diff options
Diffstat (limited to 'src')
1555 files changed, 305866 insertions, 0 deletions
diff --git a/src/libs/7zip/7zip.pri b/src/libs/7zip/7zip.pri new file mode 100644 index 000000000..1aa92d294 --- /dev/null +++ b/src/libs/7zip/7zip.pri @@ -0,0 +1,4 @@ +7ZIP_BASE=$$PWD + +win32:include($$7ZIP_BASE/win/win.pri) #this is 7zip +unix:include($$7ZIP_BASE/unix/unix.pri) #this is p7zip diff --git a/src/libs/7zip/unix/C/7zBuf.h b/src/libs/7zip/unix/C/7zBuf.h new file mode 100644 index 000000000..e9f2f316d --- /dev/null +++ b/src/libs/7zip/unix/C/7zBuf.h @@ -0,0 +1,39 @@ +/* 7zBuf.h -- Byte Buffer +2009-02-07 : Igor Pavlov : Public domain */ + +#ifndef __7Z_BUF_H +#define __7Z_BUF_H + +#include "Types.h" + +#ifdef __cplusplus +extern "C" { +#endif + +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); + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/src/libs/7zip/unix/C/7zBuf2.c b/src/libs/7zip/unix/C/7zBuf2.c new file mode 100644 index 000000000..8d17e0dcf --- /dev/null +++ b/src/libs/7zip/unix/C/7zBuf2.c @@ -0,0 +1,45 @@ +/* 7zBuf2.c -- Byte Buffer +2008-10-04 : Igor Pavlov : Public domain */ + +#include <string.h> +#include "7zBuf.h" + +void DynBuf_Construct(CDynBuf *p) +{ + p->data = 0; + p->size = 0; + p->pos = 0; +} + +void DynBuf_SeekToBeg(CDynBuf *p) +{ + p->pos = 0; +} + +int DynBuf_Write(CDynBuf *p, const Byte *buf, size_t size, 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; + } + 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/libs/7zip/unix/C/7zCrc.c b/src/libs/7zip/unix/C/7zCrc.c new file mode 100644 index 000000000..404090ef3 --- /dev/null +++ b/src/libs/7zip/unix/C/7zCrc.c @@ -0,0 +1,76 @@ +/* 7zCrc.c -- CRC32 calculation +2009-11-23 : Igor Pavlov : Public domain */ + +#include "7zCrc.h" +#include "CpuArch.h" + +#define kCrcPoly 0xEDB88320 + +#ifdef MY_CPU_LE +#define CRC_NUM_TABLES 8 +#else +#define CRC_NUM_TABLES 1 +#endif + +typedef UInt32 (MY_FAST_CALL *CRC_FUNC)(UInt32 v, const void *data, size_t size, const UInt32 *table); + +static CRC_FUNC g_CrcUpdate; +UInt32 g_CrcTable[256 * CRC_NUM_TABLES]; + +#if CRC_NUM_TABLES == 1 + +#define CRC_UPDATE_BYTE_2(crc, b) (table[((crc) ^ (b)) & 0xFF] ^ ((crc) >> 8)) + +static UInt32 MY_FAST_CALL CrcUpdateT1(UInt32 v, const void *data, size_t size, const UInt32 *table) +{ + const Byte *p = (const Byte *)data; + for (; size > 0; size--, p++) + v = CRC_UPDATE_BYTE_2(v, *p); + return v; +} + +#else + +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 + +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; +} + +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; + } + #if CRC_NUM_TABLES == 1 + g_CrcUpdate = CrcUpdateT1; + #else + for (; i < 256 * CRC_NUM_TABLES; i++) + { + UInt32 r = g_CrcTable[i - 256]; + g_CrcTable[i] = g_CrcTable[r & 0xFF] ^ (r >> 8); + } + g_CrcUpdate = CrcUpdateT4; +/* FIXME + #ifdef MY_CPU_X86_OR_AMD64 + if (!CPU_Is_InOrder()) + g_CrcUpdate = CrcUpdateT8; + #endif +*/ + #endif +} diff --git a/src/libs/7zip/unix/C/7zCrc.h b/src/libs/7zip/unix/C/7zCrc.h new file mode 100644 index 000000000..38e3e5fbc --- /dev/null +++ b/src/libs/7zip/unix/C/7zCrc.h @@ -0,0 +1,25 @@ +/* 7zCrc.h -- CRC32 calculation +2009-11-21 : Igor Pavlov : Public domain */ + +#ifndef __7Z_CRC_H +#define __7Z_CRC_H + +#include "Types.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/src/libs/7zip/unix/C/7zCrcOpt.c b/src/libs/7zip/unix/C/7zCrcOpt.c new file mode 100644 index 000000000..6c766a209 --- /dev/null +++ b/src/libs/7zip/unix/C/7zCrcOpt.c @@ -0,0 +1,34 @@ +/* 7zCrcOpt.c -- CRC32 calculation : optimized version +2009-11-23 : Igor Pavlov : Public domain */ + +#include "CpuArch.h" + +#ifdef MY_CPU_LE + +#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) +{ + return CrcUpdateT4(v, data, size, table); +} + +#endif diff --git a/src/libs/7zip/unix/C/7zCrcT8.c b/src/libs/7zip/unix/C/7zCrcT8.c new file mode 100644 index 000000000..cd80e262b --- /dev/null +++ b/src/libs/7zip/unix/C/7zCrcT8.c @@ -0,0 +1,43 @@ +/* 7zCrcT8.c -- CRC32 calculation with 8 tables +2008-03-19 +Igor Pavlov +Public domain */ + +#include "7zCrc.h" + +#define kCrcPoly 0xEDB88320 +#define CRC_NUM_TABLES 8 + +UInt32 g_CrcTable[256 * CRC_NUM_TABLES]; + +void MY_FAST_CALL CrcGenerateTable() +{ + UInt32 i; + for (i = 0; i < 256; i++) + { + UInt32 r = i; + int j; + for (j = 0; j < 8; j++) + r = (r >> 1) ^ (kCrcPoly & ~((r & 1) - 1)); + g_CrcTable[i] = r; + } + #if CRC_NUM_TABLES > 1 + for (; i < 256 * CRC_NUM_TABLES; i++) + { + UInt32 r = g_CrcTable[i - 256]; + g_CrcTable[i] = g_CrcTable[r & 0xFF] ^ (r >> 8); + } + #endif +} + +UInt32 MY_FAST_CALL CrcUpdateT8(UInt32 v, const void *data, size_t size, const UInt32 *table); + +UInt32 MY_FAST_CALL CrcUpdate(UInt32 v, const void *data, size_t size) +{ + return CrcUpdateT8(v, data, size, g_CrcTable); +} + +UInt32 MY_FAST_CALL CrcCalc(const void *data, size_t size) +{ + return CrcUpdateT8(CRC_INIT_VAL, data, size, g_CrcTable) ^ 0xFFFFFFFF; +} diff --git a/src/libs/7zip/unix/C/7zStream.c b/src/libs/7zip/unix/C/7zStream.c new file mode 100644 index 000000000..0ebb7b5f9 --- /dev/null +++ b/src/libs/7zip/unix/C/7zStream.c @@ -0,0 +1,169 @@ +/* 7zStream.c -- 7z Stream functions +2010-03-11 : Igor Pavlov : Public domain */ + +#include <string.h> + +#include "Types.h" + +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/libs/7zip/unix/C/7zVersion.h b/src/libs/7zip/unix/C/7zVersion.h new file mode 100644 index 000000000..9d99c5dff --- /dev/null +++ b/src/libs/7zip/unix/C/7zVersion.h @@ -0,0 +1,7 @@ +#define MY_VER_MAJOR 9 +#define MY_VER_MINOR 20 +#define MY_VER_BUILD 0 +#define MY_VERSION "9.20" +#define MY_DATE "2010-11-18" +#define MY_COPYRIGHT ": Igor Pavlov : Public domain" +#define MY_VERSION_COPYRIGHT_DATE MY_VERSION " " MY_COPYRIGHT " : " MY_DATE diff --git a/src/libs/7zip/unix/C/Aes.c b/src/libs/7zip/unix/C/Aes.c new file mode 100644 index 000000000..1bb658793 --- /dev/null +++ b/src/libs/7zip/unix/C/Aes.c @@ -0,0 +1,284 @@ +/* Aes.c -- AES encryption / decryption +2009-11-23 : Igor Pavlov : Public domain */ + +#include "Aes.h" +#include "CpuArch.h" + +static UInt32 T[256 * 4]; +static 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 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; +/* FIXME + #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] +/* such order (2031) in HT16 is for VC6/K8 speed optimization) */ +#define HT16(m, s, p) \ + HT4(m, 2, s, p); \ + HT4(m, 0, s, p); \ + HT4(m, 3, s, p); \ + HT4(m, 1, 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]; +/* such order (0231) in HD16 is for VC6/K8 speed optimization) */ +#define HD16(m, s, p) \ + HD4(m, 0, s, p); \ + HD4(m, 2, s, p); \ + HD4(m, 3, s, p); \ + HD4(m, 1, 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[ Sbox[gb0(r)]] ^ + D[0x100 + Sbox[gb1(r)]] ^ + D[0x200 + Sbox[gb2(r)]] ^ + D[0x300 + Sbox[gb3(r)]]; + } +} + +/* Aes_Encode and Aes_Decode functions work with little-endian words. + src and dest are pointers to 4 UInt32 words. + arc 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/libs/7zip/unix/C/Aes.h b/src/libs/7zip/unix/C/Aes.h new file mode 100644 index 000000000..c9b0677c8 --- /dev/null +++ b/src/libs/7zip/unix/C/Aes.h @@ -0,0 +1,38 @@ +/* Aes.h -- AES encryption / decryption +2009-11-23 : Igor Pavlov : Public domain */ + +#ifndef __AES_H +#define __AES_H + +#include "Types.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/src/libs/7zip/unix/C/Alloc.c b/src/libs/7zip/unix/C/Alloc.c new file mode 100644 index 000000000..be02a9299 --- /dev/null +++ b/src/libs/7zip/unix/C/Alloc.c @@ -0,0 +1,280 @@ +/* Alloc.c -- Memory allocation functions +2008-09-24 +Igor Pavlov +Public domain */ + +#ifdef _WIN32 +#include <windows.h> +#endif +#include <stdio.h> +#include <stdlib.h> + +#ifdef _7ZIP_LARGE_PAGES +#ifdef __linux__ +#ifndef _7ZIP_ST +#include <pthread.h> +#endif +#include <errno.h> +#include <string.h> +#include <unistd.h> +#include <sys/mman.h> +#include <mntent.h> +#endif +#endif + +#include "Alloc.h" + +/* #define _SZ_ALLOC_DEBUG */ + +/* use _SZ_ALLOC_DEBUG to debug alloc/free operations */ +#ifdef _SZ_ALLOC_DEBUG +#include <stdio.h> +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); +} + +#ifndef _WIN32 + +#ifdef _7ZIP_LARGE_PAGES + +#ifdef __linux__ +#define _7ZIP_MAX_HUGE_ALLOCS 64 +static void *g_HugePageAddr[_7ZIP_MAX_HUGE_ALLOCS] = { NULL }; +static size_t g_HugePageLen[_7ZIP_MAX_HUGE_ALLOCS]; +static char *g_HugetlbPath; +#endif + +#endif + +static void *VirtualAlloc(size_t size, int memLargePages) +{ + #ifdef _7ZIP_LARGE_PAGES + if (memLargePages) + { + #ifdef __linux__ + /* huge pages support for Linux; added by Joachim Henke */ + #ifndef _7ZIP_ST + static pthread_mutex_t mutex = PTHREAD_MUTEX_INITIALIZER; + #endif + int i; + + void * address = NULL; + #ifndef _7ZIP_ST + pthread_mutex_lock(&mutex); + #endif + for (i = 0; i < _7ZIP_MAX_HUGE_ALLOCS; ++i) + { + if (g_HugePageAddr[i] == NULL) + { + int fd, pathlen = strlen(g_HugetlbPath); + char tempname[pathlen+12]; + + memcpy(tempname, g_HugetlbPath, pathlen); + memcpy(tempname + pathlen, "/7z-XXXXXX", 11); + fd = mkstemp(tempname); + unlink(tempname); + if (fd < 0) + { + fprintf(stderr,"cant't open %s (%s)\n",tempname,strerror(errno)); + break; + } + address = mmap(NULL, size, PROT_READ | PROT_WRITE, MAP_SHARED, fd, 0); + close(fd); + if (address == MAP_FAILED) + { + address = NULL; + break; + } + g_HugePageLen[i] = size; + g_HugePageAddr[i] = address; +// fprintf(stderr,"HUGE[%d]=%ld %p\n",i,(long)size,address); + break; + } + } + #ifndef _7ZIP_ST + pthread_mutex_unlock(&mutex); + #endif + return address; + #endif + } + #endif + return malloc(size); +} + +static int VirtualFree(void *address) +{ + #ifdef _7ZIP_LARGE_PAGES + #ifdef __linux__ + int i; + + for (i = 0; i < _7ZIP_MAX_HUGE_ALLOCS; ++i) + { + if (g_HugePageAddr[i] == address) + { + munmap(address, g_HugePageLen[i]); + g_HugePageAddr[i] = NULL; + return 1; + } + } + #endif + #endif + free(address); + return 1; +} + +#endif + +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(size, 0); +} + +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); +} + +#ifdef _7ZIP_LARGE_PAGES +size_t g_LargePageSize = 0; +#ifdef _WIN32 +typedef SIZE_T (WINAPI *GetLargePageMinimumP)(); +#elif defined(__linux__) +size_t largePageMinimum() +{ + size_t size; + + g_HugetlbPath = getenv("HUGETLB_PATH"); + + if (g_HugetlbPath == NULL) + { + // not defined => try to find out the directory + static char dir_hugetlbfs[1024]; + const char * filename = "/etc/mtab"; // mounted filesystems + FILE *fp; + struct mntent * info; + + dir_hugetlbfs[0]=0; + + fp = setmntent(filename,"r"); + if (fp) + { + info = getmntent(fp); + while(info) + { +/* + printf("%s:\n",info->mnt_fsname); + printf(" dir='%s'\n",info->mnt_dir); + printf(" type='%s'\n",info->mnt_type); +*/ + + if (strcmp(info->mnt_type,"hugetlbfs") == 0) + { + strcpy(dir_hugetlbfs,info->mnt_dir); + break; + } + + info = getmntent(fp); + } + endmntent(fp); + } + + if (dir_hugetlbfs[0]) + { + g_HugetlbPath = dir_hugetlbfs; + // fprintf(stderr," Found hugetlbfs = '%s'\n",g_HugetlbPath); + } + } + if (g_HugetlbPath == NULL || (size = pathconf(g_HugetlbPath, _PC_REC_MIN_XFER_SIZE)) <= getpagesize()) + return 0; + return size; +} +#else +#define largePageMinimum() 0 +#endif +#endif + +void SetLargePageSize() +{ + #ifdef _7ZIP_LARGE_PAGES + size_t size; + #ifdef _WIN32 + GetLargePageMinimumP largePageMinimum = (GetLargePageMinimumP) + GetProcAddress(GetModuleHandle(TEXT("kernel32.dll")), "GetLargePageMinimum"); + if (largePageMinimum == 0) + return; + #endif + size = largePageMinimum(); + if (size == 0 || (size & (size - 1)) != 0) + return; + g_LargePageSize = size; + // fprintf(stderr,"SetLargePageSize : %ld\n",(long)g_LargePageSize); + #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( (size + g_LargePageSize - 1) & (~(g_LargePageSize - 1)), 1); + if (res != 0) + return res; + } + #endif + return VirtualAlloc(size, 0); +} + +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); +} diff --git a/src/libs/7zip/unix/C/Alloc.h b/src/libs/7zip/unix/C/Alloc.h new file mode 100644 index 000000000..7ef372e3a --- /dev/null +++ b/src/libs/7zip/unix/C/Alloc.h @@ -0,0 +1,27 @@ +/* Alloc.h -- Memory allocation functions +2009-02-07 : Igor Pavlov : Public domain */ + +#ifndef __COMMON_ALLOC_H +#define __COMMON_ALLOC_H + +#include <stddef.h> + +#ifdef __cplusplus +extern "C" { +#endif + +void *MyAlloc(size_t size); +void MyFree(void *address); + +void SetLargePageSize(); + +void *MidAlloc(size_t size); +void MidFree(void *address); +void *BigAlloc(size_t size); +void BigFree(void *address); + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/src/libs/7zip/unix/C/Bra.c b/src/libs/7zip/unix/C/Bra.c new file mode 100644 index 000000000..2e47b1413 --- /dev/null +++ b/src/libs/7zip/unix/C/Bra.c @@ -0,0 +1,133 @@ +/* Bra.c -- Converters for RISC code +2010-04-16 : Igor Pavlov : Public domain */ + +#include "Bra.h" + +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/libs/7zip/unix/C/Bra.h b/src/libs/7zip/unix/C/Bra.h new file mode 100644 index 000000000..5748c1c05 --- /dev/null +++ b/src/libs/7zip/unix/C/Bra.h @@ -0,0 +1,68 @@ +/* Bra.h -- Branch converters for executables +2009-02-07 : Igor Pavlov : Public domain */ + +#ifndef __BRA_H +#define __BRA_H + +#include "Types.h" + +#ifdef __cplusplus +extern "C" { +#endif + +/* +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); + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/src/libs/7zip/unix/C/Bra86.c b/src/libs/7zip/unix/C/Bra86.c new file mode 100644 index 000000000..1ee0e709b --- /dev/null +++ b/src/libs/7zip/unix/C/Bra86.c @@ -0,0 +1,85 @@ +/* Bra86.c -- Converter for x86 code (BCJ) +2008-10-04 : Igor Pavlov : Public domain */ + +#include "Bra.h" + +#define Test86MSByte(b) ((b) == 0 || (b) == 0xFF) + +const Byte kMaskToAllowedStatus[8] = {1, 1, 1, 0, 1, 0, 0, 0}; +const Byte kMaskToBitNumber[8] = {0, 1, 2, 2, 3, 3, 3, 3}; + +SizeT x86_Convert(Byte *data, SizeT size, UInt32 ip, UInt32 *state, int encoding) +{ + SizeT bufferPos = 0, prevPosT; + UInt32 prevMask = *state & 0x7; + if (size < 5) + return 0; + ip += 5; + prevPosT = (SizeT)0 - 1; + + for (;;) + { + Byte *p = data + bufferPos; + Byte *limit = data + size - 4; + for (; p < limit; p++) + if ((*p & 0xFE) == 0xE8) + break; + bufferPos = (SizeT)(p - data); + if (p >= limit) + break; + prevPosT = bufferPos - prevPosT; + if (prevPosT > 3) + prevMask = 0; + else + { + prevMask = (prevMask << ((int)prevPosT - 1)) & 0x7; + if (prevMask != 0) + { + Byte b = p[4 - kMaskToBitNumber[prevMask]]; + if (!kMaskToAllowedStatus[prevMask] || Test86MSByte(b)) + { + prevPosT = bufferPos; + prevMask = ((prevMask << 1) & 0x7) | 1; + bufferPos++; + continue; + } + } + } + prevPosT = bufferPos; + + if (Test86MSByte(p[4])) + { + UInt32 src = ((UInt32)p[4] << 24) | ((UInt32)p[3] << 16) | ((UInt32)p[2] << 8) | ((UInt32)p[1]); + UInt32 dest; + for (;;) + { + Byte b; + int index; + if (encoding) + dest = (ip + (UInt32)bufferPos) + src; + else + dest = src - (ip + (UInt32)bufferPos); + if (prevMask == 0) + break; + index = kMaskToBitNumber[prevMask] * 8; + b = (Byte)(dest >> (24 - index)); + if (!Test86MSByte(b)) + break; + src = dest ^ ((1 << (32 - index)) - 1); + } + p[4] = (Byte)(~(((dest >> 24) & 1) - 1)); + p[3] = (Byte)(dest >> 16); + p[2] = (Byte)(dest >> 8); + p[1] = (Byte)dest; + bufferPos += 5; + } + else + { + prevMask = ((prevMask << 1) & 0x7) | 1; + bufferPos++; + } + } + prevPosT = bufferPos - prevPosT; + *state = ((prevPosT > 3) ? 0 : ((prevMask << ((int)prevPosT - 1)) & 0x7)); + return bufferPos; +} diff --git a/src/libs/7zip/unix/C/BraIA64.c b/src/libs/7zip/unix/C/BraIA64.c new file mode 100644 index 000000000..0b4ee85bc --- /dev/null +++ b/src/libs/7zip/unix/C/BraIA64.c @@ -0,0 +1,67 @@ +/* BraIA64.c -- Converter for IA-64 code +2008-10-04 : Igor Pavlov : Public domain */ + +#include "Bra.h" + +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/libs/7zip/unix/C/BwtSort.c b/src/libs/7zip/unix/C/BwtSort.c new file mode 100644 index 000000000..207305072 --- /dev/null +++ b/src/libs/7zip/unix/C/BwtSort.c @@ -0,0 +1,516 @@ +/* BwtSort.c -- BWT block sorting +2008-08-17 +Igor Pavlov +Public domain */ + +#include "BwtSort.h" +#include "Sort.h" + +/* #define BLOCK_SORT_USE_HEAP_SORT */ + +#define NO_INLINE MY_FAST_CALL + +/* Don't change it !!! */ +#define kNumHashBytes 2 +#define kNumHashValues (1 << (kNumHashBytes * 8)) + +/* kNumRefBitsMax must be < (kNumHashBytes * 8) = 16 */ +#define kNumRefBitsMax 12 + +#define BS_TEMP_SIZE kNumHashValues + +#ifdef BLOCK_SORT_EXTERNAL_FLAGS + +/* 32 Flags in UInt32 word */ +#define kNumFlagsBits 5 +#define kNumFlagsInWord (1 << kNumFlagsBits) +#define kFlagsMask (kNumFlagsInWord - 1) +#define kAllFlags 0xFFFFFFFF + +#else + +#define kNumBitsMax 20 +#define kIndexMask ((1 << kNumBitsMax) - 1) +#define kNumExtraBits (32 - kNumBitsMax) +#define kNumExtra0Bits (kNumExtraBits - 2) +#define kNumExtra0Mask ((1 << kNumExtra0Bits) - 1) + +#define SetFinishedGroupSize(p, size) \ + { *(p) |= ((((size) - 1) & kNumExtra0Mask) << kNumBitsMax); \ + if ((size) > (1 << kNumExtra0Bits)) { \ + *(p) |= 0x40000000; *((p) + 1) |= ((((size) - 1)>> kNumExtra0Bits) << kNumBitsMax); } } \ + +static void SetGroupSize(UInt32 *p, UInt32 size) +{ + if (--size == 0) + return; + *p |= 0x80000000 | ((size & kNumExtra0Mask) << kNumBitsMax); + if (size >= (1 << kNumExtra0Bits)) + { + *p |= 0x40000000; + p[1] |= ((size >> kNumExtra0Bits) << kNumBitsMax); + } +} + +#endif + +/* +SortGroup - is recursive Range-Sort function with HeapSort optimization for small blocks + "range" is not real range. It's only for optimization. +returns: 1 - if there are groups, 0 - no more groups +*/ + +UInt32 NO_INLINE SortGroup(UInt32 BlockSize, UInt32 NumSortedBytes, UInt32 groupOffset, UInt32 groupSize, int NumRefBits, UInt32 *Indices + #ifndef BLOCK_SORT_USE_HEAP_SORT + , UInt32 left, UInt32 range + #endif + ) +{ + UInt32 *ind2 = Indices + groupOffset; + UInt32 *Groups; + if (groupSize <= 1) + { + /* + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + SetFinishedGroupSize(ind2, 1); + #endif + */ + return 0; + } + Groups = Indices + BlockSize + BS_TEMP_SIZE; + if (groupSize <= ((UInt32)1 << NumRefBits) + #ifndef BLOCK_SORT_USE_HEAP_SORT + && groupSize <= range + #endif + ) + { + UInt32 *temp = Indices + BlockSize; + UInt32 j; + UInt32 mask, thereAreGroups, group, cg; + { + UInt32 gPrev; + UInt32 gRes = 0; + { + UInt32 sp = ind2[0] + NumSortedBytes; + if (sp >= BlockSize) sp -= BlockSize; + gPrev = Groups[sp]; + temp[0] = (gPrev << NumRefBits); + } + + for (j = 1; j < groupSize; j++) + { + UInt32 sp = ind2[j] + NumSortedBytes; + UInt32 g; + if (sp >= BlockSize) sp -= BlockSize; + g = Groups[sp]; + temp[j] = (g << NumRefBits) | j; + gRes |= (gPrev ^ g); + } + if (gRes == 0) + { + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + SetGroupSize(ind2, groupSize); + #endif + return 1; + } + } + + HeapSort(temp, groupSize); + mask = ((1 << NumRefBits) - 1); + thereAreGroups = 0; + + group = groupOffset; + cg = (temp[0] >> NumRefBits); + temp[0] = ind2[temp[0] & mask]; + + { + #ifdef BLOCK_SORT_EXTERNAL_FLAGS + UInt32 *Flags = Groups + BlockSize; + #else + UInt32 prevGroupStart = 0; + #endif + + for (j = 1; j < groupSize; j++) + { + UInt32 val = temp[j]; + UInt32 cgCur = (val >> NumRefBits); + + if (cgCur != cg) + { + cg = cgCur; + group = groupOffset + j; + + #ifdef BLOCK_SORT_EXTERNAL_FLAGS + { + UInt32 t = group - 1; + Flags[t >> kNumFlagsBits] &= ~(1 << (t & kFlagsMask)); + } + #else + SetGroupSize(temp + prevGroupStart, j - prevGroupStart); + prevGroupStart = j; + #endif + } + else + thereAreGroups = 1; + { + UInt32 ind = ind2[val & mask]; + temp[j] = ind; + Groups[ind] = group; + } + } + + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + SetGroupSize(temp + prevGroupStart, j - prevGroupStart); + #endif + } + + for (j = 0; j < groupSize; j++) + ind2[j] = temp[j]; + return thereAreGroups; + } + + /* Check that all strings are in one group (cannot sort) */ + { + UInt32 group, j; + UInt32 sp = ind2[0] + NumSortedBytes; if (sp >= BlockSize) sp -= BlockSize; + group = Groups[sp]; + for (j = 1; j < groupSize; j++) + { + sp = ind2[j] + NumSortedBytes; if (sp >= BlockSize) sp -= BlockSize; + if (Groups[sp] != group) + break; + } + if (j == groupSize) + { + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + SetGroupSize(ind2, groupSize); + #endif + return 1; + } + } + + #ifndef BLOCK_SORT_USE_HEAP_SORT + { + /* ---------- Range Sort ---------- */ + UInt32 i; + UInt32 mid; + for (;;) + { + UInt32 j; + if (range <= 1) + { + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + SetGroupSize(ind2, groupSize); + #endif + return 1; + } + mid = left + ((range + 1) >> 1); + j = groupSize; + i = 0; + do + { + UInt32 sp = ind2[i] + NumSortedBytes; if (sp >= BlockSize) sp -= BlockSize; + if (Groups[sp] >= mid) + { + for (j--; j > i; j--) + { + sp = ind2[j] + NumSortedBytes; if (sp >= BlockSize) sp -= BlockSize; + if (Groups[sp] < mid) + { + UInt32 temp = ind2[i]; ind2[i] = ind2[j]; ind2[j] = temp; + break; + } + } + if (i >= j) + break; + } + } + while (++i < j); + if (i == 0) + { + range = range - (mid - left); + left = mid; + } + else if (i == groupSize) + range = (mid - left); + else + break; + } + + #ifdef BLOCK_SORT_EXTERNAL_FLAGS + { + UInt32 t = (groupOffset + i - 1); + UInt32 *Flags = Groups + BlockSize; + Flags[t >> kNumFlagsBits] &= ~(1 << (t & kFlagsMask)); + } + #endif + + { + UInt32 j; + for (j = i; j < groupSize; j++) + Groups[ind2[j]] = groupOffset + i; + } + + { + UInt32 res = SortGroup(BlockSize, NumSortedBytes, groupOffset, i, NumRefBits, Indices, left, mid - left); + return res | SortGroup(BlockSize, NumSortedBytes, groupOffset + i, groupSize - i, NumRefBits, Indices, mid, range - (mid - left)); + } + + } + + #else + + /* ---------- Heap Sort ---------- */ + + { + UInt32 j; + for (j = 0; j < groupSize; j++) + { + UInt32 sp = ind2[j] + NumSortedBytes; if (sp >= BlockSize) sp -= BlockSize; + ind2[j] = sp; + } + + HeapSortRef(ind2, Groups, groupSize); + + /* Write Flags */ + { + UInt32 sp = ind2[0]; + UInt32 group = Groups[sp]; + + #ifdef BLOCK_SORT_EXTERNAL_FLAGS + UInt32 *Flags = Groups + BlockSize; + #else + UInt32 prevGroupStart = 0; + #endif + + for (j = 1; j < groupSize; j++) + { + sp = ind2[j]; + if (Groups[sp] != group) + { + group = Groups[sp]; + #ifdef BLOCK_SORT_EXTERNAL_FLAGS + { + UInt32 t = groupOffset + j - 1; + Flags[t >> kNumFlagsBits] &= ~(1 << (t & kFlagsMask)); + } + #else + SetGroupSize(ind2 + prevGroupStart, j - prevGroupStart); + prevGroupStart = j; + #endif + } + } + + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + SetGroupSize(ind2 + prevGroupStart, j - prevGroupStart); + #endif + } + { + /* Write new Groups values and Check that there are groups */ + UInt32 thereAreGroups = 0; + for (j = 0; j < groupSize; j++) + { + UInt32 group = groupOffset + j; + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + UInt32 subGroupSize = ((ind2[j] & ~0xC0000000) >> kNumBitsMax); + if ((ind2[j] & 0x40000000) != 0) + subGroupSize += ((ind2[j + 1] >> kNumBitsMax) << kNumExtra0Bits); + subGroupSize++; + for (;;) + { + UInt32 original = ind2[j]; + UInt32 sp = original & kIndexMask; + if (sp < NumSortedBytes) sp += BlockSize; sp -= NumSortedBytes; + ind2[j] = sp | (original & ~kIndexMask); + Groups[sp] = group; + if (--subGroupSize == 0) + break; + j++; + thereAreGroups = 1; + } + #else + UInt32 *Flags = Groups + BlockSize; + for (;;) + { + UInt32 sp = ind2[j]; if (sp < NumSortedBytes) sp += BlockSize; sp -= NumSortedBytes; + ind2[j] = sp; + Groups[sp] = group; + if ((Flags[(groupOffset + j) >> kNumFlagsBits] & (1 << ((groupOffset + j) & kFlagsMask))) == 0) + break; + j++; + thereAreGroups = 1; + } + #endif + } + return thereAreGroups; + } + } + #endif +} + +/* conditions: blockSize > 0 */ +UInt32 BlockSort(UInt32 *Indices, const Byte *data, UInt32 blockSize) +{ + UInt32 *counters = Indices + blockSize; + UInt32 i; + UInt32 *Groups; + #ifdef BLOCK_SORT_EXTERNAL_FLAGS + UInt32 *Flags; + #endif + + /* Radix-Sort for 2 bytes */ + for (i = 0; i < kNumHashValues; i++) + counters[i] = 0; + for (i = 0; i < blockSize - 1; i++) + counters[((UInt32)data[i] << 8) | data[i + 1]]++; + counters[((UInt32)data[i] << 8) | data[0]]++; + + Groups = counters + BS_TEMP_SIZE; + #ifdef BLOCK_SORT_EXTERNAL_FLAGS + Flags = Groups + blockSize; + { + UInt32 numWords = (blockSize + kFlagsMask) >> kNumFlagsBits; + for (i = 0; i < numWords; i++) + Flags[i] = kAllFlags; + } + #endif + + { + UInt32 sum = 0; + for (i = 0; i < kNumHashValues; i++) + { + UInt32 groupSize = counters[i]; + if (groupSize > 0) + { + #ifdef BLOCK_SORT_EXTERNAL_FLAGS + UInt32 t = sum + groupSize - 1; + Flags[t >> kNumFlagsBits] &= ~(1 << (t & kFlagsMask)); + #endif + sum += groupSize; + } + counters[i] = sum - groupSize; + } + + for (i = 0; i < blockSize - 1; i++) + Groups[i] = counters[((UInt32)data[i] << 8) | data[i + 1]]; + Groups[i] = counters[((UInt32)data[i] << 8) | data[0]]; + + for (i = 0; i < blockSize - 1; i++) + Indices[counters[((UInt32)data[i] << 8) | data[i + 1]]++] = i; + Indices[counters[((UInt32)data[i] << 8) | data[0]]++] = i; + + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + { + UInt32 prev = 0; + for (i = 0; i < kNumHashValues; i++) + { + UInt32 prevGroupSize = counters[i] - prev; + if (prevGroupSize == 0) + continue; + SetGroupSize(Indices + prev, prevGroupSize); + prev = counters[i]; + } + } + #endif + } + + { + int NumRefBits; + UInt32 NumSortedBytes; + for (NumRefBits = 0; ((blockSize - 1) >> NumRefBits) != 0; NumRefBits++); + NumRefBits = 32 - NumRefBits; + if (NumRefBits > kNumRefBitsMax) + NumRefBits = kNumRefBitsMax; + + for (NumSortedBytes = kNumHashBytes; ; NumSortedBytes <<= 1) + { + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + UInt32 finishedGroupSize = 0; + #endif + UInt32 newLimit = 0; + for (i = 0; i < blockSize;) + { + UInt32 groupSize; + #ifdef BLOCK_SORT_EXTERNAL_FLAGS + + if ((Flags[i >> kNumFlagsBits] & (1 << (i & kFlagsMask))) == 0) + { + i++; + continue; + } + for (groupSize = 1; + (Flags[(i + groupSize) >> kNumFlagsBits] & (1 << ((i + groupSize) & kFlagsMask))) != 0; + groupSize++); + + groupSize++; + + #else + + groupSize = ((Indices[i] & ~0xC0000000) >> kNumBitsMax); + { + Bool finishedGroup = ((Indices[i] & 0x80000000) == 0); + if ((Indices[i] & 0x40000000) != 0) + { + groupSize += ((Indices[i + 1] >> kNumBitsMax) << kNumExtra0Bits); + Indices[i + 1] &= kIndexMask; + } + Indices[i] &= kIndexMask; + groupSize++; + if (finishedGroup || groupSize == 1) + { + Indices[i - finishedGroupSize] &= kIndexMask; + if (finishedGroupSize > 1) + Indices[i - finishedGroupSize + 1] &= kIndexMask; + { + UInt32 newGroupSize = groupSize + finishedGroupSize; + SetFinishedGroupSize(Indices + i - finishedGroupSize, newGroupSize); + finishedGroupSize = newGroupSize; + } + i += groupSize; + continue; + } + finishedGroupSize = 0; + } + + #endif + + if (NumSortedBytes >= blockSize) + { + UInt32 j; + for (j = 0; j < groupSize; j++) + { + UInt32 t = (i + j); + /* Flags[t >> kNumFlagsBits] &= ~(1 << (t & kFlagsMask)); */ + Groups[Indices[t]] = t; + } + } + else + if (SortGroup(blockSize, NumSortedBytes, i, groupSize, NumRefBits, Indices + #ifndef BLOCK_SORT_USE_HEAP_SORT + , 0, blockSize + #endif + ) != 0) + newLimit = i + groupSize; + i += groupSize; + } + if (newLimit == 0) + break; + } + } + #ifndef BLOCK_SORT_EXTERNAL_FLAGS + for (i = 0; i < blockSize;) + { + UInt32 groupSize = ((Indices[i] & ~0xC0000000) >> kNumBitsMax); + if ((Indices[i] & 0x40000000) != 0) + { + groupSize += ((Indices[i + 1] >> kNumBitsMax) << kNumExtra0Bits); + Indices[i + 1] &= kIndexMask; + } + Indices[i] &= kIndexMask; + groupSize++; + i += groupSize; + } + #endif + return Groups[0]; +} + diff --git a/src/libs/7zip/unix/C/BwtSort.h b/src/libs/7zip/unix/C/BwtSort.h new file mode 100644 index 000000000..ce5598f0f --- /dev/null +++ b/src/libs/7zip/unix/C/BwtSort.h @@ -0,0 +1,30 @@ +/* BwtSort.h -- BWT block sorting +2009-02-07 : Igor Pavlov : Public domain */ + +#ifndef __BWT_SORT_H +#define __BWT_SORT_H + +#include "Types.h" + +#ifdef __cplusplus +extern "C" { +#endif + +/* use BLOCK_SORT_EXTERNAL_FLAGS if blockSize can be > 1M */ +/* #define BLOCK_SORT_EXTERNAL_FLAGS */ + +#ifdef BLOCK_SORT_EXTERNAL_FLAGS +#define BLOCK_SORT_EXTERNAL_SIZE(blockSize) ((((blockSize) + 31) >> 5)) +#else +#define BLOCK_SORT_EXTERNAL_SIZE(blockSize) 0 +#endif + +#define BLOCK_SORT_BUF_SIZE(blockSize) ((blockSize) * 2 + BLOCK_SORT_EXTERNAL_SIZE(blockSize) + (1 << 16)) + +UInt32 BlockSort(UInt32 *indices, const Byte *data, UInt32 blockSize); + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/src/libs/7zip/unix/C/CpuArch.c b/src/libs/7zip/unix/C/CpuArch.c new file mode 100644 index 000000000..260cc1f45 --- /dev/null +++ b/src/libs/7zip/unix/C/CpuArch.c @@ -0,0 +1,168 @@ +/* CpuArch.c -- CPU specific code +2010-10-26: Igor Pavlov : Public domain */ + +#include "CpuArch.h" + +#ifdef MY_CPU_X86_OR_AMD64 + +#if (defined(_MSC_VER) && !defined(MY_CPU_AMD64)) || defined(__GNUC__) +#define USE_ASM +#endif + +#if defined(USE_ASM) && !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)); + #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 + +static 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__ ( + "cpuid" + : "=a" (*a) , + "=b" (*b) , + "=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 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); + model = x86cpuid_GetModel(&p); + firm = x86cpuid_GetFirm(&p); + switch (firm) + { + case CPU_FIRM_INTEL: return (family < 6 || (family == 6 && model == 0x100C)); + 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) +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/libs/7zip/unix/C/CpuArch.h b/src/libs/7zip/unix/C/CpuArch.h new file mode 100644 index 000000000..01930c7e6 --- /dev/null +++ b/src/libs/7zip/unix/C/CpuArch.h @@ -0,0 +1,155 @@ +/* CpuArch.h -- CPU specific code +2010-10-26: Igor Pavlov : Public domain */ + +#ifndef __CPU_ARCH_H +#define __CPU_ARCH_H + +#include "Types.h" + +EXTERN_C_BEGIN + +/* +MY_CPU_LE means that CPU is LITTLE ENDIAN. +If MY_CPU_LE is not defined, we don't know about that property of platform (it can be LITTLE ENDIAN). + +MY_CPU_LE_UNALIGN means that CPU is LITTLE ENDIAN and CPU supports unaligned memory accesses. +If MY_CPU_LE_UNALIGN is not defined, we don't know about these properties of platform. +*/ + +#if defined(_M_X64) || defined(_M_AMD64) || defined(__x86_64__) +#define MY_CPU_AMD64 +#endif + +#if defined(MY_CPU_AMD64) || defined(_M_IA64) +#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) +#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) +#define MY_CPU_LE_UNALIGN +#endif + +#if defined(MY_CPU_X86_OR_AMD64) || defined(MY_CPU_ARM_LE) || defined(MY_CPU_IA64_LE) || defined(__ARMEL__) || defined(__MIPSEL__) || defined(__LITTLE_ENDIAN__) +#define MY_CPU_LE +#endif + +#if defined(__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_UNALIGN + +#define GetUi16(p) (*(const UInt16 *)(p)) +#define GetUi32(p) (*(const UInt32 *)(p)) +#define GetUi64(p) (*(const UInt64 *)(p)) +#define SetUi16(p, d) *(UInt16 *)(p) = (d); +#define SetUi32(p, d) *(UInt32 *)(p) = (d); +#define SetUi64(p, d) *(UInt64 *)(p) = (d); + +#else + +#define GetUi16(p) (((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, d) { UInt32 _x_ = (d); \ + ((Byte *)(p))[0] = (Byte)_x_; \ + ((Byte *)(p))[1] = (Byte)(_x_ >> 8); } + +#define SetUi32(p, d) { UInt32 _x_ = (d); \ + ((Byte *)(p))[0] = (Byte)_x_; \ + ((Byte *)(p))[1] = (Byte)(_x_ >> 8); \ + ((Byte *)(p))[2] = (Byte)(_x_ >> 16); \ + ((Byte *)(p))[3] = (Byte)(_x_ >> 24); } + +#define SetUi64(p, d) { UInt64 _x64_ = (d); \ + SetUi32(p, (UInt32)_x64_); \ + SetUi32(((Byte *)(p)) + 4, (UInt32)(_x64_ >> 32)); } + +#endif + +#if defined(MY_CPU_LE_UNALIGN) && defined(_WIN64) && (_MSC_VER >= 1300) + +#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)) + +#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)) + +#endif + +#define GetBe16(p) (((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 +}; + +Bool x86cpuid_CheckAndRead(Cx86cpuid *p); +int x86cpuid_GetFirm(const Cx86cpuid *p); + +#define x86cpuid_GetFamily(p) (((p)->ver >> 8) & 0xFF00F) +#define x86cpuid_GetModel(p) (((p)->ver >> 4) & 0xF00F) +#define x86cpuid_GetStepping(p) ((p)->ver & 0xF) + +Bool CPU_Is_InOrder(); +Bool CPU_Is_Aes_Supported(); + +#endif + +EXTERN_C_END + +#endif diff --git a/src/libs/7zip/unix/C/Delta.c b/src/libs/7zip/unix/C/Delta.c new file mode 100644 index 000000000..2b327f15f --- /dev/null +++ b/src/libs/7zip/unix/C/Delta.c @@ -0,0 +1,62 @@ +/* Delta.c -- Delta converter +2009-05-26 : Igor Pavlov : Public domain */ + +#include "Delta.h" + +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/libs/7zip/unix/C/Delta.h b/src/libs/7zip/unix/C/Delta.h new file mode 100644 index 000000000..0d4cd6274 --- /dev/null +++ b/src/libs/7zip/unix/C/Delta.h @@ -0,0 +1,23 @@ +/* Delta.h -- Delta converter +2009-04-15 : Igor Pavlov : Public domain */ + +#ifndef __DELTA_H +#define __DELTA_H + +#include "Types.h" + +#ifdef __cplusplus +extern "C" { +#endif + +#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); + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/src/libs/7zip/unix/C/HuffEnc.c b/src/libs/7zip/unix/C/HuffEnc.c new file mode 100644 index 000000000..561c7e5da --- /dev/null +++ b/src/libs/7zip/unix/C/HuffEnc.c @@ -0,0 +1,146 @@ +/* HuffEnc.c -- functions for Huffman encoding +2009-09-02 : Igor Pavlov : Public domain */ + +#include "HuffEnc.h" +#include "Sort.h" + +#define kMaxLen 16 +#define NUM_BITS 10 +#define MASK ((1 << NUM_BITS) - 1) + +#define NUM_COUNTERS 64 + +#define HUFFMAN_SPEED_OPT + +void Huffman_Generate(const UInt32 *freqs, UInt32 *p, Byte *lens, UInt32 numSymbols, UInt32 maxLen) +{ + UInt32 num = 0; + /* if (maxLen > 10) maxLen = 10; */ + { + UInt32 i; + + #ifdef HUFFMAN_SPEED_OPT + + UInt32 counters[NUM_COUNTERS]; + for (i = 0; i < NUM_COUNTERS; i++) + counters[i] = 0; + for (i = 0; i < numSymbols; i++) + { + UInt32 freq = freqs[i]; + counters[(freq < NUM_COUNTERS - 1) ? freq : NUM_COUNTERS - 1]++; + } + + for (i = 1; i < NUM_COUNTERS; i++) + { + UInt32 temp = counters[i]; + counters[i] = num; + num += temp; + } + + for (i = 0; i < numSymbols; i++) + { + UInt32 freq = freqs[i]; + if (freq == 0) + lens[i] = 0; + else + p[counters[((freq < NUM_COUNTERS - 1) ? freq : NUM_COUNTERS - 1)]++] = i | (freq << NUM_BITS); + } + counters[0] = 0; + HeapSort(p + counters[NUM_COUNTERS - 2], counters[NUM_COUNTERS - 1] - counters[NUM_COUNTERS - 2]); + + #else + + for (i = 0; i < numSymbols; i++) + { + UInt32 freq = freqs[i]; + if (freq == 0) + lens[i] = 0; + else + p[num++] = i | (freq << NUM_BITS); + } + HeapSort(p, num); + + #endif + } + + if (num < 2) + { + unsigned minCode = 0; + unsigned maxCode = 1; + if (num == 1) + { + maxCode = (unsigned)p[0] & MASK; + if (maxCode == 0) + maxCode++; + } + p[minCode] = 0; + p[maxCode] = 1; + lens[minCode] = lens[maxCode] = 1; + return; + } + + { + UInt32 b, e, i; + + i = b = e = 0; + do + { + UInt32 n, m, freq; + n = (i != num && (b == e || (p[i] >> NUM_BITS) <= (p[b] >> NUM_BITS))) ? i++ : b++; + freq = (p[n] & ~MASK); + p[n] = (p[n] & MASK) | (e << NUM_BITS); + m = (i != num && (b == e || (p[i] >> NUM_BITS) <= (p[b] >> NUM_BITS))) ? i++ : b++; + freq += (p[m] & ~MASK); + p[m] = (p[m] & MASK) | (e << NUM_BITS); + p[e] = (p[e] & MASK) | freq; + e++; + } + while (num - e > 1); + + { + UInt32 lenCounters[kMaxLen + 1]; + for (i = 0; i <= kMaxLen; i++) + lenCounters[i] = 0; + + p[--e] &= MASK; + lenCounters[1] = 2; + while (e > 0) + { + UInt32 len = (p[p[--e] >> NUM_BITS] >> NUM_BITS) + 1; + p[e] = (p[e] & MASK) | (len << NUM_BITS); + if (len >= maxLen) + for (len = maxLen - 1; lenCounters[len] == 0; len--); + lenCounters[len]--; + lenCounters[len + 1] += 2; + } + + { + UInt32 len; + i = 0; + for (len = maxLen; len != 0; len--) + { + UInt32 num; + for (num = lenCounters[len]; num != 0; num--) + lens[p[i++] & MASK] = (Byte)len; + } + } + + { + UInt32 nextCodes[kMaxLen + 1]; + { + UInt32 code = 0; + UInt32 len; + for (len = 1; len <= kMaxLen; len++) + nextCodes[len] = code = (code + lenCounters[len - 1]) << 1; + } + /* if (code + lenCounters[kMaxLen] - 1 != (1 << kMaxLen) - 1) throw 1; */ + + { + UInt32 i; + for (i = 0; i < numSymbols; i++) + p[i] = nextCodes[lens[i]]++; + } + } + } + } +} diff --git a/src/libs/7zip/unix/C/HuffEnc.h b/src/libs/7zip/unix/C/HuffEnc.h new file mode 100644 index 000000000..9cf4bfde8 --- /dev/null +++ b/src/libs/7zip/unix/C/HuffEnc.h @@ -0,0 +1,27 @@ +/* HuffEnc.h -- Huffman encoding +2009-02-07 : Igor Pavlov : Public domain */ + +#ifndef __HUFF_ENC_H +#define __HUFF_ENC_H + +#include "Types.h" + +#ifdef __cplusplus +extern "C" { +#endif + +/* +Conditions: + num <= 1024 = 2 ^ NUM_BITS + Sum(freqs) < 4M = 2 ^ (32 - NUM_BITS) + maxLen <= 16 = kMaxLen + Num_Items(p) >= HUFFMAN_TEMP_SIZE(num) +*/ + +void Huffman_Generate(const UInt32 *freqs, UInt32 *p, Byte *lens, UInt32 num, UInt32 maxLen); + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/src/libs/7zip/unix/C/LzFind.c b/src/libs/7zip/unix/C/LzFind.c new file mode 100644 index 000000000..e3ecb0542 --- /dev/null +++ b/src/libs/7zip/unix/C/LzFind.c @@ -0,0 +1,761 @@ +/* LzFind.c -- Match finder for LZ algorithms +2009-04-22 : Igor Pavlov : Public domain */ + +#include <string.h> + +#include "LzFind.h" +#include "LzHash.h" + +#define kEmptyHashValue 0 +#define kMaxValForNormalize ((UInt32)0xFFFFFFFF) +#define kNormalizeStepMin (1 << 10) /* it must be power of 2 */ +#define kNormalizeMask (~(kNormalizeStepMin - 1)) +#define kMaxHistorySize ((UInt32)3 << 30) + +#define kStartMaxLen 3 + +static void LzInWindow_Free(CMatchFinder *p, ISzAlloc *alloc) +{ + if (!p->directInput) + { + alloc->Free(alloc, p->bufferBase); + p->bufferBase = 0; + } +} + +/* 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 == 0 || p->blockSize != blockSize) + { + LzInWindow_Free(p, alloc); + p->blockSize = blockSize; + p->bufferBase = (Byte *)alloc->Alloc(alloc, (size_t)blockSize); + } + return (p->bufferBase != 0); +} + +Byte *MatchFinder_GetPointerToCurrentPos(CMatchFinder *p) { return p->buffer; } +Byte MatchFinder_GetIndexByte(CMatchFinder *p, Int32 index) { return p->buffer[index]; } + +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; + if (p->directInput) + { + UInt32 curSize = 0xFFFFFFFF - p->streamPos; + 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 = 0; + p->directInput = 0; + p->hash = 0; + MatchFinder_SetDefaultSettings(p); + + for (i = 0; i < 256; i++) + { + UInt32 r = i; + int 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 = 0; +} + +void MatchFinder_Free(CMatchFinder *p, ISzAlloc *alloc) +{ + MatchFinder_FreeThisClassMemory(p, alloc); + LzInWindow_Free(p, alloc); +} + +static CLzRef* AllocRefs(UInt32 num, ISzAlloc *alloc) +{ + size_t sizeInBytes = (size_t)num * sizeof(CLzRef); + if (sizeInBytes / sizeof(CLzRef) != num) + return 0; + 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)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; + } + } + 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; + } + + { + UInt32 prevSize = p->hashSizeSum + p->numSons; + UInt32 newSize; + p->historySize = historySize; + p->hashSizeSum = hs; + p->cyclicBufferSize = newCyclicBufferSize; + p->numSons = (p->btMode ? newCyclicBufferSize * 2 : newCyclicBufferSize); + newSize = p->hashSizeSum + p->numSons; + if (p->hash != 0 && prevSize == newSize) + return 1; + MatchFinder_FreeThisClassMemory(p, alloc); + p->hash = AllocRefs(newSize, alloc); + if (p->hash != 0) + { + 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(CMatchFinder *p) +{ + UInt32 i; + for (i = 0; i < p->hashSizeSum; i++) + p->hash[i] = kEmptyHashValue; + p->cyclicBufferPos = 0; + p->buffer = p->bufferBase; + p->pos = p->streamPos = p->cyclicBufferSize; + p->result = SZ_OK; + p->streamEndWasReached = 0; + MatchFinder_ReadBlock(p); + MatchFinder_SetLimits(p); +} + +static UInt32 MatchFinder_GetSubValue(CMatchFinder *p) +{ + return (p->pos - p->historySize - 1) & kNormalizeMask; +} + +void MatchFinder_Normalize3(UInt32 subValue, CLzRef *items, UInt32 numItems) +{ + UInt32 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->hashSizeSum + p->numSons); + 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 hashValue; 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; + +static UInt32 Bt2_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 offset; + GET_MATCHES_HEADER(2) + HASH2_CALC; + curMatch = p->hash[hashValue]; + p->hash[hashValue] = 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[hashValue]; + p->hash[hashValue] = p->pos; + offset = 0; + GET_MATCHES_FOOTER(offset, 2) +} + +static UInt32 Bt3_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 hash2Value, delta2, maxLen, offset; + GET_MATCHES_HEADER(3) + + HASH3_CALC; + + delta2 = p->pos - p->hash[hash2Value]; + curMatch = p->hash[kFix3HashSize + hashValue]; + + p->hash[hash2Value] = + p->hash[kFix3HashSize + hashValue] = p->pos; + + + maxLen = 2; + offset = 0; + if (delta2 < p->cyclicBufferSize && *(cur - delta2) == *cur) + { + for (; maxLen != lenLimit; maxLen++) + if (cur[(ptrdiff_t)maxLen - delta2] != cur[maxLen]) + break; + distances[0] = maxLen; + distances[1] = delta2 - 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 hash2Value, hash3Value, delta2, delta3, maxLen, offset; + GET_MATCHES_HEADER(4) + + HASH4_CALC; + + delta2 = p->pos - p->hash[ hash2Value]; + delta3 = p->pos - p->hash[kFix3HashSize + hash3Value]; + curMatch = p->hash[kFix4HashSize + hashValue]; + + p->hash[ hash2Value] = + p->hash[kFix3HashSize + hash3Value] = + p->hash[kFix4HashSize + hashValue] = p->pos; + + maxLen = 1; + offset = 0; + if (delta2 < p->cyclicBufferSize && *(cur - delta2) == *cur) + { + distances[0] = maxLen = 2; + distances[1] = delta2 - 1; + offset = 2; + } + if (delta2 != delta3 && delta3 < p->cyclicBufferSize && *(cur - delta3) == *cur) + { + maxLen = 3; + distances[offset + 1] = delta3 - 1; + offset += 2; + delta2 = delta3; + } + if (offset != 0) + { + for (; maxLen != lenLimit; maxLen++) + if (cur[(ptrdiff_t)maxLen - delta2] != cur[maxLen]) + break; + 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 Hc4_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 hash2Value, hash3Value, delta2, delta3, maxLen, offset; + GET_MATCHES_HEADER(4) + + HASH4_CALC; + + delta2 = p->pos - p->hash[ hash2Value]; + delta3 = p->pos - p->hash[kFix3HashSize + hash3Value]; + curMatch = p->hash[kFix4HashSize + hashValue]; + + p->hash[ hash2Value] = + p->hash[kFix3HashSize + hash3Value] = + p->hash[kFix4HashSize + hashValue] = p->pos; + + maxLen = 1; + offset = 0; + if (delta2 < p->cyclicBufferSize && *(cur - delta2) == *cur) + { + distances[0] = maxLen = 2; + distances[1] = delta2 - 1; + offset = 2; + } + if (delta2 != delta3 && delta3 < p->cyclicBufferSize && *(cur - delta3) == *cur) + { + maxLen = 3; + distances[offset + 1] = delta3 - 1; + offset += 2; + delta2 = delta3; + } + if (offset != 0) + { + for (; maxLen != lenLimit; maxLen++) + if (cur[(ptrdiff_t)maxLen - delta2] != cur[maxLen]) + break; + 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 +} + +UInt32 Hc3Zip_MatchFinder_GetMatches(CMatchFinder *p, UInt32 *distances) +{ + UInt32 offset; + GET_MATCHES_HEADER(3) + HASH_ZIP_CALC; + curMatch = p->hash[hashValue]; + p->hash[hashValue] = 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[hashValue]; + p->hash[hashValue] = 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[hashValue]; + p->hash[hashValue] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +static void Bt3_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 hash2Value; + SKIP_HEADER(3) + HASH3_CALC; + curMatch = p->hash[kFix3HashSize + hashValue]; + p->hash[hash2Value] = + p->hash[kFix3HashSize + hashValue] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +static void Bt4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 hash2Value, hash3Value; + SKIP_HEADER(4) + HASH4_CALC; + curMatch = p->hash[kFix4HashSize + hashValue]; + p->hash[ hash2Value] = + p->hash[kFix3HashSize + hash3Value] = p->pos; + p->hash[kFix4HashSize + hashValue] = p->pos; + SKIP_FOOTER + } + while (--num != 0); +} + +static void Hc4_MatchFinder_Skip(CMatchFinder *p, UInt32 num) +{ + do + { + UInt32 hash2Value, hash3Value; + SKIP_HEADER(4) + HASH4_CALC; + curMatch = p->hash[kFix4HashSize + hashValue]; + p->hash[ hash2Value] = + p->hash[kFix3HashSize + hash3Value] = + p->hash[kFix4HashSize + hashValue] = 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[hashValue]; + p->hash[hashValue] = 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->GetIndexByte = (Mf_GetIndexByte_Func)MatchFinder_GetIndexByte; + vTable->GetNumAvailableBytes = (Mf_GetNumAvailableBytes_Func)MatchFinder_GetNumAvailableBytes; + vTable->GetPointerToCurrentPos = (Mf_GetPointerToCurrentPos_Func)MatchFinder_GetPointerToCurrentPos; + if (!p->btMode) + { + vTable->GetMatches = (Mf_GetMatches_Func)Hc4_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Hc4_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 + { + vTable->GetMatches = (Mf_GetMatches_Func)Bt4_MatchFinder_GetMatches; + vTable->Skip = (Mf_Skip_Func)Bt4_MatchFinder_Skip; + } +} diff --git a/src/libs/7zip/unix/C/LzFind.h b/src/libs/7zip/unix/C/LzFind.h new file mode 100644 index 000000000..010c4b92b --- /dev/null +++ b/src/libs/7zip/unix/C/LzFind.h @@ -0,0 +1,115 @@ +/* LzFind.h -- Match finder for LZ algorithms +2009-04-22 : Igor Pavlov : Public domain */ + +#ifndef __LZ_FIND_H +#define __LZ_FIND_H + +#include "Types.h" + +#ifdef __cplusplus +extern "C" { +#endif + +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) */ + + UInt32 matchMaxLen; + CLzRef *hash; + CLzRef *son; + UInt32 hashMask; + UInt32 cutValue; + + Byte *bufferBase; + ISeqInStream *stream; + int streamEndWasReached; + + UInt32 blockSize; + UInt32 keepSizeBefore; + UInt32 keepSizeAfter; + + UInt32 numHashBytes; + int directInput; + size_t directInputRem; + int btMode; + int bigHash; + UInt32 historySize; + UInt32 fixedHashSize; + UInt32 hashSizeSum; + UInt32 numSons; + SRes result; + UInt32 crc[256]; +} CMatchFinder; + +#define Inline_MatchFinder_GetPointerToCurrentPos(p) ((p)->buffer) +#define Inline_MatchFinder_GetIndexByte(p, index) ((p)->buffer[(Int32)(index)]) + +#define Inline_MatchFinder_GetNumAvailableBytes(p) ((p)->streamPos - (p)->pos) + +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, UInt32 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 Byte (*Mf_GetIndexByte_Func)(void *object, Int32 index); +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_GetIndexByte_Func GetIndexByte; + 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(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); + +#ifdef __cplusplus +} +#endif + +#endif diff --git a/src/libs/7zip/unix/C/LzFindMt.c b/src/libs/7zip/unix/C/LzFindMt.c new file mode 100644 index 000000000..aa41ed98a --- /dev/null +++ b/src/libs/7zip/unix/C/LzFindMt.c @@ -0,0 +1,793 @@ +/* LzFindMt.c -- multithreaded Match finder for LZ algorithms +2009-09-20 : Igor Pavlov : Public domain */ + +#include "LzHash.h" + +#include "LzFindMt.h" + +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); +} + +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 */ + +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; +} + +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, unsigned (MY_STD_CALL *startAddress)(void *), 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, unsigned (MY_STD_CALL *startAddress)(void *), 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)), hashMask = hashMask; crc = 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) */ + +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 = MatchFinder_GetPointerToCurrentPos(mf); + const Byte *afterPtr; + MatchFinder_MoveBlock(mf); + afterPtr = MatchFinder_GetPointerToCurrentPos(mf); + mt->pointerToCurPos -= beforePtr - afterPtr; + mt->buffer -= beforePtr - afterPtr; + } + 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, 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); + } + } +} + +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 + +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 + +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; + for (; p->hashNumAvail != 0; p->hashNumAvail--) + distances[curPos++] = 0; + break; + } + { + 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; +} + +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); + |