From 0e1344eedd1a176106d84311b9c699d9a4e2426f Mon Sep 17 00:00:00 2001 From: andon Date: Wed, 1 Feb 2017 03:34:07 -0500 Subject: [PATCH] Add texture mod support --- include/DLL_VERSION.H | 4 +- include/config.h | 2 + include/lzma/7z.h | 202 ++++ include/lzma/7zAlloc.h | 23 + include/lzma/7zBuf.h | 35 + include/lzma/7zCrc.h | 25 + include/lzma/7zFile.h | 83 ++ include/lzma/7zTypes.h | 256 ++++ include/lzma/7zVersion.h | 19 + include/lzma/7zVersion.rc | 55 + include/lzma/Aes.h | 38 + include/lzma/Alloc.h | 37 + include/lzma/Bcj2.h | 146 +++ include/lzma/Bra.h | 64 + include/lzma/Compiler.h | 32 + include/lzma/CpuArch.h | 222 ++++ include/lzma/LzFind.h | 117 ++ include/lzma/LzFindMt.h | 101 ++ include/lzma/LzHash.h | 57 + include/lzma/Lzma2Dec.h | 80 ++ include/lzma/Lzma2Enc.h | 62 + include/lzma/Lzma86.h | 111 ++ include/lzma/LzmaDec.h | 227 ++++ include/lzma/LzmaEnc.h | 78 ++ include/lzma/LzmaLib.h | 131 +++ include/lzma/MtCoder.h | 98 ++ include/lzma/Ppmd.h | 85 ++ include/lzma/Ppmd7.h | 140 +++ include/lzma/Precomp.h | 10 + include/lzma/RotateDefs.h | 30 + include/lzma/Sha256.h | 26 + include/lzma/Sort.h | 18 + include/lzma/Threads.h | 67 ++ include/lzma/Xz.h | 275 +++++ include/lzma/XzCrc64.h | 26 + include/lzma/XzEnc.h | 39 + include/lzma/delta.h | 19 + resource/symbols.def | 4 +- src/config.cpp | 13 + src/dllmain.cpp | 57 +- src/framerate.cpp | 4 - src/lzma/7zAlloc.c | 78 ++ src/lzma/7zArcIn.c | 1773 ++++++++++++++++++++++++++++ src/lzma/7zBuf.c | 36 + src/lzma/7zBuf2.c | 51 + src/lzma/7zCrc.c | 128 ++ src/lzma/7zCrcOpt.c | 114 ++ src/lzma/7zDec.c | 591 ++++++++++ src/lzma/7zFile.c | 286 +++++ src/lzma/7zStream.c | 171 +++ src/lzma/Aes.c | 295 +++++ src/lzma/AesOpt.c | 184 +++ src/lzma/Alloc.c | 136 +++ src/lzma/Bcj2.c | 256 ++++ src/lzma/Bcj2Enc.c | 312 +++++ src/lzma/Bra.c | 135 +++ src/lzma/Bra86.c | 82 ++ src/lzma/BraIA64.c | 69 ++ src/lzma/CpuArch.c | 200 ++++ src/lzma/Delta.c | 64 + src/lzma/LzFind.c | 1044 ++++++++++++++++ src/lzma/LzFindMt.c | 803 +++++++++++++ src/lzma/Lzma2Dec.c | 378 ++++++ src/lzma/Lzma2Enc.c | 520 ++++++++ src/lzma/Lzma86Dec.c | 54 + src/lzma/Lzma86Enc.c | 106 ++ src/lzma/LzmaDec.c | 1100 +++++++++++++++++ src/lzma/LzmaEnc.c | 2351 +++++++++++++++++++++++++++++++++++++ src/lzma/LzmaLib.c | 40 + src/lzma/MtCoder.c | 327 ++++++ src/lzma/Ppmd7.c | 710 +++++++++++ src/lzma/Ppmd7Dec.c | 189 +++ src/lzma/Ppmd7Enc.c | 187 +++ src/lzma/Sha256.c | 248 ++++ src/lzma/Sort.c | 141 +++ src/lzma/Threads.c | 93 ++ src/lzma/Xz.c | 90 ++ src/lzma/XzCrc64.c | 86 ++ src/lzma/XzCrc64Opt.c | 69 ++ src/lzma/XzDec.c | 913 ++++++++++++++ src/lzma/XzEnc.c | 538 +++++++++ src/lzma/XzIn.c | 313 +++++ src/render.cpp | 25 +- src/sound.cpp | 66 +- src/tbt/control_panel.cpp | 61 +- src/textures.cpp | 567 +++++++-- tbf.vcxproj | 59 +- tbf.vcxproj.filters | 126 ++ version.ini | Bin 8010 -> 7926 bytes 89 files changed, 19095 insertions(+), 188 deletions(-) create mode 100644 include/lzma/7z.h create mode 100644 include/lzma/7zAlloc.h create mode 100644 include/lzma/7zBuf.h create mode 100644 include/lzma/7zCrc.h create mode 100644 include/lzma/7zFile.h create mode 100644 include/lzma/7zTypes.h create mode 100644 include/lzma/7zVersion.h create mode 100644 include/lzma/7zVersion.rc create mode 100644 include/lzma/Aes.h create mode 100644 include/lzma/Alloc.h create mode 100644 include/lzma/Bcj2.h create mode 100644 include/lzma/Bra.h create mode 100644 include/lzma/Compiler.h create mode 100644 include/lzma/CpuArch.h create mode 100644 include/lzma/LzFind.h create mode 100644 include/lzma/LzFindMt.h create mode 100644 include/lzma/LzHash.h create mode 100644 include/lzma/Lzma2Dec.h create mode 100644 include/lzma/Lzma2Enc.h create mode 100644 include/lzma/Lzma86.h create mode 100644 include/lzma/LzmaDec.h create mode 100644 include/lzma/LzmaEnc.h create mode 100644 include/lzma/LzmaLib.h create mode 100644 include/lzma/MtCoder.h create mode 100644 include/lzma/Ppmd.h create mode 100644 include/lzma/Ppmd7.h create mode 100644 include/lzma/Precomp.h create mode 100644 include/lzma/RotateDefs.h create mode 100644 include/lzma/Sha256.h create mode 100644 include/lzma/Sort.h create mode 100644 include/lzma/Threads.h create mode 100644 include/lzma/Xz.h create mode 100644 include/lzma/XzCrc64.h create mode 100644 include/lzma/XzEnc.h create mode 100644 include/lzma/delta.h create mode 100644 src/lzma/7zAlloc.c create mode 100644 src/lzma/7zArcIn.c create mode 100644 src/lzma/7zBuf.c create mode 100644 src/lzma/7zBuf2.c create mode 100644 src/lzma/7zCrc.c create mode 100644 src/lzma/7zCrcOpt.c create mode 100644 src/lzma/7zDec.c create mode 100644 src/lzma/7zFile.c create mode 100644 src/lzma/7zStream.c create mode 100644 src/lzma/Aes.c create mode 100644 src/lzma/AesOpt.c create mode 100644 src/lzma/Alloc.c create mode 100644 src/lzma/Bcj2.c create mode 100644 src/lzma/Bcj2Enc.c create mode 100644 src/lzma/Bra.c create mode 100644 src/lzma/Bra86.c create mode 100644 src/lzma/BraIA64.c create mode 100644 src/lzma/CpuArch.c create mode 100644 src/lzma/Delta.c create mode 100644 src/lzma/LzFind.c create mode 100644 src/lzma/LzFindMt.c create mode 100644 src/lzma/Lzma2Dec.c create mode 100644 src/lzma/Lzma2Enc.c create mode 100644 src/lzma/Lzma86Dec.c create mode 100644 src/lzma/Lzma86Enc.c create mode 100644 src/lzma/LzmaDec.c create mode 100644 src/lzma/LzmaEnc.c create mode 100644 src/lzma/LzmaLib.c create mode 100644 src/lzma/MtCoder.c create mode 100644 src/lzma/Ppmd7.c create mode 100644 src/lzma/Ppmd7Dec.c create mode 100644 src/lzma/Ppmd7Enc.c create mode 100644 src/lzma/Sha256.c create mode 100644 src/lzma/Sort.c create mode 100644 src/lzma/Threads.c create mode 100644 src/lzma/Xz.c create mode 100644 src/lzma/XzCrc64.c create mode 100644 src/lzma/XzCrc64Opt.c create mode 100644 src/lzma/XzDec.c create mode 100644 src/lzma/XzEnc.c create mode 100644 src/lzma/XzIn.c diff --git a/include/DLL_VERSION.H b/include/DLL_VERSION.H index 536fc27..3835783 100644 --- a/include/DLL_VERSION.H +++ b/include/DLL_VERSION.H @@ -2,8 +2,8 @@ #define TBF_MAJOR 0 -#define TBF_MINOR 5 -#define TBF_BUILD 3 +#define TBF_MINOR 6 +#define TBF_BUILD 0 #define TBF_REV 0 diff --git a/include/config.h b/include/config.h index df08074..acdf56d 100644 --- a/include/config.h +++ b/include/config.h @@ -65,6 +65,7 @@ struct tbf_config_t int32_t shadow_rescale = -2; int32_t env_shadow_rescale = 1; bool dump_shaders = false; + bool fix_map_res = false; } render; struct { @@ -75,6 +76,7 @@ struct tbf_config_t float lod_bias = -0.2666f; int32_t max_cache_in_mib = 2048L; int32_t worker_threads = 3; + int32_t max_decomp_jobs = 6; bool show_loading_text = false; bool quick_load = false; } textures; diff --git a/include/lzma/7z.h b/include/lzma/7z.h new file mode 100644 index 0000000..4768151 --- /dev/null +++ b/include/lzma/7z.h @@ -0,0 +1,202 @@ +/* 7z.h -- 7z interface +2015-11-18 : 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; +} 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, + ISzAlloc *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, ISzAlloc *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 */ + ISzAlloc *allocMain, + ISzAlloc *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, + ISzAlloc *allocMain, ISzAlloc *allocTemp); + +EXTERN_C_END + +#endif diff --git a/include/lzma/7zAlloc.h b/include/lzma/7zAlloc.h new file mode 100644 index 0000000..2fd5bdb --- /dev/null +++ b/include/lzma/7zAlloc.h @@ -0,0 +1,23 @@ +/* 7zAlloc.h -- Allocation functions +2013-03-25 : Igor Pavlov : Public domain */ + +#ifndef __7Z_ALLOC_H +#define __7Z_ALLOC_H + +#include + +#ifdef __cplusplus +extern "C" { +#endif + +void *SzAlloc(void *p, size_t size); +void SzFree(void *p, void *address); + +void *SzAllocTemp(void *p, size_t size); +void SzFreeTemp(void *p, void *address); + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/include/lzma/7zBuf.h b/include/lzma/7zBuf.h new file mode 100644 index 0000000..65f1d7a --- /dev/null +++ b/include/lzma/7zBuf.h @@ -0,0 +1,35 @@ +/* 7zBuf.h -- Byte Buffer +2013-01-18 : 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, ISzAlloc *alloc); +void Buf_Free(CBuf *p, ISzAlloc *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, ISzAlloc *alloc); +void DynBuf_Free(CDynBuf *p, ISzAlloc *alloc); + +EXTERN_C_END + +#endif diff --git a/include/lzma/7zCrc.h b/include/lzma/7zCrc.h new file mode 100644 index 0000000..8fd5795 --- /dev/null +++ b/include/lzma/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/include/lzma/7zFile.h b/include/lzma/7zFile.h new file mode 100644 index 0000000..658987e --- /dev/null +++ b/include/lzma/7zFile.h @@ -0,0 +1,83 @@ +/* 7zFile.h -- File IO +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __7Z_FILE_H +#define __7Z_FILE_H + +#ifdef _WIN32 +#define USE_WINDOWS_FILE +#endif + +#ifdef USE_WINDOWS_FILE +#include +#else +#include +#endif + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +/* ---------- File ---------- */ + +typedef struct +{ + #ifdef USE_WINDOWS_FILE + HANDLE handle; + #else + FILE *file; + #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 s; + CSzFile file; +} CFileSeqInStream; + +void FileSeqInStream_CreateVTable(CFileSeqInStream *p); + + +typedef struct +{ + ISeekInStream s; + CSzFile file; +} CFileInStream; + +void FileInStream_CreateVTable(CFileInStream *p); + + +typedef struct +{ + ISeqOutStream s; + CSzFile file; +} CFileOutStream; + +void FileOutStream_CreateVTable(CFileOutStream *p); + +EXTERN_C_END + +#endif diff --git a/include/lzma/7zTypes.h b/include/lzma/7zTypes.h new file mode 100644 index 0000000..778413e --- /dev/null +++ b/include/lzma/7zTypes.h @@ -0,0 +1,256 @@ +/* 7zTypes.h -- Basic types +2013-11-12 : Igor Pavlov : Public domain */ + +#ifndef __7Z_TYPES_H +#define __7Z_TYPES_H + +#ifdef _WIN32 +/* #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 _WIN32 +/* typedef DWORD WRes; */ +typedef unsigned WRes; +#else +typedef int WRes; +#endif + +#ifndef RINOK +#define RINOK(x) { int __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 + +#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 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_CDECL __cdecl +#define MY_FAST_CALL __fastcall + +#else + +#define MY_NO_INLINE +#define MY_CDECL +#define MY_FAST_CALL + +#endif + + +/* The following interfaces use first parameter as pointer to structure */ + +typedef struct +{ + Byte (*Read)(void *p); /* reads one byte, returns 0 in case of EOF or error */ +} IByteIn; + +typedef struct +{ + void (*Write)(void *p, Byte b); +} IByteOut; + +typedef struct +{ + SRes (*Read)(void *p, void *buf, size_t *size); + /* if (input(*size) != 0 && output(*size) == 0) means end_of_stream. + (output(*size) < input(*size)) is allowed */ +} ISeqInStream; + +/* it can return SZ_ERROR_INPUT_EOF */ +SRes SeqInStream_Read(ISeqInStream *stream, void *buf, size_t size); +SRes SeqInStream_Read2(ISeqInStream *stream, void *buf, size_t size, SRes errorType); +SRes SeqInStream_ReadByte(ISeqInStream *stream, Byte *buf); + +typedef struct +{ + size_t (*Write)(void *p, const void *buf, size_t size); + /* Returns: result - the number of actually written bytes. + (result < size) means error */ +} ISeqOutStream; + +typedef enum +{ + SZ_SEEK_SET = 0, + SZ_SEEK_CUR = 1, + SZ_SEEK_END = 2 +} ESzSeek; + +typedef struct +{ + SRes (*Read)(void *p, void *buf, size_t *size); /* same as ISeqInStream::Read */ + SRes (*Seek)(void *p, Int64 *pos, ESzSeek origin); +} ISeekInStream; + +typedef struct +{ + SRes (*Look)(void *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)(void *p, size_t offset); + /* offset must be <= output(*size) of Look */ + + SRes (*Read)(void *p, void *buf, size_t *size); + /* reads directly (without buffer). It's same as ISeqInStream::Read */ + SRes (*Seek)(void *p, Int64 *pos, ESzSeek origin); +} ILookInStream; + +SRes LookInStream_LookRead(ILookInStream *stream, void *buf, size_t *size); +SRes LookInStream_SeekTo(ILookInStream *stream, UInt64 offset); + +/* reads via ILookInStream::Read */ +SRes LookInStream_Read2(ILookInStream *stream, void *buf, size_t size, SRes errorType); +SRes LookInStream_Read(ILookInStream *stream, void *buf, size_t size); + +#define LookToRead_BUF_SIZE (1 << 14) + +typedef struct +{ + ILookInStream s; + ISeekInStream *realStream; + size_t pos; + size_t size; + Byte buf[LookToRead_BUF_SIZE]; +} CLookToRead; + +void LookToRead_CreateVTable(CLookToRead *p, int lookahead); +void LookToRead_Init(CLookToRead *p); + +typedef struct +{ + ISeqInStream s; + ILookInStream *realStream; +} CSecToLook; + +void SecToLook_CreateVTable(CSecToLook *p); + +typedef struct +{ + ISeqInStream s; + ILookInStream *realStream; +} CSecToRead; + +void SecToRead_CreateVTable(CSecToRead *p); + +typedef struct +{ + SRes (*Progress)(void *p, UInt64 inSize, UInt64 outSize); + /* Returns: result. (result != SZ_OK) means break. + Value (UInt64)(Int64)-1 for size means unknown value. */ +} ICompressProgress; + +typedef struct +{ + void *(*Alloc)(void *p, size_t size); + void (*Free)(void *p, void *address); /* address can be 0 */ +} ISzAlloc; + +#define IAlloc_Alloc(p, size) (p)->Alloc((p), size) +#define IAlloc_Free(p, a) (p)->Free((p), 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/include/lzma/7zVersion.h b/include/lzma/7zVersion.h new file mode 100644 index 0000000..e884fc6 --- /dev/null +++ b/include/lzma/7zVersion.h @@ -0,0 +1,19 @@ +#define MY_VER_MAJOR 16 +#define MY_VER_MINOR 02 +#define MY_VER_BUILD 0 +#define MY_VERSION_NUMBERS "16.02" +#define MY_VERSION "16.02" +#define MY_DATE "2016-05-21" +#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) 1999-2016 Igor Pavlov" + +#ifdef USE_COPYRIGHT_CR + #define MY_COPYRIGHT MY_COPYRIGHT_CR +#else + #define MY_COPYRIGHT MY_COPYRIGHT_PD +#endif + +#define MY_VERSION_COPYRIGHT_DATE MY_VERSION " : " MY_COPYRIGHT " : " MY_DATE diff --git a/include/lzma/7zVersion.rc b/include/lzma/7zVersion.rc new file mode 100644 index 0000000..e520995 --- /dev/null +++ b/include/lzma/7zVersion.rc @@ -0,0 +1,55 @@ +#define MY_VS_FFI_FILEFLAGSMASK 0x0000003FL +#define MY_VOS_NT_WINDOWS32 0x00040004L +#define MY_VOS_CE_WINDOWS32 0x00050004L + +#define MY_VFT_APP 0x00000001L +#define MY_VFT_DLL 0x00000002L + +// #include + +#ifndef MY_VERSION +#include "7zVersion.h" +#endif + +#define MY_VER MY_VER_MAJOR,MY_VER_MINOR,MY_VER_BUILD,0 + +#ifdef DEBUG +#define DBG_FL VS_FF_DEBUG +#else +#define DBG_FL 0 +#endif + +#define MY_VERSION_INFO(fileType, descr, intName, origName) \ +LANGUAGE 9, 1 \ +1 VERSIONINFO \ + FILEVERSION MY_VER \ + PRODUCTVERSION MY_VER \ + FILEFLAGSMASK MY_VS_FFI_FILEFLAGSMASK \ + FILEFLAGS DBG_FL \ + FILEOS MY_VOS_NT_WINDOWS32 \ + FILETYPE fileType \ + FILESUBTYPE 0x0L \ +BEGIN \ + BLOCK "StringFileInfo" \ + BEGIN \ + BLOCK "040904b0" \ + BEGIN \ + VALUE "CompanyName", "Igor Pavlov" \ + VALUE "FileDescription", descr \ + VALUE "FileVersion", MY_VERSION \ + VALUE "InternalName", intName \ + VALUE "LegalCopyright", MY_COPYRIGHT \ + VALUE "OriginalFilename", origName \ + VALUE "ProductName", "7-Zip" \ + VALUE "ProductVersion", MY_VERSION \ + END \ + END \ + BLOCK "VarFileInfo" \ + BEGIN \ + VALUE "Translation", 0x409, 1200 \ + END \ +END + +#define MY_VERSION_INFO_APP(descr, intName) MY_VERSION_INFO(MY_VFT_APP, descr, intName, intName ".exe") + +#define MY_VERSION_INFO_DLL(descr, intName) MY_VERSION_INFO(MY_VFT_DLL, descr, intName, intName ".dll") diff --git a/include/lzma/Aes.h b/include/lzma/Aes.h new file mode 100644 index 0000000..64979b5 --- /dev/null +++ b/include/lzma/Aes.h @@ -0,0 +1,38 @@ +/* Aes.h -- AES encryption / decryption +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __AES_H +#define __AES_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define AES_BLOCK_SIZE 16 + +/* Call AesGenTables one time before other AES functions */ +void AesGenTables(void); + +/* UInt32 pointers must be 16-byte aligned */ + +/* 16-byte (4 * 32-bit words) blocks: 1 (IV) + 1 (keyMode) + 15 (AES-256 roundKeys) */ +#define AES_NUM_IVMRK_WORDS ((1 + 1 + 15) * 4) + +/* aes - 16-byte aligned pointer to keyMode+roundKeys sequence */ +/* keySize = 16 or 24 or 32 (bytes) */ +typedef void (MY_FAST_CALL *AES_SET_KEY_FUNC)(UInt32 *aes, const Byte *key, unsigned keySize); +void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *aes, const Byte *key, unsigned keySize); +void MY_FAST_CALL Aes_SetKey_Dec(UInt32 *aes, const Byte *key, unsigned keySize); + +/* ivAes - 16-byte aligned pointer to iv+keyMode+roundKeys sequence: UInt32[AES_NUM_IVMRK_WORDS] */ +void AesCbc_Init(UInt32 *ivAes, const Byte *iv); /* iv size is AES_BLOCK_SIZE */ +/* data - 16-byte aligned pointer to data */ +/* numBlocks - the number of 16-byte blocks in data array */ +typedef void (MY_FAST_CALL *AES_CODE_FUNC)(UInt32 *ivAes, Byte *data, size_t numBlocks); +extern AES_CODE_FUNC g_AesCbc_Encode; +extern AES_CODE_FUNC g_AesCbc_Decode; +extern AES_CODE_FUNC g_AesCtr_Code; + +EXTERN_C_END + +#endif diff --git a/include/lzma/Alloc.h b/include/lzma/Alloc.h new file mode 100644 index 0000000..bac0d87 --- /dev/null +++ b/include/lzma/Alloc.h @@ -0,0 +1,37 @@ +/* Alloc.h -- Memory allocation functions +2015-02-21 : 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 *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 ISzAlloc g_Alloc; +extern ISzAlloc g_BigAlloc; + +EXTERN_C_END + +#endif diff --git a/include/lzma/Bcj2.h b/include/lzma/Bcj2.h new file mode 100644 index 0000000..8824080 --- /dev/null +++ b/include/lzma/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/include/lzma/Bra.h b/include/lzma/Bra.h new file mode 100644 index 0000000..855e37a --- /dev/null +++ b/include/lzma/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/include/lzma/Compiler.h b/include/lzma/Compiler.h new file mode 100644 index 0000000..5bba7ee --- /dev/null +++ b/include/lzma/Compiler.h @@ -0,0 +1,32 @@ +/* Compiler.h +2015-08-02 : Igor Pavlov : Public domain */ + +#ifndef __7Z_COMPILER_H +#define __7Z_COMPILER_H + +#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 : 4786) // identifier was truncated to '255' characters in the debug information + #endif + +#endif + +#define UNUSED_VAR(x) (void)x; +/* #define UNUSED_VAR(x) x=x; */ + +#endif diff --git a/include/lzma/CpuArch.h b/include/lzma/CpuArch.h new file mode 100644 index 0000000..f6a28ba --- /dev/null +++ b/include/lzma/CpuArch.h @@ -0,0 +1,222 @@ +/* CpuArch.h -- CPU specific code +2015-12-01: 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. +*/ + +#if defined(_M_X64) \ + || defined(_M_AMD64) \ + || defined(__x86_64__) \ + || defined(__AMD64__) \ + || defined(__amd64__) + #define MY_CPU_AMD64 +#endif + +#if defined(MY_CPU_AMD64) \ + || defined(_M_IA64) \ + || defined(__AARCH64EL__) \ + || defined(__AARCH64EB__) + #define MY_CPU_64BIT +#endif + +#if defined(_M_IX86) || defined(__i386__) +#define MY_CPU_X86 +#endif + +#if defined(MY_CPU_X86) || defined(MY_CPU_AMD64) +#define MY_CPU_X86_OR_AMD64 +#endif + +#if defined(MY_CPU_X86) \ + || defined(_M_ARM) \ + || defined(__ARMEL__) \ + || defined(__THUMBEL__) \ + || defined(__ARMEB__) \ + || defined(__THUMBEB__) + #define MY_CPU_32BIT +#endif + +#if defined(_WIN32) && defined(_M_ARM) +#define MY_CPU_ARM_LE +#endif + +#if defined(_WIN32) && defined(_M_IA64) +#define MY_CPU_IA64_LE +#endif + +#if defined(MY_CPU_X86_OR_AMD64) \ + || defined(MY_CPU_ARM_LE) \ + || defined(MY_CPU_IA64_LE) \ + || defined(__LITTLE_ENDIAN__) \ + || defined(__ARMEL__) \ + || defined(__THUMBEL__) \ + || defined(__AARCH64EL__) \ + || defined(__MIPSEL__) \ + || defined(__MIPSEL) \ + || defined(_MIPSEL) \ + || (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) +Stop_Compiling_Bad_Endian +#endif + + +#ifdef MY_CPU_LE + #if defined(MY_CPU_X86_OR_AMD64) \ + /* || defined(__AARCH64EL__) */ + #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)) +#define GetUi64(p) (*(const UInt64 *)(const void *)(p)) + +#define SetUi16(p, v) { *(UInt16 *)(p) = (v); } +#define SetUi32(p, v) { *(UInt32 *)(p) = (v); } +#define SetUi64(p, v) { *(UInt64 *)(p) = (v); } + +#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 GetUi64(p) (GetUi32(p) | ((UInt64)GetUi32(((const Byte *)(p)) + 4) << 32)) + +#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); } + +#define SetUi64(p, v) { Byte *_ppp2_ = (Byte *)(p); UInt64 _vvv2_ = (v); \ + SetUi32(_ppp2_ , (UInt32)_vvv2_); \ + SetUi32(_ppp2_ + 4, (UInt32)(_vvv2_ >> 32)); } + +#endif + + +#if defined(MY_CPU_LE_UNALIGN) && /* defined(_WIN64) && */ (_MSC_VER >= 1300) + +/* Note: we use bswap instruction, that is unsupported in 386 cpu */ + +#include + +#pragma intrinsic(_byteswap_ulong) +#pragma intrinsic(_byteswap_uint64) +#define GetBe32(p) _byteswap_ulong(*(const UInt32 *)(const Byte *)(p)) +#define GetBe64(p) _byteswap_uint64(*(const UInt64 *)(const Byte *)(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)) + +#define GetBe32(p) __builtin_bswap32(*(const UInt32 *)(const Byte *)(p)) +#define GetBe64(p) __builtin_bswap64(*(const UInt64 *)(const Byte *)(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 + + +#define GetBe16(p) ( (UInt16) ( \ + ((UInt16)((const Byte *)(p))[0] << 8) | \ + ((const Byte *)(p))[1] )) + + + +#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); + +Bool 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) + +Bool CPU_Is_InOrder(); +Bool CPU_Is_Aes_Supported(); + +#endif + +EXTERN_C_END + +#endif diff --git a/include/lzma/LzFind.h b/include/lzma/LzFind.h new file mode 100644 index 0000000..d119944 --- /dev/null +++ b/include/lzma/LzFind.h @@ -0,0 +1,117 @@ +/* LzFind.h -- Match finder for LZ algorithms +2015-10-15 : 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; + 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; +} CMatchFinder; + +#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((p)->buffer) + +#define Inline_MatchFinder_GetNumAvailableBytes(p) ((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, + ISzAlloc *alloc); +void MatchFinder_Free(CMatchFinder *p, ISzAlloc *alloc); +void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems); +void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue); + +UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *buffer, CLzRef *son, + UInt32 _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; +} IMatchFinder; + +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder *vTable); + +void MatchFinder_Init_2(CMatchFinder *p, int readData); +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); + +EXTERN_C_END + +#endif diff --git a/include/lzma/LzFindMt.h b/include/lzma/LzFindMt.h new file mode 100644 index 0000000..89b91fe --- /dev/null +++ b/include/lzma/LzFindMt.h @@ -0,0 +1,101 @@ +/* LzFindMt.h -- multithreaded Match finder for LZ algorithms +2015-05-03 : Igor Pavlov : Public domain */ + +#ifndef __LZ_FIND_MT_H +#define __LZ_FIND_MT_H + +#include "LzFind.h" +#include "Threads.h" + +EXTERN_C_BEGIN + +#define kMtHashBlockSize (1 << 13) +#define kMtHashNumBlocks (1 << 3) +#define kMtHashNumBlocksMask (kMtHashNumBlocks - 1) + +#define kMtBtBlockSize (1 << 14) +#define kMtBtNumBlocks (1 << 6) +#define kMtBtNumBlocksMask (kMtBtNumBlocks - 1) + +typedef struct _CMtSync +{ + Bool wasCreated; + Bool needStart; + Bool exit; + Bool stopWriting; + + CThread thread; + CAutoResetEvent canStart; + CAutoResetEvent wasStarted; + CAutoResetEvent wasStopped; + CSemaphore freeSemaphore; + CSemaphore filledSemaphore; + Bool csWasInitialized; + Bool csWasEntered; + CCriticalSection cs; + UInt32 numProcessedBlocks; +} CMtSync; + +typedef UInt32 * (*Mf_Mix_Matches)(void *p, UInt32 matchMinPos, UInt32 *distances); + +/* kMtCacheLineDummy must be >= size_of_CPU_cache_line */ +#define kMtCacheLineDummy 128 + +typedef void (*Mf_GetHeads)(const Byte *buffer, UInt32 pos, + UInt32 *hash, UInt32 hashMask, UInt32 *heads, UInt32 numHeads, const UInt32 *crc); + +typedef struct _CMatchFinderMt +{ + /* LZ */ + const Byte *pointerToCurPos; + UInt32 *btBuf; + UInt32 btBufPos; + UInt32 btBufPosLimit; + UInt32 lzPos; + UInt32 btNumAvailBytes; + + UInt32 *hash; + UInt32 fixedHashSize; + UInt32 historySize; + const UInt32 *crc; + + Mf_Mix_Matches MixMatchesFunc; + + /* LZ + BT */ + CMtSync btSync; + Byte btDummy[kMtCacheLineDummy]; + + /* BT */ + UInt32 *hashBuf; + UInt32 hashBufPos; + UInt32 hashBufPosLimit; + UInt32 hashNumAvail; + + CLzRef *son; + UInt32 matchMaxLen; + UInt32 numHashBytes; + UInt32 pos; + const Byte *buffer; + UInt32 cyclicBufferPos; + UInt32 cyclicBufferSize; /* it must be historySize + 1 */ + UInt32 cutValue; + + /* BT + Hash */ + CMtSync hashSync; + /* Byte hashDummy[kMtCacheLineDummy]; */ + + /* Hash */ + Mf_GetHeads GetHeadsFunc; + CMatchFinder *MatchFinder; +} CMatchFinderMt; + +void MatchFinderMt_Construct(CMatchFinderMt *p); +void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAlloc *alloc); +SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore, + UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAlloc *alloc); +void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder *vTable); +void MatchFinderMt_ReleaseStream(CMatchFinderMt *p); + +EXTERN_C_END + +#endif diff --git a/include/lzma/LzHash.h b/include/lzma/LzHash.h new file mode 100644 index 0000000..e7c9423 --- /dev/null +++ b/include/lzma/LzHash.h @@ -0,0 +1,57 @@ +/* LzHash.h -- HASH functions for LZ algorithms +2015-04-12 : Igor Pavlov : Public domain */ + +#ifndef __LZ_HASH_H +#define __LZ_HASH_H + +#define kHash2Size (1 << 10) +#define kHash3Size (1 << 16) +#define kHash4Size (1 << 20) + +#define kFix3HashSize (kHash2Size) +#define kFix4HashSize (kHash2Size + kHash3Size) +#define kFix5HashSize (kHash2Size + kHash3Size + kHash4Size) + +#define HASH2_CALC hv = cur[0] | ((UInt32)cur[1] << 8); + +#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]] << 5)) & 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]] << 5); \ + h4 = temp & (kHash4Size - 1); \ + hv = (temp ^ (p->crc[cur[4]] << 3)) & p->hashMask; } + +/* #define HASH_ZIP_CALC hv = ((cur[0] | ((UInt32)cur[1] << 8)) ^ p->crc[cur[2]]) & 0xFFFF; */ +#define HASH_ZIP_CALC hv = ((cur[2] | ((UInt32)cur[0] << 8)) ^ p->crc[cur[1]]) & 0xFFFF; + + +#define MT_HASH2_CALC \ + h2 = (p->crc[cur[0]] ^ cur[1]) & (kHash2Size - 1); + +#define MT_HASH3_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + h3 = (temp ^ ((UInt32)cur[2] << 8)) & (kHash3Size - 1); } + +#define MT_HASH4_CALC { \ + UInt32 temp = p->crc[cur[0]] ^ cur[1]; \ + h2 = temp & (kHash2Size - 1); \ + temp ^= ((UInt32)cur[2] << 8); \ + h3 = temp & (kHash3Size - 1); \ + h4 = (temp ^ (p->crc[cur[3]] << 5)) & (kHash4Size - 1); } + +#endif diff --git a/include/lzma/Lzma2Dec.h b/include/lzma/Lzma2Dec.h new file mode 100644 index 0000000..e6a0f6e --- /dev/null +++ b/include/lzma/Lzma2Dec.h @@ -0,0 +1,80 @@ +/* Lzma2Dec.h -- LZMA2 Decoder +2015-05-13 : Igor Pavlov : Public domain */ + +#ifndef __LZMA2_DEC_H +#define __LZMA2_DEC_H + +#include "LzmaDec.h" + +EXTERN_C_BEGIN + +/* ---------- State Interface ---------- */ + +typedef struct +{ + CLzmaDec decoder; + UInt32 packSize; + UInt32 unpackSize; + unsigned state; + Byte control; + Bool needInitDic; + Bool needInitState; + Bool needInitProp; +} 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, ISzAlloc *alloc); +SRes Lzma2Dec_Allocate(CLzma2Dec *p, Byte prop, ISzAlloc *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); + + +/* ---------- 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, ISzAlloc *alloc); + +EXTERN_C_END + +#endif diff --git a/include/lzma/Lzma2Enc.h b/include/lzma/Lzma2Enc.h new file mode 100644 index 0000000..f409f18 --- /dev/null +++ b/include/lzma/Lzma2Enc.h @@ -0,0 +1,62 @@ +/* Lzma2Enc.h -- LZMA2 Encoder +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __LZMA2_ENC_H +#define __LZMA2_ENC_H + +#include "LzmaEnc.h" + +EXTERN_C_BEGIN + +typedef struct +{ + CLzmaEncProps lzmaProps; + size_t blockSize; + int numBlockThreads; + int numTotalThreads; +} CLzma2EncProps; + +void Lzma2EncProps_Init(CLzma2EncProps *p); +void Lzma2EncProps_Normalize(CLzma2EncProps *p); + +/* ---------- CLzmaEnc2Handle Interface ---------- */ + +/* Lzma2Enc_* functions can return the following exit codes: +Returns: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater in props + SZ_ERROR_WRITE - Write callback error + SZ_ERROR_PROGRESS - some break from progress callback + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) +*/ + +typedef void * CLzma2EncHandle; + +CLzma2EncHandle Lzma2Enc_Create(ISzAlloc *alloc, ISzAlloc *allocBig); +void Lzma2Enc_Destroy(CLzma2EncHandle p); +SRes Lzma2Enc_SetProps(CLzma2EncHandle p, const CLzma2EncProps *props); +Byte Lzma2Enc_WriteProperties(CLzma2EncHandle p); +SRes Lzma2Enc_Encode(CLzma2EncHandle p, + ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress); + +/* ---------- One Call Interface ---------- */ + +/* Lzma2Encode +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) +*/ + +/* +SRes Lzma2Encode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, int writeEndMark, + ICompressProgress *progress, ISzAlloc *alloc, ISzAlloc *allocBig); +*/ + +EXTERN_C_END + +#endif diff --git a/include/lzma/Lzma86.h b/include/lzma/Lzma86.h new file mode 100644 index 0000000..bebed5c --- /dev/null +++ b/include/lzma/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/include/lzma/LzmaDec.h b/include/lzma/LzmaDec.h new file mode 100644 index 0000000..cc44dae --- /dev/null +++ b/include/lzma/LzmaDec.h @@ -0,0 +1,227 @@ +/* LzmaDec.h -- LZMA Decoder +2013-01-18 : 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 */ + +#ifdef _LZMA_PROB32 +#define CLzmaProb UInt32 +#else +#define CLzmaProb UInt16 +#endif + + +/* ---------- LZMA Properties ---------- */ + +#define LZMA_PROPS_SIZE 5 + +typedef struct _CLzmaProps +{ + unsigned lc, lp, pb; + 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 +{ + CLzmaProps prop; + CLzmaProb *probs; + Byte *dic; + const Byte *buf; + UInt32 range, code; + SizeT dicPos; + SizeT dicBufSize; + UInt32 processedPos; + UInt32 checkDicSize; + unsigned state; + UInt32 reps[4]; + unsigned remainLen; + int needFlush; + int needInitState; + UInt32 numProbs; + unsigned tempBufSize; + Byte tempBuf[LZMA_REQUIRED_INPUT_MAX]; +} CLzmaDec; + +#define LzmaDec_Construct(p) { (p)->dic = 0; (p)->probs = 0; } + +void LzmaDec_Init(CLzmaDec *p); + +/* There are two types of LZMA streams: + 0) Stream with end mark. That end mark adds about 6 bytes to compressed size. + 1) 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, ISzAlloc *alloc); +void LzmaDec_FreeProbs(CLzmaDec *p, ISzAlloc *alloc); + +SRes LzmaDec_Allocate(CLzmaDec *state, const Byte *prop, unsigned propsSize, ISzAlloc *alloc); +void LzmaDec_Free(CLzmaDec *state, ISzAlloc *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_Constr() + 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 +*/ + +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). +*/ + +SRes LzmaDecode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + const Byte *propData, unsigned propSize, ELzmaFinishMode finishMode, + ELzmaStatus *status, ISzAlloc *alloc); + +EXTERN_C_END + +#endif diff --git a/include/lzma/LzmaEnc.h b/include/lzma/LzmaEnc.h new file mode 100644 index 0000000..cffe220 --- /dev/null +++ b/include/lzma/LzmaEnc.h @@ -0,0 +1,78 @@ +/* LzmaEnc.h -- LZMA Encoder +2013-01-18 : 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 <= (1 << 30) for 64-bit version + default = (1 << 24) */ + UInt64 reduceSize; /* estimated size of data that will be compressed. default = 0xFFFFFFFF. + Encoder uses this value to reduce dictionary size */ + 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 */ +} 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: +Returns: + SZ_OK - OK + SZ_ERROR_MEM - Memory allocation error + SZ_ERROR_PARAM - Incorrect paramater in props + SZ_ERROR_WRITE - Write callback error. + SZ_ERROR_PROGRESS - some break from progress callback + SZ_ERROR_THREAD - errors in multithreading functions (only for Mt version) +*/ + +typedef void * CLzmaEncHandle; + +CLzmaEncHandle LzmaEnc_Create(ISzAlloc *alloc); +void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAlloc *alloc, ISzAlloc *allocBig); +SRes LzmaEnc_SetProps(CLzmaEncHandle p, const CLzmaEncProps *props); +SRes LzmaEnc_WriteProperties(CLzmaEncHandle p, Byte *properties, SizeT *size); +SRes LzmaEnc_Encode(CLzmaEncHandle p, ISeqOutStream *outStream, ISeqInStream *inStream, + ICompressProgress *progress, ISzAlloc *alloc, ISzAlloc *allocBig); +SRes LzmaEnc_MemEncode(CLzmaEncHandle p, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + int writeEndMark, ICompressProgress *progress, ISzAlloc *alloc, ISzAlloc *allocBig); + +/* ---------- One Call Interface ---------- */ + +/* LzmaEncode +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) +*/ + +SRes LzmaEncode(Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + const CLzmaEncProps *props, Byte *propsEncoded, SizeT *propsSize, int writeEndMark, + ICompressProgress *progress, ISzAlloc *alloc, ISzAlloc *allocBig); + +EXTERN_C_END + +#endif diff --git a/include/lzma/LzmaLib.h b/include/lzma/LzmaLib.h new file mode 100644 index 0000000..88fa87d --- /dev/null +++ b/include/lzma/LzmaLib.h @@ -0,0 +1,131 @@ +/* LzmaLib.h -- LZMA library interface +2013-01-18 : 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: 16 KB 0 32 + 1: 64 KB 0 32 + 2: 256 KB 0 32 + 3: 1 MB 0 32 + 4: 4 MB 0 32 + 5: 16 MB 1 32 + 6: 32 MB 1 32 + 7+: 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. + +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 + destLen - output data 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/include/lzma/MtCoder.h b/include/lzma/MtCoder.h new file mode 100644 index 0000000..f0f06da --- /dev/null +++ b/include/lzma/MtCoder.h @@ -0,0 +1,98 @@ +/* MtCoder.h -- Multi-thread Coder +2009-11-19 : Igor Pavlov : Public domain */ + +#ifndef __MT_CODER_H +#define __MT_CODER_H + +#include "Threads.h" + +EXTERN_C_BEGIN + +typedef struct +{ + CThread thread; + CAutoResetEvent startEvent; + CAutoResetEvent finishedEvent; + int stop; + + THREAD_FUNC_TYPE func; + LPVOID param; + THREAD_FUNC_RET_TYPE res; +} CLoopThread; + +void LoopThread_Construct(CLoopThread *p); +void LoopThread_Close(CLoopThread *p); +WRes LoopThread_Create(CLoopThread *p); +WRes LoopThread_StopAndWait(CLoopThread *p); +WRes LoopThread_StartSubThread(CLoopThread *p); +WRes LoopThread_WaitSubThread(CLoopThread *p); + +#ifndef _7ZIP_ST +#define NUM_MT_CODER_THREADS_MAX 32 +#else +#define NUM_MT_CODER_THREADS_MAX 1 +#endif + +typedef struct +{ + UInt64 totalInSize; + UInt64 totalOutSize; + ICompressProgress *progress; + SRes res; + CCriticalSection cs; + UInt64 inSizes[NUM_MT_CODER_THREADS_MAX]; + UInt64 outSizes[NUM_MT_CODER_THREADS_MAX]; +} CMtProgress; + +SRes MtProgress_Set(CMtProgress *p, unsigned index, UInt64 inSize, UInt64 outSize); + +struct _CMtCoder; + +typedef struct +{ + struct _CMtCoder *mtCoder; + Byte *outBuf; + size_t outBufSize; + Byte *inBuf; + size_t inBufSize; + unsigned index; + CLoopThread thread; + + Bool stopReading; + Bool stopWriting; + CAutoResetEvent canRead; + CAutoResetEvent canWrite; +} CMtThread; + +typedef struct +{ + SRes (*Code)(void *p, unsigned index, Byte *dest, size_t *destSize, + const Byte *src, size_t srcSize, int finished); +} IMtCoderCallback; + +typedef struct _CMtCoder +{ + size_t blockSize; + size_t destBlockSize; + unsigned numThreads; + + ISeqInStream *inStream; + ISeqOutStream *outStream; + ICompressProgress *progress; + ISzAlloc *alloc; + + IMtCoderCallback *mtCallback; + CCriticalSection cs; + SRes res; + + CMtProgress mtProgress; + CMtThread threads[NUM_MT_CODER_THREADS_MAX]; +} CMtCoder; + +void MtCoder_Construct(CMtCoder* p); +void MtCoder_Destruct(CMtCoder* p); +SRes MtCoder_Code(CMtCoder *p); + +EXTERN_C_END + +#endif diff --git a/include/lzma/Ppmd.h b/include/lzma/Ppmd.h new file mode 100644 index 0000000..5655b26 --- /dev/null +++ b/include/lzma/Ppmd.h @@ -0,0 +1,85 @@ +/* Ppmd.h -- PPMD codec common code +2016-05-16 : 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 + +#ifdef MY_CPU_32BIT + #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) + +#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 <<= 1; (p)->Count = (Byte)(3 << (p)->Shift++); } + +typedef struct +{ + Byte Symbol; + Byte Freq; + UInt16 SuccessorLow; + UInt16 SuccessorHigh; +} CPpmd_State; + +#pragma pack(pop) + +typedef + #ifdef PPMD_32BIT + CPpmd_State * + #else + UInt32 + #endif + CPpmd_State_Ref; + +typedef + #ifdef PPMD_32BIT + void * + #else + UInt32 + #endif + CPpmd_Void_Ref; + +typedef + #ifdef PPMD_32BIT + Byte * + #else + UInt32 + #endif + CPpmd_Byte_Ref; + +#define PPMD_SetAllBitsIn256Bytes(p) \ + { unsigned 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/include/lzma/Ppmd7.h b/include/lzma/Ppmd7.h new file mode 100644 index 0000000..87eefde --- /dev/null +++ b/include/lzma/Ppmd7.h @@ -0,0 +1,140 @@ +/* Ppmd7.h -- PPMdH compression codec +2016-05-21 : Igor Pavlov : Public domain +This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ + +/* This code supports virtual RangeDecoder and includes the implementation +of RangeCoder from 7z, instead of RangeCoder from original PPMd var.H. +If you need the compatibility with original PPMd var.H, you can use external RangeDecoder */ + +#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 + #ifdef PPMD_32BIT + struct CPpmd7_Context_ * + #else + UInt32 + #endif + CPpmd7_Context_Ref; + +typedef struct CPpmd7_Context_ +{ + UInt16 NumStats; + UInt16 SummFreq; + CPpmd_State_Ref Stats; + CPpmd7_Context_Ref Suffix; +} CPpmd7_Context; + +#define Ppmd7Context_OneState(p) ((CPpmd_State *)&(p)->SummFreq) + +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; + Byte *Base, *LoUnit, *HiUnit, *Text, *UnitsStart; + UInt32 AlignOffset; + + Byte Indx2Units[PPMD_NUM_INDEXES]; + Byte Units2Indx[128]; + CPpmd_Void_Ref FreeList[PPMD_NUM_INDEXES]; + Byte NS2Indx[256], NS2BSIndx[256], HB2Flag[256]; + CPpmd_See DummySee, See[25][16]; + UInt16 BinSumm[128][64]; +} CPpmd7; + +void Ppmd7_Construct(CPpmd7 *p); +Bool Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAlloc *alloc); +void Ppmd7_Free(CPpmd7 *p, ISzAlloc *alloc); +void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder); +#define Ppmd7_WasAllocated(p) ((p)->Base != NULL) + + +/* ---------- Internal Functions ---------- */ + +extern const Byte PPMD7_kExpEscape[16]; + +#ifdef PPMD_32BIT + #define Ppmd7_GetPtr(p, ptr) (ptr) + #define Ppmd7_GetContext(p, ptr) (ptr) + #define Ppmd7_GetStats(p, ctx) ((ctx)->Stats) +#else + #define Ppmd7_GetPtr(p, offs) ((void *)((p)->Base + (offs))) + #define Ppmd7_GetContext(p, offs) ((CPpmd7_Context *)Ppmd7_GetPtr((p), (offs))) + #define Ppmd7_GetStats(p, ctx) ((CPpmd_State *)Ppmd7_GetPtr((p), ((ctx)->Stats))) +#endif + +void Ppmd7_Update1(CPpmd7 *p); +void Ppmd7_Update1_0(CPpmd7 *p); +void Ppmd7_Update2(CPpmd7 *p); +void Ppmd7_UpdateBin(CPpmd7 *p); + +#define Ppmd7_GetBinSumm(p) \ + &p->BinSumm[(unsigned)Ppmd7Context_OneState(p->MinContext)->Freq - 1][p->PrevSuccess + \ + p->NS2BSIndx[Ppmd7_GetContext(p, p->MinContext->Suffix)->NumStats - 1] + \ + (p->HiBitsFlag = p->HB2Flag[p->FoundState->Symbol]) + \ + 2 * p->HB2Flag[(unsigned)Ppmd7Context_OneState(p->MinContext)->Symbol] + \ + ((p->RunLength >> 26) & 0x20)] + +CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *scale); + + +/* ---------- Decode ---------- */ + +typedef struct +{ + UInt32 (*GetThreshold)(void *p, UInt32 total); + void (*Decode)(void *p, UInt32 start, UInt32 size); + UInt32 (*DecodeBit)(void *p, UInt32 size0); +} IPpmd7_RangeDec; + +typedef struct +{ + IPpmd7_RangeDec p; + UInt32 Range; + UInt32 Code; + IByteIn *Stream; +} CPpmd7z_RangeDec; + +void Ppmd7z_RangeDec_CreateVTable(CPpmd7z_RangeDec *p); +Bool Ppmd7z_RangeDec_Init(CPpmd7z_RangeDec *p); +#define Ppmd7z_RangeDec_IsFinishedOK(p) ((p)->Code == 0) + +int Ppmd7_DecodeSymbol(CPpmd7 *p, IPpmd7_RangeDec *rc); + + +/* ---------- Encode ---------- */ + +typedef struct +{ + UInt64 Low; + UInt32 Range; + Byte Cache; + UInt64 CacheSize; + IByteOut *Stream; +} CPpmd7z_RangeEnc; + +void Ppmd7z_RangeEnc_Init(CPpmd7z_RangeEnc *p); +void Ppmd7z_RangeEnc_FlushData(CPpmd7z_RangeEnc *p); + +void Ppmd7_EncodeSymbol(CPpmd7 *p, CPpmd7z_RangeEnc *rc, int symbol); + +EXTERN_C_END + +#endif diff --git a/include/lzma/Precomp.h b/include/lzma/Precomp.h new file mode 100644 index 0000000..e8ff8b4 --- /dev/null +++ b/include/lzma/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/include/lzma/RotateDefs.h b/include/lzma/RotateDefs.h new file mode 100644 index 0000000..8f01d1a --- /dev/null +++ b/include/lzma/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/include/lzma/Sha256.h b/include/lzma/Sha256.h new file mode 100644 index 0000000..3f455db --- /dev/null +++ b/include/lzma/Sha256.h @@ -0,0 +1,26 @@ +/* Sha256.h -- SHA-256 Hash +2013-01-18 : Igor Pavlov : Public domain */ + +#ifndef __CRYPTO_SHA256_H +#define __CRYPTO_SHA256_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +#define SHA256_DIGEST_SIZE 32 + +typedef struct +{ + UInt32 state[8]; + UInt64 count; + Byte buffer[64]; +} CSha256; + +void Sha256_Init(CSha256 *p); +void Sha256_Update(CSha256 *p, const Byte *data, size_t size); +void Sha256_Final(CSha256 *p, Byte *digest); + +EXTERN_C_END + +#endif diff --git a/include/lzma/Sort.h b/include/lzma/Sort.h new file mode 100644 index 0000000..2e2963a --- /dev/null +++ b/include/lzma/Sort.h @@ -0,0 +1,18 @@ +/* Sort.h -- Sort functions +2014-04-05 : Igor Pavlov : Public domain */ + +#ifndef __7Z_SORT_H +#define __7Z_SORT_H + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +void HeapSort(UInt32 *p, size_t size); +void HeapSort64(UInt64 *p, size_t size); + +/* void HeapSortRef(UInt32 *p, UInt32 *vals, size_t size); */ + +EXTERN_C_END + +#endif diff --git a/include/lzma/Threads.h b/include/lzma/Threads.h new file mode 100644 index 0000000..9b3e1c5 --- /dev/null +++ b/include/lzma/Threads.h @@ -0,0 +1,67 @@ +/* Threads.h -- multithreading library +2013-11-12 : Igor Pavlov : Public domain */ + +#ifndef __7Z_THREADS_H +#define __7Z_THREADS_H + +#ifdef _WIN32 +#include +#endif + +#include "7zTypes.h" + +EXTERN_C_BEGIN + +WRes HandlePtr_Close(HANDLE *h); +WRes Handle_WaitObject(HANDLE h); + +typedef HANDLE CThread; +#define Thread_Construct(p) *(p) = NULL +#define Thread_WasCreated(p) (*(p) != NULL) +#define Thread_Close(p) HandlePtr_Close(p) +#define Thread_Wait(p) Handle_WaitObject(*(p)) + +typedef +#ifdef UNDER_CE + DWORD +#else + unsigned +#endif + THREAD_FUNC_RET_TYPE; + +#define THREAD_FUNC_CALL_TYPE MY_STD_CALL +#define THREAD_FUNC_DECL THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE +typedef THREAD_FUNC_RET_TYPE (THREAD_FUNC_CALL_TYPE * THREAD_FUNC_TYPE)(void *); +WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param); + +typedef HANDLE CEvent; +typedef CEvent CAutoResetEvent; +typedef CEvent CManualResetEvent; +#define Event_Construct(p) *(p) = NULL +#define Event_IsCreated(p) (*(p) != NULL) +#define Event_Close(p) HandlePtr_Close(p) +#define Event_Wait(p) Handle_WaitObject(*(p)) +WRes Event_Set(CEvent *p); +WRes Event_Reset(CEvent *p); +WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled); +WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p); +WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled); +WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p); + +typedef HANDLE CSemaphore; +#define Semaphore_Construct(p) (*p) = NULL +#define Semaphore_Close(p) HandlePtr_Close(p) +#define Semaphore_Wait(p) Handle_WaitObject(*(p)) +WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount); +WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num); +WRes Semaphore_Release1(CSemaphore *p); + +typedef CRITICAL_SECTION CCriticalSection; +WRes CriticalSection_Init(CCriticalSection *p); +#define CriticalSection_Delete(p) DeleteCriticalSection(p) +#define CriticalSection_Enter(p) EnterCriticalSection(p) +#define CriticalSection_Leave(p) LeaveCriticalSection(p) + +EXTERN_C_END + +#endif diff --git a/include/lzma/Xz.h b/include/lzma/Xz.h new file mode 100644 index 0000000..be3a1c3 --- /dev/null +++ b/include/lzma/Xz.h @@ -0,0 +1,275 @@ +/* Xz.h - Xz interface +2015-05-01 : 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) (((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) + +SRes XzBlock_Parse(CXzBlock *p, const Byte *header); +SRes XzBlock_ReadHeader(CXzBlock *p, ISeqInStream *inStream, Bool *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_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; + size_t numBlocksAllocated; + CXzBlockSizes *blocks; + UInt64 startOffset; +} CXzStream; + +void Xz_Construct(CXzStream *p); +void Xz_Free(CXzStream *p, ISzAlloc *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, ISzAlloc *alloc); +SRes Xzs_ReadBackward(CXzs *p, ILookInStream *inStream, Int64 *startOffset, ICompressProgress *progress, ISzAlloc *alloc); + +UInt64 Xzs_GetNumBlocks(const CXzs *p); +UInt64 Xzs_GetUnpackSize(const CXzs *p); + +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; + +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, ISzAlloc *alloc); + SRes (*SetProps)(void *p, const Byte *props, size_t propSize, ISzAlloc *alloc); + void (*Init)(void *p); + SRes (*Code)(void *p, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, int *wasFinished); +} IStateCoder; + +#define MIXCODER_NUM_FILTERS_MAX 4 + +typedef struct +{ + ISzAlloc *alloc; + Byte *buf; + unsigned numCoders; + 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]; + IStateCoder coders[MIXCODER_NUM_FILTERS_MAX]; +} CMixCoder; + +void MixCoder_Construct(CMixCoder *p, ISzAlloc *alloc); +void MixCoder_Free(CMixCoder *p); +void MixCoder_Init(CMixCoder *p); +SRes MixCoder_SetFromMethod(CMixCoder *p, unsigned coderIndex, UInt64 methodId); +SRes MixCoder_Code(CMixCoder *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, int srcWasFinished, + ECoderFinishMode finishMode, ECoderStatus *status); + +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; + UInt64 indexSize; + UInt64 indexPos; + UInt64 padSize; + + UInt64 numStartedStreams; + UInt64 numFinishedStreams; + UInt64 numTotalBlocks; + + UInt32 crc; + CMixCoder decoder; + CXzBlock block; + CXzCheck check; + CSha256 sha; + Byte shaDigest[SHA256_DIGEST_SIZE]; + Byte buf[XZ_BLOCK_HEADER_SIZE_MAX]; +} CXzUnpacker; + +void XzUnpacker_Construct(CXzUnpacker *p, ISzAlloc *alloc); +void XzUnpacker_Init(CXzUnpacker *p); +void XzUnpacker_Free(CXzUnpacker *p); + +/* +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 - maybe there are more xz streams, + call XzUnpacker_IsStreamWasFinished to check that current stream was finished + 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, ECoderFinishMode finishMode, + ECoderStatus *status); + +Bool XzUnpacker_IsStreamWasFinished(CXzUnpacker *p); + +/* +Call XzUnpacker_GetExtraSize after XzUnpacker_Code function to detect real size of +xz stream in two cases: +XzUnpacker_Code() returns: + res == SZ_OK && status == CODER_STATUS_NEEDS_MORE_INPUT + res == SZ_ERROR_NO_ARCHIVE +*/ + +UInt64 XzUnpacker_GetExtraSize(CXzUnpacker *p); + +EXTERN_C_END + +#endif diff --git a/include/lzma/XzCrc64.h b/include/lzma/XzCrc64.h new file mode 100644 index 0000000..08dbc33 --- /dev/null +++ b/include/lzma/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/include/lzma/XzEnc.h b/include/lzma/XzEnc.h new file mode 100644 index 0000000..c3c19ec --- /dev/null +++ b/include/lzma/XzEnc.h @@ -0,0 +1,39 @@ +/* XzEnc.h -- Xz Encode +2011-02-07 : Igor Pavlov : Public domain */ + +#ifndef __XZ_ENC_H +#define __XZ_ENC_H + +#include "Lzma2Enc.h" + +#include "Xz.h" + +EXTERN_C_BEGIN + +typedef struct +{ + UInt32 id; + UInt32 delta; + UInt32 ip; + int ipDefined; +} CXzFilterProps; + +void XzFilterProps_Init(CXzFilterProps *p); + +typedef struct +{ + const CLzma2EncProps *lzma2Props; + const CXzFilterProps *filterProps; + unsigned checkId; +} CXzProps; + +void XzProps_Init(CXzProps *p); + +SRes Xz_Encode(ISeqOutStream *outStream, ISeqInStream *inStream, + const CXzProps *props, ICompressProgress *progress); + +SRes Xz_EncodeEmpty(ISeqOutStream *outStream); + +EXTERN_C_END + +#endif diff --git a/include/lzma/delta.h b/include/lzma/delta.h new file mode 100644 index 0000000..2fa54ad --- /dev/null +++ b/include/lzma/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/resource/symbols.def b/resource/symbols.def index d6f29b3..7b87306 100644 --- a/resource/symbols.def +++ b/resource/symbols.def @@ -4,5 +4,7 @@ LIBRARY "tbfix" EXPORTS -SKPlugIn_Init = SKPlugIn_Init @1 +SKPlugIn_Init = SKPlugIn_Init @1 +SKPlugIn_Shutdown = SKPlugIn_Shutdown @2 + ;TBT_Main = tbt_main @2 \ No newline at end of file diff --git a/src/config.cpp b/src/config.cpp index 7496617..9555edc 100644 --- a/src/config.cpp +++ b/src/config.cpp @@ -72,6 +72,7 @@ struct { tbf::ParameterInt* rescale_env_shadows; tbf::ParameterBool* dump_shaders; + tbf::ParameterBool* fix_map_res; } render; struct { @@ -315,6 +316,16 @@ TBF_LoadConfig (std::wstring name) L"Shadow.Quality", L"RescaleEnvShadows" ); + render.fix_map_res = + static_cast + (g_ParameterFactory.create_parameter ( + L"Fix crazy low-resolution map menu...") + ); + render.fix_map_res->register_to_ini ( + render_ini, + L"Resolution.Experimental", + L"FixMap" ); + framerate.replace_limiter = static_cast (g_ParameterFactory.create_parameter ( @@ -426,6 +437,7 @@ TBF_LoadConfig (std::wstring name) render.rescale_env_shadows->load (config.render.env_shadow_rescale); render.dump_shaders->load (config.render.dump_shaders); + render.fix_map_res->load (config.render.fix_map_res); framerate.replace_limiter->load (config.framerate.replace_limiter); framerate.tolerance->load (config.framerate.tolerance); @@ -463,6 +475,7 @@ TBF_SaveConfig (std::wstring name, bool close_config) render.rescale_shadows->store (config.render.shadow_rescale); render.rescale_env_shadows->store (config.render.env_shadow_rescale); + render.fix_map_res->store (config.render.fix_map_res); textures.remaster->store (config.textures.remaster); textures.uncompressed->store (config.textures.uncompressed); diff --git a/src/dllmain.cpp b/src/dllmain.cpp index c57df89..4efe829 100644 --- a/src/dllmain.cpp +++ b/src/dllmain.cpp @@ -50,6 +50,41 @@ HMODULE hInjectorDLL = { 0 }; // Handle to Special K SKX_SetPluginName_pfn SKX_SetPluginName = nullptr; + + +__declspec (dllexport) +BOOL +WINAPI +SKPlugIn_Shutdown (LPVOID* lpReserved) +{ + UNREFERENCED_PARAMETER (lpReserved); + + if (dll_log != nullptr) + { + tbf::SoundFix::Shutdown (); + //tbf::FileIO::Shutdown (); + //tbf::SteamFix::Shutdown (); + tbf::FrameRateFix::Shutdown (); + tbf::RenderFix::Shutdown (); + //tbf::KeyboardFix::Shutdown (); + + TBF_SaveConfig (); + TBF_UnInit_MinHook (); + + + dll_log->LogEx ( false, L"=========== (Version: v %s) " + L"===========\n", + TBF_VER_STR.c_str () ); + dll_log->LogEx ( true, L"End TBFix Plug-In\n" ); + dll_log->LogEx ( false, L"------- [Tales of Berseria \"Fix\"] " + L"-------\n" ); + + dll_log->close (); + } + + return TRUE; +} + __declspec (dllexport) BOOL WINAPI @@ -197,28 +232,6 @@ DllMain (HMODULE hModule, case DLL_PROCESS_DETACH: { - if (dll_log != nullptr) - { - tbf::SoundFix::Shutdown (); - //tbf::FileIO::Shutdown (); - //tbf::SteamFix::Shutdown (); - tbf::RenderFix::Shutdown (); - tbf::FrameRateFix::Shutdown (); - //tbf::KeyboardFix::Shutdown (); - - TBF_UnInit_MinHook (); - TBF_SaveConfig (); - - - dll_log->LogEx ( false, L"=========== (Version: v %s) " - L"===========\n", - TBF_VER_STR.c_str () ); - dll_log->LogEx ( true, L"End TBFix Plug-In\n" ); - dll_log->LogEx ( false, L"------- [Tales of Berseria \"Fix\"] " - L"-------\n" ); - - dll_log->close (); - } } break; } diff --git a/src/framerate.cpp b/src/framerate.cpp index fabd390..b1bad19 100644 --- a/src/framerate.cpp +++ b/src/framerate.cpp @@ -77,10 +77,6 @@ extern "C" unsigned int hde64_disasm(const void *code, hde64s *hs); -// -// @TODO: Develop a heuristic to scan for this memory address; -// hardcoding it is going to break stuff :) -// volatile uintptr_t TICK_ADDR_BASE = 0x0UL; int32_t tbf::FrameRateFix::tick_scale = INT32_MAX; // 0 FPS diff --git a/src/lzma/7zAlloc.c b/src/lzma/7zAlloc.c new file mode 100644 index 0000000..83c5c9f --- /dev/null +++ b/src/lzma/7zAlloc.c @@ -0,0 +1,78 @@ +/* 7zAlloc.c -- Allocation functions +2015-11-09 : Igor Pavlov : Public domain */ + +#include + +#include + +//#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(void *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(void *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(void *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(void *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/src/lzma/7zArcIn.c b/src/lzma/7zArcIn.c new file mode 100644 index 0000000..eed47bc --- /dev/null +++ b/src/lzma/7zArcIn.c @@ -0,0 +1,1773 @@ +/* 7zArcIn.c -- 7z Input functions +2016-05-16 : Igor Pavlov : Public domain */ + +#include + +#include + +#include +#include +#include +#include + +#define MY_ALLOC(T, p, size, alloc) { \ + if ((p = (T *)IAlloc_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) p = 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, ISzAlloc *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; +} + +void SzBitUi32s_Free(CSzBitUi32s *p, ISzAlloc *alloc) +{ + IAlloc_Free(alloc, p->Defs); p->Defs = NULL; + IAlloc_Free(alloc, p->Vals); p->Vals = NULL; +} + +#define SzBitUi64s_Init(p) { (p)->Defs = NULL; (p)->Vals = NULL; } + +void SzBitUi64s_Free(CSzBitUi64s *p, ISzAlloc *alloc) +{ + IAlloc_Free(alloc, p->Defs); p->Defs = NULL; + IAlloc_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; +} + +static void SzAr_Free(CSzAr *p, ISzAlloc *alloc) +{ + IAlloc_Free(alloc, p->PackPositions); + SzBitUi32s_Free(&p->FolderCRCs, alloc); + + IAlloc_Free(alloc, p->FoCodersOffsets); + IAlloc_Free(alloc, p->FoStartPackStreamIndex); + IAlloc_Free(alloc, p->FoToCoderUnpackSizes); + IAlloc_Free(alloc, p->FoToMainUnpackSizeIndex); + IAlloc_Free(alloc, p->CoderUnpackSizes); + + IAlloc_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, ISzAlloc *alloc) +{ + IAlloc_Free(alloc, p->UnpackPositions); + IAlloc_Free(alloc, p->IsDirs); + + IAlloc_Free(alloc, p->FolderToFile); + IAlloc_Free(alloc, p->FileToFolder); + + IAlloc_Free(alloc, p->FileNameOffsets); + IAlloc_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, ISzAlloc *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[numBytes - 1] = (Byte)((((UInt32)1 << numBits) - 1) << (8 - numBits)); + } + return SZ_OK; +} + +static MY_NO_INLINE SRes ReadUi32s(CSzData *sd2, UInt32 numItems, CSzBitUi32s *crcs, ISzAlloc *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, ISzAlloc *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, ISzAlloc *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 = 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, + ISzAlloc *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(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] = 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; + + { + size_t dataSize = 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 = 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 = 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 = 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, + ISzAlloc *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)); + RINOK(ReadPackInfo(p, sd, alloc)); + 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, + ISzAlloc *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 *)p != 0 + #else + p[0] != 0 || p[1] != 0 + #endif + ; p += 2); + pos = 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, + ISzAlloc *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, + ISzAlloc *allocMain, + ISzAlloc *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); + 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, + ISzAlloc *allocMain, + ISzAlloc *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, + ISzAlloc *allocMain, + ISzAlloc *allocTemp) +{ + Byte header[k7zStartHeaderSize]; + Int64 startArcPos; + UInt64 nextHeaderOffset, nextHeaderSize; + size_t nextHeaderSizeT; + UInt32 nextHeaderCRC; + CBuf buf; + SRes res; + + startArcPos = 0; + RINOK(inStream->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 = startArcPos + k7zStartHeaderSize; + + if (CrcCalc(header + 12, 20) != GetUi32(header + 8)) + return SZ_ERROR_CRC; + + 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(inStream->Seek(inStream, &pos, SZ_SEEK_END)); + if ((UInt64)pos < startArcPos + nextHeaderOffset || + (UInt64)pos < startArcPos + k7zStartHeaderSize + nextHeaderOffset || + (UInt64)pos < startArcPos + k7zStartHeaderSize + nextHeaderOffset + nextHeaderSize) + return SZ_ERROR_INPUT_EOF; + } + + RINOK(LookInStream_SeekTo(inStream, 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); + 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, + ISzAlloc *allocMain, ISzAlloc *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, + ISzAlloc *allocMain, + ISzAlloc *allocTemp) +{ + UInt32 folderIndex = p->FileToFolder[fileIndex]; + SRes res = SZ_OK; + + *offset = 0; + *outSizeProcessed = 0; + + if (folderIndex == (UInt32)-1) + { + IAlloc_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[folderIndex + 1]] - + p->UnpackPositions[p->FolderToFile[folderIndex]]; + */ + size_t unpackSize = (size_t)unpackSizeSpec; + + if (unpackSize != unpackSizeSpec) + return SZ_ERROR_MEM; +#if 0 + *blockIndex = folderIndex; + IAlloc_Free(allocMain, *tempBuf); + *tempBuf = NULL; +#endif + if (res == SZ_OK) + { +#if 0 + *outBufferSize = unpackSize; + if (unpackSize != 0) + { + *tempBuf = (Byte *)IAlloc_Alloc(allocMain, unpackSize); + if (*tempBuf == NULL) + res = SZ_ERROR_MEM; + } +#endif + 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[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) +{ + Bool 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/src/lzma/7zBuf.c b/src/lzma/7zBuf.c new file mode 100644 index 0000000..15fcd64 --- /dev/null +++ b/src/lzma/7zBuf.c @@ -0,0 +1,36 @@ +/* 7zBuf.c -- Byte Buffer +2013-01-21 : Igor Pavlov : Public domain */ + +#include + +#include + +void Buf_Init(CBuf *p) +{ + p->data = 0; + p->size = 0; +} + +int Buf_Create(CBuf *p, size_t size, ISzAlloc *alloc) +{ + p->size = 0; + if (size == 0) + { + p->data = 0; + return 1; + } + p->data = (Byte *)alloc->Alloc(alloc, size); + if (p->data != 0) + { + p->size = size; + return 1; + } + return 0; +} + +void Buf_Free(CBuf *p, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->data); + p->data = 0; + p->size = 0; +} diff --git a/src/lzma/7zBuf2.c b/src/lzma/7zBuf2.c new file mode 100644 index 0000000..c234cb7 --- /dev/null +++ b/src/lzma/7zBuf2.c @@ -0,0 +1,51 @@ +/* 7zBuf2.c -- Byte Buffer +2014-08-22 : Igor Pavlov : Public domain */ + +#include + +#include + +#include + +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, ISzAlloc *alloc) +{ + if (size > p->size - p->pos) + { + size_t newSize = p->pos + size; + Byte *data; + newSize += newSize / 4; + data = (Byte *)alloc->Alloc(alloc, newSize); + if (data == 0) + return 0; + p->size = newSize; + memcpy(data, p->data, p->pos); + alloc->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, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->data); + p->data = 0; + p->size = 0; + p->pos = 0; +} diff --git a/src/lzma/7zCrc.c b/src/lzma/7zCrc.c new file mode 100644 index 0000000..89dc569 --- /dev/null +++ b/src/lzma/7zCrc.c @@ -0,0 +1,128 @@ +/* 7zCrc.c -- CRC32 init +2015-03-10 : Igor Pavlov : Public domain */ + +#include + +#include +#include + +#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); + +CRC_FUNC g_CrcUpdateT4; +CRC_FUNC g_CrcUpdateT8; +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) +{ + const Byte *p = (const Byte *)data; + const Byte *pEnd = p + size; + for (; p != pEnd; p++) + v = CRC_UPDATE_BYTE_2(v, *p); + return v; +} + +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 & ~((r & 1) - 1)); + g_CrcTable[i] = r; + } + for (; i < 256 * CRC_NUM_TABLES; i++) + { + UInt32 r = g_CrcTable[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()) + g_CrcUpdate = CrcUpdateT8; + #endif + #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[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 +} diff --git a/src/lzma/7zCrcOpt.c b/src/lzma/7zCrcOpt.c new file mode 100644 index 0000000..f7ab7e8 --- /dev/null +++ b/src/lzma/7zCrcOpt.c @@ -0,0 +1,114 @@ +/* 7zCrcOpt.c -- CRC32 calculation +2015-03-01 : Igor Pavlov : Public domain */ + +#include +#include + +#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) +{ + 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 *)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) +{ + 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 *)p; + v = + table[0x700 + ((v ) & 0xFF)] + ^ table[0x600 + ((v >> 8) & 0xFF)] + ^ table[0x500 + ((v >> 16) & 0xFF)] + ^ table[0x400 + ((v >> 24))]; + d = *((const UInt32 *)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 *)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 *)p; + v = + table[0x400 + ((v ) & 0xFF)] + ^ table[0x500 + ((v >> 8) & 0xFF)] + ^ table[0x600 + ((v >> 16) & 0xFF)] + ^ table[0x700 + ((v >> 24))]; + d = *((const UInt32 *)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/src/lzma/7zDec.c b/src/lzma/7zDec.c new file mode 100644 index 0000000..ed69d97 --- /dev/null +++ b/src/lzma/7zDec.c @@ -0,0 +1,591 @@ +/* 7zDec.c -- Decoding from 7z folder +2015-11-18 : Igor Pavlov : Public domain */ + +#include + +#include + +/* #define _7ZIP_PPMD_SUPPPORT */ + +#include +#include + +#include +#include +#include +#include +#include +#include +#ifdef _7ZIP_PPMD_SUPPPORT +#include +#endif + +#define k_Copy 0 +#define k_Delta 3 +#define k_LZMA2 0x21 +#define k_LZMA 0x30101 +#define k_BCJ 0x3030103 +#define k_BCJ2 0x303011B +#define k_PPC 0x3030205 +#define k_IA64 0x3030401 +#define k_ARM 0x3030501 +#define k_ARMT 0x3030701 +#define k_SPARC 0x3030805 + + +#ifdef _7ZIP_PPMD_SUPPPORT + +#define k_PPMD 0x30401 + +typedef struct +{ + IByteIn p; + const Byte *cur; + const Byte *end; + const Byte *begin; + UInt64 processed; + Bool extra; + SRes res; + ILookInStream *inStream; +} CByteInToLook; + +static Byte ReadByte(void *pp) +{ + CByteInToLook *p = (CByteInToLook *)pp; + if (p->cur != p->end) + return *p->cur++; + if (p->res == SZ_OK) + { + size_t size = p->cur - p->begin; + p->processed += size; + p->res = p->inStream->Skip(p->inStream, size); + size = (1 << 25); + p->res = p->inStream->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, ILookInStream *inStream, + Byte *outBuffer, SizeT outSize, ISzAlloc *allocMain) +{ + CPpmd7 ppmd; + CByteInToLook s; + SRes res = SZ_OK; + + s.p.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); + } + { + CPpmd7z_RangeDec rc; + Ppmd7z_RangeDec_CreateVTable(&rc); + rc.Stream = &s.p; + if (!Ppmd7z_RangeDec_Init(&rc)) + res = SZ_ERROR_DATA; + else if (s.extra) + res = (s.res != SZ_OK ? s.res : SZ_ERROR_DATA); + else + { + SizeT i; + for (i = 0; i < outSize; i++) + { + int sym = Ppmd7_DecodeSymbol(&ppmd, &rc.p); + if (s.extra || sym < 0) + break; + outBuffer[i] = (Byte)sym; + } + if (i != outSize) + res = (s.res != SZ_OK ? s.res : SZ_ERROR_DATA); + else if (s.processed + (s.cur - s.begin) != inSize || !Ppmd7z_RangeDec_IsFinishedOK(&rc)) + 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, ISzAlloc *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 = inStream->Look(inStream, &inBuf, &lookahead); + if (res != SZ_OK) + break; + + { + SizeT inProcessed = (SizeT)lookahead, dicPos = state.dicPos; + ELzmaStatus status; + res = LzmaDec_DecodeToDic(&state, outSize, 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 = inStream->Skip((void *)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, ISzAlloc *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 = inStream->Look(inStream, &inBuf, &lookahead); + if (res != SZ_OK) + break; + + { + SizeT inProcessed = (SizeT)lookahead, dicPos = state.decoder.dicPos; + ELzmaStatus status; + res = Lzma2Dec_DecodeToDic(&state, outSize, 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 = inStream->Skip((void *)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(inStream->Look(inStream, &inBuf, &curSize)); + if (curSize == 0) + return SZ_ERROR_INPUT_EOF; + memcpy(outBuffer, inBuf, curSize); + outBuffer += curSize; + inSize -= curSize; + RINOK(inStream->Skip((void *)inStream, curSize)); + } + return SZ_OK; +} + +static Bool 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 Bool 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; +} + +#define CASE_BRA_CONV(isa) case k_ ## isa: isa ## _Convert(outBuffer, outSize, 0, 0); break; + +static SRes SzFolder_Decode2(const CSzFolder *folder, + const Byte *propsData, + const UInt64 *unpackSizes, + const UInt64 *packPositions, + ILookInStream *inStream, UInt64 startPos, + Byte *outBuffer, SizeT outSize, ISzAlloc *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 *)IAlloc_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[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 *)IAlloc_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, + ISzAlloc *allocMain) +{ + SRes res; + CSzFolder folder; + CSzData sd; + + const Byte *data = p->CodersData + p->FoCodersOffsets[folderIndex]; + sd.Data = data; + sd.Size = p->FoCodersOffsets[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++) + IAlloc_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/src/lzma/7zFile.c b/src/lzma/7zFile.c new file mode 100644 index 0000000..65c112b --- /dev/null +++ b/src/lzma/7zFile.c @@ -0,0 +1,286 @@ +/* 7zFile.c -- File IO +2009-11-24 : Igor Pavlov : Public domain */ + +#include + +#include + +#ifndef USE_WINDOWS_FILE + +#ifndef UNDER_CE +#include +#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" +*/ + +#define kChunkSizeMax (1 << 22) + +#endif + +void File_Construct(CSzFile *p) +{ + #ifdef USE_WINDOWS_FILE + p->handle = INVALID_HANDLE_VALUE; + #else + p->file = NULL; + #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(); + #else + p->file = fopen(name, writeMode ? "wb+" : "rb"); + return (p->file != 0) ? 0 : + #ifdef UNDER_CE + 2; /* ENOENT */ + #else + errno; + #endif + #endif +} + +WRes InFile_Open(CSzFile *p, const char *name) { return File_Open(p, name, 0); } +WRes OutFile_Open(CSzFile *p, const char *name) { return File_Open(p, name, 1); } +#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; + } + #else + if (p->file != NULL) + { + int res = fclose(p->file); + if (res != 0) + return res; + p->file = NULL; + } + #endif + return 0; +} + +WRes File_Read(CSzFile *p, void *data, size_t *size) +{ + size_t originalSize = *size; + if (originalSize == 0) + return 0; + + #ifdef USE_WINDOWS_FILE + + *size = 0; + do + { + DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize; + DWORD processed = 0; + BOOL res = ReadFile(p->handle, data, curSize, &processed, NULL); + data = (void *)((Byte *)data + processed); + originalSize -= processed; + *size += processed; + if (!res) + return GetLastError(); + if (processed == 0) + break; + } + while (originalSize > 0); + return 0; + + #else + + *size = fread(data, 1, originalSize, p->file); + if (*size == originalSize) + return 0; + return ferror(p->file); + + #endif +} + +WRes File_Write(CSzFile *p, const void *data, size_t *size) +{ + size_t originalSize = *size; + if (originalSize == 0) + return 0; + + #ifdef USE_WINDOWS_FILE + + *size = 0; + do + { + DWORD curSize = (originalSize > kChunkSizeMax) ? kChunkSizeMax : (DWORD)originalSize; + DWORD processed = 0; + BOOL res = WriteFile(p->handle, data, curSize, &processed, NULL); + data = (void *)((Byte *)data + processed); + originalSize -= processed; + *size += processed; + if (!res) + return GetLastError(); + if (processed == 0) + break; + } + while (originalSize > 0); + return 0; + + #else + + *size = fwrite(data, 1, originalSize, p->file); + if (*size == originalSize) + return 0; + return ferror(p->file); + + #endif +} + +WRes File_Seek(CSzFile *p, Int64 *pos, ESzSeek origin) +{ + #ifdef USE_WINDOWS_FILE + + LARGE_INTEGER value; + DWORD moveMethod; + value.LowPart = (DWORD)*pos; + value.HighPart = (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; + } + value.LowPart = SetFilePointer(p->handle, value.LowPart, &value.HighPart, moveMethod); + if (value.LowPart == 0xFFFFFFFF) + { + WRes res = GetLastError(); + if (res != NO_ERROR) + return res; + } + *pos = ((Int64)value.HighPart << 32) | value.LowPart; + return 0; + + #else + + int moveMethod; + int res; + 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 1; + } + res = fseek(p->file, (long)*pos, moveMethod); + *pos = ftell(p->file); + return res; + + #endif +} + +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; + + #else + + 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; + + #endif +} + + +/* ---------- FileSeqInStream ---------- */ + +static SRes FileSeqInStream_Read(void *pp, void *buf, size_t *size) +{ + CFileSeqInStream *p = (CFileSeqInStream *)pp; + return File_Read(&p->file, buf, size) == 0 ? SZ_OK : SZ_ERROR_READ; +} + +void FileSeqInStream_CreateVTable(CFileSeqInStream *p) +{ + p->s.Read = FileSeqInStream_Read; +} + + +/* ---------- FileInStream ---------- */ + +static SRes FileInStream_Read(void *pp, void *buf, size_t *size) +{ + CFileInStream *p = (CFileInStream *)pp; + return (File_Read(&p->file, buf, size) == 0) ? SZ_OK : SZ_ERROR_READ; +} + +static SRes FileInStream_Seek(void *pp, Int64 *pos, ESzSeek origin) +{ + CFileInStream *p = (CFileInStream *)pp; + return File_Seek(&p->file, pos, origin); +} + +void FileInStream_CreateVTable(CFileInStream *p) +{ + p->s.Read = FileInStream_Read; + p->s.Seek = FileInStream_Seek; +} + + +/* ---------- FileOutStream ---------- */ + +static size_t FileOutStream_Write(void *pp, const void *data, size_t size) +{ + CFileOutStream *p = (CFileOutStream *)pp; + File_Write(&p->file, data, &size); + return size; +} + +void FileOutStream_CreateVTable(CFileOutStream *p) +{ + p->s.Write = FileOutStream_Write; +} diff --git a/src/lzma/7zStream.c b/src/lzma/7zStream.c new file mode 100644 index 0000000..d8467a6 --- /dev/null +++ b/src/lzma/7zStream.c @@ -0,0 +1,171 @@ +/* 7zStream.c -- 7z Stream functions +2013-11-12 : Igor Pavlov : Public domain */ + +#include + +#include + +#include + +SRes SeqInStream_Read2(ISeqInStream *stream, void *buf, size_t size, SRes errorType) +{ + while (size != 0) + { + size_t processed = size; + RINOK(stream->Read(stream, buf, &processed)); + if (processed == 0) + return errorType; + buf = (void *)((Byte *)buf + processed); + size -= processed; + } + return SZ_OK; +} + +SRes SeqInStream_Read(ISeqInStream *stream, void *buf, size_t size) +{ + return SeqInStream_Read2(stream, buf, size, SZ_ERROR_INPUT_EOF); +} + +SRes SeqInStream_ReadByte(ISeqInStream *stream, Byte *buf) +{ + size_t processed = 1; + RINOK(stream->Read(stream, buf, &processed)); + return (processed == 1) ? SZ_OK : SZ_ERROR_INPUT_EOF; +} + +SRes LookInStream_SeekTo(ILookInStream *stream, UInt64 offset) +{ + Int64 t = offset; + return stream->Seek(stream, &t, SZ_SEEK_SET); +} + +SRes LookInStream_LookRead(ILookInStream *stream, void *buf, size_t *size) +{ + const void *lookBuf; + if (*size == 0) + return SZ_OK; + RINOK(stream->Look(stream, &lookBuf, size)); + memcpy(buf, lookBuf, *size); + return stream->Skip(stream, *size); +} + +SRes LookInStream_Read2(ILookInStream *stream, void *buf, size_t size, SRes errorType) +{ + while (size != 0) + { + size_t processed = size; + RINOK(stream->Read(stream, buf, &processed)); + if (processed == 0) + return errorType; + buf = (void *)((Byte *)buf + processed); + size -= processed; + } + return SZ_OK; +} + +SRes LookInStream_Read(ILookInStream *stream, void *buf, size_t size) +{ + return LookInStream_Read2(stream, buf, size, SZ_ERROR_INPUT_EOF); +} + +static SRes LookToRead_Look_Lookahead(void *pp, const void **buf, size_t *size) +{ + SRes res = SZ_OK; + CLookToRead *p = (CLookToRead *)pp; + size_t size2 = p->size - p->pos; + if (size2 == 0 && *size > 0) + { + p->pos = 0; + size2 = LookToRead_BUF_SIZE; + res = p->realStream->Read(p->realStream, p->buf, &size2); + p->size = size2; + } + if (size2 < *size) + *size = size2; + *buf = p->buf + p->pos; + return res; +} + +static SRes LookToRead_Look_Exact(void *pp, const void **buf, size_t *size) +{ + SRes res = SZ_OK; + CLookToRead *p = (CLookToRead *)pp; + size_t size2 = p->size - p->pos; + if (size2 == 0 && *size > 0) + { + p->pos = 0; + if (*size > LookToRead_BUF_SIZE) + *size = LookToRead_BUF_SIZE; + res = p->realStream->Read(p->realStream, p->buf, size); + size2 = p->size = *size; + } + if (size2 < *size) + *size = size2; + *buf = p->buf + p->pos; + return res; +} + +static SRes LookToRead_Skip(void *pp, size_t offset) +{ + CLookToRead *p = (CLookToRead *)pp; + p->pos += offset; + return SZ_OK; +} + +static SRes LookToRead_Read(void *pp, void *buf, size_t *size) +{ + CLookToRead *p = (CLookToRead *)pp; + size_t rem = p->size - p->pos; + if (rem == 0) + return p->realStream->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 LookToRead_Seek(void *pp, Int64 *pos, ESzSeek origin) +{ + CLookToRead *p = (CLookToRead *)pp; + p->pos = p->size = 0; + return p->realStream->Seek(p->realStream, pos, origin); +} + +void LookToRead_CreateVTable(CLookToRead *p, int lookahead) +{ + p->s.Look = lookahead ? + LookToRead_Look_Lookahead : + LookToRead_Look_Exact; + p->s.Skip = LookToRead_Skip; + p->s.Read = LookToRead_Read; + p->s.Seek = LookToRead_Seek; +} + +void LookToRead_Init(CLookToRead *p) +{ + p->pos = p->size = 0; +} + +static SRes SecToLook_Read(void *pp, void *buf, size_t *size) +{ + CSecToLook *p = (CSecToLook *)pp; + return LookInStream_LookRead(p->realStream, buf, size); +} + +void SecToLook_CreateVTable(CSecToLook *p) +{ + p->s.Read = SecToLook_Read; +} + +static SRes SecToRead_Read(void *pp, void *buf, size_t *size) +{ + CSecToRead *p = (CSecToRead *)pp; + return p->realStream->Read(p->realStream, buf, size); +} + +void SecToRead_CreateVTable(CSecToRead *p) +{ + p->s.Read = SecToRead_Read; +} diff --git a/src/lzma/Aes.c b/src/lzma/Aes.c new file mode 100644 index 0000000..ae03d90 --- /dev/null +++ b/src/lzma/Aes.c @@ -0,0 +1,295 @@ +/* Aes.c -- AES encryption / decryption +2016-05-21 : Igor Pavlov : Public domain */ + +#include + +#include +#include + +static UInt32 T[256 * 4]; +static const Byte Sbox[256] = { + 0x63, 0x7c, 0x77, 0x7b, 0xf2, 0x6b, 0x6f, 0xc5, 0x30, 0x01, 0x67, 0x2b, 0xfe, 0xd7, 0xab, 0x76, + 0xca, 0x82, 0xc9, 0x7d, 0xfa, 0x59, 0x47, 0xf0, 0xad, 0xd4, 0xa2, 0xaf, 0x9c, 0xa4, 0x72, 0xc0, + 0xb7, 0xfd, 0x93, 0x26, 0x36, 0x3f, 0xf7, 0xcc, 0x34, 0xa5, 0xe5, 0xf1, 0x71, 0xd8, 0x31, 0x15, + 0x04, 0xc7, 0x23, 0xc3, 0x18, 0x96, 0x05, 0x9a, 0x07, 0x12, 0x80, 0xe2, 0xeb, 0x27, 0xb2, 0x75, + 0x09, 0x83, 0x2c, 0x1a, 0x1b, 0x6e, 0x5a, 0xa0, 0x52, 0x3b, 0xd6, 0xb3, 0x29, 0xe3, 0x2f, 0x84, + 0x53, 0xd1, 0x00, 0xed, 0x20, 0xfc, 0xb1, 0x5b, 0x6a, 0xcb, 0xbe, 0x39, 0x4a, 0x4c, 0x58, 0xcf, + 0xd0, 0xef, 0xaa, 0xfb, 0x43, 0x4d, 0x33, 0x85, 0x45, 0xf9, 0x02, 0x7f, 0x50, 0x3c, 0x9f, 0xa8, + 0x51, 0xa3, 0x40, 0x8f, 0x92, 0x9d, 0x38, 0xf5, 0xbc, 0xb6, 0xda, 0x21, 0x10, 0xff, 0xf3, 0xd2, + 0xcd, 0x0c, 0x13, 0xec, 0x5f, 0x97, 0x44, 0x17, 0xc4, 0xa7, 0x7e, 0x3d, 0x64, 0x5d, 0x19, 0x73, + 0x60, 0x81, 0x4f, 0xdc, 0x22, 0x2a, 0x90, 0x88, 0x46, 0xee, 0xb8, 0x14, 0xde, 0x5e, 0x0b, 0xdb, + 0xe0, 0x32, 0x3a, 0x0a, 0x49, 0x06, 0x24, 0x5c, 0xc2, 0xd3, 0xac, 0x62, 0x91, 0x95, 0xe4, 0x79, + 0xe7, 0xc8, 0x37, 0x6d, 0x8d, 0xd5, 0x4e, 0xa9, 0x6c, 0x56, 0xf4, 0xea, 0x65, 0x7a, 0xae, 0x08, + 0xba, 0x78, 0x25, 0x2e, 0x1c, 0xa6, 0xb4, 0xc6, 0xe8, 0xdd, 0x74, 0x1f, 0x4b, 0xbd, 0x8b, 0x8a, + 0x70, 0x3e, 0xb5, 0x66, 0x48, 0x03, 0xf6, 0x0e, 0x61, 0x35, 0x57, 0xb9, 0x86, 0xc1, 0x1d, 0x9e, + 0xe1, 0xf8, 0x98, 0x11, 0x69, 0xd9, 0x8e, 0x94, 0x9b, 0x1e, 0x87, 0xe9, 0xce, 0x55, 0x28, 0xdf, + 0x8c, 0xa1, 0x89, 0x0d, 0xbf, 0xe6, 0x42, 0x68, 0x41, 0x99, 0x2d, 0x0f, 0xb0, 0x54, 0xbb, 0x16}; + +void MY_FAST_CALL AesCbc_Encode(UInt32 *ivAes, Byte *data, size_t numBlocks); +void MY_FAST_CALL AesCbc_Decode(UInt32 *ivAes, Byte *data, size_t numBlocks); +void MY_FAST_CALL AesCtr_Code(UInt32 *ivAes, Byte *data, size_t numBlocks); + +void MY_FAST_CALL AesCbc_Encode_Intel(UInt32 *ivAes, Byte *data, size_t numBlocks); +void MY_FAST_CALL AesCbc_Decode_Intel(UInt32 *ivAes, Byte *data, size_t numBlocks); +void MY_FAST_CALL AesCtr_Code_Intel(UInt32 *ivAes, Byte *data, size_t numBlocks); + +AES_CODE_FUNC g_AesCbc_Encode; +AES_CODE_FUNC g_AesCbc_Decode; +AES_CODE_FUNC g_AesCtr_Code; + +static UInt32 D[256 * 4]; +static Byte InvS[256]; + +static const Byte Rcon[11] = { 0x00, 0x01, 0x02, 0x04, 0x08, 0x10, 0x20, 0x40, 0x80, 0x1b, 0x36 }; + +#define xtime(x) ((((x) << 1) ^ (((x) & 0x80) != 0 ? 0x1B : 0)) & 0xFF) + +#define Ui32(a0, a1, a2, a3) ((UInt32)(a0) | ((UInt32)(a1) << 8) | ((UInt32)(a2) << 16) | ((UInt32)(a3) << 24)) + +#define gb0(x) ( (x) & 0xFF) +#define gb1(x) (((x) >> ( 8)) & 0xFF) +#define gb2(x) (((x) >> (16)) & 0xFF) +#define gb3(x) (((x) >> (24)) & 0xFF) + +void AesGenTables(void) +{ + unsigned i; + for (i = 0; i < 256; i++) + InvS[Sbox[i]] = (Byte)i; + + for (i = 0; i < 256; i++) + { + { + UInt32 a1 = Sbox[i]; + UInt32 a2 = xtime(a1); + UInt32 a3 = a2 ^ a1; + T[ i] = Ui32(a2, a1, a1, a3); + T[0x100 + i] = Ui32(a3, a2, a1, a1); + T[0x200 + i] = Ui32(a1, a3, a2, a1); + T[0x300 + i] = Ui32(a1, a1, a3, a2); + } + { + UInt32 a1 = InvS[i]; + UInt32 a2 = xtime(a1); + UInt32 a4 = xtime(a2); + UInt32 a8 = xtime(a4); + UInt32 a9 = a8 ^ a1; + UInt32 aB = a8 ^ a2 ^ a1; + UInt32 aD = a8 ^ a4 ^ a1; + UInt32 aE = a8 ^ a4 ^ a2; + D[ i] = Ui32(aE, a9, aD, aB); + D[0x100 + i] = Ui32(aB, aE, a9, aD); + D[0x200 + i] = Ui32(aD, aB, aE, a9); + D[0x300 + i] = Ui32(a9, aD, aB, aE); + } + } + + g_AesCbc_Encode = AesCbc_Encode; + g_AesCbc_Decode = AesCbc_Decode; + g_AesCtr_Code = AesCtr_Code; + + #ifdef MY_CPU_X86_OR_AMD64 + if (CPU_Is_Aes_Supported()) + { + g_AesCbc_Encode = AesCbc_Encode_Intel; + g_AesCbc_Decode = AesCbc_Decode_Intel; + g_AesCtr_Code = AesCtr_Code_Intel; + } + #endif +} + + +#define HT(i, x, s) (T + (x << 8))[gb ## x(s[(i + x) & 3])] + +#define HT4(m, i, s, p) m[i] = \ + HT(i, 0, s) ^ \ + HT(i, 1, s) ^ \ + HT(i, 2, s) ^ \ + HT(i, 3, s) ^ w[p + i] + +#define HT16(m, s, p) \ + HT4(m, 0, s, p); \ + HT4(m, 1, s, p); \ + HT4(m, 2, s, p); \ + HT4(m, 3, s, p); \ + +#define FT(i, x) Sbox[gb ## x(m[(i + x) & 3])] +#define FT4(i) dest[i] = Ui32(FT(i, 0), FT(i, 1), FT(i, 2), FT(i, 3)) ^ w[i]; + + +#define HD(i, x, s) (D + (x << 8))[gb ## x(s[(i - x) & 3])] + +#define HD4(m, i, s, p) m[i] = \ + HD(i, 0, s) ^ \ + HD(i, 1, s) ^ \ + HD(i, 2, s) ^ \ + HD(i, 3, s) ^ w[p + i]; + +#define HD16(m, s, p) \ + HD4(m, 0, s, p); \ + HD4(m, 1, s, p); \ + HD4(m, 2, s, p); \ + HD4(m, 3, s, p); \ + +#define FD(i, x) InvS[gb ## x(m[(i - x) & 3])] +#define FD4(i) dest[i] = Ui32(FD(i, 0), FD(i, 1), FD(i, 2), FD(i, 3)) ^ w[i]; + +void MY_FAST_CALL Aes_SetKey_Enc(UInt32 *w, const Byte *key, unsigned keySize) +{ + unsigned i, wSize; + wSize = keySize + 28; + keySize /= 4; + w[0] = ((UInt32)keySize / 2) + 3; + w += 4; + + for (i = 0; i < keySize; i++, key += 4) + w[i] = GetUi32(key); + + for (; i < wSize; i++) + { + UInt32 t = w[i - 1]; + unsigned rem = i % keySize; + if (rem == 0) + t = Ui32(Sbox[gb1(t)] ^ Rcon[i / keySize], Sbox[gb2(t)], Sbox[gb3(t)], Sbox[gb0(t)]); + else if (keySize > 6 && rem == 4) + t = Ui32(Sbox[gb0(t)], Sbox[gb1(t)], Sbox[gb2(t)], Sbox[gb3(t)]); + w[i] = w[i - keySize] ^ t; + } +} + +void MY_FAST_CALL Aes_SetKey_Dec(UInt32 *w, const Byte *key, unsigned keySize) +{ + unsigned i, num; + Aes_SetKey_Enc(w, key, keySize); + num = keySize + 20; + w += 8; + for (i = 0; i < num; i++) + { + UInt32 r = w[i]; + w[i] = + D[ (unsigned)Sbox[gb0(r)]] ^ + D[0x100 + (unsigned)Sbox[gb1(r)]] ^ + D[0x200 + (unsigned)Sbox[gb2(r)]] ^ + D[0x300 + (unsigned)Sbox[gb3(r)]]; + } +} + +/* Aes_Encode and Aes_Decode functions work with little-endian words. + src and dest are pointers to 4 UInt32 words. + src and dest can point to same block */ + +static void Aes_Encode(const UInt32 *w, UInt32 *dest, const UInt32 *src) +{ + UInt32 s[4]; + UInt32 m[4]; + UInt32 numRounds2 = w[0]; + w += 4; + s[0] = src[0] ^ w[0]; + s[1] = src[1] ^ w[1]; + s[2] = src[2] ^ w[2]; + s[3] = src[3] ^ w[3]; + w += 4; + for (;;) + { + HT16(m, s, 0); + if (--numRounds2 == 0) + break; + HT16(s, m, 4); + w += 8; + } + w += 4; + FT4(0); FT4(1); FT4(2); FT4(3); +} + +static void Aes_Decode(const UInt32 *w, UInt32 *dest, const UInt32 *src) +{ + UInt32 s[4]; + UInt32 m[4]; + UInt32 numRounds2 = w[0]; + w += 4 + numRounds2 * 8; + s[0] = src[0] ^ w[0]; + s[1] = src[1] ^ w[1]; + s[2] = src[2] ^ w[2]; + s[3] = src[3] ^ w[3]; + for (;;) + { + w -= 8; + HD16(m, s, 4); + if (--numRounds2 == 0) + break; + HD16(s, m, 0); + } + FD4(0); FD4(1); FD4(2); FD4(3); +} + +void AesCbc_Init(UInt32 *p, const Byte *iv) +{ + unsigned i; + for (i = 0; i < 4; i++) + p[i] = GetUi32(iv + i * 4); +} + +void MY_FAST_CALL AesCbc_Encode(UInt32 *p, Byte *data, size_t numBlocks) +{ + for (; numBlocks != 0; numBlocks--, data += AES_BLOCK_SIZE) + { + p[0] ^= GetUi32(data); + p[1] ^= GetUi32(data + 4); + p[2] ^= GetUi32(data + 8); + p[3] ^= GetUi32(data + 12); + + Aes_Encode(p + 4, p, p); + + SetUi32(data, p[0]); + SetUi32(data + 4, p[1]); + SetUi32(data + 8, p[2]); + SetUi32(data + 12, p[3]); + } +} + +void MY_FAST_CALL AesCbc_Decode(UInt32 *p, Byte *data, size_t numBlocks) +{ + UInt32 in[4], out[4]; + for (; numBlocks != 0; numBlocks--, data += AES_BLOCK_SIZE) + { + in[0] = GetUi32(data); + in[1] = GetUi32(data + 4); + in[2] = GetUi32(data + 8); + in[3] = GetUi32(data + 12); + + Aes_Decode(p + 4, out, in); + + SetUi32(data, p[0] ^ out[0]); + SetUi32(data + 4, p[1] ^ out[1]); + SetUi32(data + 8, p[2] ^ out[2]); + SetUi32(data + 12, p[3] ^ out[3]); + + p[0] = in[0]; + p[1] = in[1]; + p[2] = in[2]; + p[3] = in[3]; + } +} + +void MY_FAST_CALL AesCtr_Code(UInt32 *p, Byte *data, size_t numBlocks) +{ + for (; numBlocks != 0; numBlocks--) + { + UInt32 temp[4]; + Byte buf[16]; + int i; + + if (++p[0] == 0) + p[1]++; + + Aes_Encode(p + 4, temp, p); + + SetUi32(buf, temp[0]); + SetUi32(buf + 4, temp[1]); + SetUi32(buf + 8, temp[2]); + SetUi32(buf + 12, temp[3]); + + for (i = 0; i < 16; i++) + *data++ ^= buf[i]; + } +} diff --git a/src/lzma/AesOpt.c b/src/lzma/AesOpt.c new file mode 100644 index 0000000..01e847d --- /dev/null +++ b/src/lzma/AesOpt.c @@ -0,0 +1,184 @@ +/* AesOpt.c -- Intel's AES +2013-11-12 : Igor Pavlov : Public domain */ + +#include + +#include + +#ifdef MY_CPU_X86_OR_AMD64 +#if _MSC_VER >= 1500 +#define USE_INTEL_AES +#endif +#endif + +#ifdef USE_INTEL_AES + +#include + +void MY_FAST_CALL AesCbc_Encode_Intel(__m128i *p, __m128i *data, size_t numBlocks) +{ + __m128i m = *p; + for (; numBlocks != 0; numBlocks--, data++) + { + UInt32 numRounds2 = *(const UInt32 *)(p + 1) - 1; + const __m128i *w = p + 3; + m = _mm_xor_si128(m, *data); + m = _mm_xor_si128(m, p[2]); + do + { + m = _mm_aesenc_si128(m, w[0]); + m = _mm_aesenc_si128(m, w[1]); + w += 2; + } + while (--numRounds2 != 0); + m = _mm_aesenc_si128(m, w[0]); + m = _mm_aesenclast_si128(m, w[1]); + *data = m; + } + *p = m; +} + +#define NUM_WAYS 3 + +#define AES_OP_W(op, n) { \ + const __m128i t = w[n]; \ + m0 = op(m0, t); \ + m1 = op(m1, t); \ + m2 = op(m2, t); \ + } + +#define AES_DEC(n) AES_OP_W(_mm_aesdec_si128, n) +#define AES_DEC_LAST(n) AES_OP_W(_mm_aesdeclast_si128, n) +#define AES_ENC(n) AES_OP_W(_mm_aesenc_si128, n) +#define AES_ENC_LAST(n) AES_OP_W(_mm_aesenclast_si128, n) + +void MY_FAST_CALL AesCbc_Decode_Intel(__m128i *p, __m128i *data, size_t numBlocks) +{ + __m128i iv = *p; + for (; numBlocks >= NUM_WAYS; numBlocks -= NUM_WAYS, data += NUM_WAYS) + { + UInt32 numRounds2 = *(const UInt32 *)(p + 1); + const __m128i *w = p + numRounds2 * 2; + __m128i m0, m1, m2; + { + const __m128i t = w[2]; + m0 = _mm_xor_si128(t, data[0]); + m1 = _mm_xor_si128(t, data[1]); + m2 = _mm_xor_si128(t, data[2]); + } + numRounds2--; + do + { + AES_DEC(1) + AES_DEC(0) + w -= 2; + } + while (--numRounds2 != 0); + AES_DEC(1) + AES_DEC_LAST(0) + + { + __m128i t; + t = _mm_xor_si128(m0, iv); iv = data[0]; data[0] = t; + t = _mm_xor_si128(m1, iv); iv = data[1]; data[1] = t; + t = _mm_xor_si128(m2, iv); iv = data[2]; data[2] = t; + } + } + for (; numBlocks != 0; numBlocks--, data++) + { + UInt32 numRounds2 = *(const UInt32 *)(p + 1); + const __m128i *w = p + numRounds2 * 2; + __m128i m = _mm_xor_si128(w[2], *data); + numRounds2--; + do + { + m = _mm_aesdec_si128(m, w[1]); + m = _mm_aesdec_si128(m, w[0]); + w -= 2; + } + while (--numRounds2 != 0); + m = _mm_aesdec_si128(m, w[1]); + m = _mm_aesdeclast_si128(m, w[0]); + + m = _mm_xor_si128(m, iv); + iv = *data; + *data = m; + } + *p = iv; +} + +void MY_FAST_CALL AesCtr_Code_Intel(__m128i *p, __m128i *data, size_t numBlocks) +{ + __m128i ctr = *p; + __m128i one; + one.m128i_u64[0] = 1; + one.m128i_u64[1] = 0; + for (; numBlocks >= NUM_WAYS; numBlocks -= NUM_WAYS, data += NUM_WAYS) + { + UInt32 numRounds2 = *(const UInt32 *)(p + 1) - 1; + const __m128i *w = p; + __m128i m0, m1, m2; + { + const __m128i t = w[2]; + ctr = _mm_add_epi64(ctr, one); m0 = _mm_xor_si128(ctr, t); + ctr = _mm_add_epi64(ctr, one); m1 = _mm_xor_si128(ctr, t); + ctr = _mm_add_epi64(ctr, one); m2 = _mm_xor_si128(ctr, t); + } + w += 3; + do + { + AES_ENC(0) + AES_ENC(1) + w += 2; + } + while (--numRounds2 != 0); + AES_ENC(0) + AES_ENC_LAST(1) + data[0] = _mm_xor_si128(data[0], m0); + data[1] = _mm_xor_si128(data[1], m1); + data[2] = _mm_xor_si128(data[2], m2); + } + for (; numBlocks != 0; numBlocks--, data++) + { + UInt32 numRounds2 = *(const UInt32 *)(p + 1) - 1; + const __m128i *w = p; + __m128i m; + ctr = _mm_add_epi64(ctr, one); + m = _mm_xor_si128(ctr, p[2]); + w += 3; + do + { + m = _mm_aesenc_si128(m, w[0]); + m = _mm_aesenc_si128(m, w[1]); + w += 2; + } + while (--numRounds2 != 0); + m = _mm_aesenc_si128(m, w[0]); + m = _mm_aesenclast_si128(m, w[1]); + *data = _mm_xor_si128(*data, m); + } + *p = ctr; +} + +#else + +void MY_FAST_CALL AesCbc_Encode(UInt32 *ivAes, Byte *data, size_t numBlocks); +void MY_FAST_CALL AesCbc_Decode(UInt32 *ivAes, Byte *data, size_t numBlocks); +void MY_FAST_CALL AesCtr_Code(UInt32 *ivAes, Byte *data, size_t numBlocks); + +void MY_FAST_CALL AesCbc_Encode_Intel(UInt32 *p, Byte *data, size_t numBlocks) +{ + AesCbc_Encode(p, data, numBlocks); +} + +void MY_FAST_CALL AesCbc_Decode_Intel(UInt32 *p, Byte *data, size_t numBlocks) +{ + AesCbc_Decode(p, data, numBlocks); +} + +void MY_FAST_CALL AesCtr_Code_Intel(UInt32 *p, Byte *data, size_t numBlocks) +{ + AesCtr_Code(p, data, numBlocks); +} + +#endif diff --git a/src/lzma/Alloc.c b/src/lzma/Alloc.c new file mode 100644 index 0000000..b5a47fa --- /dev/null +++ b/src/lzma/Alloc.c @@ -0,0 +1,136 @@ +/* Alloc.c -- Memory allocation functions +2015-02-21 : Igor Pavlov : Public domain */ + +#include + +#ifdef _WIN32 +#include +#endif +#include + +#include + +/* #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; +#endif + +void *MyAlloc(size_t size) +{ + if (size == 0) + return 0; + #ifdef _SZ_ALLOC_DEBUG + { + void *p = malloc(size); + fprintf(stderr, "\nAlloc %10d bytes, count = %10d, addr = %8X", size, g_allocCount++, (unsigned)p); + return p; + } + #else + return malloc(size); + #endif +} + +void MyFree(void *address) +{ + #ifdef _SZ_ALLOC_DEBUG + if (address != 0) + fprintf(stderr, "\nFree; count = %10d, addr = %8X", --g_allocCount, (unsigned)address); + #endif + free(address); +} + +#ifdef _WIN32 + +void *MidAlloc(size_t size) +{ + if (size == 0) + return 0; + #ifdef _SZ_ALLOC_DEBUG + fprintf(stderr, "\nAlloc_Mid %10d bytes; count = %10d", size, g_allocCountMid++); + #endif + return VirtualAlloc(0, size, MEM_COMMIT, PAGE_READWRITE); +} + +void MidFree(void *address) +{ + #ifdef _SZ_ALLOC_DEBUG + if (address != 0) + fprintf(stderr, "\nFree_Mid; count = %10d", --g_allocCountMid); + #endif + if (address == 0) + return; + VirtualFree(address, 0, MEM_RELEASE); +} + +#ifndef MEM_LARGE_PAGES +#undef _7ZIP_LARGE_PAGES +#endif + +#ifdef _7ZIP_LARGE_PAGES +SIZE_T g_LargePageSize = 0; +typedef SIZE_T (WINAPI *GetLargePageMinimumP)(); +#endif + +void SetLargePageSize() +{ + #ifdef _7ZIP_LARGE_PAGES + SIZE_T size = 0; + GetLargePageMinimumP largePageMinimum = (GetLargePageMinimumP) + GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "GetLargePageMinimum"); + if (largePageMinimum == 0) + return; + size = largePageMinimum(); + if (size == 0 || (size & (size - 1)) != 0) + return; + g_LargePageSize = size; + #endif +} + + +void *BigAlloc(size_t size) +{ + if (size == 0) + return 0; + #ifdef _SZ_ALLOC_DEBUG + fprintf(stderr, "\nAlloc_Big %10d bytes; count = %10d", size, g_allocCountBig++); + #endif + + #ifdef _7ZIP_LARGE_PAGES + if (g_LargePageSize != 0 && g_LargePageSize <= (1 << 30) && size >= (1 << 18)) + { + void *res = VirtualAlloc(0, (size + g_LargePageSize - 1) & (~(g_LargePageSize - 1)), + MEM_COMMIT | MEM_LARGE_PAGES, PAGE_READWRITE); + if (res != 0) + return res; + } + #endif + return VirtualAlloc(0, size, MEM_COMMIT, PAGE_READWRITE); +} + +void BigFree(void *address) +{ + #ifdef _SZ_ALLOC_DEBUG + if (address != 0) + fprintf(stderr, "\nFree_Big; count = %10d", --g_allocCountBig); + #endif + + if (address == 0) + return; + VirtualFree(address, 0, MEM_RELEASE); +} + +#endif + + +static void *SzAlloc(void *p, size_t size) { UNUSED_VAR(p); return MyAlloc(size); } +static void SzFree(void *p, void *address) { UNUSED_VAR(p); MyFree(address); } +ISzAlloc g_Alloc = { SzAlloc, SzFree }; + +static void *SzBigAlloc(void *p, size_t size) { UNUSED_VAR(p); return BigAlloc(size); } +static void SzBigFree(void *p, void *address) { UNUSED_VAR(p); BigFree(address); } +ISzAlloc g_BigAlloc = { SzBigAlloc, SzBigFree }; diff --git a/src/lzma/Bcj2.c b/src/lzma/Bcj2.c new file mode 100644 index 0000000..4f13e48 --- /dev/null +++ b/src/lzma/Bcj2.c @@ -0,0 +1,256 @@ +/* Bcj2.c -- BCJ2 Decoder (Converter for x86 code) +2015-08-01 : Igor Pavlov : Public domain */ + +#include + +#include +#include + +#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[p->state++ - BCJ2_DEC_STATE_ORIG_0]; + 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 = 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 = 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 = 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 = 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; + 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/src/lzma/Bcj2Enc.c b/src/lzma/Bcj2Enc.c new file mode 100644 index 0000000..01c38c0 --- /dev/null +++ b/src/lzma/Bcj2Enc.c @@ -0,0 +1,312 @@ +/* Bcj2Enc.c -- BCJ2 Encoder (Converter for x86 code) +2014-11-10 : Igor Pavlov : Public domain */ + +#include + +/* #define SHOW_STAT */ + +#ifdef SHOW_STAT +#include +#define PRF(x) x +#else +#define PRF(x) +#endif + +#include +#include + +#include +#include + +#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 Bool 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 = 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 = 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 = 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]); + Bool 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; + unsigned 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[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/src/lzma/Bra.c b/src/lzma/Bra.c new file mode 100644 index 0000000..2300e2d --- /dev/null +++ b/src/lzma/Bra.c @@ -0,0 +1,135 @@ +/* Bra.c -- Converters for RISC code +2010-04-16 : Igor Pavlov : Public domain */ + +#include + +#include + +SizeT ARM_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + SizeT i; + if (size < 4) + return 0; + size -= 4; + ip += 8; + for (i = 0; i <= size; i += 4) + { + if (data[i + 3] == 0xEB) + { + UInt32 dest; + UInt32 src = ((UInt32)data[i + 2] << 16) | ((UInt32)data[i + 1] << 8) | (data[i + 0]); + src <<= 2; + if (encoding) + dest = ip + (UInt32)i + src; + else + dest = src - (ip + (UInt32)i); + dest >>= 2; + data[i + 2] = (Byte)(dest >> 16); + data[i + 1] = (Byte)(dest >> 8); + data[i + 0] = (Byte)dest; + } + } + return i; +} + +SizeT ARMT_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + SizeT i; + if (size < 4) + return 0; + size -= 4; + ip += 4; + for (i = 0; i <= size; i += 2) + { + if ((data[i + 1] & 0xF8) == 0xF0 && + (data[i + 3] & 0xF8) == 0xF8) + { + UInt32 dest; + UInt32 src = + (((UInt32)data[i + 1] & 0x7) << 19) | + ((UInt32)data[i + 0] << 11) | + (((UInt32)data[i + 3] & 0x7) << 8) | + (data[i + 2]); + + src <<= 1; + if (encoding) + dest = ip + (UInt32)i + src; + else + dest = src - (ip + (UInt32)i); + dest >>= 1; + + data[i + 1] = (Byte)(0xF0 | ((dest >> 19) & 0x7)); + data[i + 0] = (Byte)(dest >> 11); + data[i + 3] = (Byte)(0xF8 | ((dest >> 8) & 0x7)); + data[i + 2] = (Byte)dest; + i += 2; + } + } + return i; +} + +SizeT PPC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + SizeT i; + if (size < 4) + return 0; + size -= 4; + for (i = 0; i <= size; i += 4) + { + if ((data[i] >> 2) == 0x12 && (data[i + 3] & 3) == 1) + { + UInt32 src = ((UInt32)(data[i + 0] & 3) << 24) | + ((UInt32)data[i + 1] << 16) | + ((UInt32)data[i + 2] << 8) | + ((UInt32)data[i + 3] & (~3)); + + UInt32 dest; + if (encoding) + dest = ip + (UInt32)i + src; + else + dest = src - (ip + (UInt32)i); + data[i + 0] = (Byte)(0x48 | ((dest >> 24) & 0x3)); + data[i + 1] = (Byte)(dest >> 16); + data[i + 2] = (Byte)(dest >> 8); + data[i + 3] &= 0x3; + data[i + 3] |= dest; + } + } + return i; +} + +SizeT SPARC_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + UInt32 i; + if (size < 4) + return 0; + size -= 4; + for (i = 0; i <= size; i += 4) + { + if ((data[i] == 0x40 && (data[i + 1] & 0xC0) == 0x00) || + (data[i] == 0x7F && (data[i + 1] & 0xC0) == 0xC0)) + { + UInt32 src = + ((UInt32)data[i + 0] << 24) | + ((UInt32)data[i + 1] << 16) | + ((UInt32)data[i + 2] << 8) | + ((UInt32)data[i + 3]); + UInt32 dest; + + src <<= 2; + if (encoding) + dest = ip + i + src; + else + dest = src - (ip + i); + dest >>= 2; + + dest = (((0 - ((dest >> 22) & 1)) << 22) & 0x3FFFFFFF) | (dest & 0x3FFFFF) | 0x40000000; + + data[i + 0] = (Byte)(dest >> 24); + data[i + 1] = (Byte)(dest >> 16); + data[i + 2] = (Byte)(dest >> 8); + data[i + 3] = (Byte)dest; + } + } + return i; +} diff --git a/src/lzma/Bra86.c b/src/lzma/Bra86.c new file mode 100644 index 0000000..034e60c --- /dev/null +++ b/src/lzma/Bra86.c @@ -0,0 +1,82 @@ +/* Bra86.c -- Converter for x86 code (BCJ) +2013-11-12 : Igor Pavlov : Public domain */ + +#include + +#include + +#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[(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/src/lzma/BraIA64.c b/src/lzma/BraIA64.c new file mode 100644 index 0000000..50726ea --- /dev/null +++ b/src/lzma/BraIA64.c @@ -0,0 +1,69 @@ +/* BraIA64.c -- Converter for IA-64 code +2013-11-12 : Igor Pavlov : Public domain */ + +#include + +#include + +static const Byte kBranchTable[32] = +{ + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 4, 4, 6, 6, 0, 0, 7, 7, + 4, 4, 0, 0, 4, 4, 0, 0 +}; + +SizeT IA64_Convert(Byte *data, SizeT size, UInt32 ip, int encoding) +{ + SizeT i; + if (size < 16) + return 0; + size -= 16; + for (i = 0; i <= size; i += 16) + { + UInt32 instrTemplate = data[i] & 0x1F; + UInt32 mask = kBranchTable[instrTemplate]; + UInt32 bitPos = 5; + int slot; + for (slot = 0; slot < 3; slot++, bitPos += 41) + { + UInt32 bytePos, bitRes; + UInt64 instruction, instNorm; + int j; + if (((mask >> slot) & 1) == 0) + continue; + bytePos = (bitPos >> 3); + bitRes = bitPos & 0x7; + instruction = 0; + for (j = 0; j < 6; j++) + instruction += (UInt64)data[i + j + bytePos] << (8 * j); + + instNorm = instruction >> bitRes; + if (((instNorm >> 37) & 0xF) == 0x5 && ((instNorm >> 9) & 0x7) == 0) + { + UInt32 src = (UInt32)((instNorm >> 13) & 0xFFFFF); + UInt32 dest; + src |= ((UInt32)(instNorm >> 36) & 1) << 20; + + src <<= 4; + + if (encoding) + dest = ip + (UInt32)i + src; + else + dest = src - (ip + (UInt32)i); + + dest >>= 4; + + instNorm &= ~((UInt64)(0x8FFFFF) << 13); + instNorm |= ((UInt64)(dest & 0xFFFFF) << 13); + instNorm |= ((UInt64)(dest & 0x100000) << (36 - 20)); + + instruction &= (1 << bitRes) - 1; + instruction |= (instNorm << bitRes); + for (j = 0; j < 6; j++) + data[i + j + bytePos] = (Byte)(instruction >> (8 * j)); + } + } + } + return i; +} diff --git a/src/lzma/CpuArch.c b/src/lzma/CpuArch.c new file mode 100644 index 0000000..865095f --- /dev/null +++ b/src/lzma/CpuArch.c @@ -0,0 +1,200 @@ +/* CpuArch.c -- CPU specific code +2016-02-25: Igor Pavlov : Public domain */ + +#include + +#include + +#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 + +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)) ; + + #endif + + #else + + int CPUInfo[4]; + __cpuid(CPUInfo, function); + *a = CPUInfo[0]; + *b = CPUInfo[1]; + *c = CPUInfo[2]; + *d = CPUInfo[3]; + + #endif +} + +Bool 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; +} + +Bool 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 Bool 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 + +Bool CPU_Is_Aes_Supported() +{ + Cx86cpuid p; + CHECK_SYS_SSE_SUPPORT + if (!x86cpuid_CheckAndRead(&p)) + return False; + return (p.c >> 25) & 1; +} + +#endif diff --git a/src/lzma/Delta.c b/src/lzma/Delta.c new file mode 100644 index 0000000..ce610fe --- /dev/null +++ b/src/lzma/Delta.c @@ -0,0 +1,64 @@ +/* Delta.c -- Delta converter +2009-05-26 : Igor Pavlov : Public domain */ + +#include + +#include + +void Delta_Init(Byte *state) +{ + unsigned i; + for (i = 0; i < DELTA_STATE_SIZE; i++) + state[i] = 0; +} + +static void MyMemCpy(Byte *dest, const Byte *src, unsigned size) +{ + unsigned i; + for (i = 0; i < size; i++) + dest[i] = src[i]; +} + +void Delta_Encode(Byte *state, unsigned delta, Byte *data, SizeT size) +{ + Byte buf[DELTA_STATE_SIZE]; + unsigned j = 0; + MyMemCpy(buf, state, delta); + { + SizeT i; + for (i = 0; i < size;) + { + for (j = 0; j < delta && i < size; i++, j++) + { + Byte b = data[i]; + data[i] = (Byte)(b - buf[j]); + buf[j] = b; + } + } + } + if (j == delta) + j = 0; + MyMemCpy(state, buf + j, delta - j); + MyMemCpy(state + delta - j, buf, j); +} + +void Delta_Decode(Byte *state, unsigned delta, Byte *data, SizeT size) +{ + Byte buf[DELTA_STATE_SIZE]; + unsigned j = 0; + MyMemCpy(buf, state, delta); + { + SizeT i; + for (i = 0; i < size;) + { + for (j = 0; j < delta && i < size; i++, j++) + { + buf[j] = data[i] = (Byte)(buf[j] + data[i]); + } + } + } + if (j == delta) + j = 0; + MyMemCpy(state, buf + j, delta - j); + MyMemCpy(state + delta - j, buf, j); +} diff --git a/src/lzma/LzFind.c b/src/lzma/LzFind.c new file mode 100644 index 0000000..50d3eed --- /dev/null +++ b/src/lzma/LzFind.c @@ -0,0 +1,1044 @@ +/* LzFind.c -- Match finder for LZ algorithms +2015-10-15 : Igor Pavlov : Public domain */ + +#include + +#include + +#include +#include + +#define kEmptyHashValue 0 +#define kMaxValForNormalize ((UInt32)0xFFFFFFFF) +#define kNormalizeStepMin (1 << 10) /* it must be power of 2 */ +#define kNormalizeMask (~(UInt32)(kNormalizeStepMin - 1)) +#define kMaxHistorySize ((UInt32)7 << 29) + +#define kStartMaxLen 3 + +static void LzInWindow_Free(CMatchFinder *p, ISzAlloc *alloc) +{ + if (!p->directInput) + { + alloc->Free(alloc, p->bufferBase); + p->bufferBase = NULL; + } +} + +/* keepSizeBefore + keepSizeAfter + keepSizeReserv must be < 4G) */ + +static int LzInWindow_Create(CMatchFinder *p, UInt32 keepSizeReserv, ISzAlloc *alloc) +{ + UInt32 blockSize = p->keepSizeBefore + p->keepSizeAfter + keepSizeReserv; + if (p->directInput) + { + p->blockSize = blockSize; + return 1; + } + if (!p->bufferBase || p->blockSize != blockSize) + { + LzInWindow_Free(p, alloc); + p->blockSize = blockSize; + p->bufferBase = (Byte *)alloc->Alloc(alloc, (size_t)blockSize); + } + return (p->bufferBase != NULL); +} + +Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p) { return p->buffer; } + +UInt32 MatchFinder_GetNumAvailableBytes(CMatchFinder *p) { return p->streamPos - p->pos; } + +void MatchFinder_ReduceOffsets(CMatchFinder *p, UInt32 subValue) +{ + p->posLimit -= subValue; + p->pos -= subValue; + p->streamPos -= subValue; +} + +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 - (p->streamPos - p->pos); + 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 + (p->streamPos - p->pos); + size_t size = (p->bufferBase + p->blockSize - dest); + if (size == 0) + return; + + p->result = p->stream->Read(p->stream, dest, &size); + if (p->result != SZ_OK) + return; + if (size == 0) + { + p->streamEndWasReached = 1; + return; + } + p->streamPos += (UInt32)size; + if (p->streamPos - p->pos > p->keepSizeAfter) + return; + } +} + +void MatchFinder_MoveBlock(CMatchFinder *p) +{ + memmove(p->bufferBase, + p->buffer - p->keepSizeBefore, + (size_t)(p->streamPos - p->pos) + p->keepSizeBefore); + p->buffer = p->bufferBase + p->keepSizeBefore; +} + +int MatchFinder_NeedMove(CMatchFinder *p) +{ + if (p->directInput) + return 0; + /* if (p->streamEndWasReached) return 0; */ + return ((size_t)(p->bufferBase + p->blockSize - p->buffer) <= p->keepSizeAfter); +} + +void MatchFinder_ReadIfRequired(CMatchFinder *p) +{ + if (p->streamEndWasReached) + return; + if (p->keepSizeAfter >= p->streamPos - p->pos) + MatchFinder_ReadBlock(p); +} + +static void MatchFinder_CheckAndMoveAndRead(CMatchFinder *p) +{ + if (MatchFinder_NeedMove(p)) + MatchFinder_MoveBlock(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) +{ + UInt32 i; + p->bufferBase = NULL; + p->directInput = 0; + p->hash = NULL; + MatchFinder_SetDefaultSettings(p); + + for (i = 0; i < 256; i++) + { + UInt32 r = i; + unsigned j; + for (j = 0; j < 8; j++) + r = (r >> 1) ^ (kCrcPoly & ~((r & 1) - 1)); + p->crc[i] = r; + } +} + +static void MatchFinder_FreeThisClassMemory(CMatchFinder *p, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->hash); + p->hash = NULL; +} + +void MatchFinder_Free(CMatchFinder *p, ISzAlloc *alloc) +{ + MatchFinder_FreeThisClassMemory(p, alloc); + LzInWindow_Free(p, alloc); +} + +static CLzRef* AllocRefs(size_t num, ISzAlloc *alloc) +{ + size_t sizeInBytes = (size_t)num * sizeof(CLzRef); + if (sizeInBytes / sizeof(CLzRef) != num) + return NULL; + return (CLzRef *)alloc->Alloc(alloc, sizeInBytes); +} + +int MatchFinder_Create(CMatchFinder *p, UInt32 historySize, + UInt32 keepAddBufferBefore, UInt32 matchMaxLen, UInt32 keepAddBufferAfter, + ISzAlloc *alloc) +{ + UInt32 sizeReserv; + + if (historySize > kMaxHistorySize) + { + MatchFinder_Free(p, alloc); + return 0; + } + + sizeReserv = historySize >> 1; + if (historySize >= ((UInt32)3 << 30)) sizeReserv = historySize >> 3; + else if (historySize >= ((UInt32)2 << 30)) sizeReserv = historySize >> 2; + + sizeReserv += (keepAddBufferBefore + matchMaxLen + keepAddBufferAfter) / 2 + (1 << 19); + + p->keepSizeBefore = historySize + keepAddBufferBefore + 1; + p->keepSizeAfter = matchMaxLen + keepAddBufferAfter; + + /* we need one additional byte, since we use MoveBlock after pos++ and before dictionary using */ + + if (LzInWindow_Create(p, sizeReserv, alloc)) + { + UInt32 newCyclicBufferSize = historySize + 1; + UInt32 hs; + p->matchMaxLen = matchMaxLen; + { + p->fixedHashSize = 0; + if (p->numHashBytes == 2) + hs = (1 << 16) - 1; + else + { + hs = historySize - 1; + hs |= (hs >> 1); + hs |= (hs >> 2); + hs |= (hs >> 4); + hs |= (hs >> 8); + hs >>= 1; + hs |= 0xFFFF; /* don't change it! It's required for Deflate */ + 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))) */ + } + } + p->hashMask = hs; + hs++; + if (p->numHashBytes > 2) p->fixedHashSize += kHash2Size; + if (p->numHashBytes > 3) p->fixedHashSize += kHash3Size; + if (p->numHashBytes > 4) p->fixedHashSize += kHash4Size; + hs += p->fixedHashSize; + } + + { + size_t newSize; + size_t numSons; + p->historySize = historySize; + p->hashSizeSum = hs; + p->cyclicBufferSize = newCyclicBufferSize; + + numSons = newCyclicBufferSize; + if (p->btMode) + numSons <<= 1; + newSize = hs + numSons; + + 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 limit = kMaxValForNormalize - p->pos; + UInt32 limit2 = p->cyclicBufferSize - p->cyclicBufferPos; + + if (limit2 < limit) + limit = limit2; + limit2 = p->streamPos - p->pos; + + if (limit2 <= p->keepSizeAfter) + { + if (limit2 > 0) + limit2 = 1; + } + else + limit2 -= p->keepSizeAfter; + + if (limit2 < limit) + limit = limit2; + + { + UInt32 lenLimit = p->streamPos - p->pos; + if (lenLimit > p->matchMaxLen) + lenLimit = p->matchMaxLen; + p->lenLimit = lenLimit; + } + p->posLimit = p->pos + limit; +} + +void MatchFinder_Init_2(CMatchFinder *p, int readData) +{ + UInt32 i; + UInt32 *hash = p->hash; + UInt32 num = p->hashSizeSum; + for (i = 0; i < num; i++) + hash[i] = kEmptyHashValue; + + p->cyclicBufferPos = 0; + p->buffer = p->bufferBase; + p->pos = p->streamPos = p->cyclicBufferSize; + p->result = SZ_OK; + p->streamEndWasReached = 0; + + if (readData) + MatchFinder_ReadBlock(p); + + MatchFinder_SetLimits(p); +} + +void MatchFinder_Init(CMatchFinder *p) +{ + MatchFinder_Init_2(p, True); +} + +static UInt32 MatchFinder_GetSubValue(CMatchFinder *p) +{ + return (p->pos - p->historySize - 1) & kNormalizeMask; +} + +void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, size_t numItems) +{ + size_t i; + for (i = 0; i < numItems; i++) + { + UInt32 value = items[i]; + if (value <= subValue) + value = kEmptyHashValue; + else + value -= subValue; + items[i] = value; + } +} + +static void MatchFinder_Normalize(CMatchFinder *p) +{ + UInt32 subValue = MatchFinder_GetSubValue(p); + MatchFinder_Normalize3(subValue, p->hash, p->numRefs); + MatchFinder_ReduceOffsets(p, subValue); +} + +static void MatchFinder_CheckLimits(CMatchFinder *p) +{ + if (p->pos == kMaxValForNormalize) + MatchFinder_Normalize(p); + if (!p->streamEndWasReached && p->keepSizeAfter == p->streamPos - p->pos) + MatchFinder_CheckAndMoveAndRead(p); + if (p->cyclicBufferPos == p->cyclicBufferSize) + p->cyclicBufferPos = 0; + MatchFinder_SetLimits(p); +} + +static UInt32 * Hc_GetMatchesSpec(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *distances, UInt32 maxLen) +{ + son[_cyclicBufferPos] = curMatch; + for (;;) + { + UInt32 delta = pos - curMatch; + if (cutValue-- == 0 || delta >= _cyclicBufferSize) + return distances; + { + 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) + { + *distances++ = maxLen = len; + *distances++ = delta - 1; + if (len == lenLimit) + return distances; + } + } + } + } +} + +UInt32 * GetMatchesSpec1(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue, + UInt32 *distances, UInt32 maxLen) +{ + CLzRef *ptr0 = son + (_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + (_cyclicBufferPos << 1); + UInt32 len0 = 0, len1 = 0; + for (;;) + { + UInt32 delta = pos - curMatch; + if (cutValue-- == 0 || delta >= _cyclicBufferSize) + { + *ptr0 = *ptr1 = kEmptyHashValue; + return distances; + } + { + CLzRef *pair = son + ((_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); + const Byte *pb = cur - delta; + UInt32 len = (len0 < len1 ? len0 : len1); + if (pb[len] == cur[len]) + { + if (++len != lenLimit && pb[len] == cur[len]) + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + *distances++ = maxLen = len; + *distances++ = delta - 1; + if (len == lenLimit) + { + *ptr1 = pair[0]; + *ptr0 = pair[1]; + return distances; + } + } + } + if (pb[len] < cur[len]) + { + *ptr1 = curMatch; + ptr1 = pair + 1; + curMatch = *ptr1; + len1 = len; + } + else + { + *ptr0 = curMatch; + ptr0 = pair; + curMatch = *ptr0; + len0 = len; + } + } + } +} + +static void SkipMatchesSpec(UInt32 lenLimit, UInt32 curMatch, UInt32 pos, const Byte *cur, CLzRef *son, + UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 cutValue) +{ + CLzRef *ptr0 = son + (_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + (_cyclicBufferPos << 1); + UInt32 len0 = 0, len1 = 0; + for (;;) + { + UInt32 delta = pos - curMatch; + if (cutValue-- == 0 || delta >= _cyclicBufferSize) + { + *ptr0 = *ptr1 = kEmptyHashValue; + return; + } + { + CLzRef *pair = son + ((_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); + const Byte *pb = cur - delta; + UInt32 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; + ptr1 = pair + 1; + curMatch = *ptr1; + len1 = len; + } + else + { + *ptr0 = curMatch; + ptr0 = pair; + curMatch = *ptr0; + len0 = len; + } + } + } +} + +#define MOVE_POS \ + ++p->cyclicBufferPos; \ + p->buffer++; \ + if (++p->pos == p->posLimit) MatchFinder_CheckLimits(p); + +#define MOVE_POS_RET MOVE_POS return offset; + +static void MatchFinder_MovePos(CMatchFinder *p) { MOVE_POS; } + +#define GET_MATCHES_HEADER2(minLen, ret_op) \ + UInt32 lenLimit; UInt32 hv; const Byte *cur; UInt32 curMatch; \ + lenLimit = p->lenLimit; { if (lenLimit < minLen) { MatchFinder_MovePos(p); ret_op; }} \ + cur = p->buffer; + +#define GET_MATCHES_HEADER(minLen) GET_MATCHES_HEADER2(minLen, return 0) +#define SKIP_HEADER(minLen) GET_MATCHES_HEADER2(minLen, continue) + +#define MF_PARAMS(p) p->pos, p->buffer, p->son, p->cyclicBufferPos, p->cyclicBufferSize, p->cutValue + +#define GET_MATCHES_FOOTER(offset, maxLen) \ + offset = (UInt32)(GetMatchesSpec1(lenLimit, curMatch, MF_PARAMS(p), \ + distances + offset, maxLen) - distances); MOVE_POS_RET; + +#define SKIP_FOOTER \ + SkipMatchesSpec(lenLimit, curMatch, MF_PARAMS(p)); MOVE_POS; + +#define UPDATE_maxLen { \ + ptrdiff_t diff = (ptrdiff_t)0 - d2; \ + const Byte *c = cur + maxLen; \ + const Byte *lim = cur + lenLimit; \ + for (; c != lim; c++) if (*(c + diff) != *c) break; \ + maxLen = (UInt32)(c - cur); } + +static UInt32 Bt2_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 offset; + GET_MATCHES_HEADER(2) + HASH2_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + offset = 0; + GET_MATCHES_FOOTER(offset, 1) +} + +UInt32 Bt3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 offset; + GET_MATCHES_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + offset = 0; + GET_MATCHES_FOOTER(offset, 2) +} + +static UInt32 Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, d2, maxLen, offset, pos; + 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; + + maxLen = 2; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + UPDATE_maxLen + distances[0] = maxLen; + distances[1] = d2 - 1; + offset = 2; + if (maxLen == lenLimit) + { + SkipMatchesSpec(lenLimit, curMatch, MF_PARAMS(p)); + MOVE_POS_RET; + } + } + + GET_MATCHES_FOOTER(offset, maxLen) +} + +static UInt32 Bt4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, h3, d2, d3, maxLen, offset, pos; + 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; + + maxLen = 0; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + distances[0] = maxLen = 2; + distances[1] = d2 - 1; + offset = 2; + } + + if (d2 != d3 && d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + maxLen = 3; + distances[offset + 1] = d3 - 1; + offset += 2; + d2 = d3; + } + + if (offset != 0) + { + UPDATE_maxLen + distances[offset - 2] = maxLen; + if (maxLen == lenLimit) + { + SkipMatchesSpec(lenLimit, curMatch, MF_PARAMS(p)); + MOVE_POS_RET; + } + } + + if (maxLen < 3) + maxLen = 3; + + GET_MATCHES_FOOTER(offset, maxLen) +} + +/* +static UInt32 Bt5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, h3, h4, d2, d3, d4, maxLen, offset, 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; + + maxLen = 0; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + distances[0] = maxLen = 2; + distances[1] = d2 - 1; + offset = 2; + if (*(cur - d2 + 2) == cur[2]) + distances[0] = maxLen = 3; + else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + distances[2] = maxLen = 3; + distances[3] = d3 - 1; + offset = 4; + d2 = d3; + } + } + else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + distances[0] = maxLen = 3; + distances[1] = d3 - 1; + offset = 2; + d2 = d3; + } + + if (d2 != d4 && d4 < p->cyclicBufferSize + && *(cur - d4) == *cur + && *(cur - d4 + 3) == *(cur + 3)) + { + maxLen = 4; + distances[offset + 1] = d4 - 1; + offset += 2; + d2 = d4; + } + + if (offset != 0) + { + UPDATE_maxLen + distances[offset - 2] = maxLen; + if (maxLen == lenLimit) + { + SkipMatchesSpec(lenLimit, curMatch, MF_PARAMS(p)); + MOVE_POS_RET; + } + } + + if (maxLen < 4) + maxLen = 4; + + GET_MATCHES_FOOTER(offset, maxLen) +} +*/ + +static UInt32 Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, h3, d2, d3, maxLen, offset, pos; + 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; + + maxLen = 0; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + distances[0] = maxLen = 2; + distances[1] = d2 - 1; + offset = 2; + } + + if (d2 != d3 && d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + maxLen = 3; + distances[offset + 1] = d3 - 1; + offset += 2; + d2 = d3; + } + + if (offset != 0) + { + UPDATE_maxLen + distances[offset - 2] = maxLen; + if (maxLen == lenLimit) + { + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS_RET; + } + } + + if (maxLen < 3) + maxLen = 3; + + offset = (UInt32)(Hc_GetMatchesSpec(lenLimit, curMatch, MF_PARAMS(p), + distances + offset, maxLen) - (distances)); + MOVE_POS_RET +} + +/* +static UInt32 Hc5_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 h2, h3, h4, d2, d3, d4, maxLen, offset, 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; + + maxLen = 0; + offset = 0; + + if (d2 < p->cyclicBufferSize && *(cur - d2) == *cur) + { + distances[0] = maxLen = 2; + distances[1] = d2 - 1; + offset = 2; + if (*(cur - d2 + 2) == cur[2]) + distances[0] = maxLen = 3; + else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + distances[2] = maxLen = 3; + distances[3] = d3 - 1; + offset = 4; + d2 = d3; + } + } + else if (d3 < p->cyclicBufferSize && *(cur - d3) == *cur) + { + distances[0] = maxLen = 3; + distances[1] = d3 - 1; + offset = 2; + d2 = d3; + } + + if (d2 != d4 && d4 < p->cyclicBufferSize + && *(cur - d4) == *cur + && *(cur - d4 + 3) == *(cur + 3)) + { + maxLen = 4; + distances[offset + 1] = d4 - 1; + offset += 2; + d2 = d4; + } + + if (offset != 0) + { + UPDATE_maxLen + distances[offset - 2] = maxLen; + if (maxLen == lenLimit) + { + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS_RET; + } + } + + if (maxLen < 4) + maxLen = 4; + + offset = (UInt32)(Hc_GetMatchesSpec(lenLimit, curMatch, MF_PARAMS(p), + distances + offset, maxLen) - (distances)); + MOVE_POS_RET +} +*/ + +UInt32 Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 offset; + GET_MATCHES_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + offset = (UInt32)(Hc_GetMatchesSpec(lenLimit, curMatch, MF_PARAMS(p), + distances, 2) - (distances)); + MOVE_POS_RET +} + +static void Bt2_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + SKIP_HEADER(2) + HASH2_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +void Bt3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + SKIP_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +static void Bt3_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2; + UInt32 *hash; + SKIP_HEADER(3) + HASH3_CALC; + hash = p->hash; + curMatch = hash[kFix3HashSize + hv]; + hash[h2] = + hash[kFix3HashSize + hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +static void Bt4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2, h3; + UInt32 *hash; + SKIP_HEADER(4) + HASH4_CALC; + hash = p->hash; + curMatch = hash[kFix4HashSize + hv]; + hash[ h2] = + hash[kFix3HashSize + h3] = + hash[kFix4HashSize + hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +/* +static void Bt5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2, h3, h4; + UInt32 *hash; + SKIP_HEADER(5) + HASH5_CALC; + hash = p->hash; + curMatch = hash[kFix5HashSize + hv]; + hash[ h2] = + hash[kFix3HashSize + h3] = + hash[kFix4HashSize + h4] = + hash[kFix5HashSize + hv] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} +*/ + +static void Hc4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2, h3; + UInt32 *hash; + SKIP_HEADER(4) + HASH4_CALC; + hash = p->hash; + curMatch = hash[kFix4HashSize + hv]; + hash[ h2] = + hash[kFix3HashSize + h3] = + hash[kFix4HashSize + hv] = p->pos; + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS + } + while (--num != 0); +} + +/* +static void Hc5_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 h2, h3, h4; + UInt32 *hash; + SKIP_HEADER(5) + HASH5_CALC; + hash = p->hash; + curMatch = p->hash[kFix5HashSize + hv]; + hash[ h2] = + hash[kFix3HashSize + h3] = + hash[kFix4HashSize + h4] = + hash[kFix5HashSize + hv] = p->pos; + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS + } + while (--num != 0); +} +*/ + +void Hc3Zip_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + SKIP_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hv]; + p->hash[hv] = p->pos; + p->son[p->cyclicBufferPos] = curMatch; + MOVE_POS + } + while (--num != 0); +} + +void MatchFinder_CreateVTable(CMatchFinder *p, IMatchFinder *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; + } + */ +} diff --git a/src/lzma/LzFindMt.c b/src/lzma/LzFindMt.c new file mode 100644 index 0000000..c737504 --- /dev/null +++ b/src/lzma/LzFindMt.c @@ -0,0 +1,803 @@ +/* LzFindMt.c -- multithreaded Match finder for LZ algorithms +2015-10-15 : Igor Pavlov : Public domain */ + +#include + +#include + +#include + +static void MtSync_Construct(CMtSync *p) +{ + p->wasCreated = False; + p->csWasInitialized = False; + p->csWasEntered = False; + Thread_Construct(&p->thread); + Event_Construct(&p->canStart); + Event_Construct(&p->wasStarted); + Event_Construct(&p->wasStopped); + Semaphore_Construct(&p->freeSemaphore); + Semaphore_Construct(&p->filledSemaphore); +} + +static void MtSync_GetNextBlock(CMtSync *p) +{ + if (p->needStart) + { + p->numProcessedBlocks = 1; + p->needStart = False; + p->stopWriting = False; + p->exit = False; + Event_Reset(&p->wasStarted); + Event_Reset(&p->wasStopped); + + Event_Set(&p->canStart); + Event_Wait(&p->wasStarted); + } + else + { + CriticalSection_Leave(&p->cs); + p->csWasEntered = False; + p->numProcessedBlocks++; + Semaphore_Release1(&p->freeSemaphore); + } + Semaphore_Wait(&p->filledSemaphore); + CriticalSection_Enter(&p->cs); + p->csWasEntered = True; +} + +/* MtSync_StopWriting must be called if Writing was started */ + +static void MtSync_StopWriting(CMtSync *p) +{ + UInt32 myNumBlocks = p->numProcessedBlocks; + if (!Thread_WasCreated(&p->thread) || p->needStart) + return; + p->stopWriting = True; + if (p->csWasEntered) + { + CriticalSection_Leave(&p->cs); + p->csWasEntered = False; + } + Semaphore_Release1(&p->freeSemaphore); + + Event_Wait(&p->wasStopped); + + while (myNumBlocks++ != p->numProcessedBlocks) + { + Semaphore_Wait(&p->filledSemaphore); + Semaphore_Release1(&p->freeSemaphore); + } + p->needStart = True; +} + +static void MtSync_Destruct(CMtSync *p) +{ + if (Thread_WasCreated(&p->thread)) + { + MtSync_StopWriting(p); + p->exit = True; + if (p->needStart) + Event_Set(&p->canStart); + Thread_Wait(&p->thread); + Thread_Close(&p->thread); + } + if (p->csWasInitialized) + { + CriticalSection_Delete(&p->cs); + p->csWasInitialized = False; + } + + Event_Close(&p->canStart); + Event_Close(&p->wasStarted); + Event_Close(&p->wasStopped); + Semaphore_Close(&p->freeSemaphore); + Semaphore_Close(&p->filledSemaphore); + + p->wasCreated = False; +} + +#define RINOK_THREAD(x) { if ((x) != 0) return SZ_ERROR_THREAD; } + +static SRes MtSync_Create2(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj, UInt32 numBlocks) +{ + if (p->wasCreated) + return SZ_OK; + + RINOK_THREAD(CriticalSection_Init(&p->cs)); + p->csWasInitialized = True; + + RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->canStart)); + RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->wasStarted)); + RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->wasStopped)); + + RINOK_THREAD(Semaphore_Create(&p->freeSemaphore, numBlocks, numBlocks)); + RINOK_THREAD(Semaphore_Create(&p->filledSemaphore, 0, numBlocks)); + + p->needStart = True; + + RINOK_THREAD(Thread_Create(&p->thread, startAddress, obj)); + p->wasCreated = True; + return SZ_OK; +} + +static SRes MtSync_Create(CMtSync *p, THREAD_FUNC_TYPE startAddress, void *obj, UInt32 numBlocks) +{ + SRes res = MtSync_Create2(p, startAddress, obj, numBlocks); + if (res != SZ_OK) + MtSync_Destruct(p); + return res; +} + +void MtSync_Init(CMtSync *p) { p->needStart = True; } + +#define kMtMaxValForNormalize 0xFFFFFFFF + +#define DEF_GetHeads2(name, v, action) \ + static void GetHeads ## name(const Byte *p, UInt32 pos, \ + UInt32 *hash, UInt32 hashMask, UInt32 *heads, UInt32 numHeads, const UInt32 *crc) \ + { action; for (; numHeads != 0; numHeads--) { \ + const UInt32 value = (v); p++; *heads++ = pos - hash[value]; hash[value] = pos++; } } + +#define DEF_GetHeads(name, v) DEF_GetHeads2(name, v, ;) + +DEF_GetHeads2(2, (p[0] | ((UInt32)p[1] << 8)), UNUSED_VAR(hashMask); UNUSED_VAR(crc); ) +DEF_GetHeads(3, (crc[p[0]] ^ p[1] ^ ((UInt32)p[2] << 8)) & hashMask) +DEF_GetHeads(4, (crc[p[0]] ^ p[1] ^ ((UInt32)p[2] << 8) ^ (crc[p[3]] << 5)) & hashMask) +DEF_GetHeads(4b, (crc[p[0]] ^ p[1] ^ ((UInt32)p[2] << 8) ^ ((UInt32)p[3] << 16)) & hashMask) +/* DEF_GetHeads(5, (crc[p[0]] ^ p[1] ^ ((UInt32)p[2] << 8) ^ (crc[p[3]] << 5) ^ (crc[p[4]] << 3)) & hashMask) */ + +static void HashThreadFunc(CMatchFinderMt *mt) +{ + CMtSync *p = &mt->hashSync; + for (;;) + { + UInt32 numProcessedBlocks = 0; + Event_Wait(&p->canStart); + Event_Set(&p->wasStarted); + for (;;) + { + if (p->exit) + return; + if (p->stopWriting) + { + p->numProcessedBlocks = numProcessedBlocks; + Event_Set(&p->wasStopped); + break; + } + + { + CMatchFinder *mf = mt->MatchFinder; + if (MatchFinder_NeedMove(mf)) + { + CriticalSection_Enter(&mt->btSync.cs); + CriticalSection_Enter(&mt->hashSync.cs); + { + const Byte *beforePtr = Inline_MatchFinder_GetPointerToCurrentPos(mf); + ptrdiff_t offset; + MatchFinder_MoveBlock(mf); + offset = beforePtr - Inline_MatchFinder_GetPointerToCurrentPos(mf); + mt->pointerToCurPos -= offset; + mt->buffer -= offset; + } + CriticalSection_Leave(&mt->btSync.cs); + CriticalSection_Leave(&mt->hashSync.cs); + continue; + } + + Semaphore_Wait(&p->freeSemaphore); + + MatchFinder_ReadIfRequired(mf); + if (mf->pos > (kMtMaxValForNormalize - kMtHashBlockSize)) + { + UInt32 subValue = (mf->pos - mf->historySize - 1); + MatchFinder_ReduceOffsets(mf, subValue); + MatchFinder_Normalize3(subValue, mf->hash + mf->fixedHashSize, (size_t)mf->hashMask + 1); + } + { + UInt32 *heads = mt->hashBuf + ((numProcessedBlocks++) & kMtHashNumBlocksMask) * kMtHashBlockSize; + UInt32 num = mf->streamPos - mf->pos; + heads[0] = 2; + heads[1] = num; + if (num >= mf->numHashBytes) + { + num = num - mf->numHashBytes + 1; + if (num > kMtHashBlockSize - 2) + num = kMtHashBlockSize - 2; + mt->GetHeadsFunc(mf->buffer, mf->pos, mf->hash + mf->fixedHashSize, mf->hashMask, heads + 2, num, mf->crc); + heads[0] += num; + } + mf->pos += num; + mf->buffer += num; + } + } + + Semaphore_Release1(&p->filledSemaphore); + } + } +} + +static void MatchFinderMt_GetNextBlock_Hash(CMatchFinderMt *p) +{ + MtSync_GetNextBlock(&p->hashSync); + p->hashBufPosLimit = p->hashBufPos = ((p->hashSync.numProcessedBlocks - 1) & kMtHashNumBlocksMask) * kMtHashBlockSize; + p->hashBufPosLimit += p->hashBuf[p->hashBufPos++]; + p->hashNumAvail = p->hashBuf[p->hashBufPos++]; +} + +#define kEmptyHashValue 0 + +/* #define MFMT_GM_INLINE */ + +#ifdef MFMT_GM_INLINE + +#define NO_INLINE MY_FAST_CALL + +static Int32 NO_INLINE GetMatchesSpecN(UInt32 lenLimit, UInt32 pos, const Byte *cur, CLzRef *son, + UInt32 _cyclicBufferPos, UInt32 _cyclicBufferSize, UInt32 _cutValue, + UInt32 *_distances, UInt32 _maxLen, const UInt32 *hash, Int32 limit, UInt32 size, UInt32 *posRes) +{ + do + { + UInt32 *distances = _distances + 1; + UInt32 curMatch = pos - *hash++; + + CLzRef *ptr0 = son + (_cyclicBufferPos << 1) + 1; + CLzRef *ptr1 = son + (_cyclicBufferPos << 1); + UInt32 len0 = 0, len1 = 0; + UInt32 cutValue = _cutValue; + UInt32 maxLen = _maxLen; + for (;;) + { + UInt32 delta = pos - curMatch; + if (cutValue-- == 0 || delta >= _cyclicBufferSize) + { + *ptr0 = *ptr1 = kEmptyHashValue; + break; + } + { + CLzRef *pair = son + ((_cyclicBufferPos - delta + ((delta > _cyclicBufferPos) ? _cyclicBufferSize : 0)) << 1); + const Byte *pb = cur - delta; + UInt32 len = (len0 < len1 ? len0 : len1); + if (pb[len] == cur[len]) + { + if (++len != lenLimit && pb[len] == cur[len]) + while (++len != lenLimit) + if (pb[len] != cur[len]) + break; + if (maxLen < len) + { + *distances++ = maxLen = len; + *distances++ = delta - 1; + if (len == lenLimit) + { + *ptr1 = pair[0]; + *ptr0 = pair[1]; + break; + } + } + } + if (pb[len] < cur[len]) + { + *ptr1 = curMatch; + ptr1 = pair + 1; + curMatch = *ptr1; + len1 = len; + } + else + { + *ptr0 = curMatch; + ptr0 = pair; + curMatch = *ptr0; + len0 = len; + } + } + } + pos++; + _cyclicBufferPos++; + cur++; + { + UInt32 num = (UInt32)(distances - _distances); + *_distances = num - 1; + _distances += num; + limit -= num; + } + } + while (limit > 0 && --size != 0); + *posRes = pos; + return limit; +} + +#endif + +static void BtGetMatches(CMatchFinderMt *p, UInt32 *distances) +{ + UInt32 numProcessed = 0; + UInt32 curPos = 2; + UInt32 limit = kMtBtBlockSize - (p->matchMaxLen * 2); + + distances[1] = p->hashNumAvail; + + while (curPos < limit) + { + if (p->hashBufPos == p->hashBufPosLimit) + { + MatchFinderMt_GetNextBlock_Hash(p); + distances[1] = numProcessed + p->hashNumAvail; + if (p->hashNumAvail >= p->numHashBytes) + continue; + distances[0] = curPos + p->hashNumAvail; + distances += curPos; + for (; p->hashNumAvail != 0; p->hashNumAvail--) + *distances++ = 0; + return; + } + { + UInt32 size = p->hashBufPosLimit - p->hashBufPos; + UInt32 lenLimit = p->matchMaxLen; + UInt32 pos = p->pos; + UInt32 cyclicBufferPos = p->cyclicBufferPos; + if (lenLimit >= p->hashNumAvail) + lenLimit = p->hashNumAvail; + { + UInt32 size2 = p->hashNumAvail - lenLimit + 1; + if (size2 < size) + size = size2; + size2 = p->cyclicBufferSize - cyclicBufferPos; + if (size2 < size) + size = size2; + } + + #ifndef MFMT_GM_INLINE + while (curPos < limit && size-- != 0) + { + UInt32 *startDistances = distances + curPos; + UInt32 num = (UInt32)(GetMatchesSpec1(lenLimit, pos - p->hashBuf[p->hashBufPos++], + pos, p->buffer, p->son, cyclicBufferPos, p->cyclicBufferSize, p->cutValue, + startDistances + 1, p->numHashBytes - 1) - startDistances); + *startDistances = num - 1; + curPos += num; + cyclicBufferPos++; + pos++; + p->buffer++; + } + #else + { + UInt32 posRes; + curPos = limit - GetMatchesSpecN(lenLimit, pos, p->buffer, p->son, cyclicBufferPos, p->cyclicBufferSize, p->cutValue, + distances + curPos, p->numHashBytes - 1, p->hashBuf + p->hashBufPos, (Int32)(limit - curPos), size, &posRes); + p->hashBufPos += posRes - pos; + cyclicBufferPos += posRes - pos; + p->buffer += posRes - pos; + pos = posRes; + } + #endif + + numProcessed += pos - p->pos; + p->hashNumAvail -= pos - p->pos; + p->pos = pos; + if (cyclicBufferPos == p->cyclicBufferSize) + cyclicBufferPos = 0; + p->cyclicBufferPos = cyclicBufferPos; + } + } + + distances[0] = curPos; +} + +static void BtFillBlock(CMatchFinderMt *p, UInt32 globalBlockIndex) +{ + CMtSync *sync = &p->hashSync; + if (!sync->needStart) + { + CriticalSection_Enter(&sync->cs); + sync->csWasEntered = True; + } + + BtGetMatches(p, p->btBuf + (globalBlockIndex & kMtBtNumBlocksMask) * kMtBtBlockSize); + + if (p->pos > kMtMaxValForNormalize - kMtBtBlockSize) + { + UInt32 subValue = p->pos - p->cyclicBufferSize; + MatchFinder_Normalize3(subValue, p->son, (size_t)p->cyclicBufferSize * 2); + p->pos -= subValue; + } + + if (!sync->needStart) + { + CriticalSection_Leave(&sync->cs); + sync->csWasEntered = False; + } +} + +void BtThreadFunc(CMatchFinderMt *mt) +{ + CMtSync *p = &mt->btSync; + for (;;) + { + UInt32 blockIndex = 0; + Event_Wait(&p->canStart); + Event_Set(&p->wasStarted); + for (;;) + { + if (p->exit) + return; + if (p->stopWriting) + { + p->numProcessedBlocks = blockIndex; + MtSync_StopWriting(&mt->hashSync); + Event_Set(&p->wasStopped); + break; + } + Semaphore_Wait(&p->freeSemaphore); + BtFillBlock(mt, blockIndex++); + Semaphore_Release1(&p->filledSemaphore); + } + } +} + +void MatchFinderMt_Construct(CMatchFinderMt *p) +{ + p->hashBuf = NULL; + MtSync_Construct(&p->hashSync); + MtSync_Construct(&p->btSync); +} + +static void MatchFinderMt_FreeMem(CMatchFinderMt *p, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->hashBuf); + p->hashBuf = NULL; +} + +void MatchFinderMt_Destruct(CMatchFinderMt *p, ISzAlloc *alloc) +{ + MtSync_Destruct(&p->hashSync); + MtSync_Destruct(&p->btSync); + MatchFinderMt_FreeMem(p, alloc); +} + +#define kHashBufferSize (kMtHashBlockSize * kMtHashNumBlocks) +#define kBtBufferSize (kMtBtBlockSize * kMtBtNumBlocks) + +static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE HashThreadFunc2(void *p) { HashThreadFunc((CMatchFinderMt *)p); return 0; } +static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE BtThreadFunc2(void *p) +{ + Byte allocaDummy[0x180]; + unsigned i = 0; + for (i = 0; i < 16; i++) + allocaDummy[i] = (Byte)0; + if (allocaDummy[0] == 0) + BtThreadFunc((CMatchFinderMt *)p); + return 0; +} + +SRes MatchFinderMt_Create(CMatchFinderMt *p, UInt32 historySize, UInt32 keepAddBufferBefore, + UInt32 matchMaxLen, UInt32 keepAddBufferAfter, ISzAlloc *alloc) +{ + CMatchFinder *mf = p->MatchFinder; + p->historySize = historySize; + if (kMtBtBlockSize <= matchMaxLen * 4) + return SZ_ERROR_PARAM; + if (!p->hashBuf) + { + p->hashBuf = (UInt32 *)alloc->Alloc(alloc, (kHashBufferSize + kBtBufferSize) * sizeof(UInt32)); + if (!p->hashBuf) + return SZ_ERROR_MEM; + p->btBuf = p->hashBuf + kHashBufferSize; + } + keepAddBufferBefore += (kHashBufferSize + kBtBufferSize); + keepAddBufferAfter += kMtHashBlockSize; + if (!MatchFinder_Create(mf, historySize, keepAddBufferBefore, matchMaxLen, keepAddBufferAfter, alloc)) + return SZ_ERROR_MEM; + + RINOK(MtSync_Create(&p->hashSync, HashThreadFunc2, p, kMtHashNumBlocks)); + RINOK(MtSync_Create(&p->btSync, BtThreadFunc2, p, kMtBtNumBlocks)); + return SZ_OK; +} + +/* Call it after ReleaseStream / SetStream */ +void MatchFinderMt_Init(CMatchFinderMt *p) +{ + CMatchFinder *mf = p->MatchFinder; + p->btBufPos = p->btBufPosLimit = 0; + p->hashBufPos = p->hashBufPosLimit = 0; + + /* Init without data reading. We don't want to read data in this thread */ + MatchFinder_Init_2(mf, False); + + p->pointerToCurPos = Inline_MatchFinder_GetPointerToCurrentPos(mf); + p->btNumAvailBytes = 0; + p->lzPos = p->historySize + 1; + + p->hash = mf->hash; + p->fixedHashSize = mf->fixedHashSize; + p->crc = mf->crc; + + p->son = mf->son; + p->matchMaxLen = mf->matchMaxLen; + p->numHashBytes = mf->numHashBytes; + p->pos = mf->pos; + p->buffer = mf->buffer; + p->cyclicBufferPos = mf->cyclicBufferPos; + p->cyclicBufferSize = mf->cyclicBufferSize; + p->cutValue = mf->cutValue; +} + +/* ReleaseStream is required to finish multithreading */ +void MatchFinderMt_ReleaseStream(CMatchFinderMt *p) +{ + MtSync_StopWriting(&p->btSync); + /* p->MatchFinder->ReleaseStream(); */ +} + +static void MatchFinderMt_Normalize(CMatchFinderMt *p) +{ + MatchFinder_Normalize3(p->lzPos - p->historySize - 1, p->hash, p->fixedHashSize); + p->lzPos = p->historySize + 1; +} + +static void MatchFinderMt_GetNextBlock_Bt(CMatchFinderMt *p) +{ + UInt32 blockIndex; + MtSync_GetNextBlock(&p->btSync); + blockIndex = ((p->btSync.numProcessedBlocks - 1) & kMtBtNumBlocksMask); + p->btBufPosLimit = p->btBufPos = blockIndex * kMtBtBlockSize; + p->btBufPosLimit += p->btBuf[p->btBufPos++]; + p->btNumAvailBytes = p->btBuf[p->btBufPos++]; + if (p->lzPos >= kMtMaxValForNormalize - kMtBtBlockSize) + MatchFinderMt_Normalize(p); +} + +static const Byte * MatchFinderMt_GetPointerToCurrentPos(CMatchFinderMt *p) +{ + return p->pointerToCurPos; +} + +#define GET_NEXT_BLOCK_IF_REQUIRED if (p->btBufPos == p->btBufPosLimit) MatchFinderMt_GetNextBlock_Bt(p); + +static UInt32 MatchFinderMt_GetNumAvailableBytes(CMatchFinderMt *p) +{ + GET_NEXT_BLOCK_IF_REQUIRED; + return p->btNumAvailBytes; +} + +static UInt32 * MixMatches2(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *distances) +{ + UInt32 h2, curMatch2; + UInt32 *hash = p->hash; + const Byte *cur = p->pointerToCurPos; + UInt32 lzPos = p->lzPos; + MT_HASH2_CALC + + curMatch2 = hash[h2]; + hash[h2] = lzPos; + + if (curMatch2 >= matchMinPos) + if (cur[(ptrdiff_t)curMatch2 - lzPos] == cur[0]) + { + *distances++ = 2; + *distances++ = lzPos - curMatch2 - 1; + } + + return distances; +} + +static UInt32 * MixMatches3(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *distances) +{ + UInt32 h2, h3, curMatch2, curMatch3; + UInt32 *hash = p->hash; + const Byte *cur = p->pointerToCurPos; + UInt32 lzPos = p->lzPos; + MT_HASH3_CALC + + curMatch2 = hash[ h2]; + curMatch3 = hash[kFix3HashSize + h3]; + + hash[ h2] = lzPos; + hash[kFix3HashSize + h3] = lzPos; + + if (curMatch2 >= matchMinPos && cur[(ptrdiff_t)curMatch2 - lzPos] == cur[0]) + { + distances[1] = lzPos - curMatch2 - 1; + if (cur[(ptrdiff_t)curMatch2 - lzPos + 2] == cur[2]) + { + distances[0] = 3; + return distances + 2; + } + distances[0] = 2; + distances += 2; + } + + if (curMatch3 >= matchMinPos && cur[(ptrdiff_t)curMatch3 - lzPos] == cur[0]) + { + *distances++ = 3; + *distances++ = lzPos - curMatch3 - 1; + } + + return distances; +} + +/* +static UInt32 *MixMatches4(CMatchFinderMt *p, UInt32 matchMinPos, UInt32 *distances) +{ + UInt32 h2, h3, h4, curMatch2, curMatch3, curMatch4; + UInt32 *hash = p->hash; + const Byte *cur = p->pointerToCurPos; + UInt32 lzPos = p->lzPos; + MT_HASH4_CALC + + curMatch2 = hash[ h2]; + curMatch3 = hash[kFix3HashSize + h3]; + curMatch4 = hash[kFix4HashSize + h4]; + + hash[ h2] = lzPos; + hash[kFix3HashSize + h3] = lzPos; + hash[kFix4HashSize + h4] = lzPos; + + if (curMatch2 >= matchMinPos && cur[(ptrdiff_t)curMatch2 - lzPos] == cur[0]) + { + distances[1] = lzPos - curMatch2 - 1; + if (cur[(ptrdiff_t)curMatch2 - lzPos + 2] == cur[2]) + { + distances[0] = (cur[(ptrdiff_t)curMatch2 - lzPos + 3] == cur[3]) ? 4 : 3; + return distances + 2; + } + distances[0] = 2; + distances += 2; + } + + if (curMatch3 >= matchMinPos && cur[(ptrdiff_t)curMatch3 - lzPos] == cur[0]) + { + distances[1] = lzPos - curMatch3 - 1; + if (cur[(ptrdiff_t)curMatch3 - lzPos + 3] == cur[3]) + { + distances[0] = 4; + return distances + 2; + } + distances[0] = 3; + distances += 2; + } + + if (curMatch4 >= matchMinPos) + if ( + cur[(ptrdiff_t)curMatch4 - lzPos] == cur[0] && + cur[(ptrdiff_t)curMatch4 - lzPos + 3] == cur[3] + ) + { + *distances++ = 4; + *distances++ = lzPos - curMatch4 - 1; + } + + return distances; +} +*/ + +#define INCREASE_LZ_POS p->lzPos++; p->pointerToCurPos++; + +static UInt32 MatchFinderMt2_GetMatches(CMatchFinderMt *p, UInt32 *distances) +{ + const UInt32 *btBuf = p->btBuf + p->btBufPos; + UInt32 len = *btBuf++; + p->btBufPos += 1 + len; + p->btNumAvailBytes--; + { + UInt32 i; + for (i = 0; i < len; i += 2) + { + *distances++ = *btBuf++; + *distances++ = *btBuf++; + } + } + INCREASE_LZ_POS + return len; +} + +static UInt32 MatchFinderMt_GetMatches(CMatchFinderMt *p, UInt32 *distances) +{ + const UInt32 *btBuf = p->btBuf + p->btBufPos; + UInt32 len = *btBuf++; + p->btBufPos += 1 + len; + + if (len == 0) + { + /* change for bt5 ! */ + if (p->btNumAvailBytes-- >= 4) + len = (UInt32)(p->MixMatchesFunc(p, p->lzPos - p->historySize, distances) - (distances)); + } + else + { + /* Condition: there are matches in btBuf with length < p->numHashBytes */ + UInt32 *distances2; + p->btNumAvailBytes--; + distances2 = p->MixMatchesFunc(p, p->lzPos - btBuf[1], distances); + do + { + *distances2++ = *btBuf++; + *distances2++ = *btBuf++; + } + while ((len -= 2) != 0); + len = (UInt32)(distances2 - (distances)); + } + INCREASE_LZ_POS + return len; +} + +#define SKIP_HEADER2_MT do { GET_NEXT_BLOCK_IF_REQUIRED +#define SKIP_HEADER_MT(n) SKIP_HEADER2_MT if (p->btNumAvailBytes-- >= (n)) { const Byte *cur = p->pointerToCurPos; UInt32 *hash = p->hash; +#define SKIP_FOOTER_MT } INCREASE_LZ_POS p->btBufPos += p->btBuf[p->btBufPos] + 1; } while (--num != 0); + +static void MatchFinderMt0_Skip(CMatchFinderMt *p, UInt32 num) +{ + SKIP_HEADER2_MT { p->btNumAvailBytes--; + SKIP_FOOTER_MT +} + +static void MatchFinderMt2_Skip(CMatchFinderMt *p, UInt32 num) +{ + SKIP_HEADER_MT(2) + UInt32 h2; + MT_HASH2_CALC + hash[h2] = p->lzPos; + SKIP_FOOTER_MT +} + +static void MatchFinderMt3_Skip(CMatchFinderMt *p, UInt32 num) +{ + SKIP_HEADER_MT(3) + UInt32 h2, h3; + MT_HASH3_CALC + hash[kFix3HashSize + h3] = + hash[ h2] = + p->lzPos; + SKIP_FOOTER_MT +} + +/* +static void MatchFinderMt4_Skip(CMatchFinderMt *p, UInt32 num) +{ + SKIP_HEADER_MT(4) + UInt32 h2, h3, h4; + MT_HASH4_CALC + hash[kFix4HashSize + h4] = + hash[kFix3HashSize + h3] = + hash[ h2] = + p->lzPos; + SKIP_FOOTER_MT +} +*/ + +void MatchFinderMt_CreateVTable(CMatchFinderMt *p, IMatchFinder *vTable) +{ + vTable->Init = (Mf_Init_Func)MatchFinderMt_Init; + vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinderMt_GetNumAvailableBytes; + vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinderMt_GetPointerToCurrentPos; + vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt_GetMatches; + + switch (p->MatchFinder->numHashBytes) + { + case 2: + p->GetHeadsFunc = GetHeads2; + p->MixMatchesFunc = (Mf_Mix_Matches)0; + vTable->Skip = (Mf_Skip_Func)MatchFinderMt0_Skip; + vTable->GetMatches = (Mf_GetMatches_Func)MatchFinderMt2_GetMatches; + break; + case 3: + p->GetHeadsFunc = GetHeads3; + p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches2; + vTable->Skip = (Mf_Skip_Func)MatchFinderMt2_Skip; + break; + default: + /* case 4: */ + p->GetHeadsFunc = p->MatchFinder->bigHash ? GetHeads4b : GetHeads4; + p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches3; + vTable->Skip = (Mf_Skip_Func)MatchFinderMt3_Skip; + break; + /* + default: + p->GetHeadsFunc = GetHeads5; + p->MixMatchesFunc = (Mf_Mix_Matches)MixMatches4; + vTable->Skip = (Mf_Skip_Func)MatchFinderMt4_Skip; + break; + */ + } +} diff --git a/src/lzma/Lzma2Dec.c b/src/lzma/Lzma2Dec.c new file mode 100644 index 0000000..6a08121 --- /dev/null +++ b/src/lzma/Lzma2Dec.c @@ -0,0 +1,378 @@ +/* Lzma2Dec.c -- LZMA2 Decoder +2015-11-09 : Igor Pavlov : Public domain */ + +/* #define SHOW_DEBUG_INFO */ + +#include + +#ifdef SHOW_DEBUG_INFO +#include +#endif + +#include + +#include + +/* +00000000 - EOS +00000001 U U - Uncompressed Reset Dic +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 + new prop +111uuuuu U U P P S - LZMA reset state + new prop + reset dic + + u, U - Unpack Size + P - Pack Size + S - Props +*/ + +#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_IS_UNCOMPRESSED_STATE(p) (((p)->control & LZMA2_CONTROL_LZMA) == 0) + +#define LZMA2_GET_LZMA_MODE(p) (((p)->control >> 5) & 3) +#define LZMA2_IS_THERE_PROP(mode) ((mode) >= 2) + +#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, ISzAlloc *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, ISzAlloc *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->needInitDic = True; + p->needInitState = True; + p->needInitProp = True; + LzmaDec_Init(&p->decoder); +} + +static ELzma2State Lzma2Dec_UpdateState(CLzma2Dec *p, Byte b) +{ + switch (p->state) + { + case LZMA2_STATE_CONTROL: + p->control = b; + PRF(printf("\n %4X ", (unsigned)p->decoder.dicPos)); + PRF(printf(" %2X", (unsigned)b)); + if (p->control == 0) + return LZMA2_STATE_FINISHED; + if (LZMA2_IS_UNCOMPRESSED_STATE(p)) + { + if ((p->control & 0x7F) > 2) + return LZMA2_STATE_ERROR; + p->unpackSize = 0; + } + else + p->unpackSize = (UInt32)(p->control & 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(" %8u", (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++; + PRF(printf(" %8u", (unsigned)p->packSize)); + return LZMA2_IS_THERE_PROP(LZMA2_GET_LZMA_MODE(p)) ? LZMA2_STATE_PROP: + (p->needInitProp ? LZMA2_STATE_ERROR : 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 = b / 5; + lp = b % 5; + if (lc + lp > LZMA2_LCLP_MAX) + return LZMA2_STATE_ERROR; + p->decoder.prop.lc = lc; + p->decoder.prop.lp = lp; + p->needInitProp = False; + 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, Bool initDic, Bool 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_FINISHED) + { + SizeT dicPos = p->decoder.dicPos; + + if (p->state == LZMA2_STATE_ERROR) + return SZ_ERROR_DATA; + + 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) + { + p->state = LZMA2_STATE_ERROR; + return SZ_ERROR_DATA; + } + continue; + } + + { + SizeT destSizeCur = dicLimit - dicPos; + SizeT srcSizeCur = inSize - *srcLen; + ELzmaFinishMode curFinishMode = LZMA_FINISH_ANY; + + if (p->unpackSize <= destSizeCur) + { + destSizeCur = (SizeT)p->unpackSize; + curFinishMode = LZMA_FINISH_END; + } + + if (LZMA2_IS_UNCOMPRESSED_STATE(p)) + { + if (*srcLen == inSize) + { + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + + if (p->state == LZMA2_STATE_DATA) + { + Bool initDic = (p->control == LZMA2_CONTROL_COPY_RESET_DIC); + if (initDic) + p->needInitProp = p->needInitState = True; + else if (p->needInitDic) + { + p->state = LZMA2_STATE_ERROR; + return SZ_ERROR_DATA; + } + p->needInitDic = False; + LzmaDec_InitDicAndState(&p->decoder, initDic, False); + } + + if (srcSizeCur > destSizeCur) + srcSizeCur = destSizeCur; + + if (srcSizeCur == 0) + { + p->state = LZMA2_STATE_ERROR; + return SZ_ERROR_DATA; + } + + LzmaDec_UpdateWithUncompressed(&p->decoder, src, srcSizeCur); + + src += srcSizeCur; + *srcLen += srcSizeCur; + p->unpackSize -= (UInt32)srcSizeCur; + p->state = (p->unpackSize == 0) ? LZMA2_STATE_CONTROL : LZMA2_STATE_DATA_CONT; + } + else + { + SizeT outSizeProcessed; + SRes res; + + if (p->state == LZMA2_STATE_DATA) + { + unsigned mode = LZMA2_GET_LZMA_MODE(p); + Bool initDic = (mode == 3); + Bool initState = (mode != 0); + if ((!initDic && p->needInitDic) || (!initState && p->needInitState)) + { + p->state = LZMA2_STATE_ERROR; + return SZ_ERROR_DATA; + } + + LzmaDec_InitDicAndState(&p->decoder, initDic, initState); + p->needInitDic = False; + p->needInitState = False; + p->state = LZMA2_STATE_DATA_CONT; + } + + if (srcSizeCur > p->packSize) + srcSizeCur = (SizeT)p->packSize; + + res = LzmaDec_DecodeToDic(&p->decoder, dicPos + destSizeCur, src, &srcSizeCur, curFinishMode, status); + + src += srcSizeCur; + *srcLen += srcSizeCur; + p->packSize -= (UInt32)srcSizeCur; + + outSizeProcessed = p->decoder.dicPos - dicPos; + p->unpackSize -= (UInt32)outSizeProcessed; + + RINOK(res); + if (*status == LZMA_STATUS_NEEDS_MORE_INPUT) + return res; + + if (srcSizeCur == 0 && outSizeProcessed == 0) + { + if (*status != LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK + || p->unpackSize != 0 + || p->packSize != 0) + { + p->state = LZMA2_STATE_ERROR; + return SZ_ERROR_DATA; + } + p->state = LZMA2_STATE_CONTROL; + } + + if (*status == LZMA_STATUS_MAYBE_FINISHED_WITHOUT_MARK) + *status = LZMA_STATUS_NOT_FINISHED; + } + } + } + + *status = LZMA_STATUS_FINISHED_WITH_MARK; + return SZ_OK; +} + +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 srcSizeCur = inSize, outSizeCur, dicPos; + ELzmaFinishMode curFinishMode; + SRes res; + if (p->decoder.dicPos == p->decoder.dicBufSize) + p->decoder.dicPos = 0; + dicPos = p->decoder.dicPos; + if (outSize > p->decoder.dicBufSize - dicPos) + { + outSizeCur = p->decoder.dicBufSize; + curFinishMode = LZMA_FINISH_ANY; + } + else + { + outSizeCur = dicPos + outSize; + curFinishMode = finishMode; + } + + res = Lzma2Dec_DecodeToDic(p, outSizeCur, src, &srcSizeCur, curFinishMode, status); + src += srcSizeCur; + inSize -= srcSizeCur; + *srcLen += srcSizeCur; + outSizeCur = p->decoder.dicPos - dicPos; + memcpy(dest, p->decoder.dic + dicPos, outSizeCur); + dest += outSizeCur; + outSize -= outSizeCur; + *destLen += outSizeCur; + if (res != 0) + return res; + if (outSizeCur == 0 || outSize == 0) + return SZ_OK; + } +} + +SRes Lzma2Decode(Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + Byte prop, ELzmaFinishMode finishMode, ELzmaStatus *status, ISzAlloc *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/src/lzma/Lzma2Enc.c b/src/lzma/Lzma2Enc.c new file mode 100644 index 0000000..dccf404 --- /dev/null +++ b/src/lzma/Lzma2Enc.c @@ -0,0 +1,520 @@ +/* Lzma2Enc.c -- LZMA2 Encoder +2015-10-04 : Igor Pavlov : Public domain */ + +#include + +/* #include */ +#include + +/* #define _7ZIP_ST */ + +#include + +#ifndef _7ZIP_ST +#include +#else +#define NUM_MT_CODER_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 */ + +/* ---------- CLzma2EncInt ---------- */ + +typedef struct +{ + CLzmaEncHandle enc; + UInt64 srcPos; + Byte props; + Bool needInitState; + Bool needInitProp; +} CLzma2EncInt; + +static SRes Lzma2EncInt_Init(CLzma2EncInt *p, const CLzma2EncProps *props) +{ + Byte propsEncoded[LZMA_PROPS_SIZE]; + SizeT propsSize = LZMA_PROPS_SIZE; + RINOK(LzmaEnc_SetProps(p->enc, &props->lzmaProps)); + RINOK(LzmaEnc_WriteProperties(p->enc, propsEncoded, &propsSize)); + p->srcPos = 0; + p->props = propsEncoded[0]; + p->needInitState = True; + p->needInitProp = True; + return SZ_OK; +} + +SRes LzmaEnc_PrepareForLzma2(CLzmaEncHandle pp, ISeqInStream *inStream, UInt32 keepWindowSize, + ISzAlloc *alloc, ISzAlloc *allocBig); +SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, + UInt32 keepWindowSize, ISzAlloc *alloc, ISzAlloc *allocBig); +SRes LzmaEnc_CodeOneMemBlock(CLzmaEncHandle pp, Bool 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); + + +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); + Bool 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 (outStream->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->props; + + p->needInitProp = False; + p->needInitState = False; + destPos += packSize; + p->srcPos += unpackSize; + + if (outStream) + if (outStream->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->numTotalThreads = -1; + p->numBlockThreads = -1; + p->blockSize = 0; +} + +void Lzma2EncProps_Normalize(CLzma2EncProps *p) +{ + int t1, t1n, t2, t3; + { + CLzmaEncProps lzmaProps = p->lzmaProps; + LzmaEncProps_Normalize(&lzmaProps); + t1n = lzmaProps.numThreads; + } + + t1 = p->lzmaProps.numThreads; + t2 = p->numBlockThreads; + t3 = p->numTotalThreads; + + if (t2 > NUM_MT_CODER_THREADS_MAX) + t2 = NUM_MT_CODER_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 > NUM_MT_CODER_THREADS_MAX) + t2 = NUM_MT_CODER_THREADS_MAX; + } + else if (t1 <= 0) + { + t1 = t3 / t2; + if (t1 == 0) + t1 = 1; + } + else + t3 = t1n * t2; + + p->lzmaProps.numThreads = t1; + + LzmaEncProps_Normalize(&p->lzmaProps); + + t1 = p->lzmaProps.numThreads; + + if (p->blockSize == 0) + { + UInt32 dictSize = p->lzmaProps.dictSize; + UInt64 blockSize = (UInt64)dictSize << 2; + const UInt32 kMinSize = (UInt32)1 << 20; + const UInt32 kMaxSize = (UInt32)1 << 28; + if (blockSize < kMinSize) blockSize = kMinSize; + if (blockSize > kMaxSize) blockSize = kMaxSize; + if (blockSize < dictSize) blockSize = dictSize; + p->blockSize = (size_t)blockSize; + } + + if (t2 > 1 && p->lzmaProps.reduceSize != (UInt64)(Int64)-1) + { + UInt64 temp = p->lzmaProps.reduceSize + p->blockSize - 1; + if (temp > p->lzmaProps.reduceSize) + { + UInt64 numBlocks = temp / p->blockSize; + if (numBlocks < (unsigned)t2) + { + t2 = (unsigned)numBlocks; + if (t2 == 0) + t2 = 1; + t3 = t1 * t2; + } + } + } + + p->numBlockThreads = t2; + p->numTotalThreads = t3; +} + + +static SRes Progress(ICompressProgress *p, UInt64 inSize, UInt64 outSize) +{ + return (p && p->Progress(p, inSize, outSize) != SZ_OK) ? SZ_ERROR_PROGRESS : SZ_OK; +} + + +/* ---------- Lzma2 ---------- */ + +typedef struct +{ + Byte propEncoded; + CLzma2EncProps props; + + Byte *outBuf; + + ISzAlloc *alloc; + ISzAlloc *allocBig; + + CLzma2EncInt coders[NUM_MT_CODER_THREADS_MAX]; + + #ifndef _7ZIP_ST + CMtCoder mtCoder; + #endif + +} CLzma2Enc; + + +/* ---------- Lzma2EncThread ---------- */ + +static SRes Lzma2Enc_EncodeMt1(CLzma2EncInt *p, CLzma2Enc *mainEncoder, + ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress) +{ + UInt64 packTotal = 0; + SRes res = SZ_OK; + + if (!mainEncoder->outBuf) + { + mainEncoder->outBuf = (Byte *)IAlloc_Alloc(mainEncoder->alloc, LZMA2_CHUNK_SIZE_COMPRESSED_MAX); + if (!mainEncoder->outBuf) + return SZ_ERROR_MEM; + } + + RINOK(Lzma2EncInt_Init(p, &mainEncoder->props)); + RINOK(LzmaEnc_PrepareForLzma2(p->enc, inStream, LZMA2_KEEP_WINDOW_SIZE, + mainEncoder->alloc, mainEncoder->allocBig)); + + for (;;) + { + size_t packSize = LZMA2_CHUNK_SIZE_COMPRESSED_MAX; + res = Lzma2EncInt_EncodeSubblock(p, mainEncoder->outBuf, &packSize, outStream); + if (res != SZ_OK) + break; + packTotal += packSize; + res = Progress(progress, p->srcPos, packTotal); + if (res != SZ_OK) + break; + if (packSize == 0) + break; + } + + LzmaEnc_Finish(p->enc); + + if (res == SZ_OK) + { + Byte b = 0; + if (outStream->Write(outStream, &b, 1) != 1) + return SZ_ERROR_WRITE; + } + + return res; +} + + +#ifndef _7ZIP_ST + +typedef struct +{ + IMtCoderCallback funcTable; + CLzma2Enc *lzma2Enc; +} CMtCallbackImp; + +static SRes MtCallbackImp_Code(void *pp, unsigned index, Byte *dest, size_t *destSize, + const Byte *src, size_t srcSize, int finished) +{ + CMtCallbackImp *imp = (CMtCallbackImp *)pp; + CLzma2Enc *mainEncoder = imp->lzma2Enc; + CLzma2EncInt *p = &mainEncoder->coders[index]; + + SRes res = SZ_OK; + { + size_t destLim = *destSize; + *destSize = 0; + + if (srcSize != 0) + { + RINOK(Lzma2EncInt_Init(p, &mainEncoder->props)); + + RINOK(LzmaEnc_MemPrepare(p->enc, src, srcSize, LZMA2_KEEP_WINDOW_SIZE, + mainEncoder->alloc, mainEncoder->allocBig)); + + while (p->srcPos < srcSize) + { + size_t packSize = destLim - *destSize; + res = Lzma2EncInt_EncodeSubblock(p, dest + *destSize, &packSize, NULL); + if (res != SZ_OK) + break; + *destSize += packSize; + + if (packSize == 0) + { + res = SZ_ERROR_FAIL; + break; + } + + if (MtProgress_Set(&mainEncoder->mtCoder.mtProgress, index, p->srcPos, *destSize) != SZ_OK) + { + res = SZ_ERROR_PROGRESS; + break; + } + } + + LzmaEnc_Finish(p->enc); + if (res != SZ_OK) + return res; + } + + if (finished) + { + if (*destSize == destLim) + return SZ_ERROR_OUTPUT_EOF; + dest[(*destSize)++] = 0; + } + } + return res; +} + +#endif + + +/* ---------- Lzma2Enc ---------- */ + +CLzma2EncHandle Lzma2Enc_Create(ISzAlloc *alloc, ISzAlloc *allocBig) +{ + CLzma2Enc *p = (CLzma2Enc *)alloc->Alloc(alloc, sizeof(CLzma2Enc)); + if (!p) + return NULL; + Lzma2EncProps_Init(&p->props); + Lzma2EncProps_Normalize(&p->props); + p->outBuf = 0; + p->alloc = alloc; + p->allocBig = allocBig; + { + unsigned i; + for (i = 0; i < NUM_MT_CODER_THREADS_MAX; i++) + p->coders[i].enc = 0; + } + + #ifndef _7ZIP_ST + MtCoder_Construct(&p->mtCoder); + #endif + + return p; +} + +void Lzma2Enc_Destroy(CLzma2EncHandle pp) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + unsigned i; + for (i = 0; i < NUM_MT_CODER_THREADS_MAX; i++) + { + CLzma2EncInt *t = &p->coders[i]; + if (t->enc) + { + LzmaEnc_Destroy(t->enc, p->alloc, p->allocBig); + t->enc = 0; + } + } + + #ifndef _7ZIP_ST + MtCoder_Destruct(&p->mtCoder); + #endif + + IAlloc_Free(p->alloc, p->outBuf); + IAlloc_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; +} + +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; +} + +SRes Lzma2Enc_Encode(CLzma2EncHandle pp, + ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress) +{ + CLzma2Enc *p = (CLzma2Enc *)pp; + int i; + + for (i = 0; i < p->props.numBlockThreads; i++) + { + CLzma2EncInt *t = &p->coders[(unsigned)i]; + if (!t->enc) + { + t->enc = LzmaEnc_Create(p->alloc); + if (!t->enc) + return SZ_ERROR_MEM; + } + } + + #ifndef _7ZIP_ST + if (p->props.numBlockThreads > 1) + { + CMtCallbackImp mtCallback; + + mtCallback.funcTable.Code = MtCallbackImp_Code; + mtCallback.lzma2Enc = p; + + p->mtCoder.progress = progress; + p->mtCoder.inStream = inStream; + p->mtCoder.outStream = outStream; + p->mtCoder.alloc = p->alloc; + p->mtCoder.mtCallback = &mtCallback.funcTable; + + p->mtCoder.blockSize = p->props.blockSize; + p->mtCoder.destBlockSize = p->props.blockSize + (p->props.blockSize >> 10) + 16; + if (p->mtCoder.destBlockSize < p->props.blockSize) + { + p->mtCoder.destBlockSize = (size_t)0 - 1; + if (p->mtCoder.destBlockSize < p->props.blockSize) + return SZ_ERROR_FAIL; + } + p->mtCoder.numThreads = p->props.numBlockThreads; + + return MtCoder_Code(&p->mtCoder); + } + #endif + + return Lzma2Enc_EncodeMt1(&p->coders[0], p, outStream, inStream, progress); +} diff --git a/src/lzma/Lzma86Dec.c b/src/lzma/Lzma86Dec.c new file mode 100644 index 0000000..3e2c4d7 --- /dev/null +++ b/src/lzma/Lzma86Dec.c @@ -0,0 +1,54 @@ +/* Lzma86Dec.c -- LZMA + x86 (BCJ) Filter Decoder +2016-05-16 : Igor Pavlov : Public domain */ + +#include + +#include + +#include +#include +#include + +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/src/lzma/Lzma86Enc.c b/src/lzma/Lzma86Enc.c new file mode 100644 index 0000000..8be0fe7 --- /dev/null +++ b/src/lzma/Lzma86Enc.c @@ -0,0 +1,106 @@ +/* Lzma86Enc.c -- LZMA + x86 (BCJ) Filter Encoder +2016-05-16 : Igor Pavlov : Public domain */ + +#include + +#include + +#include + +#include +#include +#include + +#define SZE_OUT_OVERFLOW SZE_DATA_ERROR + +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; + Bool 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; + Bool 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; + Bool 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/src/lzma/LzmaDec.c b/src/lzma/LzmaDec.c new file mode 100644 index 0000000..103c339 --- /dev/null +++ b/src/lzma/LzmaDec.c @@ -0,0 +1,1100 @@ +/* LzmaDec.c -- LZMA Decoder +2016-05-16 : Igor Pavlov : Public domain */ + +#include + +#include + +#include + +#define kNumTopBits 24 +#define kTopValue ((UInt32)1 << kNumTopBits) + +#define kNumBitModelTotalBits 11 +#define kBitModelTotal (1 << kNumBitModelTotalBits) +#define kNumMoveBits 5 + +#define RC_INIT_SIZE 5 + +#define NORMALIZE if (range < kTopValue) { range <<= 8; code = (code << 8) | (*buf++); } + +#define IF_BIT_0(p) ttt = *(p); NORMALIZE; bound = (range >> kNumBitModelTotalBits) * 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 GET_BIT(p, i) GET_BIT2(p, i, ; , ;) + +#define TREE_GET_BIT(probs, i) { GET_BIT((probs + i), i); } +#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 GET_BIT(prob + symbol, symbol) +#define MATCHED_LITER_DEC \ + matchByte <<= 1; \ + bit = (matchByte & offs); \ + probLit = prob + offs + bit + symbol; \ + GET_BIT2(probLit, symbol, offs &= ~bit, offs &= bit) + +#define NORMALIZE_CHECK if (range < kTopValue) { if (buf >= bufLimit) return DUMMY_ERROR; range <<= 8; code = (code << 8) | (*buf++); } + +#define IF_BIT_0_CHECK(p) ttt = *(p); NORMALIZE_CHECK; bound = (range >> kNumBitModelTotalBits) * 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 kNumPosBitsMax 4 +#define kNumPosStatesMax (1 << kNumPosBitsMax) + +#define kLenNumLowBits 3 +#define kLenNumLowSymbols (1 << kLenNumLowBits) +#define kLenNumMidBits 3 +#define kLenNumMidSymbols (1 << kLenNumMidBits) +#define kLenNumHighBits 8 +#define kLenNumHighSymbols (1 << kLenNumHighBits) + +#define LenChoice 0 +#define LenChoice2 (LenChoice + 1) +#define LenLow (LenChoice2 + 1) +#define LenMid (LenLow + (kNumPosStatesMax << kLenNumLowBits)) +#define LenHigh (LenMid + (kNumPosStatesMax << kLenNumMidBits)) +#define kNumLenProbs (LenHigh + kLenNumHighSymbols) + + +#define kNumStates 12 +#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 + kLenNumMidSymbols + kLenNumHighSymbols) + +#define IsMatch 0 +#define IsRep (IsMatch + (kNumStates << kNumPosBitsMax)) +#define IsRepG0 (IsRep + kNumStates) +#define IsRepG1 (IsRepG0 + kNumStates) +#define IsRepG2 (IsRepG1 + kNumStates) +#define IsRep0Long (IsRepG2 + kNumStates) +#define PosSlot (IsRep0Long + (kNumStates << kNumPosBitsMax)) +#define SpecPos (PosSlot + (kNumLenToPosStates << kNumPosSlotBits)) +#define Align (SpecPos + kNumFullDistances - kEndPosModelIndex) +#define LenCoder (Align + kAlignTableSize) +#define RepLenCoder (LenCoder + kNumLenProbs) +#define Literal (RepLenCoder + kNumLenProbs) + +#define LZMA_BASE_SIZE 1846 +#define LZMA_LIT_SIZE 0x300 + +#if Literal != LZMA_BASE_SIZE +StopCompilingDueBUG +#endif + +#define LzmaProps_GetNumProbs(p) (Literal + ((UInt32)LZMA_LIT_SIZE << ((p)->lc + (p)->lp))) + +#define LZMA_DIC_MIN (1 << 12) + +/* First LZMA-symbol is always decoded. +And it decodes new LZMA-symbols while (buf < bufLimit), but "buf" is without last normalization +Out: + Result: + SZ_OK - OK + SZ_ERROR_DATA - Error + p->remainLen: + < kMatchSpecLenStart : normal remain + = kMatchSpecLenStart : finished + = kMatchSpecLenStart + 1 : Flush marker (unused now) + = kMatchSpecLenStart + 2 : State Init Marker (unused now) +*/ + +static int MY_FAST_CALL LzmaDec_DecodeReal(CLzmaDec *p, SizeT limit, const Byte *bufLimit) +{ + CLzmaProb *probs = p->probs; + + unsigned state = 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 lpMask = ((unsigned)1 << (p->prop.lp)) - 1; + unsigned lc = p->prop.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 = processedPos & pbMask; + + prob = probs + IsMatch + (state << kNumPosBitsMax) + posState; + IF_BIT_0(prob) + { + unsigned symbol; + UPDATE_0(prob); + prob = probs + Literal; + if (processedPos != 0 || checkDicSize != 0) + prob += ((UInt32)LZMA_LIT_SIZE * (((processedPos & lpMask) << lc) + + (dic[(dicPos == 0 ? dicBufSize : dicPos) - 1] >> (8 - 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); + if (checkDicSize == 0 && processedPos == 0) + return SZ_ERROR_DATA; + prob = probs + IsRepG0 + state; + IF_BIT_0(prob) + { + UPDATE_0(prob); + prob = probs + IsRep0Long + (state << kNumPosBitsMax) + posState; + IF_BIT_0(prob) + { + UPDATE_0(prob); + 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 + (posState << kLenNumLowBits); + offset = 0; + lim = (1 << kLenNumLowBits); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenChoice2; + IF_BIT_0(probLen) + { + UPDATE_0(probLen); + probLen = prob + LenMid + (posState << kLenNumMidBits); + offset = kLenNumLowSymbols; + lim = (1 << kLenNumMidBits); + } + else + { + UPDATE_1(probLen); + probLen = prob + LenHigh; + offset = kLenNumLowSymbols + kLenNumMidSymbols; + 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 + (posState << kLenNumLowBits); + 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 + LenMid + (posState << kLenNumMidBits); + 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 + kLenNumMidSymbols; + } + } + } + #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 + distance - posSlot - 1; + { + UInt32 mask = 1; + unsigned i = 1; + do + { + GET_BIT2(prob + i, i, ; , distance |= mask); + mask <<= 1; + } + while (--numDirectBits != 0); + } + } + 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 != 0); + prob = probs + Align; + distance <<= kNumAlignBits; + { + unsigned i = 1; + GET_BIT2(prob + i, i, ; , distance |= 1); + GET_BIT2(prob + i, i, ; , distance |= 2); + GET_BIT2(prob + i, i, ; , distance |= 4); + GET_BIT2(prob + i, i, ; , distance |= 8); + } + if (distance == (UInt32)0xFFFFFFFF) + { + len += kMatchSpecLenStart; + state -= kNumStates; + break; + } + } + } + + rep3 = rep2; + rep2 = rep1; + rep1 = rep0; + rep0 = distance + 1; + if (checkDicSize == 0) + { + if (distance >= processedPos) + { + p->dicPos = dicPos; + return SZ_ERROR_DATA; + } + } + else if (distance >= checkDicSize) + { + p->dicPos = dicPos; + return SZ_ERROR_DATA; + } + state = (state < kNumStates + kNumLitStates) ? kNumLitStates : kNumLitStates + 3; + } + + len += kMatchMinLen; + + { + SizeT rem; + unsigned curLen; + SizeT pos; + + if ((rem = limit - dicPos) == 0) + { + p->dicPos = dicPos; + return SZ_ERROR_DATA; + } + + curLen = ((rem < len) ? (unsigned)rem : len); + pos = dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0); + + processedPos += 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 += 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 = len; + p->dicPos = dicPos; + p->processedPos = processedPos; + p->reps[0] = rep0; + p->reps[1] = rep1; + p->reps[2] = rep2; + p->reps[3] = rep3; + p->state = state; + + return SZ_OK; +} + +static void MY_FAST_CALL LzmaDec_WriteRem(CLzmaDec *p, SizeT limit) +{ + if (p->remainLen != 0 && p->remainLen < kMatchSpecLenStart) + { + Byte *dic = p->dic; + SizeT dicPos = p->dicPos; + SizeT dicBufSize = p->dicBufSize; + unsigned len = p->remainLen; + SizeT rep0 = p->reps[0]; /* we use SizeT to avoid the BUG of VC14 for AMD64 */ + SizeT rem = limit - dicPos; + if (rem < len) + len = (unsigned)(rem); + + if (p->checkDicSize == 0 && p->prop.dicSize - p->processedPos <= len) + p->checkDicSize = p->prop.dicSize; + + p->processedPos += len; + p->remainLen -= len; + while (len != 0) + { + len--; + dic[dicPos] = dic[dicPos - rep0 + (dicPos < rep0 ? dicBufSize : 0)]; + dicPos++; + } + p->dicPos = dicPos; + } +} + +static int MY_FAST_CALL LzmaDec_DecodeReal2(CLzmaDec *p, SizeT limit, const Byte *bufLimit) +{ + do + { + SizeT limit2 = limit; + if (p->checkDicSize == 0) + { + UInt32 rem = p->prop.dicSize - p->processedPos; + if (limit - p->dicPos > rem) + limit2 = p->dicPos + rem; + } + + RINOK(LzmaDec_DecodeReal(p, limit2, bufLimit)); + + if (p->checkDicSize == 0 && p->processedPos >= p->prop.dicSize) + p->checkDicSize = p->prop.dicSize; + + LzmaDec_WriteRem(p, limit); + } + while (p->dicPos < limit && p->buf < bufLimit && p->remainLen < kMatchSpecLenStart); + + if (p->remainLen > kMatchSpecLenStart) + p->remainLen = kMatchSpecLenStart; + + return 0; +} + +typedef enum +{ + DUMMY_ERROR, /* unexpected end of input stream */ + DUMMY_LIT, + DUMMY_MATCH, + DUMMY_REP +} ELzmaDummy; + +static ELzmaDummy LzmaDec_TryDummy(const CLzmaDec *p, const Byte *buf, SizeT inSize) +{ + UInt32 range = p->range; + UInt32 code = p->code; + const Byte *bufLimit = buf + inSize; + const CLzmaProb *probs = p->probs; + unsigned state = p->state; + ELzmaDummy res; + + { + const CLzmaProb *prob; + UInt32 bound; + unsigned ttt; + unsigned posState = (p->processedPos) & ((1 << p->prop.pb) - 1); + + prob = probs + IsMatch + (state << kNumPosBitsMax) + posState; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK + + /* if (bufLimit - buf >= 7) return DUMMY_LIT; */ + + prob = probs + Literal; + if (p->checkDicSize != 0 || p->processedPos != 0) + prob += ((UInt32)LZMA_LIT_SIZE * + ((((p->processedPos) & ((1 << (p->prop.lp)) - 1)) << p->prop.lc) + + (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 <<= 1; + bit = (matchByte & offs); + probLit = prob + offs + bit + symbol; + GET_BIT2_CHECK(probLit, symbol, offs &= ~bit, 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 + (state << kNumPosBitsMax) + posState; + IF_BIT_0_CHECK(prob) + { + UPDATE_0_CHECK; + NORMALIZE_CHECK; + return DUMMY_REP; + } + 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 + (posState << kLenNumLowBits); + offset = 0; + limit = 1 << kLenNumLowBits; + } + else + { + UPDATE_1_CHECK; + probLen = prob + LenChoice2; + IF_BIT_0_CHECK(probLen) + { + UPDATE_0_CHECK; + probLen = prob + LenMid + (posState << kLenNumMidBits); + offset = kLenNumLowSymbols; + limit = 1 << kLenNumMidBits; + } + else + { + UPDATE_1_CHECK; + probLen = prob + LenHigh; + offset = kLenNumLowSymbols + kLenNumMidSymbols; + limit = 1 << kLenNumHighBits; + } + } + TREE_DECODE_CHECK(probLen, limit, len); + len += offset; + } + + if (state < 4) + { + unsigned posSlot; + prob = probs + PosSlot + + ((len < kNumLenToPosStates ? len : kNumLenToPosStates - 1) << + kNumPosSlotBits); + TREE_DECODE_CHECK(prob, 1 << kNumPosSlotBits, posSlot); + if (posSlot >= kStartPosModelIndex) + { + unsigned numDirectBits = ((posSlot >> 1) - 1); + + /* if (bufLimit - buf >= 8) return DUMMY_MATCH; */ + + if (posSlot < kEndPosModelIndex) + { + prob = probs + SpecPos + ((2 | (posSlot & 1)) << numDirectBits) - posSlot - 1; + } + else + { + numDirectBits -= kNumAlignBits; + do + { + NORMALIZE_CHECK + range >>= 1; + code -= range & (((code - range) >> 31) - 1); + /* if (code >= range) code -= range; */ + } + while (--numDirectBits != 0); + prob = probs + Align; + numDirectBits = kNumAlignBits; + } + { + unsigned i = 1; + do + { + GET_BIT_CHECK(prob + i, i); + } + while (--numDirectBits != 0); + } + } + } + } + } + NORMALIZE_CHECK; + return res; +} + + +void LzmaDec_InitDicAndState(CLzmaDec *p, Bool initDic, Bool initState) +{ + p->needFlush = 1; + p->remainLen = 0; + p->tempBufSize = 0; + + if (initDic) + { + p->processedPos = 0; + p->checkDicSize = 0; + p->needInitState = 1; + } + if (initState) + p->needInitState = 1; +} + +void LzmaDec_Init(CLzmaDec *p) +{ + p->dicPos = 0; + LzmaDec_InitDicAndState(p, True, True); +} + +static void LzmaDec_InitStateReal(CLzmaDec *p) +{ + 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->needInitState = 0; +} + +SRes LzmaDec_DecodeToDic(CLzmaDec *p, SizeT dicLimit, const Byte *src, SizeT *srcLen, + ELzmaFinishMode finishMode, ELzmaStatus *status) +{ + SizeT inSize = *srcLen; + (*srcLen) = 0; + LzmaDec_WriteRem(p, dicLimit); + + *status = LZMA_STATUS_NOT_SPECIFIED; + + while (p->remainLen != kMatchSpecLenStart) + { + int checkEndMarkNow; + + if (p->needFlush) + { + for (; inSize > 0 && p->tempBufSize < RC_INIT_SIZE; (*srcLen)++, inSize--) + p->tempBuf[p->tempBufSize++] = *src++; + if (p->tempBufSize < RC_INIT_SIZE) + { + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + if (p->tempBuf[0] != 0) + return SZ_ERROR_DATA; + p->code = + ((UInt32)p->tempBuf[1] << 24) + | ((UInt32)p->tempBuf[2] << 16) + | ((UInt32)p->tempBuf[3] << 8) + | ((UInt32)p->tempBuf[4]); + p->range = 0xFFFFFFFF; + p->needFlush = 0; + p->tempBufSize = 0; + } + + 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) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_ERROR_DATA; + } + checkEndMarkNow = 1; + } + + if (p->needInitState) + LzmaDec_InitStateReal(p); + + if (p->tempBufSize == 0) + { + SizeT processed; + const Byte *bufLimit; + if (inSize < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) + { + int dummyRes = LzmaDec_TryDummy(p, src, inSize); + if (dummyRes == DUMMY_ERROR) + { + memcpy(p->tempBuf, src, inSize); + p->tempBufSize = (unsigned)inSize; + (*srcLen) += inSize; + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + if (checkEndMarkNow && dummyRes != DUMMY_MATCH) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_ERROR_DATA; + } + bufLimit = src; + } + else + bufLimit = src + inSize - LZMA_REQUIRED_INPUT_MAX; + p->buf = src; + if (LzmaDec_DecodeReal2(p, dicLimit, bufLimit) != 0) + return SZ_ERROR_DATA; + processed = (SizeT)(p->buf - src); + (*srcLen) += processed; + src += processed; + inSize -= processed; + } + else + { + unsigned rem = p->tempBufSize, lookAhead = 0; + while (rem < LZMA_REQUIRED_INPUT_MAX && lookAhead < inSize) + p->tempBuf[rem++] = src[lookAhead++]; + p->tempBufSize = rem; + if (rem < LZMA_REQUIRED_INPUT_MAX || checkEndMarkNow) + { + int dummyRes = LzmaDec_TryDummy(p, p->tempBuf, rem); + if (dummyRes == DUMMY_ERROR) + { + (*srcLen) += lookAhead; + *status = LZMA_STATUS_NEEDS_MORE_INPUT; + return SZ_OK; + } + if (checkEndMarkNow && dummyRes != DUMMY_MATCH) + { + *status = LZMA_STATUS_NOT_FINISHED; + return SZ_ERROR_DATA; + } + } + p->buf = p->tempBuf; + if (LzmaDec_DecodeReal2(p, dicLimit, p->buf) != 0) + return SZ_ERROR_DATA; + + { + unsigned kkk = (unsigned)(p->buf - p->tempBuf); + if (rem < kkk) + return SZ_ERROR_FAIL; /* some internal error */ + rem -= kkk; + if (lookAhead < rem) + return SZ_ERROR_FAIL; /* some internal error */ + lookAhead -= rem; + } + (*srcLen) += lookAhead; + src += lookAhead; + inSize -= lookAhead; + p->tempBufSize = 0; + } + } + if (p->code == 0) + *status = LZMA_STATUS_FINISHED_WITH_MARK; + return (p->code == 0) ? SZ_OK : SZ_ERROR_DATA; +} + +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, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->probs); + p->probs = NULL; +} + +static void LzmaDec_FreeDict(CLzmaDec *p, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->dic); + p->dic = NULL; +} + +void LzmaDec_Free(CLzmaDec *p, ISzAlloc *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 = d % 9; + d /= 9; + p->pb = d / 5; + p->lp = d % 5; + + return SZ_OK; +} + +static SRes LzmaDec_AllocateProbs2(CLzmaDec *p, const CLzmaProps *propNew, ISzAlloc *alloc) +{ + UInt32 numProbs = LzmaProps_GetNumProbs(propNew); + if (!p->probs || numProbs != p->numProbs) + { + LzmaDec_FreeProbs(p, alloc); + p->probs = (CLzmaProb *)alloc->Alloc(alloc, numProbs * sizeof(CLzmaProb)); + p->numProbs = numProbs; + if (!p->probs) + return SZ_ERROR_MEM; + } + return SZ_OK; +} + +SRes LzmaDec_AllocateProbs(CLzmaDec *p, const Byte *props, unsigned propsSize, ISzAlloc *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, ISzAlloc *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 *)alloc->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, ISzAlloc *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/src/lzma/LzmaEnc.c b/src/lzma/LzmaEnc.c new file mode 100644 index 0000000..83511f7 --- /dev/null +++ b/src/lzma/LzmaEnc.c @@ -0,0 +1,2351 @@ +/* LzmaEnc.c -- LZMA Encoder +2016-05-16 : Igor Pavlov : Public domain */ + +#include + +#include + +/* #define SHOW_STAT */ +/* #define SHOW_STAT2 */ + +#if defined(SHOW_STAT) || defined(SHOW_STAT2) +#include +#endif + +#include + +#include +#ifndef _7ZIP_ST +#include +#endif + +#ifdef SHOW_STAT +static unsigned g_STAT_OFFSET = 0; +#endif + +#define kMaxHistorySize ((UInt32)3 << 29) +/* #define kMaxHistorySize ((UInt32)7 << 29) */ + +#define kBlockSizeMax ((1 << LZMA_NUM_BLOCK_SIZE_BITS) - 1) + +#define kBlockSize (9 << 10) +#define kUnpackBlockSize (1 << 18) +#define kMatchArraySize (1 << 21) +#define kMatchRecordMaxSize ((LZMA_MATCH_LEN_MAX * 2 + 3) * LZMA_MATCH_LEN_MAX) + +#define kNumMaxDirectBits (31) + +#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) + +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; +} + +void LzmaEncProps_Normalize(CLzmaEncProps *p) +{ + int level = p->level; + if (level < 0) level = 5; + p->level = level; + + if (p->dictSize == 0) p->dictSize = (level <= 5 ? (1 << (level * 2 + 14)) : (level == 6 ? (1 << 25) : (1 << 26))); + if (p->dictSize > p->reduceSize) + { + unsigned i; + for (i = 11; i <= 30; i++) + { + if ((UInt32)p->reduceSize <= ((UInt32)2 << i)) { p->dictSize = ((UInt32)2 << i); break; } + if ((UInt32)p->reduceSize <= ((UInt32)3 << i)) { p->dictSize = ((UInt32)3 << i); break; } + } + } + + 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 = 4; + if (p->mc == 0) p->mc = (16 + (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; +} + +#if (_MSC_VER >= 1400) +/* BSR code is fast for some new CPUs */ +/* #define LZMA_LOG_BSR */ +#endif + +#ifdef LZMA_LOG_BSR + +#define kDicLogSizeMaxCompress 32 + +#define BSR2_RET(pos, res) { unsigned long zz; _BitScanReverse(&zz, (pos)); res = (zz + zz) + ((pos >> (zz - 1)) & 1); } + +static UInt32 GetPosSlot1(UInt32 pos) +{ + UInt32 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 + +#define kNumLogBits (9 + sizeof(size_t) / 2) +/* #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) { UInt32 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) { UInt32 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) { UInt32 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]; else BSR2_RET(pos, res); } + +#endif + + +#define LZMA_NUM_REPS 4 + +typedef unsigned CState; + +typedef struct +{ + UInt32 price; + + CState state; + int prev1IsChar; + int prev2; + + UInt32 posPrev2; + UInt32 backPrev2; + + UInt32 posPrev; + UInt32 backPrev; + UInt32 backs[LZMA_NUM_REPS]; +} COptimal; + +#define kNumOpts (1 << 12) + +#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 kNumPosModels (kEndPosModelIndex - kStartPosModelIndex) + +#define kNumFullDistances (1 << (kEndPosModelIndex >> 1)) + +#ifdef _LZMA_PROB32 +#define CLzmaProb UInt32 +#else +#define CLzmaProb UInt16 +#endif + +#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 kLenNumMidBits 3 +#define kLenNumMidSymbols (1 << kLenNumMidBits) +#define kLenNumHighBits 8 +#define kLenNumHighSymbols (1 << kLenNumHighBits) + +#define kLenNumSymbolsTotal (kLenNumLowSymbols + kLenNumMidSymbols + kLenNumHighSymbols) + +#define LZMA_MATCH_LEN_MIN 2 +#define LZMA_MATCH_LEN_MAX (LZMA_MATCH_LEN_MIN + kLenNumSymbolsTotal - 1) + +#define kNumStates 12 + + +typedef struct +{ + CLzmaProb choice; + CLzmaProb choice2; + CLzmaProb low[LZMA_NUM_PB_STATES_MAX << kLenNumLowBits]; + CLzmaProb mid[LZMA_NUM_PB_STATES_MAX << kLenNumMidBits]; + CLzmaProb high[kLenNumHighSymbols]; +} CLenEnc; + + +typedef struct +{ + CLenEnc p; + UInt32 tableSize; + UInt32 prices[LZMA_NUM_PB_STATES_MAX][kLenNumSymbolsTotal]; + UInt32 counters[LZMA_NUM_PB_STATES_MAX]; +} CLenPriceEnc; + + +typedef struct +{ + UInt32 range; + Byte cache; + UInt64 low; + UInt64 cacheSize; + Byte *buf; + Byte *bufLim; + Byte *bufBase; + ISeqOutStream *outStream; + UInt64 processed; + SRes res; +} CRangeEnc; + + +typedef struct +{ + CLzmaProb *litProbs; + + UInt32 state; + UInt32 reps[LZMA_NUM_REPS]; + + CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb isRep[kNumStates]; + CLzmaProb isRepG0[kNumStates]; + CLzmaProb isRepG1[kNumStates]; + CLzmaProb isRepG2[kNumStates]; + CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; + + CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; + CLzmaProb posEncoders[kNumFullDistances - kEndPosModelIndex]; + CLzmaProb posAlignEncoder[1 << kNumAlignBits]; + + CLenPriceEnc lenEnc; + CLenPriceEnc repLenEnc; +} CSaveState; + + +typedef struct +{ + void *matchFinderObj; + IMatchFinder matchFinder; + + UInt32 optimumEndIndex; + UInt32 optimumCurrentIndex; + + UInt32 longestMatchLength; + UInt32 numPairs; + UInt32 numAvail; + + UInt32 numFastBytes; + UInt32 additionalOffset; + UInt32 reps[LZMA_NUM_REPS]; + UInt32 state; + + unsigned lc, lp, pb; + unsigned lpMask, pbMask; + unsigned lclp; + + CLzmaProb *litProbs; + + Bool fastMode; + Bool writeEndMark; + Bool finished; + Bool multiThread; + Bool needInit; + + UInt64 nowPos64; + + UInt32 matchPriceCount; + UInt32 alignPriceCount; + + UInt32 distTableSize; + + UInt32 dictSize; + SRes result; + + CRangeEnc rc; + + #ifndef _7ZIP_ST + Bool mtMode; + CMatchFinderMt matchFinderMt; + #endif + + CMatchFinder matchFinderBase; + + #ifndef _7ZIP_ST + Byte pad[128]; + #endif + + COptimal opt[kNumOpts]; + + #ifndef LZMA_LOG_BSR + Byte g_FastPos[1 << kNumLogBits]; + #endif + + UInt32 ProbPrices[kBitModelTotal >> kNumMoveReducingBits]; + UInt32 matches[LZMA_MATCH_LEN_MAX * 2 + 2 + 1]; + + UInt32 posSlotPrices[kNumLenToPosStates][kDistTableSizeMax]; + UInt32 distancesPrices[kNumLenToPosStates][kNumFullDistances]; + UInt32 alignPrices[kAlignTableSize]; + + CLzmaProb isMatch[kNumStates][LZMA_NUM_PB_STATES_MAX]; + CLzmaProb isRep[kNumStates]; + CLzmaProb isRepG0[kNumStates]; + CLzmaProb isRepG1[kNumStates]; + CLzmaProb isRepG2[kNumStates]; + CLzmaProb isRep0Long[kNumStates][LZMA_NUM_PB_STATES_MAX]; + + CLzmaProb posSlotEncoder[kNumLenToPosStates][1 << kNumPosSlotBits]; + CLzmaProb posEncoders[kNumFullDistances - kEndPosModelIndex]; + CLzmaProb posAlignEncoder[1 << kNumAlignBits]; + + CLenPriceEnc lenEnc; + CLenPriceEnc repLenEnc; + + CSaveState saveState; + + #ifndef _7ZIP_ST + Byte pad2[128]; + #endif +} CLzmaEnc; + + +void LzmaEnc_SaveState(CLzmaEncHandle pp) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + CSaveState *dest = &p->saveState; + int i; + dest->lenEnc = p->lenEnc; + dest->repLenEnc = p->repLenEnc; + dest->state = p->state; + + for (i = 0; i < kNumStates; i++) + { + memcpy(dest->isMatch[i], p->isMatch[i], sizeof(p->isMatch[i])); + memcpy(dest->isRep0Long[i], p->isRep0Long[i], sizeof(p->isRep0Long[i])); + } + for (i = 0; i < kNumLenToPosStates; i++) + memcpy(dest->posSlotEncoder[i], p->posSlotEncoder[i], sizeof(p->posSlotEncoder[i])); + memcpy(dest->isRep, p->isRep, sizeof(p->isRep)); + memcpy(dest->isRepG0, p->isRepG0, sizeof(p->isRepG0)); + memcpy(dest->isRepG1, p->isRepG1, sizeof(p->isRepG1)); + memcpy(dest->isRepG2, p->isRepG2, sizeof(p->isRepG2)); + memcpy(dest->posEncoders, p->posEncoders, sizeof(p->posEncoders)); + memcpy(dest->posAlignEncoder, p->posAlignEncoder, sizeof(p->posAlignEncoder)); + memcpy(dest->reps, p->reps, sizeof(p->reps)); + 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; + int i; + dest->lenEnc = p->lenEnc; + dest->repLenEnc = p->repLenEnc; + dest->state = p->state; + + for (i = 0; i < kNumStates; i++) + { + memcpy(dest->isMatch[i], p->isMatch[i], sizeof(p->isMatch[i])); + memcpy(dest->isRep0Long[i], p->isRep0Long[i], sizeof(p->isRep0Long[i])); + } + for (i = 0; i < kNumLenToPosStates; i++) + memcpy(dest->posSlotEncoder[i], p->posSlotEncoder[i], sizeof(p->posSlotEncoder[i])); + memcpy(dest->isRep, p->isRep, sizeof(p->isRep)); + memcpy(dest->isRepG0, p->isRepG0, sizeof(p->isRepG0)); + memcpy(dest->isRepG1, p->isRepG1, sizeof(p->isRepG1)); + memcpy(dest->isRepG2, p->isRepG2, sizeof(p->isRepG2)); + memcpy(dest->posEncoders, p->posEncoders, sizeof(p->posEncoders)); + memcpy(dest->posAlignEncoder, p->posAlignEncoder, sizeof(p->posAlignEncoder)); + memcpy(dest->reps, p->reps, sizeof(p->reps)); + 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 + || props.dictSize > ((UInt64)1 << kDicLogSizeMaxCompress) + || props.dictSize > kMaxHistorySize) + return SZ_ERROR_PARAM; + + p->dictSize = props.dictSize; + { + unsigned fb = props.fb; + if (fb < 5) + fb = 5; + if (fb > LZMA_MATCH_LEN_MAX) + fb = LZMA_MATCH_LEN_MAX; + p->numFastBytes = fb; + } + p->lc = props.lc; + p->lp = props.lp; + p->pb = props.pb; + p->fastMode = (props.algo == 0); + p->matchFinderBase.btMode = (Byte)(props.btMode ? 1 : 0); + { + UInt32 numHashBytes = 4; + if (props.btMode) + { + if (props.numHashBytes < 2) + numHashBytes = 2; + else if (props.numHashBytes < 4) + numHashBytes = props.numHashBytes; + } + p->matchFinderBase.numHashBytes = numHashBytes; + } + + p->matchFinderBase.cutValue = props.mc; + + p->writeEndMark = props.writeEndMark; + + #ifndef _7ZIP_ST + /* + if (newMultiThread != _multiThread) + { + ReleaseMatchFinder(); + _multiThread = newMultiThread; + } + */ + p->multiThread = (props.numThreads > 1); + #endif + + return SZ_OK; +} + +static const int kLiteralNextStates[kNumStates] = {0, 0, 0, 0, 1, 2, 3, 4, 5, 6, 4, 5}; +static const int kMatchNextStates[kNumStates] = {7, 7, 7, 7, 7, 7, 7, 10, 10, 10, 10, 10}; +static const int kRepNextStates[kNumStates] = {8, 8, 8, 8, 8, 8, 8, 11, 11, 11, 11, 11}; +static const int kShortRepNextStates[kNumStates]= {9, 9, 9, 9, 9, 9, 9, 11, 11, 11, 11, 11}; + +#define IsCharState(s) ((s) < 7) + +#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 + ((p)->buf - (p)->bufBase) + (p)->cacheSize) + +#define RC_BUF_SIZE (1 << 16) +static int RangeEnc_Alloc(CRangeEnc *p, ISzAlloc *alloc) +{ + if (!p->bufBase) + { + p->bufBase = (Byte *)alloc->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, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->bufBase); + p->bufBase = 0; +} + +static void RangeEnc_Init(CRangeEnc *p) +{ + /* Stream.Init(); */ + p->low = 0; + p->range = 0xFFFFFFFF; + p->cacheSize = 1; + p->cache = 0; + + p->buf = p->bufBase; + + p->processed = 0; + p->res = SZ_OK; +} + +static void RangeEnc_FlushStream(CRangeEnc *p) +{ + size_t num; + if (p->res != SZ_OK) + return; + num = p->buf - p->bufBase; + if (num != p->outStream->Write(p->outStream, p->bufBase, num)) + p->res = SZ_ERROR_WRITE; + p->processed += num; + p->buf = p->bufBase; +} + +static void MY_FAST_CALL RangeEnc_ShiftLow(CRangeEnc *p) +{ + if ((UInt32)p->low < (UInt32)0xFF000000 || (unsigned)(p->low >> 32) != 0) + { + Byte temp = p->cache; + do + { + Byte *buf = p->buf; + *buf++ = (Byte)(temp + (Byte)(p->low >> 32)); + p->buf = buf; + if (buf == p->bufLim) + RangeEnc_FlushStream(p); + temp = 0xFF; + } + while (--p->cacheSize != 0); + p->cache = (Byte)((UInt32)p->low >> 24); + } + p->cacheSize++; + p->low = (UInt32)p->low << 8; +} + +static void RangeEnc_FlushData(CRangeEnc *p) +{ + int i; + for (i = 0; i < 5; i++) + RangeEnc_ShiftLow(p); +} + +static void RangeEnc_EncodeDirectBits(CRangeEnc *p, UInt32 value, unsigned numBits) +{ + do + { + p->range >>= 1; + p->low += p->range & (0 - ((value >> --numBits) & 1)); + if (p->range < kTopValue) + { + p->range <<= 8; + RangeEnc_ShiftLow(p); + } + } + while (numBits != 0); +} + +static void RangeEnc_EncodeBit(CRangeEnc *p, CLzmaProb *prob, UInt32 symbol) +{ + UInt32 ttt = *prob; + UInt32 newBound = (p->range >> kNumBitModelTotalBits) * ttt; + if (symbol == 0) + { + p->range = newBound; + ttt += (kBitModelTotal - ttt) >> kNumMoveBits; + } + else + { + p->low += newBound; + p->range -= newBound; + ttt -= ttt >> kNumMoveBits; + } + *prob = (CLzmaProb)ttt; + if (p->range < kTopValue) + { + p->range <<= 8; + RangeEnc_ShiftLow(p); + } +} + +static void LitEnc_Encode(CRangeEnc *p, CLzmaProb *probs, UInt32 symbol) +{ + symbol |= 0x100; + do + { + RangeEnc_EncodeBit(p, probs + (symbol >> 8), (symbol >> 7) & 1); + symbol <<= 1; + } + while (symbol < 0x10000); +} + +static void LitEnc_EncodeMatched(CRangeEnc *p, CLzmaProb *probs, UInt32 symbol, UInt32 matchByte) +{ + UInt32 offs = 0x100; + symbol |= 0x100; + do + { + matchByte <<= 1; + RangeEnc_EncodeBit(p, probs + (offs + (matchByte & offs) + (symbol >> 8)), (symbol >> 7) & 1); + symbol <<= 1; + offs &= ~(matchByte ^ symbol); + } + while (symbol < 0x10000); +} + +static void LzmaEnc_InitPriceTables(UInt32 *ProbPrices) +{ + UInt32 i; + for (i = (1 << kNumMoveReducingBits) / 2; i < kBitModelTotal; i += (1 << kNumMoveReducingBits)) + { + const int kCyclesBits = kNumBitPriceShiftBits; + UInt32 w = i; + UInt32 bitCount = 0; + int j; + for (j = 0; j < kCyclesBits; j++) + { + w = w * w; + bitCount <<= 1; + while (w >= ((UInt32)1 << 16)) + { + w >>= 1; + bitCount++; + } + } + ProbPrices[i >> kNumMoveReducingBits] = ((kNumBitModelTotalBits << kCyclesBits) - 15 - bitCount); + } +} + + +#define GET_PRICE(prob, symbol) \ + p->ProbPrices[((prob) ^ (((-(int)(symbol))) & (kBitModelTotal - 1))) >> kNumMoveReducingBits]; + +#define GET_PRICEa(prob, symbol) \ + ProbPrices[((prob) ^ ((-((int)(symbol))) & (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_PRICE_0a(prob) ProbPrices[(prob) >> kNumMoveReducingBits] +#define GET_PRICE_1a(prob) ProbPrices[((prob) ^ (kBitModelTotal - 1)) >> kNumMoveReducingBits] + +static UInt32 LitEnc_GetPrice(const CLzmaProb *probs, UInt32 symbol, const UInt32 *ProbPrices) +{ + UInt32 price = 0; + symbol |= 0x100; + do + { + price += GET_PRICEa(probs[symbol >> 8], (symbol >> 7) & 1); + symbol <<= 1; + } + while (symbol < 0x10000); + return price; +} + +static UInt32 LitEnc_GetPriceMatched(const CLzmaProb *probs, UInt32 symbol, UInt32 matchByte, const UInt32 *ProbPrices) +{ + UInt32 price = 0; + UInt32 offs = 0x100; + symbol |= 0x100; + do + { + matchByte <<= 1; + price += GET_PRICEa(probs[offs + (matchByte & offs) + (symbol >> 8)], (symbol >> 7) & 1); + symbol <<= 1; + offs &= ~(matchByte ^ symbol); + } + while (symbol < 0x10000); + return price; +} + + +static void RcTree_Encode(CRangeEnc *rc, CLzmaProb *probs, int numBitLevels, UInt32 symbol) +{ + UInt32 m = 1; + int i; + for (i = numBitLevels; i != 0;) + { + UInt32 bit; + i--; + bit = (symbol >> i) & 1; + RangeEnc_EncodeBit(rc, probs + m, bit); + m = (m << 1) | bit; + } +} + +static void RcTree_ReverseEncode(CRangeEnc *rc, CLzmaProb *probs, int numBitLevels, UInt32 symbol) +{ + UInt32 m = 1; + int i; + for (i = 0; i < numBitLevels; i++) + { + UInt32 bit = symbol & 1; + RangeEnc_EncodeBit(rc, probs + m, bit); + m = (m << 1) | bit; + symbol >>= 1; + } +} + +static UInt32 RcTree_GetPrice(const CLzmaProb *probs, int numBitLevels, UInt32 symbol, const UInt32 *ProbPrices) +{ + UInt32 price = 0; + symbol |= (1 << numBitLevels); + while (symbol != 1) + { + price += GET_PRICEa(probs[symbol >> 1], symbol & 1); + symbol >>= 1; + } + return price; +} + +static UInt32 RcTree_ReverseGetPrice(const CLzmaProb *probs, int numBitLevels, UInt32 symbol, const UInt32 *ProbPrices) +{ + UInt32 price = 0; + UInt32 m = 1; + int i; + for (i = numBitLevels; i != 0; i--) + { + UInt32 bit = symbol & 1; + symbol >>= 1; + price += GET_PRICEa(probs[m], bit); + m = (m << 1) | bit; + } + return price; +} + + +static void LenEnc_Init(CLenEnc *p) +{ + unsigned i; + p->choice = p->choice2 = kProbInitValue; + for (i = 0; i < (LZMA_NUM_PB_STATES_MAX << kLenNumLowBits); i++) + p->low[i] = kProbInitValue; + for (i = 0; i < (LZMA_NUM_PB_STATES_MAX << kLenNumMidBits); i++) + p->mid[i] = kProbInitValue; + for (i = 0; i < kLenNumHighSymbols; i++) + p->high[i] = kProbInitValue; +} + +static void LenEnc_Encode(CLenEnc *p, CRangeEnc *rc, UInt32 symbol, UInt32 posState) +{ + if (symbol < kLenNumLowSymbols) + { + RangeEnc_EncodeBit(rc, &p->choice, 0); + RcTree_Encode(rc, p->low + (posState << kLenNumLowBits), kLenNumLowBits, symbol); + } + else + { + RangeEnc_EncodeBit(rc, &p->choice, 1); + if (symbol < kLenNumLowSymbols + kLenNumMidSymbols) + { + RangeEnc_EncodeBit(rc, &p->choice2, 0); + RcTree_Encode(rc, p->mid + (posState << kLenNumMidBits), kLenNumMidBits, symbol - kLenNumLowSymbols); + } + else + { + RangeEnc_EncodeBit(rc, &p->choice2, 1); + RcTree_Encode(rc, p->high, kLenNumHighBits, symbol - kLenNumLowSymbols - kLenNumMidSymbols); + } + } +} + +static void LenEnc_SetPrices(CLenEnc *p, UInt32 posState, UInt32 numSymbols, UInt32 *prices, const UInt32 *ProbPrices) +{ + UInt32 a0 = GET_PRICE_0a(p->choice); + UInt32 a1 = GET_PRICE_1a(p->choice); + UInt32 b0 = a1 + GET_PRICE_0a(p->choice2); + UInt32 b1 = a1 + GET_PRICE_1a(p->choice2); + UInt32 i = 0; + for (i = 0; i < kLenNumLowSymbols; i++) + { + if (i >= numSymbols) + return; + prices[i] = a0 + RcTree_GetPrice(p->low + (posState << kLenNumLowBits), kLenNumLowBits, i, ProbPrices); + } + for (; i < kLenNumLowSymbols + kLenNumMidSymbols; i++) + { + if (i >= numSymbols) + return; + prices[i] = b0 + RcTree_GetPrice(p->mid + (posState << kLenNumMidBits), kLenNumMidBits, i - kLenNumLowSymbols, ProbPrices); + } + for (; i < numSymbols; i++) + prices[i] = b1 + RcTree_GetPrice(p->high, kLenNumHighBits, i - kLenNumLowSymbols - kLenNumMidSymbols, ProbPrices); +} + +static void MY_FAST_CALL LenPriceEnc_UpdateTable(CLenPriceEnc *p, UInt32 posState, const UInt32 *ProbPrices) +{ + LenEnc_SetPrices(&p->p, posState, p->tableSize, p->prices[posState], ProbPrices); + p->counters[posState] = p->tableSize; +} + +static void LenPriceEnc_UpdateTables(CLenPriceEnc *p, UInt32 numPosStates, const UInt32 *ProbPrices) +{ + UInt32 posState; + for (posState = 0; posState < numPosStates; posState++) + LenPriceEnc_UpdateTable(p, posState, ProbPrices); +} + +static void LenEnc_Encode2(CLenPriceEnc *p, CRangeEnc *rc, UInt32 symbol, UInt32 posState, Bool updatePrice, const UInt32 *ProbPrices) +{ + LenEnc_Encode(&p->p, rc, symbol, posState); + if (updatePrice) + if (--p->counters[posState] == 0) + LenPriceEnc_UpdateTable(p, posState, ProbPrices); +} + + + + +static void MovePos(CLzmaEnc *p, UInt32 num) +{ + #ifdef SHOW_STAT + g_STAT_OFFSET += num; + printf("\n MovePos %u", num); + #endif + + if (num != 0) + { + p->additionalOffset += num; + p->matchFinder.Skip(p->matchFinderObj, num); + } +} + +static UInt32 ReadMatchDistances(CLzmaEnc *p, UInt32 *numDistancePairsRes) +{ + UInt32 lenRes = 0, numPairs; + p->numAvail = p->matchFinder.GetNumAvailableBytes(p->matchFinderObj); + numPairs = p->matchFinder.GetMatches(p->matchFinderObj, p->matches); + + #ifdef SHOW_STAT + printf("\n i = %u numPairs = %u ", g_STAT_OFFSET, numPairs / 2); + g_STAT_OFFSET++; + { + UInt32 i; + for (i = 0; i < numPairs; i += 2) + printf("%2u %6u | ", p->matches[i], p->matches[i + 1]); + } + #endif + + if (numPairs > 0) + { + lenRes = p->matches[numPairs - 2]; + if (lenRes == p->numFastBytes) + { + UInt32 numAvail = p->numAvail; + if (numAvail > LZMA_MATCH_LEN_MAX) + numAvail = LZMA_MATCH_LEN_MAX; + { + const Byte *pbyCur = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + const Byte *pby = pbyCur + lenRes; + ptrdiff_t dif = (ptrdiff_t)-1 - p->matches[numPairs - 1]; + const Byte *pbyLim = pbyCur + numAvail; + for (; pby != pbyLim && *pby == pby[dif]; pby++); + lenRes = (UInt32)(pby - pbyCur); + } + } + } + p->additionalOffset++; + *numDistancePairsRes = numPairs; + return lenRes; +} + + +#define MakeAsChar(p) (p)->backPrev = (UInt32)(-1); (p)->prev1IsChar = False; +#define MakeAsShortRep(p) (p)->backPrev = 0; (p)->prev1IsChar = False; +#define IsShortRep(p) ((p)->backPrev == 0) + +static UInt32 GetRepLen1Price(CLzmaEnc *p, UInt32 state, UInt32 posState) +{ + return + GET_PRICE_0(p->isRepG0[state]) + + GET_PRICE_0(p->isRep0Long[state][posState]); +} + +static UInt32 GetPureRepPrice(CLzmaEnc *p, UInt32 repIndex, UInt32 state, UInt32 posState) +{ + UInt32 price; + if (repIndex == 0) + { + price = GET_PRICE_0(p->isRepG0[state]); + price += GET_PRICE_1(p->isRep0Long[state][posState]); + } + else + { + price = GET_PRICE_1(p->isRepG0[state]); + if (repIndex == 1) + price += GET_PRICE_0(p->isRepG1[state]); + else + { + price += GET_PRICE_1(p->isRepG1[state]); + price += GET_PRICE(p->isRepG2[state], repIndex - 2); + } + } + return price; +} + +static UInt32 GetRepPrice(CLzmaEnc *p, UInt32 repIndex, UInt32 len, UInt32 state, UInt32 posState) +{ + return p->repLenEnc.prices[posState][len - LZMA_MATCH_LEN_MIN] + + GetPureRepPrice(p, repIndex, state, posState); +} + +static UInt32 Backward(CLzmaEnc *p, UInt32 *backRes, UInt32 cur) +{ + UInt32 posMem = p->opt[cur].posPrev; + UInt32 backMem = p->opt[cur].backPrev; + p->optimumEndIndex = cur; + do + { + if (p->opt[cur].prev1IsChar) + { + MakeAsChar(&p->opt[posMem]) + p->opt[posMem].posPrev = posMem - 1; + if (p->opt[cur].prev2) + { + p->opt[posMem - 1].prev1IsChar = False; + p->opt[posMem - 1].posPrev = p->opt[cur].posPrev2; + p->opt[posMem - 1].backPrev = p->opt[cur].backPrev2; + } + } + { + UInt32 posPrev = posMem; + UInt32 backCur = backMem; + + backMem = p->opt[posPrev].backPrev; + posMem = p->opt[posPrev].posPrev; + + p->opt[posPrev].backPrev = backCur; + p->opt[posPrev].posPrev = cur; + cur = posPrev; + } + } + while (cur != 0); + *backRes = p->opt[0].backPrev; + p->optimumCurrentIndex = p->opt[0].posPrev; + return p->optimumCurrentIndex; +} + +#define LIT_PROBS(pos, prevByte) (p->litProbs + ((((pos) & p->lpMask) << p->lc) + ((prevByte) >> (8 - p->lc))) * (UInt32)0x300) + +static UInt32 GetOptimum(CLzmaEnc *p, UInt32 position, UInt32 *backRes) +{ + UInt32 lenEnd, cur; + UInt32 reps[LZMA_NUM_REPS], repLens[LZMA_NUM_REPS]; + UInt32 *matches; + + { + + UInt32 numAvail, mainLen, numPairs, repMaxIndex, i, posState, len; + UInt32 matchPrice, repMatchPrice, normalMatchPrice; + const Byte *data; + Byte curByte, matchByte; + + if (p->optimumEndIndex != p->optimumCurrentIndex) + { + const COptimal *opt = &p->opt[p->optimumCurrentIndex]; + UInt32 lenRes = opt->posPrev - p->optimumCurrentIndex; + *backRes = opt->backPrev; + p->optimumCurrentIndex = opt->posPrev; + return lenRes; + } + p->optimumCurrentIndex = p->optimumEndIndex = 0; + + if (p->additionalOffset == 0) + mainLen = ReadMatchDistances(p, &numPairs); + else + { + mainLen = p->longestMatchLength; + numPairs = p->numPairs; + } + + numAvail = p->numAvail; + if (numAvail < 2) + { + *backRes = (UInt32)(-1); + 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++) + { + UInt32 lenTest; + const Byte *data2; + reps[i] = p->reps[i]; + data2 = data - reps[i] - 1; + if (data[0] != data2[0] || data[1] != data2[1]) + { + repLens[i] = 0; + continue; + } + for (lenTest = 2; lenTest < numAvail && data[lenTest] == data2[lenTest]; lenTest++); + repLens[i] = lenTest; + if (lenTest > repLens[repMaxIndex]) + repMaxIndex = i; + } + if (repLens[repMaxIndex] >= p->numFastBytes) + { + UInt32 lenRes; + *backRes = repMaxIndex; + lenRes = repLens[repMaxIndex]; + MovePos(p, lenRes - 1); + return lenRes; + } + + matches = p->matches; + if (mainLen >= p->numFastBytes) + { + *backRes = matches[numPairs - 1] + LZMA_NUM_REPS; + MovePos(p, mainLen - 1); + return mainLen; + } + curByte = *data; + matchByte = *(data - (reps[0] + 1)); + + if (mainLen < 2 && curByte != matchByte && repLens[repMaxIndex] < 2) + { + *backRes = (UInt32)-1; + 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]) + + (!IsCharState(p->state) ? + LitEnc_GetPriceMatched(probs, curByte, matchByte, p->ProbPrices) : + LitEnc_GetPrice(probs, curByte, p->ProbPrices)); + } + + MakeAsChar(&p->opt[1]); + + matchPrice = GET_PRICE_1(p->isMatch[p->state][posState]); + repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[p->state]); + + if (matchByte == curByte) + { + UInt32 shortRepPrice = repMatchPrice + GetRepLen1Price(p, p->state, posState); + if (shortRepPrice < p->opt[1].price) + { + p->opt[1].price = shortRepPrice; + MakeAsShortRep(&p->opt[1]); + } + } + lenEnd = ((mainLen >= repLens[repMaxIndex]) ? mainLen : repLens[repMaxIndex]); + + if (lenEnd < 2) + { + *backRes = p->opt[1].backPrev; + return 1; + } + + p->opt[1].posPrev = 0; + for (i = 0; i < LZMA_NUM_REPS; i++) + p->opt[0].backs[i] = reps[i]; + + len = lenEnd; + do + p->opt[len--].price = kInfinityPrice; + while (len >= 2); + + for (i = 0; i < LZMA_NUM_REPS; i++) + { + UInt32 repLen = repLens[i]; + UInt32 price; + if (repLen < 2) + continue; + price = repMatchPrice + GetPureRepPrice(p, i, p->state, posState); + do + { + UInt32 curAndLenPrice = price + p->repLenEnc.prices[posState][repLen - 2]; + COptimal *opt = &p->opt[repLen]; + if (curAndLenPrice < opt->price) + { + opt->price = curAndLenPrice; + opt->posPrev = 0; + opt->backPrev = i; + opt->prev1IsChar = False; + } + } + while (--repLen >= 2); + } + + normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[p->state]); + + len = ((repLens[0] >= 2) ? repLens[0] + 1 : 2); + if (len <= mainLen) + { + UInt32 offs = 0; + while (len > matches[offs]) + offs += 2; + for (; ; len++) + { + COptimal *opt; + UInt32 distance = matches[offs + 1]; + + UInt32 curAndLenPrice = normalMatchPrice + p->lenEnc.prices[posState][len - LZMA_MATCH_LEN_MIN]; + UInt32 lenToPosState = GetLenToPosState(len); + if (distance < kNumFullDistances) + curAndLenPrice += p->distancesPrices[lenToPosState][distance]; + else + { + UInt32 slot; + GetPosSlot2(distance, slot); + curAndLenPrice += p->alignPrices[distance & kAlignMask] + p->posSlotPrices[lenToPosState][slot]; + } + opt = &p->opt[len]; + if (curAndLenPrice < opt->price) + { + opt->price = curAndLenPrice; + opt->posPrev = 0; + opt->backPrev = distance + LZMA_NUM_REPS; + opt->prev1IsChar = False; + } + 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 <= lenEnd; i++) + printf("\nprice[%4X] = %u", position - cur + i, p->opt[i].price); + } + #endif + + } + + for (;;) + { + UInt32 numAvail; + UInt32 numAvailFull, newLen, numPairs, posPrev, state, posState, startLen; + UInt32 curPrice, curAnd1Price, matchPrice, repMatchPrice; + Bool nextIsChar; + Byte curByte, matchByte; + const Byte *data; + COptimal *curOpt; + COptimal *nextOpt; + + cur++; + if (cur == lenEnd) + return Backward(p, backRes, cur); + + newLen = ReadMatchDistances(p, &numPairs); + if (newLen >= p->numFastBytes) + { + p->numPairs = numPairs; + p->longestMatchLength = newLen; + return Backward(p, backRes, cur); + } + position++; + curOpt = &p->opt[cur]; + posPrev = curOpt->posPrev; + if (curOpt->prev1IsChar) + { + posPrev--; + if (curOpt->prev2) + { + state = p->opt[curOpt->posPrev2].state; + if (curOpt->backPrev2 < LZMA_NUM_REPS) + state = kRepNextStates[state]; + else + state = kMatchNextStates[state]; + } + else + state = p->opt[posPrev].state; + state = kLiteralNextStates[state]; + } + else + state = p->opt[posPrev].state; + if (posPrev == cur - 1) + { + if (IsShortRep(curOpt)) + state = kShortRepNextStates[state]; + else + state = kLiteralNextStates[state]; + } + else + { + UInt32 pos; + const COptimal *prevOpt; + if (curOpt->prev1IsChar && curOpt->prev2) + { + posPrev = curOpt->posPrev2; + pos = curOpt->backPrev2; + state = kRepNextStates[state]; + } + else + { + pos = curOpt->backPrev; + if (pos < LZMA_NUM_REPS) + state = kRepNextStates[state]; + else + state = kMatchNextStates[state]; + } + prevOpt = &p->opt[posPrev]; + if (pos < LZMA_NUM_REPS) + { + UInt32 i; + reps[0] = prevOpt->backs[pos]; + for (i = 1; i <= pos; i++) + reps[i] = prevOpt->backs[i - 1]; + for (; i < LZMA_NUM_REPS; i++) + reps[i] = prevOpt->backs[i]; + } + else + { + UInt32 i; + reps[0] = (pos - LZMA_NUM_REPS); + for (i = 1; i < LZMA_NUM_REPS; i++) + reps[i] = prevOpt->backs[i - 1]; + } + } + curOpt->state = (CState)state; + + curOpt->backs[0] = reps[0]; + curOpt->backs[1] = reps[1]; + curOpt->backs[2] = reps[2]; + curOpt->backs[3] = reps[3]; + + curPrice = curOpt->price; + nextIsChar = False; + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + curByte = *data; + matchByte = *(data - (reps[0] + 1)); + + posState = (position & p->pbMask); + + curAnd1Price = curPrice + GET_PRICE_0(p->isMatch[state][posState]); + { + const CLzmaProb *probs = LIT_PROBS(position, *(data - 1)); + curAnd1Price += + (!IsCharState(state) ? + LitEnc_GetPriceMatched(probs, curByte, matchByte, p->ProbPrices) : + LitEnc_GetPrice(probs, curByte, p->ProbPrices)); + } + + nextOpt = &p->opt[cur + 1]; + + if (curAnd1Price < nextOpt->price) + { + nextOpt->price = curAnd1Price; + nextOpt->posPrev = cur; + MakeAsChar(nextOpt); + nextIsChar = True; + } + + matchPrice = curPrice + GET_PRICE_1(p->isMatch[state][posState]); + repMatchPrice = matchPrice + GET_PRICE_1(p->isRep[state]); + + if (matchByte == curByte && !(nextOpt->posPrev < cur && nextOpt->backPrev == 0)) + { + UInt32 shortRepPrice = repMatchPrice + GetRepLen1Price(p, state, posState); + if (shortRepPrice <= nextOpt->price) + { + nextOpt->price = shortRepPrice; + nextOpt->posPrev = cur; + MakeAsShortRep(nextOpt); + nextIsChar = True; + } + } + numAvailFull = p->numAvail; + { + UInt32 temp = kNumOpts - 1 - cur; + if (temp < numAvailFull) + numAvailFull = temp; + } + + if (numAvailFull < 2) + continue; + numAvail = (numAvailFull <= p->numFastBytes ? numAvailFull : p->numFastBytes); + + if (!nextIsChar && matchByte != curByte) /* speed optimization */ + { + /* try Literal + rep0 */ + UInt32 temp; + UInt32 lenTest2; + const Byte *data2 = data - reps[0] - 1; + UInt32 limit = p->numFastBytes + 1; + if (limit > numAvailFull) + limit = numAvailFull; + + for (temp = 1; temp < limit && data[temp] == data2[temp]; temp++); + lenTest2 = temp - 1; + if (lenTest2 >= 2) + { + UInt32 state2 = kLiteralNextStates[state]; + UInt32 posStateNext = (position + 1) & p->pbMask; + UInt32 nextRepMatchPrice = curAnd1Price + + GET_PRICE_1(p->isMatch[state2][posStateNext]) + + GET_PRICE_1(p->isRep[state2]); + /* for (; lenTest2 >= 2; lenTest2--) */ + { + UInt32 curAndLenPrice; + COptimal *opt; + UInt32 offset = cur + 1 + lenTest2; + while (lenEnd < offset) + p->opt[++lenEnd].price = kInfinityPrice; + curAndLenPrice = nextRepMatchPrice + GetRepPrice(p, 0, lenTest2, state2, posStateNext); + opt = &p->opt[offset]; + if (curAndLenPrice < opt->price) + { + opt->price = curAndLenPrice; + opt->posPrev = cur + 1; + opt->backPrev = 0; + opt->prev1IsChar = True; + opt->prev2 = False; + } + } + } + } + + startLen = 2; /* speed optimization */ + { + UInt32 repIndex; + for (repIndex = 0; repIndex < LZMA_NUM_REPS; repIndex++) + { + UInt32 lenTest; + UInt32 lenTestTemp; + UInt32 price; + const Byte *data2 = data - reps[repIndex] - 1; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + for (lenTest = 2; lenTest < numAvail && data[lenTest] == data2[lenTest]; lenTest++); + while (lenEnd < cur + lenTest) + p->opt[++lenEnd].price = kInfinityPrice; + lenTestTemp = lenTest; + price = repMatchPrice + GetPureRepPrice(p, repIndex, state, posState); + do + { + UInt32 curAndLenPrice = price + p->repLenEnc.prices[posState][lenTest - 2]; + COptimal *opt = &p->opt[cur + lenTest]; + if (curAndLenPrice < opt->price) + { + opt->price = curAndLenPrice; + opt->posPrev = cur; + opt->backPrev = repIndex; + opt->prev1IsChar = False; + } + } + while (--lenTest >= 2); + lenTest = lenTestTemp; + + if (repIndex == 0) + startLen = lenTest + 1; + + /* if (_maxMode) */ + { + UInt32 lenTest2 = lenTest + 1; + UInt32 limit = lenTest2 + p->numFastBytes; + if (limit > numAvailFull) + limit = numAvailFull; + for (; lenTest2 < limit && data[lenTest2] == data2[lenTest2]; lenTest2++); + lenTest2 -= lenTest + 1; + if (lenTest2 >= 2) + { + UInt32 nextRepMatchPrice; + UInt32 state2 = kRepNextStates[state]; + UInt32 posStateNext = (position + lenTest) & p->pbMask; + UInt32 curAndLenCharPrice = + price + p->repLenEnc.prices[posState][lenTest - 2] + + GET_PRICE_0(p->isMatch[state2][posStateNext]) + + LitEnc_GetPriceMatched(LIT_PROBS(position + lenTest, data[lenTest - 1]), + data[lenTest], data2[lenTest], p->ProbPrices); + state2 = kLiteralNextStates[state2]; + posStateNext = (position + lenTest + 1) & p->pbMask; + nextRepMatchPrice = curAndLenCharPrice + + GET_PRICE_1(p->isMatch[state2][posStateNext]) + + GET_PRICE_1(p->isRep[state2]); + + /* for (; lenTest2 >= 2; lenTest2--) */ + { + UInt32 curAndLenPrice; + COptimal *opt; + UInt32 offset = cur + lenTest + 1 + lenTest2; + while (lenEnd < offset) + p->opt[++lenEnd].price = kInfinityPrice; + curAndLenPrice = nextRepMatchPrice + GetRepPrice(p, 0, lenTest2, state2, posStateNext); + opt = &p->opt[offset]; + if (curAndLenPrice < opt->price) + { + opt->price = curAndLenPrice; + opt->posPrev = cur + lenTest + 1; + opt->backPrev = 0; + opt->prev1IsChar = True; + opt->prev2 = True; + opt->posPrev2 = cur; + opt->backPrev2 = repIndex; + } + } + } + } + } + } + /* for (UInt32 lenTest = 2; lenTest <= newLen; lenTest++) */ + if (newLen > numAvail) + { + newLen = numAvail; + for (numPairs = 0; newLen > matches[numPairs]; numPairs += 2); + matches[numPairs] = newLen; + numPairs += 2; + } + if (newLen >= startLen) + { + UInt32 normalMatchPrice = matchPrice + GET_PRICE_0(p->isRep[state]); + UInt32 offs, curBack, posSlot; + UInt32 lenTest; + while (lenEnd < cur + newLen) + p->opt[++lenEnd].price = kInfinityPrice; + + offs = 0; + while (startLen > matches[offs]) + offs += 2; + curBack = matches[offs + 1]; + GetPosSlot2(curBack, posSlot); + for (lenTest = /*2*/ startLen; ; lenTest++) + { + UInt32 curAndLenPrice = normalMatchPrice + p->lenEnc.prices[posState][lenTest - LZMA_MATCH_LEN_MIN]; + { + UInt32 lenToPosState = GetLenToPosState(lenTest); + COptimal *opt; + if (curBack < kNumFullDistances) + curAndLenPrice += p->distancesPrices[lenToPosState][curBack]; + else + curAndLenPrice += p->posSlotPrices[lenToPosState][posSlot] + p->alignPrices[curBack & kAlignMask]; + + opt = &p->opt[cur + lenTest]; + if (curAndLenPrice < opt->price) + { + opt->price = curAndLenPrice; + opt->posPrev = cur; + opt->backPrev = curBack + LZMA_NUM_REPS; + opt->prev1IsChar = False; + } + } + + if (/*_maxMode && */lenTest == matches[offs]) + { + /* Try Match + Literal + Rep0 */ + const Byte *data2 = data - curBack - 1; + UInt32 lenTest2 = lenTest + 1; + UInt32 limit = lenTest2 + p->numFastBytes; + if (limit > numAvailFull) + limit = numAvailFull; + for (; lenTest2 < limit && data[lenTest2] == data2[lenTest2]; lenTest2++); + lenTest2 -= lenTest + 1; + if (lenTest2 >= 2) + { + UInt32 nextRepMatchPrice; + UInt32 state2 = kMatchNextStates[state]; + UInt32 posStateNext = (position + lenTest) & p->pbMask; + UInt32 curAndLenCharPrice = curAndLenPrice + + GET_PRICE_0(p->isMatch[state2][posStateNext]) + + LitEnc_GetPriceMatched(LIT_PROBS(position + lenTest, data[lenTest - 1]), + data[lenTest], data2[lenTest], p->ProbPrices); + state2 = kLiteralNextStates[state2]; + posStateNext = (posStateNext + 1) & p->pbMask; + nextRepMatchPrice = curAndLenCharPrice + + GET_PRICE_1(p->isMatch[state2][posStateNext]) + + GET_PRICE_1(p->isRep[state2]); + + /* for (; lenTest2 >= 2; lenTest2--) */ + { + UInt32 offset = cur + lenTest + 1 + lenTest2; + UInt32 curAndLenPrice2; + COptimal *opt; + while (lenEnd < offset) + p->opt[++lenEnd].price = kInfinityPrice; + curAndLenPrice2 = nextRepMatchPrice + GetRepPrice(p, 0, lenTest2, state2, posStateNext); + opt = &p->opt[offset]; + if (curAndLenPrice2 < opt->price) + { + opt->price = curAndLenPrice2; + opt->posPrev = cur + lenTest + 1; + opt->backPrev = 0; + opt->prev1IsChar = True; + opt->prev2 = True; + opt->posPrev2 = cur; + opt->backPrev2 = curBack + LZMA_NUM_REPS; + } + } + } + offs += 2; + if (offs == numPairs) + break; + curBack = matches[offs + 1]; + if (curBack >= kNumFullDistances) + GetPosSlot2(curBack, posSlot); + } + } + } + } +} + +#define ChangePair(smallDist, bigDist) (((bigDist) >> 7) > (smallDist)) + +static UInt32 GetOptimumFast(CLzmaEnc *p, UInt32 *backRes) +{ + UInt32 numAvail, mainLen, mainDist, numPairs, repIndex, repLen, i; + const Byte *data; + const UInt32 *matches; + + if (p->additionalOffset == 0) + mainLen = ReadMatchDistances(p, &numPairs); + else + { + mainLen = p->longestMatchLength; + numPairs = p->numPairs; + } + + numAvail = p->numAvail; + *backRes = (UInt32)-1; + if (numAvail < 2) + return 1; + 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++) + { + UInt32 len; + const Byte *data2 = data - p->reps[i] - 1; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + for (len = 2; len < numAvail && data[len] == data2[len]; len++); + if (len >= p->numFastBytes) + { + *backRes = i; + MovePos(p, len - 1); + return len; + } + if (len > repLen) + { + repIndex = i; + repLen = len; + } + } + + matches = p->matches; + if (mainLen >= p->numFastBytes) + { + *backRes = matches[numPairs - 1] + LZMA_NUM_REPS; + MovePos(p, mainLen - 1); + return mainLen; + } + + mainDist = 0; /* for GCC */ + if (mainLen >= 2) + { + mainDist = matches[numPairs - 1]; + while (numPairs > 2 && mainLen == matches[numPairs - 4] + 1) + { + if (!ChangePair(matches[numPairs - 3], mainDist)) + break; + numPairs -= 2; + mainLen = matches[numPairs - 2]; + mainDist = matches[numPairs - 1]; + } + if (mainLen == 2 && mainDist >= 0x80) + mainLen = 1; + } + + if (repLen >= 2 && ( + (repLen + 1 >= mainLen) || + (repLen + 2 >= mainLen && mainDist >= (1 << 9)) || + (repLen + 3 >= mainLen && mainDist >= (1 << 15)))) + { + *backRes = repIndex; + MovePos(p, repLen - 1); + return repLen; + } + + if (mainLen < 2 || numAvail <= 2) + return 1; + + p->longestMatchLength = ReadMatchDistances(p, &p->numPairs); + if (p->longestMatchLength >= 2) + { + UInt32 newDistance = matches[p->numPairs - 1]; + if ((p->longestMatchLength >= mainLen && newDistance < mainDist) || + (p->longestMatchLength == mainLen + 1 && !ChangePair(mainDist, newDistance)) || + (p->longestMatchLength > mainLen + 1) || + (p->longestMatchLength + 1 >= mainLen && mainLen >= 3 && ChangePair(newDistance, mainDist))) + return 1; + } + + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - 1; + for (i = 0; i < LZMA_NUM_REPS; i++) + { + UInt32 len, limit; + const Byte *data2 = data - p->reps[i] - 1; + if (data[0] != data2[0] || data[1] != data2[1]) + continue; + limit = mainLen - 1; + for (len = 2; len < limit && data[len] == data2[len]; len++); + if (len >= limit) + return 1; + } + *backRes = mainDist + LZMA_NUM_REPS; + MovePos(p, mainLen - 2); + return mainLen; +} + +static void WriteEndMarker(CLzmaEnc *p, UInt32 posState) +{ + UInt32 len; + RangeEnc_EncodeBit(&p->rc, &p->isMatch[p->state][posState], 1); + RangeEnc_EncodeBit(&p->rc, &p->isRep[p->state], 0); + p->state = kMatchNextStates[p->state]; + len = LZMA_MATCH_LEN_MIN; + LenEnc_Encode2(&p->lenEnc, &p->rc, len - LZMA_MATCH_LEN_MIN, posState, !p->fastMode, p->ProbPrices); + RcTree_Encode(&p->rc, p->posSlotEncoder[GetLenToPosState(len)], kNumPosSlotBits, (1 << kNumPosSlotBits) - 1); + RangeEnc_EncodeDirectBits(&p->rc, (((UInt32)1 << 30) - 1) >> kNumAlignBits, 30 - kNumAlignBits); + RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, kAlignMask); +} + +static SRes CheckErrors(CLzmaEnc *p) +{ + if (p->result != SZ_OK) + return p->result; + if (p->rc.res != SZ_OK) + p->result = SZ_ERROR_WRITE; + if (p->matchFinderBase.result != SZ_OK) + p->result = SZ_ERROR_READ; + if (p->result != SZ_OK) + p->finished = True; + return p->result; +} + +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); +} + +static void FillAlignPrices(CLzmaEnc *p) +{ + UInt32 i; + for (i = 0; i < kAlignTableSize; i++) + p->alignPrices[i] = RcTree_ReverseGetPrice(p->posAlignEncoder, kNumAlignBits, i, p->ProbPrices); + p->alignPriceCount = 0; +} + +static void FillDistancesPrices(CLzmaEnc *p) +{ + UInt32 tempPrices[kNumFullDistances]; + UInt32 i, lenToPosState; + for (i = kStartPosModelIndex; i < kNumFullDistances; i++) + { + UInt32 posSlot = GetPosSlot1(i); + UInt32 footerBits = ((posSlot >> 1) - 1); + UInt32 base = ((2 | (posSlot & 1)) << footerBits); + tempPrices[i] = RcTree_ReverseGetPrice(p->posEncoders + base - posSlot - 1, footerBits, i - base, p->ProbPrices); + } + + for (lenToPosState = 0; lenToPosState < kNumLenToPosStates; lenToPosState++) + { + UInt32 posSlot; + const CLzmaProb *encoder = p->posSlotEncoder[lenToPosState]; + UInt32 *posSlotPrices = p->posSlotPrices[lenToPosState]; + for (posSlot = 0; posSlot < p->distTableSize; posSlot++) + posSlotPrices[posSlot] = RcTree_GetPrice(encoder, kNumPosSlotBits, posSlot, p->ProbPrices); + for (posSlot = kEndPosModelIndex; posSlot < p->distTableSize; posSlot++) + posSlotPrices[posSlot] += ((((posSlot >> 1) - 1) - kNumAlignBits) << kNumBitPriceShiftBits); + + { + UInt32 *distancesPrices = p->distancesPrices[lenToPosState]; + for (i = 0; i < kStartPosModelIndex; i++) + distancesPrices[i] = posSlotPrices[i]; + for (; i < kNumFullDistances; i++) + distancesPrices[i] = posSlotPrices[GetPosSlot1(i)] + tempPrices[i]; + } + } + p->matchPriceCount = 0; +} + +void LzmaEnc_Construct(CLzmaEnc *p) +{ + RangeEnc_Construct(&p->rc); + MatchFinder_Construct(&p->matchFinderBase); + + #ifndef _7ZIP_ST + MatchFinderMt_Construct(&p->matchFinderMt); + p->matchFinderMt.MatchFinder = &p->matchFinderBase; + #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(ISzAlloc *alloc) +{ + void *p; + p = alloc->Alloc(alloc, sizeof(CLzmaEnc)); + if (p) + LzmaEnc_Construct((CLzmaEnc *)p); + return p; +} + +void LzmaEnc_FreeLits(CLzmaEnc *p, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->litProbs); + alloc->Free(alloc, p->saveState.litProbs); + p->litProbs = NULL; + p->saveState.litProbs = NULL; +} + +void LzmaEnc_Destruct(CLzmaEnc *p, ISzAlloc *alloc, ISzAlloc *allocBig) +{ + #ifndef _7ZIP_ST + MatchFinderMt_Destruct(&p->matchFinderMt, allocBig); + #endif + + MatchFinder_Free(&p->matchFinderBase, allocBig); + LzmaEnc_FreeLits(p, alloc); + RangeEnc_Free(&p->rc, alloc); +} + +void LzmaEnc_Destroy(CLzmaEncHandle p, ISzAlloc *alloc, ISzAlloc *allocBig) +{ + LzmaEnc_Destruct((CLzmaEnc *)p, alloc, allocBig); + alloc->Free(alloc, p); +} + +static SRes LzmaEnc_CodeOneBlock(CLzmaEnc *p, Bool useLimits, UInt32 maxPackSize, UInt32 maxUnpackSize) +{ + UInt32 nowPos32, startPos32; + if (p->needInit) + { + 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) + { + UInt32 numPairs; + Byte curByte; + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) + return Flush(p, nowPos32); + ReadMatchDistances(p, &numPairs); + RangeEnc_EncodeBit(&p->rc, &p->isMatch[p->state][0], 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 pos, len, posState; + + if (p->fastMode) + len = GetOptimumFast(p, &pos); + else + len = GetOptimum(p, nowPos32, &pos); + + #ifdef SHOW_STAT2 + printf("\n pos = %4X, len = %u pos = %u", nowPos32, len, pos); + #endif + + posState = nowPos32 & p->pbMask; + if (len == 1 && pos == (UInt32)-1) + { + Byte curByte; + CLzmaProb *probs; + const Byte *data; + + RangeEnc_EncodeBit(&p->rc, &p->isMatch[p->state][posState], 0); + data = p->matchFinder.GetPointerToCurrentPos(p->matchFinderObj) - p->additionalOffset; + curByte = *data; + probs = LIT_PROBS(nowPos32, *(data - 1)); + if (IsCharState(p->state)) + LitEnc_Encode(&p->rc, probs, curByte); + else + LitEnc_EncodeMatched(&p->rc, probs, curByte, *(data - p->reps[0] - 1)); + p->state = kLiteralNextStates[p->state]; + } + else + { + RangeEnc_EncodeBit(&p->rc, &p->isMatch[p->state][posState], 1); + if (pos < LZMA_NUM_REPS) + { + RangeEnc_EncodeBit(&p->rc, &p->isRep[p->state], 1); + if (pos == 0) + { + RangeEnc_EncodeBit(&p->rc, &p->isRepG0[p->state], 0); + RangeEnc_EncodeBit(&p->rc, &p->isRep0Long[p->state][posState], ((len == 1) ? 0 : 1)); + } + else + { + UInt32 distance = p->reps[pos]; + RangeEnc_EncodeBit(&p->rc, &p->isRepG0[p->state], 1); + if (pos == 1) + RangeEnc_EncodeBit(&p->rc, &p->isRepG1[p->state], 0); + else + { + RangeEnc_EncodeBit(&p->rc, &p->isRepG1[p->state], 1); + RangeEnc_EncodeBit(&p->rc, &p->isRepG2[p->state], pos - 2); + if (pos == 3) + p->reps[3] = p->reps[2]; + p->reps[2] = p->reps[1]; + } + p->reps[1] = p->reps[0]; + p->reps[0] = distance; + } + if (len == 1) + p->state = kShortRepNextStates[p->state]; + else + { + LenEnc_Encode2(&p->repLenEnc, &p->rc, len - LZMA_MATCH_LEN_MIN, posState, !p->fastMode, p->ProbPrices); + p->state = kRepNextStates[p->state]; + } + } + else + { + UInt32 posSlot; + RangeEnc_EncodeBit(&p->rc, &p->isRep[p->state], 0); + p->state = kMatchNextStates[p->state]; + LenEnc_Encode2(&p->lenEnc, &p->rc, len - LZMA_MATCH_LEN_MIN, posState, !p->fastMode, p->ProbPrices); + pos -= LZMA_NUM_REPS; + GetPosSlot(pos, posSlot); + RcTree_Encode(&p->rc, p->posSlotEncoder[GetLenToPosState(len)], kNumPosSlotBits, posSlot); + + if (posSlot >= kStartPosModelIndex) + { + UInt32 footerBits = ((posSlot >> 1) - 1); + UInt32 base = ((2 | (posSlot & 1)) << footerBits); + UInt32 posReduced = pos - base; + + if (posSlot < kEndPosModelIndex) + RcTree_ReverseEncode(&p->rc, p->posEncoders + base - posSlot - 1, footerBits, posReduced); + else + { + RangeEnc_EncodeDirectBits(&p->rc, posReduced >> kNumAlignBits, footerBits - kNumAlignBits); + RcTree_ReverseEncode(&p->rc, p->posAlignEncoder, kNumAlignBits, posReduced & kAlignMask); + p->alignPriceCount++; + } + } + p->reps[3] = p->reps[2]; + p->reps[2] = p->reps[1]; + p->reps[1] = p->reps[0]; + p->reps[0] = pos; + p->matchPriceCount++; + } + } + p->additionalOffset -= len; + nowPos32 += len; + if (p->additionalOffset == 0) + { + UInt32 processed; + if (!p->fastMode) + { + if (p->matchPriceCount >= (1 << 7)) + FillDistancesPrices(p); + if (p->alignPriceCount >= kAlignTableSize) + FillAlignPrices(p); + } + if (p->matchFinder.GetNumAvailableBytes(p->matchFinderObj) == 0) + break; + processed = nowPos32 - startPos32; + if (useLimits) + { + if (processed + kNumOpts + 300 >= maxUnpackSize || + RangeEnc_GetProcessed(&p->rc) + kNumOpts * 2 >= 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, ISzAlloc *alloc, ISzAlloc *allocBig) +{ + UInt32 beforeSize = kNumOpts; + if (!RangeEnc_Alloc(&p->rc, alloc)) + return SZ_ERROR_MEM; + + #ifndef _7ZIP_ST + p->mtMode = (p->multiThread && !p->fastMode && (p->matchFinderBase.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 *)alloc->Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); + p->saveState.litProbs = (CLzmaProb *)alloc->Alloc(alloc, ((UInt32)0x300 << lclp) * sizeof(CLzmaProb)); + if (!p->litProbs || !p->saveState.litProbs) + { + LzmaEnc_FreeLits(p, alloc); + return SZ_ERROR_MEM; + } + p->lclp = lclp; + } + } + + p->matchFinderBase.bigHash = (Byte)(p->dictSize > kBigHashDicLimit ? 1 : 0); + + if (beforeSize + p->dictSize < keepWindowSize) + beforeSize = keepWindowSize - p->dictSize; + + #ifndef _7ZIP_ST + if (p->mtMode) + { + RINOK(MatchFinderMt_Create(&p->matchFinderMt, p->dictSize, beforeSize, p->numFastBytes, LZMA_MATCH_LEN_MAX, allocBig)); + p->matchFinderObj = &p->matchFinderMt; + MatchFinderMt_CreateVTable(&p->matchFinderMt, &p->matchFinder); + } + else + #endif + { + if (!MatchFinder_Create(&p->matchFinderBase, p->dictSize, beforeSize, p->numFastBytes, LZMA_MATCH_LEN_MAX, allocBig)) + return SZ_ERROR_MEM; + p->matchFinderObj = &p->matchFinderBase; + MatchFinder_CreateVTable(&p->matchFinderBase, &p->matchFinder); + } + + return SZ_OK; +} + +void LzmaEnc_Init(CLzmaEnc *p) +{ + UInt32 i; + p->state = 0; + for (i = 0 ; i < LZMA_NUM_REPS; i++) + p->reps[i] = 0; + + RangeEnc_Init(&p->rc); + + + for (i = 0; i < kNumStates; i++) + { + UInt32 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; + } + + { + UInt32 num = (UInt32)0x300 << (p->lp + p->lc); + CLzmaProb *probs = p->litProbs; + for (i = 0; i < num; i++) + probs[i] = kProbInitValue; + } + + { + for (i = 0; i < kNumLenToPosStates; i++) + { + CLzmaProb *probs = p->posSlotEncoder[i]; + UInt32 j; + for (j = 0; j < (1 << kNumPosSlotBits); j++) + probs[j] = kProbInitValue; + } + } + { + for (i = 0; i < kNumFullDistances - kEndPosModelIndex; i++) + p->posEncoders[i] = kProbInitValue; + } + + LenEnc_Init(&p->lenEnc.p); + LenEnc_Init(&p->repLenEnc.p); + + for (i = 0; i < (1 << kNumAlignBits); i++) + p->posAlignEncoder[i] = kProbInitValue; + + p->optimumEndIndex = 0; + p->optimumCurrentIndex = 0; + p->additionalOffset = 0; + + p->pbMask = (1 << p->pb) - 1; + p->lpMask = (1 << p->lp) - 1; +} + +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; + LenPriceEnc_UpdateTables(&p->lenEnc, 1 << p->pb, p->ProbPrices); + LenPriceEnc_UpdateTables(&p->repLenEnc, 1 << p->pb, p->ProbPrices); +} + +static SRes LzmaEnc_AllocAndInit(CLzmaEnc *p, UInt32 keepWindowSize, ISzAlloc *alloc, ISzAlloc *allocBig) +{ + UInt32 i; + for (i = 0; i < (UInt32)kDicLogSizeMaxCompress; 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, + ISzAlloc *alloc, ISzAlloc *allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + p->matchFinderBase.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, + ISzAlloc *alloc, ISzAlloc *allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + p->matchFinderBase.stream = inStream; + p->needInit = 1; + return LzmaEnc_AllocAndInit(p, keepWindowSize, alloc, allocBig); +} + +static void LzmaEnc_SetInputBuf(CLzmaEnc *p, const Byte *src, SizeT srcLen) +{ + p->matchFinderBase.directInput = 1; + p->matchFinderBase.bufferBase = (Byte *)src; + p->matchFinderBase.directInputRem = srcLen; +} + +SRes LzmaEnc_MemPrepare(CLzmaEncHandle pp, const Byte *src, SizeT srcLen, + UInt32 keepWindowSize, ISzAlloc *alloc, ISzAlloc *allocBig) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + LzmaEnc_SetInputBuf(p, src, srcLen); + p->needInit = 1; + + 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 funcTable; + Byte *data; + SizeT rem; + Bool overflow; +} CSeqOutStreamBuf; + +static size_t MyWrite(void *pp, const void *data, size_t size) +{ + CSeqOutStreamBuf *p = (CSeqOutStreamBuf *)pp; + if (p->rem < size) + { + size = p->rem; + p->overflow = True; + } + 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, Bool reInit, + Byte *dest, size_t *destLen, UInt32 desiredPackSize, UInt32 *unpackSize) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + UInt64 nowPos64; + SRes res; + CSeqOutStreamBuf outStream; + + outStream.funcTable.Write = MyWrite; + 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.funcTable; + + res = LzmaEnc_CodeOneBlock(p, True, desiredPackSize, *unpackSize); + + *unpackSize = (UInt32)(p->nowPos64 - nowPos64); + *destLen -= outStream.rem; + if (outStream.overflow) + return SZ_ERROR_OUTPUT_EOF; + + return res; +} + + +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, False, 0, 0); + if (res != SZ_OK || p->finished) + break; + if (progress) + { + res = progress->Progress(progress, p->nowPos64, RangeEnc_GetProcessed(&p->rc)); + if (res != SZ_OK) + { + res = SZ_ERROR_PROGRESS; + break; + } + } + } + + LzmaEnc_Finish(p); + + /* + if (res == S_OK && !Inline_MatchFinder_IsFinishedOK(&p->matchFinderBase)) + res = SZ_ERROR_FAIL; + } + */ + + return res; +} + + +SRes LzmaEnc_Encode(CLzmaEncHandle pp, ISeqOutStream *outStream, ISeqInStream *inStream, ICompressProgress *progress, + ISzAlloc *alloc, ISzAlloc *allocBig) +{ + RINOK(LzmaEnc_Prepare(pp, outStream, inStream, alloc, allocBig)); + return LzmaEnc_Encode2((CLzmaEnc *)pp, progress); +} + + +SRes LzmaEnc_WriteProperties(CLzmaEncHandle pp, Byte *props, SizeT *size) +{ + CLzmaEnc *p = (CLzmaEnc *)pp; + unsigned i; + UInt32 dictSize = p->dictSize; + if (*size < LZMA_PROPS_SIZE) + return SZ_ERROR_PARAM; + *size = LZMA_PROPS_SIZE; + props[0] = (Byte)((p->pb * 5 + p->lp) * 9 + p->lc); + + if (dictSize >= ((UInt32)1 << 22)) + { + UInt32 kDictMask = ((UInt32)1 << 20) - 1; + if (dictSize < (UInt32)0xFFFFFFFF - kDictMask) + dictSize = (dictSize + kDictMask) & ~kDictMask; + } + else for (i = 11; i <= 30; i++) + { + if (dictSize <= ((UInt32)2 << i)) { dictSize = (2 << i); break; } + if (dictSize <= ((UInt32)3 << i)) { dictSize = (3 << i); break; } + } + + for (i = 0; i < 4; i++) + props[1 + i] = (Byte)(dictSize >> (8 * i)); + return SZ_OK; +} + + +SRes LzmaEnc_MemEncode(CLzmaEncHandle pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT srcLen, + int writeEndMark, ICompressProgress *progress, ISzAlloc *alloc, ISzAlloc *allocBig) +{ + SRes res; + CLzmaEnc *p = (CLzmaEnc *)pp; + + CSeqOutStreamBuf outStream; + + outStream.funcTable.Write = MyWrite; + outStream.data = dest; + outStream.rem = *destLen; + outStream.overflow = False; + + p->writeEndMark = writeEndMark; + p->rc.outStream = &outStream.funcTable; + + 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, ISzAlloc *alloc, ISzAlloc *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; +} diff --git a/src/lzma/LzmaLib.c b/src/lzma/LzmaLib.c new file mode 100644 index 0000000..76ecdcb --- /dev/null +++ b/src/lzma/LzmaLib.c @@ -0,0 +1,40 @@ +/* LzmaLib.c -- LZMA library wrapper +2015-06-13 : Igor Pavlov : Public domain */ + +#include +#include +#include +#include + +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/src/lzma/MtCoder.c b/src/lzma/MtCoder.c new file mode 100644 index 0000000..3cb6780 --- /dev/null +++ b/src/lzma/MtCoder.c @@ -0,0 +1,327 @@ +/* MtCoder.c -- Multi-thread Coder +2015-10-13 : Igor Pavlov : Public domain */ + +#include + +#include + +void LoopThread_Construct(CLoopThread *p) +{ + Thread_Construct(&p->thread); + Event_Construct(&p->startEvent); + Event_Construct(&p->finishedEvent); +} + +void LoopThread_Close(CLoopThread *p) +{ + Thread_Close(&p->thread); + Event_Close(&p->startEvent); + Event_Close(&p->finishedEvent); +} + +static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE LoopThreadFunc(void *pp) +{ + CLoopThread *p = (CLoopThread *)pp; + for (;;) + { + if (Event_Wait(&p->startEvent) != 0) + return SZ_ERROR_THREAD; + if (p->stop) + return 0; + p->res = p->func(p->param); + if (Event_Set(&p->finishedEvent) != 0) + return SZ_ERROR_THREAD; + } +} + +WRes LoopThread_Create(CLoopThread *p) +{ + p->stop = 0; + RINOK(AutoResetEvent_CreateNotSignaled(&p->startEvent)); + RINOK(AutoResetEvent_CreateNotSignaled(&p->finishedEvent)); + return Thread_Create(&p->thread, LoopThreadFunc, p); +} + +WRes LoopThread_StopAndWait(CLoopThread *p) +{ + p->stop = 1; + if (Event_Set(&p->startEvent) != 0) + return SZ_ERROR_THREAD; + return Thread_Wait(&p->thread); +} + +WRes LoopThread_StartSubThread(CLoopThread *p) { return Event_Set(&p->startEvent); } +WRes LoopThread_WaitSubThread(CLoopThread *p) { return Event_Wait(&p->finishedEvent); } + +static SRes Progress(ICompressProgress *p, UInt64 inSize, UInt64 outSize) +{ + return (p && p->Progress(p, inSize, outSize) != SZ_OK) ? SZ_ERROR_PROGRESS : SZ_OK; +} + +static void MtProgress_Init(CMtProgress *p, ICompressProgress *progress) +{ + unsigned i; + for (i = 0; i < NUM_MT_CODER_THREADS_MAX; i++) + p->inSizes[i] = p->outSizes[i] = 0; + p->totalInSize = p->totalOutSize = 0; + p->progress = progress; + p->res = SZ_OK; +} + +static void MtProgress_Reinit(CMtProgress *p, unsigned index) +{ + p->inSizes[index] = 0; + p->outSizes[index] = 0; +} + +#define UPDATE_PROGRESS(size, prev, total) \ + if (size != (UInt64)(Int64)-1) { total += size - prev; prev = size; } + +SRes MtProgress_Set(CMtProgress *p, unsigned index, UInt64 inSize, UInt64 outSize) +{ + SRes res; + CriticalSection_Enter(&p->cs); + UPDATE_PROGRESS(inSize, p->inSizes[index], p->totalInSize) + UPDATE_PROGRESS(outSize, p->outSizes[index], p->totalOutSize) + if (p->res == SZ_OK) + p->res = Progress(p->progress, p->totalInSize, p->totalOutSize); + res = p->res; + CriticalSection_Leave(&p->cs); + return res; +} + +static void MtProgress_SetError(CMtProgress *p, SRes res) +{ + CriticalSection_Enter(&p->cs); + if (p->res == SZ_OK) + p->res = res; + CriticalSection_Leave(&p->cs); +} + +static void MtCoder_SetError(CMtCoder* p, SRes res) +{ + CriticalSection_Enter(&p->cs); + if (p->res == SZ_OK) + p->res = res; + CriticalSection_Leave(&p->cs); +} + +/* ---------- MtThread ---------- */ + +void CMtThread_Construct(CMtThread *p, CMtCoder *mtCoder) +{ + p->mtCoder = mtCoder; + p->outBuf = 0; + p->inBuf = 0; + Event_Construct(&p->canRead); + Event_Construct(&p->canWrite); + LoopThread_Construct(&p->thread); +} + +#define RINOK_THREAD(x) { if ((x) != 0) return SZ_ERROR_THREAD; } + +static void CMtThread_CloseEvents(CMtThread *p) +{ + Event_Close(&p->canRead); + Event_Close(&p->canWrite); +} + +static void CMtThread_Destruct(CMtThread *p) +{ + CMtThread_CloseEvents(p); + + if (Thread_WasCreated(&p->thread.thread)) + { + LoopThread_StopAndWait(&p->thread); + LoopThread_Close(&p->thread); + } + + if (p->mtCoder->alloc) + IAlloc_Free(p->mtCoder->alloc, p->outBuf); + p->outBuf = 0; + + if (p->mtCoder->alloc) + IAlloc_Free(p->mtCoder->alloc, p->inBuf); + p->inBuf = 0; +} + +#define MY_BUF_ALLOC(buf, size, newSize) \ + if (buf == 0 || size != newSize) \ + { IAlloc_Free(p->mtCoder->alloc, buf); \ + size = newSize; buf = (Byte *)IAlloc_Alloc(p->mtCoder->alloc, size); \ + if (buf == 0) return SZ_ERROR_MEM; } + +static SRes CMtThread_Prepare(CMtThread *p) +{ + MY_BUF_ALLOC(p->inBuf, p->inBufSize, p->mtCoder->blockSize) + MY_BUF_ALLOC(p->outBuf, p->outBufSize, p->mtCoder->destBlockSize) + + p->stopReading = False; + p->stopWriting = False; + RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->canRead)); + RINOK_THREAD(AutoResetEvent_CreateNotSignaled(&p->canWrite)); + + return SZ_OK; +} + +static SRes FullRead(ISeqInStream *stream, Byte *data, size_t *processedSize) +{ + size_t size = *processedSize; + *processedSize = 0; + while (size != 0) + { + size_t curSize = size; + SRes res = stream->Read(stream, data, &curSize); + *processedSize += curSize; + data += curSize; + size -= curSize; + RINOK(res); + if (curSize == 0) + return SZ_OK; + } + return SZ_OK; +} + +#define GET_NEXT_THREAD(p) &p->mtCoder->threads[p->index == p->mtCoder->numThreads - 1 ? 0 : p->index + 1] + +static SRes MtThread_Process(CMtThread *p, Bool *stop) +{ + CMtThread *next; + *stop = True; + if (Event_Wait(&p->canRead) != 0) + return SZ_ERROR_THREAD; + + next = GET_NEXT_THREAD(p); + + if (p->stopReading) + { + next->stopReading = True; + return Event_Set(&next->canRead) == 0 ? SZ_OK : SZ_ERROR_THREAD; + } + + { + size_t size = p->mtCoder->blockSize; + size_t destSize = p->outBufSize; + + RINOK(FullRead(p->mtCoder->inStream, p->inBuf, &size)); + next->stopReading = *stop = (size != p->mtCoder->blockSize); + if (Event_Set(&next->canRead) != 0) + return SZ_ERROR_THREAD; + + RINOK(p->mtCoder->mtCallback->Code(p->mtCoder->mtCallback, p->index, + p->outBuf, &destSize, p->inBuf, size, *stop)); + + MtProgress_Reinit(&p->mtCoder->mtProgress, p->index); + + if (Event_Wait(&p->canWrite) != 0) + return SZ_ERROR_THREAD; + if (p->stopWriting) + return SZ_ERROR_FAIL; + if (p->mtCoder->outStream->Write(p->mtCoder->outStream, p->outBuf, destSize) != destSize) + return SZ_ERROR_WRITE; + return Event_Set(&next->canWrite) == 0 ? SZ_OK : SZ_ERROR_THREAD; + } +} + +static THREAD_FUNC_RET_TYPE THREAD_FUNC_CALL_TYPE ThreadFunc(void *pp) +{ + CMtThread *p = (CMtThread *)pp; + for (;;) + { + Bool stop; + CMtThread *next = GET_NEXT_THREAD(p); + SRes res = MtThread_Process(p, &stop); + if (res != SZ_OK) + { + MtCoder_SetError(p->mtCoder, res); + MtProgress_SetError(&p->mtCoder->mtProgress, res); + next->stopReading = True; + next->stopWriting = True; + Event_Set(&next->canRead); + Event_Set(&next->canWrite); + return res; + } + if (stop) + return 0; + } +} + +void MtCoder_Construct(CMtCoder* p) +{ + unsigned i; + p->alloc = 0; + for (i = 0; i < NUM_MT_CODER_THREADS_MAX; i++) + { + CMtThread *t = &p->threads[i]; + t->index = i; + CMtThread_Construct(t, p); + } + CriticalSection_Init(&p->cs); + CriticalSection_Init(&p->mtProgress.cs); +} + +void MtCoder_Destruct(CMtCoder* p) +{ + unsigned i; + for (i = 0; i < NUM_MT_CODER_THREADS_MAX; i++) + CMtThread_Destruct(&p->threads[i]); + CriticalSection_Delete(&p->cs); + CriticalSection_Delete(&p->mtProgress.cs); +} + +SRes MtCoder_Code(CMtCoder *p) +{ + unsigned i, numThreads = p->numThreads; + SRes res = SZ_OK; + p->res = SZ_OK; + + MtProgress_Init(&p->mtProgress, p->progress); + + for (i = 0; i < numThreads; i++) + { + RINOK(CMtThread_Prepare(&p->threads[i])); + } + + for (i = 0; i < numThreads; i++) + { + CMtThread *t = &p->threads[i]; + CLoopThread *lt = &t->thread; + + if (!Thread_WasCreated(<->thread)) + { + lt->func = ThreadFunc; + lt->param = t; + + if (LoopThread_Create(lt) != SZ_OK) + { + res = SZ_ERROR_THREAD; + break; + } + } + } + + if (res == SZ_OK) + { + unsigned j; + for (i = 0; i < numThreads; i++) + { + CMtThread *t = &p->threads[i]; + if (LoopThread_StartSubThread(&t->thread) != SZ_OK) + { + res = SZ_ERROR_THREAD; + p->threads[0].stopReading = True; + break; + } + } + + Event_Set(&p->threads[0].canWrite); + Event_Set(&p->threads[0].canRead); + + for (j = 0; j < i; j++) + LoopThread_WaitSubThread(&p->threads[j].thread); + } + + for (i = 0; i < numThreads; i++) + CMtThread_CloseEvents(&p->threads[i]); + return (res == SZ_OK) ? p->res : res; +} diff --git a/src/lzma/Ppmd7.c b/src/lzma/Ppmd7.c new file mode 100644 index 0000000..621b9ff --- /dev/null +++ b/src/lzma/Ppmd7.c @@ -0,0 +1,710 @@ +/* Ppmd7.c -- PPMdH codec +2016-05-21 : Igor Pavlov : Public domain +This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ + +#include + +#include + +#include + +const Byte PPMD7_kExpEscape[16] = { 25, 14, 9, 7, 5, 5, 4, 4, 4, 3, 3, 3, 2, 2, 2, 2 }; +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[(nu) - 1]) +#define I2U(indx) (p->Indx2Units[indx]) + +#ifdef PPMD_32BIT + #define REF(ptr) (ptr) +#else + #define REF(ptr) ((UInt32)((Byte *)(ptr) - (p)->Base)) +#endif + +#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 + #ifdef PPMD_32BIT + struct CPpmd7_Node_ * + #else + UInt32 + #endif + 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; + +#ifdef PPMD_32BIT + #define NODE(ptr) (ptr) +#else + #define NODE(offs) ((CPpmd7_Node *)(p->Base + (offs))) +#endif + +void Ppmd7_Construct(CPpmd7 *p) +{ + unsigned i, k, m; + + p->Base = 0; + + 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; + } + + memset(p->HB2Flag, 0, 0x40); + memset(p->HB2Flag + 0x40, 8, 0x100 - 0x40); +} + +void Ppmd7_Free(CPpmd7 *p, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->Base); + p->Size = 0; + p->Base = 0; +} + +Bool Ppmd7_Alloc(CPpmd7 *p, UInt32 size, ISzAlloc *alloc) +{ + if (p->Base == 0 || p->Size != size) + { + Ppmd7_Free(p, alloc); + p->AlignOffset = + #ifdef PPMD_32BIT + (4 - size) & 3; + #else + 4 - (size & 3); + #endif + if ((p->Base = (Byte *)alloc->Alloc(alloc, p->AlignOffset + size + #ifndef PPMD_32BIT + + UNIT_SIZE + #endif + )) == 0) + return False; + p->Size = size; + } + return True; +} + +static void InsertNode(CPpmd7 *p, void *node, unsigned indx) +{ + *((CPpmd_Void_Ref *)node) = 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; + 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); +} + +static void GlueFreeBlocks(CPpmd7 *p) +{ + #ifdef PPMD_32BIT + CPpmd7_Node headItem; + CPpmd7_Node_Ref head = &headItem; + #else + CPpmd7_Node_Ref head = p->AlignOffset + p->Size; + #endif + + CPpmd7_Node_Ref n = head; + unsigned i; + + p->GlueCount = 255; + + /* create doubly-linked list of free blocks */ + for (i = 0; i < PPMD_NUM_INDEXES; i++) + { + UInt16 nu = I2U(i); + CPpmd7_Node_Ref next = (CPpmd7_Node_Ref)p->FreeList[i]; + p->FreeList[i] = 0; + while (next != 0) + { + CPpmd7_Node *node = NODE(next); + node->Next = n; + n = NODE(n)->Prev = next; + next = *(const CPpmd7_Node_Ref *)node; + node->Stamp = 0; + node->NU = (UInt16)nu; + } + } + NODE(head)->Stamp = 1; + NODE(head)->Next = n; + NODE(n)->Prev = head; + if (p->LoUnit != p->HiUnit) + ((CPpmd7_Node *)p->LoUnit)->Stamp = 1; + + /* Glue free blocks */ + while (n != head) + { + CPpmd7_Node *node = NODE(n); + UInt32 nu = (UInt32)node->NU; + for (;;) + { + CPpmd7_Node *node2 = NODE(n) + nu; + nu += node2->NU; + if (node2->Stamp != 0 || nu >= 0x10000) + break; + NODE(node2->Prev)->Next = node2->Next; + NODE(node2->Next)->Prev = node2->Prev; + node->NU = (UInt16)nu; + } + n = node->Next; + } + + /* Fill lists of free blocks */ + for (n = NODE(head)->Next; n != head;) + { + CPpmd7_Node *node = NODE(n); + unsigned nu; + CPpmd7_Node_Ref next = node->Next; + for (nu = node->NU; 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, nu - k - 1); + } + InsertNode(p, node, i); + n = next; + } +} + +static void *AllocUnitsRare(CPpmd7 *p, unsigned indx) +{ + unsigned i; + void *retVal; + 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)); + p->GlueCount--; + return ((UInt32)(p->UnitsStart - p->Text) > numBytes) ? (p->UnitsStart -= numBytes) : (NULL); + } + } + while (p->FreeList[i] == 0); + retVal = RemoveNode(p, i); + SplitBlock(p, retVal, i, indx); + return retVal; +} + +static void *AllocUnits(CPpmd7 *p, unsigned indx) +{ + UInt32 numBytes; + if (p->FreeList[indx] != 0) + return RemoveNode(p, indx); + numBytes = U2B(I2U(indx)); + if (numBytes <= (UInt32)(p->HiUnit - p->LoUnit)) + { + void *retVal = p->LoUnit; + p->LoUnit += numBytes; + return retVal; + } + return AllocUnitsRare(p, indx); +} + +#define MyMem12Cpy(dest, src, num) \ + { UInt32 *d = (UInt32 *)dest; const UInt32 *s = (const UInt32 *)src; UInt32 n = num; \ + do { d[0] = s[0]; d[1] = s[1]; d[2] = s[2]; s += 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) ((CPpmd_Void_Ref)((p)->SuccessorLow | ((UInt32)(p)->SuccessorHigh << 16))) + +static void SetSuccessor(CPpmd_State *p, CPpmd_Void_Ref v) +{ + (p)->SuccessorLow = (UInt16)((UInt32)(v) & 0xFFFF); + (p)->SuccessorHigh = (UInt16)(((UInt32)(v) >> 16) & 0xFFFF); +} + +static void RestartModel(CPpmd7 *p) +{ + unsigned i, k, m; + + 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; + + p->MinContext = p->MaxContext = (CTX_PTR)(p->HiUnit -= UNIT_SIZE); /* AllocContext(p); */ + p->MinContext->Suffix = 0; + p->MinContext->NumStats = 256; + p->MinContext->SummFreq = 256 + 1; + p->FoundState = (CPpmd_State *)p->LoUnit; /* AllocUnits(p, PPMD_NUM_INDEXES - 1); */ + p->LoUnit += U2B(256 / 2); + p->MinContext->Stats = REF(p->FoundState); + for (i = 0; i < 256; i++) + { + CPpmd_State *s = &p->FoundState[i]; + s->Symbol = (Byte)i; + s->Freq = 1; + SetSuccessor(s, 0); + } + + for (i = 0; i < 128; i++) + for (k = 0; k < 8; k++) + { + 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++) + for (k = 0; k < 16; k++) + { + CPpmd_See *s = &p->See[i][k]; + s->Summ = (UInt16)((5 * i + 10) << (s->Shift = PPMD_PERIOD_BITS - 4)); + s->Count = 4; + } +} + +void Ppmd7_Init(CPpmd7 *p, unsigned maxOrder) +{ + p->MaxOrder = maxOrder; + RestartModel(p); + p->DummySee.Shift = PPMD_PERIOD_BITS; + p->DummySee.Summ = 0; /* unused */ + p->DummySee.Count = 64; /* unused */ +} + +static CTX_PTR CreateSuccessors(CPpmd7 *p, Bool skip) +{ + CPpmd_State upState; + CTX_PTR c = p->MinContext; + CPpmd_Byte_Ref upBranch = (CPpmd_Byte_Ref)SUCCESSOR(p->FoundState); + CPpmd_State *ps[PPMD7_MAX_ORDER]; + unsigned numPs = 0; + + if (!skip) + ps[numPs++] = p->FoundState; + + while (c->Suffix) + { + CPpmd_Void_Ref successor; + CPpmd_State *s; + c = SUFFIX(c); + if (c->NumStats != 1) + { + for (s = STATS(c); s->Symbol != p->FoundState->Symbol; s++); + } + else + s = ONE_STATE(c); + successor = SUCCESSOR(s); + if (successor != upBranch) + { + c = CTX(successor); + if (numPs == 0) + return c; + break; + } + ps[numPs++] = s; + } + + upState.Symbol = *(const Byte *)Ppmd7_GetPtr(p, upBranch); + SetSuccessor(&upState, upBranch + 1); + + if (c->NumStats == 1) + upState.Freq = ONE_STATE(c)->Freq; + else + { + UInt32 cf, s0; + CPpmd_State *s; + for (s = STATS(c); s->Symbol != upState.Symbol; s++); + cf = s->Freq - 1; + s0 = c->SummFreq - c->NumStats - cf; + upState.Freq = (Byte)(1 + ((2 * cf <= s0) ? (5 * cf > s0) : ((2 * cf + 3 * s0 - 1) / (2 * s0)))); + } + + do + { + /* Create Child */ + CTX_PTR c1; /* = AllocContext(p); */ + if (p->HiUnit != p->LoUnit) + c1 = (CTX_PTR)(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) = upState; + c1->Suffix = REF(c); + SetSuccessor(ps[--numPs], REF(c1)); + c = c1; + } + while (numPs != 0); + + return c; +} + +static void SwapStates(CPpmd_State *t1, CPpmd_State *t2) +{ + CPpmd_State tmp = *t1; + *t1 = *t2; + *t2 = tmp; +} + +static void UpdateModel(CPpmd7 *p) +{ + CPpmd_Void_Ref successor, fSuccessor = SUCCESSOR(p->FoundState); + CTX_PTR c; + unsigned s0, ns; + + if (p->FoundState->Freq < MAX_FREQ / 4 && p->MinContext->Suffix != 0) + { + 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); + if (s->Symbol != p->FoundState->Symbol) + { + do { s++; } while (s->Symbol != p->FoundState->Symbol); + if (s[0].Freq >= s[-1].Freq) + { + SwapStates(&s[0], &s[-1]); + s--; + } + } + if (s->Freq < MAX_FREQ - 9) + { + s->Freq += 2; + c->SummFreq += 2; + } + } + } + + if (p->OrderFall == 0) + { + p->MinContext = p->MaxContext = CreateSuccessors(p, True); + if (p->MinContext == 0) + { + RestartModel(p); + return; + } + SetSuccessor(p->FoundState, REF(p->MinContext)); + return; + } + + *p->Text++ = p->FoundState->Symbol; + successor = REF(p->Text); + if (p->Text >= p->UnitsStart) + { + RestartModel(p); + return; + } + + if (fSuccessor) + { + if (fSuccessor <= successor) + { + CTX_PTR cs = CreateSuccessors(p, False); + if (cs == NULL) + { + RestartModel(p); + return; + } + fSuccessor = REF(cs); + } + if (--p->OrderFall == 0) + { + successor = fSuccessor; + p->Text -= (p->MaxContext != p->MinContext); + } + } + else + { + SetSuccessor(p->FoundState, successor); + fSuccessor = REF(p->MinContext); + } + + s0 = p->MinContext->SummFreq - (ns = p->MinContext->NumStats) - (p->FoundState->Freq - 1); + + for (c = p->MaxContext; c != p->MinContext; c = SUFFIX(c)) + { + unsigned ns1; + UInt32 cf, sf; + if ((ns1 = c->NumStats) != 1) + { + if ((ns1 & 1) == 0) + { + /* Expand for one UNIT */ + unsigned oldNU = ns1 >> 1; + unsigned i = U2I(oldNU); + if (i != U2I(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->Stats = STATS_REF(ptr); + } + } + c->SummFreq = (UInt16)(c->SummFreq + (2 * ns1 < ns) + 2 * ((4 * ns1 <= ns) & (c->SummFreq <= 8 * ns1))); + } + else + { + CPpmd_State *s = (CPpmd_State*)AllocUnits(p, 0); + if (!s) + { + RestartModel(p); + return; + } + *s = *ONE_STATE(c); + c->Stats = REF(s); + if (s->Freq < MAX_FREQ / 4 - 1) + s->Freq <<= 1; + else + s->Freq = MAX_FREQ - 4; + c->SummFreq = (UInt16)(s->Freq + p->InitEsc + (ns > 3)); + } + cf = 2 * (UInt32)p->FoundState->Freq * (c->SummFreq + 6); + sf = (UInt32)s0 + c->SummFreq; + if (cf < 6 * sf) + { + cf = 1 + (cf > sf) + (cf >= 4 * sf); + c->SummFreq += 3; + } + else + { + cf = 4 + (cf >= 9 * sf) + (cf >= 12 * sf) + (cf >= 15 * sf); + c->SummFreq = (UInt16)(c->SummFreq + cf); + } + { + CPpmd_State *s = STATS(c) + ns1; + SetSuccessor(s, successor); + s->Symbol = p->FoundState->Symbol; + s->Freq = (Byte)cf; + c->NumStats = (UInt16)(ns1 + 1); + } + } + p->MaxContext = p->MinContext = CTX(fSuccessor); +} + +static void Rescale(CPpmd7 *p) +{ + unsigned i, adder, sumFreq, escFreq; + CPpmd_State *stats = STATS(p->MinContext); + CPpmd_State *s = p->FoundState; + { + CPpmd_State tmp = *s; + for (; s != stats; s--) + s[0] = s[-1]; + *s = tmp; + } + escFreq = p->MinContext->SummFreq - s->Freq; + s->Freq += 4; + adder = (p->OrderFall != 0); + s->Freq = (Byte)((s->Freq + adder) >> 1); + sumFreq = s->Freq; + + i = p->MinContext->NumStats - 1; + do + { + escFreq -= (++s)->Freq; + s->Freq = (Byte)((s->Freq + adder) >> 1); + sumFreq += s->Freq; + if (s[0].Freq > s[-1].Freq) + { + CPpmd_State *s1 = s; + CPpmd_State tmp = *s1; + do + s1[0] = s1[-1]; + while (--s1 != stats && tmp.Freq > s1[-1].Freq); + *s1 = tmp; + } + } + while (--i); + + if (s->Freq == 0) + { + unsigned numStats = p->MinContext->NumStats; + unsigned n0, n1; + do { i++; } while ((--s)->Freq == 0); + escFreq += i; + p->MinContext->NumStats = (UInt16)(p->MinContext->NumStats - i); + if (p->MinContext->NumStats == 1) + { + CPpmd_State tmp = *stats; + do + { + tmp.Freq = (Byte)(tmp.Freq - (tmp.Freq >> 1)); + escFreq >>= 1; + } + while (escFreq > 1); + InsertNode(p, stats, U2I(((numStats + 1) >> 1))); + *(p->FoundState = ONE_STATE(p->MinContext)) = tmp; + return; + } + n0 = (numStats + 1) >> 1; + n1 = (p->MinContext->NumStats + 1) >> 1; + if (n0 != n1) + p->MinContext->Stats = STATS_REF(ShrinkUnits(p, stats, n0, n1)); + } + p->MinContext->SummFreq = (UInt16)(sumFreq + escFreq - (escFreq >> 1)); + p->FoundState = STATS(p->MinContext); +} + +CPpmd_See *Ppmd7_MakeEscFreq(CPpmd7 *p, unsigned numMasked, UInt32 *escFreq) +{ + CPpmd_See *see; + unsigned nonMasked = p->MinContext->NumStats - numMasked; + if (p->MinContext->NumStats != 256) + { + see = p->See[(unsigned)p->NS2Indx[nonMasked - 1]] + + (nonMasked < (unsigned)SUFFIX(p->MinContext)->NumStats - p->MinContext->NumStats) + + 2 * (unsigned)(p->MinContext->SummFreq < 11 * p->MinContext->NumStats) + + 4 * (unsigned)(numMasked > nonMasked) + + p->HiBitsFlag; + { + unsigned r = (see->Summ >> see->Shift); + see->Summ = (UInt16)(see->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 && (Byte *)c > p->Text) + p->MinContext = p->MaxContext = c; + else + UpdateModel(p); +} + +void Ppmd7_Update1(CPpmd7 *p) +{ + CPpmd_State *s = p->FoundState; + s->Freq += 4; + p->MinContext->SummFreq += 4; + if (s[0].Freq > s[-1].Freq) + { + SwapStates(&s[0], &s[-1]); + p->FoundState = --s; + if (s->Freq > MAX_FREQ) + Rescale(p); + } + NextContext(p); +} + +void Ppmd7_Update1_0(CPpmd7 *p) +{ + p->PrevSuccess = (2 * p->FoundState->Freq > p->MinContext->SummFreq); + p->RunLength += p->PrevSuccess; + p->MinContext->SummFreq += 4; + if ((p->FoundState->Freq += 4) > MAX_FREQ) + Rescale(p); + NextContext(p); +} + +void Ppmd7_UpdateBin(CPpmd7 *p) +{ + p->FoundState->Freq = (Byte)(p->FoundState->Freq + (p->FoundState->Freq < 128 ? 1: 0)); + p->PrevSuccess = 1; + p->RunLength++; + NextContext(p); +} + +void Ppmd7_Update2(CPpmd7 *p) +{ + p->MinContext->SummFreq += 4; + if ((p->FoundState->Freq += 4) > MAX_FREQ) + Rescale(p); + p->RunLength = p->InitRL; + UpdateModel(p); +} diff --git a/src/lzma/Ppmd7Dec.c b/src/lzma/Ppmd7Dec.c new file mode 100644 index 0000000..54080ba --- /dev/null +++ b/src/lzma/Ppmd7Dec.c @@ -0,0 +1,189 @@ +/* Ppmd7Dec.c -- PPMdH Decoder +2010-03-12 : Igor Pavlov : Public domain +This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ + +#include + +#include + +#define kTopValue (1 << 24) + +Bool Ppmd7z_RangeDec_Init(CPpmd7z_RangeDec *p) +{ + unsigned i; + p->Code = 0; + p->Range = 0xFFFFFFFF; + if (p->Stream->Read((void *)p->Stream) != 0) + return False; + for (i = 0; i < 4; i++) + p->Code = (p->Code << 8) | p->Stream->Read((void *)p->Stream); + return (p->Code < 0xFFFFFFFF); +} + +static UInt32 Range_GetThreshold(void *pp, UInt32 total) +{ + CPpmd7z_RangeDec *p = (CPpmd7z_RangeDec *)pp; + return (p->Code) / (p->Range /= total); +} + +static void Range_Normalize(CPpmd7z_RangeDec *p) +{ + if (p->Range < kTopValue) + { + p->Code = (p->Code << 8) | p->Stream->Read((void *)p->Stream); + p->Range <<= 8; + if (p->Range < kTopValue) + { + p->Code = (p->Code << 8) | p->Stream->Read((void *)p->Stream); + p->Range <<= 8; + } + } +} + +static void Range_Decode(void *pp, UInt32 start, UInt32 size) +{ + CPpmd7z_RangeDec *p = (CPpmd7z_RangeDec *)pp; + p->Code -= start * p->Range; + p->Range *= size; + Range_Normalize(p); +} + +static UInt32 Range_DecodeBit(void *pp, UInt32 size0) +{ + CPpmd7z_RangeDec *p = (CPpmd7z_RangeDec *)pp; + UInt32 newBound = (p->Range >> 14) * size0; + UInt32 symbol; + if (p->Code < newBound) + { + symbol = 0; + p->Range = newBound; + } + else + { + symbol = 1; + p->Code -= newBound; + p->Range -= newBound; + } + Range_Normalize(p); + return symbol; +} + +void Ppmd7z_RangeDec_CreateVTable(CPpmd7z_RangeDec *p) +{ + p->p.GetThreshold = Range_GetThreshold; + p->p.Decode = Range_Decode; + p->p.DecodeBit = Range_DecodeBit; +} + + +#define MASK(sym) ((signed char *)charMask)[sym] + +int Ppmd7_DecodeSymbol(CPpmd7 *p, IPpmd7_RangeDec *rc) +{ + 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; + if ((count = rc->GetThreshold(rc, p->MinContext->SummFreq)) < (hiCnt = s->Freq)) + { + Byte symbol; + rc->Decode(rc, 0, s->Freq); + p->FoundState = s; + symbol = s->Symbol; + Ppmd7_Update1_0(p); + return symbol; + } + p->PrevSuccess = 0; + i = p->MinContext->NumStats - 1; + do + { + if ((hiCnt += (++s)->Freq) > count) + { + Byte symbol; + rc->Decode(rc, hiCnt - s->Freq, s->Freq); + p->FoundState = s; + symbol = s->Symbol; + Ppmd7_Update1(p); + return symbol; + } + } + while (--i); + if (count >= p->MinContext->SummFreq) + return -2; + p->HiBitsFlag = p->HB2Flag[p->FoundState->Symbol]; + rc->Decode(rc, hiCnt, p->MinContext->SummFreq - hiCnt); + PPMD_SetAllBitsIn256Bytes(charMask); + MASK(s->Symbol) = 0; + i = p->MinContext->NumStats - 1; + do { MASK((--s)->Symbol) = 0; } while (--i); + } + else + { + UInt16 *prob = Ppmd7_GetBinSumm(p); + if (rc->DecodeBit(rc, *prob) == 0) + { + Byte symbol; + *prob = (UInt16)PPMD_UPDATE_PROB_0(*prob); + symbol = (p->FoundState = Ppmd7Context_OneState(p->MinContext))->Symbol; + Ppmd7_UpdateBin(p); + return symbol; + } + *prob = (UInt16)PPMD_UPDATE_PROB_1(*prob); + p->InitEsc = PPMD7_kExpEscape[*prob >> 10]; + PPMD_SetAllBitsIn256Bytes(charMask); + MASK(Ppmd7Context_OneState(p->MinContext)->Symbol) = 0; + p->PrevSuccess = 0; + } + for (;;) + { + CPpmd_State *ps[256], *s; + UInt32 freqSum, count, hiCnt; + CPpmd_See *see; + unsigned i, num, numMasked = p->MinContext->NumStats; + do + { + p->OrderFall++; + if (!p->MinContext->Suffix) + return -1; + p->MinContext = Ppmd7_GetContext(p, p->MinContext->Suffix); + } + while (p->MinContext->NumStats == numMasked); + hiCnt = 0; + s = Ppmd7_GetStats(p, p->MinContext); + i = 0; + num = p->MinContext->NumStats - numMasked; + do + { + int k = (int)(MASK(s->Symbol)); + hiCnt += (s->Freq & k); + ps[i] = s++; + i -= k; + } + while (i != num); + + see = Ppmd7_MakeEscFreq(p, numMasked, &freqSum); + freqSum += hiCnt; + count = rc->GetThreshold(rc, freqSum); + + if (count < hiCnt) + { + Byte symbol; + CPpmd_State **pps = ps; + for (hiCnt = 0; (hiCnt += (*pps)->Freq) <= count; pps++); + s = *pps; + rc->Decode(rc, hiCnt - s->Freq, s->Freq); + Ppmd_See_Update(see); + p->FoundState = s; + symbol = s->Symbol; + Ppmd7_Update2(p); + return symbol; + } + if (count >= freqSum) + return -2; + rc->Decode(rc, hiCnt, freqSum - hiCnt); + see->Summ = (UInt16)(see->Summ + freqSum); + do { MASK(ps[--i]->Symbol) = 0; } while (i != 0); + } +} diff --git a/src/lzma/Ppmd7Enc.c b/src/lzma/Ppmd7Enc.c new file mode 100644 index 0000000..fbfd116 --- /dev/null +++ b/src/lzma/Ppmd7Enc.c @@ -0,0 +1,187 @@ +/* Ppmd7Enc.c -- PPMdH Encoder +2015-09-28 : Igor Pavlov : Public domain +This code is based on PPMd var.H (2001): Dmitry Shkarin : Public domain */ + +#include + +#include + +#define kTopValue (1 << 24) + +void Ppmd7z_RangeEnc_Init(CPpmd7z_RangeEnc *p) +{ + p->Low = 0; + p->Range = 0xFFFFFFFF; + p->Cache = 0; + p->CacheSize = 1; +} + +static void RangeEnc_ShiftLow(CPpmd7z_RangeEnc *p) +{ + if ((UInt32)p->Low < (UInt32)0xFF000000 || (unsigned)(p->Low >> 32) != 0) + { + Byte temp = p->Cache; + do + { + p->Stream->Write(p->Stream, (Byte)(temp + (Byte)(p->Low >> 32))); + temp = 0xFF; + } + while (--p->CacheSize != 0); + p->Cache = (Byte)((UInt32)p->Low >> 24); + } + p->CacheSize++; + p->Low = (UInt32)p->Low << 8; +} + +static void RangeEnc_Encode(CPpmd7z_RangeEnc *p, UInt32 start, UInt32 size, UInt32 total) +{ + p->Low += start * (p->Range /= total); + p->Range *= size; + while (p->Range < kTopValue) + { + p->Range <<= 8; + RangeEnc_ShiftLow(p); + } +} + +static void RangeEnc_EncodeBit_0(CPpmd7z_RangeEnc *p, UInt32 size0) +{ + p->Range = (p->Range >> 14) * size0; + while (p->Range < kTopValue) + { + p->Range <<= 8; + RangeEnc_ShiftLow(p); + } +} + +static void RangeEnc_EncodeBit_1(CPpmd7z_RangeEnc *p, UInt32 size0) +{ + UInt32 newBound = (p->Range >> 14) * size0; + p->Low += newBound; + p->Range -= newBound; + while (p->Range < kTopValue) + { + p->Range <<= 8; + RangeEnc_ShiftLow(p); + } +} + +void Ppmd7z_RangeEnc_FlushData(CPpmd7z_RangeEnc *p) +{ + unsigned i; + for (i = 0; i < 5; i++) + RangeEnc_ShiftLow(p); +} + + +#define MASK(sym) ((signed char *)charMask)[sym] + +void Ppmd7_EncodeSymbol(CPpmd7 *p, CPpmd7z_RangeEnc *rc, 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; + if (s->Symbol == symbol) + { + RangeEnc_Encode(rc, 0, s->Freq, p->MinContext->SummFreq); + p->FoundState = s; + Ppmd7_Update1_0(p); + return; + } + p->PrevSuccess = 0; + sum = s->Freq; + i = p->MinContext->NumStats - 1; + do + { + if ((++s)->Symbol == symbol) + { + RangeEnc_Encode(rc, sum, s->Freq, p->MinContext->SummFreq); + p->FoundState = s; + Ppmd7_Update1(p); + return; + } + sum += s->Freq; + } + while (--i); + + p->HiBitsFlag = p->HB2Flag[p->FoundState->Symbol]; + PPMD_SetAllBitsIn256Bytes(charMask); + MASK(s->Symbol) = 0; + i = p->MinContext->NumStats - 1; + do { MASK((--s)->Symbol) = 0; } while (--i); + RangeEnc_Encode(rc, sum, p->MinContext->SummFreq - sum, p->MinContext->SummFreq); + } + else + { + UInt16 *prob = Ppmd7_GetBinSumm(p); + CPpmd_State *s = Ppmd7Context_OneState(p->MinContext); + if (s->Symbol == symbol) + { + RangeEnc_EncodeBit_0(rc, *prob); + *prob = (UInt16)PPMD_UPDATE_PROB_0(*prob); + p->FoundState = s; + Ppmd7_UpdateBin(p); + return; + } + else + { + RangeEnc_EncodeBit_1(rc, *prob); + *prob = (UInt16)PPMD_UPDATE_PROB_1(*prob); + p->InitEsc = PPMD7_kExpEscape[*prob >> 10]; + PPMD_SetAllBitsIn256Bytes(charMask); + MASK(s->Symbol) = 0; + p->PrevSuccess = 0; + } + } + for (;;) + { + UInt32 escFreq; + CPpmd_See *see; + CPpmd_State *s; + UInt32 sum; + unsigned i, numMasked = p->MinContext->NumStats; + do + { + p->OrderFall++; + if (!p->MinContext->Suffix) + return; /* EndMarker (symbol = -1) */ + p->MinContext = Ppmd7_GetContext(p, p->MinContext->Suffix); + } + while (p->MinContext->NumStats == numMasked); + + see = Ppmd7_MakeEscFreq(p, numMasked, &escFreq); + s = Ppmd7_GetStats(p, p->MinContext); + sum = 0; + i = p->MinContext->NumStats; + do + { + int cur = s->Symbol; + if (cur == symbol) + { + UInt32 low = sum; + CPpmd_State *s1 = s; + do + { + sum += (s->Freq & (int)(MASK(s->Symbol))); + s++; + } + while (--i); + RangeEnc_Encode(rc, low, s1->Freq, sum + escFreq); + Ppmd_See_Update(see); + p->FoundState = s1; + Ppmd7_Update2(p); + return; + } + sum += (s->Freq & (int)(MASK(cur))); + MASK(cur) = 0; + s++; + } + while (--i); + + RangeEnc_Encode(rc, sum, escFreq, sum + escFreq); + see->Summ = (UInt16)(see->Summ + sum + escFreq); + } +} diff --git a/src/lzma/Sha256.c b/src/lzma/Sha256.c new file mode 100644 index 0000000..021e549 --- /dev/null +++ b/src/lzma/Sha256.c @@ -0,0 +1,248 @@ +/* Crypto/Sha256.c -- SHA-256 Hash +2015-11-14 : Igor Pavlov : Public domain +This code is based on public domain code from Wei Dai's Crypto++ library. */ + +#include + +#include + +#include +#include +#include + +/* define it for speed optimization */ +#ifndef _SFX +#define _SHA256_UNROLL +#define _SHA256_UNROLL2 +#endif + +/* #define _SHA256_UNROLL2 */ + +void Sha256_Init(CSha256 *p) +{ + 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; + p->count = 0; +} + +#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 blk0(i) (W[i]) +#define blk2(i) (W[i] += s1(W[((i)-2)&15]) + W[((i)-7)&15] + s0(W[((i)-15)&15])) + +#define Ch(x,y,z) (z^(x&(y^z))) +#define Maj(x,y,z) ((x&y)|(z&(x|y))) + +#ifdef _SHA256_UNROLL2 + +#define R(a,b,c,d,e,f,g,h, i) \ + h += S1(e) + Ch(e,f,g) + K[(i)+(j)] + (j ? blk2(i) : blk0(i)); \ + d += h; \ + h += S0(a) + Maj(a, b, c) + +#define RX_8(i) \ + R(a,b,c,d,e,f,g,h, i); \ + R(h,a,b,c,d,e,f,g, i+1); \ + R(g,h,a,b,c,d,e,f, i+2); \ + R(f,g,h,a,b,c,d,e, i+3); \ + R(e,f,g,h,a,b,c,d, i+4); \ + R(d,e,f,g,h,a,b,c, i+5); \ + R(c,d,e,f,g,h,a,b, i+6); \ + R(b,c,d,e,f,g,h,a, i+7) + +#define RX_16 RX_8(0); RX_8(8); + +#else + +#define a(i) T[(0-(i))&7] +#define b(i) T[(1-(i))&7] +#define c(i) T[(2-(i))&7] +#define d(i) T[(3-(i))&7] +#define e(i) T[(4-(i))&7] +#define f(i) T[(5-(i))&7] +#define g(i) T[(6-(i))&7] +#define h(i) T[(7-(i))&7] + +#define R(i) \ + h(i) += S1(e(i)) + Ch(e(i),f(i),g(i)) + K[(i)+(j)] + (j ? blk2(i) : blk0(i)); \ + d(i) += h(i); \ + h(i) += S0(a(i)) + Maj(a(i), b(i), c(i)) \ + +#ifdef _SHA256_UNROLL + +#define RX_8(i) R(i+0); R(i+1); R(i+2); R(i+3); R(i+4); R(i+5); R(i+6); R(i+7); +#define RX_16 RX_8(0); RX_8(8); + +#else + +#define RX_16 unsigned i; for (i = 0; i < 16; i++) { R(i); } + +#endif + +#endif + +static const UInt32 K[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 +}; + +static void Sha256_WriteByteBlock(CSha256 *p) +{ + UInt32 W[16]; + unsigned j; + UInt32 *state; + + #ifdef _SHA256_UNROLL2 + UInt32 a,b,c,d,e,f,g,h; + #else + UInt32 T[8]; + #endif + + for (j = 0; j < 16; j += 4) + { + const Byte *ccc = p->buffer + j * 4; + W[j ] = GetBe32(ccc); + W[j + 1] = GetBe32(ccc + 4); + W[j + 2] = GetBe32(ccc + 8); + W[j + 3] = GetBe32(ccc + 12); + } + + state = p->state; + + #ifdef _SHA256_UNROLL2 + a = state[0]; + b = state[1]; + c = state[2]; + d = state[3]; + e = state[4]; + f = state[5]; + g = state[6]; + h = state[7]; + #else + for (j = 0; j < 8; j++) + T[j] = state[j]; + #endif + + for (j = 0; j < 64; j += 16) + { + RX_16 + } + + #ifdef _SHA256_UNROLL2 + state[0] += a; + state[1] += b; + state[2] += c; + state[3] += d; + state[4] += e; + state[5] += f; + state[6] += g; + state[7] += h; + #else + for (j = 0; j < 8; j++) + state[j] += T[j]; + #endif + + /* Wipe variables */ + /* memset(W, 0, sizeof(W)); */ + /* memset(T, 0, sizeof(T)); */ +} + +#undef S0 +#undef S1 +#undef s0 +#undef s1 + +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; + } + + size -= num; + memcpy(p->buffer + pos, data, num); + data += num; + } + + for (;;) + { + Sha256_WriteByteBlock(p); + if (size < 64) + break; + size -= 64; + memcpy(p->buffer, data, 64); + data += 64; + } + + if (size != 0) + memcpy(p->buffer, data, size); +} + +void Sha256_Final(CSha256 *p, Byte *digest) +{ + unsigned pos = (unsigned)p->count & 0x3F; + unsigned i; + + p->buffer[pos++] = 0x80; + + while (pos != (64 - 8)) + { + pos &= 0x3F; + if (pos == 0) + Sha256_WriteByteBlock(p); + p->buffer[pos++] = 0; + } + + { + UInt64 numBits = (p->count << 3); + SetBe32(p->buffer + 64 - 8, (UInt32)(numBits >> 32)); + SetBe32(p->buffer + 64 - 4, (UInt32)(numBits)); + } + + Sha256_WriteByteBlock(p); + + for (i = 0; i < 8; i += 2) + { + UInt32 v0 = p->state[i]; + UInt32 v1 = p->state[i + 1]; + SetBe32(digest , v0); + SetBe32(digest + 4, v1); + digest += 8; + } + + Sha256_Init(p); +} diff --git a/src/lzma/Sort.c b/src/lzma/Sort.c new file mode 100644 index 0000000..6bb11bb --- /dev/null +++ b/src/lzma/Sort.c @@ -0,0 +1,141 @@ +/* Sort.c -- Sort functions +2014-04-05 : Igor Pavlov : Public domain */ + +#include + +#include + +#define HeapSortDown(p, k, size, temp) \ + { for (;;) { \ + size_t s = (k << 1); \ + if (s > size) break; \ + if (s < size && p[s + 1] > p[s]) s++; \ + if (temp >= p[s]) break; \ + p[k] = p[s]; k = s; \ + } p[k] = temp; } + +void HeapSort(UInt32 *p, size_t size) +{ + if (size <= 1) + return; + p--; + { + size_t i = size / 2; + do + { + UInt32 temp = p[i]; + size_t k = i; + HeapSortDown(p, k, size, temp) + } + while (--i != 0); + } + /* + do + { + size_t k = 1; + UInt32 temp = p[size]; + p[size--] = p[1]; + HeapSortDown(p, k, size, temp) + } + while (size > 1); + */ + while (size > 3) + { + UInt32 temp = p[size]; + size_t k = (p[3] > p[2]) ? 3 : 2; + p[size--] = p[1]; + p[1] = p[k]; + HeapSortDown(p, k, size, temp) + } + { + UInt32 temp = p[size]; + p[size] = p[1]; + if (size > 2 && p[2] < temp) + { + p[1] = p[2]; + p[2] = temp; + } + else + p[1] = temp; + } +} + +void HeapSort64(UInt64 *p, size_t size) +{ + if (size <= 1) + return; + p--; + { + size_t i = size / 2; + do + { + UInt64 temp = p[i]; + size_t k = i; + HeapSortDown(p, k, size, temp) + } + while (--i != 0); + } + /* + do + { + size_t k = 1; + UInt64 temp = p[size]; + p[size--] = p[1]; + HeapSortDown(p, k, size, temp) + } + while (size > 1); + */ + while (size > 3) + { + UInt64 temp = p[size]; + size_t k = (p[3] > p[2]) ? 3 : 2; + p[size--] = p[1]; + p[1] = p[k]; + HeapSortDown(p, k, size, temp) + } + { + UInt64 temp = p[size]; + p[size] = p[1]; + if (size > 2 && p[2] < temp) + { + p[1] = p[2]; + p[2] = temp; + } + else + p[1] = temp; + } +} + +/* +#define HeapSortRefDown(p, vals, n, size, temp) \ + { size_t k = n; UInt32 val = vals[temp]; for (;;) { \ + size_t s = (k << 1); \ + if (s > size) break; \ + if (s < size && vals[p[s + 1]] > vals[p[s]]) s++; \ + if (val >= vals[p[s]]) break; \ + p[k] = p[s]; k = s; \ + } p[k] = temp; } + +void HeapSortRef(UInt32 *p, UInt32 *vals, size_t size) +{ + if (size <= 1) + return; + p--; + { + size_t i = size / 2; + do + { + UInt32 temp = p[i]; + HeapSortRefDown(p, vals, i, size, temp); + } + while (--i != 0); + } + do + { + UInt32 temp = p[size]; + p[size--] = p[1]; + HeapSortRefDown(p, vals, 1, size, temp); + } + while (size > 1); +} +*/ diff --git a/src/lzma/Threads.c b/src/lzma/Threads.c new file mode 100644 index 0000000..3ae7f92 --- /dev/null +++ b/src/lzma/Threads.c @@ -0,0 +1,93 @@ +/* Threads.c -- multithreading library +2014-09-21 : Igor Pavlov : Public domain */ + +#include + +#ifndef UNDER_CE +#include +#endif + +#include + +static WRes GetError() +{ + DWORD res = GetLastError(); + return (res) ? (WRes)(res) : 1; +} + +WRes HandleToWRes(HANDLE h) { return (h != 0) ? 0 : GetError(); } +WRes BOOLToWRes(BOOL v) { return v ? 0 : GetError(); } + +WRes HandlePtr_Close(HANDLE *p) +{ + if (*p != NULL) + if (!CloseHandle(*p)) + return GetError(); + *p = NULL; + return 0; +} + +WRes Handle_WaitObject(HANDLE h) { return (WRes)WaitForSingleObject(h, INFINITE); } + +WRes Thread_Create(CThread *p, THREAD_FUNC_TYPE func, LPVOID param) +{ + /* Windows Me/98/95: threadId parameter may not be NULL in _beginthreadex/CreateThread functions */ + + #ifdef UNDER_CE + + DWORD threadId; + *p = CreateThread(0, 0, func, param, 0, &threadId); + + #else + + unsigned threadId; + *p = (HANDLE)_beginthreadex(NULL, 0, func, param, 0, &threadId); + + #endif + + /* maybe we must use errno here, but probably GetLastError() is also OK. */ + return HandleToWRes(*p); +} + +WRes Event_Create(CEvent *p, BOOL manualReset, int signaled) +{ + *p = CreateEvent(NULL, manualReset, (signaled ? TRUE : FALSE), NULL); + return HandleToWRes(*p); +} + +WRes Event_Set(CEvent *p) { return BOOLToWRes(SetEvent(*p)); } +WRes Event_Reset(CEvent *p) { return BOOLToWRes(ResetEvent(*p)); } + +WRes ManualResetEvent_Create(CManualResetEvent *p, int signaled) { return Event_Create(p, TRUE, signaled); } +WRes AutoResetEvent_Create(CAutoResetEvent *p, int signaled) { return Event_Create(p, FALSE, signaled); } +WRes ManualResetEvent_CreateNotSignaled(CManualResetEvent *p) { return ManualResetEvent_Create(p, 0); } +WRes AutoResetEvent_CreateNotSignaled(CAutoResetEvent *p) { return AutoResetEvent_Create(p, 0); } + + +WRes Semaphore_Create(CSemaphore *p, UInt32 initCount, UInt32 maxCount) +{ + *p = CreateSemaphore(NULL, (LONG)initCount, (LONG)maxCount, NULL); + return HandleToWRes(*p); +} + +static WRes Semaphore_Release(CSemaphore *p, LONG releaseCount, LONG *previousCount) + { return BOOLToWRes(ReleaseSemaphore(*p, releaseCount, previousCount)); } +WRes Semaphore_ReleaseN(CSemaphore *p, UInt32 num) + { return Semaphore_Release(p, (LONG)num, NULL); } +WRes Semaphore_Release1(CSemaphore *p) { return Semaphore_ReleaseN(p, 1); } + +WRes CriticalSection_Init(CCriticalSection *p) +{ + /* InitializeCriticalSection can raise only STATUS_NO_MEMORY exception */ + #ifdef _MSC_VER + __try + #endif + { + InitializeCriticalSection(p); + /* InitializeCriticalSectionAndSpinCount(p, 0); */ + } + #ifdef _MSC_VER + __except (EXCEPTION_EXECUTE_HANDLER) { return 1; } + #endif + return 0; +} diff --git a/src/lzma/Xz.c b/src/lzma/Xz.c new file mode 100644 index 0000000..2a0f84e --- /dev/null +++ b/src/lzma/Xz.c @@ -0,0 +1,90 @@ +/* Xz.c - Xz +2015-05-01 : Igor Pavlov : Public domain */ + +#include + +#include +#include +#include +#include + +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[i - 1] &= 0x7F; + return i; +} + +void Xz_Construct(CXzStream *p) +{ + p->numBlocks = p->numBlocksAllocated = 0; + p->blocks = 0; + p->flags = 0; +} + +void Xz_Free(CXzStream *p, ISzAlloc *alloc) +{ + alloc->Free(alloc, p->blocks); + p->numBlocks = p->numBlocksAllocated = 0; + p->blocks = 0; +} + +unsigned XzFlags_GetCheckSize(CXzStreamFlags f) +{ + unsigned t = XzFlags_GetCheckType(f); + return (t == 0) ? 0 : (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/src/lzma/XzCrc64.c b/src/lzma/XzCrc64.c new file mode 100644 index 0000000..c5c7704 --- /dev/null +++ b/src/lzma/XzCrc64.c @@ -0,0 +1,86 @@ +/* XzCrc64.c -- CRC64 calculation +2015-03-01 : Igor Pavlov : Public domain */ + +#include + +#include +#include + +#define kCrc64Poly UINT64_CONST(0xC96C5795D7870F42) + +#ifdef MY_CPU_LE + #define CRC_NUM_TABLES 4 +#else + #define CRC_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 *CRC_FUNC)(UInt64 v, const void *data, size_t size, const UInt64 *table); + +static CRC_FUNC g_Crc64Update; +UInt64 g_Crc64Table[256 * CRC_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 & ~((r & 1) - 1)); + g_Crc64Table[i] = r; + } + for (; i < 256 * CRC_NUM_TABLES; i++) + { + UInt64 r = g_Crc64Table[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 * CRC_NUM_TABLES - 1; i >= 256; i--) + { + UInt64 x = g_Crc64Table[i - 256]; + g_Crc64Table[i] = CRC_UINT64_SWAP(x); + } + g_Crc64Update = XzCrc64UpdateT1_BeT4; + } + } + #endif +} diff --git a/src/lzma/XzCrc64Opt.c b/src/lzma/XzCrc64Opt.c new file mode 100644 index 0000000..fbbfd93 --- /dev/null +++ b/src/lzma/XzCrc64Opt.c @@ -0,0 +1,69 @@ +/* XzCrc64Opt.c -- CRC64 calculation +2015-03-01 : Igor Pavlov : Public domain */ + +#include + +#include + +#ifndef MY_CPU_BE + +#define CRC_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) +{ + 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) + { + UInt32 d = (UInt32)v ^ *(const UInt32 *)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 = CRC_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 CRC_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) +{ + 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 = CRC_UPDATE_BYTE_2_BE(v, *p); + for (; size >= 4; size -= 4, p += 4) + { + UInt32 d = (UInt32)(v >> 32) ^ *(const UInt32 *)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 = CRC_UPDATE_BYTE_2_BE(v, *p); + return CRC_UINT64_SWAP(v); +} + +#endif diff --git a/src/lzma/XzDec.c b/src/lzma/XzDec.c new file mode 100644 index 0000000..cf4d5ed --- /dev/null +++ b/src/lzma/XzDec.c @@ -0,0 +1,913 @@ +/* XzDec.c -- Xz Decode +2015-11-09 : Igor Pavlov : Public domain */ + +#include + +/* #define XZ_DUMP */ + +#ifdef XZ_DUMP +#include +#endif + +#include +#include + +#include +#include +#include +#include +#include +#include + +#ifdef USE_SUBBLOCK +#include +#include +#endif + +#include + +#define XZ_CHECK_SIZE_MAX 64 + +#define CODER_BUF_SIZE (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; + + UInt32 methodId; + int encodeMode; + UInt32 delta; + UInt32 ip; + UInt32 x86State; + Byte deltaState[DELTA_STATE_SIZE]; + + Byte buf[BRA_BUF_SIZE]; +} CBraState; + +static void BraState_Free(void *pp, ISzAlloc *alloc) +{ + alloc->Free(alloc, pp); +} + +static SRes BraState_SetProps(void *pp, const Byte *props, size_t propSize, ISzAlloc *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: p->bufConv = isa ## _Convert(p->buf, p->bufTotal, p->ip, p->encodeMode); break; + +static SRes BraState_Code(void *pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, int *wasFinished) +{ + CBraState *p = ((CBraState *)pp); + SizeT destLenOrig = *destLen; + SizeT srcLenOrig = *srcLen; + UNUSED_VAR(finishMode); + *destLen = 0; + *srcLen = 0; + *wasFinished = 0; + while (destLenOrig > 0) + { + if (p->bufPos != p->bufConv) + { + size_t curSize = p->bufConv - p->bufPos; + if (curSize > destLenOrig) + curSize = destLenOrig; + memcpy(dest, p->buf + p->bufPos, curSize); + p->bufPos += curSize; + *destLen += curSize; + dest += curSize; + destLenOrig -= curSize; + continue; + } + p->bufTotal -= p->bufPos; + memmove(p->buf, p->buf + p->bufPos, p->bufTotal); + p->bufPos = 0; + p->bufConv = 0; + { + size_t curSize = BRA_BUF_SIZE - p->bufTotal; + if (curSize > srcLenOrig) + curSize = srcLenOrig; + memcpy(p->buf + p->bufTotal, src, curSize); + *srcLen += curSize; + src += curSize; + srcLenOrig -= curSize; + p->bufTotal += curSize; + } + if (p->bufTotal == 0) + break; + switch (p->methodId) + { + case XZ_ID_Delta: + if (p->encodeMode) + Delta_Encode(p->deltaState, p->delta, p->buf, p->bufTotal); + else + Delta_Decode(p->deltaState, p->delta, p->buf, p->bufTotal); + p->bufConv = p->bufTotal; + break; + case XZ_ID_X86: + p->bufConv = x86_Convert(p->buf, p->bufTotal, 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) + default: + return SZ_ERROR_UNSUPPORTED; + } + p->ip += (UInt32)p->bufConv; + + if (p->bufConv == 0) + { + if (!srcWasFinished) + break; + p->bufConv = p->bufTotal; + } + } + if (p->bufTotal == p->bufPos && srcLenOrig == 0 && srcWasFinished) + *wasFinished = 1; + return SZ_OK; +} + +SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAlloc *alloc) +{ + CBraState *decoder; + if (id != XZ_ID_Delta && + id != XZ_ID_X86 && + id != XZ_ID_PPC && + id != XZ_ID_IA64 && + id != XZ_ID_ARM && + id != XZ_ID_ARMT && + id != XZ_ID_SPARC) + return SZ_ERROR_UNSUPPORTED; + p->p = 0; + decoder = (CBraState *)alloc->Alloc(alloc, sizeof(CBraState)); + if (decoder == 0) + return SZ_ERROR_MEM; + decoder->methodId = (UInt32)id; + decoder->encodeMode = encodeMode; + p->p = decoder; + p->Free = BraState_Free; + p->SetProps = BraState_SetProps; + p->Init = BraState_Init; + p->Code = BraState_Code; + return SZ_OK; +} + +/* ---------- SbState ---------- */ + +#ifdef USE_SUBBLOCK + +static void SbState_Free(void *pp, ISzAlloc *alloc) +{ + CSbDec *p = (CSbDec *)pp; + SbDec_Free(p); + alloc->Free(alloc, pp); +} + +static SRes SbState_SetProps(void *pp, const Byte *props, size_t propSize, ISzAlloc *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_Code(void *pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, int *wasFinished) +{ + 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 */ + return res; +} + +SRes SbState_SetFromMethod(IStateCoder *p, ISzAlloc *alloc) +{ + CSbDec *decoder; + p->p = 0; + decoder = alloc->Alloc(alloc, sizeof(CSbDec)); + if (decoder == 0) + return SZ_ERROR_MEM; + p->p = decoder; + p->Free = SbState_Free; + p->SetProps = SbState_SetProps; + p->Init = SbState_Init; + p->Code = SbState_Code; + SbDec_Construct(decoder); + SbDec_SetAlloc(decoder, alloc); + return SZ_OK; +} +#endif + +/* ---------- Lzma2State ---------- */ + +static void Lzma2State_Free(void *pp, ISzAlloc *alloc) +{ + Lzma2Dec_Free((CLzma2Dec *)pp, alloc); + alloc->Free(alloc, pp); +} + +static SRes Lzma2State_SetProps(void *pp, const Byte *props, size_t propSize, ISzAlloc *alloc) +{ + if (propSize != 1) + return SZ_ERROR_UNSUPPORTED; + return Lzma2Dec_Allocate((CLzma2Dec *)pp, props[0], alloc); +} + +static void Lzma2State_Init(void *pp) +{ + Lzma2Dec_Init((CLzma2Dec *)pp); +} + +static SRes Lzma2State_Code(void *pp, Byte *dest, SizeT *destLen, const Byte *src, SizeT *srcLen, + int srcWasFinished, ECoderFinishMode finishMode, int *wasFinished) +{ + ELzmaStatus status; + /* ELzmaFinishMode fm = (finishMode == LZMA_FINISH_ANY) ? LZMA_FINISH_ANY : LZMA_FINISH_END; */ + SRes res = Lzma2Dec_DecodeToBuf((CLzma2Dec *)pp, dest, destLen, src, srcLen, (ELzmaFinishMode)finishMode, &status); + UNUSED_VAR(srcWasFinished); + *wasFinished = (status == LZMA_STATUS_FINISHED_WITH_MARK); + return res; +} + +static SRes Lzma2State_SetFromMethod(IStateCoder *p, ISzAlloc *alloc) +{ + CLzma2Dec *decoder = (CLzma2Dec *)alloc->Alloc(alloc, sizeof(CLzma2Dec)); + p->p = decoder; + if (decoder == 0) + return SZ_ERROR_MEM; + p->Free = Lzma2State_Free; + p->SetProps = Lzma2State_SetProps; + p->Init = Lzma2State_Init; + p->Code = Lzma2State_Code; + Lzma2Dec_Construct(decoder); + return SZ_OK; +} + + +void MixCoder_Construct(CMixCoder *p, ISzAlloc *alloc) +{ + unsigned i; + p->alloc = alloc; + p->buf = NULL; + p->numCoders = 0; + for (i = 0; i < MIXCODER_NUM_FILTERS_MAX; i++) + p->coders[i].p = NULL; +} + +void MixCoder_Free(CMixCoder *p) +{ + unsigned i; + for (i = 0; i < p->numCoders; i++) + { + IStateCoder *sc = &p->coders[i]; + if (p->alloc && sc->p) + sc->Free(sc->p, p->alloc); + } + p->numCoders = 0; + if (p->buf) + { + p->alloc->Free(p->alloc, p->buf); + p->buf = NULL; /* 9.31: the BUG was fixed */ + } +} + +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); + } +} + +SRes MixCoder_SetFromMethod(CMixCoder *p, unsigned coderIndex, UInt64 methodId) +{ + IStateCoder *sc = &p->coders[coderIndex]; + p->ids[coderIndex] = methodId; + switch (methodId) + { + case XZ_ID_LZMA2: return Lzma2State_SetFromMethod(sc, 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); +} + +SRes MixCoder_Code(CMixCoder *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, int srcWasFinished, + ECoderFinishMode finishMode, ECoderStatus *status) +{ + SizeT destLenOrig = *destLen; + SizeT srcLenOrig = *srcLen; + Bool allFinished = True; + *destLen = 0; + *srcLen = 0; + *status = CODER_STATUS_NOT_FINISHED; + + if (!p->buf) + { + p->buf = (Byte *)p->alloc->Alloc(p->alloc, CODER_BUF_SIZE * (MIXCODER_NUM_FILTERS_MAX - 1)); + if (!p->buf) + return SZ_ERROR_MEM; + } + + if (p->numCoders != 1) + finishMode = CODER_FINISH_ANY; + + for (;;) + { + Bool processed = False; + unsigned i; + /* + 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 *destCur; + SizeT destLenCur, srcLenCur; + const Byte *srcCur; + int srcFinishedCur; + int encodingWasFinished; + + if (i == 0) + { + srcCur = src; + srcLenCur = srcLenOrig - *srcLen; + srcFinishedCur = srcWasFinished; + } + else + { + srcCur = p->buf + (CODER_BUF_SIZE * (i - 1)) + p->pos[i - 1]; + srcLenCur = p->size[i - 1] - p->pos[i - 1]; + srcFinishedCur = p->finished[i - 1]; + } + + if (i == p->numCoders - 1) + { + destCur = dest; + destLenCur = destLenOrig - *destLen; + } + else + { + if (p->pos[i] != p->size[i]) + continue; + destCur = p->buf + (CODER_BUF_SIZE * i); + destLenCur = CODER_BUF_SIZE; + } + + res = coder->Code(coder->p, destCur, &destLenCur, srcCur, &srcLenCur, srcFinishedCur, finishMode, &encodingWasFinished); + + if (!encodingWasFinished) + allFinished = False; + + if (i == 0) + { + *srcLen += srcLenCur; + src += srcLenCur; + } + else + { + p->pos[i - 1] += srcLenCur; + } + + if (i == p->numCoders - 1) + { + *destLen += destLenCur; + dest += destLenCur; + } + else + { + p->size[i] = destLenCur; + p->pos[i] = 0; + p->finished[i] = encodingWasFinished; + } + + if (res != SZ_OK) + return res; + + if (destLenCur != 0 || srcLenCur != 0) + processed = True; + } + if (!processed) + break; + } + if (allFinished) + *status = CODER_STATUS_FINISHED_WITH_MARK; + return SZ_OK; +} + +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 Bool 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) && + memcmp(buf + 10, XZ_FOOTER_SIG, XZ_FOOTER_SIG_SIZE) == 0); +} + +#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_Parse(CXzBlock *p, const Byte *header) +{ + unsigned pos; + unsigned numFilters, i; + unsigned headerSize = (unsigned)header[0] << 2; + + if (CrcCalc(header, headerSize) != GetUi32(header + headerSize)) + return SZ_ERROR_ARCHIVE; + + pos = 1; + if (pos == headerSize) + return SZ_ERROR_ARCHIVE; + p->flags = header[pos++]; + + 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; + } + + 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 + } + + while (pos < headerSize) + if (header[pos++] != 0) + return SZ_ERROR_ARCHIVE; + return SZ_OK; +} + +SRes XzDec_Init(CMixCoder *p, const CXzBlock *block) +{ + unsigned i; + Bool needReInit = True; + unsigned numFilters = XzBlock_GetNumFilters(block); + + if (numFilters == p->numCoders) + { + for (i = 0; i < numFilters; i++) + if (p->ids[i] != block->filters[numFilters - 1 - i].id) + break; + needReInit = (i != numFilters); + } + + if (needReInit) + { + MixCoder_Free(p); + p->numCoders = numFilters; + for (i = 0; i < numFilters; i++) + { + const CXzFilter *f = &block->filters[numFilters - 1 - i]; + RINOK(MixCoder_SetFromMethod(p, i, f->id)); + } + } + + 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; +} + +void XzUnpacker_Construct(CXzUnpacker *p, ISzAlloc *alloc) +{ + MixCoder_Construct(&p->decoder, alloc); + XzUnpacker_Init(p); +} + +void XzUnpacker_Free(CXzUnpacker *p) +{ + MixCoder_Free(&p->decoder); +} + +SRes XzUnpacker_Code(CXzUnpacker *p, Byte *dest, SizeT *destLen, + const Byte *src, SizeT *srcLen, ECoderFinishMode finishMode, ECoderStatus *status) +{ + SizeT destLenOrig = *destLen; + SizeT srcLenOrig = *srcLen; + *destLen = 0; + *srcLen = 0; + *status = CODER_STATUS_NOT_SPECIFIED; + for (;;) + { + SizeT srcRem = srcLenOrig - *srcLen; + + if (p->state == XZ_STATE_BLOCK) + { + SizeT destLen2 = destLenOrig - *destLen; + SizeT srcLen2 = srcLenOrig - *srcLen; + SRes res; + if (srcLen2 == 0 && destLen2 == 0) + { + *status = CODER_STATUS_NOT_FINISHED; + return SZ_OK; + } + + res = MixCoder_Code(&p->decoder, dest, &destLen2, src, &srcLen2, False, finishMode, status); + XzCheck_Update(&p->check, dest, destLen2); + + (*srcLen) += srcLen2; + src += srcLen2; + p->packSize += srcLen2; + + (*destLen) += destLen2; + dest += destLen2; + p->unpackSize += destLen2; + + RINOK(res); + + if (*status == CODER_STATUS_FINISHED_WITH_MARK) + { + Byte temp[32]; + unsigned num = Xz_WriteVarInt(temp, p->packSize + p->blockHeaderSize + XzFlags_GetCheckSize(p->streamFlags)); + num += Xz_WriteVarInt(temp + num, p->unpackSize); + Sha256_Update(&p->sha, temp, num); + p->indexSize += num; + p->numBlocks++; + + p->state = XZ_STATE_BLOCK_FOOTER; + p->pos = 0; + p->alignPos = 0; + } + else if (srcLen2 == 0 && destLen2 == 0) + return SZ_OK; + + continue; + } + + if (srcRem == 0) + { + *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; + p->buf[p->pos++] = *src++; + (*srcLen)++; + } + else + { + RINOK(Xz_ParseHeader(&p->streamFlags, p->buf)); + p->numStartedStreams++; + p->state = XZ_STATE_BLOCK_HEADER; + Sha256_Init(&p->sha); + p->indexSize = 0; + p->numBlocks = 0; + p->pos = 0; + } + break; + } + + case XZ_STATE_BLOCK_HEADER: + { + if (p->pos == 0) + { + p->buf[p->pos++] = *src++; + (*srcLen)++; + if (p->buf[0] == 0) + { + 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; + } + p->blockHeaderSize = ((UInt32)p->buf[0] << 2) + 4; + } + else 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)); + p->numTotalBlocks++; + p->state = XZ_STATE_BLOCK; + p->packSize = 0; + p->unpackSize = 0; + XzCheck_Init(&p->check, XzFlags_GetCheckType(p->streamFlags)); + RINOK(XzDec_Init(&p->decoder, &p->block)); + } + break; + } + + case XZ_STATE_BLOCK_FOOTER: + { + if (((p->packSize + p->alignPos) & 3) != 0) + { + (*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 (cur > srcRem) + cur = (UInt32)srcRem; + memcpy(p->buf + p->pos, src, cur); + p->pos += cur; + (*srcLen) += cur; + src += cur; + } + else + { + 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; + } + } + 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 + { + p->state = XZ_STATE_STREAM_FOOTER; + p->pos = 0; + if (CRC_GET_DIGEST(p->crc) != GetUi32(p->buf)) + 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; + */ +} + +Bool XzUnpacker_IsStreamWasFinished(CXzUnpacker *p) +{ + return (p->state == XZ_STATE_STREAM_PADDING) && (((UInt32)p->padSize & 3) == 0); +} + +UInt64 XzUnpacker_GetExtraSize(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; +} diff --git a/src/lzma/XzEnc.c b/src/lzma/XzEnc.c new file mode 100644 index 0000000..01f342d --- /dev/null +++ b/src/lzma/XzEnc.c @@ -0,0 +1,538 @@ +/* XzEnc.c -- Xz Encode +2015-09-16 : Igor Pavlov : Public domain */ + +#include + +#include +#include + +#include +#include +#include +#include + +#ifdef USE_SUBBLOCK +#include +#include +#include +#endif + +#include + +#define XzBlock_ClearFlags(p) (p)->flags = 0; +#define XzBlock_SetNumFilters(p, n) (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, UInt32 size) +{ + return (s->Write(s, buf, size) == size) ? SZ_OK : SZ_ERROR_WRITE; +} + +static SRes WriteBytesAndCrc(ISeqOutStream *s, const void *buf, UInt32 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); +} + + +static SRes Xz_WriteFooter(CXzStream *p, ISeqOutStream *s) +{ + Byte buf[32]; + UInt64 globalPos; + { + UInt32 crc = CRC_INIT_VAL; + unsigned pos = 1 + Xz_WriteVarInt(buf + 1, p->numBlocks); + size_t i; + + globalPos = pos; + buf[0] = 0; + RINOK(WriteBytesAndCrc(s, buf, pos, &crc)); + + for (i = 0; i < p->numBlocks; i++) + { + const CXzBlockSizes *block = &p->blocks[i]; + pos = Xz_WriteVarInt(buf, block->totalSize); + pos += Xz_WriteVarInt(buf + pos, block->unpackSize); + globalPos += pos; + RINOK(WriteBytesAndCrc(s, buf, pos, &crc)); + } + + pos = ((unsigned)globalPos & 3); + + if (pos != 0) + { + buf[0] = buf[1] = buf[2] = 0; + RINOK(WriteBytesAndCrc(s, buf, 4 - pos, &crc)); + globalPos += 4 - pos; + } + { + SetUi32(buf, CRC_GET_DIGEST(crc)); + RINOK(WriteBytes(s, buf, 4)); + globalPos += 4; + } + } + + { + UInt32 indexSize = (UInt32)((globalPos >> 2) - 1); + SetUi32(buf + 4, indexSize); + buf[8] = (Byte)(p->flags >> 8); + buf[9] = (Byte)(p->flags & 0xFF); + SetUi32(buf, CrcCalc(buf + 4, 6)); + memcpy(buf + 10, XZ_FOOTER_SIG, XZ_FOOTER_SIG_SIZE); + return WriteBytes(s, buf, 12); + } +} + + +static SRes Xz_AddIndexRecord(CXzStream *p, UInt64 unpackSize, UInt64 totalSize, ISzAlloc *alloc) +{ + if (!p->blocks || p->numBlocksAllocated == p->numBlocks) + { + size_t num = p->numBlocks * 2 + 1; + size_t newSize = sizeof(CXzBlockSizes) * num; + CXzBlockSizes *blocks; + if (newSize / sizeof(CXzBlockSizes) != num) + return SZ_ERROR_MEM; + blocks = (CXzBlockSizes *)alloc->Alloc(alloc, newSize); + if (!blocks) + return SZ_ERROR_MEM; + if (p->numBlocks != 0) + { + memcpy(blocks, p->blocks, p->numBlocks * sizeof(CXzBlockSizes)); + alloc->Free(alloc, p->blocks); + } + p->blocks = blocks; + p->numBlocksAllocated = num; + } + { + CXzBlockSizes *block = &p->blocks[p->numBlocks++]; + block->unpackSize = unpackSize; + block->totalSize = totalSize; + } + return SZ_OK; +} + + +/* ---------- CSeqCheckInStream ---------- */ + +typedef struct +{ + ISeqInStream p; + ISeqInStream *realStream; + UInt64 processed; + CXzCheck check; +} CSeqCheckInStream; + +static void SeqCheckInStream_Init(CSeqCheckInStream *p, unsigned mode) +{ + p->processed = 0; + XzCheck_Init(&p->check, mode); +} + +static void SeqCheckInStream_GetDigest(CSeqCheckInStream *p, Byte *digest) +{ + XzCheck_Final(&p->check, digest); +} + +static SRes SeqCheckInStream_Read(void *pp, void *data, size_t *size) +{ + CSeqCheckInStream *p = (CSeqCheckInStream *)pp; + SRes res = p->realStream->Read(p->realStream, data, size); + XzCheck_Update(&p->check, data, *size); + p->processed += *size; + return res; +} + + +/* ---------- CSeqSizeOutStream ---------- */ + +typedef struct +{ + ISeqOutStream p; + ISeqOutStream *realStream; + UInt64 processed; +} CSeqSizeOutStream; + +static size_t MyWrite(void *pp, const void *data, size_t size) +{ + CSeqSizeOutStream *p = (CSeqSizeOutStream *)pp; + size = p->realStream->Write(p->realStream, 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; + +static SRes SeqInFilter_Read(void *pp, void *data, size_t *size) +{ + CSeqInFilter *p = (CSeqInFilter *)pp; + 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(p->realStream->Read(p->realStream, p->buf, &p->endPos)); + if (p->endPos == 0) + p->srcWasFinished = 1; + } + { + SizeT srcLen = p->endPos - p->curPos; + int wasFinished; + SRes res; + *size = sizeOriginal; + res = p->StateCoder.Code(p->StateCoder.p, data, size, p->buf + p->curPos, &srcLen, + p->srcWasFinished, CODER_FINISH_ANY, &wasFinished); + p->curPos += srcLen; + if (*size != 0 || srcLen == 0 || res != 0) + return res; + } + } +} + +static void SeqInFilter_Construct(CSeqInFilter *p) +{ + p->buf = NULL; + p->p.Read = SeqInFilter_Read; +} + +static void SeqInFilter_Free(CSeqInFilter *p) +{ + if (p->buf) + { + g_Alloc.Free(&g_Alloc, p->buf); + p->buf = NULL; + } +} + +SRes BraState_SetFromMethod(IStateCoder *p, UInt64 id, int encodeMode, ISzAlloc *alloc); + +static SRes SeqInFilter_Init(CSeqInFilter *p, const CXzFilter *props) +{ + if (!p->buf) + { + p->buf = g_Alloc.Alloc(&g_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, &g_Alloc)); + RINOK(p->StateCoder.SetProps(p->StateCoder.p, props->props, props->propsSize, &g_Alloc)); + p->StateCoder.Init(p->StateCoder.p); + return SZ_OK; +} + + +/* ---------- CSbEncInStream ---------- */ + +#ifdef USE_SUBBLOCK + +typedef struct +{ + ISeqInStream p; + ISeqInStream *inStream; + CSbEnc enc; +} CSbEncInStream; + +static SRes SbEncInStream_Read(void *pp, void *data, size_t *size) +{ + CSbEncInStream *p = (CSbEncInStream *)pp; + size_t sizeOriginal = *size; + if (sizeOriginal == 0) + return S_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 S_OK; + } +} + +void SbEncInStream_Construct(CSbEncInStream *p, ISzAlloc *alloc) +{ + SbEnc_Construct(&p->enc, alloc); + p->p.Read = SbEncInStream_Read; +} + +SRes SbEncInStream_Init(CSbEncInStream *p) +{ + return SbEnc_Init(&p->enc); +} + +void SbEncInStream_Free(CSbEncInStream *p) +{ + SbEnc_Free(&p->enc); +} + +#endif + + +typedef struct +{ + CLzma2EncHandle lzma2; + #ifdef USE_SUBBLOCK + CSbEncInStream sb; + #endif + CSeqInFilter filter; + ISzAlloc *alloc; + ISzAlloc *bigAlloc; +} CLzma2WithFilters; + + +static void Lzma2WithFilters_Construct(CLzma2WithFilters *p, ISzAlloc *alloc, ISzAlloc *bigAlloc) +{ + p->alloc = alloc; + p->bigAlloc = bigAlloc; + p->lzma2 = NULL; + #ifdef USE_SUBBLOCK + SbEncInStream_Construct(&p->sb, alloc); + #endif + SeqInFilter_Construct(&p->filter); +} + +static SRes Lzma2WithFilters_Create(CLzma2WithFilters *p) +{ + p->lzma2 = Lzma2Enc_Create(p->alloc, p->bigAlloc); + if (!p->lzma2) + return SZ_ERROR_MEM; + return SZ_OK; +} + +static void Lzma2WithFilters_Free(CLzma2WithFilters *p) +{ + SeqInFilter_Free(&p->filter); + #ifdef USE_SUBBLOCK + SbEncInStream_Free(&p->sb); + #endif + if (p->lzma2) + { + Lzma2Enc_Destroy(p->lzma2); + p->lzma2 = NULL; + } +} + + +void XzProps_Init(CXzProps *p) +{ + p->lzma2Props = NULL; + p->filterProps = NULL; + p->checkId = XZ_CHECK_CRC32; +} + +void XzFilterProps_Init(CXzFilterProps *p) +{ + p->id = 0; + p->delta = 0; + p->ip = 0; + p->ipDefined = False; +} + + +static SRes Xz_Compress(CXzStream *xz, CLzma2WithFilters *lzmaf, + ISeqOutStream *outStream, ISeqInStream *inStream, + const CXzProps *props, ICompressProgress *progress) +{ + xz->flags = (Byte)props->checkId; + + RINOK(Lzma2Enc_SetProps(lzmaf->lzma2, props->lzma2Props)); + RINOK(Xz_WriteHeader(xz->flags, outStream)); + + { + CSeqCheckInStream checkInStream; + CSeqSizeOutStream seqSizeOutStream; + CXzBlock block; + unsigned filterIndex = 0; + CXzFilter *filter = NULL; + const CXzFilterProps *fp = props->filterProps; + + 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) + { + SetUi32(filter->props, 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.p.Write = MyWrite; + seqSizeOutStream.realStream = outStream; + seqSizeOutStream.processed = 0; + + RINOK(XzBlock_WriteHeader(&block, &seqSizeOutStream.p)); + + checkInStream.p.Read = SeqCheckInStream_Read; + checkInStream.realStream = inStream; + SeqCheckInStream_Init(&checkInStream, XzFlags_GetCheckType(xz->flags)); + + if (fp) + { + #ifdef USE_SUBBLOCK + if (fp->id == XZ_ID_Subblock) + { + lzmaf->sb.inStream = &checkInStream.p; + RINOK(SbEncInStream_Init(&lzmaf->sb)); + } + else + #endif + { + lzmaf->filter.realStream = &checkInStream.p; + RINOK(SeqInFilter_Init(&lzmaf->filter, filter)); + } + } + + { + UInt64 packPos = seqSizeOutStream.processed; + + SRes res = Lzma2Enc_Encode(lzmaf->lzma2, &seqSizeOutStream.p, + fp ? + #ifdef USE_SUBBLOCK + (fp->id == XZ_ID_Subblock) ? &lzmaf->sb.p: + #endif + &lzmaf->filter.p: + &checkInStream.p, + progress); + + RINOK(res); + block.unpackSize = checkInStream.processed; + block.packSize = seqSizeOutStream.processed - packPos; + } + + { + unsigned padSize = 0; + Byte buf[128]; + while ((((unsigned)block.packSize + padSize) & 3) != 0) + buf[padSize++] = 0; + SeqCheckInStream_GetDigest(&checkInStream, buf + padSize); + RINOK(WriteBytes(&seqSizeOutStream.p, buf, padSize + XzFlags_GetCheckSize(xz->flags))); + RINOK(Xz_AddIndexRecord(xz, block.unpackSize, seqSizeOutStream.processed - padSize, &g_Alloc)); + } + } + return Xz_WriteFooter(xz, outStream); +} + + +SRes Xz_Encode(ISeqOutStream *outStream, ISeqInStream *inStream, + const CXzProps *props, ICompressProgress *progress) +{ + SRes res; + CXzStream xz; + CLzma2WithFilters lzmaf; + Xz_Construct(&xz); + Lzma2WithFilters_Construct(&lzmaf, &g_Alloc, &g_BigAlloc); + res = Lzma2WithFilters_Create(&lzmaf); + if (res == SZ_OK) + res = Xz_Compress(&xz, &lzmaf, outStream, inStream, props, progress); + Lzma2WithFilters_Free(&lzmaf); + Xz_Free(&xz, &g_Alloc); + return res; +} + + +SRes Xz_EncodeEmpty(ISeqOutStream *outStream) +{ + SRes res; + CXzStream xz; + Xz_Construct(&xz); + res = Xz_WriteHeader(xz.flags, outStream); + if (res == SZ_OK) + res = Xz_WriteFooter(&xz, outStream); + Xz_Free(&xz, &g_Alloc); + return res; +} diff --git a/src/lzma/XzIn.c b/src/lzma/XzIn.c new file mode 100644 index 0000000..9dffa8b --- /dev/null +++ b/src/lzma/XzIn.c @@ -0,0 +1,313 @@ +/* XzIn.c - Xz input +2015-11-08 : Igor Pavlov : Public domain */ + +#include + +#include + +#include +#include +#include + +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, Bool *isIndex, UInt32 *headerSizeRes) +{ + Byte header[XZ_BLOCK_HEADER_SIZE_MAX]; + unsigned headerSize; + *headerSizeRes = 0; + RINOK(SeqInStream_ReadByte(inStream, &header[0])); + headerSize = ((unsigned)header[0] << 2) + 4; + if (headerSize == 0) + { + *headerSizeRes = 1; + *isIndex = True; + return SZ_OK; + } + + *isIndex = False; + *headerSizeRes = headerSize; + RINOK(SeqInStream_Read(inStream, header + 1, headerSize - 1)); + return XzBlock_Parse(p, header); +} + +#define ADD_SIZE_CHECH(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_CHECH(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_CHECH(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, ISzAlloc *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->numBlocksAllocated = numBlocks; + p->blocks = alloc->Alloc(alloc, sizeof(CXzBlockSizes) * numBlocks); + if (p->blocks == 0) + 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, ISzAlloc *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 = alloc->Alloc(alloc, size); + if (buf == 0) + return SZ_ERROR_MEM; + res = LookInStream_Read2(stream, buf, size, SZ_ERROR_UNSUPPORTED); + if (res == SZ_OK) + res = Xz_ReadIndex2(p, buf, size, alloc); + alloc->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, ISzAlloc *alloc) +{ + UInt64 indexSize; + Byte buf[XZ_STREAM_FOOTER_SIZE]; + UInt64 pos = *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 (memcmp(buf + 10, XZ_FOOTER_SIG, XZ_FOOTER_SIG_SIZE) != 0) + { + 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 (memcmp(buf + 10, XZ_FOOTER_SIG, XZ_FOOTER_SIG_SIZE) != 0) + return SZ_ERROR_NO_ARCHIVE; + } + + p->flags = (CXzStreamFlags)GetBe16(buf + 8); + + if (!XzFlags_IsSupported(p->flags)) + return SZ_ERROR_UNSUPPORTED; + + if (GetUi32(buf) != 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 = pos; + } + { + CXzStreamFlags headerFlags; + CSecToRead secToRead; + SecToRead_CreateVTable(&secToRead); + secToRead.realStream = stream; + + RINOK(Xz_ReadHeader(&headerFlags, &secToRead.s)); + 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, ISzAlloc *alloc) +{ + size_t i; + for (i = 0; i < p->num; i++) + Xz_Free(&p->streams[i], alloc); + alloc->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_CHECH(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_CHECH(size, Xz_GetTotalSize(&p->streams[i])); + return size; +} +*/ + +SRes Xzs_ReadBackward(CXzs *p, ILookInStream *stream, Int64 *startOffset, ICompressProgress *progress, ISzAlloc *alloc) +{ + Int64 endOffset = 0; + RINOK(stream->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 = *startOffset; + RINOK(res); + if (p->num == p->numAllocated) + { + size_t newNum = p->num + p->num / 4 + 1; + Byte *data = (Byte *)alloc->Alloc(alloc, newNum * sizeof(CXzStream)); + if (data == 0) + return SZ_ERROR_MEM; + p->numAllocated = newNum; + if (p->num != 0) + memcpy(data, p->streams, p->num * sizeof(CXzStream)); + alloc->Free(alloc, p->streams); + p->streams = (CXzStream *)data; + } + p->streams[p->num++] = st; + if (*startOffset == 0) + break; + RINOK(LookInStream_SeekTo(stream, *startOffset)); + if (progress && progress->Progress(progress, endOffset - *startOffset, (UInt64)(Int64)-1) != SZ_OK) + return SZ_ERROR_PROGRESS; + } + return SZ_OK; +} diff --git a/src/render.cpp b/src/render.cpp index 71cde40..ab12b4d 100644 --- a/src/render.cpp +++ b/src/render.cpp @@ -856,11 +856,34 @@ D3D9SetViewport_Detour (IDirect3DDevice9* This, if (This != tbf::RenderFix::pDevice) return D3D9SetViewport_Original (This, pViewport); + // In-Game Map + // + // + if ( config.render.fix_map_res && ( + ( pViewport->Width == tbf::RenderFix::width/2 && pViewport->Height == tbf::RenderFix::height/2 ) || + ( pViewport->Width == tbf::RenderFix::width/3 && pViewport->Height == tbf::RenderFix::height/3 ) || + ( pViewport->Width == tbf::RenderFix::width/4 && pViewport->Height == tbf::RenderFix::height/4 ) || + ( pViewport->Width == tbf::RenderFix::width/5 && pViewport->Height == tbf::RenderFix::height/5 ) || + ( pViewport->Width == tbf::RenderFix::width/6 && pViewport->Height == tbf::RenderFix::height/6 ) || + ( pViewport->Width == tbf::RenderFix::width/7 && pViewport->Height == tbf::RenderFix::height/7 ) || + ( pViewport->Width == tbf::RenderFix::width/8 && pViewport->Height == tbf::RenderFix::height/8 ) || + ( pViewport->Width == tbf::RenderFix::width/9 && pViewport->Height == tbf::RenderFix::height/9 ) || + ( pViewport->Width == tbf::RenderFix::width/10 && pViewport->Height == tbf::RenderFix::height/10 ) ) ) + { + D3DVIEWPORT9 rescaled_map = *pViewport; + + rescaled_map.Width = tbf::RenderFix::width; + rescaled_map.Height = tbf::RenderFix::height; + + return D3D9SetViewport_Original (This, &rescaled_map); + } + // // Adjust Character Drop Shadows // if (pViewport->Width == pViewport->Height && - (pViewport->Width == 64 || pViewport->Width == 128 || pViewport->Width == 256)) { + (pViewport->Width == 64 || pViewport->Width == 128 || pViewport->Width == 256)) +{ D3DVIEWPORT9 rescaled_shadow = *pViewport; uint32_t shift = TBF_MakeShadowBitShift (pViewport->Width); diff --git a/src/sound.cpp b/src/sound.cpp index a6d0cc4..7e3016d 100644 --- a/src/sound.cpp +++ b/src/sound.cpp @@ -1,12 +1,12 @@ /** - * This file is part of Tales of Zestiria "Fix". + * This file is part of Tales of Berseria "Fix". * - * Tales of Zestiria "Fix" is free software : you can redistribute it + * Tales of Berseria "Fix" is free software : you can redistribute it * and/or modify it under the terms of the GNU General Public License * as published by The Free Software Foundation, either version 3 of * the License, or (at your option) any later version. * - * Tales of Zestiria "Fix" is distributed in the hope that it will be + * Tales of Berseria "Fix" is distributed in the hope that it will be * useful, * * But WITHOUT ANY WARRANTY; without even the implied warranty of @@ -14,7 +14,7 @@ * GNU General Public License for more details. * * You should have received a copy of the GNU General Public License - * along with Tales of Zestiria "Fix". + * along with Tales of Berseria "Fix". * * If not, see . * @@ -59,7 +59,7 @@ HMODULE tbf::SoundFix::ole32_dll = 0; WAVEFORMATEXTENSIBLE g_DeviceFormat; const wchar_t* -tbf_DescribeHRESULT (HRESULT result) +TBF_DescribeHRESULT (HRESULT result) { switch (result) { @@ -173,7 +173,7 @@ tbf_DescribeHRESULT (HRESULT result) #define DSOUND_CALL(_Ret, _Call) { \ audio_log->LogEx (true, L" Calling original function: "); \ (_Ret) = (_Call); \ - audio_log->LogEx (false, L"(ret=%s)\n\n", tbf_DescribeHRESULT (_Ret));\ + audio_log->LogEx (false, L"(ret=%s)\n\n", TBF_DescribeHRESULT (_Ret));\ } typedef HRESULT (WINAPI *DSound_GetSpeakerConfig_t) @@ -281,8 +281,8 @@ DirectSoundCreate_Detour (_In_opt_ LPCGUID pcGuidDevice, _Outptr_ LPDIRECTSOUND *ppDS, _Pre_null_ LPUNKNOWN pUnkOuter) { - new_session = false; - g_DeviceFormat.Format.cbSize = 0; + //new_session = false; + //g_DeviceFormat.Format.cbSize = 0; audio_log->Log ( L"[!] %s (%08Xh, %08Xh, %08Xh) - " L"[Calling Thread: 0x%04x]", @@ -327,7 +327,7 @@ DirectSoundCreate_Detour (_In_opt_ LPCGUID pcGuidDevice, } else { _com_error ce (ret); - audio_log->Log ( L" > FAILURE %s - (%s)", tbf_DescribeHRESULT (ret), + audio_log->Log ( L" > FAILURE %s - (%s)", TBF_DescribeHRESULT (ret), ce.ErrorMessage () ); } @@ -446,13 +446,14 @@ IAudioClient_GetMixFormat_Detour (IAudioClient *This, #define TARGET_SAMPLE_RATE config.audio.sample_hz - if (pMixFormat->nSamplesPerSec != TARGET_SAMPLE_RATE) { + if (TARGET_SAMPLE_RATE != -1 && pMixFormat->nSamplesPerSec != TARGET_SAMPLE_RATE) { audio_log->Log ( L" ** Resampling Audiostream from %lu Hz to %lu Hz", pMixFormat->nSamplesPerSec, TARGET_SAMPLE_RATE ); pMixFormat->nSamplesPerSec = TARGET_SAMPLE_RATE; } pMixFormat->nAvgBytesPerSec = (pMixFormat->nSamplesPerSec * pMixFormat->nChannels * pMixFormat->wBitsPerSample) >> 3; + pMixFormat->nBlockAlign = (pMixFormat->wBitsPerSample * pMixFormat->nChannels) >> 3; g_DeviceFormat.Format.cbSize = 22; g_DeviceFormat.Format.nSamplesPerSec = pMixFormat->nSamplesPerSec; @@ -460,16 +461,15 @@ IAudioClient_GetMixFormat_Detour (IAudioClient *This, g_DeviceFormat.Format.nBlockAlign = pMixFormat->nBlockAlign; g_DeviceFormat.Format.nAvgBytesPerSec = pMixFormat->nAvgBytesPerSec; g_DeviceFormat.Format.wBitsPerSample = pMixFormat->wBitsPerSample; - //g_DeviceFormat.Format.wBitsPerSample = ((PWAVEFORMATEXTENSIBLE)pMixFormat)->Samples.wValidBitsPerSample; // We may have gotten an extensible wave format, but... we need to // truncate this sucker to a plain old WAVEFORMATEX - IAudioClient_GetMixFormat_Original (This, ppDeviceFormat); + HRESULT hr = IAudioClient_GetMixFormat_Original (This, ppDeviceFormat); memcpy (*ppDeviceFormat, pMixFormat, (*ppDeviceFormat)->cbSize); pStore->Release (); - return S_OK; + return hr; } pStore->Release (); @@ -509,7 +509,6 @@ IAudioClient_Initialize_Detour (IAudioClient *This, L" >> Channels: %lu, Samples Per Sec: %lu, Bits Per Sample: %hu\n", pFormat->nChannels, pFormat->nSamplesPerSec, pFormat->wBitsPerSample ); -#if 0 WAVEFORMATEX format; format.cbSize = pFormat->cbSize; @@ -520,16 +519,15 @@ IAudioClient_Initialize_Detour (IAudioClient *This, format.wBitsPerSample = pFormat->wBitsPerSample; format.wFormatTag = pFormat->wFormatTag; - WAVEFORMATEX *pClosestMatch = - (PWAVEFORMATEX)CoTaskMemAlloc (sizeof (WAVEFORMATEXTENSIBLE)); - +#if 0 if (This->IsFormatSupported (AUDCLNT_SHAREMODE_SHARED, pFormat, &pClosestMatch) == S_OK) { CoTaskMemFree (pClosestMatch); pClosestMatch = &format; } #endif - WAVEFORMATEX *pClosestMatch = (WAVEFORMATEX *)pFormat; + WAVEFORMATEX *pClosestMatch = + (PWAVEFORMATEX)CoTaskMemAlloc (sizeof (WAVEFORMATEXTENSIBLE)); #define AUDCLNT_STREAMFLAGS_AUTOCONVERTPCM 0x80000000 #define AUDCLNT_STREAMFLAGS_SRC_DEFAULT_QUALITY 0x08000000 @@ -551,14 +549,12 @@ IAudioClient_Initialize_Detour (IAudioClient *This, AUDCLNT_STREAMFLAGS_UNKNOWN4000000 ); } - if (g_DeviceFormat.Format.cbSize == 22) { - pClosestMatch->nChannels = g_DeviceFormat.Format.nChannels; - pClosestMatch->nSamplesPerSec = g_DeviceFormat.Format.nSamplesPerSec; - //pClosestMatch->nBlockAlign = g_DeviceFormat.Format.nBlockAlign; - //pClosestMatch->nAvgBytesPerSec = g_DeviceFormat.Format.nAvgBytesPerSec; - //pClosestMatch->wBitsPerSample = g_DeviceFormat.Format.wBitsPerSample; + if (This->IsFormatSupported (AUDCLNT_SHAREMODE_SHARED, &format, &pClosestMatch) != S_OK) { + CoTaskMemFree (pClosestMatch); + pClosestMatch = (WAVEFORMATEX *)pFormat; } + HRESULT ret = IAudioClient_Initialize_Original (This, AUDCLNT_SHAREMODE_SHARED, StreamFlags, hnsBufferDuration, @@ -570,15 +566,13 @@ IAudioClient_Initialize_Detour (IAudioClient *This, _com_error error (ret); -#if 0 - if (pClosestMatch != &format) + if (pClosestMatch != pFormat) CoTaskMemFree (pClosestMatch); -#endif audio_log->Log ( L" Result: 0x%04X (%s)\n", ret - AUDCLNT_ERR (0x0000), error.ErrorMessage () ); - //new_session = false; + new_session = false; //DSOUND_VIRTUAL_OVERRIDE ( ppAudioClient, 8, "IAudioClient::GetMixFormat", //IAudioClient_GetMixFormat_Original, //IAudioClient_GetMixFormat_Original, @@ -755,6 +749,7 @@ tbf::SoundFix::Init (void) dsound_dll = LoadLibrary (L"dsound.dll"); ole32_dll = LoadLibrary (L"Ole32.dll"); +#if 0 audio_log->LogEx (true, L"@ Hooking DirectSoundCreate... "); TBF_CreateDLLHook ( L"dsound.dll", "DirectSoundCreate", @@ -770,6 +765,7 @@ tbf::SoundFix::Init (void) // when the code was originally written -- test this in the future. DirectSoundCreate_Detour (NULL, &g_pDS, NULL); g_pDS->SetCooperativeLevel (NULL, DSSCL_EXCLUSIVE); +#endif new_session = true; @@ -788,17 +784,19 @@ tbf::SoundFix::Init (void) void tbf::SoundFix::Shutdown (void) { - if (! config.audio.enable_fix) - return; + ////if (! config.audio.enable_fix) + ////return; - //tbf_RemoveHook (pfnCoCreateInstance); - //tbf_RemoveHook (pfnDirectSoundCreate); + //TBF_RemoveHook (pfnCoCreateInstance); + //TBF_RemoveHook (pfnDirectSoundCreate); FreeLibrary (dsound_dll); FreeLibrary (ole32_dll); - audio_log->Log (L"Closing log file..."); - audio_log->close (); + if (audio_log) { + audio_log->Log (L"Closing log file..."); + audio_log->close (); + } } diff --git a/src/tbt/control_panel.cpp b/src/tbt/control_panel.cpp index efe96cb..b79e580 100644 --- a/src/tbt/control_panel.cpp +++ b/src/tbt/control_panel.cpp @@ -325,6 +325,25 @@ TBFix_DrawConfigUI (void) { ImGui::TreePush (""); + if (ImGui::CollapsingHeader ("Experimental")) { + ImGui::TreePush (""); + + need_restart |= ImGui::Checkbox ("Enable map menu resolution fix", &config.render.fix_map_res); + + if (ImGui::IsItemHovered ()) + { + ImGui::BeginTooltip (); + ImGui::Text ( "This feature is experimental and only works at 16:9 resolutions, I am not positive that it" + " will not break something unrelated." ); + ImGui::Separator (); + ImGui::Bullet (); ImGui::SameLine (); + ImGui::Text ("That is where you come in, my faithful guinea pig >:)" ); + ImGui::EndTooltip (); + } + + ImGui::TreePop (); + } + if (ImGui::CollapsingHeader ("Quality")) { ImGui::TreePush (""); @@ -481,7 +500,41 @@ TBFix_DrawConfigUI (void) ImGui::Text ("Enabling this will cause the game to run slower and waste disk space, only enable if you know what you are doing."); ImGui::EndTooltip (); } - ImGui::TreePop (); + + ImGui::BeginChild ("ModDescription", ImVec2 (750, 325), true); + ImGui::TextColored (ImVec4 (0.9f, 0.7f, 0.5f, 1.0f), "Texture Modding Overview"); ImGui::SameLine (); + ImGui::Text (" (Documentation Pending)"); + + ImGui::Separator (); + + ImGui::TextWrapped ("\nReplacement textures go in (TBFix_Res\\inject\\textures\\{blocking|streaming}\\.dds)\n\n"); + + ImGui::TreePush (""); + ImGui::BulletText ("Blocking textures have a high performance penalty, but zero chance of visible pop-in."); + ImGui::BulletText ("Streaming textures will replace the game's original texture whenever Disk/CPU loads finish."); + ImGui::TreePush (""); + ImGui::PushStyleColor (ImGuiCol_Text, ImVec4 (0.6f, 0.9f, 0.2f, 1.0f)); + ImGui::BulletText ("Use streaming whenever possible or performance will bite you in the ass."); + ImGui::PopStyleColor (); + ImGui::TreePop ( ); + ImGui::TreePop (); + + ImGui::TextWrapped ("\n\nLoading modified textures from separate files is inefficient; entire groups of textures may also be packaged into \".7z\" files (See TBFix_Res\\inject\\00_License.7z as an example, and use low/no compression ratio or you will kill the game's performance).\n"); + + ImGui::Separator (); + + ImGui::PushStyleColor (ImGuiCol_Text, ImVec4 (0.9f, 0.6f, 0.3f, 1.0f)); + ImGui::TextWrapped ( "\n\nA more detailed synopsis will follow in future versions, for now please refer to the GitHub release notes for Tales of Symphonia " + "\"Fix\" v 0.9.0 for a thorough description on authoring texture mods.\n\n" ); + ImGui::PopStyleColor (); + + ImGui::Separator (); + + ImGui::Bullet (); ImGui::SameLine (); + ImGui::TextWrapped ( "\nIf texture mods are enabled, you can click on the Injected and Base buttons on the texture cache " + "summary pannel to compare modified and unmodified." ); + ImGui::EndChild (); + ImGui::TreePop (); } ImGui::TreePop (); } @@ -544,12 +597,12 @@ TBFix_DrawConfigUI (void) ImGui::TreePush (""); ImGui::Checkbox ("Swap WASD and Arrow Keys", &config.keyboard.swap_wasd); ImGui::TreePop ( ); - - ImGui::TreePop (); } if (ImGui::CollapsingHeader ("Audio Configuration", ImGuiTreeNodeFlags_CollapsingHeader | ImGuiTreeNodeFlags_DefaultOpen)) { + ImGui::TreePush (""); + if (tbf::SoundFix::wasapi_init) { ImGui::PushStyleVar (ImGuiStyleVar_ChildWindowRounding, 16.0f); @@ -607,6 +660,8 @@ TBFix_DrawConfigUI (void) ImGui::SetTooltip ("May reduce audio quality, but can help with some weird USB headsets and Windows 7 / Older."); ImGui::TreePop (); } + + ImGui::TreePop (); } ImGui::PopItemWidth (); diff --git a/src/textures.cpp b/src/textures.cpp index 18df841..cf37c2a 100644 --- a/src/textures.cpp +++ b/src/textures.cpp @@ -37,6 +37,13 @@ #include "command.h" +#include +#include +#include +#include +#include +#include + #define TBFIX_TEXTURE_DIR L"TBFix_Res" #define TBFIX_TEXTURE_EXT L".dds" @@ -58,21 +65,21 @@ typedef HRESULT (STDMETHODCALLTYPE *SetRenderState_pfn) ); -static D3DXSaveTextureToFile_pfn D3DXSaveTextureToFile = nullptr; -static D3DXCreateTextureFromFileInMemoryEx_pfn D3DXCreateTextureFromFileInMemoryEx = nullptr; +static D3DXSaveTextureToFile_pfn D3DXSaveTextureToFile = nullptr; +static D3DXCreateTextureFromFileInMemoryEx_pfn D3DXCreateTextureFromFileInMemoryEx_Original = nullptr; -static BeginScene_pfn D3D9BeginScene = nullptr; -static EndScene_pfn D3D9EndScene = nullptr; - SetRenderState_pfn D3D9SetRenderState = nullptr; +static BeginScene_pfn D3D9BeginScene = nullptr; +static EndScene_pfn D3D9EndScene = nullptr; + SetRenderState_pfn D3D9SetRenderState = nullptr; -static StretchRect_pfn D3D9StretchRect = nullptr; -static CreateTexture_pfn D3D9CreateTexture = nullptr; -static CreateRenderTarget_pfn D3D9CreateRenderTarget = nullptr; -static CreateDepthStencilSurface_pfn D3D9CreateDepthStencilSurface = nullptr; +static StretchRect_pfn D3D9StretchRect = nullptr; +static CreateTexture_pfn D3D9CreateTexture = nullptr; +static CreateRenderTarget_pfn D3D9CreateRenderTarget = nullptr; +static CreateDepthStencilSurface_pfn D3D9CreateDepthStencilSurface = nullptr; -static SetTexture_pfn D3D9SetTexture = nullptr; -static SetRenderTarget_pfn D3D9SetRenderTarget = nullptr; -static SetDepthStencilSurface_pfn D3D9SetDepthStencilSurface = nullptr; +static SetTexture_pfn D3D9SetTexture = nullptr; +static SetRenderTarget_pfn D3D9SetRenderTarget = nullptr; +static SetDepthStencilSurface_pfn D3D9SetDepthStencilSurface = nullptr; extern uint32_t TBF_MakeShadowBitShift (uint32_t dim); @@ -103,10 +110,10 @@ enum tbf_load_method_t { }; struct tbf_tex_record_s { - int archive = -1; - int fileno = 0UL; - enum tbf_load_method_t method = DontCare; - size_t size = 0UL; + unsigned int archive = std::numeric_limits ::max (); + int fileno = 0UL; + enum tbf_load_method_t method = DontCare; + size_t size = 0UL; }; bool pending_loads (void); @@ -127,6 +134,10 @@ std::set dumped_textures; std::vector textures_last_frame; std::set textures_used; +// Textures that we will not allow injection for +// (primarily to speed things up, but also for EULA-related reasons). +std::set inject_blacklist; + std::wstring SK_D3D9_UsageToStr (DWORD dwUsage) { @@ -686,20 +697,53 @@ D3D9CreateTexture_Detour (IDirect3DDevice9 *This, L"%lu, %lu, %08Xh, %08Xh)", Width, Height, Levels, Usage, Format, Pool, ppTexture, pSharedHandle); -#endif -#if 0 - //if (config.textures.log) { - tex_log->Log ( L"[Load Trace] >> Creating Texture: " - L"(%d x %d), Format: %s, Usage: [%s], Pool: %s", - Width, Height, - SK_D3D9_FormatToStr (Format), - SK_D3D9_UsageToStr (Usage).c_str (), - SK_D3D9_PoolToStr (Pool) ); - //} -//#endif + tex_log->Log ( L"[Load Trace] >> Creating Texture: " + L"(%d x %d), Format: %s, Usage: [%s], Pool: %s", + Width, Height, + SK_D3D9_FormatToStr (Format), + SK_D3D9_UsageToStr (Usage).c_str (), + SK_D3D9_PoolToStr (Pool) ); #endif + if (Usage == D3DUSAGE_RENDERTARGET && config.render.fix_map_res) + { + if (Width == tbf::RenderFix::width / 10 && Height == tbf::RenderFix::height / 10) { + Width = tbf::RenderFix::width; Height = tbf::RenderFix::height; + } + if (Width == tbf::RenderFix::width / 9 && Height == tbf::RenderFix::height / 9) { + Width = tbf::RenderFix::width; Height = tbf::RenderFix::height; + } + if (Width == tbf::RenderFix::width / 8 && Height == tbf::RenderFix::height / 8) { + Width = tbf::RenderFix::width; Height = tbf::RenderFix::height; + } + if (Width == tbf::RenderFix::width / 7 && Height == tbf::RenderFix::height / 7) { + Width = tbf::RenderFix::width; Height = tbf::RenderFix::height; + } + if (Width == tbf::RenderFix::width / 6 && Height == tbf::RenderFix::height / 6) { + Width = tbf::RenderFix::width; Height = tbf::RenderFix::height; + } + if (Width == tbf::RenderFix::width / 5 && Height == tbf::RenderFix::height / 5) { + Width = tbf::RenderFix::width; Height = tbf::RenderFix::height; + } + if (Width == tbf::RenderFix::width / 4 && Height == tbf::RenderFix::height / 4) { + Width = tbf::RenderFix::width; Height = tbf::RenderFix::height; + } + if (Width == tbf::RenderFix::width / 3 && Height == tbf::RenderFix::height / 3) { + Width = tbf::RenderFix::width; Height = tbf::RenderFix::height; + } + if (Width == tbf::RenderFix::width / 2 && Height == tbf::RenderFix::height / 2) + { + Width = tbf::RenderFix::width; Height = tbf::RenderFix::height; + } + //if (Width == 960 && Height == 540) { + //Width = 3840; Height = 2160; + //} + //if (Width == 640 && Height == 360) { + //Width = 3840; Height = 2160; + //} + } + // // Model Shadows // @@ -1265,6 +1309,62 @@ friend class SK_TextureWorkerThread; HANDLE spool_thread_; } *resample_pool = nullptr; +// +// Split stream jobs into small and large in order to prevent +// starvation from wreaking havoc on load times. +// +// This is a simple, but remarkably effective approach and +// further optimization work probably will not be done. +// +struct SK_StreamSplitter { + bool working (void) { + if (lrg_tex && lrg_tex->working ()) + return true; + + if (sm_tex && sm_tex->working ()) + return true; + + return false; + } + + int queueLength (void) { + int len = 0; + + if (lrg_tex) len += lrg_tex->queueLength (); + if (sm_tex) len += sm_tex->queueLength (); + + return len; + } + + std::vector getFinished (void) + { + std::vector results; + + std::vector lrg_loads; + std::vector sm_loads; + + if (lrg_tex) lrg_loads = lrg_tex->getFinished (); + if (sm_tex) sm_loads = sm_tex->getFinished (); + + results.insert (results.begin (), lrg_loads.begin (), lrg_loads.end ()); + results.insert (results.begin (), sm_loads.begin (), sm_loads.end ()); + + return results; + } + + void postJob (tbf_tex_load_s* job) + { + // A "Large" load is one >= 128 KiB + if (job->SrcDataSize > (128 * 1024)) + lrg_tex->postJob (job); + else + sm_tex->postJob (job); + } + + SK_TextureThreadPool* lrg_tex = nullptr; + SK_TextureThreadPool* sm_tex = nullptr; +} stream_pool; + std::queue textures_to_stream; std::map @@ -1294,7 +1394,9 @@ pending_loads (void) { bool ret = false; - return resample_pool != nullptr && resample_pool->working (); + return + ( stream_pool.working () || + ( resample_pool != nullptr && resample_pool->working () ) ); // EnterCriticalSection (&cs_tex_inject); // ret = (! finished_loads.empty ()); @@ -1329,27 +1431,9 @@ pending_streams (void) { bool ret = false; - bool stream = false; - bool resample = false; - - - if (InterlockedIncrement (&streaming) > 1) - stream = true; - - InterlockedDecrement (&streaming); - - if (stream) + if (streaming || stream_pool.queueLength () || (resample_pool && resample_pool->queueLength ())) ret = true; - - if (InterlockedIncrement (&resampling) > 1) - resample = true; - - InterlockedDecrement (&resampling); - - if (resample || resample_pool->queueLength ()) - return true; - return ret; } @@ -1438,19 +1522,34 @@ namespace streaming_memory { HRESULT InjectTexture (tbf_tex_load_s* load) { - D3DXIMAGE_INFO img_info = { 0 }; + D3DXIMAGE_INFO img_info; bool streamed; - DWORD size = 0; + size_t size = 0; HRESULT hr = E_FAIL; + auto inject = injectable_textures.find (load->checksum); + + if (inject == injectable_textures.end ()) + { + tex_log->Log ( L"[Inject Tex] >> Load Request for Checksum: %X " + L"has no Injection Record !!", + load->checksum ); + + return E_NOT_VALID_STATE; + } + + const tbf_tex_record_s* inj_tex = + &(*inject).second; + streamed = - injectable_textures [load->checksum].method == Streaming; + (inj_tex->method == Streaming); // // Load: From Regular Filesystem // - if (injectable_textures [load->checksum].archive == -1) { + if ( inj_tex->archive == std::numeric_limits ::max () ) + { HANDLE hTexFile = CreateFile ( load->wszFilename, GENERIC_READ, @@ -1465,14 +1564,15 @@ InjectTexture (tbf_tex_load_s* load) if (hTexFile != INVALID_HANDLE_VALUE) { size = GetFileSize (hTexFile, nullptr); - load->pSrcData = new uint8_t [size]; - if (load->pSrcData != nullptr) { - ReadFile (hTexFile, load->pSrcData, size, &read, nullptr); + if (streaming_memory::alloc (size)) { + load->pSrcData = streaming_memory::data [GetCurrentThreadId ()]; + + ReadFile (hTexFile, load->pSrcData, (DWORD)size, &read, nullptr); load->SrcDataSize = read; - if (streamed && size > (16 * 1024)) { + if (streamed && size > (32 * 1024)) { SetThreadPriority ( GetCurrentThread (), THREAD_PRIORITY_BELOW_NORMAL | THREAD_MODE_BACKGROUND_BEGIN ); @@ -1483,7 +1583,7 @@ InjectTexture (tbf_tex_load_s* load) load->SrcDataSize, &img_info ); - hr = D3DXCreateTextureFromFileInMemoryEx ( + hr = D3DXCreateTextureFromFileInMemoryEx_Original ( load->pDevice, load->pSrcData, load->SrcDataSize, D3DX_DEFAULT, D3DX_DEFAULT, img_info.MipLevels, @@ -1494,7 +1594,6 @@ InjectTexture (tbf_tex_load_s* load) &img_info, nullptr, &load->pSrc ); - delete [] load->pSrcData; load->pSrcData = nullptr; } else { // OUT OF MEMORY ?! @@ -1507,10 +1606,129 @@ InjectTexture (tbf_tex_load_s* load) // // Load: From (Compressed) Archive (.7z or .zip) // - else { + else + { + wchar_t arc_name [MAX_PATH]; + CFileInStream arc_stream; + CLookToRead look_stream; + ISzAlloc thread_alloc; + ISzAlloc thread_tmp_alloc; + + FileInStream_CreateVTable (&arc_stream); + LookToRead_CreateVTable (&look_stream, False); + + look_stream.realStream = &arc_stream.s; + LookToRead_Init (&look_stream); + + thread_alloc.Alloc = SzAlloc; + thread_alloc.Free = SzFree; + + thread_tmp_alloc.Alloc = SzAllocTemp; + thread_tmp_alloc.Free = SzFreeTemp; + + CSzArEx arc; + size = inj_tex->size; + int fileno = inj_tex->fileno; + + if (inj_tex->archive <= archives.size ()) + wcscpy (arc_name, archives [inj_tex->archive].c_str ()); + else + wcscpy (arc_name, L"INVALID"); + + if (streamed && size > (32 * 1024)) { + SetThreadPriority ( GetCurrentThread (), + THREAD_PRIORITY_LOWEST | + THREAD_MODE_BACKGROUND_BEGIN ); + } + + if (InFile_OpenW (&arc_stream.file, arc_name)) + { + tex_log->Log ( L"[Inject Tex] ** Cannot open archive file: %s", + arc_name ); + return E_FAIL; + } + + SzArEx_Init (&arc); + + if (SzArEx_Open (&arc, &look_stream.s, &thread_alloc, &thread_tmp_alloc) != SZ_OK) + { + tex_log->Log ( L"[Inject Tex] ** Cannot open archive file: %s", + arc_name ); + return E_FAIL; + } + + if (streaming_memory::alloc (size)) + { + load->pSrcData = streaming_memory::data [GetCurrentThreadId ()]; + + bool wait = true; + + while (wait) + { + DWORD dwResult = WAIT_OBJECT_0; + + if (streamed && size > (32 * 1024)) { + dwResult = + WaitForSingleObject ( decomp_semaphore, INFINITE ); + } + + switch (dwResult) + { + case WAIT_OBJECT_0: + { + uint32_t block_idx = 0xFFFFFFFF; + Byte* out = (Byte *)streaming_memory::data [GetCurrentThreadId ()]; + size_t out_len = streaming_memory::data_len [GetCurrentThreadId ()]; + size_t offset = 0; + size_t decomp_size = 0; + + SzArEx_Extract ( &arc, &look_stream.s, fileno, + &block_idx, &out, &out_len, + &offset, &decomp_size, + &thread_alloc, &thread_tmp_alloc ); + + if (streamed && size > (32 * 1024)) + ReleaseSemaphore (decomp_semaphore, 1, nullptr); + + wait = false; + + load->pSrcData = (Byte *)streaming_memory::data [GetCurrentThreadId ()] + offset; + load->SrcDataSize = (UINT)decomp_size; + + D3DXGetImageInfoFromFileInMemory ( + load->pSrcData, + load->SrcDataSize, + &img_info ); + + hr = D3DXCreateTextureFromFileInMemoryEx_Original ( + load->pDevice, + load->pSrcData, load->SrcDataSize, + img_info.Width, img_info.Height, img_info.MipLevels, + 0, img_info.Format, + D3DPOOL_DEFAULT, + D3DX_DEFAULT, D3DX_DEFAULT, + 0, + &img_info, nullptr, + &load->pSrc ); + } break; + + default: + tex_log->Log ( L"[ Tex. Mgr ] Unexpected Wait Status: %X (crc32=%x)", + dwResult, + load->checksum ); + wait = false; + break; + } + } + + load->pSrcData = nullptr; + } + + File_Close (&arc_stream.file); + SzArEx_Free (&arc, &thread_alloc); } - if (streamed && size > (16 * 1024)) { + if (streamed && size > (32 * 1024)) { SetThreadPriority ( GetCurrentThread (), THREAD_MODE_BACKGROUND_END ); } @@ -1534,7 +1752,7 @@ TBFix_LoadQueuedTextures (void) spin = 193; if (config.textures.show_loading_text) { - if (InterlockedExchangeAdd (&resampling, 0)) { + if (resample_pool->working ()) { mod_text += spin; char szFormatted [64]; @@ -1555,7 +1773,7 @@ TBFix_LoadQueuedTextures (void) mod_text += "\n\n"; } - if (InterlockedExchangeAdd (&streaming, 0)) { + if (stream_pool.working ()) { mod_text += spin; char szFormatted [64]; @@ -1580,12 +1798,72 @@ TBFix_LoadQueuedTextures (void) int loads = 0; - std::vector finished; + std::vector finished_resamples; + std::vector finished_streams = stream_pool.getFinished (); if (resample_pool != nullptr) - finished = resample_pool->getFinished (); + finished_resamples = resample_pool->getFinished (); - for (auto it = finished.begin (); it != finished.end (); /*it++*/) { + for (auto it = finished_resamples.begin (); it != finished_resamples.end (); /*it++*/) { + tbf_tex_load_s* load = + *it; + + QueryPerformanceCounter (&load->end); + + if (true) { + tex_log->Log ( L"[%s] Finished %s texture %08x (%5.2f MiB in %9.4f ms)", + (load->type == tbf_tex_load_s::Stream) ? L"Inject Tex" : + (load->type == tbf_tex_load_s::Immediate) ? L"Inject Tex" : + L" Resample ", + (load->type == tbf_tex_load_s::Stream) ? L"streaming" : + (load->type == tbf_tex_load_s::Immediate) ? L"loading" : + L"filtering", + load->checksum, + (double)load->SrcDataSize / (1024.0f * 1024.0f), + 1000.0f * (double)(load->end.QuadPart - load->start.QuadPart) / + (double)load->freq.QuadPart ); + } + + tbf::RenderFix::Texture* pTex = + tbf::RenderFix::tex_mgr.getTexture (load->checksum); + + if (pTex != nullptr) { + pTex->load_time = (float)(1000.0 * (double)(load->end.QuadPart - load->start.QuadPart) / + (double)load->freq.QuadPart); + } + + ISKTextureD3D9* pSKTex = + (ISKTextureD3D9 *)load->pDest; + + if (pSKTex != nullptr) { + if (pSKTex->refs == 0 && load->pSrc != nullptr) { + tex_log->Log (L"[ Tex. Mgr ] >> Original texture no longer referenced, discarding new one!"); + load->pSrc->Release (); + } else { + QueryPerformanceCounter (&pSKTex->last_used); + + pSKTex->pTexOverride = load->pSrc; + pSKTex->override_size = load->SrcDataSize; + + tbf::RenderFix::tex_mgr.addInjected (load->SrcDataSize); + } + + finished_streaming (load->checksum); + + tbf::RenderFix::tex_mgr.updateOSD (); + + ++loads; + + // Remove the temporary reference + load->pDest->Release (); + } + + ++it; + + delete load; + } + + for (auto it = finished_streams.begin (); it != finished_streams.end (); /*it++*/) { tbf_tex_load_s* load = *it; @@ -1699,7 +1977,8 @@ TBFix_ReloadPadButtons (void) L"TBFix_Res\\Gamepads\\%s\\Buttons.dds", config.input.gamepad.texture_set.c_str () ); - if (GetFileAttributesW (wszFile) != INVALID_FILE_ATTRIBUTES) { + if (GetFileAttributesW (wszFile) != INVALID_FILE_ATTRIBUTES) + { tex_log->LogEx (true, L"[Inject Tex] Injecting custom gamepad buttons... "); pD3DTex->GetDevice (&pDevice); @@ -1795,7 +2074,7 @@ D3DXCreateTextureFromFileInMemoryEx_Detour ( // Injection would recurse slightly and cause impossible to diagnose reference counting problems // with texture caching if we did not check for this! if (inject_thread) { - return D3DXCreateTextureFromFileInMemoryEx ( + return D3DXCreateTextureFromFileInMemoryEx_Original ( pDevice, pSrcData, SrcDataSize, Width, Height, MipLevels, @@ -1935,7 +2214,7 @@ D3DXCreateTextureFromFileInMemoryEx_Detour ( //tex_log->Log (L"D3DXCreateTextureFromFileInMemoryEx (... MipLevels=%lu ...)", MipLevels); hr = - D3DXCreateTextureFromFileInMemoryEx ( pDevice, + D3DXCreateTextureFromFileInMemoryEx_Original ( pDevice, pSrcData, SrcDataSize, Width, Height, will_replace ? 1 : MipLevels, Usage, Format, Pool, @@ -2009,7 +2288,27 @@ D3DXCreateTextureFromFileInMemoryEx_Detour ( L"TBFix_Res\\Gamepads\\%s\\Buttons.dds", config.input.gamepad.texture_set.c_str () ); - if (GetFileAttributesW (wszFile) != INVALID_FILE_ATTRIBUTES) { + if (GetFileAttributesW (wszFile) != INVALID_FILE_ATTRIBUTES) + { + WIN32_FILE_ATTRIBUTE_DATA + file_attrib_data = { 0 }; + + GetFileAttributesEx ( wszFile, + GetFileExInfoStandard, + &file_attrib_data ); + + tbf_tex_record_s rec; + rec.size = ULARGE_INTEGER { file_attrib_data.nFileSizeLow, + file_attrib_data.nFileSizeHigh }.QuadPart; + rec.archive = -1; + rec.method = Blocking; + + if (! injectable_textures.count (checksum)) + injectable_textures.insert (std::make_pair (checksum, rec)); + else { + injectable_textures [checksum] = rec; + } + tex_log->LogEx (true, L"[Inject Tex] Injecting custom gamepad buttons... "); load_op = new tbf_tex_load_s; @@ -2067,7 +2366,8 @@ D3DXCreateTextureFromFileInMemoryEx_Detour ( textures_in_flight.insert ( std::make_pair ( load_op->checksum, load_op ) ); - resample_pool->postJob (load_op); + stream_pool.postJob (load_op); + //resample_pool->postJob (load_op); } LeaveCriticalSection (&cs_tex_stream); @@ -2121,7 +2421,6 @@ D3DXCreateTextureFromFileInMemoryEx_Detour ( } #endif -// Temporarily disabled while mipmap-related issues are debugged... else if (resample) { load_op = new tbf_tex_load_s; @@ -2382,6 +2681,17 @@ tbf::RenderFix::TextureManager::Init (void) tex_log = TBF_CreateLog (L"logs/textures.log"); + CrcGenerateTable (); + + CFileInStream arc_stream; + CLookToRead look_stream; + + FileInStream_CreateVTable (&arc_stream); + LookToRead_CreateVTable (&look_stream, False); + + look_stream.realStream = &arc_stream.s; + LookToRead_Init (&look_stream); + d3dx9_43_dll = LoadLibrary (L"D3DX9_43.DLL"); // @@ -2503,45 +2813,73 @@ tbf::RenderFix::TextureManager::Init (void) FindClose (hFind); } -#if 0 hFind = FindFirstFileW (TBFIX_TEXTURE_DIR L"\\inject\\*.*", &fd); - if (hFind != INVALID_HANDLE_VALUE) { + if (hFind != INVALID_HANDLE_VALUE) + { int archive = 0; - do { - if (fd.dwFileAttributes != INVALID_FILE_ATTRIBUTES) { + do + { + if (fd.dwFileAttributes != INVALID_FILE_ATTRIBUTES) + { wchar_t* wszArchiveNameLwr = _wcslwr (_wcsdup (fd.cFileName)); - if ( wcsstr (wszArchiveNameLwr, L".zip") || - wcsstr (wszArchiveNameLwr, L".7z") ) { - struct archive *a; - struct archive_entry *entry; - int r, tex_count = 0; + if ( wcsstr (wszArchiveNameLwr, L".7z") ) + { + int tex_count = 0; + + CSzArEx arc; + ISzAlloc thread_alloc; + ISzAlloc thread_tmp_alloc; - a = archive_read_new (); + thread_alloc.Alloc = SzAlloc; + thread_alloc.Free = SzFree; - archive_read_support_filter_all (a); - archive_read_support_format_all (a); + thread_tmp_alloc.Alloc = SzAllocTemp; + thread_tmp_alloc.Free = SzFreeTemp; wchar_t wszQualifiedArchiveName [MAX_PATH]; _swprintf ( wszQualifiedArchiveName, L"%s\\inject\\%s", - TSFIX_TEXTURE_DIR, + TBFIX_TEXTURE_DIR, fd.cFileName ); - r = archive_read_open_filename_w (a, wszQualifiedArchiveName, 10240); + if (InFile_OpenW (&arc_stream.file, wszQualifiedArchiveName)) + { + tex_log->Log ( L"[Inject Tex] ** Cannot open archive file: %s", + wszQualifiedArchiveName ); + continue; + } - if (r == ARCHIVE_OK) { - int fileno = 0; + SzArEx_Init (&arc); + + if ( SzArEx_Open ( &arc, + &look_stream.s, + &thread_alloc, + &thread_tmp_alloc ) == SZ_OK ) + { + uint32_t i; + + wchar_t wszEntry [MAX_PATH]; + + for (i = 0; i < arc.NumFiles; i++) + { + if (SzArEx_IsDir (&arc, i)) + continue; + + SzArEx_GetFileNameUtf16 (&arc, i, (UInt16 *)wszEntry); + + // Truncate to 32-bits --> there's no way in hell a texture will ever be >= 2 GiB + size_t fileSize = SzArEx_GetFileSize (&arc, i); - while (archive_read_next_header (a, &entry) == ARCHIVE_OK) { wchar_t* wszFullName = - _wcslwr (_wcsdup (archive_entry_pathname_w (entry))); + _wcslwr (_wcsdup (wszEntry)); - if ( wcsstr ( wszFullName, TSFIX_TEXTURE_EXT) ) { - tsf_load_method_t method = DontCare; + if ( wcsstr ( wszFullName, TBFIX_TEXTURE_EXT) ) + { + tbf_load_method_t method = DontCare; uint32_t checksum; wchar_t* wszUnqualifiedEntry = @@ -2555,13 +2893,12 @@ tbf::RenderFix::TextureManager::Init (void) if (*wszUnqualifiedEntry == L'/') ++wszUnqualifiedEntry; - swscanf (wszUnqualifiedEntry, L"%x" TSFIX_TEXTURE_EXT, &checksum); + swscanf (wszUnqualifiedEntry, L"%x" TBFIX_TEXTURE_EXT, &checksum); // Already got this texture... - if (injectable_textures.count (checksum)) { + if ( injectable_textures.count (checksum) || + inject_blacklist.count (checksum) ) { free (wszFullName); - archive_read_data_skip (a); - ++fileno; continue; } @@ -2570,10 +2907,10 @@ tbf::RenderFix::TextureManager::Init (void) else if (wcsstr (wszFullName, L"blocking")) method = Blocking; - tsf_tex_record_s rec; - rec.size = (uint32_t)archive_entry_size (entry); + tbf_tex_record_s rec; + rec.size = (uint32_t)fileSize; rec.archive = archive; - rec.fileno = fileno; + rec.fileno = i; rec.method = method; injectable_textures.insert (std::make_pair (checksum, rec)); @@ -2585,28 +2922,24 @@ tbf::RenderFix::TextureManager::Init (void) } free (wszFullName); - - archive_read_data_skip (a); - ++fileno; } - if (tex_count == 0) { - } else { + if (tex_count > 0) { ++archive; archives.push_back (wszQualifiedArchiveName); } - - r = archive_read_free (a); } - free (wszArchiveNameLwr); + SzArEx_Free (&arc, &thread_alloc); + File_Close (&arc_stream.file); } + + free (wszArchiveNameLwr); } } while (FindNextFileW (hFind, &fd) != 0); FindClose (hFind); } -#endif tex_log->LogEx ( false, L" %lu files (%3.1f MiB)\n", files, (double)liSize.QuadPart / (1024.0 * 1024.0) ); @@ -2707,7 +3040,7 @@ tbf::RenderFix::TextureManager::Init (void) TBF_CreateDLLHook ( L"D3DX9_43.DLL", "D3DXCreateTextureFromFileInMemoryEx", D3DXCreateTextureFromFileInMemoryEx_Detour, - (LPVOID *)&D3DXCreateTextureFromFileInMemoryEx ); + (LPVOID *)&D3DXCreateTextureFromFileInMemoryEx_Original ); D3DXSaveTextureToFile = (D3DXSaveTextureToFile_pfn) @@ -2766,7 +3099,10 @@ tbf::RenderFix::TextureManager::Init (void) config.textures.worker_threads, nullptr ); - resample_pool = new SK_TextureThreadPool (); + resample_pool = new SK_TextureThreadPool (); + + stream_pool.lrg_tex = new SK_TextureThreadPool (); + stream_pool.sm_tex = new SK_TextureThreadPool (); SK_ICommandProcessor& command = *SK_GetCommandProcessor (); @@ -3174,7 +3510,7 @@ ResampleTexture (tbf_tex_load_s* load) if (img_info.Depth == 1) { hr = - D3DXCreateTextureFromFileInMemoryEx ( + D3DXCreateTextureFromFileInMemoryEx_Original ( load->pDevice, load->pSrcData, load->SrcDataSize, img_info.Width, img_info.Height, 0, @@ -3230,10 +3566,10 @@ SK_TextureWorkerThread::ThreadProc (LPVOID user) // Ghetto sync. barrier, since Windows 7 does not support them... while ( InterlockedCompareExchange ( &num_threads_init, - config.textures.worker_threads, - config.textures.worker_threads - ) < (ULONG)config.textures.worker_threads ) { - Sleep (15); + config.textures.worker_threads * 3, + config.textures.worker_threads * 3 + ) < (ULONG)config.textures.worker_threads * 3 ) { + SwitchToThread (); } SK_TextureWorkerThread* pThread = @@ -3302,7 +3638,10 @@ SK_TextureWorkerThread::ThreadProc (LPVOID user) } - else if (dwWaitStatus == (wait.mem_trim)) { + else if (dwWaitStatus == (wait.mem_trim)) + { + // Yay for magic numbers :P ==> (8 MiB Min Size, 5 Seconds Between Trims) + // const size_t MIN_SIZE = 8192 * 1024; const uint32_t MIN_AGE = 5000UL; diff --git a/tbf.vcxproj b/tbf.vcxproj index 7dc38b2..ed08ce5 100644 --- a/tbf.vcxproj +++ b/tbf.vcxproj @@ -44,7 +44,7 @@ - true + false tbfix C:\Program Files (x86)\Steam\steamapps\common\Tales of Berseria\ @@ -67,12 +67,13 @@ true NotSet false - Disabled + AnySuitable true Speed false true include\ + ProgramDatabase Windows @@ -85,8 +86,9 @@ Default false false - true + false false + DllMain @@ -96,10 +98,10 @@ Full true - true + false NDEBUG;_WINDOWS;_USRDLL;TBF_EXPORTS;%(PreprocessorDefinitions);IMGUI_DISABLE_INCLUDE_IMCONFIG_H MultiThreadedDLL - Disabled + Default false false false @@ -107,7 +109,7 @@ Speed false include\ - EditAndContinue + ProgramDatabase false true @@ -120,12 +122,14 @@ true false false - true + false false Default kernel32.lib;user32.lib;gdi32.lib;winspool.lib;comdlg32.lib;advapi32.lib;shell32.lib;ole32.lib;oleaut32.lib;uuid.lib;odbc32.lib;odbccp32.lib;%(AdditionalDependencies);C:\Program Files (x86)\Steam\steamapps\common\Tales of Berseria\d3d9.lib + + @@ -162,6 +166,47 @@ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + diff --git a/tbf.vcxproj.filters b/tbf.vcxproj.filters index b942292..3ce7ea7 100644 --- a/tbf.vcxproj.filters +++ b/tbf.vcxproj.filters @@ -46,6 +46,129 @@ Source Files\ImGui + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + + + Source Files\lzma + @@ -129,6 +252,9 @@ {9a0b79ba-cdc0-45eb-8818-c78ab4eab12c} + + {ba7bd273-40d1-40c0-a04f-e18ec8166b92} + diff --git a/version.ini b/version.ini index 2b2ad1d788db221fcf4de0f4d8600777f9ba0723..d7afd04c886006c5d7e1dd074485e9d423892ac2 100644 GIT binary patch delta 118 zcmX?Q_sw=g9W$#LgC2vy#6s=K>zMTz%_iSvR#pvRNM)#CC}AjNC<3w+78B&0B zFi@lb2=jsBC6n!0>zU0M3??6BHRd*Bh-WZh&|xrQu-N>cbukmrRG`3QNA5