOK(p->StateCoder.SetProps(p->StateCoder.p, props->props, props->propsSize, alloc)); + p->StateCoder.Init(p->StateCoder.p); + return SZ_OK; +} + + +static SRes SeqInFilter_Read(const ISeqInStream *pp, void *data, size_t *size) +{ + CSeqInFilter *p = CONTAINER_FROM_VTBL(pp, CSeqInFilter, p); + size_t sizeOriginal = *size; + if (sizeOriginal == 0) + return SZ_OK; + *size = 0; + + for (;;) + { + if (!p->srcWasFinished && p->curPos == p->endPos) + { + p->curPos = 0; + p->endPos = FILTER_BUF_SIZE; + RINOK(ISeqInStream_Read(p->realStream, p->buf, &p->endPos)); + if (p->endPos == 0) + p->srcWasFinished = 1; + } + { + SizeT srcLen = p->endPos - p->curPos; + ECoderStatus status; + SRes res; + *size = sizeOriginal; + res = p->StateCoder.Code2(p->StateCoder.p, + (Byte *)data, size, + p->buf + p->curPos, &srcLen, + p->srcWasFinished, CODER_FINISH_ANY, + &status); + p->curPos += srcLen; + if (*size != 0 || srcLen == 0 || res != SZ_OK) + return res; + } + } +} + +static void SeqInFilter_Construct(CSeqInFilter *p) +{ + p->buf = NULL; + p->StateCoder.p = NULL; + p->p.Read = SeqInFilter_Read; +} + +static void SeqInFilter_Free(CSeqInFilter *p, ISzAllocPtr alloc) +{ + if (p->StateCoder.p) + { + p->StateCoder.Free(p->StateCoder.p, alloc); + p->StateCoder.p = NULL; + } + if (p->buf) + { + ISzAlloc_Free(alloc, p->buf); + p->buf = NULL; + } +} + + +/* ---------- CSbEncInStream ---------- */ + +#ifdef USE_SUBBLOCK + +typedef struct +{ + ISeqInStream vt; + ISeqInStream *inStream; + CSbEnc enc; +} CSbEncInStream; + +static SRes SbEncInStream_Read(const ISeqInStream *pp, void *data, size_t *size) +{ + CSbEncInStream *p = CONTAINER_FROM_VTBL(pp, CSbEncInStream, vt); + size_t sizeOriginal = *size; + if (sizeOriginal == 0) + return SZ_OK; + + for (;;) + { + if (p->enc.needRead && !p->enc.readWasFinished) + { + size_t processed = p->enc.needReadSizeMax; + RINOK(p->inStream->Read(p->inStream, p->enc.buf + p->enc.readPos, &processed)); + p->enc.readPos += processed; + if (processed == 0) + { + p->enc.readWasFinished = True; + p->enc.isFinalFinished = True; + } + p->enc.needRead = False; + } + + *size = sizeOriginal; + RINOK(SbEnc_Read(&p->enc, data, size)); + if (*size != 0 || !p->enc.needRead) + return SZ_OK; + } +} + +void SbEncInStream_Construct(CSbEncInStream *p, ISzAllocPtr alloc) +{ + SbEnc_Construct(&p->enc, alloc); + p->vt.Read = SbEncInStream_Read; +} + +SRes SbEncInStream_Init(CSbEncInStream *p) +{ + return SbEnc_Init(&p->enc); +} + +void SbEncInStream_Free(CSbEncInStream *p) +{ + SbEnc_Free(&p->enc); +} + +#endif + + + +/* ---------- CXzProps ---------- */ + + +void XzFilterProps_Init(CXzFilterProps *p) +{ + p->id = 0; + p->delta = 0; + p->ip = 0; + p->ipDefined = False; +} + +void XzProps_Init(CXzProps *p) +{ + p->checkId = XZ_CHECK_CRC32; + p->blockSize = XZ_PROPS__BLOCK_SIZE__AUTO; + p->numBlockThreads_Reduced = -1; + p->numBlockThreads_Max = -1; + p->numTotalThreads = -1; + p->reduceSize = (UInt64)(Int64)-1; + p->forceWriteSizesInHeader = 0; + // p->forceWriteSizesInHeader = 1; + + XzFilterProps_Init(&p->filterProps); + Lzma2EncProps_Init(&p->lzma2Props); +} + + +static void XzEncProps_Normalize_Fixed(CXzProps *p) +{ + UInt64 fileSize; + int t1, t1n, t2, t2r, t3; + { + CLzma2EncProps tp = p->lzma2Props; + if (tp.numTotalThreads <= 0) + tp.numTotalThreads = p->numTotalThreads; + Lzma2EncProps_Normalize(&tp); + t1n = tp.numTotalThreads; + } + + t1 = p->lzma2Props.numTotalThreads; + t2 = p->numBlockThreads_Max; + t3 = p->numTotalThreads; + + if (t2 > MTCODER__THREADS_MAX) + t2 = MTCODER__THREADS_MAX; + + if (t3 <= 0) + { + if (t2 <= 0) + t2 = 1; + t3 = t1n * t2; + } + else if (t2 <= 0) + { + t2 = t3 / t1n; + if (t2 == 0) + { + t1 = 1; + t2 = t3; + } + if (t2 > MTCODER__THREADS_MAX) + t2 = MTCODER__THREADS_MAX; + } + else if (t1 <= 0) + { + t1 = t3 / t2; + if (t1 == 0) + t1 = 1; + } + else + t3 = t1n * t2; + + p->lzma2Props.numTotalThreads = t1; + + t2r = t2; + + fileSize = p->reduceSize; + + if ((p->blockSize < fileSize || fileSize == (UInt64)(Int64)-1)) + p->lzma2Props.lzmaProps.reduceSize = p->blockSize; + + Lzma2EncProps_Normalize(&p->lzma2Props); + + t1 = p->lzma2Props.numTotalThreads; + + { + if (t2 > 1 && fileSize != (UInt64)(Int64)-1) + { + UInt64 numBlocks = fileSize / p->blockSize; + if (numBlocks * p->blockSize != fileSize) + numBlocks++; + if (numBlocks < (unsigned)t2) + { + t2r = (int)numBlocks; + if (t2r == 0) + t2r = 1; + t3 = t1 * t2r; + } + } + } + + p->numBlockThreads_Max = t2; + p->numBlockThreads_Reduced = t2r; + p->numTotalThreads = t3; +} + + +static void XzProps_Normalize(CXzProps *p) +{ + /* we normalize xzProps properties, but we normalize only some of CXzProps::lzma2Props properties. + Lzma2Enc_SetProps() will normalize lzma2Props later. */ + + if (p->blockSize == XZ_PROPS__BLOCK_SIZE__SOLID) + { + p->lzma2Props.lzmaProps.reduceSize = p->reduceSize; + p->numBlockThreads_Reduced = 1; + p->numBlockThreads_Max = 1; + if (p->lzma2Props.numTotalThreads <= 0) + p->lzma2Props.numTotalThreads = p->numTotalThreads; + return; + } + else + { + CLzma2EncProps *lzma2 = &p->lzma2Props; + if (p->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) + { + // xz-auto + p->lzma2Props.lzmaProps.reduceSize = p->reduceSize; + + if (lzma2->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) + { + // if (xz-auto && lzma2-solid) - we use solid for both + p->blockSize = XZ_PROPS__BLOCK_SIZE__SOLID; + p->numBlockThreads_Reduced = 1; + p->numBlockThreads_Max = 1; + if (p->lzma2Props.numTotalThreads <= 0) + p->lzma2Props.numTotalThreads = p->numTotalThreads; + } + else + { + // if (xz-auto && (lzma2-auto || lzma2-fixed_) + // we calculate block size for lzma2 and use that block size for xz, lzma2 uses single-chunk per block + CLzma2EncProps tp = p->lzma2Props; + if (tp.numTotalThreads <= 0) + tp.numTotalThreads = p->numTotalThreads; + + Lzma2EncProps_Normalize(&tp); + + p->blockSize = tp.blockSize; // fixed or solid + p->numBlockThreads_Reduced = tp.numBlockThreads_Reduced; + p->numBlockThreads_Max = tp.numBlockThreads_Max; + if (lzma2->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) + lzma2->blockSize = tp.blockSize; // fixed or solid, LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID + if (lzma2->lzmaProps.reduceSize > tp.blockSize && tp.blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) + lzma2->lzmaProps.reduceSize = tp.blockSize; + lzma2->numBlockThreads_Reduced = 1; + lzma2->numBlockThreads_Max = 1; + return; + } + } + else + { + // xz-fixed + // we can use xz::reduceSize or xz::blockSize as base for lzmaProps::reduceSize + + p->lzma2Props.lzmaProps.reduceSize = p->reduceSize; + { + UInt64 r = p->reduceSize; + if (r > p->blockSize || r == (UInt64)(Int64)-1) + r = p->blockSize; + lzma2->lzmaProps.reduceSize = r; + } + if (lzma2->blockSize == LZMA2_ENC_PROPS__BLOCK_SIZE__AUTO) + lzma2->blockSize = LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID; + else if (lzma2->blockSize > p->blockSize && lzma2->blockSize != LZMA2_ENC_PROPS__BLOCK_SIZE__SOLID) + lzma2->blockSize = p->blockSize; + + XzEncProps_Normalize_Fixed(p); + } + } +} + + +/* ---------- CLzma2WithFilters ---------- */ + +typedef struct +{ + CLzma2EncHandle lzma2; + CSeqInFilter filter; + + #ifdef USE_SUBBLOCK + CSbEncInStream sb; + #endif +} CLzma2WithFilters; + + +static void Lzma2WithFilters_Construct(CLzma2WithFilters *p) +{ + p->lzma2 = NULL; + SeqInFilter_Construct(&p->filter); + + #ifdef USE_SUBBLOCK + SbEncInStream_Construct(&p->sb, alloc); + #endif +} + + +static SRes Lzma2WithFilters_Create(CLzma2WithFilters *p, ISzAllocPtr alloc, ISzAllocPtr bigAlloc) +{ + if (!p->lzma2) + { + p->lzma2 = Lzma2Enc_Create(alloc, bigAlloc); + if (!p->lzma2) + return SZ_ERROR_MEM; + } + return SZ_OK; +} + + +static void Lzma2WithFilters_Free(CLzma2WithFilters *p, ISzAllocPtr alloc) +{ + #ifdef USE_SUBBLOCK + SbEncInStream_Free(&p->sb); + #endif + + SeqInFilter_Free(&p->filter, alloc); + if (p->lzma2) + { + Lzma2Enc_Destroy(p->lzma2); + p->lzma2 = NULL; + } +} + + +typedef struct +{ + UInt64 unpackSize; + UInt64 totalSize; + size_t headerSize; +} CXzEncBlockInfo; + + +static SRes Xz_CompressBlock( + CLzma2WithFilters *lzmaf, + + ISeqOutStream *outStream, + Byte *outBufHeader, + Byte *outBufData, size_t outBufDataLimit, + + ISeqInStream *inStream, + // UInt64 expectedSize, + const Byte *inBuf, // used if (!inStream) + size_t inBufSize, // used if (!inStream), it's block size, props->blockSize is ignored + + const CXzProps *props, + ICompressProgress *progress, + int *inStreamFinished, /* only for inStream version */ + CXzEncBlockInfo *blockSizes, + ISzAllocPtr alloc, + ISzAllocPtr allocBig) +{ + CSeqCheckInStream checkInStream; + CSeqSizeOutStream seqSizeOutStream; + CXzBlock block; + unsigned filterIndex = 0; + CXzFilter *filter = NULL; + const CXzFilterProps *fp = &props->filterProps; + if (fp->id == 0) + fp = NULL; + + *inStreamFinished = False; + + RINOK(Lzma2WithFilters_Create(lzmaf, alloc, allocBig)); + + RINOK(Lzma2Enc_SetProps(lzmaf->lzma2, &props->lzma2Props)); + + XzBlock_ClearFlags(&block); + XzBlock_SetNumFilters(&block, 1 + (fp ? 1 : 0)); + + if (fp) + { + filter = &block.filters[filterIndex++]; + filter->id = fp->id; + filter->propsSize = 0; + + if (fp->id == XZ_ID_Delta) + { + filter->props[0] = (Byte)(fp->delta - 1); + filter->propsSize = 1; + } + else if (fp->ipDefined) + { + Byte *ptr = filter->props; + SetUi32(ptr, fp->ip); + filter->propsSize = 4; + } + } + + { + CXzFilter *f = &block.filters[filterIndex++]; + f->id = XZ_ID_LZMA2; + f->propsSize = 1; + f->props[0] = Lzma2Enc_WriteProperties(lzmaf->lzma2); + } + + seqSizeOutStream.vt.Write = SeqSizeOutStream_Write; + seqSizeOutStream.realStream = outStream; + seqSizeOutStream.outBuf = outBufData; + seqSizeOutStream.outBufLimit = outBufDataLimit; + seqSizeOutStream.processed = 0; + + /* + if (expectedSize != (UInt64)(Int64)-1) + { + block.unpackSize = expectedSize; + if (props->blockSize != (UInt64)(Int64)-1) + if (expectedSize > props->blockSize) + block.unpackSize = props->blockSize; + XzBlock_SetHasUnpackSize(&block); + } + */ + + if (outStream) + { + RINOK(XzBlock_WriteHeader(&block, &seqSizeOutStream.vt)); + } + + checkInStream.vt.Read = SeqCheckInStream_Read; + SeqCheckInStream_Init(&checkInStream, props->checkId); + + checkInStream.realStream = inStream; + checkInStream.data = inBuf; + checkInStream.limit = props->blockSize; + if (!inStream) + checkInStream.limit = inBufSize; + + if (fp) + { + #ifdef USE_SUBBLOCK + if (fp->id == XZ_ID_Subblock) + { + lzmaf->sb.inStream = &checkInStream.vt; + RINOK(SbEncInStream_Init(&lzmaf->sb)); + } + else + #endif + { + lzmaf->filter.realStream = &checkInStream.vt; + RINOK(SeqInFilter_Init(&lzmaf->filter, filter, alloc)); + } + } + + { + SRes res; + Byte *outBuf = NULL; + size_t outSize = 0; + BoolInt useStream = (fp || inStream); + // useStream = True; + + if (!useStream) + { + XzCheck_Update(&checkInStream.check, inBuf, inBufSize); + checkInStream.processed = inBufSize; + } + + if (!outStream) + { + outBuf = seqSizeOutStream.outBuf; // + (size_t)seqSizeOutStream.processed; + outSize = seqSizeOutStream.outBufLimit; // - (size_t)seqSizeOutStream.processed; + } + + res = Lzma2Enc_Encode2(lzmaf->lzma2, + outBuf ? NULL : &seqSizeOutStream.vt, + outBuf, + outBuf ? &outSize : NULL, + + useStream ? + (fp ? + ( + #ifdef USE_SUBBLOCK + (fp->id == XZ_ID_Subblock) ? &lzmaf->sb.vt: + #endif + &lzmaf->filter.p) : + &checkInStream.vt) : NULL, + + useStream ? NULL : inBuf, + useStream ? 0 : inBufSize, + + progress); + + if (outBuf) + seqSizeOutStream.processed += outSize; + + RINOK(res); + blockSizes->unpackSize = checkInStream.processed; + } + { + Byte buf[4 + 64]; + unsigned padSize = XZ_GET_PAD_SIZE(seqSizeOutStream.processed); + UInt64 packSize = seqSizeOutStream.processed; + + buf[0] = 0; + buf[1] = 0; + buf[2] = 0; + buf[3] = 0; + + SeqCheckInStream_GetDigest(&checkInStream, buf + 4); + RINOK(WriteBytes(&seqSizeOutStream.vt, buf + (4 - padSize), padSize + XzFlags_GetCheckSize((CXzStreamFlags)props->checkId))); + + blockSizes->totalSize = seqSizeOutStream.processed - padSize; + + if (!outStream) + { + seqSizeOutStream.outBuf = outBufHeader; + seqSizeOutStream.outBufLimit = XZ_BLOCK_HEADER_SIZE_MAX; + seqSizeOutStream.processed = 0; + + block.unpackSize = blockSizes->unpackSize; + XzBlock_SetHasUnpackSize(&block); + + block.packSize = packSize; + XzBlock_SetHasPackSize(&block); + + RINOK(XzBlock_WriteHeader(&block, &seqSizeOutStream.vt)); + + blockSizes->headerSize = (size_t)seqSizeOutStream.processed; + blockSizes->totalSize += seqSizeOutStream.processed; + } + } + + if (inStream) + *inStreamFinished = checkInStream.realStreamFinished; + else + { + *inStreamFinished = False; + if (checkInStream.processed != inBufSize) + return SZ_ERROR_FAIL; + } + + return SZ_OK; +} + + + +typedef struct +{ + ICompressProgress vt; + ICompressProgress *progress; + UInt64 inOffset; + UInt64 outOffset; +} CCompressProgress_XzEncOffset; + + +static SRes CompressProgress_XzEncOffset_Progress(const ICompressProgress *pp, UInt64 inSize, UInt64 outSize) +{ + const CCompressProgress_XzEncOffset *p = CONTAINER_FROM_VTBL(pp, CCompressProgress_XzEncOffset, vt); + inSize += p->inOffset; + outSize += p->outOffset; + return ICompressProgress_Progress(p->progress, inSize, outSize); +} + + + + +typedef struct +{ + ISzAllocPtr alloc; + ISzAllocPtr allocBig; + + CXzProps xzProps; + UInt64 expectedDataSize; + + CXzEncIndex xzIndex; + + CLzma2WithFilters lzmaf_Items[MTCODER__THREADS_MAX]; + + size_t outBufSize; /* size of allocated outBufs[i] */ + Byte *outBufs[MTCODER__BLOCKS_MAX]; + + #ifndef _7ZIP_ST + unsigned checkType; + ISeqOutStream *outStream; + BoolInt mtCoder_WasConstructed; + CMtCoder mtCoder; + CXzEncBlockInfo EncBlocks[MTCODER__BLOCKS_MAX]; + #endif + +} CXzEnc; + + +static void XzEnc_Construct(CXzEnc *p) +{ + unsigned i; + + XzEncIndex_Construct(&p->xzIndex); + + for (i = 0; i < MTCODER__THREADS_MAX; i++) + Lzma2WithFilters_Construct(&p->lzmaf_Items[i]); + + #ifndef _7ZIP_ST + p->mtCoder_WasConstructed = False; + { + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + p->outBufs[i] = NULL; + p->outBufSize = 0; + } + #endif +} + + +static void XzEnc_FreeOutBufs(CXzEnc *p) +{ + unsigned i; + for (i = 0; i < MTCODER__BLOCKS_MAX; i++) + if (p->outBufs[i]) + { + ISzAlloc_Free(p->alloc, p->outBufs[i]); + p->outBufs[i] = NULL; + } + p->outBufSize = 0; +} + + +static void XzEnc_Free(CXzEnc *p, ISzAllocPtr alloc) +{ + unsigned i; + + XzEncIndex_Free(&p->xzIndex, alloc); + + for (i = 0; i < MTCODER__THREADS_MAX; i++) + Lzma2WithFilters_Free(&p->lzmaf_Items[i], alloc); + + #ifndef _7ZIP_ST + if (p->mtCoder_WasConstructed) + { + MtCoder_Destruct(&p->mtCoder); + p->mtCoder_WasConstructed = False; + } + XzEnc_FreeOutBufs(p); + #endif +} + + +CXzEncHandle XzEnc_Create(ISzAllocPtr alloc, ISzAllocPtr allocBig) +{ + CXzEnc *p = (CXzEnc *)ISzAlloc_Alloc(alloc, sizeof(CXzEnc)); + if (!p) + return NULL; + XzEnc_Construct(p); + XzProps_Init(&p->xzProps); + XzProps_Normalize(&p->xzProps); + p->expectedDataSize = (UInt64)(Int64)-1; + p->alloc = alloc; + p->allocBig = allocBig; + return p; +} + + +void XzEnc_Destroy(CXzEncHandle pp) +{ + CXzEnc *p = (CXzEnc *)pp; + XzEnc_Free(p, p->alloc); + ISzAlloc_Free(p->alloc, p); +} + + +SRes XzEnc_SetProps(CXzEncHandle pp, const CXzProps *props) +{ + CXzEnc *p = (CXzEnc *)pp; + p->xzProps = *props; + XzProps_Normalize(&p->xzProps); + return SZ_OK; +} + + +void XzEnc_SetDataSize(CXzEncHandle pp, UInt64 expectedDataSiize) +{ + CXzEnc *p = (CXzEnc *)pp; + p->expectedDataSize = expectedDataSiize; +} + + + + +#ifndef _7ZIP_ST + +static SRes XzEnc_MtCallback_Code(void *pp, unsigned coderIndex, unsigned outBufIndex, + const Byte *src, size_t srcSize, int finished) +{ + CXzEnc *me = (CXzEnc *)pp; + SRes res; + CMtProgressThunk progressThunk; + + Byte *dest = me->outBufs[outBufIndex]; + + UNUSED_VAR(finished) + + { + CXzEncBlockInfo *bInfo = &me->EncBlocks[outBufIndex]; + bInfo->totalSize = 0; + bInfo->unpackSize = 0; + bInfo->headerSize = 0; + } + + if (!dest) + { + dest = (Byte *)ISzAlloc_Alloc(me->alloc, me->outBufSize); + if (!dest) + return SZ_ERROR_MEM; + me->outBufs[outBufIndex] = dest; + } + + MtProgressThunk_CreateVTable(&progressThunk); + progressThunk.mtProgress = &me->mtCoder.mtProgress; + MtProgressThunk_Init(&progressThunk); + + { + CXzEncBlockInfo blockSizes; + int inStreamFinished; + + res = Xz_CompressBlock( + &me->lzmaf_Items[coderIndex], + + NULL, + dest, + dest + XZ_BLOCK_HEADER_SIZE_MAX, me->outBufSize - XZ_BLOCK_HEADER_SIZE_MAX, + + NULL, + // srcSize, // expectedSize + src, srcSize, + + &me->xzProps, + &progressThunk.vt, + &inStreamFinished, + &blockSizes, + me->alloc, + me->allocBig); + + if (res == SZ_OK) + me->EncBlocks[outBufIndex] = blockSizes; + + return res; + } +} + + +static SRes XzEnc_MtCallback_Write(void *pp, unsigned outBufIndex) +{ + CXzEnc *me = (CXzEnc *)pp; + + const CXzEncBlockInfo *bInfo = &me->EncBlocks[outBufIndex]; + const Byte *data = me->outBufs[outBufIndex]; + + RINOK(WriteBytes(me->outStream, data, bInfo->headerSize)); + + { + UInt64 totalPackFull = bInfo->totalSize + XZ_GET_PAD_SIZE(bInfo->totalSize); + RINOK(WriteBytes(me->outStream, data + XZ_BLOCK_HEADER_SIZE_MAX, (size_t)totalPackFull - bInfo->headerSize)); + } + + return XzEncIndex_AddIndexRecord(&me->xzIndex, bInfo->unpackSize, bInfo->totalSize, me->alloc); +} + +#endif + + + +SRes XzEnc_Encode(CXzEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress) +{ + CXzEnc *p = (CXzEnc *)pp; + + const CXzProps *props = &p->xzProps; + + XzEncIndex_Init(&p->xzIndex); + { + UInt64 numBlocks = 1; + UInt64 blockSize = props->blockSize; + + if (blockSize != XZ_PROPS__BLOCK_SIZE__SOLID + && props->reduceSize != (UInt64)(Int64)-1) + { + numBlocks = props->reduceSize / blockSize; + if (numBlocks * blockSize != props->reduceSize) + numBlocks++; + } + else + blockSize = (UInt64)1 << 62; + + RINOK(XzEncIndex_PreAlloc(&p->xzIndex, numBlocks, blockSize, XZ_GET_ESTIMATED_BLOCK_TOTAL_PACK_SIZE(blockSize), p->alloc)); + } + + RINOK(Xz_WriteHeader((CXzStreamFlags)props->checkId, outStream)); + + + #ifndef _7ZIP_ST + if (props->numBlockThreads_Reduced > 1) + { + IMtCoderCallback2 vt; + + if (!p->mtCoder_WasConstructed) + { + p->mtCoder_WasConstructed = True; + MtCoder_Construct(&p->mtCoder); + } + + vt.Code = XzEnc_MtCallback_Code; + vt.Write = XzEnc_MtCallback_Write; + + p->checkType = props->checkId; + p->xzProps = *props; + + p->outStream = outStream; + + p->mtCoder.allocBig = p->allocBig; + p->mtCoder.progress = progress; + p->mtCoder.inStream = inStream; + p->mtCoder.inData = NULL; + p->mtCoder.inDataSize = 0; + p->mtCoder.mtCallback = &vt; + p->mtCoder.mtCallbackObject = p; + + if ( props->blockSize == XZ_PROPS__BLOCK_SIZE__SOLID + || props->blockSize == XZ_PROPS__BLOCK_SIZE__AUTO) + return SZ_ERROR_FAIL; + + p->mtCoder.blockSize = (size_t)props->blockSize; + if (p->mtCoder.blockSize != props->blockSize) + return SZ_ERROR_PARAM; /* SZ_ERROR_MEM */ + + { + size_t destBlockSize = XZ_BLOCK_HEADER_SIZE_MAX + XZ_GET_MAX_BLOCK_PACK_SIZE(p->mtCoder.blockSize); + if (destBlockSize < p->mtCoder.blockSize) + return SZ_ERROR_PARAM; + if (p->outBufSize != destBlockSize) + XzEnc_FreeOutBufs(p); + p->outBufSize = destBlockSize; + } + + p->mtCoder.numThreadsMax = (unsigned)props->numBlockThreads_Max; + p->mtCoder.expectedDataSize = p->expectedDataSize; + + RINOK(MtCoder_Code(&p->mtCoder)); + } + else + #endif + { + int writeStartSizes; + CCompressProgress_XzEncOffset progress2; + Byte *bufData = NULL; + size_t bufSize = 0; + + progress2.vt.Progress = CompressProgress_XzEncOffset_Progress; + progress2.inOffset = 0; + progress2.outOffset = 0; + progress2.progress = progress; + + writeStartSizes = 0; + + if (props->blockSize != XZ_PROPS__BLOCK_SIZE__SOLID) + { + writeStartSizes = (props->forceWriteSizesInHeader > 0); + + if (writeStartSizes) + { + size_t t2; + size_t t = (size_t)props->blockSize; + if (t != props->blockSize) + return SZ_ERROR_PARAM; + t = XZ_GET_MAX_BLOCK_PACK_SIZE(t); + if (t < props->blockSize) + return SZ_ERROR_PARAM; + t2 = XZ_BLOCK_HEADER_SIZE_MAX + t; + if (!p->outBufs[0] || t2 != p->outBufSize) + { + XzEnc_FreeOutBufs(p); + p->outBufs[0] = (Byte *)ISzAlloc_Alloc(p->alloc, t2); + if (!p->outBufs[0]) + return SZ_ERROR_MEM; + p->outBufSize = t2; + } + bufData = p->outBufs[0] + XZ_BLOCK_HEADER_SIZE_MAX; + bufSize = t; + } + } + + for (;;) + { + CXzEncBlockInfo blockSizes; + int inStreamFinished; + + /* + UInt64 rem = (UInt64)(Int64)-1; + if (props->reduceSize != (UInt64)(Int64)-1 + && props->reduceSize >= progress2.inOffset) + rem = props->reduceSize - progress2.inOffset; + */ + + blockSizes.headerSize = 0; // for GCC + + RINOK(Xz_CompressBlock( + &p->lzmaf_Items[0], + + writeStartSizes ? NULL : outStream, + writeStartSizes ? p->outBufs[0] : NULL, + bufData, bufSize, + + inStream, + // rem, + NULL, 0, + + props, + progress ? &progress2.vt : NULL, + &inStreamFinished, + &blockSizes, + p->alloc, + p->allocBig)); + + { + UInt64 totalPackFull = blockSizes.totalSize + XZ_GET_PAD_SIZE(blockSizes.totalSize); + + if (writeStartSizes) + { + RINOK(WriteBytes(outStream, p->outBufs[0], blockSizes.headerSize)); + RINOK(WriteBytes(outStream, bufData, (size_t)totalPackFull - blockSizes.headerSize)); + } + + RINOK(XzEncIndex_AddIndexRecord(&p->xzIndex, blockSizes.unpackSize, blockSizes.totalSize, p->alloc)); + + progress2.inOffset += blockSizes.unpackSize; + progress2.outOffset += totalPackFull; + } + + if (inStreamFinished) + break; + } + } + + return XzEncIndex_WriteFooter(&p->xzIndex, (CXzStreamFlags)props->checkId, outStream); +} + + +#include "Alloc.h" + +SRes Xz_Encode(ISeqOutStream *outStream, ISeqInStream *inStream, + const CXzProps *props, ICompressProgress *progress) +{ + SRes res; + CXzEncHandle xz = XzEnc_Create(&g_Alloc, &g_BigAlloc); + if (!xz) + return SZ_ERROR_MEM; + res = XzEnc_SetProps(xz, props); + if (res == SZ_OK) + res = XzEnc_Encode(xz, outStream, inStream, progress); + XzEnc_Destroy(xz); + return res; +} + + +SRes Xz_EncodeEmpty(ISeqOutStream *outStream) +{ + SRes res; + CXzEncIndex xzIndex; + XzEncIndex_Construct(&xzIndex); + res = Xz_WriteHeader((CXzStreamFlags)0, outStream); + if (res == SZ_OK) + res = XzEncIndex_WriteFooter(&xzIndex, (CXzStreamFlags)0, outStream); + XzEncIndex_Free(&xzIndex, NULL); // g_Alloc + return res; +} diff --git a/3rdparty/lzma/src/XzIn.c b/3rdparty/lzma/src/XzIn.c new file mode 100644 index 0000000000..84f868ec65 --- /dev/null +++ b/3rdparty/lzma/src/XzIn.c @@ -0,0 +1,325 @@ +/* XzIn.c - Xz input +2021-09-04 : Igor Pavlov : Public domain */ + +#include "Precomp.h" + +#include + +#include "7zCrc.h" +#include "CpuArch.h" +#include "Xz.h" + +/* +#define XZ_FOOTER_SIG_CHECK(p) (memcmp((p), XZ_FOOTER_SIG, XZ_FOOTER_SIG_SIZE) == 0) +*/ +#define XZ_FOOTER_SIG_CHECK(p) ((p)[0] == XZ_FOOTER_SIG_0 && (p)[1] == XZ_FOOTER_SIG_1) + + +SRes Xz_ReadHeader(CXzStreamFlags *p, ISeqInStream *inStream) +{ + Byte sig[XZ_STREAM_HEADER_SIZE]; + RINOK(SeqInStream_Read2(inStream, sig, XZ_STREAM_HEADER_SIZE, SZ_ERROR_NO_ARCHIVE)); + if (memcmp(sig, XZ_SIG, XZ_SIG_SIZE) != 0) + return SZ_ERROR_NO_ARCHIVE; + return Xz_ParseHeader(p, sig); +} + +#define READ_VARINT_AND_CHECK(buf, pos, size, res) \ + { unsigned s = Xz_ReadVarInt(buf + pos, size - pos, res); \ + if (s == 0) return SZ_ERROR_ARCHIVE; \ + pos += s; } + +SRes XzBlock_ReadHeader(CXzBlock *p, ISeqInStream *inStream, BoolInt *isIndex, UInt32 *headerSizeRes) +{ + Byte header[XZ_BLOCK_HEADER_SIZE_MAX]; + unsigned headerSize; + *headerSizeRes = 0; + RINOK(SeqInStream_ReadByte(inStream, &header[0])); + headerSize = (unsigned)header[0]; + if (headerSize == 0) + { + *headerSizeRes = 1; + *isIndex = True; + return SZ_OK; + } + + *isIndex = False; + headerSize = (headerSize << 2) + 4; + *headerSizeRes = headerSize; + RINOK(SeqInStream_Read(inStream, header + 1, headerSize - 1)); + return XzBlock_Parse(p, header); +} + +#define ADD_SIZE_CHECK(size, val) \ + { UInt64 newSize = size + (val); if (newSize < size) return XZ_SIZE_OVERFLOW; size = newSize; } + +UInt64 Xz_GetUnpackSize(const CXzStream *p) +{ + UInt64 size = 0; + size_t i; + for (i = 0; i < p->numBlocks; i++) + ADD_SIZE_CHECK(size, p->blocks[i].unpackSize); + return size; +} + +UInt64 Xz_GetPackSize(const CXzStream *p) +{ + UInt64 size = 0; + size_t i; + for (i = 0; i < p->numBlocks; i++) + ADD_SIZE_CHECK(size, (p->blocks[i].totalSize + 3) & ~(UInt64)3); + return size; +} + +/* +SRes XzBlock_ReadFooter(CXzBlock *p, CXzStreamFlags f, ISeqInStream *inStream) +{ + return SeqInStream_Read(inStream, p->check, XzFlags_GetCheckSize(f)); +} +*/ + +static SRes Xz_ReadIndex2(CXzStream *p, const Byte *buf, size_t size, ISzAllocPtr alloc) +{ + size_t numBlocks, pos = 1; + UInt32 crc; + + if (size < 5 || buf[0] != 0) + return SZ_ERROR_ARCHIVE; + + size -= 4; + crc = CrcCalc(buf, size); + if (crc != GetUi32(buf + size)) + return SZ_ERROR_ARCHIVE; + + { + UInt64 numBlocks64; + READ_VARINT_AND_CHECK(buf, pos, size, &numBlocks64); + numBlocks = (size_t)numBlocks64; + if (numBlocks != numBlocks64 || numBlocks * 2 > size) + return SZ_ERROR_ARCHIVE; + } + + Xz_Free(p, alloc); + if (numBlocks != 0) + { + size_t i; + p->numBlocks = numBlocks; + p->blocks = (CXzBlockSizes *)ISzAlloc_Alloc(alloc, sizeof(CXzBlockSizes) * numBlocks); + if (!p->blocks) + return SZ_ERROR_MEM; + for (i = 0; i < numBlocks; i++) + { + CXzBlockSizes *block = &p->blocks[i]; + READ_VARINT_AND_CHECK(buf, pos, size, &block->totalSize); + READ_VARINT_AND_CHECK(buf, pos, size, &block->unpackSize); + if (block->totalSize == 0) + return SZ_ERROR_ARCHIVE; + } + } + while ((pos & 3) != 0) + if (buf[pos++] != 0) + return SZ_ERROR_ARCHIVE; + return (pos == size) ? SZ_OK : SZ_ERROR_ARCHIVE; +} + +static SRes Xz_ReadIndex(CXzStream *p, ILookInStream *stream, UInt64 indexSize, ISzAllocPtr alloc) +{ + SRes res; + size_t size; + Byte *buf; + if (indexSize > ((UInt32)1 << 31)) + return SZ_ERROR_UNSUPPORTED; + size = (size_t)indexSize; + if (size != indexSize) + return SZ_ERROR_UNSUPPORTED; + buf = (Byte *)ISzAlloc_Alloc(alloc, size); + if (!buf) + return SZ_ERROR_MEM; + res = LookInStream_Read2(stream, buf, size, SZ_ERROR_UNSUPPORTED); + if (res == SZ_OK) + res = Xz_ReadIndex2(p, buf, size, alloc); + ISzAlloc_Free(alloc, buf); + return res; +} + +static SRes LookInStream_SeekRead_ForArc(ILookInStream *stream, UInt64 offset, void *buf, size_t size) +{ + RINOK(LookInStream_SeekTo(stream, offset)); + return LookInStream_Read(stream, buf, size); + /* return LookInStream_Read2(stream, buf, size, SZ_ERROR_NO_ARCHIVE); */ +} + +static SRes Xz_ReadBackward(CXzStream *p, ILookInStream *stream, Int64 *startOffset, ISzAllocPtr alloc) +{ + UInt64 indexSize; + Byte buf[XZ_STREAM_FOOTER_SIZE]; + UInt64 pos = (UInt64)*startOffset; + + if ((pos & 3) != 0 || pos < XZ_STREAM_FOOTER_SIZE) + return SZ_ERROR_NO_ARCHIVE; + + pos -= XZ_STREAM_FOOTER_SIZE; + RINOK(LookInStream_SeekRead_ForArc(stream, pos, buf, XZ_STREAM_FOOTER_SIZE)); + + if (!XZ_FOOTER_SIG_CHECK(buf + 10)) + { + UInt32 total = 0; + pos += XZ_STREAM_FOOTER_SIZE; + + for (;;) + { + size_t i; + #define TEMP_BUF_SIZE (1 << 10) + Byte temp[TEMP_BUF_SIZE]; + + i = (pos > TEMP_BUF_SIZE) ? TEMP_BUF_SIZE : (size_t)pos; + pos -= i; + RINOK(LookInStream_SeekRead_ForArc(stream, pos, temp, i)); + total += (UInt32)i; + for (; i != 0; i--) + if (temp[i - 1] != 0) + break; + if (i != 0) + { + if ((i & 3) != 0) + return SZ_ERROR_NO_ARCHIVE; + pos += i; + break; + } + if (pos < XZ_STREAM_FOOTER_SIZE || total > (1 << 16)) + return SZ_ERROR_NO_ARCHIVE; + } + + if (pos < XZ_STREAM_FOOTER_SIZE) + return SZ_ERROR_NO_ARCHIVE; + pos -= XZ_STREAM_FOOTER_SIZE; + RINOK(LookInStream_SeekRead_ForArc(stream, pos, buf, XZ_STREAM_FOOTER_SIZE)); + if (!XZ_FOOTER_SIG_CHECK(buf + 10)) + return SZ_ERROR_NO_ARCHIVE; + } + + p->flags = (CXzStreamFlags)GetBe16(buf + 8); + + if (!XzFlags_IsSupported(p->flags)) + return SZ_ERROR_UNSUPPORTED; + + { + /* to eliminate GCC 6.3 warning: + dereferencing type-punned pointer will break strict-aliasing rules */ + const Byte *buf_ptr = buf; + if (GetUi32(buf_ptr) != CrcCalc(buf + 4, 6)) + return SZ_ERROR_ARCHIVE; + } + + indexSize = ((UInt64)GetUi32(buf + 4) + 1) << 2; + + if (pos < indexSize) + return SZ_ERROR_ARCHIVE; + + pos -= indexSize; + RINOK(LookInStream_SeekTo(stream, pos)); + RINOK(Xz_ReadIndex(p, stream, indexSize, alloc)); + + { + UInt64 totalSize = Xz_GetPackSize(p); + if (totalSize == XZ_SIZE_OVERFLOW + || totalSize >= ((UInt64)1 << 63) + || pos < totalSize + XZ_STREAM_HEADER_SIZE) + return SZ_ERROR_ARCHIVE; + pos -= (totalSize + XZ_STREAM_HEADER_SIZE); + RINOK(LookInStream_SeekTo(stream, pos)); + *startOffset = (Int64)pos; + } + { + CXzStreamFlags headerFlags; + CSecToRead secToRead; + SecToRead_CreateVTable(&secToRead); + secToRead.realStream = stream; + + RINOK(Xz_ReadHeader(&headerFlags, &secToRead.vt)); + return (p->flags == headerFlags) ? SZ_OK : SZ_ERROR_ARCHIVE; + } +} + + +/* ---------- Xz Streams ---------- */ + +void Xzs_Construct(CXzs *p) +{ + p->num = p->numAllocated = 0; + p->streams = 0; +} + +void Xzs_Free(CXzs *p, ISzAllocPtr alloc) +{ + size_t i; + for (i = 0; i < p->num; i++) + Xz_Free(&p->streams[i], alloc); + ISzAlloc_Free(alloc, p->streams); + p->num = p->numAllocated = 0; + p->streams = 0; +} + +UInt64 Xzs_GetNumBlocks(const CXzs *p) +{ + UInt64 num = 0; + size_t i; + for (i = 0; i < p->num; i++) + num += p->streams[i].numBlocks; + return num; +} + +UInt64 Xzs_GetUnpackSize(const CXzs *p) +{ + UInt64 size = 0; + size_t i; + for (i = 0; i < p->num; i++) + ADD_SIZE_CHECK(size, Xz_GetUnpackSize(&p->streams[i])); + return size; +} + +/* +UInt64 Xzs_GetPackSize(const CXzs *p) +{ + UInt64 size = 0; + size_t i; + for (i = 0; i < p->num; i++) + ADD_SIZE_CHECK(size, Xz_GetTotalSize(&p->streams[i])); + return size; +} +*/ + +SRes Xzs_ReadBackward(CXzs *p, ILookInStream *stream, Int64 *startOffset, ICompressProgress *progress, ISzAllocPtr alloc) +{ + Int64 endOffset = 0; + RINOK(ILookInStream_Seek(stream, &endOffset, SZ_SEEK_END)); + *startOffset = endOffset; + for (;;) + { + CXzStream st; + SRes res; + Xz_Construct(&st); + res = Xz_ReadBackward(&st, stream, startOffset, alloc); + st.startOffset = (UInt64)*startOffset; + RINOK(res); + if (p->num == p->numAllocated) + { + const size_t newNum = p->num + p->num / 4 + 1; + void *data = ISzAlloc_Alloc(alloc, newNum * sizeof(CXzStream)); + if (!data) + return SZ_ERROR_MEM; + p->numAllocated = newNum; + if (p->num != 0) + memcpy(data, p->streams, p->num * sizeof(CXzStream)); + ISzAlloc_Free(alloc, p->streams); + p->streams = (CXzStream *)data; + } + p->streams[p->num++] = st; + if (*startOffset == 0) + break; + RINOK(LookInStream_SeekTo(stream, (UInt64)*startOffset)); + if (progress && ICompressProgress_Progress(progress, (UInt64)(endOffset - *startOffset), (UInt64)(Int64)-1) != SZ_OK) + return SZ_ERROR_PROGRESS; + } + return SZ_OK; +} diff --git a/PCSX2_qt.sln b/PCSX2_qt.sln index fa7a984c3d..c980ab1867 100644 --- a/PCSX2_qt.sln +++ b/PCSX2_qt.sln @@ -62,6 +62,8 @@ Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "zstd", "3rdparty\zstd\zstd. EndProject Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "d3d12memalloc", "3rdparty\d3d12memalloc\d3d12memalloc.vcxproj", "{D45CEC7A-3171-40DD-975D-E1544CF16139}" EndProject +Project("{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}") = "lzma", "3rdparty\lzma\lzma.vcxproj", "{A4323327-3F2B-4271-83D9-7F9A3C66B6B2}" +EndProject Global GlobalSection(SolutionConfigurationPlatforms) = preSolution Debug AVX2|x64 = Debug AVX2|x64 @@ -384,6 +386,18 @@ Global {D45CEC7A-3171-40DD-975D-E1544CF16139}.Release AVX2|x64.Build.0 = Release|x64 {D45CEC7A-3171-40DD-975D-E1544CF16139}.Release|x64.ActiveCfg = Release|x64 {D45CEC7A-3171-40DD-975D-E1544CF16139}.Release|x64.Build.0 = Release|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Debug AVX2|x64.ActiveCfg = Debug|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Debug AVX2|x64.Build.0 = Debug|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Debug|x64.ActiveCfg = Debug|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Debug|x64.Build.0 = Debug|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Devel AVX2|x64.ActiveCfg = Devel|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Devel AVX2|x64.Build.0 = Devel|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Devel|x64.ActiveCfg = Devel|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Devel|x64.Build.0 = Devel|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Release AVX2|x64.ActiveCfg = Release|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Release AVX2|x64.Build.0 = Release|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Release|x64.ActiveCfg = Release|x64 + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2}.Release|x64.Build.0 = Release|x64 EndGlobalSection GlobalSection(SolutionProperties) = preSolution HideSolutionNode = FALSE @@ -412,6 +426,7 @@ Global {20B2E9FE-F020-42A0-B324-956F5B06EA68} = {78EBE642-7A4D-4EA7-86BE-5639C6646C38} {52244028-937A-44E9-A76B-2BEA18FD239A} = {78EBE642-7A4D-4EA7-86BE-5639C6646C38} {D45CEC7A-3171-40DD-975D-E1544CF16139} = {78EBE642-7A4D-4EA7-86BE-5639C6646C38} + {A4323327-3F2B-4271-83D9-7F9A3C66B6B2} = {78EBE642-7A4D-4EA7-86BE-5639C6646C38} EndGlobalSection GlobalSection(ExtensibilityGlobals) = postSolution SolutionGuid = {0BC474EA-3628-45D3-9DBC-E22D0B7E0F77} diff --git a/cmake/SearchForStuff.cmake b/cmake/SearchForStuff.cmake index 6f6a902d53..fba6a42fc5 100644 --- a/cmake/SearchForStuff.cmake +++ b/cmake/SearchForStuff.cmake @@ -17,6 +17,7 @@ if (WIN32) add_subdirectory(3rdparty/wxwidgets3.0 EXCLUDE_FROM_ALL) add_subdirectory(3rdparty/xz EXCLUDE_FROM_ALL) add_subdirectory(3rdparty/D3D12MemAlloc EXCLUDE_FROM_ALL) + add_subdirectory(3rdparty/lzma EXCLUDE_FROM_ALL) else() ## Use cmake package to find module if (Linux)